1000 Federal Register /Vol. 65, No. 4/Thursday, … Federal Register/Vol. 65, No. 4/Thursday,...

1000 Federal Register / Vol. 65, No. 4 / Thursday, January 6, 2000 / Rules and Regulations

DEPARTMENT OF HEALTH ANDHUMAN SERVICES

Food and Drug Administration

21 CFR Part 101

[Docket No. 98N–0044]

RIN 0910–AB97

Regulations on Statements Made forDietary Supplements Concerning theEffect of the Product on the Structureor Function of the Body

AGENCY: Food and Drug Administration,HHS.ACTION: Final rule.

SUMMARY: The Food and DrugAdministration (FDA) is issuing finalregulations defining the types ofstatements that can be made concerningthe effect of a dietary supplement on thestructure or function of the body. Theregulations also establish criteria fordetermining when a statement about adietary supplement is a claim todiagnose, cure, mitigate, treat, orprevent disease. This action is intendedto clarify the types of claims that maybe made for dietary supplementswithout prior review by FDA and thetypes of claims that require priorauthorization as health claims or priorapproval as drug claims.DATES: The final rule will becomeeffective February 7, 2000.FOR FURTHER INFORMATION CONTACT: AnnMarlin Witt, Office of Policy, Planning,and Legislation (HF–11), Food and DrugAdministration, 5600 Fishers Lane,Rockville, MD 20857, 301–827–0084.SUPPLEMENTARY INFORMATION:

I. Introduction

In the Federal Register of April 29,1998 (63 FR 23624), FDA proposedregulations to identify the types ofstatements that may be made withoutprior FDA review about the effects ofdietary supplements on the structure orfunction of the body (‘‘structure/function claims’’), and to distinguishthese claims from claims that a productdiagnoses, treats, prevents, cures, ormitigates disease (disease claims). FDAreceived over 235,000 submissions inresponse to the proposed rule. Many ofthese were form letters, but over 22,000were individual letters from the dietarysupplement industry, trade associations,health professional groups, andconsumers. Almost all the commentsfrom the dietary supplement industryand from individuals, which made upthe vast majority of the comments,objected to all or part of the proposedrule, arguing that it inappropriately

restricted the structure/function claimsthat could be made for dietarysupplements. Most of the commentsfrom health professional groups andgroups devoted to particular diseasessupported the proposed rule, orbelieved it did not go far enough inlimiting structure/function claims fordietary supplements.

After reviewing the comments, FDAconcluded that the comments had raisedsignificant questions about some of thekey provisions of the proposal such thata public meeting was warranted. In theFederal Register of July 8, 1999 (64 FR36824), FDA announced a publicmeeting to be held on August 4, 1999,at which representatives of the dietarysupplement industry, consumer groups,and health professionals were asked toaddress three major issues raised by thecomments. The three issues, describedin the Federal Register notice, were: (1)Whether to finalize the proposeddefinition of ‘‘disease’’ or retain a 1993definition of ‘‘disease or health-relatedcondition’’ that was in effect at the timethe Dietary Supplement Health andEducation Act (DSHEA) was enacted; (2)whether to modify one of the proposedcriteria for assessing disease claims topermit structure/function claims relatedto certain conditions associated withnatural states, such as hot flashesassociated with menopause anddecreased sexual function associatedwith aging; and (3) whether to permitimplied disease claims structure/function claims. The July 8, 1999, noticealso reopened the comment period untilAugust 4, 1999, to receive writtencomments on these three issues.

This document addresses thecomments received on the proposedrule, as well as comments received inresponse to the July 8, 1999, FederalRegister notice. A few comments raisedissues that are beyond the scope of thisrule and generally will not be addressedin this document.

A. Highlights of the Final RuleLike the proposed rule, the final rule

contains criteria to determine when alabeling statement made about a dietarysupplement constitutes a structure/function claim for which no prior FDAreview is required and when itconstitutes a disease-related claim thatrequires either authorization of a healthclaim or review under the drugprovisions of Federal Food, Drug, andCosmetic Act (the act). FDA has,however, made several importantchanges in the final rule in response tocomments.

First, the agency has deleted theproposed definition of ‘‘disease.’’ Ratherthan creating a new definition of

disease, FDA will use the preexistingdefinition of ‘‘disease or health-relatedcondition’’ in § 101.14(a)(5) (21 CFR101.14(a)(5)) (formerly § 101.14(a)(6)),which was issued as part of theimplementation of the health claimsprovisions of the Nutrition Labeling andEducation Act (NLEA). This change hasbeen made in response to the largenumber of comments that objected tothe proposed definition and urged thatFDA retain the NLEA definition.

Second, FDA has revised the criterionthat applies to conditions associatedwith such natural states or processes asmenopause, aging, adolescence, andpregnancy. The proposed rule statedthat menopause, aging, and pregnancyare not themselves diseases but thatcertain conditions associated with themare diseases if they are recognizable toconsumers or health professionals asabnormal. Many comments objected toclassifying as diseases such commonconditions as hot flashes, premenstrualsyndrome (PMS), and decreased sexualfunction associated with aging. Inresponse to these comments, FDA hasrevised proposed § 101.93(g)(2)(iii).Common conditions associated withnatural states or processes that do notcause significant or permanent harmwill not be treated as diseases under thefinal rule. For example, hot flashes,common symptoms associated with themenstrual cycle, ordinary morningsickness associated with pregnancy,mild memory problems associated withaging, hair loss associated with aging,and noncystic acne will not be treatedas diseases under this provision.Uncommon or serious conditions likesenile dementia, toxemia of pregnancy,severe depression associated with themenstrual cycle, and cystic acne willcontinue to be treated as diseases underthe final rule.

Third, FDA has revised the criterionthat relates to the use in labeling of thetitles of publications that refer todiseases. In response to commentsobjecting that, as proposed, thiscriterion would hamper manufacturersfrom providing consumers withinformation substantiating their claims,FDA has revised this criterion. Underthe revised criterion, the use in labelingof a publication title that refers to adisease will be considered a diseaseclaim only if, in context, it implies thatthe product may be used to diagnose,treat, mitigate, cure, or prevent disease.Highlighting, bolding, using large typesize, or prominent placement of acitation that refers to a disease use inthe title could suggest that the producthas an effect on disease. Placing acitation to a scientific reference thatrefers to a disease in the title on the

VerDate 04-JAN-2000 16:16 Jan 05, 2000 Jkt 190000 PO 00000 Frm 00002 Fmt 4701 Sfmt 4700 E:\FR\FM\06JAR4.XXX pfrm11 PsN: 06JAR4

1001Federal Register / Vol. 65, No. 4 / Thursday, January 6, 2000 / Rules and Regulations

immediate product label or packagingwill be considered a disease claim forthat product. The agency will alsoconsider whether the cited articleprovides legitimate support for theexpress structure/function statementmade for that dietary supplement.Enhancing the bibliography withcitations to scientific references thatrefer to a disease in the title and thathave no reasonable relation to thestatement made will be considered adisease claim. Similarly, the agency willconsider whether citations are to bonafide research.

B. BackgroundDSHEA created a new regime for the

regulation of dietary supplements.These products were previouslyregulated either as foods or as drugs,depending upon whether they had theattributes of food and upon theirintended uses. Before the passage ofDSHEA, a dietary supplement for whicha health-related claim was made wasregulated either as a drug, which had tobe shown to be safe and effective beforemarketing, or as a food, for which priorauthorization to make a health claimwas required if the claim concerned adisease or health-related condition. Ifthe claim concerned a non-disease-related effect on the structure orfunction of the body and the claimedeffect derived from a food attribute,such as nutritive value, the claim wasconsidered a food claim, and priorauthorization was not required. Undersection 201(g)(1)(B) and (g)(1)(C) of theact (21 U.S.C. 321(g)(1)(B) and (g)(1)(C)),a drug is defined as ‘‘an article intendedfor use in the diagnosis, cure,mitigation, treatment, or prevention ofdisease,’’ or ‘‘an article (other than food)intended to affect the structure or anyfunction of the body.’’ Section 505 ofthe act (21 U.S.C. 355) requires that newdrugs (see section 201(p) of the act) beshown to be safe and effective for theirintended uses before marketing. Undersections 403(r)(1)(B) and (r)(5)(D) of theact (21 U.S.C. 343(r)(1)(B) and (r)(5)(D))and § 101.14, prior authorization isrequired to make a health claim for adietary supplement. A health claim is aclaim that ‘‘characterizes therelationship of any nutrient * * * in thefood to a disease or health-relatedcondition’’ (section 403(r)(1)(B) of theact; see § 101.14(a)(1)).

DSHEA specifically authorizedcertain types of claims about the uses ofdietary supplements, including someclaims that formerly would haverequired review by FDA before theclaim is made. Section 403(r)(6) of theact, added by DSHEA, allows dietarysupplement labeling to bear, among

other types of statements, a statementthat ‘‘describes the role of a nutrient ordietary ingredient intended to affect thestructure or function in humans’’ or that‘‘characterizes the documentedmechanism by which a nutrient ordietary ingredient acts to maintain suchstructure or function.’’ Such statementsare generally referred to as ‘‘structure/function claims.’’ Because many of theseclaims would previously have beencovered by the drug definition insection 201(g)(1)(C) of the act, section201(g)(1) was amended by DSHEA toprovide that a dietary supplement ‘‘forwhich a truthful and not misleadingstatement is made in accordance withsection 403(r)(6) is not a drug underclause (C) solely because the label or thelabeling contains such a statement.’’

Although a dietary supplementmanufacturer who wishes to make astatement permitted under section403(r)(6) of the act need not obtain priorreview of the statement, themanufacturer must possesssubstantiation that the statement istruthful and not misleading, and mustinclude in the statement the followingdisclaimer: ‘‘This statement has notbeen evaluated by the Food and DrugAdministration. This product is notintended to diagnose, treat, cure, orprevent any disease.’’ DSHEA alsorequires the manufacturer of a dietarysupplement bearing a statement undersection 403(r)(6) of the act to notifyFDA, no later than 30 days after the firstmarketing of the dietary supplementwith the statement, that such astatement is being made for the product.Regulations implementing theserequirements were published in theFederal Register of September 23, 1997,and are codified at § 101.93 (21 CFR101.93) (62 FR 49883 at 49886,September 23, 1997).

DSHEA did not alter the statutorytreatment of dietary supplement claimsrelated to disease (‘‘disease claims’’).Section 403(r)(6) of the act, specificallyprovides that statements permittedunder that section ‘‘may not claim todiagnose, mitigate, treat, cure, orprevent a specific disease or class ofdiseases,’’ except that such statementsmay claim a benefit related to a classicalnutrient deficiency disease, providedthat they also disclose the prevalence ofthe disease in the United States.Consistent with the quoted provision,Congress did not modify section201(g)(1)(B) of the act to exclude diseaseclaims for dietary supplements from useas evidence of intended use as a drug,as it had done for section 201(g)(1)(C) ofthe act. Thus, dietary supplements‘‘intended for use in the diagnosis, cure,mitigation, treatment, or prevention of

disease’’ remain within the definition ofa ‘‘drug.’’ In enacting DSHEA, Congressalso maintained the requirement of priorauthorization of a claim thatcharacterizes the relationship of anutrient in a dietary supplement to adisease (section 403(r)(1)(B) and(r)(5)(D) of the act). An interested personmay submit a petition to FDA requestingthe agency to issue a regulationauthorizing the health claim (see§ 101.70 (21 CFR 101.70)). Thepetitioner must demonstrate, amongother things, that the use of thesubstance at levels necessary to justifythe claim is safe and that there is‘‘significant scientific agreement’’among qualified experts that the claimis supported by the totality of publiclyavailable scientific evidence(§ 101.14(b)(3)(ii) and (c)). The agencynotes that for health claims to be usedon conventional foods, an interestedperson may submit to FDA anotification of an authoritativestatement by one of certain designatedscientific bodies concerning thesubstance-disease relationship to whichthe claim refers (see section 403(r)(3)(C)of the act). Unless FDA issues aregulation modifying or prohibiting theclaim, or a Federal district court findsthat applicable statutory requirementshave not been met, the claim may beused 120 days after the notification hasbeen submitted (see section403(r)(3)(C)(ii) and (r)(3)(D) of the act).This alternative authorization proceduredoes not apply to dietary supplementsby statute, but FDA has proposed toextend it to dietary supplements byregulation (see 64 FR 3250, January 21,1999).

Although FDA believes that dietarysupplements have potential benefits forconsumers, dietary supplements labeledwith unproven disease claims, i.e., thosethat have not met the requirements forhealth claim authorization or new drugapproval, can pose serious risks. Suchclaims may encourage consumers toself-treat for a serious disease withoutbenefit of a medical diagnosis ortreatment. They may also causeconsumers to substitute potentiallyineffective products for proven ones,foregoing or delaying effective treatmentfor serious and life-threatening illnesses.Reliance on disease prevention claimsmay encourage consumers to feelsufficiently protected from developingserious diseases (e.g., cancer or humanimmunodeficiency virus (HIV)infection) that they delay or foregoregular screening, and forfeit theopportunity for early medical treatmentthat may be critical to survival. Finally,use of dietary supplements to treat



disease may increase the risk of adversereactions due to the interaction of thedietary supplement with othercompounds a consumer is taking forthat disease or for other conditions, e.g.,prescription medications.

This final rule is intended to applyonly to structure/function claims anddisease claims within the meaning ofsection 403(r)(6) of the act. DSHEA,generally, and section 403(r)(6) of theact, specifically, apply only to dietarysupplements for human consumptionand were enacted to provide a uniqueregulatory regime for these products.Thus, this rule is neither intended toapply to products other than dietarysupplements for human consumptionnor to interpret other provisions of theact.

The final rule establishes criteria fordetermining whether a statement madeabout a dietary supplement isacceptable as a structure/function claimunder section 403(r)(6) of the act. Therule is neither intended to establishwhether any particular structure/function claim is appropriate for anyspecific product, nor whether the claimwould be permitted under otherprovisions of the act. Like the labelingof any other FDA-regulated product, thelabeling of dietary supplements mustcomply with all applicable requirementsof the act and regulations. For example,an otherwise acceptable structure/function claim might nevertheless befalse or misleading for other reasons,causing the product to be misbrandedunder section 403(a)(1) of the act.

C. The Proposed RuleThe proposed rule defined criteria for

determining when a statement about adietary supplement is a claim todiagnose, cure, mitigate, treat, orprevent disease (‘‘disease claim’’), andthus requires prior approval as a drug orprior authorization as a health claim.The proposed rule included a definitionof ‘‘disease,’’ which was to replace adefinition of ‘‘disease or health-relatedcondition’’ issued for implementation ofthe health claims regulations, and 10criteria for identifying express orimplied disease claims. FDA proposedto treat a statement about a dietarysupplement as a disease claim if thestatement claimed, explicitly orimplicitly, that the product: (1) Has aneffect on a specific disease or class ofdiseases; (2) has an effect, usingscientific or lay terminology, on one ormore signs or symptoms that arerecognizable to health care professionalsor consumers as being characteristic ofa specific disease or of a number ofdifferent specific diseases; (3) has aneffect on a consequence of a natural

state that presents a characteristic set ofsigns or symptoms recognizable tohealth care professionals or consumersas constituting an abnormality of thebody; (4) has an effect on diseasethrough one or more of the followingfactors: (a) The name of the product; (b)a statement about the formulation of theproduct, including a claim that theproduct contains an ingredient that hasbeen regulated by FDA as a drug and iswell known to consumers for its use inpreventing or treating a disease; (c)citation of a publication or reference, ifthe citation refers to a disease use; (d)use of the term ‘‘disease’’ or ‘‘diseased;’’or (e) use of pictures, vignettes, symbols,or other means; (5) belongs to a class ofproducts that is intended to diagnose,mitigate, treat, cure, or prevent adisease; (6) is a substitute for a productthat is a therapy for a disease; (7)augments a particular therapy or drugaction; (8) has a role in the body’sresponse to a disease or to a vector ofdisease; (9) treats, prevents, or mitigatesadverse events associated with a therapyfor a disease and manifested by acharacteristic set of signs or symptoms;or (10) otherwise suggests an effect ona disease or diseases.

Claims that did not fall within theproposed criteria for disease claims andthat otherwise complied with thenotification and disclaimer provisionsof § 101.93(a) through (e) were to beeligible for use as structure/functionclaims. The proposed rule also providedexamples of claims that would bepermitted as structure/function claimsand those that would require priorreview as disease claims under each ofthe 10 criteria.

The basis for the proposed rule wasthe agency’s experience inimplementing section 403(r)(6) of theact, and the final report (the report) ofthe President’s Commission on DietarySupplement Labels (Ref. 1), whichincluded a number of recommendationsfor distinguishing structure/functionand disease claims and suggested thatFDA issue further guidance onacceptable structure/function claims.

II. Comments

A. General Comments

(1.) Many comments focused on theimpact of the rule on consumers. Manycomments opposing the proposed rulesaid that consumers should be able toreceive truthful and non-misleadinginformation and that the proposed rulewould curtail or restrict suchinformation or restrict the focus ofdietary supplements to preventive careand wellness. Some comments addedthat DSHEA, through the dissemination

of truthful and non-misleadinginformation on health and promotionand disease prevention, makesconsumers responsible for their ownhealth. Other comments said that FDAshould let the public educate itself.Other comments suggested that FDAsimply adopt a ‘‘truthful and non-misleading’’ standard. Some commentsadded that full disclosure of allpertinent information (such as thepreliminary status of scientific studiessubstantiating the claim) would besufficient. Another comment questionedwhether consumers would, as thepreamble to the proposed rule stated,benefit from not having to search forinformation and from gettingappropriate information. The commentargued that consumers would receiveless information under the rule andwould have to search more extensivelyfor information.

Many comments supporting theproposed rule, including commentsfrom nutrition counselors and healthprofessionals, said that the proposalwould reduce confusion amongpatients, prevent consumers from beingmisled, diminish the number ofinappropriate disease claims, and helpconsumers decide when to seek medicalattention. One comment added that,while it supported the need forconsumers to have choice regardingdietary supplements, the choice shouldbe made based on accurate informationthat is supported by appropriatescientific investigations. One commentargued that in the absence of valideffectiveness data, which does not existfor most dietary supplements, it is notpossible to provide ‘‘truthful’’information about the effects of theseproducts. Some comments said that theproposal would protect consumers fromharmful or potentially harmful productsand save consumers from needlesssuffering and financial loss; othersexpressed concern that inappropriatestatements would expose consumers topotentially harmful drug-supplementinteractions, create ‘‘false hopes,’’ andlead consumers to stop complying withadvice from health care professionals orto avoid proven treatments.

FDA agrees that DSHEA encouragesthe dissemination of truthful and non-misleading information about the usesof dietary supplements to affect thestructure or function of the body, andencourages full disclosure ofinformation about claims authorized bythe statute. To the extent that truthfuland non-misleading information isbeing withheld from consumers in thecontext of structure/function claims fordietary supplements, it is the statutethat, in the first instance, precludes



certain information from being includedin such claims. Section 403(r)(6) of theact permits dietary supplement labels tocarry structure/function claims withoutmeeting the requirements for drugapproval or health claim authorization,but precludes them from carryingunreviewed claims that the productdiagnoses, treats, mitigates, cures, orprevents disease. (The statute does notultimately prevent dissemination ofinformation about disease uses to theconsumer in labeling claims orotherwise. Instead, it requires thatclaims about disease uses meet certainstandards of substantiation and undergoagency review.) This final ruledifferentiates between structure/function claims authorized by section403(r)(6) of the act and disease claimsthat may not be made in dietarysupplement labeling under the authorityof section 403(r)(6). The agency notesthat, in response to comments, the finalrule classifies many more claims asstructure/function claims than wouldhave been so classified under theproposed rule, thus increasing theamount of information available to theconsumer without prior FDA review.

The agency also declines to adopt a‘‘truthful and non-misleading’’ standardinstead of the final rule. Section403(a)(1) of the act already subjects allfood claims, including structure/function claims on dietary supplements,to the ‘‘truthful and non-misleading’’standard, so promulgating the samestandard through regulations isunnecessary. In addition, section403(r)(6)(B) of the act already requiresdietary supplement manufacturers tohave substantiation that their statementsare truthful and non-misleading. Finallya fundamental problem with thisapproach is that a ‘‘truthful and non-misleading’’ standard, unlike the finalrule, would not provide any criteria fordifferentiating between structure/function claims and disease claims.

(2.) Some comments focused onproduct safety. One comment said thatregulation of claims is unnecessarybecause dietary supplements are safe.Similarly, another comment claimedthat ‘‘one million peer-reviewedstudies’’ showed that dietarysupplements provide benefits, whereasa recent medical journal reported deathsand other injuries to patients who useprescription drugs. Other commentsdeclared that dietary supplements aresafer than most regularly-used drugproducts. In contrast, other commentsargued that the safety of many dietarysupplements is unknown, and that riskshave been documented with somesupplements. Some comments claimedthat dietary supplements pose risks

because they can cause consumers toavoid or delay more effective treatment.One comment stated that there is asubstantial potential for public harmbecause of the unknown or unregulatedsource materials for many dietarysupplements, the variety of suppliers,and the lack of regulatory productionstandards and quality control.

Although this final rule may notappear to be a safety measure because itaddresses the labeling of dietarysupplements rather than theircomposition, protecting consumerhealth and safety is one of its majorpurposes. Because structure/functionclaims are not subject to the new drugapproval standard or the health claimauthorization standard and do notundergo FDA review before marketing,FDA believes it is important to ensurethat such claims do not promoteproducts for disease treatment orprevention claims. Disease treatment orprevention claims can pose serious risksto consumers if they induce consumersto substitute ineffective or less effectivetreatments for proven ones, especially ifthe disease involved is serious or life-threatening. Therefore, the agencybelieves that ensuring that such claimscannot be made without ademonstration of safety andeffectiveness will protect and promotepublic health.

FDA also believes that the safety andthe effectiveness of products intended topromote health, including both dietarysupplements and drugs, cannot beviewed independently of each other.FDA agrees that prescription drugs canand do cause adverse reactions. It isimportant to remember, however, that‘‘safety’’ is relative. Products that arecapable of treating diseases havepowerful effects on the body andfrequently carry risks. Beforeprescription drugs are marketed, boththeir risks and their benefits must becarefully investigated and documentedin adequately designed clinical trials.Prescription drugs are permitted to bemarketed only when the agencyconcludes that their documentedbenefits outweigh their known andpotential risks. Those with significantrisks are approved for marketing only ifthe benefits warrant those risks. Andthey are marketed as ‘‘prescription’’drugs to ensure that health professionalsmanage their risks. Even over-the-counter (OTC) drugs are evaluated forboth benefits and risks and arepermitted to be marketed only whentheir established benefits outweigh theirrisks. There is no comparable testingand approval process for dietarysupplements marketed with structure/function claims. The manufacturer must

have substantiation of the structure/function claim, but this substantiation isnot reviewed before the product ismarketed with the claim. Contrary to thesuggestion in the comment, few dietarysupplements have been the subjects ofadequately designed clinical trials.

This does not mean that dietarysupplements are unsafe or that they donot have benefits. Some have alreadybeen shown to be safe and to havebenefits, and the safety andeffectiveness of others are likely to beshown in the future. At this time,however, many marketed supplementshave not been the subjects of adequatestudies to establish whether or not theyare safe or effective, or the nature of thebenefits they may provide.

(3.) Many comments asserted thatFDA had no authority to issue theproposed rule because it wasinconsistent with DSHEA andcongressional intent, in that it restrictedrather than increased the amount ofinformation given to consumers. Somecomments said that Congress enactedDSHEA to reverse FDA’s ‘‘overlyrestrictive’’ approach towards healthclaims and to increase thedissemination of truthful and non-misleading health information and thatCongress repeatedly expressed itsdispleasure with FDA’s regulatoryapproach. One comment said FDA mustdetermine whether a proposed action isconsistent with its statutory authoritybefore it takes any regulatory action.The comment cited excerpts fromcongressional documents ‘‘condemningthe agency’s repeated penchant’’ forrestricting statements on dietarysupplement labels and labeling, andsaid that, given congressional intent andthe act’s language, FDA has no authorityto proceed with rulemaking without agrant of authority from Congress. Onecomment cited section 403B of the act(21 U.S.C. 343–2) as evidence thatCongress, by exempting certainpublications from the definition oflabeling, barred FDA from restricting in‘‘any way whatsoever’’ thedissemination of such publications andinformation.

FDA agrees that DSHEA was intendedto authorize the dissemination of moretruthful and non-misleadinginformation in dietary supplementlabeling without the need for prioragency review. In response to commentsthat the proposed rule was toorestrictive, FDA has modified the finalrule to incorporate many of the changesrequested by the comments, including areturn to the preexisting definition of‘‘disease or health-related condition,’’and a less restrictive interpretation ofthe types of structure/function claims



that can be made about conditionsassociated with such natural states asaging, pregnancy, and the menstrualcycle. The final rule classifies manymore claims as structure/functionclaims than the proposed rule wouldhave.

The agency does not agree, however,that section 403(r)(6) of the actauthorizes dissemination of any and allinformation about dietary supplementswithout prior review. That sectionauthorizes statements about the effectsof dietary supplements on the structureor function of the body, but notstatements that claim to diagnose,mitigate, treat, cure, or prevent aspecific disease or class of diseases.Section 403B of the act exempts frombeing considered labeling certainbalanced, third-party publications thatare physically separate from productlabeling and do not promote a particularbrand or product. This provision doesnot authorize dietary supplementmanufacturers to ignore the restrictionsin section 403(r)(6) of the act on whatstructure/function claims may be madeby a manufacturer about its product onthe product label and in materials thatare indisputably part of the product’slabeling.

The agency also disagrees with theassertion that separate congressionalauthority is needed for this rulemaking.FDA issued the proposed rule, and thisfinal rule, to implement section403(r)(6) of the act. No independentauthority to issue these regulations isnecessary because section 701(a) of theact (21 U.S.C 371(a)) expressly givesFDA ‘‘the authority to promulgateregulations for the efficient enforcementof this Act, except as otherwiseprovided in (section 701 of the act) * **.’’ The proposed rule identified section701(a) of the act as being part of theagency’s legal authority (see 63 FR23624 at 23628 and 23631), and there isno exception in the act that restricts orlimits, either expressly or impliedly, theagency’s ability to issue regulations toimplement section 403(r)(6) of the act.Therefore, the rule is authorized by lawand consistent with FDA’s statutoryauthority.

(4.) Some comments contended thatFDA did not provide a sufficientjustification for issuing the rule. Twocomments challenged FDA’s assertionthat the rule would reduce substantialconfusion among manufacturers. Thecomments referred to statements in thepreamble to the proposed rule whichsaid FDA received approximately 2,300notifications of structure/functionclaims and sent objection letters toapproximately 150 notifications. Onecomment said the low objection rate did

not indicate ‘‘substantial confusion’’among manufacturers, while the othercomment hypothesized that, if FDAobjected to a small number of claims ineach notification, the number ofobjectionable claims was very small.Other comments contended that theCommission report did not support theproposed rule. These comments weredivided in their reasons. Somecomments argued that the Commissionexceeded its statutory mandate undersection 12 of DSHEA or failed toperform its statutory obligations. Thus,the comments stated, FDA cannot baseany regulation on the Commission’sfindings, guidance, or recommendationsand has no authority to proceed withthe rulemaking. Other comments statedthat FDA relied on statements fromindividual Commission members ratherthan the report itself, that the report didnot suggest that FDA issue regulations,and that the report did not suggest thatFDA issue a new definition of disease.One comment said that the Commissiondid not support a need for regulations.Another comment noted that theCommission did not recommendregulations and asserted that FDA hadpublicly said that DSHEA is self-implementing.

FDA does not agree that there isinsufficient support for this rule. FDA’sexperience, the Commission report, andFDA’s authority under section 701(a) ofthe act to issue regulationsimplementing statutory requirementsprovide more than adequate support forthe rule. The preamble to the proposedrule referred to substantial confusionamong manufacturers and consumers,rather than manufacturers alone.Comments received from other sources,particularly physicians, dieticians, andhealth professional organizations,agreed that consumers are confused andmisled by claims. In addition, thenumber of objection letters is not thesole indicator of manufacturerconfusion, for three reasons. First,manufacturers and consumers haveasked FDA to provide clarification onstructure/function and disease claims,and such requests for clarificationwould not necessarily have resulted inan objection letter from FDA. Second,the agency has repeatedly said that theabsence of an objection letter does notnecessarily indicate acceptance of theclaim. Third, there are apparently alarge number of marketed dietarysupplement products making claims forwhich FDA has not received 30-daynotification letters under section403(r)(6) of the act. (In the proposedrule, FDA estimated that approximately22,500 dietary supplement labels

carried structure/function claims. FDAhad received 2,300 notifications at thetime of the proposed rule. While somenotifications contain more than oneclaim, they do not average 10 claims pernotification.)

FDA also does not agree that theCommission report was necessary toprovide support for this rule. Theproposal was based not only on theCommission report, but also on theagency’s experience in reviewing 30-daynotification letters submitted undersection 403(r)(6) of the act (63 FR 23624at 23625). Although FDA believes therule is consistent with the viewsexpressed in the Commission report, theCommission report was not a necessaryprerequisite for the agency to issue therule. FDA issued the proposal undersection 403(r)(6) of the act (section 6 ofDSHEA) and the rulemaking authorityof section 701(a) of the act, not undersection 12 of DSHEA. FDA takes noview on whether the Commission metits statutory obligations in issuing itsreport. To the extent that the report isbeyond the Commission’s authority,FDA’s experience and section 701(a) ofthe act provide adequate support for therule. Thus, whether or not theCommission exceeded its mandate isirrelevant to the validity of the rule.

With regard to the issues raised aboutthe consistency of the agency’sapproach with the Commission report, itis true that the Commission did notspecifically recommend regulations, butthe Commission did express the viewthat FDA guidance on claims undersection 403(r)(6) of the act would be‘‘appropriate and helpful in clarifyingthe appropriate scope’’ of such claims(the report, p. 38).

As to the agency’s public statementsthat DSHEA is self-implementing, thecomment took those statements out ofcontext. When DSHEA was passed,there was confusion in the industryabout whether the types of statementspermitted by section 403(r)(6) of the actcould be made under the authority ofthe statute alone, in the absence ofimplementing regulations. To clear upthis confusion, at least one agencyofficial publicly said that DSHEA was‘‘self-implementing.’’ Agency statementsto this effect were intended to clarifythat manufacturers were not required towait for FDA to issue implementingregulations before making claims undersection 403(r)(6) of the act; however,they were in no way intended to implythat the agency lacked authority to issueimplementing regulations.

Contrary to the suggestion in one ofthe comments, FDA did not rely on theviews of individual Commissionmembers, but on the official 7-point



‘‘guidance’’ developed by theCommission ‘‘as to what constitutes anacceptable statement of nutritionalsupport of the structure function type’’(the report at pp. 38 and 39). Thecriteria developed by FDA are highlyconsistent with the Commission’sguidance. FDA also agrees that theCommission did not make any findingsor recommendations on the definition ofdisease. As described elsewhere in thisrule, the final rule does not modify theexisting definition of disease found inFDA’s health claims regulations.

(5.) One comment said that FDAshould have admitted that there is andwill be some overlap between diseaseand structure/function claims and thatthe agency should have drafted a rule toprevent extreme overlap betweenstructure/function claims and drug orhealth claims.

FDA disagrees with this comment. Inthe proposed rule, FDA recognized thatsection 403(r)(6) of the act leaves openquestions concerning the distinctionbetween structure/function claims anddisease claims. Diseases cause, and canbe characterized as, abnormalities in thestructure or function of the body. Itwould therefore be possible to describealmost all products intended to treat orprevent disease in terms of their effectson the structure or function of the body,without mentioning the disease itself.

The language of DSHEA, however,does not support treating thosestructure/function claims that are alsodisease claims as statements permittedunder section 403(r)(6) of the act. Asnoted above, section 403(r)(6) of the actcontains two passages that indicateCongress’ intent to exclude from thescope of structure/function claims anyclaim that is also a disease claim.Section 403(r)(6) of the act provides thatstructure/function statements ‘‘may notclaim to diagnose, mitigate, treat, cure,or prevent a specific disease or class ofdiseases.’’ It also requires structure/function claims to be accompanied by adisclaimer stating that the product ‘‘isnot intended to diagnose, treat, cure, orprevent any disease.’’

In light of the statutory framework,FDA concluded in the preamble to theproposed rule that section 403(r)(6) ofthe act authorizes claims related to theeffect of a product on the structure orfunction of the body only if they are notalso disease claims. FDA’s conclusionwas consistent with the policy guidanceoffered by the President’s Commissionon Dietary Supplement Labels. In thereport the Commission offered generalguidance on structure/function claims,including the following:

3. Statements indicating the role of anutrient or dietary ingredient in affecting the

structure or function of humans may be madewhen the statements do not suggest diseaseprevention or treatment.(The report, p. 38)

Accordingly, FDA believes that it isappropriate to define the universe ofpermitted structure/function claims byfirst identifying those claims that shouldbe considered disease claims.Remaining claims about the effect of adietary supplement on the structure orfunction of the body may be acceptablestructure/function claims under section403(r)(6) of the act, provided that theyare consistent with the requirement insection 201(ff)(1) of the act that a dietarysupplement be ‘‘intended to supplementthe diet.’’

(6.) Some comments, particularlythose received at the public hearing orduring the reopened comment period,argued that it is difficult or impossibleto draw principled distinctions betweenstructure/function claims and diseaseclaims. Some of these comments saidthat section 403(r)(6) of the act, whichis premised on such a distinction, is notscientifically based. Other commentsargued that it is not necessary orpractical to draw clear lines betweendisease claims and structure/functionclaims, and that dietary supplementlabeling should instead focus oneducating consumers about theconditions for which a product may beused. According to these comments, ifthere are disease conditions that mightbe implied by a particular claim, thelabeling should, for example, informconsumers of the symptoms of suchconditions, the importance of seekingmedical attention for them, and theirhealth-related consequences. Othercomments argued that consumersreading the labels of dietarysupplements will incorrectly assumethat the information provided thereinhas been reviewed by the governmentand that the claims, express or implied,are supported by the kind of scientificevidence that supports drugs withsimilar claims.

FDA agrees that it may be verydifficult to draw clear lines betweenstructure/function claims and diseaseclaims. Despite the difficulty,implementing section 403(r)(6) of theact requires the agency to draw theselines. FDA would not be carrying out itsstatutory obligations if it abdicatedresponsibility for distinguishingbetween the two types of claims, andinstead permitted dietary supplementsto disseminate information aboutspecific disease states. FDA agrees thatscientifically valid information aboutdiseases is helpful to consumers, if it isdelivered consistently and accurately,but does not agree that section 403(r)(6)

of the act authorizes suchdissemination. FDA strongly believesthat the dissemination of suchinformation on dietary supplementlabels increases the likelihood thatconsumers will believe that thesupplements are intended to treat orprevent the diseases described in thelabeling. Therefore, it is important thatany disease claims in dietarysupplement labeling continue to besubject to prior FDA review to evaluatethe safety and effectiveness of theproduct for the use described orsuggested by the claim.

The agency also notes that there maybe important health-relatedconsequences associated with taking adietary supplement, even if the productdoes not bear disease claims. For thelabeling of a dietary supplement to beconsidered truthful and non-misleading(see sections 403(a) and (r)(6) and201(g)(1) of the act), it must include allinformation that is material in light ofthe claims made for the product and theconsequences that may result from itsuse (see section 201(m)) of the act.

(7.) Many comments discussed therule’s effect on scientific research. Somecomments argued that the proposalwould discourage scientific research ondietary supplements. One commentcontended that such research mightprompt FDA to consider a dietarysupplement to be a drug. Anothercomment said the proposal would‘‘chill’’ the availability of third-partyinformation on dietary supplements.

The agency disagrees with thecomments. The comments provided noevidence, and the agency is aware ofnone, that establishing criteria fordistinguishing structure/function claimsand disease claims will adversely affectthe conduct or use of scientific research.In the agency’s experience, establishingregulatory standards has generated moreresearch rather than less. As describedbelow, some comments frompharmaceutical companies and frompatient organizations expressed thecontrary concern that allowing dietarysupplements to make disease claimswithout FDA review would undermineincentives for rigorous scientificresearch. The agency also notes thatnothing in this rule would treatscientific research or the publication ofresearch results in a scientific journal asevidence that a product is marketed asa dietary supplement or is a drug.

(8.) Several comments addressed therelationship between dietarysupplements and drug products, and theeffects of this regulation on drugproducts and drug development. Somecomments suggested that the proposalrepresented an attempt by FDA to



regulate dietary supplements in amanner that benefits pharmaceuticalinterests or to regulate dietarysupplements in a manner that is similarto European regulatory systems thatapply drug requirements to suchproducts.

In contrast, other comments expressedconcern over the negative effects ofDSHEA and the proposed rule onincentives for pharmaceutical drugdevelopment. One comment asked FDAto provide an ‘‘unambiguousdemarcation’’ that would preserveresearch and development incentives fordrug products and permit evaluation ofopportunities in the dietary supplementmarketplace. According to thiscomment, section 403(r)(6) of the act,and DSHEA generally, were intended tocreate ‘‘parity’’ between the dietarysupplement and food industries withoutundermining research and developmentincentives for the pharmaceuticalindustry and to address a perceivedfailure by FDA to implement the healthclaims provision for dietarysupplements in section 403(r)(5)(D) ofthe act. The comment contended thatsection 403(r)(6) of the act is intendedto provide a limited statutory safeharbor for certain dietary supplementsthat might otherwise be subject toregulation under the health claim rulesfor food or as unapproved new drugs,but it does not permit any and allstructure/function statements for dietarysupplements. Thus, the comment saidFDA should have ‘‘parallelinterpretations’’ of sections 201(g)(1)(C)and 403(r)(6) of the act. The commentsuggested that FDA enforce therequirement of a ‘‘documentedmechanism’’ imposed in section403(r)(6)(A) of the act, which permitsclaims that ‘‘characterize thedocumented mechanism by which anutrient or dietary supplement acts tomaintain’’ structure or function and thatFDA limit claims to ‘‘maintaining,’’rather than ‘‘promoting’’ or ‘‘improving’’structure or function.

FDA does not agree that this rule wasdesigned to benefit the pharmaceuticalindustry or to establish rules that areconsistent with European regulation ofdietary supplements. As noted above,some pharmaceutical companies believethat the rule will harm them bypermitting competition by products thathave not had to undergo rigorous testingor review. Other pharmaceuticalcompanies already produce dietarysupplements and expressed the samereservations about the rule as otherdietary supplement manufacturers.There was also no attempt to model thisrule after European regulation of dietarysupplements.

FDA recognizes the importance ofmaintaining incentives for research andproduct innovation. By establishingcriteria for determining when astatement may be a disease claim, thefinal rule indirectly contributes towardspreserving the incentives forpharmaceutical research anddevelopment by ensuring that productsmarketed for treatment or prevention ofdiseases must all meet the sameregulatory standards. As stated below,FDA believes that if the rule were topermit dietary supplements to carryimplied disease claims, the incentivesfor new drug development could besignificantly undermined.

FDA agrees with the comment that thestructure/function provisions of sections403(r)(6) and 201(g)(1)(C) of the act aresimilar in scope. FDA also agrees that tomake a statement about the mechanismby which a dietary supplementmaintains structure or function, themechanism of action must be‘‘documented.’’ FDA does not agree,however, that this is the only provisionunder which a dietary supplement mayclaim to maintain healthy structure orfunction. Maintenance claims also canbe made under the provision thatauthorizes statements that ‘‘describe therole’’ of a supplement ‘‘intended toaffect the structure or function’’ of thebody (section 403(r)(6)(A) of the act).

In response to the comment askingFDA to limit claims to ‘‘maintaining,’’rather than ‘‘promoting’’ or‘‘improving,’’ structure/function, theagency agrees that ‘‘improving’’ oftensuggests some abnormality or deficiencythat can be treated, so a claim to‘‘improve’’ a structure or function of thebody would be more likely to be adisease claim. On the other hand, aclaim to improve memory or strengthwould be a permitted structure/functionclaim, unless disease treatment wereimplied. Use of the term ‘‘promote’’ maybe acceptable under the portion ofsection 403(r)(6)(A) of the act whichauthorizes claims that ‘‘describe[] therole of a * * * dietary ingredientintended to affect the structure orfunction.’’ Whether a claim for‘‘promoting’’ structure or function is adisease claim will depend on thecontext and nature of the claim. Forexample, a claim that a product ‘‘helpspromote digestion’’ would be astructure/function claim because it doesnot refer explicitly or implicitly to aneffect on a disease state, but a claim thata product promotes low blood pressurewould be considered a disease claim.Both the preamble to the proposed ruleand the Commission recognized thatstatements using the word ‘‘promote’’can be appropriate when the statements

do not suggest disease prevention ortreatment or use for a serious healthcondition that consumers cannotevaluate (see 63 FR 23624 at 23626).

(9.) A few comments objected to thestatement that a dietary supplementbearing an appropriate structure/function claim may be subject toregulation as a drug if there is otherevidence that it is intended for thediagnosis, cure, mitigation, treatment, orprevention of disease. One commentargued that many dietary supplementsare used for medicinal purposes and itwould be ‘‘easy’’ for FDA to findevidence that they were intended forthis purpose based on consumer use ofthe product.

Although FDA’s longstandinginterpretation of section 201(g)(1)(B) ofthe act authorizes the agency to rely onevidence outside the labeling andadvertising of a product to establish itsintended use, FDA does not rely onsuch evidence alone except in unusualcircumstances. For example, the courtshave suggested that if the agency seeksto rely solely on evidence thatconsumers use a product for a particularpurpose to support a finding of intendeduse for that purpose, consumers mustuse the product predominantly or nearlyexclusively for that purpose. (See, e.g.,Action on Smoking and Health (ASH) v.Harris, 655 F.2d 236, 239–240 (D.C. Cir.1980); National Nutritional Foods(NNFA) v. Weinberger, 512 F.2d 688,702 (2d Cir. 1975), cert. denied, 423 U.S.827 (1975).) The fact that someconsumers used a dietary supplementfor medicinal purposes would not byitself be sufficient to establish intendeduse as a drug, if use for medicinalpurposes was not the predominant use.

FDA reiterates, however, that inappropriate circumstances, FDA mayfind that a dietary supplement for whichonly structure/function claims are madein labeling may nevertheless be a drugif there is other evidence of intendeduse to prevent or treat disease.

(10.) Some comments discussed the‘‘disclaimer’’ statement required bysection 403(r)(6)(C) of the act. Thedisclaimer reads as follows: ‘‘Thisstatement has not been evaluated by theFood and Drug Administration. Thisproduct is not intended to diagnose,treat, cure, or prevent any disease.’’ Onecomment said the disclaimer resolvesany consumer confusion betweendietary supplement claims and drugclaims. Another comment said theproposed rule showed that FDA wasimplicitly rejecting the disclaimer’smeaning because the proposed rulewould restrict the amount ofinformation flowing to consumers. Onecomment said the disclaimer reflects



Congress’ understanding of a tensionbetween structure/function and diseaseclaims, while another comment assertedthat the disclaimers required on a labelare an attempt to decrease the amountof space on a label for a structure/function claim.

Section 403(r)(6) of the act requiresdietary supplement manufacturers whowish to make a structure/functionstatement to include the disclaimer,and, since 1997, FDA regulationsregarding the disclaimer have beencodified at § 101.93. However, thedisclaimer’s role does not eliminate theneed for this final rule to establishcriteria for determining whether astatement is a disease claim. Section403(r)(6) of the act provides that astatement for a dietary supplement thatis made under section 403(r)(6) ‘‘maynot claim to diagnose, mitigate, treat,cure, or prevent a specific disease orclass of diseases.’’ Had Congress thoughtthe disclaimer, alone, was sufficient todistinguish between structure/functionclaims and disease claims, it would nothave enacted the restriction againstdisease claims in section 403(r)(6) of theact.

FDA does not agree with the assertionthat the disclaimer, which is expresslyrequired by the act, is a scheme todecrease the space for structure/function claims on a label. FDA believesthat the disclaimer is intended to makesure that consumers understand thatstructure/function claims, unlike healthclaims and claims that appear on thelabels of drugs, are not reviewed by FDAprior to marketing, and to cautionconsumers that dietary supplementsbearing such claims are not fortherapeutic uses.

(11.) Several comments soughtadditional statements or language onproduct labels. One comment supportedthe marketing of dietary supplementsand other substances whoseeffectiveness has not been establishedand that have no appreciable toxicity aslong as the product’s label stated thateffectiveness had not been proven.Another comment said precautions,such as adverse reactions andcontraindications to certain diseasesand medications, are importantinformation for labels. The commentalso sought a description of a dietarysupplement product’s contents as apercentage of a person’s recommendeddaily intake (RDI) and in actual units.

FDA declines to revise the rule assuggested by the comments. With regardto the marketing of dietary supplementswith a label statement that the product’seffectiveness has not been proven, theagency advises that dietary supplementsthat do not do what they claim to do are

misbranded. The act forbids false andmisleading labeling and advertisingclaims and requires businesses to havesubstantiation for any structure/functionclaims they make for dietarysupplements in labeling (see section403(a) and (r)(6)(B)) of the act). Thepresence of a disclaimer indicating thateffectiveness has not been establishedcannot vitiate these statutoryobligations. Therefore, it would beinappropriate for FDA to sanction theuse of effectiveness disclaimers.

Although the act does not prescribeany specific statements concerningadverse reactions or contraindicationsthat dietary supplements must carry, theagency notes that dietary supplementlabeling, like the labeling of all otherFDA-regulated products, is required toinclude all information that is materialin light of consequences that may resultfrom the use of the product orrepresentations made about it (seesections 403(a)(1) and 201(n) of the act).

As for requiring information on thepercentage of RDI and actual units fordietary ingredients in dietarysupplements, FDA agrees that suchinformation is useful. In fact, FDA’snutrition labeling regulations for dietarysupplements generally require thepercentage of the RDI or daily referencevalue (DRV) that a dietary supplementcontains to be given for dietaryingredients that have an RDI or DRV(see § 101.36(b)(2)(iii) (21 CFR101.36(b)(2)(iii))). In addition, theamount in units must be given,regardless of whether an RDI or DRV hasbeen established (see § 101.36(b)(2) and(b)(3) (21 CFR 101.36(b)(2) and (b)(3)).This information can be found on theSupplement Facts panel of dietarysupplements.

(12.) One comment objected toreferring to structure/functionstatements as ‘‘claims.’’ The commentsaid that, under section 403(r)(6) of theact, such statements must be truthfuland non-misleading, so they should becalled ‘‘statements’’ instead of ‘‘claims.’’

FDA has traditionally used the term‘‘claim’’ to refer to any statement madeby a manufacturer that recommends orsuggests a particular use of a product.This term is used for all productsregulated by FDA, including drugs,foods, devices, and dietarysupplements. Use of the term ‘‘claim’’ isnot intended to suggest that a statementis untrue or misleading in any way.

(13.) One comment said that anysubstance used with ‘‘pharmacologicintent’’ should be classified as a drug orbiologic in order to ensure the efficacy,potency, and purity of medicines. Thecomment explained that suchsubstances have a potential for

therapeutic benefit as well as harm, andsuggested that existing and new dietarysupplements that are marketed withhealth-related claims be required toprovide scientific evidence of theirsafety and efficacy as a condition oftheir being marketed as a drug orbiologic.

FDA declines to adopt the comment’ssuggestion. Section 403(r)(6) of the actexpressly authorizes certain structure/function claims for dietary supplements.Many of these claims may be said to be‘‘health-related.’’ (The agency isuncertain what is meant by‘‘pharmacologic intent.’’) Thus, the actdoes not require all substances withhealth-related claims to be classified asa drug or biologic.

Regarding safety and effectivenessevidence for dietary supplements thatbear health-related claims, FDA agreesthat such evidence should continue tobe required where the claim is a healthclaim within the meaning of§ 101.14(a)(1) or a claim that subjectsthe product to regulation as a drugunder section 201(g)(1)(B) of the act.With regard to health-related claims thatare authorized by section 403(r)(6) of theact, section 403(r)(6)(B) does requiremanufacturers to have substantiation fortheir claims. However, the act does notgenerally require dietary supplementmanufacturers that make claims for theirproducts under section 403(r)(6) of theact to provide a premarketdemonstration of safety andeffectiveness to FDA.

(14.) One comment recommended thatFDA not finalize the proposed rulebecause it claimed that the proposal’scriteria were based on a subjectiveevaluation of claims and not onobjective information from marketresearch studies to determine whetherconsumers are confused by the claim.The comment also argued that FDA didnot provide data and informationregarding consumer confusion, and thatall interested parties should be able toevaluate and comment on any databefore FDA finalizes the proposal. Thecomment asserted that a significantlyrevised and limited final rule couldprovide a basic regulatory definition ofdisease and a ‘‘construct’’ for structure/function claims so that detailedregulatory criteria would beunnecessary.

The act does not require marketresearch studies to determine whether aparticular statement is a structure/function claim or disease claim, and itwould be both impractical andinefficient to require such studies todecide the status of every possible claimthat could be made under section403(r)(6) of the act. FDA also does not



believe that market research studies arenecessary to provide a reasonable basisfor the agency’s determinationsconcerning the meaning of labelingclaims. The agency has extensiveexperience in interpreting such claims.The agency has, however, modified thesecond criterion in § 101.93(g)(2)(ii) toeliminate reference to recognition ofsigns and symptoms by consumers orhealth professionals because manycomments objected that this standardwould appear to require consumertesting. FDA has replaced therecognition standard with an objectivestandard.

(15.) One comment said that it wouldbe inappropriate for FDA to issue anyregulation that restricted the scope ofstatements of nutritional support relatedto a nutrient content claim or claimspertaining to a classical nutrientdeficiency-related disease. Thecomment said that claims such as‘‘calcium builds strong bones’’ areacceptable and that FDA should clarifythis fact in the final rule.

FDA agrees that dietary supplementsmay carry structure/function statementsconcerning the relationship of nutrientsand the structure or function of thebody, such as ‘‘calcium builds strongbones.’’ The preamble to the proposedrule also specifically acknowledged thatalthough statements under section403(r)(6) of the act generally may notclaim to diagnose, mitigate, treat, cure,or prevent a specific disease or class ofdiseases, ‘‘such statements may claim abenefit related to a classical nutrientdeficiency disease, provided that theyalso disclose the prevalence of thedisease in the United States’’ (63 FR23624). The final rule codifies thisexception at § 101.93(g)(2), which statesthat ‘‘FDA will find that a statementabout a product claims to diagnose,mitigate, treat, cure, or otherwiseprevent disease (other than a classicalnutrient deficiency disease) * * *’’(emphasis added). Classical nutrientdiseases are also specifically excludedfrom the definition of disease in§ 101.93(g)(1). Thus, because the finalrule already contains the exception, nochange to the rule is necessary.

(16.) Many comments suggested thatFDA issue a guidance document insteadof regulations. Some of the commentsstated that regulations are neitherdesirable nor necessary. Others statedthat a guidance document would beappropriate because it would permitnew information to support newstructure/function claims or because itwould enable FDA to conduct consumerresearch and industry outreachprograms before imposing new rules.Some comments also requested separate

guidance documents for specific claimsor recommended that FDA create or useadvisory committees to help draftguidance documents. Two commentssaid that the Commission report onlyprovided guidance and suggestions, soFDA did not have to issue the proposedrule. Another comment said thatpublishing a guidance document wouldconsume fewer agency resources andthat a rule is unnecessary because theindustry already knows the permissiblescope of statements for dietarysupplements.

FDA disagrees with the comments.The final rule creates uniform,enforceable requirements for structure/function claims. By doing so, the finalrule establishes a ‘‘level playing field’’for all members of the dietarysupplement industry, and permitsrational use of FDA’s limitedenforcement resources. In contrast,guidance documents, although theyrepresent FDA’s best advice on aparticular matter, are not binding on anyparty. Relying solely on guidancedocuments would not be as effective inachieving consistency in the regulationof structure/function claims on dietarysupplements and would lead to case-by-case enforcement.

FDA does, however, intend to issue aguidance document to provideadditional information regardingstructure/function and disease claims.The guidance document wouldcomplement, rather than substitute for,the final rule.

As for those comments stating that aguidance document would permit newinformation to support new structure/function claims or that outreachprograms are necessary, FDA notes thatinterested persons may generate suchinformation regardless of the rule. FDAmay also conduct research or otherprograms or consult advisorycommittees or other persons if suchactions would be helpful. In short,gathering more information orconducting research and other programsis not dependent on whether FDA issuesa guidance document instead of a rule.

(17.) A few comments stated that FDAshould enforce existing laws andregulations, remove unsafe productsfrom the market, take action againstdietary supplements that make‘‘extravagant, unsubstantiated’’ claims,or promote educational activitiesinstead of issuing regulations. Onecomment suggested that FDA resourceswould be better spent reviewing noticessent to the agency instead of issuingregulations. Another commentsuggested that FDA continue to clarifyissues on a case-by-case basis.

FDA disagrees with the comments.Regulations offer several importantadvantages that case-by-caseclarification, individual enforcementactions, and educational activitiesgenerally cannot. For example, whenFDA develops a regulation, it providesnotice, obtains public comment,considers alternatives, and evaluates therule’s potential impacts, costs, andbenefits. Individual enforcement actionsand educational activities are notsubject to these considerations.

Regulations also establish uniform,industry-wide requirements in a singleadministrative proceeding (rulemaking).In contrast, individual enforcementactions focus on distinct facts that maynot lend themselves to uniformapplication to an entire industry.Moreover, enforcement actions areresource-intensive and require multiplesteps, such as inspections, warningletters, and sometimes litigation, beforethey are completed. Educationalactivities may deal with general topicsand provide valuable opportunities fordiscussing issues with FDA, but they donot create uniform requirements.

Regulations are also easier to locatebecause they are published in theFederal Register when they are issued,are codified and published in the Codeof Federal Regulations (CFR) and can befound in libraries and on governmentInternet sites (such as the GovernmentPrinting Office’s website atwww.gpo.gov). In contrast, agencycorrespondence and results ofindividual enforcement actions are notas widely available and may be difficultfor some regulated entities andconsumers to obtain.

Thus, when it comes to establishinguniform, industry-wide requirements,conserving agency resources, andproviding public notice and anopportunity to comment, regulations arepreferable to individual enforcementactions and educational activities.

(18.) A comment suggested that FDAadopt an approach like hazard analysiscritical control point (HACCP) insteadof issuing the rule.

FDA disagrees with the comment.HACCP is best suited for issues relatingto how a product is manufactured. Here,the principal issue is the claims madefor a product rather than how theproduct is made.

(19.) A comment stated that FDAlacks the expertise to determine whethera botanical is a drug or a dietarysupplement. The comment explainedthat botanicals can be used formedicinal purposes, but that they canalso be used for promoting general wellbeing and supporting the structure orfunction of the body. According to the



comment, FDA declared Yellowdock, anherb, to have medicinal purposes only,when the herb also had a long historyof use as a food source.

The comment may havemisinterpreted the rule. The focus ofthis rule is not on whether a substancehas a history of use as a food but onclaims made in the product’s labeling.The rule defines the types of statementsthat may be made concerning a dietarysupplement’s effect on the structure orfunction of the body. FDA has manyyears of experience in regulating andinterpreting health-related productclaims.

(20.) One comment said othercountries (naming several Europeannations) and the World HealthOrganization have established lists ofingredients and botanical products thatare safe and permitted for therapeuticpurposes. The comment suggested thatFDA consider assembling a committeeto establish a similar list for the UnitedStates.

A list of dietary ingredients andbotanical products and their therapeuticuses might provide valuableinformation. Nevertheless, section403(r)(6) of the act permits onlystructure/function claims for dietarysupplements that are not also diseaseclaims, and so such a list would not berelevant to this rulemaking.

(21.) Two comments suggested thatFDA list examples of structure/functionclaims in order to reduce confusion.Another comment would have FDAdescribe both disease claims andstructure/function claims.

FDA intends to issue a guidancedocument that will provide examples ofclaims that would and would not beconsidered disease claims. This finalrule also includes many examples ofstructure/function and disease claims.

B. Permitted Structure/FunctionStatements (§ 101.93(f))

Proposed § 101.93(f) stated thatdietary supplement labels and labelingmay bear structure/function statementsthat are not disease claims within themeaning of proposed § 101.93(g) andthat otherwise comply with thenotification and disclaimer provisionsof § 101.93(a) through (e). FDA isrevising § 101.93(f) on its own initiativeto make it clear that a dietarysupplement may bear a disease claim ifit is the subject of an authorized healthclaim, but that otherwise disease claimswill subject the product to regulation asa drug.

C. Definition of Disease (§ 101.93(g)(1))To assist in describing what

constitutes a disease claim, the

proposed rule contained a definition of‘‘disease.’’ The proposed definition wasbased on standard medical and legaldefinitions of the term (Refs. 2, 3, 4, and5). Proposed § 101.93(g)(1) defined‘‘disease’’ as:

any deviation from, impairment of, orinterruption of the normal structure orfunction of any part, organ, or system (orcombination thereof) of the body that ismanifested by a characteristic set of one ormore signs or symptoms, includinglaboratory or clinical measurements that arecharacteristic of a disease.

The proposed definition would havereplaced an earlier definition issued in1993 as part of the regulationsimplementing the health claimsprovisions of NLEA. The implementingregulations require dietary supplementmanufacturers to obtain priorauthorization of any labeling statementthat characterizes the relationshipbetween a substance in the supplementto a ‘‘disease or a health-relatedcondition’’ (section 403(r)(1)(B) of theact; § 101.14(a)(1)). The phrase ‘‘diseaseor health-related condition’’ wasdefined in those regulations as:

damage to an organ, part, structure, orsystem of the body such that it does notfunction properly (e.g., cardiovasculardisease), or a state of health leading to suchdysfunctioning (e.g., hypertension); exceptthat diseases resulting from essential nutrientdeficiencies (e.g., scurvy, pellagra) are notincluded in this definition * * *.Section 101.14(a)(5) (formerly§ 101.14(a)(6)). The definition wasredesignated as § 101.14(a)(5) effectiveMarch 23, 1999 (see 62 FR 49859,49867).

FDA tentatively concluded that it didnot want to retain the older healthclaims definition because its use of theterm ‘‘damage’’ could be interpreted tolimit the definition to serious or long-term diseases, and could imply thatthere needed to be pathologicalevidence of damage, which is notalways present. For example, mostmental illnesses have no evidence ofanatomic damage, yet are clearlydiseases.

In the July 8, 1999, Federal Registernotice announcing a public meeting andreopening the comment period, FDArequested additional comment on thedefinition of disease. The notice listedfour questions on which it soughtspecific comment: (1) What are theconsequences, with respect to the rangeof acceptable structure/function claims,of adopting: (a) The 1993 definition in§ 101.14(a)(5), or (b) the definition in theproposed rule? (2) If FDA were to retainthe 1993 definition, does the referenceto ‘‘damage’’ exclude any conditionsthat are medically understood to bediseases? Please provide examples. (3) If

it does not exclude any such conditions,is the 1993 definition otherwiseconsistent with current medicaldefinitions of disease? (4) If it doesexclude conditions that are medicallyunderstood to be diseases, could it berevised in a way that would includesuch conditions?

(22.) Almost all of the comments fromthe dietary supplement industry andfrom individuals objected to the newdefinition of disease. Most of thesecomments argued that the newdefinition is too broad, sweeping inmany minor deviations or abnormalitiesthat are not diseases. (Many of thesecomments did not appear to haveunderstood that the definition requirednot only a deviation, but one that ‘‘ismanifested by a characteristic set of oneor more signs or symptoms.’’) Onecomment said that under the newdefinition wrinkles and gray hair wouldqualify as diseases. Some commentsobjected to the fact that the proposeddefinition was not limited to adversedeviations from normal structure orfunction. Other comments argued thatthe breadth of the proposed definition isinconsistent with the intent of DSHEA.Some comments objected to thedistinction between normal andabnormal functions, and argued thatCongress did not intend to limitstructure/function claims to normalstructure or function. Some commentscontended that the definition of diseaseshould not include the phrase‘‘structure or function.’’ Othercomments said that Congress should bepresumed to have been aware of the1993 definition of ‘‘disease or health-related condition’’ and to have intendedFDA to use that definition. Severalcomments argued that the newdefinition of ‘‘disease or health-relatedcondition’’ for health claims wouldinappropriately broaden the scope ofhealth claims for conventional foodsand concomitantly narrow the scope ofacceptable structure/function claims forfoods. One comment said thatredefining ‘‘disease or health-relatedcondition’’ in § 101.14(a)(5) wouldundermine the existing definition of‘‘statement of nutritional support,’’ andwould violate DSHEA and the FirstAmendment. Most of the commentsfrom the dietary supplement industryand from individuals recommended thatFDA return to the 1993 definition.

Most of the comments from healthprofessional groups and groups devotedto specific diseases, including thosewho participated in the August 4, 1999,public meeting, supported the newdefinition of disease as more consistentwith a medical understanding of diseasethan the NLEA definition. Some of these



comments criticized the 1993 definitionbecause of its reliance on ‘‘damage’’ anddysfunction and because of its failure torefer to signs and symptoms. Whilemany comments from the dietarysupplement industry said that norecognized diseases would be excludedby requiring evidence of ‘‘damage,’’comments from health professionalspointed out a number of recognizeddisease conditions for which it is notcurrently possible to identify physicaldamage to an organ, part, or system ofthe body, including most psychiatricdiseases (depression, bipolar disorder,schizophrenia, and obsessivecompulsive disorder, among others),and the early stages of certain metabolicdiseases, including diabetes, geneticdiseases, and nutritional deficiencydiseases.

A few comments offered alternativedefinitions of disease. A major medicalassociation contended that the proposeddefinition would be improved by theaddition of the phrase ‘‘or a state ofhealth leading to such deviation,impairment, or interruption.’’ An OTCdrug and dietary supplement tradeassociation offered the followingalternative definition of disease, whichwould modify the proposed definition:

A disease is any adverse deviation from, orimpairment of, or interruption of the normalstructure or function of any part, organ, orsystem (or combination thereof) of the bodythat is manifested by a characteristic set ofone or more signs or symptoms that are notcharacteristic of a natural state or process.According to this comment, the additionof the word ‘‘adverse’’ appropriatelynarrows the nature of the deviation,‘‘laboratory or clinical measurements’’are appropriately deleted because theyare already included under the conceptof ‘‘signs,’’ and the exclusion of naturalstates ‘‘encompasses Congress’ intent toallow health promotion/maintenanceclaims.’’ One comment suggested that, ifFDA were to retain the 1993 definition,it add the word ‘‘impairment’’ after‘‘damage’’ to cover those recognizeddisease conditions for which evidenceof damage is missing. A pharmaceuticaltrade association urged FDA to convenea small workshop of physicians,patients, and other stakeholders todevelop a consensus on the distinctionbetween disease claims and structure/function claims.

In response to the comments, FDA hasreconsidered the proposed definition ofdisease in § 101.93(g)(1), and hasconcluded that it is not necessary tochange the 1993 health claimsdefinition, because it can be construedin a manner that covers conditions thatare medically understood to be diseases.In light of Congress’ desire to increase

the number of claims that could bemade for dietary supplements withoutsubjecting them to drug regulation, FDAis persuaded that it is thereforeappropriate to retain a narrowerdefinition of disease at this time.

FDA has concluded that the olderhealth claims definition, read as awhole, will not exclude any significantconditions that are medicallyunderstood to be diseases. For example,the requirement of ‘‘damage to an organ,part, structure, or system of the bodysuch that it does not function properly’’indicates that a condition may beconsidered a disease if there is directevidence of structural damage to anorgan, part, structure, or system of thebody, or indirect evidence of damage,indicated by the failure of the organ,part, structure, or system of the body tofunction properly. This interpretation isappropriate because otherwise well-recognized psychiatric diseases,migraine headaches, hypertension,blood lipid disorders, and many otherwell-accepted diseases, could beexcluded from coverage due to the lackof direct evidence of physical damage.The reference to ‘‘a state of healthleading to such dysfunctioning’’ alsopermits the agency to look at evidenceother than actual damage to an organ,part, structure, or system of the body.

FDA does not believe that it would beconstructive to defer a decision on thedefinition of disease and seek a‘‘consensus’’ of stakeholders. Theagency believes that it is unlikely thatdiverse, strongly-held views expressedin written comments and at the publichearing could be forged into aconsensus on this issue. FDA alsobelieves that it is important to reach adecision as soon as possible to permitthe issuance of clear, uniform rules thatwill apply to all dietary supplementlabeling.

Accordingly, the final rule does notinclude a new definition of disease, butincorporates the definition of ‘‘diseaseor health-related condition’’ in§ 101.14(a)(5). If experience shows apublic health need for a different orbroader definition, however, FDA willconsider initiating a rulemaking toamend that definition.

(23.) One comment argued that it isunnecessary for FDA to define disease atall, but that the agency should use a‘‘common sense’’ approach todistinguishing structure/function claimsfrom disease claims. According to thiscomment, dietary supplements shouldbe allowed to make any claim that doesnot contain express references ‘‘tospecific diseases * * * or which canonly be reasonably interpreted to refer

to a specific disease (e.g., ‘helps preventtumors’).’’

FDA does not agree that a definitionof disease is unnecessary. The commentthat made this argument went on to usethe term disease in its ‘‘common sense’’principle, apparently assuming thatthere is some common senseunderstanding of the term. FDA is notaware of any common senseunderstanding of ‘‘disease,’’ and thediversity of comments received in thisrulemaking on the appropriatedefinition of disease supports FDA’sview that a definition is needed if FDAis to enforce section 403(r)(6) of the actfairly and consistently.

(24.) One comment argued that anydefinition of disease should excludesymptoms or diseases that do notnormally require a drug or doctor’s carebecause these states could be consideredpart of ‘‘normal’’ living.

FDA does not agree that DSHEA wasintended to permit structure/functionclaims about diseases that can normallybe treated without a physician’s care.Nothing in the statute or its legislativehistory suggests that Congress intendedto accord different treatment to thissubset of diseases. Diseases that do notordinarily require a physician’s care aregenerally those for which drugs may besold over OTC. (OTC drug claimsinclude both disease claims andstructure/function claims.) Drugscarrying OTC claims are alreadyregulated under rules different fromthose applicable to prescription drugs.FDA has undertaken a comprehensivereview of OTC drug claims andpublished monographs on these claims.Had Congress intended to permit dietarysupplements to make all OTC claims(both disease claims and structure/function claims) without prior review, itcould easily have so indicated. BecauseCongress did not do so, FDA does notbelieve that there is support for treatingthis subset of diseases differently fromother diseases. As discussed elsewherein this document, the structure/functionclaims made for OTC drugs also may bemade, in appropriate circumstances, fordietary supplements under section403(r)(6) of the act.

(25.) One comment argued that it wasirrelevant whether the 1993 definitionexcluded conditions that weremedically understood to be diseases.According to this comment, thedefinition of disease should be based onconsumer understanding rather thanmedical understanding, because DSHEAwas intended to educate consumers.

FDA does not agree that itsinterpretation of a medical term like‘‘disease’’ should ignore medicaldefinitions of the term, unless there is



clear guidance from Congress that itintended a nonmedical definition of theterm. In any case, the commentprovided no argument or evidence thatthe 1993 definition was based on, orreflects, consumer understanding of theterm ‘‘disease.’’

D. Disease Claims (§ 101.93(g)(2))(26.) Many comments agreed with the

statement in proposed § 101.93(g)(2)that, in determining whether astatement is a disease claim, it isappropriate to consider the context inwhich the claim is presented. Onecomment argued, however, thatlanguage of the regulation and preambleshowed that FDA was biased becausethe agency would only consider thecontext of a claim to convert a dietarysupplement to a drug.

FDA does not agree that it willconsider context only to convert anotherwise acceptable structure/functionclaim to a disease claim. The context inwhich a claim appears can provideevidence in either direction.

(27.) One comment argued that therule should have only the followingthree criteria: (1) The words ‘‘diagnose,’’‘‘prevent,’’ ‘‘treat,’’ ‘‘cure,’’ and‘‘mitigate’’ should not be used in astructure/function claim; (2) the words‘‘stimulate,’’ ‘‘maintain,’’ ‘‘support,’’‘‘regulate,’’ and ‘‘promote’’—or othersimilar words—may be used in astructure/function claim to distinguishthe claim from a specific disease claim;and (3) clinical endpoints that arerecognizable to health professionals orconsumers as being related to a diseasemay be used in a structure/functionclaim.

FDA does not believe that the threesuggested criteria provide a sufficientbasis to distinguish between structure/function claims and disease claims.Nothing in these criteria would preventa structure/function claim fromdiscussing a specific disease, explicitlyor implicitly, as long as the claim didnot contain the specific verbs‘‘diagnose,’’ ‘‘prevent,’’ ‘‘treat,’’ ‘‘cure,’’or ‘‘mitigate.’’

(28.) Several comments from medicaland consumer groups supported theestablishment of criteria for structure/function claims, but were concernedthat the criteria in the proposed rulewere too vague and would fail to protectconsumers from misleading claims. Amajor medical association contendedthat some of the structure/functionclaims listed as acceptable in theproposal were debatable and expresseddoubt that the public health would beadequately protected. Some of thesecomments expressed the view that someof the structure/function claims listed in

the proposal in fact imply diseaseprevention. For example, some of thesecomments argued that healthmaintenance claims imply diseaseprevention. On the other hand, acomment from a major dietarysupplement trade association arguedthat the overall impact of the criteriarestricts the value of structure/functionclaims in providing consumers withuseful information about dietarysupplements.

FDA agrees that consumers shouldhave access to, and be allowed toevaluate for themselves, as muchtruthful information about dietarysupplements as is possible, consistentwith the statutory restrictions on diseasetreatment and prevention claims. FDAbelieves that the criteria in this rulestrike a reasonable balance betweenthese competing goals. Undoubtedly,the criteria will not satisfy everyone. Forexample, some of the claims consideredto be structure/function claims mayimply specific disease prevention tosome consumers. Because of theimportance of the context in which aclaim is presented, it will not always bepossible to draw a line betweenstructure/function and disease claims inthis rule with great specificity. FDAbelieves that, within these constraints,the criteria, as finalized, adequatelydistinguish between structure/functionclaims and disease claims. Indeveloping final criteria, the agency hastried to pay particularly close attentionto claims that might relate to serioushealth conditions that patients cannotsafely evaluate on their own. Thequestion of whether health maintenanceclaims necessarily imply diseaseprevention is discussed in more detailbelow.

(29.) One comment, from aCommission member, said the ‘‘dietaryrelationship’’ of a structure/functionclaim is relevant in considering whethersuch a claim is appropriate. Thecomment said that statements fordietary ingredients should ‘‘relate to therole of the dietary ingredient in the dietin achieving effects like those associatedwith the effects of foods.’’ The commentadded that the claim ‘‘should be for aneffect that is similar to the non-diseaseeffects of a food on the body’’ and‘‘phrased to indicate the role of thedietary ingredient in the diet inmaintaining or supporting the ordinaryfunctioning of the body in a mannersimilar to that achieved through foods.’’Thus, the comment would consider aclaim such as ‘‘promotes relaxation’’ tobe appropriate ‘‘only if it is indicated tobe similar to the effects achieved fromfoods, such as by indicating that itprovides a relaxing calming effect like a

cup of tea.’’ While the preamble to theproposed rule considered the claim of‘‘improves absentmindedness’’ to be astructure/function claim, the commentviewed the same claim as a diseaseclaim ‘‘because of the association ofabsentmindedness with Alzheimer’sdisease.’’ The comment continued,‘‘That claim should not be permissiblefor the same reason that a claim that adietary supplement is an ‘oralcontraceptive’ is not permissible—theclaim is simply not one for the effectsof a dietary ingredient.’’

FDA agrees that dietary supplementsmust be ‘‘intended to supplement thediet’’ (section 201(ff) of the act). Ininterpreting section 403(r)(6) of the act,however, FDA believes that it isappropriate to focus on the claims madefor the product. Unlike section201(g)(1)(C) of the act, section 403(r)(6)of the act does not limit authorization tomake structure/function claims (withouttriggering drug approval requirements)to substances that are ‘‘food.’’ FDA notesthat it is developing an overall dietarysupplement strategy and will, when adocument incorporating the strategy isreleased, state how the agency plans toaddress the requirement that dietarysupplements be ‘‘intended tosupplement the diet.’’

(30.) One comment said FDA shoulddevelop a list of ‘‘acceptable subclinical,pre-disease, and normal states’’ that maybe used in structure/function claims.

FDA declines to adopt the comment’ssuggestion. However, this rule containsmany examples of acceptable structure/function claims and FDA intends toissue further guidance listing acceptableclaims.

(31.) One comment argued that allstatements about effects on structure orfunction should be deemed permissibleunless they are already approved drugclaims. The comment noted that‘‘reduces joint pain’’ and ‘‘relievesheadache’’ would not be structure/function claims because they are OTCmonograph claims.

FDA does not agree that such acriterion would appropriatelydiscriminate between structure/functionclaims and disease claims. One kind ofvalid drug claim is a claim related to theeffect of the product on the structure orfunction of the body (section201(g)(1)(C) of the act) but not related todisease prevention or treatment. In otherwords, not all drug claims are diseaseclaims. Congress specifically providedthat structure/function claimsauthorized by section 403(r)(6) of the actdo not, in themselves, subject a dietarysupplement to regulation as a drugunder 201(g)(1)(C) of the act. It thuswould not be appropriate to exclude



from the scope of acceptable structure/function claims OTC monograph claimsor other approved claims for productsclassified as drugs under section201(g)(1)(C) of the act.

(32.) A national pharmacy groupstated that the examples of structure/function and disease claims in theproposal were reasonable and based ongood science and logic, but should beevaluated and revised as necessary overtime.

FDA agrees that it will be necessaryto evaluate the examples over time andto revise them as experience dictates.

(33.) Some comments argued that thetypes of claims permitted under theproposal may discourage seriousapproaches to substantiation becausethe terms used are not scientificallyverifiable. Stating that the preferredmethod of substantiation is an adequateand well-controlled trial, one commentcontended that the claims permittedunder the rule are not amenable to suchproof. According to this comment, thisrule may preclude companies frommeeting the substantiation rules of theFederal Trade Commission (FTC). A fewcomments said that manufacturerscannot substantiate claims that aproduct maintains healthy status. Oneof these comments stated that it wasimpossible to show by adequate studiesthat ‘‘cranberry extract supports healthyurinary tract functioning,’’ and thatcompanies should instead be able toshow that cranberry extract reducesfrequency of urinary tract infections insusceptible people. Similarly, because itis ‘‘impossible’’ to test whether St.John’s Wort ‘‘supports mood’’ in thegeneral population, companies need tobe able to test its effect on depressedpeople.

FDA agrees that some structure/function claims that are acceptableunder DSHEA may be difficult tosubstantiate. For example, somestructure/function claims currently inthe marketplace use terms that do nothave clear scientific meaning. Otherclaims concern health maintenance inthe general population and thereforecould require studies in a largepopulation for substantiation. FDAbelieves, however, that such claims arewithin the intended scope of section403(r)(6) of the act. Difficulty insubstantiating them does not alter theterms of the statute. Manufacturers areresponsible for determining whetherclaims for their products can beappropriately substantiated, and to useonly those claims for which they havesubstantiation. FDA does not agree thatdifficulty in substantiating a particularclaim justifies the use of express orimplied disease claims for which

methods of substantiation may be morestraightforward. Such an approachwould turn section 403(r)(6) of the acton its head.

FDA also does not agree that it isimpossible to substantiate the claimsdescribed in the comments. Forexample, to substantiate the claim‘‘supports mood,’’ it is not necessary tostudy the effects of a substance onclinical depression. Instead, it is quitepossible to assess the effects of asubstance on mood changes that do notconstitute clinical depression.

E. Effect on Disease or Class of Diseases(§ 101.93(g)(2)(i))

Under proposed § 101.93(g)(2)(i), astatement would be considered adisease claim if it explicitly orimplicitly claimed an effect on aspecific disease or class of diseases.FDA included the following examples ofsuch disease claims: ‘‘Protective againstthe development of cancer,’’ ‘‘reducesthe pain and stiffness associated witharthritis,’’ ‘‘decreases the effects ofalcohol intoxication,’’ or ‘‘alleviatesconstipation.’’ FDA included thefollowing examples of claims that donot refer explicitly or implicitly to aneffect on a specific disease state: ‘‘Helpspromote urinary tract health,’’ ‘‘helpsmaintain cardiovascular function and ahealthy circulatory system,’’ ‘‘helpsmaintain intestinal flora,’’ and‘‘promotes relaxation.’’ FDA proposed totreat both express and implied diseaseclaims as disease claims that could notbe made for dietary supplementswithout prior review either as healthclaims or as drug claims. Implieddisease claims do not mention the nameof a specific disease, but refer toidentifiable characteristics of a diseasefrom which the disease itself may beinferred. There are many possible waysto imply treatment or prevention ofdisease, from listing the characteristicsigns and symptoms of the disease toproviding images of people sufferingfrom the disease. Nine of the 10 criteriaproposed by FDA for identifying diseaseclaims could be considered methods ofimplying disease treatment orprevention.

In the July 8, 1999, Federal Registernotice announcing a public meeting andreopening the comment period, FDAsought additional comment on theapplicability of the rule to implieddisease claims. The discussion in thenotice offered three examples ofpossible implied disease claims: (1)‘‘shrinks tumors of the lung’’ or‘‘prevents development of malignanttumors’’ (‘‘treats cancer’’ would be thecorresponding express claim); (2)‘‘prevention of seizures’’ (‘‘treatment of

epilepsy’’ would be the correspondingexpress claim); (3) ‘‘relief of sneezing,runny nose, and itchy watery eyescaused by exposure to pollen or otherallergens’’ (‘‘treatment of hayfever’’would be the corresponding expressclaim). The notice listed four questionsrelated to implied disease claims onwhich the agency sought specificcomments: (1) If implied disease claimsshould be permitted, has FDA correctlydrawn the line between what constitutesan express disease claim and whatconstitutes a permitted implied claim?(2) If such claims should be permitted,what are representative examples of thetypes of implied disease claims thatshould be permitted without priorreview? (3) Are the examples of impliedclaims mentioned in the July 8 noticeappropriate structure/function claims?(4) Is a claim that a product ‘‘maintainshealthy function’’ an implied diseaseclaim in all cases? If not, under whatcircumstances is such a claim not animplied disease claim?

(34.) Many comments agreed withproposed § 101.93(g)(2)(I) that structure/function statements should notexplicitly or implicitly mention specificdiseases or class of diseases. Thesecomments contended that consumerscannot distinguish between implied andexpress disease claims and thatpermitting implied disease claims posessignificant dangers to consumers withdiseases. According to these comments,permitting implied disease claims ondietary supplements may causeconsumers to delay or forego effectivetreatment for serious diseases withoutassurance that the dietary supplementthat has been substituted is safe oreffective for the disease. Somecomments also argued that permittingimplied disease claims on dietarysupplements will undermine the drugapproval process by permitting dietarysupplement manufacturers to marketproducts for essentially the sameindications for which pharmaceuticalcompanies have spent millions ofdollars obtaining approval.

Many other comments objected totreating implied disease claims asdisease claims, arguing that dietarysupplements should be allowed to carryany truthful claim that does notexplicitly refer to a specific disease.Some comments argued that Congressintended consumers to have access to asmuch information about supplements aspossible. Other comments contendedthat barring implied disease claimseliminates any meaningful claims fordietary supplements. Other commentsargued that treating implied claims asdisease claims gives FDA ‘‘unlimiteddiscretion’’ to treat structure/function



claims as disease claims. Somecomments, however, agreed that diseaseclaims may be implied as well asexpress, and said that it is appropriateto consider a structure/functionstatement in context to determinewhether it conveys a disease claim.

FDA continues to believe thatstructure/function claims should notimply disease treatment or prevention.Most disease treatment or preventionclaims, including claims about seriousand life-threatening diseases, can bedescribed in a manner that will be easilyunderstood by consumers withoutexpress reference to a specific disease.The following examples of implieddisease claims demonstrate that it is notdifficult to convey prevention ortreatment of a specific disease or classof diseases without actually mentioningthe name of the disease, which are givenin parentheses: ‘‘Relieves crushing chestpain’’ (angina or heart attack), ‘‘preventsbone fragility in post-menopausalwomen’’ (osteoporosis), ‘‘improves jointmobility and reduces joint inflammationand pain’’ (rheumatoid arthritis), ‘‘healsstomach or duodenal lesions andbleeding’’ (ulcers), ‘‘anticonvulsant’’(epilepsy), ‘‘relief of bronchospasm’’(asthma), ‘‘prevents wasting in personswith weakened immune systems’’(AIDS) (acquired immune deficiencysyndrome), ‘‘prevents irregularheartbeat’’ (arrhythmias), ‘‘controlsblood sugar in persons with insufficientinsulin’’ (diabetes), ‘‘prevents the spreadof neoplastic cells’’ (prevention ofcancer metastases); ‘‘antibiotic’’(infections), ‘‘herbal Prozac’’(depression). The distinction betweenimplied and express disease claims isthus, in many cases, a semantic one thathas little, if any, practical meaning toconsumers. The argument that Congressintended to encourage the free flow ofinformation about dietary supplementsand therefore intended to permitimplied disease claims is illogical. IfCongress wanted to ensure thatconsumers receive information abouthow these products can treat or preventdiseases, it is difficult to imagine whyit would have specifically denied theright to make such claims expressly, andallowed manufacturers to make theclaims only by implication.

There are also serious public healthquestions raised by implied diseaseclaims. Treatment and prevention ofdisease are serious matters, and thestatute reflects a congressional judgmentthat consumers deserve to have claimsfor such uses reviewed by experts forproof of safety and effectiveness. Inaddition, permitting dietary supplementmanufacturers to make implied diseaseclaims without prior review would

allow them to compete unfairly withprescription and OTC drugs, which arerequired to establish their safety andeffectiveness for disease treatment andprevention before being marketed.Pharmaceutical manufacturers, facedwith this competition, might be lesslikely to undertake future research anddevelopment, compromising one of thenation’s most important sources oftherapeutic advances. Had Congressintended to allow implied diseaseclaims when it authorized dietarysupplement manufacturers to makestructure/function claims without priorreview, it could easily have made clearits intention through express statutorylanguage or legislative history. Asdiscussed below, Congress did not doso.

FDA does not agree that the final ruleeliminates all meaningful claims fordietary supplements. FDA believes thatthere are many meaningful structure/function claims that can be madewithout implying disease treatment orprevention, and has listed a number ofsuch claims in this preamble.

FDA does not agree that treatingimplied claims as disease claims givesthe agency unfettered discretion to treatall structure/function claims as diseaseclaims. The purpose of this rule is toclarify which claims are structure/function claims permitted under section403(r)(6) of the act and which aredisease claims. Both in the proposedrule and in this final rule, FDA hasprovided many examples of specificclaims that would be acceptablestructure/function claims.

(35.) Many comments pointed to threeprovisions of DSHEA as evidence thatCongress intended to include implieddisease claims among structure/functionclaims permitted under section 403(r)(6)of the act. First, the ‘‘Findings’’ sectionof DSHEA refers to the relationshipbetween dietary supplements anddisease prevention. Many commentsargued that Congress would not havemade statutory findings linking dietarysupplements to disease prevention if itintended that FDA could prohibit suchreferences.

Second, section 403(r)(6) of the actstates that structure/function statementsmay not ‘‘claim’’ to treat or preventdisease, and, according to thecomments, this term should be read torefer only to express claims. Somecomments noted that section 403(r)(6) ofthe act does not use the word ‘‘implied’’to qualify the term ‘‘claims,’’ andcontrasted the language of the drugdefinition in section 201(g)(1)(B) of theact (‘‘articles intended for use in thediagnosis, cure, mitigation, treatment, orprevention of disease’’) with the

language of section 403(r)(6)(C) of theact, which states that a structure/function statement may not ‘‘claim’’ todiagnose, cure, mitigate, treat, orprevent disease. One comment agreedwith the proposal’s statement that whileDSHEA authorizes structure/functionclaims that are not also disease claims,but nevertheless asserted that the statuteauthorizes structure/function claimsthat imply ‘‘some protection againstdisease.’’ This comment reasoned thatthe act, as amended by DSHEA, allowsdietary supplements to be ‘‘intended’’ toaffect the structure or function of thebody, provided that the product doesnot ‘‘expressly claim to prevent, etc.disease’’ (emphasis in original) and theproduct bears ‘‘an express, formaldisclaimer of an intent to prevent, etc.disease.’’ The comment also said thatthe Commission report only referred toexpress claims.

Third, DSHEA requires structure/function claims to be accompanied by adisclaimer that reads, in part: ‘‘[T]hisproduct is not intended to diagnose,treat, cure, or prevent any disease.’’According to some comments, Congressunderstood that specific diseasetreatment or prevention effects can alsobe described as effects on the structureor function of the body, and resolvedthe tension by requiring the disclaimer.In contrast, however, another commentargued that the drug definition insection 201(g)(1)(B) of the act stillapplies to dietary supplements becausethe exemption for dietary supplementsadded to section 201(g)(1) applies onlyto the structure/function definition insection 201(g)(1)(C). Many commentsargued generally that DSHEA wasintended to promote the free flow oftruthful information about dietarysupplements, and that prohibitingimplied disease claims is contrary tothis legislative goal.

FDA does not agree that DSHEAauthorizes dietary supplementmanufacturers to make implied diseaseclaims without prior review of theclaims. There is no express provision ofDSHEA that authorizes implied diseaseclaims, and a construction of DSHEAthat permitted such claims would befundamentally incompatible withimportant provisions of the act thatwere squarely before Congress when itpassed DSHEA, including thedefinitions of ‘‘drug’’ and ‘‘new drug’’and the health claims provisions ofNLEA.

As described above, Congress createda partial exemption for dietarysupplements from the definition of drugin section 201(g)(1)(C) of the act byproviding that truthful and non-misleading claims under section



403(r)(6) of the act do not in themselvestrigger drug regulation. Congress didnot, however, create any exemptionfrom section 201(g)(1)(B) of the act fordietary supplements. Thus, dietarysupplements that are ‘‘intended for usein the diagnosis, cure, mitigation,treatment, or prevention of disease’’ aresubject to regulation as drugs under theact. It has been FDA’s longstandinginterpretation of section 201(g)(1)(B) ofthe act that the phrase ‘‘intended foruse’’ refers to the objective intent of themanufacturer, which is not limited to amanufacturer’s express representations.See § 201.128 (21 CFR 201.128); NNFAv. Weinberger, 557 F.2d 325, 334 (2dCir. 1977) (‘‘the FDA is not bound by themanufacturer’s subjective claims ofintent,’’ but may establish intent ‘‘on thebasis of objective evidence’’). Evidenceof objective intent can come from avariety of sources, and may include bothimplied and express claims (UnitedStates v. Undetermined Quantities * * *Pets Smellfree, 22 F.3d 235 (10th Cir.1994); United States v. Storage SpacesDesignated Nos. ‘‘8’’ and ‘‘49’’, 777 F.2d1363, 1366 (9th Cir. 1985) (‘‘intent maybe derived or inferred from labeling,promotional material, advertising, orany other relevant source’’), cert.denied, 479 U.S. 1086 (1987); UnitedStates v. Kasz Enterprises, Inc. 855 F.Supp. 534, 539, 543–44 (D.R.I. 1994),modified on other grounds, 862 F. Supp.717 (D.R.I. 1994); United States v.Articles of Drug * * * Neptone, 568 F.Supp. 1182 (N.D. Ca. 1983); UnitedStates v. * * * Vitasafe, 226 F. Supp. 266(D.N.J. 1964); United States v. 14 105Pound Bags * * * Mineral Compound,118 F. Supp. 837 (D.C. Idaho 1953);United States v. 43 1/2 Gross RubberProphylactics, 65 F. Supp. 534, 535 (D.Minn. 1946), aff’d sub nom. Gellman v.United States, 159 F.2d 881 (8th Cir.1947); 59 FR 6084, 6088 (February 9,1994) (terms ‘‘antibacterial,’’‘‘antimicrobial,’’ ‘‘antiseptic,’’ or ‘‘killsgerms’’ constitute implied drug claimsthat cause products carrying them to bedrugs); 58 FR 47611, 47612 (September9, 1993) (labeling indicating that‘‘hormones’’ are present in a productconstitutes implied drug claim); 58 FR28194, 28204 (May 12, 1993) (productscarrying term ‘‘sunscreen’’ are drugsbecause ‘‘sunscreen’’ implies diseaseprevention, even if not expresslypromoted for prevention of skincancer)).

Thus, interpreting section 403(r)(6) ofthe act as permitting implied diseaseclaims would be irreconcilable withFDA’s longstanding interpretation ofsection 201(g)(1)(B) of the act, whichtreats such claims as drug claims.

Permitting implied disease claims asstructure/function claims would alsoconflict with the health claims schemeestablished in section 403(r)(1) through(r)(1)(5) of the act, which requires foodand dietary supplement manufacturersto obtain health claim authorizationbefore making a claim ‘‘which expresslyor by implication’’ characterizes therelationship of a nutrient to a disease orhealth-related condition. Under thisprovision, a claim that characterized, byimplication, the relationship between adietary supplement ingredient and adisease would require authorization as ahealth claim. Interpreting section403(r)(6) of the act as permitting thesame implied claim withoutauthorization of a health claim directlyconflicts with 403(r)(1) through (r)(1)(5)of the act.

None of the statutory provisionsrelied on by the comments providespersuasive support for the conclusionthat structure/function claims can implydisease treatment or prevention.

FDA agrees that the Findings sectionof DSHEA includes statements linkingdietary supplements and diseaseprevention. However, in addition to thetypes of claims authorized for dietarysupplements in section 403(r)(6) of theact, the act specifically authorizesdietary supplements to bear healthclaims. Health claims are expresslydescribed in the statute as claims thatcharacterize the link between a nutrientand a disease or health-relatedcondition (section 403(r)(1)(B) of theact). The statements in the ‘‘Findings’’section of the DSHEA are entirelyconsistent with this scheme and do notcompel the conclusion that claimslinking dietary supplements and diseaseprevention may be made as structure/function claims.

The use of the word ‘‘claim’’ ratherthan ‘‘intended for use’’ in section403(r)(6) of the act also does not showthat Congress intended to permitimplied disease claims. First, thecomment cites no authority, and FDA isaware of none, for the proposition thatthe meaning of the word ‘‘claim’’ islimited to ‘‘express claim.’’ Moreimportantly, section 403(r)(6) of the actdoes not stand by itself. As Congressrecognized when it provided thatdietary supplements making appropriateclaims under section 403(r)(6) of the actdo not thereby become drugs undersection 201(g)(1)(C) of the act, section403(r)(6) must be read in conjunctionwith section 201(g)(1). As describedabove, section 201(g)(1)(B) of the actcontinues to apply to dietarysupplements and treats them as drugs ifthey are ‘‘intended for use in thediagnosis, cure, mitigation, treatment, or

prevention of disease.’’ FDA hasinterpreted section 201(g)(1)(B) of theact to cover both express and impliedclaims for more than 50 years. HadCongress intended 403(r)(6) of the act topermit any claims covered by section201(g)(1)(B) of the act, it would havehad to provide an exemption from thelatter section.

Further, FDA does not agree that theCommission report referred only toexpress claims. In its guidance onstatements under section 403(r)(6) of theact, the Commission specifically saidthat such statements ‘‘should be distinctfrom NLEA health claims in that they donot state or imply a link between asupplement and prevention of a specificdisease or health-related condition’’ (thereport, p. 38) (emphasis added). Inaddition, the Commission cautionedthat claims using terms such as, e.g.,‘‘support,’’ ‘‘maintain,’’ or ‘‘promote’’are appropriate only if they do not‘‘suggest disease prevention or treatmentor use for a serious health condition thatis beyond the ability of the consumer toevaluate’’ (the report, p. 38) (emphasisadded). Clearly, the Commission wasconcerned about implied claims as wellas express claims.

FDA also does not agree that therequired disclaimer demonstrates anintention to permit implied claims. Tothe contrary, FDA believes that thedisclaimer language (‘‘This product isnot intended to diagnose, treat, cure, orprevent any disease’’), which is virtuallyidentical to the language of section201(g)(1)(B) of the act, provides furtherevidence that Congress did not intendsection 403(r)(6) of the act claims tooverlap section 201(g)(1)(B) claims. As apractical matter, it is unreasonable tointerpret section 403(r)(6) of the act asinviting a communication to consumerslike the following: ‘‘This productprevents bone fractures in post-menopausal women due to bone loss.This product is not intended todiagnose, treat, cure, or prevent anydisease.’’ The comments suggested thatthe addition of the disclaimer wouldsomehow clarify the product’s purposeto consumers. The comments providedno support, however, for their view thatconsumers reading the disclaimerwould interpret it as eliminatingimplications in the remainder of thelabeling that the product treats orprevents disease. FDA believes that thetwo statements simply contradict oneanother and could confuse consumers.Indeed, FDA is concerned thatjuxtaposing two such contradictorystatements is likely to cause consumersto ignore the disclaimer required bysection 403(r)(6) of the act, underminingits effectiveness.



(36.) A few comments addressed theexamples of implied claims listed in theJuly 8, 1999, Federal Register notice.Some comments said that all of theexamples were appropriate structure/function claims. Two commentssuggested that ‘‘shrinks tumors,’’‘‘prevents development of malignanttumors,’’ and ‘‘prevents seizures’’ areexpress disease claims because theyemploy ‘‘synonyms’’ for specificdiseases. According to these comments,‘‘tumor’’ is a synonym for cancer, and‘‘seizure’’ is a synonym for epilepsy.Another comment said that FDA shouldtreat as implied disease claims onlythose claims ‘‘where there is a directcausal relationship between thestructure/function parameter identifiedin the claim and a specific knowndisease.’’ According to this comment, atumor is a ‘‘direct manifestation ofcancer’’ and therefore reference to atumor is a disease claim. In contrast,risk factors for disease, in which thecomment includes elevated cholesterol,are not direct manifestations of adisease, and therefore may be thesubject of structure/function claims.Another comment contended thatdisease claims should be limited toexpress claims and to terms ormeasurements that are ‘‘surrogates forthe disease itself.’’ According to thiscomment, tumors are a surrogate forcancer, but elevated cholesterol is not asurrogate for heart disease. Onecomment argued that ‘‘relief of sneezing,runny nose, and itchy watery eyescaused by exposure to pollen or otherallergens’’ is an acceptable structure/function claim, but did not explain why.

FDA has considered these comments,but does not believe that any of themhave provided a principle thatdistinguishes between claims thatconsumers will understand as diseaseclaims and those that will not beunderstood as disease claims. Accordingto the comments, some of the claimsthat FDA offered as examples of implieddisease claims should not be allowed asstructure/function claims. FDA agreesthat claims that refer to synonyms fordisease, direct manifestations of disease,and surrogates for disease are diseaseclaims. Each of these principles,however, would permit many types ofimplied disease claims that would beclearly understood by consumers asdisease claims, e.g., ‘‘Herbal Prozac’’and ‘‘antibiotic.’’

(37.) Some comments argued that it isimpossible to construct a structure/function claim that does not implydisease prevention or treatment. Severalof these comments claimed that healthpromotion claims inevitably implydisease prevention.

FDA does not agree that everystructure/function claim implies diseaseprevention or treatment. In the proposedrule, FDA provided examples of manytypes of claims that the agency wouldnot consider implied disease claims,and has expanded that list in the finalrule.

(38.) Some comments disagreed withFDA’s examples of disease claims in theproposed rule. These comments statedthat intoxication and constipation arenot in and of themselves diseases, andthat these conditions are not readilyunderstood by consumers as diseases. Afew comments argued that alcoholintoxication is a ‘‘self-inducedcondition’’ and not a disease.

FDA continues to believe that alcoholintoxication, like all poisonings(mushroom, digitalis, or any drugoverdose), meets the definition ofdisease, albeit a transient disease. Thedefinition in § 101.14(a)(5), which FDAis incorporating in this rule, states, inpart, that a disease is ‘‘damage to anorgan, part or structure, or system of thebody such that it does not functionproperly * * *’’ All poisonings, likealcohol intoxication, cause dose-relateddysfunctioning and damage, rangingfrom mild impairments to death.Alcohol intoxication causes temporarydamage to brain function, causingimpairments of judgment, attention,reflexes, and coordination. The fact thatit is ‘‘self-induced’’ does not remove itfrom the definition of disease.Deliberate barbiturate overdoses are alsoself-induced, but clearly meet thedefinition of disease.

FDA has considered the comments onconstipation and agrees that certainconstipation claims should not betreated as disease claims. Constipationhas a variety of causes, many of themunrelated to disease. For example,constipation can be caused by changesin diet and schedule, and by travel.Constipation can also, however, be asymptom of such serious diseases asbowel obstruction and irritable bowelsyndrome. FDA is aware that there maybe differences of opinion about whetheroccasional constipation, alone,constitutes a disease, but believes thattreating it as a disease would not beconsistent with the intent of DSHEA.‘‘For relief of occasional constipation’’would therefore not be considered adisease claim under the rule. Thelabeling of a product that claimed totreat occasional constipation shouldmake clear, however, that the product isnot intended to be used to treat chronicconstipation, which may be a symptomof a serious disease.

(39.) One comment questionedwhether a claim that begins, ‘‘According

to the National Cancer Institute’’ wouldbe a disease claim because it used theword ‘‘cancer.’’

Although the National CancerInstitute (NCI) is associated with thetreatment and prevention of cancer,such a statement will be considered adisease claim only if, within the contextof the total labeling, the statement canbe reasonably understood to relate theproduct to the disease listed in theorganization’s name, e.g., cancer. Forexample, FDA would regard as a diseaseclaim ‘‘According to the NationalCancer Institute, ingredient X protectssmokers’ lungs.’’

F. Signs or Symptoms of Disease(§ 101.93(g)(2)(ii))

Under proposed § 101.93(g)(2)(ii), astatement would be considered adisease claim if it explicitly orimplicitly claimed an effect (usingscientific or lay terminology) on one ormore signs or symptoms that arerecognizable to health care professionalsor consumers as being characteristic ofa specific disease or of a number ofdiseases. FDA provided as examples ofsuch disease claims: ‘‘Improves urineflow in men over 50 years old,’’ ‘‘lowerscholesterol,’’ ‘‘reduces joint pain,’’ and‘‘relieves headache.’’ Stating that claimsof an effect on symptoms that are notrecognizable as characteristic of aspecific disease or diseases would notconstitute disease claims, FDA providedthe following examples of acceptablestructure/function claims: ‘‘Reducesstress and frustration,’’ ‘‘inhibits plateletaggregation,’’ and ‘‘improvesabsentmindedness.’’ The agency alsostated that if the context did not suggesttreatment or prevention of a disease, aclaim that a substance helps maintainnormal function would not ordinarily bea disease claim. Examples included:‘‘Helps maintain a healthy cholesterollevel,’’ or ‘‘helps maintain regularity.’’

FDA specifically requested commenton the distinction between maintainingnormal function, which is potentiallythe basis for an acceptable structure/function claim, and preventing ortreating abnormal function, which ispotentially a disease claim. FDA notedthat the members of the Commissionwere divided on this issue, but that thefinal report concluded that ‘‘statementsthat mention a body system, organ, orfunction affected by the supplementusing terms such as ‘stimulate,’‘maintain,’ ‘support,’ ‘regulate,’ or‘promote’ can be appropriate when thestatements do not suggest diseaseprevention or treatment or use for aserious health condition that is beyondthe ability of the consumer to evaluate’’(the report, p. 38). Recognizing that



claims relating to maintaining healthycholesterol levels raise particularlydifficult issues, FDA sought specificcomment on these claims.

(40.) Many comments frommanufacturers and individuals objectedto proposed § 101.93(g)(2)(ii). Some ofthese comments argued that basing thecriterion on which signs and symptomswere ‘‘recognizable’’ to health careprofessionals or consumers was toovague, and that it was unclear whatproportion of health care professionalsor consumers would be necessary toestablish recognition. Some commentsasked whether FDA expectedmanufacturers to conduct consumersurveys. Other comments urged thatFDA itself conduct consumer surveys todetermine which signs and symptomswere recognizable to consumers asimplied disease claims. Other commentsargued that the proposed provisionwould create a moving target because‘‘as soon as consumers understood thatcertain signs and symptoms arecharacteristic of a disease—that is, assoon as consumers understood why theyshould take a particular supplement—FDA could * * * prohibit a product labelfrom bearing the substantive claimsinformation.’’

FDA agrees with these comments thatthe proposal’s focus on recognition ofsigns and symptoms by consumers orhealth professionals might have madethe provision difficult to apply, both formanufacturers and for the agency.Accordingly, the agency has substituteda more objective criterion. The final ruleeliminates the reference to recognition,and focuses simply on whether thelabeling suggests that the product willproduce a change in the characteristicsigns or symptoms of a specific diseaseor class of diseases. FDA believes thatit will be easier for manufacturers toverify whether symptoms are in factcharacteristic of a disease. FDA andmanufacturers may look to medical textsand other objective sources ofinformation about disease to determinewhether a label implies treatment orprevention of disease by listing thecharacteristic signs and symptoms of adisease or class of diseases.

FDA notes that the standard in therule may be met if characteristic signsand symptoms are referred to either intechnical or lay language. It also wouldnot be necessary to mention everypossible sign or symptom of a disease tomeet this standard. Instead, the standardfocuses on whether the labeling suggeststhat the product will produce a changein a set of one or more signs orsymptoms that are characteristic of thedisease.

FDA does not agree with the commentthat objected to the recognition standardbecause it would prohibit a claim ‘‘assoon as consumers understood thatcertain signs and symptoms arecharacteristic of a disease—that is, assoon as consumers understood why theyshould take a particular supplement * **.’’ This comment assumes that the onlyreason people take dietary supplementsis to treat or prevent disease and that itis appropriate to market supplements byimplying that they can do so. Manypeople take dietary supplements forhealth-related reasons that do notinvolve treatment or prevention ofspecific diseases. As discussedelsewhere in this document, FDA doesnot believe that the act permitsstructure/function claims to implytreatment or prevention of specificdiseases.

(41.) Several comments contendedthat the recognition standard was toorestrictive because all signs orsymptoms relating to the structure orfunction of the body are potentiallyrecognizable to health care professionalsand educated consumers ascharacteristic of some specific disease.Another comment argued that theproposal to treat references to signs andsymptoms as disease claims wasarbitrary and artificial. The commentsaid that specific examples of diseaseclaims used in the proposal could aseasily refer to nondisease states, e.g.,‘‘reduces joint pain’’ could refer to over-exercise. Conversely, ‘‘stress andfrustration’’ could refer to anxiety anddepression. Another commentcontended that ‘‘reduces joint pain’’ isan acceptable structure/function claim ifother language or graphics in thelabeling clearly communicatedtreatment of conditions unrelated toarthritis. One comment asked whether‘‘helps support cartilage and jointfunction’’ would constitute apermissible structure/function claim.Some comments said that references tosigns and symptoms should not beevidence of a disease claim becausesigns and symptoms can be associatedwith a number of varying conditions.One comment claimed that ‘‘inhibitsplatelet aggregation’’ does not meananything to most consumers. On theother hand, some medical groups,groups devoted to specific diseases, andothers expressed concern that theexamples of structure/function claimsprovided by FDA permitted referencesto signs or symptoms that imply diseasetreatment or prevention. According toone comment, ‘‘inhibits plateletaggregation’’ could be interpreted tomean ‘‘prevents heart attack,’’ and

‘‘improves absentmindedness’’ could beinterpreted as a treatment forAlzheimer’s disease.

FDA believes that removing thereference to recognition by consumersor health professionals from§ 101.93(g)(2)(ii) will permit a clearerdistinction between those signs andsymptoms that imply a disease andthose that do not. The focus will be onwhether specific signs or symptoms arecharacteristic of a disease, based onobjective sources. FDA does not believethat ‘‘improves absentmindedness’’ or‘‘relieves stress and frustration’’ arecharacteristic of the specific diseasesmentioned in the comments. FDA agreesthat some signs and symptoms areassociated with such a wide variety ofdiseases and nondisease states that theymay not imply a specific disease or classof diseases. For example, FDA wouldnot interpret ‘‘improvesabsentmindedness’’ as implyingtreatment of Alzheimer’s diseasebecause absentmindedness is not asserious as the type of memory losscharacteristically suffered byAlzheimer’s patients; absentmindednessis, in fact, suffered predominantly bypeople who do not have Alzheimer’sdisease or any other disease. Stress andfrustration, while associated with someanxiety disorders, are not thecharacteristic symptoms of thosedisorders; in addition, these symptomsare equally associated with many othernondisease states.

The agency does agree, however, withthe comment that ‘‘inhibits plateletaggregation’’ is an implied diseasetreatment or prevention claim. Althoughplatelet aggregation is a normal functionneeded to maintain homeostasis,inhibiting or decreasing plateletaggregation is a well-recognized therapyfor the prevention of stroke andrecurrent heart attack (see, e.g., 63 FR56802, October 23, 1998 (final rule forprofessional labeling of aspirin forcardiovascular, cerebrovascular, andrheumatologic uses); 53 FR 46204,November 16, 1988, (internal analgesictentative final monograph)). Inhibitingor decreasing platelet aggregation is themechanism of action of a number ofdrug products approved for thetreatment or prevention of stroke andheart attack. Thus, the agency wouldconsider a claim to inhibit normalplatelet function to be an implied claimto treat or prevent these diseaseconditions.

FDA also believes that ‘‘joint pain’’ ischaracteristic of arthritis. According tothe Merck Manual, joint tenderness isthe most sensitive physical sign ofrheumatoid arthritis (Ref. 6). The claim‘‘helps support cartilage and joint



function,’’ on the other hand, would bea permissible structure/function claim,because it relates to maintaining normalfunction rather than treating joint pain.

(42.) One comment suggested thatclaims about a physiologic marker orsymptom should be regarded as diseaseclaims in two situations: (1) If thephysiologic marker or symptom of adisease is described as beingquantifiably linked to that disease in anofficial government health agencysummary statement or consensus report,or (2) if most clinicians treating patientswith the condition prescribeprescription drugs to modify the markerand historically do so without includingnutritional or dietary intervention aspart of the treatment. According to thiscomment, references to cholesterollowering or blood pressure reductionwould be regarded as disease claimsunder the first suggested criterion, andwhite cell counts and fever would bedisease claims under the second. Thiscomment also suggested that FDAdevelop a list of disease markers andsymptoms that fall under each of theproposed criteria.

FDA agrees in part and disagrees inpart with this comment. The agencyagrees that references in dietarysupplement labeling to physiologicmarkers or symptoms of a disease thatare quantifiably linked to that disease inan official government health agencysummary statement or consensus reportwould be appropriately treated asimplied disease claims. Indeed, in thecases described, elevated blood pressure(hypertension) and elevated cholesterol(hypercholesterolemia) are diseasesthemselves, with subsequent events(heart attack, stroke) the lateconsequences of those diseases.Although FDA agrees that fever andelevated white cell counts are almostalways evidence of a disease, FDA doesnot agree that the second criterionappropriately describes the remainingcircumstances in which references tosigns or symptoms should be treated asdisease claims. The appropriate test iswhether: (1) The condition to be treatedor prevented is a disease and (2) thesigns and symptoms referred to in thelabeling, in context, are characteristic ofa disease and thus permit the inferencethat the product is intended to treat orprevent the disease. The secondcriterion offered by the comment doesnot provide information on either ofthese elements.

(43.) Some comments that objected tothe proposed definition of diseaseargued that the inclusion of ‘‘signs orsymptoms’’ as part of the definition ofdisease should not mean that a referenceto the signs and symptoms of a disease

in dietary supplement labelingconstitutes a disease claim. Anothercomment argued that because signs andsymptoms do not appear in thedefinition of ‘‘drug,’’ FDA is notauthorized to treat a reference tocharacteristic signs and symptoms as adrug claim.

The health claims definition of‘‘disease or health-related condition’’ in§ 101.14(a)(5), which is being adopted asthe definition of ‘‘disease’’ in thisregulation, does not include reference tothe signs and symptoms of disease.Nonetheless, dietary supplementlabeling that refers to the characteristicsigns or symptoms of a specific diseaseor class of diseases will still beconsidered to have made an implieddisease claim. Labeling that claims aproduct ‘‘prevents bone fragility in post-menopausal women,’’ clearly impliesthat the product prevents osteoporosis.Similarly, labeling that claims a product‘‘prevents shortness of breath, anenlarged heart, inability to exercise,generalized weakness, and edema’’ hasmade a congestive heart failure claim.

The basis for determining whethersuch a reference to signs or symptomsconstitutes an implied disease claim isnot whether the definition of diseaseincludes mention of signs or symptoms.Rather, FDA looks at whether theobjective evidence shows that theproduct is ‘‘intended for use in thediagnosis, cure, mitigation, treatment, orprevention of disease’’ within themeaning of section 201(g)(1)(B) of theact and § 201.128, or the claimconstitutes a health claim within themeaning of section 403(r)(1)(B) of theact and § 101.14(a)(1). For example,§ 201.128 provides that the objectiveintent of those responsible for thelabeling of drugs ‘‘is determined by suchpersons’ expressions or may be shownby the circumstances surrounding thedistribution of the article.’’ Section101.14(a)(1) provides that ‘‘[i]mpliedhealth claims include those statements,symbols, vignettes, or other forms ofcommunication that suggest, within thecontext in which they are presented,that a relationship exists between thepresence or level of a substance in thefood and a disease or health-relatedcondition.’’ Both of these provisionspermit FDA to look at whether areference to the characteristic signs orsymptoms of a disease constitute animplied disease claim.

(44.) Many comments argued that thedistinction between claims that aproduct maintains healthy function andthat it prevents or treats abnormalfunction is artificial, and that consumersunderstand both types of claims asdisease treatment or prevention claims.

Comments from dietary supplementmanufacturers and some consumergroups argued that both types of claimsshould be permitted either because theyare not implied disease claims orbecause implied disease claims arepermissible. Conversely, most of thecomments from health professionalgroups, groups devoted to specificdiseases, pharmaceutical companies,and other consumer groups argued thatneither type of claim should bepermitted, because permitting implieddisease claims to be made without priorreview would jeopardize the publichealth by encouraging substitution ofunproven remedies for proven ones.One comment argued that analysis ofhealth maintenance claims is nodifferent than analysis of any otherstructure/function claim: They aredisease claims if they imply diseaseprevention or treatment. According tothis comment, health maintenanceclaims are permissible unless they relateto endpoints that are understood to bedisease markers, such as blood pressureand cholesterol. Comments from aformer Commission member and from aconsumer group argued that manyhealth maintenance claims will beperceived as disease treatment orprevention claims, and urged that FDAfollow the Commission’s guidelines,under which the seriousness of thecondition and the ability of theconsumer to evaluate it are key factorsin deciding whether a disease claim hasbeen made. One comment argued thatFDA may not prohibit a claim that adietary supplement ‘‘maintains normalfunction’’ even if it implies a diseaseclaim because 403(r)(6)(A) of the actexpressly authorizes such claims.

One comment said that the proposedrule would frustrate the ‘‘orphan drug’’process. The comment contended that ifdietary supplement labeling may claimto promote or maintain ‘‘healthy’’endpoints that are related to signs andsymptoms of specific diseases, thenincentives to conduct research onorphan drugs would be undermined.The comment explained that dietarysupplements do not require the samefinancial investment as drugs do(because drugs must be approved as safeand effective for their intended uses andmeet quality controls), and couldundercut sales of a more heavilyregulated and more expensive approveddrug. The comment said that a dietarysupplement manufacturer’s ability tomake a disease prevention claim bycharacterizing the product as promotinggood health ‘‘cannot become a license tosell an active ingredient in a product



that is functionally a drug but is labeledas a dietary supplement.’’

FDA has carefully considered thesecomments and has concluded that thedistinction drawn in the proposalbetween maintaining normal functionand treatment or prevention of abnormalfunction is supported by the statute andthe Commission report. FDA does notagree that health maintenance claimsmust always be treated as implieddisease claims. Section 403(r)(6)(A) ofthe act demonstrates that Congressintended to treat as structure/functionclaims some claims concerningmaintenance of normal structure orfunction, because it expressly permitsstatements that ‘‘characterize thedocumented mechanism by which anutrient or dietary ingredient acts tomaintain such structure or function.’’

FDA also believes that many claimsconcerning the maintenance of‘‘normal’’ or ‘‘healthy’’ structure orfunction do not imply diseaseprevention in the context of dietarysupplement labeling, unless otherstatements or pictures in the labelingimply prevention of a specific disease orclass of diseases. There may be cases,however, in which a statement of healthmaintenance can be understood only asa claim of prevention of a specificdisease, in which case it will beconsidered a disease claim. Thus, anyreference to ‘‘maintaining a tumor-freestate’’ would be a disease claim.Similarly, a claim to ‘‘maintain normalbone density in post-menopausalwomen’’ is a disease claim because post-menopausal women characteristicallydevelop osteoporosis, a disease whoseprincipal sign is decreased bone mass.

FDA has added a sentence to§ 101.93(g)(2) clarifying that the criteriain that paragraph are not intended topreclude structure/function claims thatrefer to the maintenance of healthystructure or function, unless they implydisease treatment or prevention.

For the reasons described elsewherein this document, however, FDA doesnot believe that DSHEA permits claimsconcerning treatment or prevention ofabnormal function, where suchabnormal function implies a specificdisease or class of diseases.Accordingly, FDA believes that thestatutory scheme is consistent withtreating many health maintenancestatements as structure/function claims,while treating as health claims or newdrug claims statements that implydisease treatment or prevention byreference to an effect on abnormalstructure or function.

The Commission report also supportsthe distinction drawn by FDA betweenmaintaining healthy function and

preventing or treating abnormalfunction. The report’s Guidance states:

4. Statements that mention a body system,organ, or function affected by the supplementusing terms such as ‘‘stimulate,’’ ‘‘maintain,’’‘‘support,’’ ‘‘regulate,’’ or ‘‘promote’’ can beappropriate when the statements do notsuggest disease prevention or treatment oruse for a serious health condition that isbeyond the ability of the consumer toevaluate.

5. Statements should not be made thatproducts ‘‘restore’’ normal or ‘‘correct’’abnormal function when the abnormalityimplies the presence of disease. An examplemight be a claim to ‘‘restore’’ normal bloodpressure when the abnormality implieshypertension.(Report at pp. 38 and 39.)

FDA agrees that if a healthmaintenance claim implies diseasetreatment or prevention, it would not beacceptable. (In FDA’s view, a claimpromoting ‘‘use for a serious healthcondition that is beyond the ability ofthe consumer to evaluate’’ is simply oneform of implied disease claim.) FDAbelieves that many health maintenanceclaims are acceptable. In some cases, ahealth maintenance claim could useterms that are so closely identified witha specific disease or that so clearly referto a particular at-risk population thatFDA would consider the claim to be animplied disease prevention claim, e.g.,‘‘maintains healthy lungs in smokers’’would imply prevention of tobacco-related lung cancer and chronic lungdisease. ‘‘Maintains healthy lungfunction,’’ alone, however, would be anacceptable structure/function claim.

In response to the commentcontending that dietary supplementsundercut sales of orphan drugs bymaking health promotion claims foractive ingredients already approved asorphan drugs, FDA notes that section201(ff)(3) of the act excludes from thedefinition of ‘‘dietary supplement’’articles that have been approved asdrugs or for which substantial clinicalinvestigations conducted under aninvestigational new drug application(IND) have been made public, beforethey were marketed as dietarysupplements or foods.

(45.) Many comments responded toFDA’s specific request for comment onwhether it is appropriate to treat‘‘maintains healthy cholesterol levels’’as a permissible structure/functionclaim, while treating ‘‘lowerscholesterol’’ as a disease claim. A fewcomments supported the distinctiondrawn in the proposed rule. Many didnot, however. One comment from amajor trade association claimed that thedistinction between lowering andmaintaining cholesterol levels isambiguous, asking ‘‘What is a healthy

cholesterol level, but a lower cholesterollevel?’’ Another comment from a foodindustry group contended that‘‘cholesterol’’ itself is a sign orsymptom, and thus that both types ofclaims refer to a sign or symptom ofdisease. Several comments argued thatlowering cholesterol is inextricablylinked to cardiovascular disease. Somecomments argued that the distinctionbetween maintaining normal cholesteroland lowering cholesterol is arbitrarybecause both have as their purposepreventing heart disease, and consumerslink cholesterol levels with diseaseprevention. Other comments, however,argued that cholesterol claims do notimply disease prevention. A commentfrom an organization devoted toprevention and treatment of heartdisease argued that if any cholesterolclaims were to be permitted, a claim like‘‘promotes cholesterol clearance’’ wouldbe a more accurate structure/functionstatement than ‘‘maintains healthycholesterol’’ and less likely to implydisease prevention. Two commentscontended that changing a claim from‘‘lowers cholesterol’’ to ‘‘maintainshealthy cholesterol levels’’ does notchange the effect of the product or itsuse. Some comments argued that‘‘lowers cholesterol’’ claims should bepermitted for cholesterol levels that arenot ‘‘abnormal’’ or are belowhypercholesterolemia.

FDA does not agree that claimsconcerning maintenance of normalcholesterol levels necessarily constituteimplied disease claims. Although anelevated cholesterol level is a sign ofhypercholesterolemia and an importantrisk factor for heart disease, acholesterol level within the normalrange is not a sign or risk factor fordisease. Moreover, maintainingcholesterol levels within the normalrange is essential to the structure andfunction of the body for reasons otherthan prevention of heart disease.Although many people think ofcholesterol solely in terms of thenegative role of elevated cholesterol inheart disease, normal cholesterol levelsplay a positive role in maintaining ahealthy body. Cholesterol is a necessaryconstituent of cell membranes and ofmyelin, the sheath that coats nerves.Cholesterol is also required for thesynthesis of steroid hormones, whichare essential for life. Finally, cholesterolis required for the production of bile inthe liver, making possible absorption ofdietary fat and fat soluble vitamins.Thus, a claim that a dietary supplementhelps maintain cholesterol levels thatare already within the normal rangedoes not necessarily imply disease



treatment. FDA also believes thatCongress intended to permit dietarysupplements to carry claims of this typeunder section 403(r)(6)(A) of the act.

The agency has concluded, however,that references to ‘‘healthy’’ cholesterolmay be misleading to consumersbecause the phrase ‘‘healthycholesterol’’ is now frequently used torefer to high density lipoproteins (HDL),a specific cholesterol fraction believedto be beneficial. To avoid thisconfusion, FDA has concluded that anappropriate structure/function claim formaintaining cholesterol would be‘‘helps to maintain cholesterol levelsthat are already within the normalrange.’’

FDA continues to believe that ‘‘lowerscholesterol,’’ however qualified, is animplied disease claim. As manycomments argued, lowering cholesterolis inextricably linked in the public mindwith treating elevated cholesterol andpreventing heart disease. The agencyalso believes that ‘‘promotes cholesterolclearance’’ is an implied disease claimbecause it is directed at loweringcholesterol rather than maintaininglevels already determined to be withina normal range. FDA will review allcholesterol claims to determine whetherthe labeling as a whole implies that theproduct is intended to lower elevatedcholesterol levels. In such cases, FDAwould consider the labeling to create animplied disease claim.

(46.) A comment from a formerSurgeon General of the United Statesargued that, given the importance ofpreventing cardiovascular disease,dietary supplements should bepermitted to make claims for cholesterolreduction, because ‘‘our citizens deservethe opportunity to know when safe andeffective dietary supplements areavailable to lower cholesterol.’’ Acomment from the Nutrition Committeeof the American Heart Associationargued that current scientific evidencedoes not support added benefits ofdietary supplementation with nutritivesubstances for prevention ofcardiovascular disease in the generalpopulation, and expressed concern thatdietary supplements also carry risks.

FDA agrees that prevention of heartdisease is an extremely important publichealth goal. Lowering cholesterol withcertain drugs has been conclusivelyshown to be effective in reducingmortality from coronary artery disease.Indeed, the evidence linking thelowering of elevated cholesterol withpreventing heart disease is so strong thatidentifying and using effective therapiesto lower cholesterol in patients withelevated cholesterol levels has becomeof compelling importance. With this in

mind, use of possibly ineffectivetherapies in persons with elevatedcholesterol, which can delay or preventeffective treatment, poses significantpublic health risks. Although DSHEArequires that manufacturers who makestructure/function claims havesubstantiation, manufacturers are notcurrently required to submit thatsubstantiation to FDA for premarketreview, nor does FDA have theresources to inspect and review thequality of the substantiation in mostcases. For this reason, FDA does notbelieve that permitting ‘‘lowerscholesterol’’ claims on dietarysupplements without prior reviewserves the public health.

(47.) A few comments argued thatFDA may not prohibit ‘‘lowerscholesterol’’ claims because the agencyhad earlier issued an advisory letterpermitting such claims if the claimstated that the product was useful in thecontext of a healthy diet. One of thesecomments contended that the agencymay not change its advice or guidancebecause it has cited no studies in thisrulemaking to support the view that‘‘lowers cholesterol’’ implies diseasetreatment.

FDA does not agree that it may notchange its position on whetherparticular cholesterol claims implydisease treatment. The record andanalysis in this rulemaking, as well asFDA’s experience in implementingDSHEA, provide an ample basis for theconclusions that the agency has reachedon cholesterol claims.

G. Conditions Associated With NaturalStates (§ 101.93(g)(2)(iii))

The proposed rule stated that naturalstates such as aging, menopause,pregnancy, and the menstrual cycle, arenot themselves diseases, but can beassociated with abnormal conditionsthat are diseases. FDA proposed in§ 101.93(g)(2)(iii) to treat as a diseaseclaim a statement that a product had aneffect on a condition associated with anatural state if the condition presented‘‘a characteristic set of signs orsymptoms recognizable to health careprofessionals or consumers’’ as an‘‘abnormality.’’ FDA provided asexamples of such abnormal conditionsthe following: Toxemia of pregnancy;premenstrual syndrome; hot flashes;and presbyopia, decreased sexualfunction, and Alzheimer’s diseaseassociated with aging.

In the July 8, 1999, Federal Registernotice announcing a public meeting andreopening the comment period, FDAasked for additional comment on thisprovision of the proposed rule. Theagency sought specific comment on the

following three questions: (1) If FDAwere to treat some conditions associatedwith natural states as diseases (e.g.,toxemia of pregnancy and Alzheimer’sdisease) but not others (e.g., hot flashes,common symptoms associated with themenstrual cycle, and decreased sexualfunction associated with aging), whatwould be an appropriate principle fordistinguishing the two groups? (2) Forexample, would it be appropriate toconsider the severity of the healthconsequences if the condition were togo without effective treatment? (3) If so,how should ‘‘severity’’ be defined?

(48.) Although some comments fromdisease-specific organizations andhealth professionals supported thisprovision, most of the commentsstrongly objected to classifying commonconditions associated with natural statesas diseases. None of the objectingcomments argued that toxemia ofpregnancy or Alzheimer’s disease arenot diseases. Almost all of thesecomments, however, contended thatPMS, hot flashes, and variousconditions associated with aging, suchas decreased sexual function, are socommon that they should be consideredneither abnormal nor diseases. Somecomments argued that any conditionsuffered by more than 50 percent of thepopulation should be considerednormal and not a disease, and gave asan example benign prostatichypertrophy. Other comments citedprevalence rates for conditions such asPMS and hot flashes, and contendedthat the cited rates were too high forthese conditions to be consideredabnormal. A large number of commentsasserted that the proposed rule wouldtreat pregnancy, menopause, and agingas diseases. A few comments arguedthat if menopause, aging, and pregnancyare not diseases, then signs andsymptoms associated with these statescannot be diseases. One commentargued that conditions related to naturalstates are not diseases but ‘‘health-related conditions’’ and that DSHEApermits statements about health-relatedconditions.

In response to the questions in theJuly 8, 1999, Federal Register notice,many comments argued that the severityof the condition associated with anatural state was not an appropriateprinciple for distinguishing diseasesfrom nondiseases. These commentsgenerally argued that the severity of thesymptoms (rather than the severity ofthe consequences of going withouteffective treatment) was not an adequatebasis to distinguish diseases fromnondiseases. One comment from a foodindustry group argued that this was aninappropriate principle because ‘‘all



natural states can have severeconsequences if left unattended.’’ Thiscomment suggested that conditions thatwere ‘‘universal’’ should not be treatedas diseases. This comment and oneother also suggested that thedistinguishing principle was whetherthe cause of the condition was‘‘pathological.’’

FDA has reconsidered proposed§ 101.93(g)(2)(iii), and has concludedthat it is not appropriate, under DSHEA,to treat certain common, nonseriousconditions associated with natural statesas diseases. There are a wide variety ofconditions representing impairedfunction of an organ or system that areassociated with particular stages of lifeor normal physiologic processes. Thesestages and processes includeadolescence, the menstrual cycle,pregnancy, menopause, and aging. (FDAnotes that, contrary to the comments,the proposed rule would not haveclassified these stages or processesthemselves as diseases; it classified onlycertain abnormal conditions associatedwith these stages or processes asdiseases.) The conditions associatedwith these stages or processes can varyfrom common, relatively mildabnormalities, for which medicalattention is not required, to seriousconditions that can cause significant orpermanent harm if not effectivelytreated.

For example, pregnancy is associatedwith common and mild abnormalitiessuch as morning sickness and leg edemathat cause no permanent harm if leftuntreated, as well as with such seriousconditions as hyperemesis gravidarum,toxemia of pregnancy, and acutepsychosis of pregnancy, which can belife-threatening if not effectively treated.The menstrual cycle is commonlyassociated with mild mood changes,edema, and cramping that do not causesignificant or permanent harm if leftuntreated, but also, more rarely, withserious cyclical depression that canresult in significant harm if noteffectively treated. Aging is almostinvariably associated with characteristicskin and scalp changes, such aswrinkles and hair loss, which do notneed medical attention. It is also,however, associated with seriousdiseases that will result in significant,often irreversible damage, many ofwhich can be effectively treated. Thesediseases include osteoporosis,glaucoma, and arteriosclerotic diseasesof coronary, cerebral, and peripheralvessels. Adolescence is commonlyassociated with mild acne, which doesnot cause significant or permanent harmif not treated, and, rarely, with cysticacne, which can produce severe

physical and psychological scars if noteffectively treated.

Whether all of these conditionsrepresent diseases is, in part, a matter ofdefinition and, in part, depends on theconsequences of the conditions if noteffectively treated, and on howcommonly they occur, i.e., whether theymay be considered ‘‘normal.’’ Althoughmost people consider the more seriousor infrequent conditions referred toabove to be diseases, views vary withrespect to the common, milderconditions. FDA has reconsidered theposition it took in the proposed rule andagrees with the comments that treatingas diseases the common, mildsymptoms associated with normal lifestages or processes would not beconsistent with the intent of DSHEA.

FDA does not believe that thefrequency with which a conditionassociated with a natural state occurs is,by itself, sufficient to distinguishdiseases from nondiseases. The severityof the consequences of disease, as wellas the consequences of ineffectivetreatment, must also be considered. Asnoted above, whether common, minorconditions associated with natural statesare diseases is a matter of debate, butFDA has decided not to treat them asdiseases because the agency believesthis approach is consistent with theintent of DSHEA. FDA does not,however, believe that DSHEA wasintended to permit unreviewed claimsabout serious conditions that couldcause significant or permanent harm,particularly where effective treatment isavailable. FDA also does not agree that‘‘all natural states can have severeconsequences if left unattended.’’ FDAhas listed a large number of conditionsassociated with natural states thatcommonly do not have seriousconsequences even if not effectivelytreated. FDA also does not agree that itis helpful in this context to distinguishbetween diseases and nondiseases byasking which have a ‘‘pathological’’basis. The term ‘‘pathological’’ is itselfdefined by reference to disease, namely,‘‘caused by or involving disease;morbid’’ (Ref. 7).

Accordingly, for purposes of this rule,mild conditions commonly associatedwith particular stages of life or normalphysiological processes will not beconsidered diseases. Therefore,§ 101.93(g)(2)(iii) now states that astatement will be considered a diseaseclaim if it claims that the product ‘‘hasan effect on an abnormal conditionassociated with a natural state orprocess, if the abnormal condition isuncommon or can cause significant orpermanent harm.’’ Ordinarily, FDAwould follow the suggestion in the

comments that conditions associatedwith a stage of life or a normalphysiological process be consideredcommon if they occur in more than one-half of those experiencing that stage orprocess.

The following are examples ofconditions about which structure/function claims could be made under§ 101.93(g)(2)(iii): (1) Morning sicknessassociated with pregnancy; (2) legedema associated with pregnancy; (3)mild mood changes, cramps, and edemaassociated with the menstrual cycle; (4)hot flashes; (5) wrinkles; (6) other signsof aging on the skin, e.g., liver spots,spider veins; (7) presbyopia (inability tochange focus from near to far and viceversa) associated with aging; (8) mildmemory problems associated withaging; (9) hair loss associated withaging; and (10) noncystic acne. Thefollowing are examples of conditionsthat would remain disease claims: (1)Toxemia of pregnancy; (2) hyperemesisgravidarum; (3) acute psychosis ofpregnancy; (4) osteoporosis; (5)Alzheimer’s disease, and other seniledementias; (6) glaucoma; (7)arteriosclerotic diseases of coronary,cerebral or peripheral blood vessels; (8)cystic acne; and (9) severe depressionassociated with the menstrual cycle.

FDA has not included benignprostatic hypertrophy (BPH) on either ofthese lists, because the agency does notbelieve that BPH should be considereda consequence of aging. Like many otherdiseases, e.g., diabetes, prostate cancer,and heart disease, the incidence of BPHis much higher among older men. Thisdoes not mean that BPH or prostatecancer is caused by the aging process.Even if BPH were considered a directconsequence of aging, however, claimsto treat or prevent it would still betreated as disease claims because failureto obtain effective treatment can causesignificant or permanent harm.

FDA notes that it does not base theexclusion of the mild commonconditions associated with natural statesfrom § 101.93(g)(2)(iii) on the argumentadvanced by one of the comments thatthese are ‘‘health-related conditions’’and that DSHEA permits structure/function claims about health-relatedconditions. FDA believes that a ‘‘health-related condition’’ is a state of healthleading to disease. As FDA has saidpreviously, ‘‘diseases’’ and ‘‘health-related conditions’’ are ‘‘so closelyrelated that no bright-line distinction ispracticable’’ (58 FR 2478, 2481 January6, 1993). There is nothing in DSHEA, itslegislative history, or in the definition of‘‘disease or health-related condition’’that would suggest that commonconditions associated with natural states



are ‘‘health-related conditions’’ withinthe meaning of section 403(r)(1)(B) ofthe act. Further, FDA does not agree thatsection 403(r)(6) of the act authorizesstructure/function claims about ‘‘health-related conditions.’’ Had Congressintended to authorize structure/functionclaims about ‘‘health-relatedconditions’’ it could easily have usedthat terminology, but did not.

(49.) Some comments concernedspecific claims under proposed§ 101.93(g)(2)(iii). One comment soughtconcurrence that the following areacceptable structure/function claims:‘‘supports a normal, healthy attitudeduring PMS’’ and ‘‘supportive formenopausal women.’’ Another commentargued that a statement that a productprovides nutrients that diminish thenormal symptomatology ofpremenstrual syndrome or menopauseis a permissible structure/functionclaim. Another comment asked whether‘‘helps to maintain normal urine flow inmen over 50 years old’’ is a permissiblestructure/function claim. One commenturged that only products proven safewhen used as directed should bepermitted for sale for enlarged prostateand that such products shouldrecommend that a man see hisphysician. Another comment arguedthat the claim ‘‘for men over 50 yearsold,’’ which FDA had proposed as anacceptable structure/function claim, isvague and ambiguous and is of no useto consumers.

FDA agrees that ‘‘supports a normal,healthy attitude during PMS’’ and‘‘supportive for menopausal women’’are appropriate structure/functionclaims. ‘‘Supports a normal, healthyattitude during PMS’’ is acceptablebecause PMS is generally a common,mild condition associated with a normalphysiologic process. ‘‘Supportive formenopausal women’’ is acceptablebecause it is a general statement thatdoes not refer to symptoms of anyconditions at all. Claims aboutdiminishing the normalsymptomatology of premenstrualsyndrome or menopause would also beacceptable structure/function claims, ifthey did not suggest, for example,prevention or treatment of osteoporosis,or another disease associated with thesestates. ‘‘Helps to maintain normal urineflow in men over 50 years old,’’however, is an implied disease claimbecause, as many comments pointedout, the average or ‘‘normal’’ state inmen over 50 years old is diminishingurine flow, in most cases due to BPH,so that the apparent ‘‘maintenance’’really represents a claim ofimprovement (treatment).

H. Generally (§ 101.93(g)(2)(iv))

Under proposed § 101.93(g)(2)(iv),FDA stated that a statement would beconsidered a disease claim if it claimedexplicitly or implicitly to have an effecton disease through one or more of thefollowing factors: (1) The name of theproduct (e.g., ‘‘Carpaltum’’ (carpaltunnel syndrome), ‘‘Raynaudin’’(Raynaud’s phenomenon), ‘‘Hepatacure’’(liver problems)). Names that did notimply an effect on a disease, such as‘‘Cardiohealth’’ and ‘‘Heart Tabs,’’would not constitute disease claims; (2)statements about the formulation of theproduct, including a claim that theproduct contained an ingredient thathas been regulated by FDApredominantly as a drug and is wellknown to consumers for its use inpreventing or treating a disease (e.g.,aspirin, digoxin, or laetrile); (3) citationof a publication or other reference, if thecitation refers to a disease use. Forexample, labeling for a vitamin Eproduct that included a citation to anarticle entitled ‘‘Serial CoronaryAngiographic Evidence ThatAntioxidant Vitamin Intake ReducesProgression of Coronary ArteryAtherosclerosis,’’ would create a diseaseclaim under this criterion; (4) use of theterm ‘‘disease’’ or ‘‘diseased;’’ or (5)otherwise suggesting an effect ondisease by use of pictures, vignettes,symbols, or other means (e.g.,electrocardiogram tracings, pictures oforgans that suggest prevention ortreatment of a disease state, or theprescription symbol (Rx)). The proposedrule stated that a picture of a bodywould not constitute a disease claimunder this criterion.

(50.) A few comments stated that thephrase ‘‘has an effect on’’ in proposed§ 101.93(g)(2)(iv) is vague and could beinterpreted by the agency to meanalmost anything. Some of thesecomments argued that disease claimsshould include only those that use thespecific terms ‘‘diagnose,’’ ‘‘prevent,’’‘‘treat,’’ ‘‘mitigate,’’ or ‘‘cure.’’

FDA does not agree that the phrase‘‘has an effect on’’ is inappropriatelyvague. FDA believes that it is necessaryto use a phrase that encompassessynonyms for the terms ‘‘diagnose,’’‘‘prevent,’’ ‘‘treat,’’ ‘‘mitigate,’’ or‘‘cure.’’ If disease claims were limited tothose that used the specific terms in thestatute, it would be possible to makeobvious and explicit disease claimssimply by using terms that are similarin meaning to the statutory terms, e.g.,‘‘relieves arthritis pain’’ rather than‘‘treats arthritis pain,’’ or ‘‘eliminatesthe risk of cancer’’ rather than ‘‘preventscancer.’’

I. Product Name (§ 101.93(g)(2)(iv)(A))(51.) One comment observed that

there is an inconsistency between thestatement in the proposed rule that‘‘Heart Tabs’’ does not imply an effecton a disease and § 101.14(a)(1), whichstates that:

Health claim means any claim made on thelabel or in the labeling of a food, includinga dietary supplement, that expressly or byimplication, including ‘‘third party’’references, written statements (e.g., a brandname including a term such as ‘‘heart’’),symbols (e.g., a heart symbol) characterizesthe relationship of any substance to a diseaseor health-related condition * * *and requested clarification.

FDA agrees, in part, and disagrees, inpart, with the comment. FDA does notagree that § 101.93(g)(2)(iv)(A) and§ 101.14(a)(1) are inconsistent. Section101.14(a)(1) was issued in 1993 toimplement the health claims provisionsof NLEA. In § 101.14(a)(1), use of theterm ‘‘heart’’ in a brand name and useof the heart symbol in labeling areoffered as examples of health claims, ifin the context of the labeling as a whole,the word or symbol suggests that thereis a relationship between the productand a disease or health-relatedcondition. Thus, according to thepreamble to that final rule (58 FR 2478at 2486), the heart symbol mightappropriately appear in the labeling ofa food product if, in context, it did notsuggest a relationship to heart disease,e.g, in conjunction with ‘‘Hey, FudgeLovers.’’ If, however, the heart symbolappeared alone on a food, withoutfurther explanation from context,consumers might conclude that the foodwas beneficial for reducing the risk ofdeveloping cardiovascular disease (id.).

Following the issuance of§ 101.14(a)(1), Congress enactedDSHEA. DSHEA created a specialregulatory regime for dietarysupplements. That regime, while closelyrelated to the regime for food, was notidentical to the food regime. Section403(r)(6) of the act specifies certaintypes of structure/function claims andgeneral well-being claims that may bemade for dietary supplements withoutfirst obtaining new drug approval orhealth claim authorization. The types ofclaims listed in section 403(r)(6) of theact are similar, but not identical to theclaims permitted for foods under section201(g)(1)(C) of the act. Under Nutrilab v.Schweiker, 713 F.2d 335 (7th Cir. 1983),conventional food claims are limited tostructure/function effects that derivefrom the taste, aroma, or nutritive valueof the food. Dietary supplement claimsare not subject to that limitation. HadCongress intended the scope of thepermitted claims to be identical, it



could simply have declared that dietarysupplements are ‘‘foods.’’ In light ofCongress’ intent to expand the types ofclaims authorized for dietarysupplements in DSHEA, FDA interprets§ 101.14(a)(1) as permitting dietarysupplements to have brand names thatinclude the word ‘‘heart’’ or otherorgans, if, in the context of the labelingas a whole, the name does not implydisease treatment or prevention.

FDA does agree, however, that under.§ 101.14(a)(1), a dietary supplementname that included the word ‘‘heart’’could be a health claim, depending onthe context. Thus, a dietary supplementcould be called ‘‘HeartTabs’’ if its claimwas ‘‘to maintain healthy circulation,’’or some other role related to thestructure or function of the heart thatdid not imply treatment or prevention ofdisease. If, however, the product namewas not qualified by any further claimin the labeling, the product could beconsidered, under § 101.14(a)(1), to beintended for treatment or prevention ofcardiovascular disease.

FDA also believes that the heartsymbol has become so widely associatedwith prevention of heart disease that itsuse in the labeling of a dietarysupplement would be ordinarilyconsidered an implied heart diseaseprevention claim. Consistent with theexamples provided in the January 6,1993, Federal Register document onhealth claims (58 FR 2486), however,there may be unusual cases in which, incontext, the use of a heart symbol doesnot imply heart disease prevention.

(52.) Several comments agreed withproposed § 101.93(g)(2)(iv)(A) thatproduct names that imply an effect ondisease, including implying cure ortreatment of a disease, should not beallowed. The comments, however,requested that the agency providefurther guidance as to what types ofproduct names are acceptable and whattypes are not. Some commentsquestioned whether product names suchas ‘‘CarpalHealth,’’ ‘‘HepatoHealth,’’‘‘HepataCare,’’ ‘‘CircuCure,’’ or‘‘Soothing Sleep’’ would be acceptableunder proposed § 101.93(g)(2)(iv)(A).Other comments disagreed with theagency’s examples and stated that it isdifficult to distinguish the reasoningbehind some of the examples cited. Forexample, a few comments stated thatboth ‘‘Cardiohealth’’ and ‘‘Heart Tabs’’imply that the product prevents heartdisease.

Two principles formed the basis forthe distinctions in the proposed rulebetween product names that wereconsidered structure/function claimsand those that were considered diseaseclaims. First, the name should not

contain the name, or a recognizableportion of the name, of a disease.Second, the name should not use termssuch as ‘‘cure,’’ ‘‘treat,’’ ‘‘correct,’’‘‘prevent’’ or other terms that suggesttreatment or prevention of a disease.Thus, ‘‘CarpalHealth’’ and ‘‘CircuCure’’would be considered disease claims. Insome cases, to determine whether aproduct name implies an effect ondisease, the agency will need toconsider the context in which a term ispresented in the labeling as a whole.Thus, ‘‘Soothing Sleep’’ could beconsidered a claim to treat insomnia,unless the labeling made clear that theproduct was intended only foroccasional sleeplessness. ‘‘HepataCare’’and ‘‘HepataHealth’’ could also beconsidered disease claims because‘‘Hepata’’ could be read as a reference tohepatitis, unless the labeling made clearthat the product was intended forgeneral liver health and not intended totreat or prevent hepatitis.

The agency notes that in the nearfuture, FDA will issue for publiccomment a draft guidance to provideadditional clarification and examples ofclaims that would and would not beconsidered disease claims under thefinal rule. FDA will include in the draftguidance examples of product names.

(53.) Another comment stated thatproposed § 101.93(g)(2)(iv)(A) wouldprohibit the use of the name of the‘‘dispensing institution’’ if it had theword ‘‘Cancer’’ in it because the agencywould interpret the labeling as implyingan effect on disease, when in fact theproduct was listing the institutionwhere the product was dispensed, e.g.,ABC Cancer Institute. Other commentswere concerned that the proposed rulewould prohibit the use of their companytrade name, which includes the use ofthe word ‘‘prescription’’ and itsabbreviation ‘‘Rx.’’

The agency reiterates that it will viewthe name in the context of the entirelabeling to determine whether a diseaseclaim is being made. However, amanufacturer may not circumvent therequirements of the act, DSHEA, or thisfinal rule by using the name of aninstitution or the manufacturer to implya disease claim.

The agency agrees that the use of theword ‘‘prescription’’ or its abbreviation‘‘Rx’’ in the name of the product shouldnot automatically be interpreted as adisease claim. Although these termsimply that the product is a prescriptiondrug, some prescription drugs areintended for nondisease conditions.Therefore, if nothing else in the labelingsuggests a disease use, the agency willnot consider the use of ‘‘prescription’’ or‘‘Rx’’ to be an implied disease claim.

The agency notes, however, that the useof these terms on dietary supplementproducts may deceive consumers intothinking that they are purchasing aprescription drug without aprescription. Thus, use of the terms‘‘prescription’’ or ‘‘Rx’’ is misleadingand will misbrand the product undersection 403(a)(1) of the act if, in thecontext of the labeling as a whole, theterms imply that the product is aprescription drug.

(54.) A few comments cited in aproposed rule published in the FederalRegister of March 27, 1974 (39 FR11298), in which FDA stated that itwould challenge brand names only insituations where clarifying language isincapable of rectifying FDA’s concernwith the brand name and that excisionof a brand name should be a last resortand should be pursued only when allother methods of qualifying the namehave failed.

The agency notes that the proposedrule cited in this comment was neverfinalized and was withdrawn onDecember 30, 1991 (56 FR 67440), aspart of an FDA initiative to reduce thebacklog of outstanding proposed rulesthat have never been finalized. Thepolicies outlined in the March 27, 1974,Federal Register notice are not in effect.

(55.) Several comments sought astatement from FDA that if a productbrand name becomes synonymous overtime with use for prevention ortreatment of a disease, it will still bepermitted. As an example, thecomments claimed that Kleenex hasbecome synonymous with treatment ofnasal congestion, but did not providesupport for this assertion.

FDA does not believe that Kleenex issynonymous with treatment of nasalcongestion and, absent any supportivedata, has no reason to believe thatconsumers believe them to besynonymous. The agency would agreethat Kleenex has become synonymouswith ‘‘tissue,’’ and that both are used inconjunction with nasal congestion.Neither tissue nor Kleenex, however,treat, prevent, or otherwise affect nasalcongestion in any way. Because theagency was not presented with anyspecific examples of, nor is it aware ofany, names of products that are notintended to treat disease but that havebecome synonymous with diseasetreatment or prevention, it does nothave reason to believe that there is a realbasis for concern.

J. Product Formulation(§ 101.93(g)(2)(iv)(B))

(56.) Several comments questionedwhether the inclusion of a dietaryingredient in the ingredient list of a



dietary supplement would beinterpreted as a disease claim underproposed § 101.93(g)(2)(iv)(B). Theyargued that to provide truthful labeling,this information must be included.Another comment stated that theproposal fails to distinguish betweentrue claims and false claims. Severalcomments further argued that ingredientinformation may be of value toconsumers to alert them to potentialadverse effects or drug interactions. Onecomment urged that the presence of aconstituent that is naturally occurring ina plant and is also regulated as a drugdoes not automatically classify thesubstance as a drug. The commentasserted that 45 percent of drugs arederived from plants, which, accordingto the comment, would classify anumber of dietary ingredients as drugs.

Listing a dietary ingredient in theingredient list of a dietary supplementwill not be considered to imply an effecton disease unless the ingredient is onethat has been regulated primarily byFDA as a drug and is well-known toconsumers for its use or claimed use inpreventing or treating a disease. (In theproposed rule, the agency gave asexamples aspirin, digoxin, and laetrile.)Very few dietary ingredients meet thistest. The agency agrees that a certainpercentage of drug products are derivedfrom plants. However, only a handful ofthese drugs are well-known toconsumers under the name of the plantor natural plant ingredient from whichthey were derived. Instead, they areknown to consumers under a brandname or generic name, e.g., aspirin.Thus, FDA does not believe that listingdietary ingredients that happen to berelated to well-known drugs will fallunder this provision, except in unusualcircumstances. In those cases where amanufacturer does add a drugingredient that is well-known to treat orprevent disease to its product and labelits presence, however, FDA mayconsider it a disease claim. The fact thatthe labeling is truthful does notnecessarily mean that it falls within thescope of claims authorized by section403(r)(6) of the act. For example, theagency believes that there are manydietary ingredients that could be shownto treat or prevent diseases, and forwhich it could thus be truthful to statethat the product treats or prevents aspecific disease. Under the act,however, if a manufacturer wants tolabel its product to treat or preventdisease, it must do so under the drugapproval provisions or the health claimprovisions of the act. It may not do sounder section 403(r)(6) of the act. Indrafting section 403(r)(6) of the act to

exclude disease claims, Congress madea judgment that the public health willbe served by requiring premarket reviewof such claims.

FDA agrees that it is important toinform consumers about potentialadverse effects or drug interactions forspecific dietary supplement ingredients.In fact, dietary supplement labeling, likethe labeling of other FDA-regulatedproducts, is required to include all factsthat are material in light ofconsequences that may result from useof the product or representations madeabout it (sections 403(a)(1) and 201(n) ofthe act). This provision is not intendedin any way to preclude truthful adverseevent or drug interaction informationfrom appearing in a dietarysupplement’s labeling.

(57.) A dietary supplementmanufacturer asked FDA to clarify theeffect of § 101.93(4)(ii) on a dietaryingredient found in common food(s),whose biological activity is firstcharacterized in a food context, butwhich is subsequently approved as adrug. The comment asked whether, ifindole-3-carbinol, a compounddiscovered in broccoli and othervegetables, were to be approved as abreast cancer drug, claims to the effectthat a vegetable-based dietarysupplement product contains indole-3-carbinol would be permitted asstructure/function claims under theproposed rule. The comment claimedthat the proposed rule would classifysuch claims as disease claims even if thebiological activity of this dietaryingredient were first identified in thefood context.

Where an ingredient has beenapproved as a drug, section 201(ff)(3) ofthe act prohibits marketing of theingredient as a dietary supplementunless the ingredient itself waspreviously marketed as a food(including a dietary supplement), orunless a food containing the ingredientwas previously marketed for thepresence of the ingredient. In theexample provided in the comment, theisolated ingredient indole-3-carbinolcould not be marketed as a dietarysupplement, unless a food containingthe ingredient had been marketed forthe presence of the ingredient before thedrug was approved or was the subject ofsubstantial investigations that had beenmade public. However, to avoid aconflict between this provision andsection 201(ff)(3) of the act in a situationwhere the ingredient was marketed as afood first, FDA has revised§ 101.93(g)(2)(iv)(B) to exclude claimsabout an ingredient that is an articleincluded in the definition of ‘‘dietary

supplement’’ under section 201(ff)(3) ofthe act.

(58.) One comment misunderstood§ 101.93(g)(2)(iv)(B) and believed thatthis provision only applies to the listingof OTC drug ingredients recognized byconsumers.

This provision is not limited to thelisting of OTC drug ingredients. Forpurposes of § 101.93(g)(2)(iv)(B), theagency may consider as a disease claima claim that the product contains aningredient that has been regulated byFDA as a drug, whether marketed over-the-counter or by prescription, and thatis well known for its use in preventingor treating a disease.

K. Citation of Publication Titles(§ 101.93(g)(2)(iv)(C))

(59.) Many comments objected to thisproposed criterion or soughtclarification. Many comments said thatthe proposed criterion underminesDSHEA by prohibiting the use of mostjournals, is not required by DSHEA, oris contrary to section 403B of the act (21U.S.C. 343–2), which, the commentsaid, exempts scientific publicationsfrom labeling rules and is intended toallow consumers to be more informedby reading scientific studies. Othercomments said that Congress intendedto encourage the dissemination ofscientific research and truthful, non-misleading information, so FDA shouldnot prohibit titles of scientific studies.Some comments stated that the issueshould not be whether a publication’stitle refers to a disease use, but ratherwhether, on balance, the entirepresentation, including the productlabel, package insert, and other labeling,represents a disease claim. Thesecomments supported the use ofcomplete citations to scientificliterature, including the titles ofscientific articles. Some commentssuggested that the proposal contradictedearlier FDA positions. One commentreferred to the September-October 1998issue of FDA Consumer which, thecomment stated, suggested thatconsumers contact companies to obtainscientific articles that the companymight have to substantiate a claim.Another comment said the proposal wascontrary to FDA policy to recognize andaccept valid science. Several commentsquestioned how to providesubstantiation of labeling claims, incompliance with 403(r)(6)(B) of the act,if the supporting articles cannot becited. One comment stated that therewill be more fraud and deception in themarketplace because companies will notcite scientific support for theirstatements. Several comments statedthat the proposed rule will restrict



access by consumers and the medicalcommunity to important new researchresults and discourage companies frominvesting in research. A dietarysupplement manufacturer suggestedrevising the provision to permitcompanies to cite ‘‘bonafide’’ textbooksand peer-reviewed scientific journalsthat mention a disease in the title.Another dietary supplementmanufacturer suggested revising thisprovision to permit citation of apublication or reference if the citation‘‘is necessary to present a balanceddiscussion of the documentedmechanism by which a nutrient ordietary ingredient acts to maintain thestructure or function of the body.’’

FDA agrees that in enacting DSHEA,Congress intended to encourage thedissemination of scientific research andtruthful, non-misleading information.FDA also agrees that consumers canbenefit from reviewing the scientificsupport used to substantiate a statementmade for a dietary supplement undersection 403(r)(6) of the act. In keepingwith these goals, FDA has modified§ 101.93(g)(2)(iv)(C) to narrow thecircumstances under which citation to ascientific reference will be considered adisease claim. Based on Congress’explicit prohibition in section 403(r)(6)of the act of claims to affect disease,however, FDA does not believe thatCongress intended to permit scientificreferences to be used in a way thatconstitutes an implied disease claim.Consequently, § 101.93(g)(2)(iv)(C) hasbeen revised to state that citation of atitle referring to a disease will be treatedas a disease claim, if, in the context ofthe labeling as a whole, the citationimplies treatment or prevention of adisease, e.g., through placement on theimmediate product label or packaging,inappropriate prominence, or lack ofrelationship to the product’s expressclaims.

The agency continues to believe thatplacing a citation to a scientificreference that mentions a disease in thetitle on the immediate product label orpackaging should be considered adisease claim for that product, becauseof the unusual and unnecessaryprominence of such placement. Forcitations to scientific references thatrefer to a disease use in the title and thatare included in other types of labeling(i.e., other than the product label orpackaging) the agency will consider thecontext in which the citation ispresented. FDA agrees with thecomments that the totality of allavailable labeling should be consideredto determine the context. One elementthat the agency will look at is theprominence of the citation in the

labeling. If, for example, the citation issimply listed in the bibliography sectionof the labeling among other titles, it willgenerally not suggest an implied diseaseclaim. On the other hand, highlighting,bolding, using large type size, orprominent placement of a citation thatrefers to a disease use in the title couldsuggest that the product has an effect ondisease. The agency will also considerwhether the cited article provideslegitimate support for a 403(r)(6) of theact statement that appears in thelabeling of the dietary supplement.Enhancing the bibliography withcitations to scientific references thatrefer to a disease in the title and thathave no reasonable relation to thestatement made will be considered adisease claim. Similarly, the agency willconsider whether citations are to bonafide research.

FDA also agrees that it is important toprovide a balanced discussion of thescientific literature regarding the claim.FDA encourages manufacturers to citereferences that provide a balanceddiscussion of the evidence supporting astructure/function claim.

The agency believes that the final rulestrikes a reasonable balance betweenencouraging the dietary supplementindustry to inform consumers about thesubstantiation for their claims andpreventing abuses of section 403(r)(6) ofthe act.

(60.) Several comments challengedthe basis for the proposed restriction ofscientific references. One comment fromindustry said the proposed restrictionon titles is outside DSHEA because theact refers to statements. The commentsaid titles could be prohibited if theywere misleading, but said the ruleshould not contain a blanketprohibition.

The comment is apparently referringto section 403(r)(6) of the act, whichprescribes the terms under which a‘‘statement’’ may be made for a dietarysupplement. FDA believes that thecomment’s reading is too literal,however. A ‘‘statement’’ does not haveto be a declaratory sentence but ratheris fairly read to include other kinds ofstatements, such as citations ofscientific authority. In keeping withDSHEA’s purpose to broaden the scopeof labeling claims that may be made fordietary supplements without subjectingthem to regulation as drugs, FDAbelieves that Congress intended‘‘statement’’ to refer to any claim madethat recommends or suggests aparticular use of a dietary supplement.In addition to being under inclusive, anarrower interpretation would notbenefit the dietary supplement industrybecause it would limit the scope of

claims authorized under section403(r)(6) of the act.

(61.) A few comments stated that theagency did not provide any support forthe assumption that citations are diseaseclaims rather than substantiation for aclaim.

FDA believes that a citation of a titlethat refers to a specific disease can serveboth as a disease claim and assubstantiation for a claim. A citation ofa publication title that links the productto a particular disease could leadconsumers to believe that the productcan be used to diagnose, prevent,mitigate, treat, or cure a disease, even ifthe title also provides substantiation forthe product claims.

As stated above, citation of a scientificreference will not be treated as a diseaseclaim if, in the context of the labelingas a whole, the reference lacksprominence and if it is appropriatesupport for the product claim.

(62.) One comment soughtclarification of the effect of thisprovision on multi-ingredient products.The comment asked whether a diseaseclaim for the entire product would becreated if the labeling cited an articleabout only one ingredient of a multi-ingredient product.

Generally, if a citation is presented inthe product labeling in such a way as toimply that a specific ingredient can treator prevent disease, the product, as awhole, will be considered to beintended to treat or prevent disease.

(63.) A few comments requested FDAto clarify how proposed§ 101.93(g)(2)(iv)(C) would operate. Thecomments questioned whether theywould have to delete a citation from alist or redact the reference to a diseasefrom the title of the article. Onecomment asked whether an article thatcontains a reference to a disease can becited if the title is not used in thecitation. The comments furtherquestioned whether they can providethe entire article, with the title on it, ifrequested by a consumer. Somecomments asked FDA to clarify that alabel may cite a title that appears in apublication whose name includes adisease (such as the publication titledCancer) or to clarify how scientificstudies may be cited. One commentrequested that the agency issue furtherguidance to clarify what is and is notcovered by § 101.93(g)(2)(iv)(C).

FDA does not expect a manufacturerto redact portions of the citation ordelete a citation from a list of referencesor bibliography if it is appropriate toinclude the reference to substantiate aclaim. As described above, if thecitation to a scientific reference refers toa disease, the agency will consider the



context in which the citation ispresented, including its prominence inthe labeling and whether there is areasonable relationship between thereference and the express claim. In mostcases, the unredacted reference title canbe included in the product labelingwithout subjecting the product toregulation as a drug, as long as theprominence of the reference does notsuggest that it is being used to implydisease treatment or prevention. Underrevised § 101.93(g)(2)(iv)(C), the onlyreason a publication title would beconsidered a disease claim regardless ofprominence would be if the reference isnot reasonably related to substantiatingthe product’s express claim. In that case,FDA believes that the reference wouldbe a disease claim, even if the name ofthe disease is redacted, because the onlypurpose of including the referencewould be to suggest use of the productfor treatment or prevention of thedisease discussed in the reference.

With regard to citation of titles fromjournals whose official names includethe name of a disease, the sameconsiderations of appropriateprominence and reasonable relationshipto the product’s express claims apply.FDA expects that accepted conventionsof scientific citation will be used for allcitations that appear in labeling.

Finally, if specific information aboutan unlabeled use of a product isrequested by a consumer, and therequest is not solicited by themanufacturer, providing articles that areresponsive to the request will not beconsidered a disease claim.

FDA will issue further guidance on§ 101.93(g)(2)(iv)(C), if necessary.

(64.) Several comments soughtmodifications to proposed§ 101.93(g)(2)(iv)(C). One commentsuggested revising the provision topermit companies to cite articles orreferences that use ‘‘intermediate terms’’(which the comment said were terms orphrases that have disease-relatedendpoints) on the label or labeling.

Whether a citation that refers to adisease-related endpoint will beconsidered a disease claim under therule will depend on the context inwhich the disease-related endpoint isreferred to and whether the referenceimplies that the product has an effect ondisease. For example, the title of anarticle that states that a product wasshown to maintain cholesterol levelsthat were already within the normalrange, with no reference to a disease,would be considered a structure/function statement about maintenancerather than a disease claim. However, ifthe title of the article states that theproduct was shown to lower elevated

cholesterol levels, this implies that theproduct can be used to have an effect onthe disease states hypercholesterolemiaand heart disease, because heart diseaseis associated with high cholesterollevels.

(65.) A trade association suggestedthat the title should not be consideredto be a disease claim unless it uses theterms ‘‘treat,’’ ‘‘cure,’’ ‘‘mitigate,’’‘‘prevent,’’ or ‘‘diagnose.’’

As stated elsewhere in this document,FDA believes that a disease claim can bemade explicitly or implicitly usingterms other than those listed in thecomment. For example, depending onhow it was used in a product’s labeling,a scientific reference entitled ‘‘UsingIngredient X For Diabetes’’ couldconstitute a claim that the product candiagnose, mitigate, treat, cure, orprevent diabetes, without using any ofthese specific terms.

(66.) A few comments argued thatcitation of articles that refer to a diseaseuse should be permitted becauseconsumers have access to these articlesin connection with the sale of dietarysupplements under section 403B(a) ofthe act.

As stated above, FDA has revised theproposed rule’s treatment of citations toscientific articles. Under the final rule,such citations will not always beconsidered disease claims. FDA doesnot agree, however, that section 403B ofthe act applies to the citation of titles inproduct labeling. Although section 403Bof the act exempts certain publicationsfrom the labeling provisions of the act,section 403B(a)(2) states that theexemption applies only when, amongother requirements, the publication is‘‘used in connection with the sale of adietary supplement to consumers whenit * * * does not promote a particularmanufacturer or brand of a dietarysupplement.’’ If the reference or the titleof the reference was disseminated by aparticular manufacturer of the dietarysupplement discussed in the reference,the agency would conclude that it wasbeing used to promote thatmanufacturer’s brand of the dietarysupplement. Therefore, the exemptionin section 403B of the act would notapply.

Furthermore, to qualify for theexemption in section 403B of the act, apublication must be ‘‘an article, achapter in a book, or an official abstract* * * reprinted in its entirety’’ and mustbe ‘‘displayed or presented, or * * *displayed or presented with other suchitems on the same subject matter, so asto present a balanced view of theavailable scientific information of adietary supplement.’’ A citation to an

article alone could not meet theserequirements.

L. Use of Disease or Diseased(§ 101.93(g)(2)(iv)(D))

(67.) Many comments agreed withproposed § 101.93(g)(2)(iv)(D), statingthat the terms ‘‘disease’’ or ‘‘diseased’’should classify a statement as a diseaseclaim. Several comments urged that astatement referring in a general way tothe concept of ‘‘health promotion anddisease prevention’’ not cause thestatement to be considered a diseaseclaim, as long as no specific disease wasmentioned. One comment asked that theagency permit general discussions of theconcept of disease prevention, citing thefollowing example from the U.S. PublicHealth Service Healthy People 2000initiative: ‘‘Better dietary and exercisepatterns can contribute significantly toreducing conditions like heart disease,stroke, diabetes, and cancer, and couldprevent 300,000 deaths.’’

FDA agrees that general statementsabout health promotion and diseaseprevention may be acceptable, as long asthe statements do not imply that aspecific product can diagnose, mitigate,cure, treat or prevent disease.Accordingly, FDA has revised§ 101.93(g)(2)(iv)(D) to permit generalstatements about disease prevention thatdo not refer explicitly or implicitly to aspecific disease or class of diseases or tothe specific product or ingredient. Forexample, the statement ‘‘a good dietpromotes good health and prevents theonset of disease’’ would not beconsidered a disease claim. On the otherhand, the claim ‘‘Promotes good healthand prevents the onset of disease’’would refer implicitly to the productand would constitute a diseaseprevention claim. FDA also believes thatthe particular statement offered by oneof the commenters would constitute adisease claim. The example cites fourspecific diseases. If that statement wereincluded in the labeling for a dietarysupplement, a consumer wouldreasonably assume that the statementapplies to the product and that takingthat dietary supplement contributes topreventing the diseases listed. If,however, the statement said ‘‘betterdietary and exercise patterns cancontribute to disease prevention andbetter health,’’ FDA would not considerit a disease claim.

M. Pictures, Vignettes, and Symbols(§ 101.93(g)(2)(iv)(E))

(68.) Many comments agreed thatcertain pictures, vignettes, and symbolscan explicitly or implicitly convey thatthe product has an effect on disease. Afew comments agreed that a diseased



organ should be considered a diseaseclaim. They argued, however, that apicture of a healthy heart, healthyartery, or other healthy organ should bepermitted because such pictures do notin and of themselves depict a disease. Afew comments stated that a healthyelectrocardiogram (EKG) tracing shouldnot be considered a disease claim. Onecomment requested that the agencyclarify whether a picture of an organ ispermitted if the claims are appropriateand within the scope of permittedstructure/function claims. The commentoffered as an example a statement thata product maintains cardiovascularhealth accompanied by a picture of aheart and circulatory system.

FDA agrees that in most cases, apicture of a healthy organ would not beconsidered a disease claim, if, in thecontext of the labeling as a whole, it didnot imply treatment or prevention ofdisease. As described in response tocomment 51 of section II.I of thisdocument, however, there may besymbols for organs, like the heartsymbol, that have become so widelyrecognized as symbols for diseasetreatment or prevention, their use inlabeling would constitute an implieddisease claim. FDA also believes that apicture of a healthy EKG tracing is animplied disease claim. Because mostconsumers cannot distinguish a healthyEKG tracing from an unhealthy one,both types may be viewed as referencesto diagnosis or treatment of unhealthyheart conditions.

N. Membership in Product Class(§ 101.93(g)(2)(v))

Some product class names are sostrongly associated with use to treat orprevent a specific disease or class ofdiseases that claiming membership inthe product class implies diseasetreatment or prevention. Underproposed § 101.93(g)(2)(v), a statementwould have been considered a diseaseclaim if it claimed that the productbelonged in a class of productsrecognizable to health care professionalsor consumers as intended for use todiagnose, mitigate, treat, cure, orprevent a disease. The preambleprovided the following examples ofclass names that would imply diseasetreatment or prevention: Claims that theproduct was an ‘‘antibiotic,’’ a‘‘laxative,’’ an ‘‘analgesic,’’ an‘‘antiviral,’’ a ‘‘diuretic,’’ an‘‘antimicrobial,’’ an ‘‘antiseptic,’’ an‘‘antidepressant,’’ or a ‘‘vaccine.’’ Theseexamples were not intended toconstitute an exclusive list of productclass names that convey disease claims.Under the proposed rule, claiming thata product was in a class that is not

recognizable to health care professionalsor consumers as intended for use todiagnose, mitigate, treat, cure or preventdisease would not have constituted adisease claim under this criterion. Thepreamble provided as examples ofacceptable structure/function claims:Claims that the product was an‘‘energizer,’’ a ‘‘rejuvenative,’’ a‘‘revitalizer,’’ or an ‘‘adaptogen.’’ In lightof the agency’s decision that claims forrelief of ‘‘occasional constipation’’should not be considered diseaseclaims, the term ‘‘laxative’’ will not beconsidered a disease claim under thefinal rule, as long as the remainder ofthe labeling makes clear that theproduct is not intended to treat chronicconstipation.

(69.) Most of the comments onproposed § 101.93(g)(2)(v) weregenerally supportive, but some wantedto ensure that the provision would beapplied in specific ways. One commenturged that ‘‘appetite suppressant’’ betreated as a disease claim, while anothercomment urged that ‘‘tonic’’ be treatedas a structure/function claim.

FDA does not agree that ‘‘appetitesuppressant’’ should be considered adisease claim. As discussed elsewherein this document, although obesity is adisease, overweight is not. An appetitesuppressant may be intended forordinary weight loss, rather than as atreatment for obesity. Therefore,‘‘appetite suppressant’’ would only beconsidered a disease claim in a contextwhere it implies use for obesity. FDAagrees that ‘‘tonic’’ is not a diseaseclaim. ‘‘Tonic’’ is commonly understoodas a general term for anything thatrefreshes, and, by itself, would not beconsidered to constitute a disease claim.

(70.) Some comments stated thatvarious class names should be allowedwhen they describe the mechanism bywhich a supplement has its effect, orwhen they are present in a product andit is truthful and not misleading to namethem. One comment offered as examplesof class names that might be used todescribe a product’s mechanism ofaction: A statement that a product thatis soothing to the stomach achieves itseffects as a result of its ‘‘carminative(antispasmodic) properties’’ or as aresult of its ‘‘anti-inflammatory effect onthe gastrointestinal tract.’’ Thiscomment stated that it is notmembership in a given class ofcompounds that should make a producta drug, but rather the intended use ofthe product. One comment askedwhether this criterion precludes astatement that daily consumption ofvitamins and minerals may prevent theonset of disease or other physicalailments.

Nothing in this provision wouldpreclude a manufacturer from truthfullydeclaring the ingredients contained in aproduct. In fact, FDA regulations requirethe ingredients in a dietary supplementto be listed on its label. (See§ 101.4(a)(1) and (g) (21 CFR 101.4(a)(1)and (g)), and § 101.36). The rationale for§ 101.93(g)(2)(v) is that certain productclass names (not particular ingredients)are so strongly associated with use todiagnose, treat, mitigate, cure, orprevent disease that claimingmembership in the class wouldconstitute a disease claim. FDA does notbelieve that claiming membership in aproduct class is necessary in order toprovide an accurate list of theingredients present in a product.

FDA agrees that dietary supplementsmay carry statements that characterize‘‘the documented mechanism of actionby which a nutrient or dietaryingredient acts to maintain * * *structure or function,’’ but only to theextent that such a statement does ‘‘notclaim to diagnose, mitigate, treat, cure,or prevent a specific disease or class ofdiseases’’ (section 403(r)(6) of the act).In the examples provided in thecomment, FDA is unaware of evidenceestablishing that the claims actuallydescribe ‘‘documented’’ mechanisms bywhich the products ‘‘maintain’’ a calmstomach. Nevertheless, assuming thatthese statements met the otherrequirements of section 403(r)(6)(A) ofthe act, FDA would not consider theterm ‘‘antispasmodic’’ to constitute adisease claim because the agency doesnot believe that it is closely associatedwith treatment or prevention ofgastrointestinal disease. The term ‘‘anti-inflammatory’’ is, however, stronglyassociated with treatment of certainserious gastrointestinal diseases, andwould constitute a disease claim.

FDA agrees with the statement that itis not membership in a given class ofcompounds that makes a product adrug, but rather the intended use of theproduct. This criterion sets forth FDA’sconclusion that claiming membership incertain product classes that are stronglyassociated with use to treat or preventdisease is evidence that the product isintended to treat or prevent disease.

Although this provision does notitself treat as a disease claim a statementby a vitamin manufacturer that theproduct prevents the onset of a disease,such a statement would be considered adisease claim under § 101.93(g)(2)(I),which covers statements that a producthas an effect on a specific disease orclass of diseases. In addition, a generalstatement that a product prevents theonset of disease would be considered adisease claim under



§ 101.93(g)(2)(iv)(D), as noted in thediscussion of that provision. Claimingmembership in the class of vitamins orminerals would not constitute a diseaseclaim under this criterion.

(71.) A food manufacturers’ tradeassociation and an individualmanufacturer opposed the provision,arguing that it goes beyond the intent ofDSHEA and would prohibit the use ofany term associated with a drugproduct.

FDA does not agree that this provisiongoes beyond the intent of DSHEA northat it would prohibit the use of anyterm associated with a drug product.DSHEA precludes statements undersection 403(r)(6) of the act fromclaiming to treat or prevent disease.This provision constitutes FDA’sconclusion that some drug class names(but not all terms associated with drugproducts) are so strongly associatedwith disease prevention or treatmentthat claiming membership in the classconstitutes a claim that the product, likeother members of the class, treats orprevents disease.

(72.) One pharmaceutical companyargued that proposed § 101.93(g)(2)(v)would violate DSHEA, because DSHEAspecifically defines as a dietarysupplement an article that is approvedas a new drug under section 505 of theact, if it was, prior to approval,marketed as a dietary supplement.

FDA agrees that the dietarysupplement definition includes theprovision cited by the comment (section201(ff)(3)(A) of the act), but believes thatthe definition and § 101.93(g)(2)(v) arenot inconsistent. Section 101.93(g)(2)(v)would treat as a disease claim a labelingstatement that the supplement is amember of a product class when thatclass is so recognizable for its diseasetreatment or prevention use that thelabeling statement would be understoodas a disease claim for the supplement.The criterion would not treat inclusionof an ingredient in a dietary supplementas a disease claim merely because theingredient had been approved undersection 505 of the act nor would itpreclude listing the ingredient in theSupplement Facts panel or ingredientlist.

O. Substitute for Disease Therapy(§ 101.93(g)(2)(vi))

Under proposed § 101.93(g)(2)(vi), astatement would have been considereda disease claim if it explicitly orimplicitly claimed that the product wasa substitute for another product that isa therapy for a disease. FDA offered‘‘Herbal Prozac’’ as an example of sucha claim. A claim that did not identify aspecific drug, drug action, or therapy

(e.g., ‘‘use as part of your weight lossplan’’) would not constitute a diseaseclaim under this criterion.

(73.) There was general support forthe provision, particularly forconsidering terms that make a directconnection with an approved drug, like‘‘Herbal Prozac’’ and ‘‘Herbal Phen-fen,’’disease claims. Several organizationsnoted that associating dietarysupplements with regulated drugproducts is deceptive and dangerousbecause it can signal to consumers thatbecause the product is ‘‘herbal’’ it issafer. Several medical associations,however, objected to the interpretationthat ‘‘use as part of your weight lossplan,’’ is nonspecific and would beacceptable. They maintained that theterm implies treatment of a disease,obesity. A comment from amanufacturer also strongly objected tothe statement in the proposal that ‘‘Useas part of your weight loss plan’’ wouldbe an acceptable structure/functionclaim. The comment contended that thelegislative history of the act shows thatCongress intended weight loss claims tobe treated as disease claims. Finally, thecomment argued that even if FDAdecides to permit weight loss claims asstructure/function claims, the legislativehistory of the act and case law requirethat FDA classify products containing‘‘antinutrients’’ as drugs.

FDA agrees with these comments thatobesity is a disease, and that obesityclaims are not acceptable structure/function claims. Being overweight, i.e.,being more than one’s ideal weight butless than obese, however, is not adisease. FDA believes that it iscommonly understood that ‘‘weight lossplans’’ relate to a broad range ofoverweight statuses. Therefore, weightloss plans are not so narrowlyassociated with disease treatment that areference to use as part of a weight lossplan should be considered a diseaseclaim.

FDA does not agree that either thelegislative history of the act or the caselaw interpreting section 201(g) of the actor DSHEA require a determination thatFDA classify as drugs products makingweight loss claims. The legislativehistory of section 201(g)(1)(C) of the actshows that Congress added thestructure/function definition of ‘‘drug’’in part to capture obesity claims thatwere not covered by section 201(g)(1)(B)because obesity was not, at that time,considered a disease. FDA believes thatthe legislative history in fact supportsFDA’s view that weight loss claims areproperly considered structure/functionclaims. Although obesity claims are nowcovered by section 201(g)(1)(B) of theact because obesity is now considered a

disease, section 201(g)(1)(C) was addedto cover conditions, like overweight,that are not considered diseases, butthat affect the structure or function ofthe body. Structure/function claimsunder section 403(r)(6) of the act areclosely related to structure/functionclaims under section 201(g)(1)(C) of theact and therefore should encompassweight loss claims.

FDA also does not agree that casescited by the comment compel theconclusion that weight loss productsmust be regulated as drugs. In Nutrilabv. Schweiker, 713 F.2d 335 (7th Cir.1983), American Health Products Co. v.Hayes, 574 F. Supp. 1498 (S.D.N.Y.1982), aff’d, 744 F.2d 912 (2d Cir. 1984),and United States of America v.Undetermined Quantities Of ‘‘CAL–BAN3000’’, 776 F. Supp. 249 (E.D.N.C.1991), the courts held that certainweight loss products were drugs undersection 201(g)(1)(C) of the act becausethey were labeled to affect the structureor function of the body, and did notqualify for the ‘‘food’’ exception tosection 201(g)(1)(C). At the time thesecases were decided, the only issue waswhether these products were ‘‘foods’’ or‘‘drugs.’’ Since then, however, DSHEAcreated a new statutory category ofproducts, dietary supplements. Section403(r)(6) of the act, which was added byDSHEA, permits structure/functionclaims to be made for dietarysupplements without subjecting them toregulation as drugs, even if they couldnot qualify for the ‘‘food’’ exception insection 201(g)(1)(C) of the act.Therefore, these cases do not establishthat dietary supplements making weightloss claims must be regulated as drugs.To the contrary, because the productswere held to be drugs under section201(g)(1)(C) of the act rather thansection 201(g)(1)(B), these cases supporttreatment of weight loss claims fordietary supplements as structure/function claims authorized undersection 403(r)(6) of the act.

Finally, FDA does not agree that,under United States v. Ten Cartons,More or Less, of an Article * * * Ener-B Vitamin B–12, 72 F.3d 285 (2d Cir.1995), dietary supplements makingweight loss claims must necessarily beregulated as drugs. The court in Ener-Bheld that a dietary supplement thatmakes a structure/function claim maynevertheless be regulated as a drug,under certain circumstances. In thatcase, the court found that FDA couldregulate a product as a drug, based onits method of intake (nasaladministration). Nothing in that casesuggests that FDA must regulate dietarysupplements making weight loss claimsas drugs.



(74.) Several comments reiterated thatgeneral statements about the nature of aproduct or its mechanism of actionshould not be disease claims, or shouldbe structure/function claims as long asthey are truthful and not misleading.One comment objected to the provisionas duplicative of proposed§ 101.93(g)(2)(v). Another commentsought to delete the provision, arguingthat dietary supplement manufacturershave the right to communicate toconsumers that their products havefewer side effects than drugs.

FDA does not believe that thisprovision precludes general statementsabout the function or mechanism ofaction of a dietary supplement. It is notnecessary to claim that the product is asubstitute for a drug or therapy todescribe its function or its mechanismof action. Nor is § 101.93(g)(2)(vi)duplicative of § 101.93(g)(2)(v).Claiming that a product is a substitutefor a specific drug or therapy, e.g.,‘‘Herbal Prozac,’’ is a different means ofcommunicating that a dietarysupplement is intended to treat adisease than claiming that the productbelongs to a class of drugs associatedwith treatment or prevention of thatdisease, e.g., ‘‘antidepressant.’’

FDA does not agree that section403(r)(6) of the act permits a dietarysupplement manufacturer to claim thatits product has fewer side effects thana drug, if the drug is intended to treator prevent disease, because the clearimplication is that the dietarysupplement is intended for treatment orprevention of the same disease. If,however, the drug is not intended totreat or prevent disease, a dietarysupplement manufacturer is free tomake truthful, non-misleadingcomparisons between the drug and thedietary supplement.

P. Augmentation of Therapy or Drug forDisease (§ 101.93(g)(2)(vii)))

Under proposed § 101.93(g)(2)(vii), astatement would have been considereda disease claim if it explicitly orimplicitly claimed that the productaugmented a particular therapy or drugaction. The preamble offered thefollowing example of a disease claimunder this criterion: ‘‘Use as part of yourdiet when taking insulin to helpmaintain a healthy blood sugar level.’’ Aclaim that did not identify a specificdrug, drug action, or therapy would notconstitute a disease claim under thiscriterion. The preamble gave thefollowing example of an acceptablestructure/function claim: ‘‘use as a partof your weight loss plan.’’

(75.) Several comments supported thisprovision. A few comments requested

that FDA withdraw the provision,arguing that dietary supplements areoften useful in providing nutritionalsupport to complement drug therapy ormedical treatment and that the agencyshould encourage such information tobe communicated to consumers. Onecomment stated that as long as thestatement makes it clear that theproduct is being recommended for itsnutritional impact on structure orfunction ‘‘as part of the therapy and notas the therapy itself,’’ FDA shouldpermit the statement. According to thecomment, ‘‘use as part of your dietwhen taking insulin to help maintain ahealthy blood sugar level’’ should beacceptable because the product is beingrecommended for its nutritional impacton structure or function as part of thetherapy and not as the therapy itself.Another comment asked whetherremoving the words ‘‘when takinginsulin’’ from the statement would makeit an acceptable structure/functionclaim.

The agency agrees that dietarysupplements may be useful in providingnutritional support. Associating such astatement with an express or impliedclaim that the dietary supplementaugments a therapy or drug action,however, implies that the dietarysupplement has a role in treating orpreventing the disease for which thedrug or other therapy is used.

The agency does not agree that theproposed claim involving insulin is anacceptable structure/function claim.Persons who take insulin have a disease,namely, diabetes. By referring to the useof the dietary supplement inconjunction with and for the samepurpose (‘‘to maintain a healthy bloodsugar level’’) as a drug (insulin), whichis used to for a disease (diabetes), thestatement implies that the dietarysupplement will help treat diabetes.

A general statement that a dietarysupplement provides nutritionalsupport would be an acceptablestructure/function claim, provided thatthe statement does not suggest that thesupplement is intended to augment orhave the same purpose as a specificdrug, drug action, or therapy for adisease. In the example, if the statementwere changed to ‘‘use as part of yourdiet to help maintain a healthy bloodsugar level,’’ the claim would beconsidered acceptable. Deleting thereference to the drug, insulin, wouldremove the implication that the dietarysupplement is used to augment theinsulin to treat, mitigate, prevent, orcure diabetes.

On its own initiative, FDA ismodifying § 101.93(g)(2)(vii) to limit itsapplicability to claims for augmentation

of drugs or therapies that are intendedto diagnose, mitigate, treat, cure, orprevent disease.

(76.) Another comment noted that theagency did not address the use ofsynonyms for ‘‘augment,’’ such as‘‘strengthen,’’ ‘‘reduce,’’ ‘‘improve,’’‘‘modify,’’ ‘‘inhibit,’’ ‘‘protect,’’ or‘‘defend.’’

Use of these terms may be appropriatein some contexts, i.e., when thestatements do not suggest diseaseprevention or treatment use. If,however, the use of these terms impliesthat the dietary supplement augments aparticular therapy or drug action orotherwise suggests an effect on disease,the agency will consider the statementa disease claim.

(77.) A trade association maintainedthat under the proposal, bread, crackers,and other baked goods used inconjunction with prescription drugsand/or other therapy would not beconsidered a food, but a drug, undercertain circumstances.

Section 101.93 is intended to provideregulatory criteria for statements madefor dietary supplements. Under section201(ff)(2)(B) of the act, a dietarysupplement does not include a productrepresented for use as a conventionalfood or as a sole item of a meal or thediet. If statements made for breads,crackers, and other baked goodscharacterize the relationship between asubstance in the food and a disease orhealth-related condition, they mustcomply with the health claimsprovisions for foods under section403(r)(1)(B) and (r)(3) through (r)(4) ofthe act.

Q. Role in Body’s Response to Diseaseor Disease Vector (§ 101.93(g)(2)(viii))

Under proposed § 101.93(g)(2)(viii), astatement would have been considereda disease claim if it explicitly orimplicitly claimed a role in the body’sresponse to a disease or to a vector ofdisease. The preamble to the proposaldefined a vector of disease as anorganism or object that is able totransport or transmit to humans anagent, such as a virus or bacterium, thatis capable of causing disease in man.The preamble offered as examples ofdisease claims under this criterionclaims that a product ‘‘supports thebody’s antiviral capabilities’’ or‘‘supports the body’s ability to resistinfection.’’ A more general reference toan effect on a body system that did notimply prevention or treatment of adisease state would not have constituteda disease claim under this criterion.FDA provided as an example of anacceptable structure/function claim



under this criterion ‘‘supports theimmune system.’’

(78.) Two comments from healthassociations supported this provision.One comment from a manufacturerargued that it should be deleted becausea number of nutrients and dietarysupplements ‘‘have a role in the body’sresponse to disease.’’ One commentargued that the body has naturaldefenses to disease, that these arenormal functions of the body, and thattherefore, statements such as ‘‘enhancesdisease resistance’’ should be allowableas structure/function claims. Commentsfrom a consumer organization and amember of the President’s Commissionon Dietary Supplement Labels assertedthat the provision made too manyclaims allowable. These commentsstated that as long as a claim includesa disease-fighting function of the body,e.g., ‘‘supports the immune system,’’ itshould be considered a disease claim,regardless of other functions that mightbe involved.

FDA agrees that nutrients and dietarysupplements may play a role in thebody’s response to disease. This doesnot mean, however, that diseaseprevention claims are acceptablestructure/function claims. The actrequires dietary supplementmanufacturers who wish to makedisease prevention claims to do so byobtaining authorization for a healthclaim or by obtaining new drugapproval. Although FDA agrees thatclaims that a product fights disease, orenhances disease-fighting functions ofthe body, are disease claims, FDA doesnot agree that claims such as ‘‘supportsthe immune system’’ are specificenough to imply prevention of disease.

(79.) Several comments argued thatthere was no significant differencebetween ‘‘supports the immune system’’(identified as a structure/function claimin the proposal) and ‘‘supports thebody’s antiviral capabilities’’ (identifiedas a disease claim in the proposal). Oneview was that both should beconsidered structure/function claims.Conversely, other comments contendedthat ‘‘supports the immune system’’ is adisease claim, because it could beinterpreted as a claim for treatment orprevention of human immunodeficiencyvirus (HIV) disease. Another commentrecommended that ‘‘supports the body’santiviral capabilities’’ be allowable as astructure/function claim, stating that thebroader ‘‘supports the immune system’’statement was vague and useless toconsumers because the immune systemhas many functions.

The distinction between the twoclaims is one of specificity. An intactimmune system has several functions.

In addition to their role in the defenseagainst pathogens, certain componentsof the immune system, namely whiteblood cells, have other importantfunctions. For example, white bloodcells play an essential role in thephagocytosis and disposal of aging redblood cells or otherwise damaged cells.A statement of support for the immunesystem, by itself, conveys no specificreference to disease treatment orprevention. The claim that vitamin A isnecessary to maintaining a healthyimmune response does not imply that aspecific disease or class of diseases willbe prevented. In contrast, a claim that aproduct ‘‘supports the body’s antiviralcapabilities’’ represents a claim oftreatment or prevention of a specificclass of diseases, those caused byviruses (e.g., colds, hepatitis, or HIVinfection).

R. Treatment/Prevention of AdverseEvents (§ 101.93(g)(2)(ix))

Under proposed § 101.93(g)(2)(ix), astatement would have been considereda disease claim if it explicitly orimplicitly claimed to treat, prevent, ormitigate adverse events associated witha therapy for a disease (e.g., ‘‘reducesnausea associated with chemotherapy,’’‘‘helps avoid diarrhea associated withantibiotic use,’’ and ‘‘to aid patientswith reduced or compromised immunefunction, such as patients undergoingchemotherapy’’). A claim that did notmention a therapy for disease (e.g.,‘‘helps maintain healthy intestinalflora’’) would not have constituted adisease claim under this criterion.

(80.) Comments from two large healthorganizations supported this provision,while two large business organizationsand several other comments criticizedit. Those opposing the provision arguedthat the proposal incorrectly categorizedadverse reactions as diseases. Opposingcomments also contended that dietarysupplements may be useful as anadjunct to therapy by counterbalancingthe effects of a drug in depleting anutrient or interfering with themetabolism of a nutrient, and that thisshould be considered a structure/function role.

FDA believes that some of thesecomments may have misconstrued theprovision. The criterion is not intendedto capture every adverse event claim,but only claims about adverse eventsthat satisfy the definition of disease. Inthe proposed rule, this limitation wasconveyed by the phrase ‘‘andmanifested by a characteristic set ofsigns or symptoms.’’ Because the finalrule uses a different definition ofdisease, § 101.93(g)(2)(ix) has beenrevised to state that claims about

adverse events are disease claims only‘‘if the adverse events constitutediseases.’’ FDA believes that a claimthat a product is useful because itcounterbalances the effects of a drug indepleting a nutrient or interfering withthe metabolism of a nutrient would beacceptable as a structure/functionstatement. Such a claim would notsuggest treatment of an adverse reactionthat meets the definition of disease.However, as discussed above, if theclaim expressly or impliedly suggeststhat the supplement is intended toaugment a specific drug, drug action, ortherapy for a disease, or serve the samepurpose as a specific drug or therapy fora disease, then the statement may beconsidered a disease claim.

(81.) A dietary supplementmanufacturer requested that FDA clarifywhy a statement that refers to a drug butnot a disease, such as ‘‘helpsindividuals using antibiotics tomaintain normal intestinal flora’’ is adisease claim, but a general statement,such as ‘‘helps maintain intestinal flora’’is a permissible structure/functionclaim.

Although the statement ‘‘helpsindividuals using antibiotics tomaintain normal intestinal flora’’ doesnot explicitly refer to a disease, there isan implicit claim that use of the dietarysupplement while taking antibiotics willprevent or mitigate a disease. Personsusing certain antibiotics are at risk ofdeveloping overgrowth in the gut of apathogenic organism because along withfighting the target organisms in the bodythe antibiotic can suppress normalintestinal flora that are used to preventinfection in the intestinal tract. A firmthat markets its product to address thisconcern, with claims that the productcan be used to maintain normalintestinal flora while taking antibiotics,is making an implied disease preventionclaim. Conversely, the statement ‘‘helpsmaintain intestinal flora’’ alone, withoutany reference to a disease, drug, drugaction, or therapy, does not imply aneffect on disease and would beconsidered a structure/function claimabout general health maintenance.

S. Otherwise Affects Disease(§ 101.93(g)(2)(x))

Under proposed § 101.93(g)(2)(x), astatement would have been considereda disease claim if it suggested an effecton a disease or class of diseases in amanner other than those specificallyenumerated in the first nine criteria.

(82.) A food manufacturers’ tradeassociation commented that thisprovision is of no regulatoryimportance, whereas a dietarysupplement trade association and



several other comments considered it anover-reaching ‘‘catch-all’’ provision thatwould allow FDA to treat any claim asa disease claim. These commentsprovided examples of a number ofclaims that they believed would bedisease claims under this provision, e.g.‘‘provides nutritional support forwomen during premenstruation bypromoting proper fluid balances andbreast health,’’ and ‘‘ginger supports thecardiovascular system by inhibitingleukotriene and thromboxane synthesis,substances associated with plateletaggregation.’’

FDA believes that this provision isnecessary to allow for implied diseaseclaims that may not fit into the nineenumerated criteria. The nine criteriaare examples, and not an exhaustive list,of types of claims that the agencybelieves would constitute diseaseclaims, based on past experience. Ratherthan attempting to evaluate or categorizestatements that have not yet beenpresented to FDA, § 101.93(g)(2)(x)recognizes the possibility that othertypes of statements may also implydisease treatment or prevention. FDAdoes not believe that the provision willcause the agency to classify anystructure/function statement as adisease claim. To regulate a statement asa disease claim under this provision, theagency would have to show that thestatement implied an effect on disease.The two examples quoted in thecomments do not appear to the agencyto constitute disease claims.

T. Specific Claims Not Mentioned in theProposed Rule

(83.) One comment contended that adietary supplement called ‘‘pain free’’ or‘‘pain product,’’ that is labeled ‘‘tosupport and maintain joints,’’ shouldnot be regulated as an internal analgesicdrug product under the OTC drugreview because it is intended tomaintain or support ‘‘normal well-beingand pain levels.’’ According to thiscomment, however, products sold as‘‘pain relief’’ or ‘‘otherwise indicated torelieve temporary occurrences ofarthritis pain’’ could be regulated asdrug products under the OTC review,because the tentative final monographfor internal analgesics requires that suchproducts be labeled for the ‘‘temporaryrelief of minor aches and pains’’ (53 FR46204). At the same time, this commentargued that pain, in and of itself, is nota disease and therefore that pain claimsshould not be regulated as diseaseclaims unless accompanied by anexplicit reference to a specific disease.

FDA agrees in part and disagrees inpart with this comment. FDA agrees thatsome minor pain relief claims may be

appropriate structure/function claimsfor dietary supplements. A claim that aproduct is intended to treat minor pain,without reference of any otherconditions, symptoms, or parts of thebody that would imply diseasetreatment or prevention, would be anappropriate structure/function claims,because minor pain, by itself, can becaused by a variety of conditions, not allof them disease-related.

FDA does not agree, however thatgeneral well-being or healthmaintenance claims would encompasssuch pain claims. Pain is not a normalstate, nor are there ‘‘normal painlevels.’’ The claim is thus clearly one ofpain treatment or prevention. FDA alsodoes not agree that section 403(r)(6) ofthe act authorizes a product whosename promises freedom from or relief ofpain (‘‘pain-free’’ or ‘‘pain product’’)and whose labeling includes claimsrelated to maintenance or support ofjoints. While the latter claims alone areappropriate structure/functionstatements, in conjunction with a namethat includes the term ‘‘pain,’’ theproduct is clearly making a claimrelated to treatment or prevention ofjoint pain. As explained elsewhere inthis document, joint pain is acharacteristic symptom of arthritis, andjoint pain claims are therefore diseaseclaims. Acceptable structure/functionclaims could be made, however, for painassociated with nondisease states, e.g.,muscle pain following exercise.

(84.) One comment listed severalclaims and sought concurrence that theywere acceptable structure/functionclaims: ‘‘Boosts stamina, helps increasemuscle size, and helps enhance muscletone’’; ‘‘deters bacteria from adhering tothe wall of the bladder and urinarytract’’; and ‘‘dietary support during thecold and flu season.’’ Another commentasked whether ‘‘promotes general well-being during the cold and flu season’’ isa permissible claim.

FDA agrees that ‘‘boosts stamina,helps increase muscle size, and helpsenhance muscle tone’’ are acceptablestructure/function claims, because theydo not refer to any disease. However,the agency notes that a claim to increasemuscle size implies an effect that maysubject the product regulation as ananabolic steroid under the ControlledSubstances Act (see 21 U.S.C. 802(41)).‘‘Deters bacteria from adhering to thewall of the bladder and urinary tract’’ isnot an acceptable structure/functionclaim because it implies prevention ofbacterial infections of the bladder andurinary tract. The claims ‘‘dietarysupport during the cold and flu season’’and ‘‘promotes general well-beingduring the cold and flu season’’ are

disease claims because they imply thatthe product will prevent colds and fluor will mitigate the symptoms of thosediseases.

(85.) One comment asked that theFDA clarify that dietary supplementscan bear ‘‘smoking-alternative’’ claims ifthey avoid references to nicotine,nicotine withdrawal symptoms, andtobacco-related disease. The commentsought concurrence that the followingtypes of claims were permitted:‘‘Smoking alternative,’’ ‘‘temporarilyreduces your desire to smoke,’’ ‘‘to beused as a dietary adjunct in conjunctionwith your smoking cessation plan;’’ and‘‘mimics the oral sensations of cigarettesmoke.’’

FDA agrees that certain smokingalternative claims may be acceptablestructure/function claims, if they do notimply treatment of nicotine addiction,relief of nicotine withdrawal symptoms,or prevention or mitigation of tobacco-related illnesses. ‘‘Smoking alternative,’’‘‘temporarily reduces your desire tosmoke’’ and ‘‘mimics the oral sensationsof cigarette smoke’’ may be acceptable(for products that otherwise meet thedefinition of a dietary supplement), ifthe context does not imply treatment ofnicotine addiction, e.g., by suggestingthat the product can be used in smokingcessation, or prevention or mitigation oftobacco-related diseases. For example,such claims would not be disease claimsif the context made clear that they werefor short-term use in situations wheresmoke is prohibited or sociallyunacceptable. ‘‘To be used as a dietaryadjunct in conjunction with yoursmoking cessation plan,’’ however, is adisease claim because it is a claim thatthe product aids in smoking cessation,thereby implying that the product isuseful in treating nicotine addiction. Asnoted earlier, a claim that the productis useful in counterbalancing the effectsof a drug in depleting a nutrient orinterfering with the metabolism of anutrient would be acceptable as astructure/function statement.

(86.) One comment offered asacceptable structure/function claims along list of OTC drug claims providedfor in the monographs for antacids,antiflatulents (antigas), antiemetics,nighttime sleep-aids, stimulants(alertness aids), daytime sedatives,aphrodisiacs, products for relief ofsymptoms of benign prostatichypertrophy, anticholinergics (productsthat, at low doses, depress salivary andbronchial secretions), and products forcertain uses. Two comments soughtclarification that inclusion of a claim inan OTC monograph does not precludeits use as a structure/function claim.



FDA agrees that some of the claims onthe comment’s list of OTC drug claimsmay be acceptable structure/functionclaims, but believes that others on thelist are disease claims. Of the claimslisted in the comment from the‘‘Antacids’’ monograph, ‘‘relief of sourstomach’’ and ‘‘upset stomach’’ areacceptable structure/function claims,because they refer to a nonspecificgroup of conditions that have a varietyof causes, many of which are notdisease-related. Thus, they are notcharacteristic of a specific disease orclass of diseases. Although ‘‘relief ofheartburn’’ and ‘‘relief of acidindigestion’’ without furtherqualification are not appropriatestructure/function claims, the agencyhas concluded that ‘‘occasionalheartburn’’ and ‘‘occasional acidindigestion’’ can also be considerednonspecific symptoms, arising as theydo in overindulgence and other sporadicsituations. These claims could beappropriate structure/function claims.In contrast, ‘‘recurrent’’ or ‘‘persistent’’heartburn and acid indigestion can behallmarks of significant illness, and aretherefore disease claims.

All of the claims listed in thecomment from the ‘‘Antiflatulents’’(antigas) monograph are acceptablestructure/function claims, because thesymptoms in the claims are notsufficiently characteristic of specificdiseases: ‘‘Alleviates the symptomsreferred to as gas,’’ ‘‘alleviates bloating,’’‘‘alleviates pressure,’’ ‘‘alleviatesfullness,’’ and ‘‘alleviates stuffedfeeling.’’ The claim listed in thecomment from the ‘‘Antiemetics’’monograph, ‘‘for the prevention andtreatment of the nausea, vomiting, ordizziness associated with motion,’’ isalso a permitted structure/functionclaim.

Of the claims listed in the commentfrom the ‘‘Nighttime’’ sleep-aidsmonograph, ‘‘for the relief of occasionalsleeplessness’’ is an acceptablestructure/function claim, becauseoccasional sleeplessness is not acharacteristic symptom of a disease.‘‘Helps you fall asleep if you havedifficulty falling asleep,’’ and ‘‘helps toreduce difficulty falling asleep’’ aredisease claims because, unless thecontext makes clear that the product isonly for occasional sleeplessness, theyimply treatment of insomnia, a disease.The claim listed in the comment fromthe ‘‘Stimulants’’ (alertness aids)monograph, ‘‘helps restore mentalalertness or wakefulness whenexperiencing fatigue or drowsiness,’’ isan acceptable structure/function claimbecause occasional fatigue anddrowsiness are not characteristic

symptoms of a specific disease or classof diseases. FDA notes, however, thatchronic fatigue or daytime drowsinesscan be symptoms of chronic fatiguesyndrome and narcolepsy, respectively.Products labeled ‘‘to help restore mentalalertness or wakefulness whenexperiencing fatigue or drowsiness’’should not imply treatment of either ofthese diseases.

Of the claims listed in the commentfrom the ‘‘Daytime’’ sedativesmonograph, almost all are acceptablestructure/function claims. ‘‘Occasionalsimple nervous tension,’’ ‘‘nervousnessdue to common every day overwork andfatigue,’’ ‘‘a relaxed feeling,’’ ‘‘calmingdown and relaxing,’’ ‘‘gently sootheaway the tension,’’ ‘‘calmative,’’‘‘resolving that irritability that ruinsyour day,’’ ‘‘helps you relax,’’‘‘restlessness,’’ ‘‘nervous irritability,’’and ‘‘when you’re under occasionalstress, helps you work relaxed’’ are allacceptable structure/function claims,because all suggest occasional ratherthan long-term or chronic moodchanges. Although occasional or acutesymptoms can be characteristic ofdiseases in other settings, none of theoccasional symptoms referred to here ischaracteristic of a specific disease.‘‘Nervous tension headache’’ is a diseaseclaim because tension headache meetsthe definition of a disease.

Of the claims listed in the commentfrom the ‘‘Aphrodisiacs’’ monograph,‘‘arouses or increases sexual desire andimproves sexual performance’’ is anacceptable structure/function claimbecause it does not imply treatment ofa disease. ‘‘Helps restore sexual vigor,potency, and performance,’’ ‘‘improvesperformance, staying power, and sexualpotency,’’ and ‘‘builds virility andsexual potency’’ are disease claimsbecause they use the term ‘‘potency,’’which implies treatment of impotence,a disease. If, however, these claimsmade clear that they were intendedsolely for decreased sexual functionassociated with aging, they could beacceptable structure/function claims.The claim from the ‘‘Products for reliefof symptoms of benign prostatichypertrophy’’ monograph (‘‘To relievethe symptoms of benign prostatichypertrophy, e.g., urinary urgency andfrequency, excessive urinating at night,and delayed urination’’) is a diseaseclaim, because benign prostatichypertrophy meets the definition of adisease.

The claim listed in the comment fromthe ‘‘Anticholinergics’’ monograph is adisease claim. ‘‘Relieve excessivesecretions of the nose and eyes’’ refersto the characteristic signs or symptomsof hay fever. Of the claims listed in the

comment from the ‘‘Products for certainuses’’ monograph, ‘‘digestive aid,’’‘‘stool softener,’’ ‘‘weight control,’’ and‘‘menstrual’’ are, by themselves,acceptable structure/function claims ifthe labeling does not otherwise implytreatment or prevention of a disease.None mentions a characteristicsymptom of a disease. ‘‘Laxative’’ is anot a disease claim, if the labelingmakes clear that the intended use is fortreatment of occasional rather thanchronic constipation. ‘‘Nasaldecongestant,’’ ‘‘expectorant,’’ and‘‘bronchodilator’’ are disease claims.‘‘Nasal decongestant’’ is a treatment fora characteristic symptom of colds, flu,and hay fever. ‘‘Expectorant’’ is atreatment for a characteristic symptomof colds, flu, and bronchitis.‘‘Bronchodilator’’ is a treatment forbronchospasm, a characteristicsymptom of asthma.

The claim from the ‘‘Products for thetreatment and/or prevention ofnocturnal leg muscle cramps’’monograph (‘‘treatment and/orprevention of nocturnal leg musclecramps, i.e., a condition of localizedpain in the lower extremities usuallyoccurring in middle life and beyondwith no regular pattern concerning timeor severity’’) is an appropriate structurefunction claim. Nocturnal leg cramps donot meet the definition of disease.

As is clear from this response, FDAagrees that inclusion of a claim in anOTC monograph does not preclude itsuse as a structure/function claim. FDAnotes, however, that in light of thestatutory requirement that dietarysupplements bear all information that ismaterial in light of consequences thatmay result from use of the product orrepresentations made about it, dietarysupplements that contain or are labeledas containing ingredients covered by anOTC monograph and that are being soldfor the claims covered by themonograph may be misbranded to theextent that they omit materialinformation required under themonograph. For example, if the OTCmonograph required a label statementthat products containing a particularingredient should not be used bypersons taking a prescriptionmonoamine oxidase inhibitor, a dietarysupplement containing that ingredientwould be misbranded if its label did notinclude such statement.

U. Substantiation of Claims(87.) Several comments requested that

the final rule explicitly state thatstructure/function statements must beadequately substantiated and that FDAprovide guidance on what constitutesadequate substantiation. One comment



maintained that adequate substantiationis critical to ensuring that consumersreceive truthful and accurateinformation about the benefits of dietarysupplements. Another comment arguedthat this final rule should focus onadequate substantiation of claims ratherthan on delineating the boundariesbetween structure/function claims anddisease claims. Other commentsmaintained that substantiation is not aseffective in preventing consumer fraudas preapproval of the claims becauseconsumers will be using the productslong before the label claims areinvestigated.

FDA agrees that the statutoryrequirement to substantiate claims isimportant. FDA does not agree,however, that it is necessary to state inthe regulatory text of the final rule thatstructure/function claims must beadequately substantiated. Section101.93(a)(3) requires a firm notifyingFDA of a claim under section 403(r)(6)of the act to certify that the firm hassubstantiation that the claim is truthfuland not misleading. FDA also does notagree that substantiation is anappropriate alternative to distinguishingstructure/function claims from diseaseclaims. The requirement that structure/function statements and otherstatements for dietary supplementsunder section 403(r)(6) of the act beadequately substantiated is distinct fromthe requirement that such statementsnot claim to diagnose, treat, mitigate,cure, or prevent disease. Both of theserequirements are imposed by the statuteand must be complied with.

(88.) Several comments offered adviceon what types of evidence shouldconstitute adequate substantiation. Aconsumer health organization suggestedthat health claims and structure/function claims for dietary supplementsbe based on the totality of the publiclyavailable scientific evidence, includingresults from well-designed studiesconducted in a manner consistent withgenerally recognized scientificprinciples and procedures. Thecomment added that consumers wouldbe better served if standards for supportapplied to both health claims andstructure/function claims. Anotherconsumer health organization suggestedthat substantiation be based on‘‘significant scientific agreement.’’

Many of the comments suggested thatthe agency adopt FTC standards forsubstantiation. A comment from FTCexplained that FTC typically applies asubstantiation standard known as‘‘competent and reliable scientificevidence’’ to claims about the safety andeffectiveness of dietary supplements,after first looking at the overall context

to determine what the claim is. Thecomment further stated that FTC’sapproach to substantiation is consistentwith the guidance provided by thePresident’s Commission on DietarySupplement Labels, and, because FDAconcurred with the Commission’sguidance on substantiation, thecomment suggested that FDA refer tothe Commission guidance in the finalrule.

As stated above, the agency does notbelieve that this final rule is theappropriate venue to address thesubstantiation requirement. FDA does,however, agree that claims undersection 403(r)(6) of the act should besupported by adequate scientificevidence and may provide additionalguidance regarding substantiation for403(r)(6) statements at a future date.

The Commission report includedguidance on what quantity and qualityof evidence should be used tosubstantiate claims made under403(r)(6) of the act. It also containedguidance on the content of thesubstantiation files for such statements,including the 30-day notification letterto FDA, identification of the product’singredients, evidence to substantiate thestatements, evidence to substantiatesafety, assurances that goodmanufacturing practices were followed,and the qualifications of the person(s)who reviewed the data on safety andefficacy. In a notice published in theFederal Register (63 FR 23624 at23633), FDA stated that it agreed withthe guidance of the Commission. FDAencourages manufacturers of dietarysupplements making a 403(r)(6) of theact statement for a dietary supplementto follow this guidance.

(89.) A food manufacturer suggestedthat the agency require dietarysupplement manufacturers makingstructure/function claims to disclose inlabeling any and all scientific studiessupporting the claim. In addition, thecomment advocated requiring that thesestudies be performed using themarketed formulation. The commentalso urged FDA to determine howcontrary studies should be addressed.

DSHEA does not require dietarysupplement labeling that carries astatement under section 403(r)(6) of theact to include in the labeling ‘‘any andall scientific studies supporting theclaim.’’ Section 403(r)(6)(B) of the actrequires only that the ‘‘manufacturerhave substantiation that such statementis truthful and not misleading.’’Contrary studies should be consideredwhen deciding whether to make andhow to word a 403(r)(6) of the actstatement to ensure that any statementsmade are truthful and not misleading.

Additionally, in response to a requestfor substantiation for the statement, theagency would expect manufacturers toprovide a requester with contrary aswell as supporting studies.

There is no specific statutoryrequirement that the studiessubstantiating the statement beperformed using the actual marketedformulation. However, many ingredientsand factors influencing the formulationcan affect the safety and effectiveness ofthe dietary supplement. Thesevariations from the marketed productshould be considered before using astudy to substantiate a statement madefor a particular product.

V. Enforcement Issues

(90.) One comment said that theproposal shifts the burden of proof tomanufacturers to show that their filesmatch and support the claims made fortheir products.

The regulations issued by this finalrule do not address or affect the burdenof proof during enforcement actions.However, section 403(r)(6)(B) of the actclearly states that manufacturers musthave substantiation to show that thestatements that they make under section403(r)(6) of the act are truthful and notmisleading. This indicates thatmanufacturers must be prepared todemonstrate to the court that they havesupport for each claim.

(91.) One comment predictedwidespread noncompliance with therule because of its complexity andlimited FDA resources.

FDA disagrees with the comment.FDA believes that most of the rule isstraightforward, and the commentsreceived on the proposed rule indicatethat dietary supplement manufacturersunderstood the provisions of the rule.Moreover, as noted in the Analysis ofImpact in section VI.E of this document,most of the claims of which FDA hasbeen notified are consistent with thefinal rule. Thus, based on what has beenprovided to FDA, most manufacturerswould appear to be already incompliance with this final rule. If itbecomes apparent that there areprovisions that are being violatedbecause of true confusion about theirapplicability, FDA will issue clarifyingguidance. FDA agrees that itsenforcement resources are limited, andis issuing this rule in part to avoidinefficient use of those resources oncase-by-case enforcement. FDA believesthat the dietary supplement industrywill make good faith efforts to complywith this rule, once it becomes effective.



W. Other Comments

(92.) One comment said FDA shouldconduct an educational campaign toenhance public awareness of thedifferences between structure/functionclaims and disease claims and themeaning of individual claims.

FDA intends to conduct variousoutreach activities on dietarysupplement matters.

(93.) One comment said FDA shouldamend the tentative final monograph onOTC laxatives to be consistent with therule. The comment explained that thetentative final monograph should permitthe words ‘‘help maintain regularity’’ onOTC labeling.

The agency disagrees with thecomment. The fact that ‘‘helps maintainregularity’’ is an acceptable structure/function claim does not mean that itsatisfies the requirements for inclusionin an OTC monograph, including therequirement of a finding of generalrecognition of safety and effectiveness.

(94.) Several comments addressedmanufacturing or related issues. Onecomment said FDA should investigateeffects of dissolution on productpotency and efficacy, while othercomments advocated using UnitedStates Pharmacopeia standards for alldietary supplements on matterspertaining to dissolution, disintegration,purity, and potency. One commentadded that poor product quality wouldpresent a health threat to consumers andresult in economic fraud.

Another comment said FDA shouldconcentrate on standardization andquality control instead of regulatinglabeling statements, but offered nospecific suggestions. Some comments,however, made specificrecommendations. One comment saidthat product labels should contain lotnumbers and expiration dates and thatmanufacturers should conduct stabilitytests to determine accurate expirationdates. Another comment said the publicshould be protected against poormanufacturing standards for herbalproducts. Other comments simply statedthat there is substantial potential forpublic harm because there are: Multiplesources of dietary supplementingredients; multiple suppliers; a lack ofregulatory production standards, orquestions concerning product safety,efficacy, and manufacturing quality;vigorous product promotion; and asizeable market. One comment simplyasked for good manufacturing practiceregulations for dietary supplements.

Manufacturing issues are outside thescope of this rule. FDA intends to issuea separate proposed rule on currentgood manufacturing practice (CGMP) for

dietary supplements, and that proposedCGMP rule may address some of theissues raised by the comments.

III. Legal Authority

A. Scope of Section 403(r)(6) of the Act

1. Relationship Between Sections403(r)(6) and 201(g)(1)(C) of the Act

(95.) Several comments stated that theproposal mistakenly suggests that thereis only one type of structure/functionclaim that may be used for dietarysupplements. Some of these commentssaid that if a structure/function claimdoes not trigger drug status for theproduct and is not a health claim, thensuch a claim may be made in labelingfor a dietary supplement so long as it istruthful and not misleading. Thesecomments asserted that such a claim isnot subject to the notice, labeling, ordisclaimer requirements in section403(r)(6) of the act. As an example, thecomments said the claim that ‘‘calciumhelps build strong bones’’ is not a healthclaim because it does not characterize arelationship between the substance anda disease, damage, or dysfunction of thebody. The comments added that FDArecognized this in the final rule that itpublished in the Federal Register onSeptember 23, 1997 (62 FR 49859,49860, 49863, and 49864), when itstated in the preamble that claims thatcranberry juice cocktail helps maintainurinary tract health or that calciumbuilds strong bones and teeth are nothealth claims because no disease ismentioned explicitly or implicitly.Some comments added that FDA cannotsay that only those claims falling undersection 406(r)(6) of the act are structure/function claims because such a resultwould be contrary to the act and wouldmean that the proposed rule must bewithdrawn.

FDA agrees with these comments inpart and disagrees in part. The agencyagrees that statements such as ‘‘calciumhelps build strong bones’’ are not healthclaims because they do not characterizethe relationship between a substanceand a disease or health-relatedcondition. Rather, such statements arestructure/function claims authorized bysection 403(r)(6) of the act.

FDA does not agree with thecomment’s statement that dietarysupplements may bear structure/function claims without complying withthe notice, disclaimer, and otherrequirements of section 403(r)(6) of theact. Section 403(r)(6) of the act, by itsterms, applies to dietary supplements.The other possible source of authority tomake structure/function claims ondietary supplements is section201(g)(1)(C) of the act, which provides

that ‘‘articles (other than food) intendedto affect the structure or any function ofthe body of man or other animals’’ aredrugs. Under this provision, foods maymake claims to affect the structure orfunction of the body without beingregulated as drugs. By its terms,however, section 201(g)(1)(C) of the actexempts a dietary supplement that bearsa structure/function claim from drugregulation only if it is also a food. Thelast sentence of section 201(ff) of the actprovides, ‘‘Except for purposes ofsection 201(g), a dietary supplementshall be deemed to be a food within themeaning of this Act.’’ The clear importof this language is that dietarysupplements are not foods under section201(g) of the act and therefore cannotqualify for the ‘‘(other than food)’’exception to the drug definition insection 201(g)(1)(C). As a result, dietarysupplements that use structure/functionclaims may do so only under section403(r)(6) of the act and are thereforesubject to the disclaimer, notification,and other requirements in that sectionand in FDA’s implementing regulation.

The agency acknowledges that it tooka contrary position in the September1997 final rule preamble referred to inthe comment. In that preamble, FDAsaid that a dietary supplement couldbear a structure/function claim underthe ‘‘(other than food)’’ exception to thedefinition of ‘‘drug’’ in section201(g)(1)(C) of the act, provided that theclaim was truthful, non-misleading, andderived from nutritive value (see 62 FR49859 at 49860, 49863, and 49864).However, the agency has nowreconsidered in light of the plainlanguage of section 201(ff) of the act andis revoking its statements on this subjectin the September 1997 preamble (i.e.,the statements at 62 FR 49859 at 49860,49863, and 49864 concerning structure/function claims for dietary supplementsunder section 201(g)(1)(C)). It should benoted, however, that the agency is notrevoking its statements in that preambleconcerning structure/function claims forconventional foods under section201(g)(1)(C) of the act. As explained inthe September 1997 preamble (62 FR49859 at 49860), conventional foodsmay make structure/function claimsunder section 201(g)(1)(C) of the act aslong as such claims are truthful, non-misleading, and derive from thenutritive value of the food.

For a limited transition period, FDAdoes not intend to take enforcementaction against firms who have relied onthe agency’s September 1997 final rulepreamble statements to make astructure/function claim for a dietarysupplement under section 201(g)(1)(C)of the act. To allow a reasonable time for



the necessary label changes, thetransition period will last until theapplicable compliance date for the restof the rule; i.e., small businesses willhave 18 months from publication tocomply, and other firms will have 12months. As of the applicablecompliance date, firms that have beenmaking structure/function claims undersection 201(g)(1)(C) of the act musteither remove the claim or comply withthe requirements of section 403(r)(6) ofthe act and § 101.93, including notifyingFDA of the claim and relabeling to addthe required disclaimer. New structure/function claims are not subject to thistransition period; any firm that makes astructure/function claim in the labelingof a dietary supplement after theeffective date of this rule must complywith section 403(r)(6) of the act and§ 101.93.

(96.) One comment objected to asentence in the introductory paragraphin the preamble to the proposed rule.The sentence stated that, before DSHEA,certain claims could have rendered aproduct a ‘‘drug’’ under the act. Thecomment argued that even beforeDSHEA, dietary supplements couldmake structure/function claims and notbe considered drugs. The comment saidthat section 201(g)(1)(C) of the actexpressly excluded food from thedefinition of drug and that dietarysupplements fell within the ‘‘food’’exception. The comment characterizedDSHEA as limiting and restricting‘‘what had been the unconditional rightof dietary supplement marketers tomake structure/function claims.’’

The agency agrees that before DSHEA,dietary supplements that were alsofoods could make structure/functionclaims under section 201(g)(1)(C) of theact without being considered drugs.However, the passage of DSHEAchanged the regulatory framework forstructure/function claims on dietarysupplements by adding sections 201(ff)and 403(r)(6) to the act. As explained inthe response to the preceding set ofcomments, section 201(ff) of the actprovides that dietary supplements arenot considered food for purposes ofsection 201(g). Therefore, dietarysupplements may no longer makestructure/function claims under the‘‘food’’ exception to the drug definitionin section 201(g)(1)(C) of the act. FDAtherefore agrees with the comment thatin one respect, DSHEA limited theability of dietary supplement marketersto make structure/function claims.

The sentence in the introductoryparagraph of the preamble to theproposed rule correctly stated that‘‘certain claims’’—structure/functionclaims for dietary supplements that

were not also foods—could haverendered the product a drug before thepassage of DSHEA (63 FR 23624). Post-DSHEA, however, dietary supplementsmay make structure/function claimsunder section 403(r)(6) of the actregardless of whether they are alsofoods. Thus, although in one wayDSHEA did limit the ability of dietarysupplement marketers to makestructure/function claims, it alsosignificantly expanded the opportunityto make structure/function claims inanother way by removing the limitationthat dietary supplements must be foodsto make structure/function claims.Under section 403(r)(6) of the act,claims may be made for nondiseaseeffects of a dietary supplement on thestructure or function of the body,regardless of whether those effects arenutritive, as long as the product isintended to supplement the diet asprovided in section 201(ff)(1) of the act.2. Structure/Function Claims forConventional Foods

(97.) Several comments soughtconsistency in the treatment ofconventional foods and dietarysupplements with respect to structure/function claims and health claims.Some of these comments contended thatthis rule would permit dietarysupplements to carry claims that wouldbe health claims if made for aconventional food. One comment statedthat differential treatment of foods anddietary supplements was inconsistentwith the Commission’srecommendations. This commentsuggested that differential treatmentwould cause consumers to perceivedietary supplements as better sourcesfor safeguarding health thanconventional foods. One commentexpressed the view that the rule shouldapply to claims for conventional foodsas well as dietary supplements andrequested FDA to clarify the rule’sscope. Other comments said that anystructure/function claims that may bemade for dietary supplements may alsobe made for conventional foods. Thecomments explained that the history ofthe act shows that claims that foodaffect the structure or function of thebody do not result in the food beingclassified as a drug, citing the districtcourt and appellate decisions inAmerican Health Products Co. v. Hayes,574 F. Supp. 1498, 1501 (S.D.N.Y.1983), aff’d, 744 F.2d 912 (2d Cir. 1984).Another comment stated thatestablished case law shows that anarticle may be a food if it is usedprimarily for taste, aroma, or nutritionalvalue, but that nutritional value is notrequired in all instances. One commentfurther noted that FDA, when it

implemented the labeling requirementsfor DSHEA (62 FR 49859, 49860, and49861) said that it was committed to ‘‘asmuch parity between dietarysupplements and conventional foods asis possible within the statute’’ and thatFDA has recognized that a dietarysupplement may lawfully be inconventional food form, but must berepresented as a dietary supplement(citing 62 FR 49826 at 49837, September23, 1997).

Given this background, the commentsargued that FDA cannot take theposition that a structure/function claimmay be made for a conventional foodonly if the effect derives from the food’snutritional value. One comment addedthat the act does not distinguish foodsbased on their nutritional value and thatDSHEA considers structure/functionclaims for all dietary ingredients to be‘‘statements of nutritional support.’’ Thecomment said FDA, therefore, shouldrecognize that structure/function claimsthat can be made for dietary ingredientswhen those ingredients are in dietarysupplements can also be made whenthose ingredients are in conventionalfood, but added that the disclaimerstatement and notification to FDA, asrequired by section 403(r)(6)(C) of theact, apply only to dietary supplementsand not to conventional food. Onecomment said that requiring structure/function claims for conventional foodsto be derived from the food’s nutritionalvalue would create a marketingdisparity and put conventional foods ata competitive disadvantage.

This rule applies to claims for dietarysupplements only. Its purpose is toimplement section 403(r)(6) of the act,which applies to dietary supplementsonly. Therefore, a detailed discussion ofthe regulatory framework applicable tostructure/function claims forconventional foods, which are madeunder section 201(g)(1)(C) of the act, isbeyond the scope of the rule. FDAadvises, however, that for consistency,the agency is likely to interpret thedividing line between structure/function claims and disease claims in asimilar manner for conventional foodsas for dietary supplements. The agencyalso notes that as discussed in theresponse to comment 1 in section II.Aof this document, FDA reaffirms thestatements about structure/functionclaims for conventional foods in theSeptember 23, 1997 (62 FR 49859), finalrule entitled ‘‘Food Labeling: NutrientContent Claims, Health Claims, andStatements of Nutritional Support forDietary Supplements.’’ As explained inthat rule (62 FR 49859 at 49860, 49861,and 49864), the fact that structure/function claims for conventional foods



are limited to effects derived fromnutritional value, while structure/function claims for dietary supplementsare not, is a result of differences in thelanguage of the exemption for foods insection 201(g)(1)(C) of the act, asinterpreted by the courts (see Nutrilab,Inc. v. Schweiker, 713 F.2d 335, 338 (7thCir. 1983)), and the language of section403(r)(6) of the act.

(98.) One comment suggested revisingthe definition of ‘‘disease or health-related condition’’ in proposed§ 101.14(a)(6) to include a reference to§ 101.93, and also recommendedrevising the definition of ‘‘health claim’’at § 101.14(a)(1) to be consistent with§ 101.93. Currently, § 101.14(a)(1) readsas follows:

Health claim means any claim made on thelabel or in labeling of a food, including adietary supplement, that expressly or byimplication, including ‘‘third party’’references, written statements (e.g., a brandname including a term such as ‘‘heart’’),symbols (e.g., a heart symbol), or vignettes,characterizes the relationship of anysubstance to a disease or health-relatedcondition. Implied health claims includethose statements, symbols, vignettes, or otherforms of communication that suggest, withinthe context in which they are presented, thata relationship exists between the presence orlevel of a substance in the food and a diseaseor health-related condition.The comment would revise thedefinition to read as follows:

Health claim means any claim made on thelabel or in labeling of a food, including adietary supplement, that expressly or byimplication, including ‘‘third party’’references, written statements (e.g., a brandname that includes or implies a disease, suchas ‘‘Raynaudin’’), symbols, or vignettes,characterizes the relationship of anysubstance to a disease or health-relatedcondition (e.g., disease-indicatingelectrocardiogram tracings, pictures of organsthat suggest prevention or treatment of adisease state, the prescription symbol, or anyreference to prescription use). Implied healthclaims include those statements, symbols,vignettes, or other forms of communicationthat suggest, within the context in whichthey are presented, that a relationship existsbetween the presence or level of a substancein the food and a disease or health-relatedcondition.

As stated in response to comment 51of section II.I of this document, FDAdoes not believe that §§ 101.14(a)(1) and101.93(g) are inconsistent. As a result ofthe special regime for dietarysupplements under DSHEA, there maybe some differences in the treatment ofdietary supplements and conventionalfoods under § 101.14(a)(1).3. Relationship Between Structure/Function Claims and Health Claims

(99.) One comment stated that theproposed rule ‘‘improperlydistinguishes between other health-

related claims and structure/functionclaims.’’ Relying in part on theintroduction to section 403(r)(6) of theact (‘‘For purposes of paragraph (r)(1)(B)* * *’’), the comment asserted thatstructure/function claims are a subset ofthe claims authorized by section403(r)(1)(B) of the act (health claims).Consequently, because claims undersection 403(r)(1)(B) of the act maycharacterize the relationship of anutrient to a disease, the commentstated that FDA cannot precludestructure/function claims from makingany contextual references to diseases.

FDA disagrees with this comment.Structure/function claims are not asubset of health claims because, clearly,there are claims about the effect of aproduct on the structure or function ofthe body that are not also health claims.To be a health claim, a claim must referto the relationship between a foodsubstance and a disease or health-related condition. FDA interprets‘‘health-related condition’’ to mean astate of health leading to disease. Claimssuch as ‘‘calcium builds strong bones’’are not health claims because they donot refer explicitly or implicitly to anydisease or health-related condition.Therefore, the comment is based on aninvalid premise.

(100.) One comment requested thatFDA revise § 101.93(f) to state that therequirements of section 403(r)(6) of theact, e.g., use of the disclaimer andsubstantiation, apply only to structure/function claims that fall within thedefinition of a ‘‘health claim’’ in§ 101.14(a)(1) and (a)(5). According tothis comment, the introduction tosection 403(r)(6) of the act (‘‘Forpurposes of paragraph (r)(1)(B) * * *’’)establishes that structure/functionclaims that do not fall within thedefinition of health claims are notsubject to section 403(r)(6), and may bemade without complying with any of itsrequirements.

FDA does not agree and, in fact,believes that the opposite is true. Asexplained elsewhere in this documentand in the proposed rule, structure/function claims that fall within thedefinition of health claims, or thatotherwise constitute disease claims, donot fall within the scope of claimsauthorized under section 403(r)(6) of theact, but other structure/function claimsdo fall within the scope of section403(r)(6) and are subject to itsrequirements. Adopting theinterpretation advocated by thecomment would bring about illogicalresults for dietary supplement labelingclaims in two ways. First, structure/function claims that are also healthclaims would not be subject to the

health claims prior authorizationrequirements, but instead could bemade simply by meeting therequirements of section 403(r)(6) of theact and FDA’s implementingregulations. The language in section403(r)(6) of the act excluding claims toaffect disease from the coverage of thatsection demonstrates that Congressmade a public health judgment thatclaims promoting dietary supplementsfor disease uses should continue torequire premarket authorization. Itwould not make sense for Congress toexclude labeling claims pertaining todisease uses in one part of section403(r)(6) of the act, while permittingsuch claims in another paragraph of thesame section. Moreover, theinterpretation advocated by thecomment would lead to confusing andcontradictory labeling. A dietarysupplement that bears a health claim—a claim that, by definition, is a claimthat a substance in the supplement insome way has an effect on a disease—would also have to bear a contradictorydisclaimer that it is not intended totreat, mitigate, or prevent any disease.Second, structure/function claims thatare not also health claims would not beauthorized under section 403(r)(6) of theact at all. In fact, a structure/functionclaim on a dietary supplement wouldsubject it to drug regulation because, asexplained in the response to comment 1in section II.A of this document, section403(r)(6) of the act is the only provisionthat authorizes the use of structure/function claims on dietary supplements.

The introductory language in section403(r)(6) (‘‘For purposes of [section403](r)(1)(B) * * *’’) does not supportthe interpretation advocated in thecomment. If Congress had wanted tosubject only structure/function claimsthat are also health claims to section403(r)(6) of the act, it could have doneso much more directly by usinglanguage such as ‘‘A statement for adietary supplement may be made if * ** and the statement is a statement of thetype governed by paragraph (r)(1)(B).’’The ambiguity of the ‘‘For purposes of(r)(1)(B)’’ language is well demonstratedby the diametrically opposedinterpretations adopted by thiscomment and the preceding comment.FDA interprets this language as acaution that the category of claimscovered by section 403(r)(6) of the act isnot to be interpreted as coextensive withhealth claims, the category covered bysection 403(r)(1)(B) of the act. Congressmay have been concerned that thehealth claims category would swallowthe category of claims under section403(r)(6) of the act because all claims



under section 403(r)(6) could becharacterized as referring to a ‘‘health-related condition’’ if that term weredefined broadly as ‘‘a state of health.’’The result would have been that allstructure/function claims, as claimsabout the relationship between asubstance and a health-relatedcondition, would also have been healthclaims and would have requiredpremarket authorization. By includingthe introductory language, Congresseffectively forestalled such aninterpretation.

(101.) Another comment said theproposed rule does not distinguishbetween structure/function statementsthat assert health claims and those thatdo not, and said the failure to make thisdistinction would mean that moreproducts would be subject to the rulethan necessary.

FDA does not agree that the rule failsto distinguish between structure/function claims that do and do notassert health claims. On the contrary,the rule makes clear that only structure/function claims that do not assert healthclaims may be made under section403(r)(6) of the act. To the extent thatthe comment may be suggesting thatstructure/function claims that are alsohealth claims should be exempt fromthe health claims authorizationrequirements, the agency disagrees forthe reasons given in the response to theprevious comment.

B. Miscellaneous Legal Issues(102.) Two comments said the

proposed rule violated theAdministrative Procedure Act because itwas arbitrary and capricious, on twogrounds. One comment asserted thatFDA failed to consider an importantaspect of the problem of distinguishingbetween drug claims and dietarysupplement claims: The application ofthe ‘‘general well-being’’ provision ofsection 403(r)(6) of the act. Thecomment argued that FDA should haveconsidered whether claims relating tonormal body functions might qualify as‘‘general well-being’’ claims undersection 403(r)(6) of the act beforedeciding to regulate them as diseaseclaims. The comment also argued thatFDA’s explanation of the need for theproposed rule ran counter to theevidence before the agency, in that theagency’s actions on notifications ofclaims under section 403(r)(6) of the actdid not support a need for furtherregulation.

The ‘‘general well-being’’ provision ofsection 403(r)(6) of the act authorizesstatements in dietary supplementlabeling that describe ‘‘general well-being from consumption of a nutrient or

dietary ingredient’’ (section 403(r)(6)(A)of the act). FDA did not considerwhether statements were authorizedunder this provision in developing theproposed rule because the purpose ofthe rule was to implement the structure/function provisions of section403(r)(6)(A) of the act, not otherprovisions. However, consideration ofthis provision as applied to normal bodyfunctions would not have led to adifferent result. The criteria in the rulewere developed to identify claims thatrefer directly or indirectly to an effecton disease and do not encompass claimsthat refer only to general well-being.Claims relating to normal bodyfunctions are authorized under the rule.

The comment’s argument about theuse of FDA’s actions on notifications ofclaims under section 403(r)(6) of the actto justify the rule is addressed incomment 4 of section II.A of thisdocument.

(103.) One comment claimed that theproposal does not require FDA to showany evidence of a manufacturer’s intentto find that a dietary supplement claimconstitutes an illegal drug claim. Thecomment argued that proposed§ 101.93(g)(2)(ii), (g)(2)(iii), (g)(2)(viii),and (g)(2)(x) run afoul of the recentappellate decision in Brown &Williamson Tobacco Corp. v. FDA, 153F.3d 155 (4th Cir. 1998), contending that‘‘a product is not a drug merely becausea consumer uses it as one’’ and that‘‘there must be proof as to themanufacturer’s intent.’’ The commentalso cited National Nutritional FoodsAss’n v. Mathews, 557 F.2d 325 (2d Cir.1977), to support its position that amanufacturer’s intent, as determinedfrom labeling or advertising, is theprimary factor in determining whether aproduct is intended to treat a disease.

Although FDA disagrees with theBrown & Williamson decision and isawaiting the outcome of Supreme Courtreview, this rule does not depend on theresolution of the legal issues in thatcase. The focus of the rule is on expressand implied claims made by the vendorin labeling. None of the provisions ofthe rule, including those mentioned inthe comment, rely on consumer use asa standard for determining whether theproduct is intended to treat or preventdisease.

The rule is consistent with thedecision in National Nutritional FoodsAss’n v. Mathews, in which the courtsaid, ‘‘FDA is not bound by themanufacturer’s subjective claims ofintent but can find actual therapeuticintent on the basis of objective evidence.Such intent also may be derived orinferred from labeling, promotionalmaterial, advertising, and ‘any other

relevant source’’’ (557 F.2d at 334(citations omitted)). See also § 201.128(listing evidence FDA will consider indetermining the intended use of a drug).

(104.) One comment said that theproposal must be withdrawn because,contrary to section 403(r)(6) of the act,it gives manufacturers the burden toprove that a claim is not a drug claimwhen, in fact, FDA has the burden, bya preponderance of relevant evidence, toestablish that a dietary supplement ismisbranded. The comment cited twocourt opinions, United States v. 29Cartons * * * an Article of Food(Oakmont), 987 F.2d 33 (1st Cir. 1993)and United States v. An Article of Food* * * Viponte Ltd. Black Currant Oil, 984F.2d 814 (7th Cir. 1993), for theproposition that, before DSHEA wasenacted, courts had invalidated an FDAenforcement theory that shifted theburden of proof to manufacturers.

FDA disagrees with this comment.Although the comment is correct thatFDA has the burden of proving that adietary supplement—or, in fact, anyfood—is misbranded, the rule does notgive manufacturers the burden ofproving that a claim is not a drug claim.The rule does not shift the burden ofproof in an enforcement action butrather sets forth criteria for what claimsare disease claims that may subject aproduct marketed as a dietarysupplement to regulation as a drug.

The two cases cited in the commentare inapposite. They concern FDA’sefforts to regulate certain dietaryingredients as food additives and do nothave any relevance to claims issues.

(105.) One comment said that theproposed rule is inconsistent with theact and congressional intent, arguingthat, by enacting DSHEA, Congress hadtaken steps to reverse FDA’s ‘‘overlyrestrictive’’ approach towards claimsand had commanded the agency toexpand, rather than restrict, the amountof health information permitted ondietary supplement labels and labeling.According to the comment, the proposal‘‘directly and substantially violates theoverall statutory scheme and theexpressed legislative intent’’ and FDA‘‘has no authority to proceed with therulemaking without a grant of authorityfrom Congress in light of the Act’slanguage and Congressional intent.’’

The agency disagrees with thiscomment and believes that the rule isconsistent with the act andcongressional intent. AlthoughCongress, in enacting DSHEA, didexpand the scope of information indietary supplement labeling byproviding for claims to affect thestructure or function of the body and theother types of claims authorized by



section 403(r)(6) of the act, Congressalso explicitly limited statements undersection 403(r)(6) to those that do notclaim to ‘‘diagnose, mitigate, treat, cure,or prevent a specific disease or class ofdiseases.’’ This rule does not create newrestrictions but merely implements theprovisions of section 403(r)(6) of the act.FDA has authority to issueimplementing regulations under section701(a) of the act, which authorizes theagency to issue regulations for theefficient enforcement of the act.

(106.) One comment declared thatFDA has no legal basis to include abroad variety of implied claims.

FDA disagrees with this comment.The agency has regulated impliedclaims in labeling for many years, inmany contexts. (See, e.g., 21 CFR104.5(b) and (d) (prohibiting certainimplied claims relating to compliancewith nutritional quality guidelines); 21CFR 101.13(a) (classifying impliedclaims to characterize the level of anutrient in food as nutrient contentclaims subject to the same requirementsas express claims); 21 CFR 101.95(prescribing conditions under whichimplied claims of freshness may bemade for foods); 21 CFR 201.10(c)(3)(prohibiting use in ingredient statementof fanciful drug or ingredient names thatfalsely imply that the drug or ingredienthas some unique effectiveness orcomposition); 21 CFR 201.302(c)(prohibiting implied claims that drugsfor internal use that contain mineral oilare for administration to infants). Theagency has also regulated impliedclaims in prescription drug advertising.(See, e.g., § 202.1(a)(3) (21 CFR202.1(a)(3)) (prohibiting use inadvertising of fanciful product oringredient names that falsely imply thatthe drug or ingredient has some uniqueeffectiveness or composition);§ 202.1(e)(6)(v) (prohibiting impliedclaims that a study represents morewidespread experience with the drugthan it actually does).) Morespecifically, the agency has repeatedlytaken the position that implied diseaseclaims in labeling subject a product toregulation as a drug. In the animal drugcontext, § 500.52 (21 CFR 500.52)provides that the use of certain terms inthe labeling of products intended for usein or on animals implies that theproduct is capable of a therapeutic effectand causes the product to be a drugwithin the meaning of section 201(g) ofthe act. In the human drug context,§ 201.56(c) (21 CFR 201.56(c)) prohibits‘‘implied claims or suggestions of druguse’’ in prescription drug labelingunless the product has been shown to besafe and effective for the implied orsuggested use. (See also § 310.530 (21

CFR 310.530) (use of the word‘‘hormone’’ in labeling is an implieddrug claim).) Moreover, courts haveupheld FDA’s authority to regulateimplied drug claims. (See, e.g., UnitedStates v. Storage Spaces DesignatedNos. ‘‘8’’ and ‘‘49’’, 777 F.2d 1363, 1366& n. 5 (9th Cir. 1985), cert. denied, 479U.S. 1086 (1987); Pasadena ResearchLabs., Inc. v. United States, 169 F.2d375, 383 (9th Cir.), cert. denied, 335U.S. 853 (1948); United States v. SixDozen Bottles * * * ‘‘Dr. Peter’s Kuriko’’,158 F.2d 667, 669 (7th Cir. 1947);United States v. John J. Fulton Co., 33F.2d 506, 507 (9th Cir. 1929); Bradley v.United States, 264 F. 79, 81–82 (5th Cir.1920); United States v. Kasz Enterprises,Inc., 855 F. Supp. 534, 539, 543–44(D.R.I. 1994), modified on othergrounds, 862 F. Supp. 717 (D.R.I.1994);United States v. 43 1/2 Gross RubberProphylactics, 65 F. Supp. 534, 535 (D.Minn. 1946), aff’d sub nom. Gellman v.United States, 159 F.2d 881 (8th Cir.1947).)

(107.) Many comments argued that theproposed rule ignored the SupremeCourt decision in Daubert v. MerrellDow Pharmaceuticals, Inc., 509 U.S. 579(1993).

FDA disagrees with these comments.The comments did not explain how therule was contrary to or even affected bythe decision. Daubert involved theadmissibility of scientific evidence in ajudicial proceeding under the FederalRules of Evidence. This rulemakingdoes not present issues regarding theadmissibility of evidence in anyproceeding, judicial or administrative,nor does it address expert testimony(which was at issue in Daubert). Thus,FDA does not agree that the rule‘‘ignores’’ or is contrary to the Daubertdecision.

C. Constitutional Issues1. First Amendment

(108.) Several comments focused onthe First Amendment. One commentargued that the rule violates the FirstAmendment because it is morerestrictive than is necessary to advanceFDA’s interests. The comment concededthat the government may regulate orprohibit commercial speech if thespeech is inherently false, deceptive, ormisleading, but argued that thegovernment can only restrictcommercial speech that is not false,deceptive, or misleading if thegovernment shows that the restrictiondirectly and materially advances asubstantial state interest in a mannerthat is no more extensive than necessaryto serve that interest (citing Ibanez v.Florida Dept. Of Bus. & Prof’lRegulation, 512 U.S. 136, 142 (1994);

Central Hudson Gas & Elec. Corp. v.Pub. Serv. Comm’n, 447 U.S. 557, 566(1980)). The comment argued that notall structure/function claims prohibitedunder the proposed rule are inherentlyfalse or misleading and that if FDA doesnot review the evidence for a claim, theclaim does not become false ormisleading. Although the commentadmitted that FDA has a substantialinterest in regulating the safety, efficacy,and labeling of dietary supplements inorder to protect the public health, thecomment claimed that the regulationwas more extensive than necessary. Thecomment argued that a disclaimer is‘‘the constitutionally mandated methodof regulating commercial speech.’’

Other comments said the proposedrule violates the First Amendmentbecause, using the analysis in CentralHudson Gas & Electric Corp. v. PublicService Commission, 447 U.S. 557(1980), it is not narrowly tailored tomeet FDA’s interests and does notdirectly and materially advance theagency’s interests. In general, thesecomments offered various reasons whythe proposed rule did not survivescrutiny under Central Hudson. Forexample, under Central Hudson, thegovernment may regulate commercialspeech that concerns unlawful activityor is misleading if, among other things,the government asserts a substantialinterest in support of its regulation. Inbrief, the comments said FDA failed toassert a substantial interest or construedthe government’s interest to beCongress’ interest in increasing theamount of information to consumers.Others said that, contrary to CentralHudson, the proposed rule was notnarrowly tailored and suppressed morespeech than necessary to protect apossible government interest inprotecting consumers from fraud andprotecting public health and eithersuggested alternatives or said FDAshould consider less restrictivealternatives. Some comments said theproposal also did not advance theasserted government interest because itblurred, instead of clarified, the linebetween drug and dietary supplementclaims.

One comment also asserted that thereis no substantial government interestinvolved, because FDA has not shown aconcern for consumer safety or a dangerto public health; according to thiscomment, the proposed rule was aresponse to confusion by manufacturersand consumers about what claims arepermitted.

Some comments also argued that FDAhas not shown that the claims aremisleading or that the commercialspeech covered by the proposed rule is



inherently misleading. One commentasserted that, if statements wereuntruthful or misleading, DSHEA wouldhave prohibited them.

Another comment said the proposal‘‘trenches on’’ the First Amendmentbecause consumers have the right toreceive, and manufacturers have theright to express, non-misleadinginformation. The comment citedWashington Legal Foundation v.Friedman, 13 F. Supp. 2d 51 (D.D.C.1998) for this proposition. Anothercomment cited the Washington LegalFoundation decision to argue that theproposed rule would ‘‘impermissiblycurtail’’ the flow of information toconsumers. The comment suggested thatless restrictive alternatives, such as‘‘allowing implicit, but not explicit,claims,’’ establishing ‘‘categories ofdiseases that clearly denoted drugclaims’’ or identifying terms thatconnote ‘‘treatment,’’ ‘‘cure,’’ or‘‘mitigation’’ exist.

A few comments simply claimed thatthe proposal violates the FirstAmendment because it would decreasethe amount of scientific information onlabels and labeling or because itrepresents a ‘‘prior restraint’’ on healthclaims. Other comments objected toparticular provisions of the proposedrule on First Amendment grounds,notably proposed § 101.93(g)(2)(iv)(C),which provided that citation of the titleof a scientific reference in dietarysupplement labeling would be a diseaseclaim if the title referred to a disease useof the product. Several comments saidthat this provision of the proposed rulewould violate the First Amendment asan unlawful restraint on commercialspeech. Others characterized theproposed provision as simply arestriction on freedom of speech,whether the restriction was on the rightof companies to provide the informationor on the right of consumers to receivethe information. One comment said thatreferences to publication titles could beprohibited if they were misleading, butthat the rule should not contain ablanket prohibition. Some commentsadded that the agency should reconsiderits position on this provision in light ofWashington Legal Foundation v.Friedman.

Finally, a comment said that theproposal was contrary to the decision ofthe U.S. Court of Appeals for the Districtof Columbia Circuit in Pearson v.Shalala, 164 F.3d 650 (D.C. Cir. 1999).According to the comment, the court ofappeals’ First Amendment ruling inPearson requires the agency to permithealth claims that do not satisfy the‘‘significant scientific agreement’’standard as long as the claim can be

rendered non-misleading by requiring adisclaimer. According to the comment,the court’s decision also requires FDA tofurther define the ‘‘significant scientificagreement’’ standard for authorizingdietary supplement health claims. Thecomment said that the proposed rulewas premature in light of the need toamend the health claims regulations toconform to the Pearson decision. Thecomment also argued that, in light ofPearson, FDA may not issue a final rulethat prohibits disease claims but rathermust choose the less restrictivealternative of permitting such claimsprovided that they are accompaniedwith disclaimers.

FDA does not believe that the ruleviolates the First Amendment. The ruledoes not prohibit any speech; rather, itclarifies the circumstances under whichFDA will consider a certain type ofspeech—labeling claims—to beevidence of intended use as a drug,absent health claim authorization. Thus,the rule does not regulate speech assuch, but rather as evidence of intendeduse. The use of speech as evidence of acompany’s intended use for its productsis constitutional because ‘‘[t]he FirstAmendment * * * does not prohibit theevidentiary use of speech * * * to provemotive or intent’’ (Wisconsin v.Mitchell, 508 U.S. 476, 489 (1993).) (Seealso Village of Hoffman Estates v.Flipside, 455 U.S. 489, 495–96 (1982)(upholding village ordinance treatingthe proximity of drug-oriented literatureas evidence that items were marketedfor use with illegal drugs). Because it isthe intent and not the speech thattriggers a regulatory burden on thespeaker, there is no First Amendmentviolation. (See Wisconsin v. Mitchell,508 U.S. at 489; United States v. Articlesof Drug * * * B–Complex CholinosCapsules, 362 F.2d 923, 927 (3d Cir.1966) (no impingement on free speechfor FDA to use statements made by alecturer employed by a manufacturer asevidence of the manufacturer’s intentthat its products be used for therapeuticpurposes).)

Even if the rule were viewed as adirect restriction on speech, it wouldnot violate the First Amendment. Themarketing in interstate commerce of adrug that has not been determined byFDA to be safe and effective is illegal(see section 301(a) and (d) of the act (21U.S.C. 331(a) and (d)) and 505 of the act.Thus, labeling claims that promote adietary supplement for disease usespromote the product for use as anunapproved new drug, which is illegal.Speech promoting an illegal activitymay be restricted without violating theFirst Amendment (Central Hudson, 447U.S. at 563–564). In Pittsburgh Press Co.

v. Pittsburgh Comm’n on HumanRelations, 413 U.S. 376 (1973), theSupreme Court held that anadvertisement could be prohibitedwhere it indicated that the advertiserwas likely to have an illegal intent whileengaging in the proposed transaction(id. at 389). There, as here, ‘‘therestriction * * * is incidental to a validlimitation on economic activity’’ (id.).

Nor does the rule create anunconstitutional prior restraint. FDAdoes not believe that the regulations in§ 101.93(f) and (g) are properly analyzedas a prior restraint at all. As explainedpreviously, the regulations do notrestrict speech but rather treat it asevidence of a product’s intended use.Using speech to infer intent does notviolate the First Amendment (Wisconsinv. Mitchell, 508 U.S. 476, 489 (1993)).Thus, the regulations do not preventspeech from happening, but, asevidence of intended use, theydetermine the consequences that resultfrom certain types of speech. (SeeVillage of Hoffman Estates v. Flipside,455 U.S. at 495–96 (rejecting headshop’s ‘‘exorbitant’’ claim that villageordinance treating the proximity ofdrug-oriented literature as evidence ofintended use was a prior restraint).)

Although the regulations cannotthemselves be considered as a directprior restraint, it is true that claimsclassified as disease claims under theregulations are subject to priorauthorization requirements that couldbe considered prior restraints—namely,the prior authorization requirement fordietary supplement health claims andthe new drug approval requirementsthat are triggered in the absence ofhealth claim authorization. In bothcases, a disease claim cannot be madeuntil FDA has evaluated the safety ofthe product and the evidencesupporting the claim. However, labelingclaims are commercial speech, and theSupreme Court has indicated that theprior restraint doctrine may not apply tocommercial speech. (See CentralHudson, 447 U.S. at 571 n.13(‘‘[C]ommercial speech is such a sturdybrand of expression that traditionalprior restraint doctrine may not apply toit.’’; Virginia State Bd. of Pharmacy v.Va. Citizens Consumer Council, 425U.S. 748, 771–72 n.24 (1976) (greaterobjectivity and hardiness of commercialspeech may make prior restraintdoctrine inapplicable). Commercialspeech is ‘‘sturdy’’ because of its profitmotive. ‘‘[S]ince advertising is the sinequa non of commercial profits, there islittle likelihood of its being chilled byproper regulation and forgone entirely’’(Virginia State Bd. of Pharmacy, 425U.S. at 771–72 n.24). The same is true



of labeling. The Supreme Court hasexpressed approval of prior reviewrequirements in commercial speechcases. (See Shapero v. Kentucky BarAss’n, 486 U.S. 466, 476 (1988) (lawyermay be required to file solicitation letterwith State in advance, to give it ‘‘ampleopportunity to supervise mailings andpenalize actual abuses’’);CentralHudson, 447 U.S. at 571 n.13 (State mayrequire ‘‘a system of previewingadvertising campaigns’’).)

If the prior authorization requirementfor dietary supplement health claimsand the approval requirement for newdrugs were to be considered priorrestraints, they would be constitutionalprior restraints. The only court ofappeals to address the issue in thehealth claims context ruled that thehealth claims authorization process isnot an unconstitutional prior restraint.In a recent case challenging the NLEAand FDA’s health claim regulations fordietary supplements, the U.S. Court ofAppeals for the Second Circuit held thatthe prior restraint doctrine did apply,but it went on to uphold the statute andregulations based on consideration ofthe Central Hudson factors. NutritionalHealth Alliance v. Shalala, 144 F.3d220, 227–28 (2d Cir.), cert. denied, 119S. Ct. 589 (1998). In Nutritional HealthAlliance, the Second Circuit held thatthe health claims authorization processis ‘‘sufficiently narrowly tailored’’ andhas adequate procedural safeguards—including a deadline for final agencyaction, a decision making standard toconstrain the agency’s discretion, andprovision for development of a recordfor judicial review—to render itconstitutionally valid (144 F.3d at 228;see § 101.70 (procedures for petitioningfor a health claim)). In upholding theregulatory scheme, the court alsostressed that matters of public healthand safety were involved (144 F.3d at228). The same considerations that thecourt in Nutritional Health Alliancerelied on also operate in the new drugapproval context: Matters of publichealth and safety are involved, and theact and implementing regulationsprovide many procedural safeguards,including a deadline, a decision makingstandard, and the development of anrecord for judicial review (see section505(c)(1), (d), and (h) of the act and; 21CFR 314.200.) Moreover, as far as FDAis aware, the constitutionality of thenew drug approval process has neverbeen challenged on First Amendmentgrounds. Therefore, FDA does notbelieve that the prior restraint argumentin the comments has merit.

Many of the comments assumed thatthe test for restrictions on commercialspeech set forth by the Supreme Court

in Central Hudson applies. FDAbelieves that it is not necessary to reachthe Central Hudson test because the ruleis constitutional under Wisconsin v.Mitchell, Pittsburgh Press, and Village ofHoffman Estates; however, the rule alsoeasily passes muster under the four-parttest in Central Hudson. Under that test,the first question is whether thecommercial speech at issue is false,misleading, or concerns unlawfulactivity, because such speech is beyondthe First Amendment’s protection andmay be prohibited. If the speech istruthful, non-misleading, and concernslawful activity, the government maynonetheless regulate it if thegovernment interest asserted to justifythe regulation is substantial; theregulation directly advances theasserted governmental interest; and theregulation is no more extensive thannecessary to serve the governmentinterest (Central Hudson, 447 U.S. at566). The Supreme Court has explainedthat the last element of the test is not a‘‘least restrictive means’’ requirement,but rather requires narrow tailoring—‘‘afit that is not necessarily perfect, butreasonable’’ between means and ends(Board of Trustees of the State Univ. ofN.Y. v.Fox, 109 S. Ct. 3028, 3032–35(1989)). In subsequent decisions, theCourt has also clarified that‘‘misleading’’ in the first element of thetest refers to speech that is inherently oractually misleading. Thus, if the speechto be regulated is not inherently oractually misleading, the remainder ofthe test applies. (See In re R.M.J., 455U.S. 191, 203 (1982).)

As previously discussed, FDAbelieves that claims for disease uses thathave not been found to be safe andeffective are speech related to anunlawful activity, and therefore there isno need to reach the remaining elementsof the Central Hudson test. The agencyalso considers such claims inherentlymisleading because, when accompaniedby a disclaimer that directly contradictsthe claim by stating that the product isnot intended to have an effect ondisease, they are inherently likely toconfuse consumers rather than providethem with useable information. Speechthat is ‘‘more likely to deceive thepublic than to inform it’’ is notprotected by the First Amendment(Central Hudson, 447 U.S. at 563). If notinherently misleading, claims fordisease uses that have not been found tobe safe and effective are at leastpotentially misleading because of theconfusion caused by the disclaimer.Such claims also may lead consumers tobelieve that the product has benefits in

treating or preventing disease, even ifthat is not the case.

Even if the remaining elements of theCentral Hudson test are reached, therule and the statutory provisions that itimplements are constitutional. Aspreviously noted, this rule restricts nospeech directly. Rather, it determineswhat types of speech in dietarysupplement labeling will trigger otherstatutory provisions and regulations thatmay be considered restrictions onspeech. To the extent that this rule, thestatute, and the drug and health claimregulations restrict speech by requiringeither health claim authorization or newdrug approval before a business maymake a disease claim for a dietarysupplement, that restriction directlyadvances the substantial governmentinterest in protecting and promoting thepublic health by helping to ensure thatproducts intended to have an effect ona disease are safe and effective for thatintended use. That interest is an interestboth in preventing direct harm fromsuch products—i.e., protecting thepublic from adverse events that suchproducts might cause—and inpreventing the indirect harm to healththat is caused when an ill personforegoes medical care in favor ofineffective self-treatment.

Requiring prior FDA review andauthorization of disease claims ensuresthat such claims will be evaluated by apublic health agency that has scientificand medical expertise so that onlyproducts that are safe and effective willbe permitted to be sold for therapeuticpurposes. As a government agency withno financial stake in either permitting ordenying claims, FDA is in a position toevaluate the strength of the safety andefficacy evidence objectively.

The rule and the other components ofthe regulatory framework for drugs andhealth claims also advance the relatedsubstantial government interest inprotecting consumers from fraud. Ifproducts are marketed for disease usesonly after they have been demonstratedto be safe and effective for such uses,consumers will not suffer economicharm from spending money onworthless remedies.

Moreover, the rule is not moreextensive than necessary. The agencydoes not believe that the alternativesmentioned in the comments, or anyother alternative, would adequatelyfurther its substantial interest inprotecting and promoting public healthby ensuring the safety and efficacy ofproducts intended to have an effect ondisease. For example, allowing implicitdisease claims, but not explicit ones,would merely allow companies to doindirectly what they cannot do



directly—to market products for diseaseuses without demonstrating their safetyand efficacy. Likewise, identifyingspecific terms that connote treatment,cure, or mitigation would notaccomplish the goal of requiring proofof the safety and effectiveness ofproducts marketed for disease uses.Merely regulating synonyms for thoseterms would leave unregulated thoseclaims that achieve the same effectwithout using such a synonym, such asthe claims ‘‘herbal Prozac’’ and ‘‘forcancer.’’ The suggestion in onecomment that FDA establish ‘‘categoriesof diseases that clearly denote drugclaims’’ is not a workable alternativeeither. Section 403(r)(6) of the actprovides that the category of structure/function claims excludes claims toaffect any category of disease, not justcertain categories.

Permitting disease claims undersection 403(r)(6) of the act as long asthey are accompanied with a disclaimer,as suggested by the comment that citedthe Pearson decision, would be anuntenable alternative. If companiescould avoid the time and expense ofcomplying with the new drugprovisions of the act merely by attachinga disclaimer to a disease treatment orprevention claim, the longstandingsystem of drug regulation in thiscountry would be eviscerated, withserious public health consequences.Nothing in Pearson requires such aresult. Indeed, the Pearson courtrecognized that its ruling did not applyto drugs (164 F.3d at 656 n. 6). Becausethe act classifies products on the basisof intended use, dietary supplementsthat make disease claims are drugs,unless the disease claim is also anauthorized health claim for which theproduct qualifies (see section 201(g)(1)of the act).

The Washington Legal Foundationdecision is not to the contrary. That caseinvolved the dissemination ofinformation on ‘‘off-label’’ (unapproved)uses for approved drugs and devices tophysicians by means of scientific andeducational symposia, reprints, andtextbooks. The U.S. District Court forthe District of Columbia held certainFDA guidance documents that describedacceptable ways of disseminating suchinformation unconstitutional under theCentral Hudson test. While recognizingthe substantial government interest inhaving off-label uses for drugs anddevices found to be safe and effective byFDA, the court held that the guidancedocuments violated the FirstAmendment because it believed thatthey ‘‘restricted’’ speech in a mannerthat was more extensive than necessaryto further that interest. (See 13 F. Supp.

2d at 73.) (Subsequent to the 1998decision cited by the comments, thecourt rendered another decision adverseto FDA (Washington Legal Foundationv. Henney, 1999 WL 557679 (D.D.C. July28, 1999)). That decision concerned theconstitutionality of certain provisions ofthe FDA Modernization Act of 1997involving the same subject matter as theguidance documents, and the court’sFirst Amendment rationale was similarto its rationale in the 1998 decisionpertaining to the guidance documents.)

FDA disagrees with the district courtdecision in Washington LegalFoundation and has appealed. In anyevent, however, the outcome inWashington Legal Foundation does notdetermine the outcome here for severalreasons. First, in Washington LegalFoundation the court found a lessrestrictive alternative that it concludedwould more precisely address thegovernment’s regulatory concerns:Requiring manufacturers whodisseminate information about off-labeluses to physicians through scientificreprints or educational symposia todisclose: (1) Their interest in drugs ordevices that are the subject of suchactivities, and (2) the fact that the usediscussed has not been approved byFDA. Here, as explained previously,there are no less restrictive alternativesto this rule that would further thegovernment’s substantial public healthinterest. Second, in Washington LegalFoundation physicians were theintended audience of the commercialspeech at issue. In contrast, consumersare the primary audience for dietarysupplement labeling. Although themarketplace includes consumers ofvarying levels of sophistication, theaverage consumer does not possess themedical and scientific expertisenecessary to evaluate claims about theeffect of a product on disease. (SeeAmerican Home Products Corp. v. FTC,695 F.2d 681, 698 (3d Cir. 1983);Association of Nat’l Advertisers, Inc. v.Lungren, 44 F.3d 726, 733–34 (9th Cir.1994), cert. denied, 516 U.S. 812(1995).) Finally, in Washington LegalFoundation, it was undisputed that theproducts involved were drugs (or, insome cases, devices) to be used intreating or preventing disease. Incontrast, the purpose of this rule is todistinguish between products that areintended to affect disease and productsthat are not.

The agency does not believe this ruleis premature in light of the need toreassess the regulatory regime for healthclaims under Pearson. Since healthclaims and structure/function claims areregulated separately, there is no need towait for any post-Pearson changes for

health claims to be complete beforeproceeding with this rulemaking onstructure/function claims. Moreover,since the agency has decided not toamend the health claims regulations aspart of this rulemaking, there is nopotential conflict between the two.

The First Amendment issues raised incomments on § 101.93(g)(4)(iii)(proposed § 101.93(g)(2)(iv)(C)),concerning citations to scientificreferences in labeling, are not differentfrom those raised by comments on therule as a whole and are addressed in thepreceding analysis. FDA also notes that,as discussed elsewhere in thisdocument, § 101.93(g)(4)(iii) has beenrevised to narrow the circumstancesunder which the agency will considercitations to scientific references inlabeling to be disease claims.

(109.) Another comment furtherasserted that the prohibition againstimplied disease claims violates the FirstAmendment because it does notadvance the safety of dietarysupplements. The commentacknowledged that some dietarysupplements ‘‘may present serioussafety risks,’’ but said ‘‘these risks willnot be lessened by prohibiting truthful,non-misleading structure/functionclaims * * *.’’ The comment suggestedthat other provisions in DSHEA addressthe safety of dietary supplements andthat FDA can bring an enforcementaction if it has safety concerns.

FDA agrees with this comment in partand disagrees in part. The agency agreesthat prohibiting truthful, non-misleading structure/function claimswould not lessen the safety risks posedby some dietary supplements. The ruleis aimed at the safety risks posed byunapproved drug claims andunauthorized health claims on dietarysupplements. Unproven disease claimson a product marketed as a dietarysupplement may induce consumers totreat themselves with the supplementinstead of seeking treatments that areknown to be effective. Such claims mayalso dissuade consumers from seeing adoctor. These are very real safety risks.To the extent that safety risks are causedby the composition of a dietarysupplement rather than by claims madefor it, the agency agrees that otherprovisions in DSHEA and the act are theappropriate remedy.2. Equal Protection

(110.) One comment claimed the ruleviolates the equal protection clause ofthe Fourteenth Amendment because itsupposedly gives more protection to the‘‘labeling rights and speech’’ ofpharmaceutical manufacturers than todietary supplement manufacturers.



First, it should be noted that the equalprotection clause of the FourteenthAmendment applies only to the States,not to the Federal Government.However, the due process clause of theFifth Amendment contains an equalprotection component that is equivalentto the equal protection clause of theFourteenth Amendment (Schweiker v.Wilson, 450 U.S. 221, 226 & n. 6 (1981)).Even if the comment is interpreted torefer to equal protection under the FifthAmendment, FDA disagrees with it.First, the comment does not explain inwhat manner the rule gives moreprotection to the labeling rights andspeech of pharmaceutical manufacturersthan to those of dietary supplementmanufacturers. Second, even if the ruledoes treat these two classes ofmanufacturers differently, treatingdifferent regulated groups differentlydoes not in itself violate the equalprotection clause. Unless a regulatoryclassification jeopardizes the exercise ofa fundamental right or classifies uponinherently suspect grounds such as raceor religion, it is subject to the leastexacting form of equal protectionreview: Whether the classification itdraws bears a rational relationship to alegitimate government interest. (SeeNordlinger v. Hahn, 505 U.S. 1, 10(1992).)

This rule neither jeopardizes theexercise of a fundamental right norcreates a suspect classification. Thepurpose of the rule is to clarify thestatutory distinction between productsthat are intended for use in treating orpreventing disease and products that areintended for use in affecting thestructure or function of the body.Products intended to treat or preventdisease are subject to regulation asdrugs, unless they qualify for anauthorized health claim. Productsintended to affect the structure orfunction of the body may be regulatedas dietary supplements, subject tocertain conditions. Products regulatedas drugs must meet strict requirementsfor a premarket demonstration of safetyand efficacy (see sections 201(p) and505 of the act); these requirements donot apply to dietary supplements. Thedistinction that the statute and this ruledraw between products that areintended to have an effect on diseaseand those that are intended only toaffect the structure or function of thebody is clearly rationally related to thelegitimate government interest ofensuring that products intended to havean effect on a disease are safe andeffective for that intended use.3. Takings Under the Fifth Amendment

(111.) Several comments claimed thatthe proposal violates the Takings Clause

of the Fifth Amendment because itwould prohibit the use of specific termsthat now appear in product names,trademarks, trade names, symbols, andcompany logos, or would harmcompanies that use such terms in theircorporate names. One comment saidFDA must provide compensation foreach taking, but that the proposal failedto do so.

FDA disagrees with these comments.The Takings Clause forbids thegovernment from taking privateproperty for public use without justcompensation. However, FDA believesthat no taking will occur as a result ofthis rule.

The first issue to be considered iswhether the categories of names, words,and symbols identified in the commentson this issue are property within themeaning of the Takings Clause. TheConstitution itself does not define whatqualifies as property. Rather, ‘‘existingrules or understandings derived from anindependent source,’’ such as State orFederal law, define the interests thatqualify for protection as property underthe Fifth Amendment (Lucas v. SouthCarolina Coastal Council, 505 U.S.1003, 1030 (1992)).

The categories of names, words, andsymbols mentioned by the commentsare intangible property interests. Asdiscussed below, trademarks and tradenames are property to the extent thatthey are associated with businessgoodwill. A trademark is a word, name,symbol, device, or combination thereofthat a person uses, or intends to use andhas applied to register, to identify anddistinguish his or her goods from otherson the market and to indicate theirsource (15 U.S.C. 1127). A trade nameis the name a person uses to identify hisor her business (15 U.S.C. 1127) andmay include corporate, partnership, andother names. Symbols and logos, whenused to identify a product or company,may be property insofar as they aretrademarks or trade names. Likewise,product names may be property if theyare protected by a trademark or tradename. For brevity, in the remainder ofthis discussion the categories of names,words, and symbols mentioned by thecomments on the takings issue will bereferred to collectively as ‘‘trademarksand trade names.’’

Trademarks and trade names areproperty, but only insofar as they areassociated with the goodwill of anongoing business. (See American SteelFoundries v. Robertson, 269 U.S. 372,380 (1926).) They have no intrinsicvalue. The purpose of a trademark ortrade name is to prevent confusion withthe products of another manufacturer.(See United Drug Co. v. Theodore

Rectanus Co., 248 U.S. 90, 97 (1918).)Trademarks and trade names are givenlegal protection to prevent onemanufacturer from passing off its goodsas the goods of another and thus takingadvantage of the latter’s goodwill(American Steel Foundries, 269 U.S. at380; United Drug, 248 U.S. at 97).

The Supreme Court has declined toprescribe a ‘‘set formula’’ for identifyingtakings and instead has characterizedtakings analysis as an ‘‘essentially adhoc, factual’’ inquiry (Penn CentralTransp. Co. v. City of New York, 438U.S. 104, 124 (1978)). Nonetheless, theCourt has identified three factors forconsideration in assessing whether aregulatory taking has occurred: Thecharacter of the governmental action;the regulation’s economic impact; andthe extent to which the regulationinterferes with reasonable investment-backed expectations (Ruckelshaus v.Monsanto Co., 467 U.S. 986, 1005(1984)). The force of any one of thesefactors may be ‘‘so overwhelming * * *that it disposes of the taking question’’(Monsanto, 467 U.S. at 1005). Whenexamined in light of these three factors,the rule does not effect a compensabletaking under the Fifth Amendment.

a. The character of the governmentaction. With respect to the first factor,the character of the government action,courts are more likely to find a takingwhen the interference with property canbe characterized as a physical invasionby government than when theinterference is caused by a regulatoryprogram that ‘‘adjust[s] the benefits andburdens of economic life to promote thecommon good’’ (Penn Central, 438 U.S.at 124). The Supreme Court has heldthat, when a governmental action istaken in order to protect the publicinterest in health, safety, and welfare,this factor weighs heavily againstfinding a taking. (See KeystoneBituminous Coal Ass’n v. DeBenedictis,480 U.S. 470, 488 (1987).) Regulatoryactions taken to protect the publichealth are rarely, if ever, held toconstitute takings. (See Porter v.DiBlasio, 93 F.3d 301, 310 (7th Cir.1996) (action taken to protect publichealth falls within class of propertydeprivations for which FifthAmendment does not requirecompensation); Jarboe-Lackey Feedlots,Inc. v. United States, 7 Cl. Ct. 329 (1985)(seizure of adulterated meat not ataking).)

Although these regulations willrestrict the use of certain terms,including terms that appear in sometrademarks and trade names, thisrestriction does not rise to the level ofa taking. Governmental restrictions onthe uses individuals can make of their



property are ‘‘properly treated as part ofthe burden of common citizenship’’(Keystone, 480 U.S. at 491 (citationomitted)). These burdens are ‘‘borne tosecure ‘the advantage of living anddoing business in a civilizedcommunity’’’ (Andrus v. Allard, 444U.S. 51, 67 (1979) (quotingPennsylvania Coal Co. v. Mahon, 260U.S. 393, 422 (1922) (Brandeis, J.,dissenting)). Moreover, these regulationsare not without benefit tomanufacturers. (See Keystone, 480 U.S.at 491 (‘‘While each of us is burdenedsomewhat by such restrictions, we, inturn, benefit greatly from the restrictionsthat are placed on others.’’).) Theregulations will help ensure a levelplaying field in the dietary supplementmarket because no manufacturer will beable to make an implied disease claimwithout prior FDA review under thehealth claim or new drug standard.Previously, unreviewed implied diseaseclaims on dietary supplementsproliferated, in part because ofuncertainty about the line betweenstructure/function claims and diseaseclaims.

These regulations are rationallyrelated to, and substantially advance,FDA’s legitimate interest in promotingand protecting the public health byensuring the safety and efficacy ofproducts promoted for use in treating orpreventing disease. (See Keystone, 480U.S. 470 at 485; Monsanto, 467 U.S. at1007.) By clarifying that such productsmay not be marketed under thestructure/function claim regime, FDA isseeking to ensure that they are regulatedthrough the drug approval or healthclaims authorization process, asappropriate.

The effect of the regulations cannot becharacterized as a taking of property.Dietary supplement companies will notbe precluded from using terms thatimply a disease claim in theirtrademarks and trade names. If theywish to continue using trademarks andtrade names that imply a disease claim,they may do so, provided that they firstmeet the safety and efficacy standardsand other regulatory requirementsapplicable to drugs or, in appropriatecases, provided that they obtainauthorization to make a health claim.(As discussed below, only non-misleading trademarks and trade namesmay be used.)

Even if these regulations could be saidto prevent a business from using atrademark or trade name on its dietarysupplements, such a result still wouldnot constitute a taking of the trademarkor trade name. The purpose of givingtrademarks and trade names legalprotection is to prevent one

manufacturer from passing off its goodsas the goods of another (American SteelFoundries, 269 U.S. at 380). Thisregulation will not allow onemanufacturer to use another’s trademarkor trade name; rather, all manufacturerswill be precluded from usingtrademarks and trade names thatcontain an implied disease claim unlessthey have obtained new drug approvalor health claim authorization. Thus,manufacturers will not suffer anycompetitive injury.

Moreover, deprivation of a trademarkalone is not a deprivation of property.Because the trademark is ‘‘merely aprotection for the good will’’ (HanoverStar Milling Co. v. Metcalf, 240 U.S. 403,414 (1916)), only if a regulation takesthe owner’s goodwill as well would theregulation be a taking. It is not apparent,however, that these regulations willdeprive manufacturers of any goodwill.Manufacturers will be faced with achoice as to whether to change theirtrademark or trade name or to seekapproval for their products as drugs. Insome cases, they will also have a thirdoption: Seeking authorization to make ahealth claim. If they are able to obtaindrug approval for the intended usesuggested by the trademark or tradename, they will not have to change thetrademark or trade name, provided thatthe name is not confusingly similar tothe name of another drug or otherwisemisleading (see section 502(a)(1) of theact (21 U.S.C. 352(a)(1)); and§ 201.10(c)(3) and (c)(5).) Similarly, ifthey are able to obtain authorization tomake a health claim for the intendeduse suggested by the trademark or tradename, they will not have to change thetrademark or trade name unless it ismisleading. (See section 403(a)(1) of theact.) Even if a manufacturer chooses tochange its trade name or trademark, itwill not be deprived of the goodwillunderlying them but only of thatparticular symbol of the goodwill. Themanufacturer will still be able totransfer the goodwill associated with itsproducts to another trade name ortrademark.

Case law on the treatment of goodwillunder the Takings Clause supports theview that no taking will occur as a resultof these regulations. The general rule isthat the owner of a place of business towhich the government takes title is notentitled to compensation for loss ofgoodwill (United States v. GeneralMotors Corp., 323 U.S. 373, 379 (1945)).The reason for the rule is that thebusiness may reopen at another locationto which the goodwill may betransferred (Kimball Laundry Co. v.United States, 338 U.S. 1, 11–12 (1949)).Only where the government operates the

business, thereby depriving the ownerof its ‘‘going-concern value,’’ is there acompensable taking of goodwill. InKimball, the Supreme Court held thatthe government owed compensation forthe loss of goodwill associated with thetemporary taking of a laundry duringWorld War II. This action was held tobe a taking of goodwill because thegovernment not only physically took butalso operated the laundry during thewar (Kimball, 338 U.S. at 12–13). Thus,during the period that the governmentoperated the laundry, there was nobusiness to whose benefit the goodwillassociated with the private laundrybusiness could inure. Here, thegovernment is not taking any trademarkor trade name for its own use, nor is itshutting down the businesses that ownthem. Therefore, the goodwillsymbolized by the trademark or tradename will remain with these businesses.

Finally, although trademarks andtrade names can be property when theysymbolize and protect the goodwillassociated with a business, there can beno property interest in an illegalproduct. Dietary supplements that bearclaims to treat or prevent disease aremisbranded and are also unapprovednew drugs (unless the claim is anauthorized health claim). As such, theymay not legally be sold in interstatecommerce (see section 301 (a) and (d) ofthe act. There can be no taking of anillegal article. (See Meserey v. UnitedStates, 447 F. Supp. 548, 554 (D. Nev.1977) (‘‘Plaintiff has not been denied hisproperty. He is denied the right tointroduce his goods into commerceunless they are in compliance with the[Federal Food, Drug, and Cosmetic]Act.’’).) Moreover, it has always beenillegal to market dietary supplements orother foods with disease claims, exceptthat since 1990 the act has permittedauthorized health claims. Theseregulations merely clarify the linebetween acceptable structure/functionclaims and prohibited disease claims.(See Lucas, 505 U.S. at 1030 (‘‘The useof [property] for what are now expresslyprohibited purposes was alwaysunlawful, and * * * it was open to theState at any point to make theimplication of those backgroundprinciples of * * * law explicit’’ withoutpaying compensation) (emphasis inoriginal).) For this reason and the otherreasons previously discussed, the firstfactor of the takings analysis indicatesthat these regulations effect no takings.

b. The economic impact of thegovernment action. The second factor toconsider is the economic impact of thegovernment action. This impact is not tobe considered piecemeal by dividing aproperty interest ‘‘into discrete



segments and attempt[ing] to determinewhether rights in a particular segmenthave been entirely abrogated’’ (PennCentral, 438 U.S. at 130). The analysisinvolves looking not just at what hasbeen lost, but at the nature and extentof the interference with rights in theproperty as a whole. (See Penn Central,438 U.S. at 130–31; Andrus v. Allard,444 U.S. at 65–66.) Thus, here the totalimpact of the regulations on propertyrights should be considered, rather thanonly whether a business can or cannotcontinue to use a particular trademarkor trade name. It is clear that aregulation’s economic impact may begreat without rising to the level of ataking. (See Pace Resources, Inc. v.Shrewsbury Township, 808 F.2d 1023,1031 (3d Cir.), cert. denied, 482 U.S. 906(1987) (citing Hadacheck v. Sebastian,239 U.S. 394 (1915) (reduction in valuefrom $800,000 to $60,000); Euclid v.Ambler Realty Co., 272 U.S. 365 (1926)(75 percent diminution in value)).)

In assessing whether a regulationeffects a taking, the Supreme Court hasconsidered whether the regulationdenies an owner the ‘‘economicallyviable’’ use of its property. (See, e.g.,Keystone, 480 U.S. at 499.) Although itis undeniable that compliance withthese regulations will cost money andmay mean that certain trademarks andtrade names must be altered, companieswill not be denied the economicallyviable use of their property. Aspreviously discussed, some firms maybe able to obtain new drug approval orhealth claim authorization for thoseproducts that bear trademarks or tradenames that include disease claims. Ifapproved as new drugs or authorized tobear a health claim, in many cases theseproducts could continue to bear theoriginal trademark or trade name. Thisapproach would, however, require thecompany involved to make significantexpenditures of time and money tosubmit a new drug application (NDA) orhealth claim petition to FDA. Thefinancial burden required to complywith such requirements is not a takingunder these circumstances, however,just as it is not a taking to require othercompanies to comply with applicablerequirements before marketing a newdrug or a food bearing a health claim.Obtaining new drug approval orauthorization to make a health claimmay be costly, but it is not the kind ofeconomic impact that leads to a taking.‘‘Requiring money to be spent is not ataking of property’’ (Atlas Corp.) v.United States, 895 F.2d 745, 756 (Fed.Cir.), cert. denied, 498 U.S. 811 (1990)).

As previously noted in the discussionof the first factor of the takings analysis,case law indicates that the regulations

will cause no loss of goodwill even incases where a trademark or trade namemust be changed because new drugapproval or health claim authorizationcannot be obtained. Even if theregulations do cause a loss of goodwill,however, FDA believes that theeconomic impact of that loss of goodwillis outweighed in the takings analysis bylack of reasonable investment-backedexpectations in being able to makedisease claims in trademarks and tradenames.

c. Interference with reasonableinvestment-backed expectations. Thefinal factor to consider is whether acompany has a reasonable investment-backed expectation in continuing to usea trademark or trade name. To bereasonable, expectations must take intoaccount the power of the state toregulate in the public interest (PaceResources, 808 F.2d at 1033).Reasonable expectations must also takeinto account the regulatoryenvironment, including theforeseeability of changes in theregulatory scheme. ‘‘In an industry thatlong has been the focus of great publicconcern and significant governmentregulation,’’ Monsanto, 467 U.S. at 1008,the possibility is substantial that therewill be modifications of the regulatoryrequirements. ‘‘Those who do businessin the regulated field cannot object’’ ifthe regulatory scheme is ‘‘buttressed * ** to achieve the legislative end’’(Connolly v. Pension Benefit Guar.Corp., 475 U.S. 211, 227 (1986) (citationomitted)). The lack of a reasonableinvestment-backed expectation canoutweigh the other takings factors andbe determinative in whether a takinghas occurred (Monsanto, 467 U.S. at1005).

Companies that use trademarks ortrade names that include disease claimslack a reasonable investment-backedexpectation that they will be able tocontinue to use those trademarks andtrade names. First, the Supreme Courthas said that it is unreasonable to havehigh expectations in personal property(i.e., property other than land): ‘‘[I]n thecase of personal property, by reason ofthe State’s traditionally high degree ofcontrol over commercial dealings, [theproperty owner] ought to be aware ofthe possibility that new regulationmight even render his propertyeconomically worthless * * *.’’ (Lucas v.South Carolina Coastal Council, 505U.S. at 1027–28). Second, the dietarysupplement and drug industries are a‘‘focus of great public concern andsignificant government regulation’’(Monsanto, 467 U.S. at 1008). A productthat bears a disease claim, whether thatclaim appears in a trademark, trade

name, or elsewhere, has been subject toregulation as a drug since 1906, exceptthat since 1990 the act has permittedconventional foods and dietarysupplements to bear authorized healthclaims without drug approval. Since1938, drugs (with certain narrowexceptions) have been subject to apremarket approval requirement. Giventhis longstanding history of closeregulation, it cannot be reasonable for amanufacturer or distributor to expect tobe able to make disease claims withoutprior authorization from FDA.

Moreover, it has always been illegal tomarket dietary supplements or otherfoods with disease claims, except thatsince 1990 authorized health claimshave been permitted. These regulationsmerely clarify the line betweenacceptable structure/function claimsand prohibited disease claims. (SeeLucas, 505 U.S. at 1030 (‘‘The use of[property] for what are now expresslyprohibited purposes was alwaysunlawful, and * * * it was open to theState at any point to make theimplication of those backgroundprinciples of * * * law explicit.’’).)Companies in the dietary supplementindustry should have been aware thatFDA was likely to issue such aclarification, not only because of theregulatory environment generally butalso for several specific reasons. First,the passage of DSHEA, which addedsection 403(r)(6) to the act, created alikelihood that FDA would issueregulations ‘‘to achieve the legislativeend’’ of permitting structure/functionclaims without premarket review, whilecontinuing to prohibit disease claimslacking FDA authorization (seeConnolly, 475 U.S. at 227 (citationomitted)). Second, the Commission onDietary Supplement Labels specificallyencouraged FDA to clarify theappropriate scope of structure/functionstatements (Ref. to Commission report,p. 38). Third, the rapidly expandingdietary supplement market and theproliferation of implied disease claimsin labeling should have put the industryon notice that FDA might take action.

For all these reasons, there can be noreasonable investment-backedexpectations with respect to trademarksand trade names that include diseaseclaims. Thus, the third factor of thetakings analysis weighs strongly againstfinding a taking of property that requirescompensation under the FifthAmendment. Moreover, the threefactors, taken together, show that theseregulations do not effect such a taking.Therefore, FDA concludes that thecomments arguing the contrary areunpersuasive.



IV. Implementation PlanThe preamble to the proposed rule

discussed FDA’s tentative conclusionsregarding the effective date of a finalrule and the agency’s implementationplan. In general, the preamble to theproposed rule stated that a final rulewould become effective 30 days afterthe date of the final rule’s publication inthe Federal Register. Any product thatis marketed for the first time afterpublication of the final rule, and anynew claims made for an existingproduct for the first time after thepublication of the final rule, will beexpected to be in compliance beginning30 days after publication of the finalrule. However, small businesses thatmarketed a product as of the date ofpublication of a final rule would havehad an additional 17 months to bringexisting claims (i.e., claims already inthe products’s labeling on January 6,2000 for those products intocompliance, provided that the smallbusiness had notified FDA of the claimas required by section 403(r)(6) of theact and § 101.93(a) and that FDA hadnot objected to the claim. For all otherproducts that were on the market as ofthe date of publication of a final rule,FDA would have allowed an additional11 months beyond the effective date tobring existing claims for those productsinto compliance, provided that the firmhad notified FDA of the claim asrequired by section 403(r)(6) of the actand § 101.93(a) and that FDA had notobjected to the claim. Any productmarketed for the first time after the dateof publication of the final rule, and anynew claim made for an existing productfor the first time after publication of thefinal rule, would have been expected tobe in compliance beginning 30 daysafter the date of publication of a finalrule.

(112.) Two comments suggestedextending the compliance period to 6months after the date of publication ofa final rule. The comments alsoadvocated that there be no distinctionbetween large and small businesses forcompliance dates. The commentsfurther suggested that FDA givebusinesses whose products were on themarket as of the date of publication ofa final rule 15 months (instead of 11 or17 months) to comply. Anothercomment suggested that the final rulebecome effective 12 months, rather than30 days, after its publication date.

FDA believes that the proposedcompliance periods of 11 and 17months following the effective date ofthe final rule are reasonable and fair,and that the distinction between largeand small businesses is appropriate.

FDA has decided, however, that it willnot treat manufacturers who have notnotified the agency of their claimsdifferently from other manufacturers. Atleast some of those manufacturers whodid not submit 30-day notifications tothe agency may have failed to do sobelieving that notification was notnecessary under section 201(g)(1)(C) ofthe act. Therefore, all manufacturerswill have 11 months after the effectivedate of the final rule to come intocompliance, and small businesses willhave 17 months after the effective dateof the final rule. The agency believesthat these compliance periods,uniformly applied, are sufficiently longthat it is not necessary to extend theeffective date to 6 months afterpublication in the Federal Register.

For a limited transition period, FDAdoes not intend to take enforcementaction against firms who have relied onthe agency’s September 1997 preamblestatements to make a structure/functionclaim for a dietary supplement undersection 201(g)(1)(C) of the act. To allowa reasonable time for the necessary labelchanges, the transition period will lastuntil the applicable compliance date forthe rest of the rule; i.e., small businesseswill have 18 months from publication tocomply, and other firms will have 12months. As of the applicablecompliance date, firms that have beenmaking structure/function claims undersection 201(g)(1)(C) must either removethe claim or comply with therequirements of section 403(r)(6) of theact and § 101.93, including notifyingFDA of the claim and relabeling to addthe required disclaimer. New structure/function claims are not subject to thistransition period; any firm that makes astructure/function claim in the labelingof a dietary supplement after theeffective date of this rule must complywith section 403(r)(6) of the act and§ 101.93.

V. Environmental Impact

The agency has determined under 21CFR 25.30(h) and (k), that this action isof a type that does not individually orcumulatively have a significant effect onthe human environment. Therefore,neither an environmental assessmentnor an environmental impact statementis required.

VI. Analysis of Impacts

A. Background

FDA has examined the impacts of thefinal rule under Executive Order 12866,the Regulatory Flexibility Act (5 U.S.C.601–612), and the Unfunded MandatesReform Act of 1995 (Public Law 104–4).Executive Order 12866 directs agencies

to assess all costs and benefits ofavailable regulatory alternatives and,when regulation is necessary, to selectregulatory approaches that maximizenet benefits (including potentialeconomic, environmental, public healthand safety, and other advantages;distributive impacts; and equity). TheRegulatory Flexibility Act requiresagencies to examine the economicimpact of a rule on small entities. TheUnfunded Mandates Reform Actrequires agencies to prepare anassessment of anticipated costs andbenefits before enacting any rule thatmay result in an expenditure in any oneyear by State, local and tribalgovernments, in the aggregate, or by theprivate sector, of $100 million (adjustedannually for inflation).

FDA concludes that this final rule isconsistent with the principles set forthin the Executive Order and in these twostatutes. The agency has determinedthat the rule is a significant regulatoryaction as defined by the ExecutiveOrder, because it raises novel policyissues. FDA has further determined thatthe final rule may have a significanteconomic impact on a substantialnumber of small entities. This sectionconstitutes the agency’s final regulatoryflexibility analysis as required under theRegulatory Flexibility Act. Because thisrule imposes no mandates ongovernment entities and will not resultin private expenditures of $100 millionin any one year, the UnfundedMandates Reform Act does not requirethe agency to prepare a cost-benefitanalysis.

B. Benefits of the Labeling RequirementsThe primary purpose of the rule is to

provide a consistent standard fordistinguishing between claims that maybe made in labeling without priorreview by FDA and claims that requireprior authorization as health claims orprior review as drug claims. The largergoal is to ensure that information aboutnon-disease-related effects of a dietarysupplement on the body may be freelydisseminated in labeling, while at thesame time guaranteeing that claims foruse of a dietary supplement to treat orprevent disease are not made withoutprior review to ensure that thesupplement is safe and effective for thatuse.

Although dietary supplements canplay a valuable role in consumer health,the agency recognizes that, wheninappropriately labeled, they can poseunnecessary risks. Such risks arisewhen the product labeling: (1)Encourages consumers to self-treat for aserious disease without the benefit of amedical diagnosis, or to self-treat for a



serious disease by substituting a dietaryproduct of uncertain value for a medicaltherapy that has been shown to be safeand effective; (2) encourages consumersto feel sufficiently protected from aserious disease (e.g., cancer) that theydelay, or possibly forego, regularscreening or early medical attention thatmay be critical to improved odds ofpatient survival; or (3) increases the riskof adverse reactions due to interactionswith other chemical compounds (e.g.,prescription medications) taken by thepatient. As consumer spending ondietary supplements continues to rise,the need for an information standardthat minimizes these risks becomesmore acute.

The rule may also benefit consumersby encouraging manufacturers of dietarysupplements to develop the safety andeffectiveness data needed to support ahealth or drug claim. Where diseaseclaims can be made without thisdemonstration of safety andeffectiveness, product manufacturershave less incentive to develop thesubstantial documentation needed toreceive this agency authorization. Theavailability of additional products withauthorized health or drug claims wouldbe extremely useful to the manyconsumers who have difficultydistinguishing among the variety ofproducts now marketed for particularhealth concerns.

The dietary supplement industry hasgrown rapidly, with estimated sales in1996 of $10.4 billion for all dietarysupplements, including $4.9 billion forvitamins and $3.0 billion fornonprescription herbal products (Ref.8). FDA has limited information on thenumber of products and quantities sold,or on the age, gender, and disease statusof persons currently using dietarysupplements. However, a 1997 survey of43,000 households, conducted by theHartman and New Hope researchorganization, indicates thatapproximately 70 percent of allhouseholds reported using vitamins,minerals, or herbal supplements in thepast 6 months (Ref. 9). Among surveyrespondents, those under age 30accounted for only 8 percent of allhouseholds with a member usingdietary supplements; ages 30 to 39accounted for 21 percent, ages 40 to 49accounted for 22 percent, ages 50 to 59accounted for 18 percent, and ages 60 orolder accounted for 30 percent (Ref. 10).Although the oldest group of surveyrespondents were, on the whole, lessknowledgeable about individualproducts, they reported more regularproduct use and more use for specificconditions than younger respondents.

FDA anticipates, therefore, that thefinal rule will clarify the dividing linebetween acceptable structure/functionclaims and disease claims, and therebyreduce the number of inappropriatedisease claims in dietary supplementlabeling. The defined standard forstructure/function claims under section403(r)(6) of the act will help to avoidinstances of inappropriate substitutionof dietary products for timely diseasescreening or medical treatment, and ofadverse interactions orcontraindications of drug-supplementcombinations. In addition, the rule maypromote the development of data andinformation for the support of newhealth or drug claims. Although FDAcannot quantify these regulatorybenefits, the agency expects that thisstandard will positively support theeffective integration of dietarysupplements into consumers’ overallprograms of wellness and self-care.

C. Costs of Compliance

The costs to industry are the directcosts of compliance, which areprimarily the costs of the neededproduct relabeling; and the indirectcosts of compliance, which include thepotential loss of product sales due to theelimination of disease claims. Thefollowing section details the agency’scalculation of the direct costs ofcompliance. FDA has been unable,however, to estimate the extent of theindirect costs of this rule. As explainedbelow, the agency estimates that over800 dietary supplement products willneed to be relabeled due to this rule.The substitution of a valid structure/function claim for a disease claim may,in fact, lead to a decrease in the sale ofcertain products. The magnitude of thisimpact, however, is unknown, as mostfirms will replace the disease claim witha structure/function claim that appealsto many of the same consumers. It isalso possible that some firms will avoida potential drop in sales by developingthe safety and effectiveness data neededto obtain either a new drug approval orauthorization from FDA to make ahealth claim. The agency cannotquantify the probability of theseoccurrences, however, and no industrycomment includes such data.

1. Proposed RuleIn the preamble to the proposed rule

(63 FR 23624), FDA had projected thatthe direct costs of compliance wouldrange from $0.1 million to $8.5 million.This figure largely reflected agencyestimates of the average cost ofrelabeling a typical dietary supplementproduct multiplied by the number ofdietary supplement products that would

need to be relabeled to conform with theproposed criteria for structure/functionclaims. The cost categories includedadministrative, analytical, andinventory disposal activities.

FDA acknowledged that estimates ofthe number of dietary supplementproducts were approximate, butprojected that the proposed rule wouldcover about 29,000 products, with about75,000 distinct labels, or stock keepingunits (SKU’s). The agency alsoexplained that the rule would directlyaffect from 500 to 850 manufacturers ofdietary supplement products.

To estimate the lower-bound costs ofthe proposed rule, FDA assumed thatthe 2,300 notifications initially receivedfrom dietary supplement manufacturersadequately represented the number ofproducts with structure/function claims.The agency had already objected to 150notifications because they containedobvious disease claims, but identifiedan additional 60 notificationscontaining one or more claims thatmight not have met the newly proposedcriteria for structure/function claims.Consequently, FDA’s lower-bounddirect cost estimate included labelchanges for 60 dietary supplementproducts. The estimated administrative,redesign, and inventory lossesassociated with these 60 label changestotaled between $91,400 and $123,400.

FDA also presented an upper-bound$8.5 million estimate of the direct costsof the proposed rule, based on thelikelihood that many additional dietarysupplements are marketed withstructure/function claims. For thisestimate, the agency concluded thatabout 30 percent, or 22,500, of theestimated universe of 75,000 dietarysupplement labels contain structure-function claims. Assuming that theproportion of disease claims on alllabels containing structure/functionclaims equals the proportion of diseaseclaims in the 2,300 notificationscontaining structure/function claims,the agency calculated that up to 585labels (60/2,300 x 22,500) could need tobe changed if the proposed rule becamefinal. The higher costs of the upper-bound estimate resulted both from thesubstantially increased assumed numberof affected labels and from the impact ofthe significantly shorter complianceperiod (30 days) for manufacturers thathad not notified FDA of their structure/function claim by the publication dateof the final rule.

2. Final RuleA number of the comments submitted

in response to the proposed rulespecifically addressed FDA’s analysis ofcompliance costs. As a result, the



agency has altered several of its costassumptions. In addition, FDA hasadjusted its analysis to reflect themodified provisions of the final rule. Asdescribed below, the agency estimatesthe total direct costs of the final rule tobe about $3.73 million, but presentssensitivity analysis to indicate that thecosts could rise to as much as $10.35million under certain worst-caseassumptions.

Although several industry commentssuggested that FDA had underestimatedthe costs of relabeling, no commentsobjected to the specific elements thatwere considered, i.e., administrative,redesign, and inventory disposalactivities. In response, FDA has retainedthis format for its analysis of the finalrule. One comment claimed that FDAhad underestimated the number ofproducts that would be affected, butprovided no evidence or basis fordetermining a more accurate count.Another comment stated that theagency’s cost estimates were not wellexplained and that all assumptions werenot disclosed. Consequently, FDA hasrevised its analysis to; (1) Simplify thecost-estimating methodology, (2) clearlypresent and describe each assumption,(3) fully explain the derivation of theestimated direct costs of compliance,and (4) conduct sensitivity analysis forthe remaining areas of significantuncertainty.

a. Cost of designing new labels.Dietary supplements will no longer beable to make claims whose status waspreviously unclear, but which now havebeen defined as disease claims. Firmsmay comply either by obtaining newdrug approval, by receivingauthorization from FDA to make ahealth claim, or by revising theirproduct labeling to eliminate diseaseclaims. Because the cost of submittingadequate documentation to obtain newdrug approval or health claimauthorization far exceeds the cost ofmodifying a label, this analysis assumesthat the direct costs of the rule will bethe costs of modifying labels withdisease claims. As explained above,FDA recognizes that some firms maychoose to obtain health claimauthorization or new drug approval asan alternative means of compliance, orto improve the marketability of theirproducts. The agency believes, however,that it is unlikely that the rule would bethe determining factor in a large numberof instances.

No public comments providedalternative estimates of the number ofaffected dietary supplement products.As noted above, FDA had estimated thatthe industry markets approximately29,000 covered products with about

75,000 distinct labels. The agency hasused this estimate for its analyses ofdietary supplement rules over the pastseveral years (e.g., 60 FR 67211December 28, 1995) and has received noindication from industry that betterestimates were available. Although theagency’s preliminary analysis reportedthat an estimated 30 percent of theproducts (8,700) carry structure/function claims, more recent data froma random survey conducted for FDA byRTI of about 3,000 dietary supplementproducts indicates that this percentagemay have been too low (Ref. 11).Although RTI notes that the surveyedsample is too small to supportquantitative inferences for thepopulation of dietary supplements, FDAfinds the data to be the best available.The RTI report actually shows that 69percent of the products in its samplehave claims, but this percentageincludes ‘‘diet supplementation’’claims. When adjusted to exclude ‘‘dietsupplementation’’ only 62 percent of theproducts in the RTI data base includerelevant claims. Even this 62 percentfigure is too high, however, because RTIover-sampled herbal products, whichhave a higher probablity of claims andwould not exceed 60 percent and hasused this figure as its final estimate.

Of the first 2,300 notifications ofstructure/function claims reviewed byFDA, no more than 60, or 2.6 percent ofthe products with claims, would haveneeded labeling changes due to thecriteria described in the proposed rule.Since that time, the total number ofnotifications with structure/functionclaims submitted to the agency hasincreased to about 5,200. A subsequentreview of all of the submitted claimsindicates that the final rule couldrequire about 1.9 times as many labelmodifications as the proposed rule,owing largely to the revised criteria forcholesterol claims in the final rule. FDAestimates that the final rule may requirerevised labels for about 4.81 percent ofthe 17,400 dietary supplement products(29,000 x 60 percent currently estimatedas marketed with structure/functionclaims (Refs. 15 and 16). (Excludingcholesterol claims would reduce thisfigure to 1.74 percent of the productswith claims.)

The resulting label cost calculationsare straightforward. First, the agencyfound that revised labels (for all claimsincluding cholesterol) may be neededfor approximately 837 products (17,400products with claims x 4.81 percent).Because each product may containroughly 2.6 distinct SKU’s (75,000SKU’s ÷ 29,000 products), labels for anestimated 2,164 SKU’s may need to bemodified (837 products x 2.6 SKU’s/

product). As described in its earlieranalysis, based on an average of theestimates provided in comments toearlier rules, FDA determined that theaverage label redesign cost is about$1,700 per dietary supplement SKU fora 12-month compliance period, and$1,300 for an 18-month complianceperiod. No industry commentquestioned the reasonableness of theseunit cost estimates.

The final rule sets compliance periodsof 1 year for large firms (revenues above$20 million) and 18 months for smallfirms (revenues below $20 million),except that new claims (i.e., claims notmade before the publication of the finalrule) must be in compliance as of theeffective date. Such claims will notnecessitate relabeling, however. FDAdoes not know the size of the firms thatwill need to make label changes. RTI(Ref. 12) reports that 95 percent of thefirms in the industry are small, but thatthe 5 percent that are large account for80 percent of industry sales. The RTIproduct data base also indicates thatapproximately 25 percent of the sampleproducts were manufactured by just 5percent of the companies. Thus, FDAhas assumed that approximately one-quarter of the affected products willcome from large firms and three-quarters from small firms.Consequently, the total estimated labelredesign costs equal about $3.03 million(i.e., $1,700 x 0.25 x 2,164 SKU’s +$1,300 x 0.75 x 2,164 SKU’s).

b. Administrative costs. One industrycomment contended that FDA had notadequately explained the basis for itscompany-specific administrative costs,estimated at $425 and $320 respectively,for 12-month and 18-month complianceperiods. These figures were derivedfrom data presented in a 1991 RTI reporton the cost of FDA’s food labelingregulations (Ref. 13). They includedcosts associated with interpreting aregulation, determining the manner ofcompliance and managing thecompliance method. RTI had estimatedthat, on average, small firms would bearadministrative costs of $850 to complywith the new food labeling rules for a1-year compliance period, and $650 fora 2-year compliance period. For itsanalysis of the proposed rule, FDAreduced this figure by fifty percent,based on the smaller administrativeeffort that would be needed to complywith the proposed rule, compared to theconventional food labeling regulationsevaluated by RTI in 1991. Theregulations that were the subject of the1991 RTI evaluation involved a broaderrange of administrative options andtasks, such as nutritional testing and



product reformulation. (The $320estimate for the 18-month complianceperiod was determined by interpolatingbetween the estimates for 12 and 24months.) The agency has raised thesecosts by about 27 percent to $540 and$407, respectively, to account for salaryinflation since 1991 (Ref. 14).

FDA had initially estimated that 500to 850 firms manufacture dietarysupplements. The recent RTI study,however, has identified 1,050manufacturers (Ref. 12). This highernumber probably overestimates the sizeof the industry covered by this rule,because it includes homeopathicproducts, which are drugs by statutorydefinition, and ‘‘functional foods’’ andsports nutrition products, which may beeither conventional foods or dietarysupplements depending on how theyare marketed and used. For this finalanalysis, FDA has assumed that 1,000companies manufacture the dietarysupplement products covered by thisrule. Although only a small fraction ofthese establishments will need toimplement changes in labeling due tothis rule, the agency anticipates mostfirms will review the final rule to assesswhether their labeling will be affected.

The administrative costs of the finalrule would likely be higher for thosefirms that will need to revise labels andlower for those firms that do not.Nevertheless, FDA assumes that, onaverage, all large dietary supplementmanufacturers would incur costs of$540 and all small dietary supplementmanufacturers would incur costs of$407. As noted above, RTI found thatabout 95 percent of the firms in thisindustry are small. Thus, the agencycalculated administrative costs to equalabout $413,000 (i.e., 950 small firms x$407 + 50 large firms x $540). FDAnotes that these estimates may overstatethe incremental administrative costs ofthis final rule, because dietarysupplement firms must already complywith DSHEA and this rule is meant toclarify the meaning of that act, ratherthan to add new requirements.Nevertheless, the agency’s sensitivityanalysis, presented below, doubles theabove cost estimates.

c. Costs of inventory losses. The finalcost component involves the value oflost inventory. FDA’s preliminaryanalysis relied on information from anearlier nutrition labeling rule thataffected the entire dietary supplementindustry. That information indicatedthat inventory disposal costs for theentire industry would be about $8million for an 18-month complianceperiod and $15 million for a 12-monthcompliance period. As explained above,FDA estimated that about 2.89 percent

of the dietary supplement products willrequire new labels as a result of this rule(837 ÷ 29,000) and that about threequarters of the affected products aremanufactured by small firms. Thus,total inventory disposal costs arecalculated at $281,000 (i.e., $8 millionx 2.89 percent x 0.75 + $15 million x2.89 percent x 0.25).

d. Total direct compliance costs. Asdescribed above, FDA has assumed thedirect compliance costs of this rule to bethe costs associated with relabelingthose dietary supplements whoselabeling claims are considered diseaseclaims under the newly defined criteria.Redesign costs are estimated at $3.03million, administrative costs at$413,000, and inventory disposal costsat $281,000. In sum, therefore, the totalestimated direct compliance costs equalalmost $3.73 million.

In addition, there may be costsassociated with the discussion in thefinal rule concerning structure/functionclaims made under section 201(g)(1)(C)of the act. (See response to comment 95in section III.A.1 of this document.) Theagency believes that some firms havebeen making structure/function claimsfor dietary supplements withoutincluding a disclaimer statement ornotifying FDA, based on FDA’sstatements in a 1997 preamble (62 FR49859 at 49860, 49863, and 49864).Because the agency has not repudiatedthese statements, any firm that hasrelied on them to make a claim for adietary supplement will need to add thedisclaimer to all applicable labels, aswell as to notify FDA, according to therequirements of this section 403(r)(6) ofthe act and § 101.93. Because firmsmaking such claims have not identifiedthemselves to FDA, the agency does nothave a reliable database on which tobase a cost estimate of the number offirms and products that may incur coststo comply with this new provision.

The costs to industry of the final ruleare substantially different from the costsof the proposed rule, because of twoimportant changes to the proposedrequirements. First, the final rulerequires more product labels to bechanged, because it includes morespecific parameters for acceptablestructure/function claims aboutcholesterol. This change increases thedirect compliance costs of the final rule.Second, the proposed rule requiredneeded label modifications to becompleted within 30 days afterpublication of the final rule, for thoseproducts without a properly submittedclaim notification. Roughly 70 percentof all products with claims may havefallen into this group (1–5,200 productswithout notifications ÷ 17,400 products

with claims). Because relabeling costsare reported to double for each halvingof the compliance period, compliancecosts would have been eight timesgreater for those products. For the finalrule, all large firms will be expected tocomply within 12 months, and all smallfirms within 18 months, regardless ofwhether the firm has notified FDA ofthe structure/function claims on itsproducts. This change significantlyreduces the direct compliance costs ofthe final rule.

e. Sensitivity analysis. Due touncertainty with respect to severalfactors in the agency’s direct cost model,FDA has prepared a sensitivity analysisof other possible cost scenarios. First,FDA tripled the percentage of productnotifications assumed to be out ofcompliance with the new criteria forstructure/function claims. This changeresults in almost tripling the total directcompliance costs of the regulation,raising the estimate from about 3.73million to about 10.35 to about $5.93million. Second, FDA doubled itsestimate of administrative costs. Thischange raises the inital cost estimate toabout $4.14 million. Changing bothassumptions simultaneously raises thetotal estimated costs to about $11million. Finally, under the initialscenario, if all of the needed labelchanges were assumed to affect onlysmall businesses, the total cost estimaterises to about $3.46 million. Thissensitivity analysis indicates that thetotal direct costs of this rule would notimpose a major burden on this industryeven if the most uncertain cost factorsare doubled or tripled from FDA’s bestestimates.

D. Other Industry Comments

Several comments insisted that FDAhad not conducted a comprehensivecost-benefit analysis of the proposedrule, as required under Executive Order12866. These comments stated thatFDA’s economic analysis ignored boththe potential savings in consumer healthcare expenditures that would be lost byrestricting important labelinginformation, as well as the likelynegative effect of the proposal on thegrowth of the dietary supplementindustry. One industry comment, forexample, declared that a substantivecost-benefit analysis ‘‘must identify thepotential health benefits that are lost asa consequence of reduced consumeraccess to useful information about thehealth-related properties of dietarysupplements and ingredients.’’ It notedthat FDA’s analysis ‘‘fails to considerthe public health benefits associatedwith ingesting dietary supplements as



well as the losses to public health thatcould result from consumers failing totake appropriate dietary supplementsdue to uninformative structure/functionclaims.’’ That comment also maintainsthat ‘‘FDA’s failure to assess andconsider such benefits (and costs)stands in contrast with the specificfinding of DSHEA that ‘appropriate useof safe nutritional supplements willlimit the incidence of chronic diseases,and reduce long-term health careexpenditures’.’’ The comment alsopoints out that FDA has performed suchanalyses in other rulemakings, e.g.,tobacco, nutrition labeling, and ephedraregulations.

FDA disagrees. Although ExecutiveOrder 12866 directs agencies to assessthe costs and benefits of economicallysignificant rules, the quantification ofthese expected costs and benefits isrequired only ‘‘to the extent feasible’’(58 FR 51735 at 51741, October 14,1993). As described above, FDA believesthat its final rule strikes the appropriatebalance with respect to health-relatedclaims in dietary supplement labeling.The rule classifies certain claims asacceptable structure/function claimsthat may be made without prior FDAreview. Although the provision ofstructure/function information toconsumers may reduce health careexpenditures, no health organization,industry association, or any otherinterested public or private group haspresented information or data thatwould allow the agency to develop aquantifiable estimate of the health carebenefits. The rule classifies other claimsas disease claims that are subject toexisting requirements for new drugapproval or health claim authorizationbefore a product may be marketed withthe claim. FDA believes that classifyingclaims into a category that requires FDAreview of safety and efficacy evidence,where appropriate, will similarly reducelong-term health care expenditures.Again, however, the agency has nomeans of quantifying the probablehealth outcomes of this aspect of therule and therefore has no means ofquantifying its impact on health careexpenditures. Because this analysisdiscusses the types of benefits and costsreasonably expected, and quantifiesthose that can be ‘‘feasibly’’ quantified,the agency has, in fact, complied withthe direction of Executive Order 12866.

FDA has attempted to quantify thebenefits of some of its previousregulations. The agency’s estimatedbenefits of the tobacco rule relied on awidely established risk assessmentpublished by the American CancerSociety. Estimated benefits of theproposed ephedra rule were based on

incidents identified in the agency’sadverse event database. Estimatedbenefits of the nutrition labeling rulewere derived from epidemiologicalstudies of the consequences of dietaryfat. In each case, the agency believedthat it had a reasonably reliable database upon which to base conclusions,and each risk assessment dealt with therisks of a single substance (tobacco,ephedra, and dietary fat). In contrast,this structure/function rule governsstructure/function claims in the labelingfor all dietary supplements. Althoughthe agency could conceivably analyze afew of the claims covered by the rule,adequate data on the benefits and risksof most of these products are notavailable. Consequently, the agencybelieves that this rule will improve thenation’s health, but concludes that itcannot feasibly quantify the effects ofthe rule on the nation’s healthexpenditures.

One industry comment suggested thatthe regulatory system could impedefirms from conducting research tosubstantiate structure/function claims, ifDSHEA is construed so narrowly that itexcludes meaningful health-relatedbenefits. This comment noted, however,that the absence of an enforceable legalstandard for substantiation woulddiscriminate against companies that doresearch to support their claims andwould deter science-based companiesfrom entering the market. Similarly, apatient organization and severalpharmaceutical companies expressedconcern that the rule would permitsome products to escape regulation asdrugs and therefore diminish incentivesfor the costly clinical researchconducted by pharmaceuticalcompanies and academic scientists.

As stated previously in the document,FDA is not aware of any evidence thatwould indicate that the establishment ofcriteria for distinguishing structure/function claims from disease claims willadversely affect the conduct of scientificresearch. In fact, FDA believes that thefinal rule accords with the intent ofDSHEA in promoting the enhancementsto consumer health expected from thebroad dissemination of structure/function information, while reducingthe risks to consumer health associatedwith the promotion of disease treatmentand/or prevention uses for productswhose safety and efficacy have not beendemonstrated.

E. Regulatory AlternativesFDA has considered several major

alternatives to the proposed rule as partof the rulemaking process. Theseinclude: (1) Taking no new regulatoryaction; (2) treating a statement about a

dietary supplement as a disease claimonly if the statement included anexpress reference to a specific disease;and (3) treating a statement about adietary supplement as a disease claim ifthe statement mentions an abnormalityof the structure or function of the body,even if the abnormality was notcharacterized by a set of signs orsymptoms recognized as the disease.These alternatives are fully discussed inthe preamble to the proposed rule (63FR 23624 at 23630) and alternative (2)is also discussed extensively in sectionII.E of this document. In brief, FDAfinds that the public comment does notinclude evidence or argumentssufficient to persuade the agency tosupport these alternatives.

Within the broad framework of thefinal rule, FDA weighed other policychanges that could affect thecompliance costs. One option wouldhave set the compliance period for allfirms at 6 months and another at 12months from the publication date of thefinal rule. Other options would haveextended the compliance period beyond18 months for small businesses, orcompletely exempted small businessesfrom the rule. Finally, the proposed rulewould have permitted firms 12 or 18months to comply, depending onwhether they were large or small firms;but only if they had submitted timelynotifications of their structure/functionclaims to FDA and FDA had notobjected to the claims. Other firms hadonly a 30-day compliance period.

Based on its model of food labelingcosts, FDA assumes that compliancecosts double for each halving of thecompliance period (Ref. 13). Thus, thefirst option, which set a 6-monthcompliance date for all firms, results inaverage relabeling costs twice as high asthat of the 12-month compliance period.FDA decided that this additional burdenwas not warranted. The option of a 12-month compliance period for small aswell as large firms was rejected becauseof the additional burden to small firms,which may find it more difficult toeffect rapid shifts in labelingprocedures. The final rule providessmall firms with an additional 6 monthsto introduce these labeling changes.Extending the compliance date for smallfirms beyond 18 months was rejected,because the agency did not believe thatthe delayed consumer benefits would bebalanced by the relatively modestadditional cost saving. Exempting allsmall firms was not acceptable, becausemost firms covered by this rule aresmall. The final option, which was toinclude the compliance periodsspecified in the proposed rule, requiredlabel changes within 30 days for



products bearing claims of which FDAhad not been notified or claims to whichFDA had already objected. This optionwas rejected because it could haveincreased costs per label for many smallfirms by a factor of eight.

F. Small Business Impacts

As stated above, the RegulatoryFlexibility Act requires agencies toanalyze regulatory options that wouldminimize any significant impact of arule on small entities, unless the rule isnot expected to have a significanteconomic impact on a substantialnumber of small entities. With this finalrule, FDA is defining the types ofstatements that can be made concerningthe effect of a dietary supplement on thestructure or function of the body. It alsoestablishes criteria for determiningwhen a statement represents a claim todiagnose, cure, mitigate, treat, orprevent disease and thus is notacceptable as a structure/function claim.The regulation was prepared inresponse to the dietary supplementindustry’s request for clarification fromFDA with respect to the distinctionbetween structure/function and diseaseclaims, and to guidance in theCommission report suggesting that FDAprovide such clarification to industry.

For its analysis of the proposed rule,FDA had estimated that between 500and 850 firms were involved in dietarysupplement manufacturing. A morerecent industry survey reports that 1,050companies manufacture dietarysupplements; although as explainedabove, some of these companies may

manufacture products not covered bythis rule. FDA has projected theindustry size for this rule at about 1,000firms. The Small BusinessAdministration (SBA) has determinedthat dietary supplement manufacturerswith fewer than 500 employees aresmall businesses. Because most datasources characterize firms in thisindustry by sales revenues rather thanemployment size, and because companyrevenues of less than $20 millioncorrelate reasonably well with a 500employee threshold, FDA has receivedapproval from the SBA to use a less-than-$20 million sales revenue standardto represent small dietary supplementmanufacturers. Table 1 displays thereported size distribution of the dietarysupplement manufacturing industry.

As described above, FDA assumesthat all small manufacturers of dietarysupplements will incur administrativecosts of about $407 per firm. Inaddition, a number of smallmanufacturers of dietary supplementswill need to alter some product labels,at an average redesign cost of about$1,300 per SKU, and an averageinventory cost of about $107 per SKU.FDA further analyzed the dietarysupplement product data base describedin the October 1999 RTI report (Ref. 11)to determine how these products may bedistributed among small businesses. Asnoted earlier, FDA estimates that about628 of the 837 products (75 percent)needing revised labels due to this ruleare manufactured by small firms. Ifthese 628 products were randomlydistributed among the 950 small

businesses, less than 0.1 percent of thesmall firms (1 firm) would be likely tohave more than 4 of these products andonly about 3 percent (30 firms) to havemore than 2 of these products.

A small firm that needs to redesignlabels for three products (about eightSKU’s) due to the rule will incurestimated one-time direct compliancecosts of about $11,650. A small firm thatneeds to redesign labels for 4 products(about 10 SKU’s) would incur costs ofabout $14,950, or roughly 1.2 percent ofaverage company revenue. Thus, theassumption that these products arerandomly distributed among small firmsindicates that very few small businesseswould be likely to incur relabeling coststhat are greater than 1 percent of averagesmall company revenue. It is possible,however, that some firms will have adisproportionate number of labels to berevised. In the RTI database of 3,000randomly selected products, only 3companies (all large) have more than 24products. Although the data basesample show a number of smallcompanies with up to 24 products, it isvery unlikely that all of these productlabels would need to be changed due tothis rule. If a small company needed torevise 10 products, however, its directcosts of compliance would be about$37,000. Moreover, although FDAcannot quantify the likelihood, somesmall firms could lose product sales dueto the necessary removal of a diseaseclaim from a product label. Thus, FDAfinds that this rule may have asignificant economic impact on asubstantial number of small companies.

TABLE 1.—ESTIMATED NUMBER OF DIETARY SUPPLEMENT MANUFACTURERS AND REVENUES, BY SIZE CATEGORY 1

Size Category Number of Companies Revenues ($ in billions) Percentage of Market

>$100 million 16 3.32 55%$20 to $100 million 38 1.54 25%<$20 million 996 1.19 20%Total 1,050 6.05 100%

1 Research Triangle Institute, ‘‘Economic Characterization of the Dietary Supplement Industry,’’ March 1999, pp. 5–15.

VII. Paperwork Reduction Act of 1995This final rule contains no collections

of information. Therefore, clearance bythe Office of Management and Budgetunder the Paperwork Reduction Act of1995 is not required.

VIII. ReferencesThe following references have been

placed on display in the DocketsManagement Branch (address above)and may be seen by interested personsbetween 9 a.m. and 4 p.m., Mondaythrough Friday.

1. Commission on Dietary SupplementLabels, Report to the President, Congress,

and the Secretary of the Department ofHealth and Human Services, November 1997.

2. Dorland’s Illustrated Medical Dictionary,28th ed., W. B. Saunders Co., Philadelphia,p. 478, 1994.

3. Stedman’s Medical Dictionary, 26th ed.,Williams & Wilkins, Baltimore, p. 492, 1995.

4. The Encyclopedia Americana,International Edition, Grolier Inc., Danbury,p. 168, 1985.

5. Black’s Law Dictionary, 6th ed., WestPublishing Co., St. Paul, p. 467, 1990.

6. The Merck Manual, 17th ed., MerckResearch Laboratories, Whitehouse Station,NJ, p. 416, 1999.

7. Webster’s Encyclopedic UnabridgedDictionary, p. 1057, 1989.

8. Economic Characterization of theDietary Supplement Industry, prepared forDHHS/FDA/CFSAN by Research TriangleInstitute, Center for Economics Researchunder Contract No. 223–96–2290: Task Order3, Final Report, p. 5–2, March 1999.

9. Herb and Supplement Usage Nears 70Percent, Natural Foods Merchandiser,www.nfm-online.com/nfm—backs/Feb—98/herbusage.html.

10. Wyngate, P., Consumers NotSupplement Brand Savvy, Natural FoodsMerchandiser, www.nfm-online.com/nfm—backs/Mar—98/brandsavvy.html.

11. Dietary Supplement Sales Information,prepared for DHHS/FDA/CFSAN by ResearchTriangle Institute, Center for EconomicsResearch Under Contract No. 223–96–2290:



Task Order 4, Final Report, pp. 5–8, October1999.

12. Research Triangle Institute, EconomicCharacterization of the Dietary SupplementIndustry, p. 5–15.

13. Research Triangle Institute,‘‘Compliance Costs of Food LabelingRegulations,’’ prepared for CFSAN/FDA byRTI under Contract No. 223–87–2097, finalreport, pp. 5–3, 5–4, January 1991.

14. U.S. Department of Labor, Bureau ofLabor Statistics, BLS.

15. Memorandum from R. J. Moore, FDA,to file, review of notifications made pursuantto 21 U.S.C. 343(r)(6).

16. Memorandum from J. Lienesch, FDA, tofile, calculation of relabeling cost estimate forfinal rule on statements made for dietarysupplements concerning the effect of theproduction on the structure or function of thebody, December 22, 1999.

List of Subjects in 21 CFR Part 101

Food labeling, Nutrition, Reportingand recordkeeping requirements.

Therefore, under the Federal Food,Drug, and Cosmetic Act, the PublicHealth Service Act, and under authoritydelegated to the Commissioner of Foodand Drugs, 21 CFR part 101 is amendedas follows:

PART 101—FOOD LABELING

1. The authority citation for 21 CFRpart 101 continues to read as follows:

Authority: 15 U.S.C. 1453, 1454, 1455; 21U.S.C. 321, 331, 342, 343, 348, 371.

2. Section 101.93 is amended byrevising the section heading and byadding paragraphs (f) and (g) to read asfollows:

§ 101.93 Certain types of statements fordietary supplements.

* * * * *(f) Permitted structure/function

statements. Dietary supplement labelsor labeling may, subject to therequirements in paragraphs (a) through(e) of this section, bear statements thatdescribe the role of a nutrient or dietaryingredient intended to affect thestructure or function in humans or thatcharacterize the documentedmechanism by which a nutrient ordietary ingredient acts to maintain such

structure or function, provided thatsuch statements are not disease claimsunder paragraph (g) of this section. Ifthe label or labeling of a productmarketed as a dietary supplement bearsa disease claim as defined in paragraph(g) of this section, the product will besubject to regulation as a drug unless theclaim is an authorized health claim forwhich the product qualifies.

(g) Disease claims. (1) For purposes of21 U.S.C. 343(r)(6), a ‘‘disease’’ isdamage to an organ, part, structure, orsystem of the body such that it does notfunction properly (e.g., cardiovasculardisease), or a state of health leading tosuch dysfunctioning (e.g.,hypertension); except that diseasesresulting from essential nutrientdeficiencies (e.g., scurvy, pellagra) arenot included in this definition.

(2) FDA will find that a statementabout a product claims to diagnose,mitigate, treat, cure, or prevent disease(other than a classical nutrientdeficiency disease) under 21 U.S.C.343(r)(6) if it meets one or more of thecriteria listed below. These criteria arenot intended to classify as diseaseclaims statements that refer to theability of a product to maintain healthystructure or function, unless thestatement implies disease prevention ortreatment. In determining whether astatement is a disease claim under thesecriteria, FDA will consider the contextin which the claim is presented. Astatement claims to diagnose, mitigate,treat, cure, or prevent disease if itclaims, explicitly or implicitly, that theproduct:

(i) Has an effect on a specific diseaseor class of diseases;

(ii) Has an effect on the characteristicsigns or symptoms of a specific diseaseor class of diseases, using scientific orlay terminology;

(iii) Has an effect on an abnormalcondition associated with a natural stateor process, if the abnormal condition isuncommon or can cause significant orpermanent harm;

(iv) Has an effect on a disease ordiseases through one or more of thefollowing factors:

(A) The name of the product;(B) A statement about the formulation

of the product, including a claim thatthe product contains an ingredient(other than an ingredient that is anarticle included in the definition of‘‘dietary supplement’’ under 21 U.S.C.321(ff)(3)) that has been regulated byFDA as a drug and is well known toconsumers for its use or claimed use inpreventing or treating a disease;

(C) Citation of a publication orreference, if the citation refers to adisease use, and if, in the context of thelabeling as a whole, the citation impliestreatment or prevention of a disease,e.g., through placement on theimmediate product label or packaging,inappropriate prominence, or lack ofrelationship to the product’s expressclaims;

(D) Use of the term ‘‘disease’’ or‘‘diseased,’’ except in general statementsabout disease prevention that do notrefer explicitly or implicitly to a specificdisease or class of diseases or to aspecific product or ingredient; or

(E) Use of pictures, vignettes,symbols, or other means;

(v) Belongs to a class of products thatis intended to diagnose, mitigate, treat,cure, or prevent a disease;

(vi) Is a substitute for a product thatis a therapy for a disease;

(vii) Augments a particular therapy ordrug action that is intended to diagnose,mitigate, treat, cure, or prevent a diseaseor class of diseases;

(viii) Has a role in the body’s responseto a disease or to a vector of disease;

(ix) Treats, prevents, or mitigatesadverse events associated with a therapyfor a disease, if the adverse eventsconstitute diseases; or

(x) Otherwise suggests an effect on adisease or diseases.

Dated: October 26, 1999.Jane E. Henney,Commissioner of Food and Drugs.Donna E. Shalala,Secretary of Health and Human Services.[FR Doc. 00–53 Filed 01–5–00; 8:45 am]BILLING CODE 4160–01–F


Date post:	13-Mar-2018
Category:	Documents
Upload:	phungmien
View:	216 times
Download:	2 times

1000 Federal Register /Vol. 65, No. 4/Thursday, … Federal Register/Vol. 65, No. 4/Thursday,...

Documents