lable at ScienceDirect

Nutrition 30 (2014) 569–574

Contents lists avai

Nutrition

journal homepage: www.nutr i t ionjrnl .com

Applied nutritional investigation

Bioelectrical impedance vector analysis in obese women beforeand after bariatric surgery: Changes in body composition

Carolina Ferreira Nicoletti R.D. a, Jos�e Simon Camelo Jr. M.D. b,Jos�e Ernesto dos Santos M.D. a, Julio Sergio Marchini M.D. a,Wilson Salgado Jr. M.D. c, Carla Barbosa Nonino R.D. a,*aDepartment of Internal Medicine, Faculty of Medicine of Ribeirao Preto, University of S~ao Paulo, Ribeir~ao Preto, BrazilbDepartment of Pediatrics, Faculty of Medicine of Ribeirao Preto, University of S~ao Paulo, Ribeir~ao Preto, BrazilcDepartment of Surgery and Anatomy, Faculty of Medicine of Ribeirao Preto, University of S~ao Paulo, Ribeir~ao Preto, Brazil

a r t i c l e i n f o

Article history:Received 9 September 2013Accepted 28 October 2013

Keywords:ObesityBariatric surgeryBioelectrical impedance analysisImpedance vectorBody composition

CFN, JEDS, and WS Jr. designed the research. CFNresearch. JSM and JSC Jr. performed the statistical ainterpretations. CFN, CBN, JSM, and JSC Jr. wrote the aapproved the final version of the manuscript. CFN asponsibility for the final content. The authors certifyinterest.* Corresponding author. Tel.: þ55 16 3 602 4810; f

E-mail address: carla@fmrp.usp.br (C. B. Nonino).

0899-9007/$ - see front matter � 2014 Elsevier Inc. Ahttp://dx.doi.org/10.1016/j.nut.2013.10.013

a b s t r a c t

Objective: Because of the inefficacy of standard methods for the evaluation of body composition ofgrade III obese individuals, it is difficult to analyze the quality of weight loss after bariatric surgeryin these patients. Electrical bioimpedance vector analysis and the RXc graph uses crude resistance(R) and reactance (Xc) values, like components of the Z vector, to monitor variations in body fluidand the nutritional status of obese individuals. Using bioelectrical impedance vector analysis(BIVA) and the RXc graph, the objective of the present study was to evaluate long-term changes inweight and body composition of obese women after Roux-en-Y bariatric surgery.Methods: A study was conducted on 43 grade III obese women submitted to bariatric surgery.Anthropometric and bioimpedance (800 mA–50 kHz) data were obtained during the preoperativeperiod and 1, 2, 3, and 4 y after surgery. BIVA was performed by plotting resistance and reactancevalues corrected for body height (R/H and Xc/H, Ohm/m) as bivariates on the RXc graph. BIVAsoftware was used to plot the vectors of the RXc plane.Results: Surgery promoted changes in body composition, with a reduction of fat mass and of fat-free mass. During the postoperative period, the vectors demonstrated migration to the rightlower quadrant of the graph, corresponding to the classification of cachexia and water retention.Conclusion: Weight loss due to surgery results in an important reduction of fat-free mass charac-terized by the position of most individuals in the cachexia quadrant throughout the postoperativeperiod.

� 2014 Elsevier Inc. All rights reserved.

Introduction

The prevalence of obesity in recent (defined by body massindex [BMI] � 30 kg/m2) has markedly increased over the pastdecades, representing a worldwide epidemic [1]. Clinical treat-ment, including behavioral changes with a reduction of energyintake and increased practice of physical activity, at times asso-ciated with drug treatment, results in modest and transitory

and CBN conducted thenalysis and BIVA and PArticle. All authors read andnd CBN take primary re-no potential conflicts of

ax: þ55 16 3 602 3375.

ll rights reserved.

effects, without resolving the problem of severe obesity [2]. Theindication of surgical treatment currently is increasing and todayis accepted as the only efficient tool for the treatment of grade IIIobesity, inducing the loss of excess weight and its later mainte-nance, as well as positive results regarding comorbidity condi-tions such as type 2 diabetes mellitus (T2 DM), hyperlipidemia,and arterial hypertension [2]. Roux-en-Y gastric bypass iscurrently the procedure most accepted by surgeons [3].

Ideally, weight loss should primarily occur due to reduction offat mass (FM), thus minimizing the loss of fat-free mass (FFM)[4]. The assessment of body composition plays an important rolein clinical evaluation and in monitoring FM and FFM changesduring specific therapeutic regimens in obese individuals as away to determine the efficacy of the interventions regardingweight loss [5]. However, the validation of body compositionmeasurements in obese individuals is still scarce [6]. Some

C. F. Nicoletti et al. / Nutrition 30 (2014) 569–574570

traditional methods used for individuals of normal weight oroverweight are ineffective for the evaluation of grade III obeseindividuals due to practical reasons such as patient size anddesign and capacity of the equipment [6].

Bioelectrical impedance (BIA) is a simple, low-cost, andnoninvasive procedure used for the evaluation of body compo-sition in clinical routine and in weight reduction programs [7].The method is based on the principle that electric flow is facili-tated through hydrated tissue and extracellular water comparedwith adipose tissue, providing reproducible and rapidly obtainedresults [8]. Although BIA is widely employed for the evaluation ofgrade III obese individuals, its use is still controversial [5]. Studieshave demonstrated that obese patients present variations of softtissue hydration, which may generate errors in the evaluation ofbody composition by standard methods such as predictiveequations [6,9,10]. It has been demonstrated that changes inbody composition occur in obese individuals, characterized by anincreased quantity of FM and a reduction of total bodywater [10].

Bioelectrical impedance vector analysis (BIVA) and the RXcgraph are useful methods for the determination of changes intissue hydration and nutritional status [11]. BIVA uses resistance(R) and reactance (Xc) values as components of the Z vector inthe RXc graph, with R and Xc being normalized for height (R/Hand Xc/H) and plotted as bivariates in the RXc graph [12]. Thus,the length of the vector indicates the state of hydration and themigration of the vector correlates with the quantity of soft tissueregardless of the measurements of body weight or predictiveequations [12].

The postoperative period of bariatric surgery is accompaniedby rapid weight loss. Due to the inefficacy of standard methodsfor the evaluation of the body composition of grade III obeseindividuals, however, it is difficult to analyze the quality ofweight loss after surgery. Thus, the objective of the present studywas to evaluate the long-term changes in weight and in bodycomposition after Roux-en-Y bariatric surgery by BIVA and theRXc graph in obese women.

Material and methods

Study design and participants

The sample consisted of womenwith grade III obesity submitted to bariatricsurgery by Roux-en-Y gastric bypass and followed up at the National ReferenceCenter of Bariatric Surgery of the University Hospital, Faculty of Medicine ofRibeir~ao Preto, University of S~ao Paulo (HC-FMRP/USP). Data were collectedretrospectively by analysis of medical/nutritional records. Among all the patientsfollowed up at the service, those who attended all return visits and for whomcomplete information was available regarding anthropometry and bodycomposition were included. The study was approved by the Research EthicsCommittee of HC-FMRP/USP and all patients gave written informed consent toparticipate.

Data collection

Body composition analysis of two compartments was obtained by standardBIA (R, Xc, FFM [kg], FM [kg], and total body water [kg]) during the preoperativeperiod and 1, 2, 3, and 4 y after surgery, as well as weight (kg), height (m), andBMI (kg/m2). As part of the protocol of the service, body weight was measuredwith a Filizola scale with 100 g precision (Filizola, Sao Paulo, Brazil), with theparticipant standing erect, barefoot, and wearing minimal light clothing. Heightwas measured with a graded metal rod with a maximum length of 2 m and0.5 cm precision. BMI was calculated as weight divided by height squared.

A monofrequency Quantum BIA 101 q-RJL Systems analyzer (Clinton Town-ship, MI, USA) was used to evaluate body composition. The analyzer providesvalues of resistance and reactance to the passage of an electrical current of 800 mAand 50 kHz by means of electrodes positioned on the dorsum of the hand and ofthe ipsilateral foot of individuals lying in dorsal decubitus. Patients were previ-ously instructed to avoid alcohol intake and exhausting physical activity on theday before the exam and to limit food or fluid intake to 4 h before the test.

FFM and FM were determined by means of linear regression equationsvalidated for adults [13] using the RJL Systems Weight Manager 2.05 a software.The phase-angle (PA) was calculated from R and Xc values according to thisequation [14]: F ¼ Xc/R x 180/p.

Based on the RXc graph method, R and Xc values were standardized forpatient height (H), with R/H and Xc/H values being obtained as Ohm/m. BIVAsoftwarewas used to plot the vectors of the RXc plane [15]. The 50%, 75%, and 95%tolerance ellipses of a reference population determined in a previous study [16]were used to evaluate the evolution of the individual vectors of the patients. Aspreviously reported [15], the ellipses were divided into four quadrants namedobese, athletic, lean, and cachectic.

Statistical analysis

Mixed-effects regression models were used for the analysis of the anthro-pometric and body composition variables during the postoperative period. Forgroup comparison, ellipses of 95% confidence were calculated for the meanimpedance vectors and the Hotelling T2 test and univariate analysis (F test) wereused. The absence of overlap of the 95% tolerance ellipses indicated a significantdifference (P < 0.05) between the mean values of the vectors positioned on theRXc plane.

Results

Longitudinal evaluation

To date, the Reference Center of Bariatric Surgery has con-ducted 550 bariatric surgeries. Of all patients undergoing thesurgery, 370 completed 4 y of postoperative period; 124 patientshad BIA data, and only 54 had complete BIA data for every 4-yperiod; however, 11 of these were men. Thus, the study wasconducted on 43 women with a mean age of 43.2 � 9.3 y and amean height of 1.62� 0.1 cm during the preoperative period. Theanthropometric and body composition data of the patients arereported in Table 1. During the preoperative period 27.9% pa-tients had T2 DM, 58.1% had hypertension, 25.6% had dyslipi-demia, 14% had hypothyroidism, and 4.6% had hyperuricemia.

A reduction of 47.4�15.7 kg (35%) of initial weight, 8.9� 4 kg(14.2%) of FFM, and 38.5 � 12 kg (52.6%) of FM was observedduring the postoperative year 1, with 81.2% of the weight lossinvolving FM and 18.8% involving FFM. PA, total body water, andXc also resulted in a significant reduction (P < 0.001) during thefirst year.

By postoperative year 4, therewas a 15.5% reduction of FFM inparallel to a 9.6% reduction of total body water, suggesting loss ofmuscle and bone tissue. The significant reduction of R and PAvalues is emphasized (P < 0.001).

Evolution of individual vectors

The distribution of vectors in the tolerance ellipses during thepreoperative period and at the different postoperative times isillustrated in Figure 1. It can be seen that during the preoperativeperiod, 88.4% of the vectorswere outside the 95% ellipse and only11.6% demonstrated adequate tissue hydration.

Short vectors tending to be located on the lower right side ofthe graph were observed at all-time points and, after bariatricsurgery, the vectors indicated a rightward shift characterizingmigration of the vectors parallel to themajor axis of the graph. Atthe end of the study, the patients were predominantly located inthe quadrant denoted as cachexia and water retention, accordingto a previous report [15].

Group comparison

Figure 2 shows the 95% tolerance ellipses for the groups.There was a significant difference between the preoperative and

Table 1Anthropometric and body composition data

Groups Preoperative period Postoperative period

1 y 2 y 3 y 4 y

Weight (kg) 135 � 19.3 87.5 � 16* 82.9 � 16.7*,y 84.8 � 16.7* 86.7 � 17.6*

BMI (kg/m2) 51.4 � 7 33.2 � 5.3* 31.5 � 5.7*,y 32.2 � 5.8* 32.9 � 6.1*

FFM (kg) 61.8 � 6.3 52.9 � 5* 51.7 � 4.9* 52.3 � 6.5* 52 � 4.9*

FFM (%) 46.1 � 2.7 61.5 � 7.1* 63.8 � 8*,y 62.7 � 6.7* 61.2 � 6.8*,z

FM (kg) 73.1 � 13.6 34.7 � 11.6* 31.2 � 12.6*,y 32.5 � 11.6* 34.7 � 13.2*

FM (%) 53.1 � 2.7 38.5 � 7.1* 36.2 � 8*,y 37.6 � 6.7*,y 38.8 � 6.8*,z

TBW (l) 44.7 � 8.8 41.9 � 6.7* 40.8 � 6.4* 40.1 � 6.3*,y 39.7 � 5.8*,y

TBW (%) 33.1 � 3.8 48.5 � 6.9* 50.3 � 8.9* 48.2 � 7.2*,z 46.6 � 6.7*,z

Phase angle (�) 6.3 � 1.1 5.2 � 1.4* 5 � 1.3* 4.8 � 0.8* 4.7 � 0.9*,y

R (Ohm) 449.2 � 90 437.4 � 52.4 447.6 � 56.6 449.4 � 65.4y 464.1 � 57.8*,y,z

R/H (Ohm/m) 277.7 � 58 270.4 � 35.6 276.6 � 37.5 278 � 43y 286.6 � 36.8*,y,z

Xc (Ohm) 49.2 � 12.8 39.2 � 10.8* 39 � 11.7* 38.5 � 8.4* 37.3 � 11.7*

Xc/H (Ohm/m) 30.4 � 8 24.1 � 6.4* 24.1 � 7.4* 23.7 � 4.9* 23.1 � 7.5*

R (R/H, Xc/H) 0.77 0.22 0.52 0.09 0.44

BMI, body mass index; FFM, fat-free mass; FM, fat mass; TBW, total body water; R, resistance; Xc, reactance; H, height* P � 0.05 compared with preoperative period.y P � 0.05 compared with postoperative period of 1 y.z P � 0.05 compared with postoperative period of 2 y.

C. F. Nicoletti et al. / Nutrition 30 (2014) 569–574 571

the postoperative conditions at 1, 2, 3, and 4 y postsurgery (P <

0.01), demonstrating that surgery promoted changes in bodycomposition characterized by loss of FM, FFM, and wateraccumulation.

Discussion

The main objective of the present study was to evaluate thechanges in weight, body composition, and tissue hydration in agroup of obese women during a 4-y study after Roux-en-Ygastricbypass using BIVA, which is based on absolute R and Xc dataplotted as vectors on an RXc graph. The results demonstratedthat the surgical procedure promoted a significant weight lossand changes in body composition characterized by FM and FFMreduction. The vectors proved migration to the lower rightquadrant of the graph, corresponding to a classification ofcachexia and water retention according to a previous interpre-tation [15].

Data from the literature consider a weight loss of more than50% of excess weight or a 30% to 40% loss of the initial weight torepresent surgical success [17,18]. The results of the presentstudy show that bariatric surgery induced a 35% loss of initialweight during postoperative year 1, which was maintained afterthis period.

Obesity is characterized by excess weight resulting fromexcess lipid accumulation in adipose tissue, which is closelyassociated with comorbidities such as cardiovascular diseases,insulin resistance, and other metabolic disorders [2]. In thetreatment of obesity, weight loss should preferentially resultfrom the reduction of adipose tissue, with preservation of skel-etal muscle, favoring greater success in the maintenance ofweight loss and lower risk for undernutrition and clinical com-plications [19]. Studies have demonstrated significant reductionsof FFM after different surgical techniques including Roux-en-Ygastric bypass, with loss of muscle proteins during the phase ofrapid weight loss [6,20].

The present findings show that weight loss was characterizedby a great reduction of adipose tissue, especially during thepostoperative year 1, although FFM loss also occurred during thisperiod. The data have shown wide variation of the influence ofFFM loss onweight reduction and have reportedweight loss ratesranging from 11% to 31% [21,22]. Some authors have reported 22%

as a maximum acceptable loss of FFM in situations of weight lossin white women [23]. The evaluation of these data is extremelyimportant because several studies have demonstrated a strongrelationship between FFM loss and unfavorable clinical andnutritional evolution [8]. One study [24] later assessed changes inbody composition by dual-energy x-ray absorptiometry (DXA)after Roux-en-Y gastric bypass and observed that those patientswho had a higher rate of weight loss had accelerated loss of FFMand FM. It has been demonstrated that a significant loss of FFM isassociated with rapid weight loss [21].

The use of BIA as a method for the evaluation of bodycomposition is based on the assumption that the hydration ofsoft tissues (FFM) is a constant factor, corresponding to 73.2% inadults [25]. However, some authors have shown that obeseindividuals have a greater quantity of total body water thannormal-weight individuals [6,26], with the abnormal geometryof the body and the variations in tissue hydration possiblyresulting in errors of prediction of body composition when us-ing standard methods, overestimating the quantity of FFM andconsequently underestimating FM [26,27]. Although DXA isconsidered the gold standard for assessment of body composi-tion; the available devices do not carry the weight of severelyobese individuals. Thus, there are no appropriate methods forevaluation of body composition in grade III obese individuals.In a study comparing body composition data by BIA anddeuterium-labeled water (gold standard) in obese children andadolescents, the authors stated that although BIA is a rapid andpractical method, the predictive formulas used must beadjusted [9].

BIVA uses crude primary R and Xc values without the need forequations or models. Thus, as a qualitative method for theevaluation of body tissue hydration, BIVA can be used for theroutinemonitoring of the variation in body fluids and nutritionalstatus of obese individuals, representing a clinically useful pro-cedure [11].

The use of the RXc graph can yield three types of results:

1. point analysis of an individual by plotting the single vector inthe reference tolerance ellipse;

2. longitudinal analysis of an individual by plotting the vectorsof successive measurement in the reference tolerance ellip-ses; and

Fig. 1. Individual vectors for 43 obese patients in the 50%, 75%, and 95% tolerance ellipses for the reference population, (A) preoperative; (B) 1-y postoperative period; (C) 2-ypostoperative period; (D) 3-y postoperative period; (E) 4-y postoperative period. Xc, reactance; R, resistance.

C. F. Nicoletti et al. / Nutrition 30 (2014) 569–574572

3. evaluation of groups of individuals using the bivariate of the95% confidence ellipse of the vectors [15].

Validation studies carried out on adults have shown thatvectors positioned outside the 75% confidence ellipse andsloping downward indicate abnormal tissue impedance, withreduced vectors sloping downward indicating a fluid overload,whereas steep and elongated vectors represent dehydration [15].

Because nowater is present in the fat droplets of adipocytes, ithas been pointed out that the quantitation of total bodywater canbe an important factor in FMmonitoring [9]. The current findingspresent short vectors during the preoperative period, demon-strating that obese individuals have tissue overhydration. Thesefindings agree with those reported previously [11], which indi-cated that 61.8% of obese individuals had vectors outside the 75%confidence ellipse. However, a study of premenopausal obesewomen (35.3 � 9.1 y), did not detect vectors less than the limitbetween obesity/edema, indicating normal tissue hydration [5].

In the present study, vectors less than the 75% tolerance el-lipse were observed after bariatric surgery, characterizingedematous individuals. However, the surgical procedure pro-moted a reduction of absolute total body water parallel to theloss of FFM. A study of obese patients after laparoscopic Roux-en-Y gastric bypass [28] did not detect differences in tissue hydra-tion between the preoperative and postoperative periods, asindicated by the unchanged length of the bioimpedance vectors.

Analysis of the individual vectors during the postoperativeperiod determined a migration to the lower right side of thegraph parallel to the smaller axis, indicating loss of cell mass andcorroborating a consistent decrease of Xc values. The present Xcdata demonstrate probable damage to cell integrity, indicating areduction in cell volume or number. The Xc component of thevector, which is neglected in most of the studies using standardBIA, was crucial for the discrimination of membrane integritybecause its consistent reduction during the postoperative perioddemonstrated undesired cell loss.

Fig. 2. Mean impedance vector with the 95% confidence ellipse for the patientsbefore and after bariatric surgery, (a) preoperative; (b) 1-y postoperative period; (c)2-y postoperative period; (d) 3-y postoperative period; (e) 4-y postoperativeperiod. Xc, reactance; R, resistance; H, height.

C. F. Nicoletti et al. / Nutrition 30 (2014) 569–574 573

The PA can be used as an indicator of nutritional status andbody mass [29]. Recent studies on adults have indicated theusefulness of the PA as a prognostic marker in clinical conditionsand as a risk marker in critically ill patients [30,31]. It has beendemonstrated that the PA is associated with low BMI and a lowFFM index. The results of the present study show a reduction ofPA values after bariatric surgery as a consequence of reduced Xc.A study using a different surgical technique from that employedin the present patients did not detect a change in PA by 6 and12 mo postsurgery, with a mean preoperative value of 6.6 � 0.6�

and values of 6.7 � 0.8� and 6.6 � 0.6� after 6 and 12 mo [5].These values were higher than those detected in the presentstudy, possibly questioning the severity of the repercussions ofthe surgical technique.

A study evaluating weight loss after a low-calorie dietdemonstrated lengthening of the impedance vector character-izing the loss of FFM after weight loss regimens [32]. In our study,a migration of vectors is observed from the classification ofobesity to cachexia, confirming FFM loss. Thus, the present studybased on the BIVA method demonstrates a change in electricalproperties of the organism submitted to bariatric surgery,reflecting a severe, although acceptable, loss of FFM.We questionwhether this could be an adverse effect that may occur afterbariatric surgery. Further studies addressing other aspects ofnutritional status, particularly biochemical parameters and/orthe use of stable isotopes should be performed to clarify theseissues.

We wish to underscore that the concept of cachexia definedpreviously [15] for the right lower quadrant has no direct rela-tionship with the concept that cachexia involves severe malnu-trition and inflammation.

Group comparison by the 95% confidence ellipse revealed adifference between the preoperative period and the post-operative periods of 1, 2, 3, and 4 y, emphasizing the effect ofweight loss on body composition postsurgery.

The evaluation of grade III obese individuals by BIVA and theRXc graph provides information about the status of tissue hy-dration and permits the discrimination between hyperhydrationand body fat, also providing information about the quantity ofcell mass and membrane integrity.

Our findings suggest that interventions to reduce FFM lossafter bariatric surgerymust be considered, as well as the need for

specific nutritional monitoring during the immediate post-operative period to reduce the loss of muscle mass. We speculatethat protein supplementation together with the practice ofphysical activity may minimize the loss of muscle mass,contributing to better patient prognosis.

Conclusion

The present data are consistent with changes in the electricalproperties of the organism, which suggest that the weight lossresulting from bariatric surgery leads to an important reductionof FM and FFM characterized by the positioning of most in-dividuals in the cachexia quadrant throughout the postoperativeperiod.

References

[1] Hurt RT, Kulisek C, Buchanan LA, McClave SA. The obesity epidemic:challenges, health initiatives, and implications for gastroenterologists.Gastroenterol Hepatol 2010;6:780–2.

[2] Donadelli SP, Salgado W Jr, Marchini JS, Schmidt A, Amato CA, Ceneviva R,et al. Change in predicted 10-year cardiovascular risk following Roux-en-Ygastric bypass surgery: who benefits? Obes Surg 2011;21:569–73.

[3] Nguyen NT, Gelfand DV, Zainabadi K. Laparoscopic Roux-en-Y gastricbypass vs. laparoscopic adjustable banding for treatment of morbid obesity.Surg Technol Int 2004;12:111–9.

[4] Palazuelos-Genis T, Mosti M, S�anchez-Leenheer S, Hern�andez R, Gardu~no O,Herrera MF. Weight loss and body composition during the first post-operative year of a laparoscopic Roux-en-Y gastric bypass. Obes Surg2008;18:1–4.

[5] Savastano S, Belfiore A, Di Somma C, Mauriello C, Rossi A, Pizza G, et al.Validity of bioelectrical impedance analysis to estimate body compositionchanges after bariatric surgery in premenopausal morbidly women. ObesSurg 2009;20:332–9.

[6] Das SK, Roberts SB, Kehayias JJ, Wang J, Hsu LK, Shikora SA, et al. Bodycomposition assessment in extreme obesity and after massive weight lossinduced by gastric bypass surgery. Am J Physiol Endocrinol Metab2003;284:1080–8.

[7] Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-freemass using bioelectrical impedance measurements of the human body. AmJ Clin Nutr 1985;41:810–7.

[8] Cunha SFC, Sanches M, Faria A, Nonino-Borges CB. Evoluc~ao da massacorporal magra ap�os 12 meses da cirurgia bari�atrica. Revista de Nutric~ao2010;23:535–41.

[9] Resende CM, Camelo J�unior JS, Vieira MN, Ferriolli E, Pfrimer K, Perdon�a GS,et al. Body composition measures of obese adolescents by the deuteriumoxide dilution method and by bioelectrical impedance. Braz J Med Biol Res2011;44:1164–70.

[10] World Health Organization. Physical status: the use and interpretation ofanthropometry. Available at: http//:www.who.int/em/. Accessed 09/05/2013.

[11] Piccoli A, Brunani A, Savia G, Pillon L, Favaro E, Berselli ME, et al.Discriminating between body fat and fluid changes in the obese adultusing bioimpedance vector analysis. Int J Obes Relat Metab Disord1998;22:97–104.

[12] Piccoli A, Rossi B, Pillon L, Bucciante G. A new method for monitoring bodyfluid variation by bioimpedance analysis: the RXc graph. Kidney Int1994;46:534–9.

[13] Jakicic JM, Wing RR, Lang W. Bioelectrical impedance analysis to assessbody composition in obese adult women: the effect of ethnicity. Int JObesity 1998;22:243–9.

[14] Baumgartner RN, Chumlea C, Roche PA. Bioelectric impedance phase angleand body composition. Am J Clin Nutr 1988;48:16–23.

[15] Piccoli A, Pillon L, Dumler F. Impedance vector distribution by sex, race,body mass index, and age in the United States: standard reference intervalsas bivariate Z scores. Nutrition 2002;18:153–67.

[16] Piccoli A, Nigrelli S, Caberlotto A, Bottazzo S, Rossi B, Pillon L, et al. Bivariatenormal value of the bioelectrical impedance vector in adult and elderlypopulations. Am J Clin Nutr 1995;61:269–70.

[17] De Maria EJ, Jamal MK. Surgical options for obesity. Gastroenterol ClinNorth Am 2005;34:127–42.

[18] Costa LD, Valezi AC, Matsuo T, Dichi I, Dichi JB. Repercuss~ao da perda depeso sobre para

ˇ

metros nutricionais e metab�olicos de pacientes obesosgraves ap�os um ano de gastroplastia em Y-de-Roux. Rev Col Bras Cir2010;37:96–101.

[19] Ciangura C, Bouillot JL, Lloret-Linares C, Poitou C, Veyrie N, Basdevant A,et al. Dynamics of change in total and regional body composition aftergastric bypass in obese patients. Obesity 2010;1:760–5.

C. F. Nicoletti et al. / Nutrition 30 (2014) 569–574574

[20] Wadström C, Backman L, Forsberg AM, Nilsson E, Hultman E, Reizenstein P,et al. Body composition and muscle constituents during weight loss: studiesin obese patients following gastroplasty. Obes Surg 2000;10:203–13.

[21] Carey DG, Pliego GJ, Raymond RL. Body composition and metaboliscchanges following bariatric surgery: effects on fat mass, lean mass andbasal metabolic rate: six months to one-year follow-up. Obes Surg2006;16:1602–8.

[22] Coupaye M, Bouillot JL, Poitou C, Schutz Y, Basdevant A, Oppert JM. Is leanbody mass decreased after obesity treatment by adjustable gastric band-ing? Obes Surg; 2007:427–33.

[23] Webster JD, Hesp R, Garrow JS. The composition of excess weight in obesewomen estimated by body density, total body water and total body po-tassium. Hum Clin Nut 1984;38:299–306.

[24] Zalesin KC, Franklin BA, Lillystone MA, Shamoun T, Krause KR,Chengelis DL, et al. Differential loss of fat and lean mass in the morbidlyobese after bariatric surgery. Metab Synd Relat Disord 2010;8:15–20.

[25] Horie LM, Barbosa-Silva MC, Torrinhas RS, de Mello MT, Cecconello I,Waitzberg DL. New body fat prediction equations for severely obese pa-tients. Clin Nutr 2008;27:350–6.

[26] Waki M, Kral JG, Mazariegos M, Wang J, Pierson RN Jr, Heymsfield SB.Relative expansion of extracellular fluid in obese vs. nonobese women. AmJ Physiol 1991;261:199–203.

[27] Coppini LZ, Waitzberg DL, Campos AC. Limitations and validation ofbioelectrical impedance analysis in morbidly obese patients. Curr Opin ClinNutr Metab Care 2005;8:329–32.

[28] Guida B, Belfiore A, Angrisani L, Micanti F, Mauriello C, Trio R, et al.Laparoscopic gastric banding and body composition in morbid obesity.Nutr Metab Cardiovasc Dis 2005;15:198–203.

[29] Schiesser M, Kirchhoff P, Müller MK, Schäfer M, Clavien PA. The correlationof nutrition risk index, nutrition risk score, and bioimpedance analysiswith postoperative complications in patients undergoing gastrointestinalsurgery. Surgery 2009;145:519–26.

[30] Van Lettow M, Kumwenda JJ, Harries AD, Whalen CC, Taha TE,Kumwenda N, et al. Undernutrition and the severity of lung disease inadults with pulmonary tuberculosis in Malawi. Int J Tuberc Lung Dis2004;8:211–7.

[31] GIBI Brazilian Group for Bioimpedance Study. Total body bioelec-trical impedance measurement as a progressive outcome predictionand therapeutic index in the comparison between septic and non-septic patients. A multicenter Brazilian study. Rev Metab Nutr1995;2:159–70.

[32] Frisard MI, Greenway FL, Delany JP. Comparison of methods to assessbody composition changes during a period of weight loss. Obes Res2005;1:845–54.