AbstractObjectiveTo validate the current criteria of visual snow and to describe its common phenotype usinga substantial clinical database
MethodsWe performed a web-based survey of patients with self-assessed visual snow (n = 1104) witheither the complete visual snow syndrome (n = 1061) or visual snow without the syndrome(n = 43) We also describe a population of patients (n = 70) with possible hallucinogenpersisting perception disorder who presented clinically with visual snow syndrome
ResultsThe visual snow population had an average age of 29 years and had no sex prevalence Thedisorder usually started in early life and asymp40 of patients had symptoms for as long as theycould remember The most commonly experienced static was black and white Floatersafterimages and photophobia were themost reported additional visual symptoms A latent classanalysis showed that visual snow does not present with specific clinical endophenotypesSeverity can be classified by the amount of visual symptoms experienced Migraine and tinnitushad a very high prevalence and were independently associated with a more severe presentationof the syndrome
ConclusionsClinical characteristics of visual snow did not differ from the previous cohort in the literaturesupporting validity of the current criteria Visual snow likely represents a clinical continuumwith different degrees of severity On the severe end of the spectrum it is more likely to presentwith its common comorbid conditions migraine and tinnitus Visual snow does not depend onthe effect of psychotropic substances on the brain
RELATED ARTICLE
EditorialVisual snow Are webeginning to see the light
Page 241
MORE ONLINE
PodcastDr Teshamae Monteithtalks with Dr FrancescaPuledda about her paperproviding a clinical andphenotypical descriptionof visual snow syndrome
NPuborgfxcblh
CME CourseNPuborgcmelist
From the Headache Group (FP PJG) Department of Basic and Clinical Neuroscience Institute of Psychiatry Psychology amp Neuroscience Kingrsquos College London NIHR-WellcomeTrust Kingrsquos Clinical Research Facility (FP PJG) SLaM Biomedical Research Centre Kingrsquos College Hospital London UK and Department of Neurology (CS) Inselspital BernUniversity Hospital University of Bern Switzerland
Go to NeurologyorgN for full disclosures Funding information and disclosures deemed relevant by the authors if any are provided at the end of the article
The Article Processing Chargewas funded by crowdfunding from the self-help group for visual snow EyeOn Vision Foundation and funding from the Visual Snow Initiative and from theSLaM Biomedical Research Centre
This is an open access article distributed under the terms of the Creative Commons Attribution License 40 (CC BY) which permits unrestricted use distribution and reproduction in anymedium provided the original work is properly cited
e564 Copyright copy 2020 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology
Visual snow (VS) is a recently identified neurologic conditionconsisting of a constant positive visual disturbance described asuncountable tiny dots over the entire visual field1 (figure 1) Inaddition to the static patients very often report visual symp-toms such as palinopsia entoptic phenomena photophobiaand nyctalopia This constitutes the VS syndrome (VSS)which is outlined by a set of specific criteria (table 1)2 From thefirst case report of VS by Liu et al3 in 1995 the recognition ofthe disorder has grown considerably to the point where VS isnow included in the appendix of the International Classificationof Headache Disorders as a complication of migraine4
Several questions concerning VS have arisen as the conditionhas become better known For example if there is a need todefine the differences between VS and the presentation asa ldquocompleterdquo syndrome What are its most common clinicalphenotypes and what is the relationship with other comorbidconditions Similarly given that hallucinogenics can producea similar disturbance5 are there clinical distinctions that mayinform understanding the biology involved
Here we describe the clinical characteristics of a substantialpopulation of patients with VS both with and without thecomplete VSS Our main objective was to test the currentcriteria and to confirm the typical presentation of the mainsymptom of VSmdashie the staticmdashin the context of a largercohort and to begin to determine any broad differences re-lated to geography We also wanted to explore a dataset large
enough to dissect possible subgroups and endophenotypesFinally we wished to observe the interaction between VS andits main comorbid conditions migraine and tinnitus and tocompare the disturbance with patients with hallucinogenpersisting perception disorder (HPPD)
MethodsParticipant selection and recruitmentThe study was advertised on the website of Eye On Vision(eyeonvisionorg) a patient self-help group for VS with whomwe have collaboratedMost of the patients involved in the studyapproached our group through a dedicated research e-mailwhich they could find on the website A smaller number ofpatients had contacted the researchers individually asking to beinvolved in research and were redirected to the website
An online survey was prepared in collaboration with the patientgroup and was made available on the Eye On Vision websiteThe survey is illustrated in table 2 it presents a series of open anddichotomous questions aimed to characterize the symptoms ofVS following the available criteria We also investigated thepresence ofmigraine and tinnitus Finally we enquired about ageat symptom onset and previous exposure to recreational drugs
The study was approved by the KCL Research Ethics PanelData were collected between April 2016 and May 2018
Figure 1 Illustration of visual snow
GlossaryDSM-V = Diagnostic and Statistical Manual of Mental Disorders 5th edition HPPD = hallucinogen persisting perceptiondisorder VS = visual snow VSS = visual snow syndrome
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e565
Patient characterizationFollowing the 2014 criteria (table 1) we defined patients whoself-reported visual symptoms corresponding to criterion A(as evaluated by responses to questions 2ndash4 of our survey)and who also fit criterion B (as evaluated by responses toquestion 5 in our survey) as having VSS Criteria C and Dwere evaluated on a case-to-case basis on the basis of answersto questions 2 6 and 7 and eventual follow-up questions bythe investigators when in doubt Patients who did not reportgt2 additional visual symptoms of the 4 main categoriestherefore lacking criterion B but who fit all the other criteriawere considered to have VS without the syndrome
To avoid confounding with HPPD patients who answeredldquoyesrdquo to question 7 in our survey were further followed up within-depth questions aimed at assessing when their symptomsappeared with respect to the intake of recreational drugs Allparticipants who reported the onset of VS symptoms in the 12months after any exposure to recreational drugs were excludedfrom the first 2 groups regardless of the remaining symptomsand were added to a third group called possible HPPD Allparticipants in this third group fit criterion A for typical VS andwere therefore included in the analysis
All data collection and patient characterization were per-formed by one of us (FP)
Statistical analysisData were tabulated (Excel 2016 for Windows) Descriptivestatistics analysis of variance or χ2 analysis for comparisons ofcontinuous and categorical variables and cluster analysis wereperformedwith SPSS Statistics Version 240 forWindows (IBM
Corp Armonk NY) Regression analysis multiple imputationsand latent class analysis were performed in Stata (Stata Statis-tical Software release 15 2017 StataCorp LLC College StationTX) Values of p lt 005 were considered significant
We separated patients into 3 different groups according to theirdiagnosis patients with VSS were coded as group 1 patientswith VS as group 2 and patients with HPPD as group 3
An ordinal variable was created to measure disease severityaccording to the number of visual symptoms experienced Forthe largest cohort (ie group 1) this outcome variable wasregressed on selected variables using ordinal logistic re-gression The variables included as covariates in this modelwere selected in a previous step based on a correlation withthe number of symptoms experienced defined by significantcorrelations at the 5 level using the Spearman correlation
A latent class analysis was performed to investigate possibleendophenotypes of VS first on group 1 only and then ongroups 1 and 2 combined
Data availabilityThe data that support the findings of this study are availablefrom the corresponding author on reasonable request
ResultsDemographic characteristicsFrom April 2016 to May 2018 patients (n = 1400) contactedthe study group through the e-mail designated to VS researchOf these 210 either gave incomplete data in the initial survey
Table 1 Criteria for the definition of the visual snow syndrome
A Visual snow dynamic continuous tiny dots in the entire visual field lasting gt3 mo
The dots are usually blackgray on white background and graywhite on black background however they can also be transparent white flashing orcolored
B Presence of at least 2 additional visual symptoms of the 4 following categories
(i) Palinopsia At least 1 of the following afterimages or trailing of moving objects
Afterimages should be different from retinal afterimages which occur only when staring at a high-contrast image and are in complementary color
(ii) Enhanced entoptic phenomena At least 1 of the following excessive floaters in both eyes excessive blue field entoptic phenomenon self-light of theeye or spontaneous photopsia
Entoptic phenomena arise from the structure of the visual system itself The blue field entoptic phenomenon is described as uncountable little graywhiteblack dots or rings shooting over the visual field in both eyes when looking at homogeneous bright surfaces such as the blue sky self-light of theeye is described as colored waves or clouds when closing the eyes in the dark spontaneous photopsia is characterized by bright flashes of light
(iii) Photophobia
(iv) Nyctalopia
C Symptoms are not consistent with typical migraine visual aura
As defined by the International Headache Society in the International Classification of Headache Disorders4
D Symptoms are not better explained by another disorder
Normal ophthalmology tests (best corrected visual acuity dilated fundus examination visual field and electroretinogram) not causedby previous intake ofpsychotropic drugs
e566 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
or never replied after having been redirected to the patientwebsite these individuals were excluded from further datacollection Two individuals had an insufficient English level 6had a serious underlying ophthalmic condition and 8 did notfulfill criterion A meaning the static they reported either wasof episodic nature or was present in only 1 part of the visualfield These patients were all excluded from the analysis
Demographic characteristics of the remaining participants (n= 1174) are presented in table 3 which also shows details ofsymptom onset and associated comorbid conditions Themajority of the cohort (n = 1061 90) had complete VSSForty-three participants in the cohort were considered to haveVS alone because they provided a very clear description of thedynamic continuous pan-field tiny dots described in criterion
A These participants however did not present at least 2 visualsymptoms from the additional categories and were grouped inthe VS category Of these 7 participants had no additionalsymptoms 20 had only 1 symptom and 16 had between 2 and4 symptoms of the same category eg palinopsia for stationaryobjects and trailing or several entoptic phenomena
Seventy participants were grouped as possible HPPD followingthe criteria previously described1 With the exception of expo-sure to recreational drugs in the 12 months before the onset ofsymptoms they all had the remaining criteria for VSS diagnosis
Features of the disorder in the cohorts andcomparison to HPPDIn table 3 the 3 groups are compared among themselves andwith the cohort from the 2014 study when the required datawere available The 4 groups did not differ with regard to ageMale and female ratios were similar for participants with VS(VSS and VS) however in the HPPD population mostpatients (71 p lt 0001) were male The large majority ofparticipants came from North America (n = 429 41 for VSSn = 23 54 for VS n = 21 31 for HPPD) and Europe (n =497 48 for VSS n = 15 35 for VS n = 41 61 for HPPD)
Participants with HPPD had a significantly later onset ofsymptoms compared to patients with VS both with and with-out the syndrome (p lt 0001) As a consequence the averageyears with disease were lower in theHPPDgroup (plt 0001) Asmall number (n = 99) of participants in the VSS groupreported a clear stepwise worsening of symptoms at some pointof the condition however this feature was not routinelyscreened for so it is not possible to infer its actual prevalenceForty percent of patients with VSS for whom data on onset agewere available reported the presence of symptoms since child-hood meaning for as long as they could recall This was higherthan the proportion found in the 2014 study About one-quarterof the participants with VS (VSS and VS) reported a suddenonset of their symptoms however the real frequency of thisform of onset might be different because participants were notinterrogated about it directly Of these spontaneous reports ofsudden onset some were related to specific conditions in-dicated in the table 3 a migraine attack was the most frequentIn the majority of cases however the participants could notrecall any specific associated event A sudden onset of symp-toms was significantly more frequent in the HPPD group (81p lt 0001) By definition all of these patients had the start ofsymptoms within a year after using recreational drugs Four ofthese participants could also recall other specific events (ie amigraine attack a new medication and a mild head trauma) instrict temporal relation to the beginning of their symptoms
Tinnitus and migraineThe presence of tinnitus in the VSS population was the highestbut overall similar to that of the HPPD and 2014 cohorts Thefrequency of this symptom was however significantly lower inthe VS group compared to the others This was also the case formigraine which was significantly less frequent in the VS
Table 2 Online survey (available on eyeonvisionorg)
Name
Address
Date of birth (daymonthyear)
Telephone number
(1) Please make a brief statement that you are willing to be contacted forresearch This is a European data protection issue Example ldquoYes pleasekeep my contact details and you may contact me for research purposesrdquo
(2) Brief description of all symptoms you relate to visual snow syndrome
(3) Date or age when your symptoms started
(4) Visual snow what type
Black and white (ie only black dots on white background white dots onblack background)
Clear (ie color of the background)
Flashing (ie always white brighter than background)
Colored
All of these
(5) Other symptoms (please only answer yes or no)
After images
Trailing of images in the vision
Blue field entoptic phenomenon (ie white squiggly lines movingpulsating on the blue sky)
Floaters in vision
Colored clouds or waves with eyes closed
Flashes of light
Impaired night vision
Sensitive to light
Tinnitus
(6) Have you ever been diagnosed with migraine or have you hada headache of moderate or severe intensity in the past (Please answer yesor no)
(7) Have you ever taken any illicit drugs in the past
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e567
population and more frequent in the VSS population Thepresence of migraine aura was not routinely investigated becausewe considered this diagnosis unreliable for a dichotomousquestionnaire It was spontaneously reported in 37 (p= 005) ofparticipants in the VSS group and in all these cases the diagnosiswas confirmed with thorough follow-up questions
Clinical characterization of VS StaticWe collected information on the type of static that patients ex-perienced and the associated visual symptoms that form the VSSTable 4 shows the frequencies of these characteristics comparingthem across the 3 groups and with the 2014 study cohort
Each of the 4 static types was reported more frequently in theVSS group This is probably accounted for by the fact that thisgroup had a significantly higher proportion of participantsreporting all 4 types of static The most common type of staticdiffered between the VSS and VS groups compared to theHPPD group When only 1 type of static was present this wasmost commonly black and white in the VSS and VS groupsand transparent in the HPPD group When 2 types of staticwere present the most frequent combination was black andwhite and transparent for the VSS and VS groups and coloredand flashing for the HPPD group When 3 types of static werepresent the combination of black and white flashing and
Table 3 Demographics comorbid conditions and characteristics of symptom onset
VSS VS HPPD Schankin et al1
Patients n ( of cohort) 1061 (90) 43 (4) 70 (6) 78
Age (mean plusmn SD) y 295 plusmn 103 294 plusmn 123 275 plusmn 72 30 plusmn 10
Femalemale within group n 521539 (p = 06) 1825 (p = 03) 2050b 3741
Region of origin n ()
Europe 497 (48) 15 (35) 41 (61)
North America 429 (41) 23 (54) 21 (31)
Central and South America 23 (2) 0 1 (2)
Central Asia and Middle East 28 (3) 1 (23) 0
Southeast Asia 17 (2) 1 (23) 0
Oceania 49 (5) 3 (7) 4 (6)
Africa 3 (03) 0 0
Age at symptom onset (mean plusmn SD) y 128 plusmn 132 87 plusmn 102 205 plusmn 67b 21 plusmn 9
Data available n 823 28 60
Symptoms present since childhood n () 326 (40) 13 (46) 1 (001)b 19 (24)
Years of disease (mean plusmn SD) 172 plusmn 136 202 plusmn 169 71 plusmn 81b mdash
Stepwise worsening of symptoms n () 99 (9) 0 4 (6) 10 (13)
Sudden onset of symptoms nN () 176691 (26) 416 (25) 3847 (81)b
Abbreviations HPPD = hallucinogen persisting perception disorder VS = visual snow VSS = visual snow syndromea p lt 005 measured accordingly by analysis of variance or χ2 testsb p lt 0001 measured accordingly by analysis of variance or χ2 tests
e568 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
Visual snow (VS) is a recently identified neurologic conditionconsisting of a constant positive visual disturbance described asuncountable tiny dots over the entire visual field1 (figure 1) Inaddition to the static patients very often report visual symp-toms such as palinopsia entoptic phenomena photophobiaand nyctalopia This constitutes the VS syndrome (VSS)which is outlined by a set of specific criteria (table 1)2 From thefirst case report of VS by Liu et al3 in 1995 the recognition ofthe disorder has grown considerably to the point where VS isnow included in the appendix of the International Classificationof Headache Disorders as a complication of migraine4
Several questions concerning VS have arisen as the conditionhas become better known For example if there is a need todefine the differences between VS and the presentation asa ldquocompleterdquo syndrome What are its most common clinicalphenotypes and what is the relationship with other comorbidconditions Similarly given that hallucinogenics can producea similar disturbance5 are there clinical distinctions that mayinform understanding the biology involved
Here we describe the clinical characteristics of a substantialpopulation of patients with VS both with and without thecomplete VSS Our main objective was to test the currentcriteria and to confirm the typical presentation of the mainsymptom of VSmdashie the staticmdashin the context of a largercohort and to begin to determine any broad differences re-lated to geography We also wanted to explore a dataset large
enough to dissect possible subgroups and endophenotypesFinally we wished to observe the interaction between VS andits main comorbid conditions migraine and tinnitus and tocompare the disturbance with patients with hallucinogenpersisting perception disorder (HPPD)
MethodsParticipant selection and recruitmentThe study was advertised on the website of Eye On Vision(eyeonvisionorg) a patient self-help group for VS with whomwe have collaboratedMost of the patients involved in the studyapproached our group through a dedicated research e-mailwhich they could find on the website A smaller number ofpatients had contacted the researchers individually asking to beinvolved in research and were redirected to the website
An online survey was prepared in collaboration with the patientgroup and was made available on the Eye On Vision websiteThe survey is illustrated in table 2 it presents a series of open anddichotomous questions aimed to characterize the symptoms ofVS following the available criteria We also investigated thepresence ofmigraine and tinnitus Finally we enquired about ageat symptom onset and previous exposure to recreational drugs
The study was approved by the KCL Research Ethics PanelData were collected between April 2016 and May 2018
Figure 1 Illustration of visual snow
GlossaryDSM-V = Diagnostic and Statistical Manual of Mental Disorders 5th edition HPPD = hallucinogen persisting perceptiondisorder VS = visual snow VSS = visual snow syndrome
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e565
Patient characterizationFollowing the 2014 criteria (table 1) we defined patients whoself-reported visual symptoms corresponding to criterion A(as evaluated by responses to questions 2ndash4 of our survey)and who also fit criterion B (as evaluated by responses toquestion 5 in our survey) as having VSS Criteria C and Dwere evaluated on a case-to-case basis on the basis of answersto questions 2 6 and 7 and eventual follow-up questions bythe investigators when in doubt Patients who did not reportgt2 additional visual symptoms of the 4 main categoriestherefore lacking criterion B but who fit all the other criteriawere considered to have VS without the syndrome
To avoid confounding with HPPD patients who answeredldquoyesrdquo to question 7 in our survey were further followed up within-depth questions aimed at assessing when their symptomsappeared with respect to the intake of recreational drugs Allparticipants who reported the onset of VS symptoms in the 12months after any exposure to recreational drugs were excludedfrom the first 2 groups regardless of the remaining symptomsand were added to a third group called possible HPPD Allparticipants in this third group fit criterion A for typical VS andwere therefore included in the analysis
All data collection and patient characterization were per-formed by one of us (FP)
Statistical analysisData were tabulated (Excel 2016 for Windows) Descriptivestatistics analysis of variance or χ2 analysis for comparisons ofcontinuous and categorical variables and cluster analysis wereperformedwith SPSS Statistics Version 240 forWindows (IBM
Corp Armonk NY) Regression analysis multiple imputationsand latent class analysis were performed in Stata (Stata Statis-tical Software release 15 2017 StataCorp LLC College StationTX) Values of p lt 005 were considered significant
We separated patients into 3 different groups according to theirdiagnosis patients with VSS were coded as group 1 patientswith VS as group 2 and patients with HPPD as group 3
An ordinal variable was created to measure disease severityaccording to the number of visual symptoms experienced Forthe largest cohort (ie group 1) this outcome variable wasregressed on selected variables using ordinal logistic re-gression The variables included as covariates in this modelwere selected in a previous step based on a correlation withthe number of symptoms experienced defined by significantcorrelations at the 5 level using the Spearman correlation
A latent class analysis was performed to investigate possibleendophenotypes of VS first on group 1 only and then ongroups 1 and 2 combined
Data availabilityThe data that support the findings of this study are availablefrom the corresponding author on reasonable request
ResultsDemographic characteristicsFrom April 2016 to May 2018 patients (n = 1400) contactedthe study group through the e-mail designated to VS researchOf these 210 either gave incomplete data in the initial survey
Table 1 Criteria for the definition of the visual snow syndrome
A Visual snow dynamic continuous tiny dots in the entire visual field lasting gt3 mo
The dots are usually blackgray on white background and graywhite on black background however they can also be transparent white flashing orcolored
B Presence of at least 2 additional visual symptoms of the 4 following categories
(i) Palinopsia At least 1 of the following afterimages or trailing of moving objects
Afterimages should be different from retinal afterimages which occur only when staring at a high-contrast image and are in complementary color
(ii) Enhanced entoptic phenomena At least 1 of the following excessive floaters in both eyes excessive blue field entoptic phenomenon self-light of theeye or spontaneous photopsia
Entoptic phenomena arise from the structure of the visual system itself The blue field entoptic phenomenon is described as uncountable little graywhiteblack dots or rings shooting over the visual field in both eyes when looking at homogeneous bright surfaces such as the blue sky self-light of theeye is described as colored waves or clouds when closing the eyes in the dark spontaneous photopsia is characterized by bright flashes of light
(iii) Photophobia
(iv) Nyctalopia
C Symptoms are not consistent with typical migraine visual aura
As defined by the International Headache Society in the International Classification of Headache Disorders4
D Symptoms are not better explained by another disorder
Normal ophthalmology tests (best corrected visual acuity dilated fundus examination visual field and electroretinogram) not causedby previous intake ofpsychotropic drugs
e566 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
or never replied after having been redirected to the patientwebsite these individuals were excluded from further datacollection Two individuals had an insufficient English level 6had a serious underlying ophthalmic condition and 8 did notfulfill criterion A meaning the static they reported either wasof episodic nature or was present in only 1 part of the visualfield These patients were all excluded from the analysis
Demographic characteristics of the remaining participants (n= 1174) are presented in table 3 which also shows details ofsymptom onset and associated comorbid conditions Themajority of the cohort (n = 1061 90) had complete VSSForty-three participants in the cohort were considered to haveVS alone because they provided a very clear description of thedynamic continuous pan-field tiny dots described in criterion
A These participants however did not present at least 2 visualsymptoms from the additional categories and were grouped inthe VS category Of these 7 participants had no additionalsymptoms 20 had only 1 symptom and 16 had between 2 and4 symptoms of the same category eg palinopsia for stationaryobjects and trailing or several entoptic phenomena
Seventy participants were grouped as possible HPPD followingthe criteria previously described1 With the exception of expo-sure to recreational drugs in the 12 months before the onset ofsymptoms they all had the remaining criteria for VSS diagnosis
Features of the disorder in the cohorts andcomparison to HPPDIn table 3 the 3 groups are compared among themselves andwith the cohort from the 2014 study when the required datawere available The 4 groups did not differ with regard to ageMale and female ratios were similar for participants with VS(VSS and VS) however in the HPPD population mostpatients (71 p lt 0001) were male The large majority ofparticipants came from North America (n = 429 41 for VSSn = 23 54 for VS n = 21 31 for HPPD) and Europe (n =497 48 for VSS n = 15 35 for VS n = 41 61 for HPPD)
Participants with HPPD had a significantly later onset ofsymptoms compared to patients with VS both with and with-out the syndrome (p lt 0001) As a consequence the averageyears with disease were lower in theHPPDgroup (plt 0001) Asmall number (n = 99) of participants in the VSS groupreported a clear stepwise worsening of symptoms at some pointof the condition however this feature was not routinelyscreened for so it is not possible to infer its actual prevalenceForty percent of patients with VSS for whom data on onset agewere available reported the presence of symptoms since child-hood meaning for as long as they could recall This was higherthan the proportion found in the 2014 study About one-quarterof the participants with VS (VSS and VS) reported a suddenonset of their symptoms however the real frequency of thisform of onset might be different because participants were notinterrogated about it directly Of these spontaneous reports ofsudden onset some were related to specific conditions in-dicated in the table 3 a migraine attack was the most frequentIn the majority of cases however the participants could notrecall any specific associated event A sudden onset of symp-toms was significantly more frequent in the HPPD group (81p lt 0001) By definition all of these patients had the start ofsymptoms within a year after using recreational drugs Four ofthese participants could also recall other specific events (ie amigraine attack a new medication and a mild head trauma) instrict temporal relation to the beginning of their symptoms
Tinnitus and migraineThe presence of tinnitus in the VSS population was the highestbut overall similar to that of the HPPD and 2014 cohorts Thefrequency of this symptom was however significantly lower inthe VS group compared to the others This was also the case formigraine which was significantly less frequent in the VS
Table 2 Online survey (available on eyeonvisionorg)
Name
Address
Date of birth (daymonthyear)
Telephone number
(1) Please make a brief statement that you are willing to be contacted forresearch This is a European data protection issue Example ldquoYes pleasekeep my contact details and you may contact me for research purposesrdquo
(2) Brief description of all symptoms you relate to visual snow syndrome
(3) Date or age when your symptoms started
(4) Visual snow what type
Black and white (ie only black dots on white background white dots onblack background)
Clear (ie color of the background)
Flashing (ie always white brighter than background)
Colored
All of these
(5) Other symptoms (please only answer yes or no)
After images
Trailing of images in the vision
Blue field entoptic phenomenon (ie white squiggly lines movingpulsating on the blue sky)
Floaters in vision
Colored clouds or waves with eyes closed
Flashes of light
Impaired night vision
Sensitive to light
Tinnitus
(6) Have you ever been diagnosed with migraine or have you hada headache of moderate or severe intensity in the past (Please answer yesor no)
(7) Have you ever taken any illicit drugs in the past
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e567
population and more frequent in the VSS population Thepresence of migraine aura was not routinely investigated becausewe considered this diagnosis unreliable for a dichotomousquestionnaire It was spontaneously reported in 37 (p= 005) ofparticipants in the VSS group and in all these cases the diagnosiswas confirmed with thorough follow-up questions
Clinical characterization of VS StaticWe collected information on the type of static that patients ex-perienced and the associated visual symptoms that form the VSSTable 4 shows the frequencies of these characteristics comparingthem across the 3 groups and with the 2014 study cohort
Each of the 4 static types was reported more frequently in theVSS group This is probably accounted for by the fact that thisgroup had a significantly higher proportion of participantsreporting all 4 types of static The most common type of staticdiffered between the VSS and VS groups compared to theHPPD group When only 1 type of static was present this wasmost commonly black and white in the VSS and VS groupsand transparent in the HPPD group When 2 types of staticwere present the most frequent combination was black andwhite and transparent for the VSS and VS groups and coloredand flashing for the HPPD group When 3 types of static werepresent the combination of black and white flashing and
Table 3 Demographics comorbid conditions and characteristics of symptom onset
VSS VS HPPD Schankin et al1
Patients n ( of cohort) 1061 (90) 43 (4) 70 (6) 78
Age (mean plusmn SD) y 295 plusmn 103 294 plusmn 123 275 plusmn 72 30 plusmn 10
Femalemale within group n 521539 (p = 06) 1825 (p = 03) 2050b 3741
Region of origin n ()
Europe 497 (48) 15 (35) 41 (61)
North America 429 (41) 23 (54) 21 (31)
Central and South America 23 (2) 0 1 (2)
Central Asia and Middle East 28 (3) 1 (23) 0
Southeast Asia 17 (2) 1 (23) 0
Oceania 49 (5) 3 (7) 4 (6)
Africa 3 (03) 0 0
Age at symptom onset (mean plusmn SD) y 128 plusmn 132 87 plusmn 102 205 plusmn 67b 21 plusmn 9
Data available n 823 28 60
Symptoms present since childhood n () 326 (40) 13 (46) 1 (001)b 19 (24)
Years of disease (mean plusmn SD) 172 plusmn 136 202 plusmn 169 71 plusmn 81b mdash
Stepwise worsening of symptoms n () 99 (9) 0 4 (6) 10 (13)
Sudden onset of symptoms nN () 176691 (26) 416 (25) 3847 (81)b
Abbreviations HPPD = hallucinogen persisting perception disorder VS = visual snow VSS = visual snow syndromea p lt 005 measured accordingly by analysis of variance or χ2 testsb p lt 0001 measured accordingly by analysis of variance or χ2 tests
e568 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
Patient characterizationFollowing the 2014 criteria (table 1) we defined patients whoself-reported visual symptoms corresponding to criterion A(as evaluated by responses to questions 2ndash4 of our survey)and who also fit criterion B (as evaluated by responses toquestion 5 in our survey) as having VSS Criteria C and Dwere evaluated on a case-to-case basis on the basis of answersto questions 2 6 and 7 and eventual follow-up questions bythe investigators when in doubt Patients who did not reportgt2 additional visual symptoms of the 4 main categoriestherefore lacking criterion B but who fit all the other criteriawere considered to have VS without the syndrome
To avoid confounding with HPPD patients who answeredldquoyesrdquo to question 7 in our survey were further followed up within-depth questions aimed at assessing when their symptomsappeared with respect to the intake of recreational drugs Allparticipants who reported the onset of VS symptoms in the 12months after any exposure to recreational drugs were excludedfrom the first 2 groups regardless of the remaining symptomsand were added to a third group called possible HPPD Allparticipants in this third group fit criterion A for typical VS andwere therefore included in the analysis
All data collection and patient characterization were per-formed by one of us (FP)
Statistical analysisData were tabulated (Excel 2016 for Windows) Descriptivestatistics analysis of variance or χ2 analysis for comparisons ofcontinuous and categorical variables and cluster analysis wereperformedwith SPSS Statistics Version 240 forWindows (IBM
Corp Armonk NY) Regression analysis multiple imputationsand latent class analysis were performed in Stata (Stata Statis-tical Software release 15 2017 StataCorp LLC College StationTX) Values of p lt 005 were considered significant
We separated patients into 3 different groups according to theirdiagnosis patients with VSS were coded as group 1 patientswith VS as group 2 and patients with HPPD as group 3
An ordinal variable was created to measure disease severityaccording to the number of visual symptoms experienced Forthe largest cohort (ie group 1) this outcome variable wasregressed on selected variables using ordinal logistic re-gression The variables included as covariates in this modelwere selected in a previous step based on a correlation withthe number of symptoms experienced defined by significantcorrelations at the 5 level using the Spearman correlation
A latent class analysis was performed to investigate possibleendophenotypes of VS first on group 1 only and then ongroups 1 and 2 combined
Data availabilityThe data that support the findings of this study are availablefrom the corresponding author on reasonable request
ResultsDemographic characteristicsFrom April 2016 to May 2018 patients (n = 1400) contactedthe study group through the e-mail designated to VS researchOf these 210 either gave incomplete data in the initial survey
Table 1 Criteria for the definition of the visual snow syndrome
A Visual snow dynamic continuous tiny dots in the entire visual field lasting gt3 mo
The dots are usually blackgray on white background and graywhite on black background however they can also be transparent white flashing orcolored
B Presence of at least 2 additional visual symptoms of the 4 following categories
(i) Palinopsia At least 1 of the following afterimages or trailing of moving objects
Afterimages should be different from retinal afterimages which occur only when staring at a high-contrast image and are in complementary color
(ii) Enhanced entoptic phenomena At least 1 of the following excessive floaters in both eyes excessive blue field entoptic phenomenon self-light of theeye or spontaneous photopsia
Entoptic phenomena arise from the structure of the visual system itself The blue field entoptic phenomenon is described as uncountable little graywhiteblack dots or rings shooting over the visual field in both eyes when looking at homogeneous bright surfaces such as the blue sky self-light of theeye is described as colored waves or clouds when closing the eyes in the dark spontaneous photopsia is characterized by bright flashes of light
(iii) Photophobia
(iv) Nyctalopia
C Symptoms are not consistent with typical migraine visual aura
As defined by the International Headache Society in the International Classification of Headache Disorders4
D Symptoms are not better explained by another disorder
Normal ophthalmology tests (best corrected visual acuity dilated fundus examination visual field and electroretinogram) not causedby previous intake ofpsychotropic drugs
e566 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
or never replied after having been redirected to the patientwebsite these individuals were excluded from further datacollection Two individuals had an insufficient English level 6had a serious underlying ophthalmic condition and 8 did notfulfill criterion A meaning the static they reported either wasof episodic nature or was present in only 1 part of the visualfield These patients were all excluded from the analysis
Demographic characteristics of the remaining participants (n= 1174) are presented in table 3 which also shows details ofsymptom onset and associated comorbid conditions Themajority of the cohort (n = 1061 90) had complete VSSForty-three participants in the cohort were considered to haveVS alone because they provided a very clear description of thedynamic continuous pan-field tiny dots described in criterion
A These participants however did not present at least 2 visualsymptoms from the additional categories and were grouped inthe VS category Of these 7 participants had no additionalsymptoms 20 had only 1 symptom and 16 had between 2 and4 symptoms of the same category eg palinopsia for stationaryobjects and trailing or several entoptic phenomena
Seventy participants were grouped as possible HPPD followingthe criteria previously described1 With the exception of expo-sure to recreational drugs in the 12 months before the onset ofsymptoms they all had the remaining criteria for VSS diagnosis
Features of the disorder in the cohorts andcomparison to HPPDIn table 3 the 3 groups are compared among themselves andwith the cohort from the 2014 study when the required datawere available The 4 groups did not differ with regard to ageMale and female ratios were similar for participants with VS(VSS and VS) however in the HPPD population mostpatients (71 p lt 0001) were male The large majority ofparticipants came from North America (n = 429 41 for VSSn = 23 54 for VS n = 21 31 for HPPD) and Europe (n =497 48 for VSS n = 15 35 for VS n = 41 61 for HPPD)
Participants with HPPD had a significantly later onset ofsymptoms compared to patients with VS both with and with-out the syndrome (p lt 0001) As a consequence the averageyears with disease were lower in theHPPDgroup (plt 0001) Asmall number (n = 99) of participants in the VSS groupreported a clear stepwise worsening of symptoms at some pointof the condition however this feature was not routinelyscreened for so it is not possible to infer its actual prevalenceForty percent of patients with VSS for whom data on onset agewere available reported the presence of symptoms since child-hood meaning for as long as they could recall This was higherthan the proportion found in the 2014 study About one-quarterof the participants with VS (VSS and VS) reported a suddenonset of their symptoms however the real frequency of thisform of onset might be different because participants were notinterrogated about it directly Of these spontaneous reports ofsudden onset some were related to specific conditions in-dicated in the table 3 a migraine attack was the most frequentIn the majority of cases however the participants could notrecall any specific associated event A sudden onset of symp-toms was significantly more frequent in the HPPD group (81p lt 0001) By definition all of these patients had the start ofsymptoms within a year after using recreational drugs Four ofthese participants could also recall other specific events (ie amigraine attack a new medication and a mild head trauma) instrict temporal relation to the beginning of their symptoms
Tinnitus and migraineThe presence of tinnitus in the VSS population was the highestbut overall similar to that of the HPPD and 2014 cohorts Thefrequency of this symptom was however significantly lower inthe VS group compared to the others This was also the case formigraine which was significantly less frequent in the VS
Table 2 Online survey (available on eyeonvisionorg)
Name
Address
Date of birth (daymonthyear)
Telephone number
(1) Please make a brief statement that you are willing to be contacted forresearch This is a European data protection issue Example ldquoYes pleasekeep my contact details and you may contact me for research purposesrdquo
(2) Brief description of all symptoms you relate to visual snow syndrome
(3) Date or age when your symptoms started
(4) Visual snow what type
Black and white (ie only black dots on white background white dots onblack background)
Clear (ie color of the background)
Flashing (ie always white brighter than background)
Colored
All of these
(5) Other symptoms (please only answer yes or no)
After images
Trailing of images in the vision
Blue field entoptic phenomenon (ie white squiggly lines movingpulsating on the blue sky)
Floaters in vision
Colored clouds or waves with eyes closed
Flashes of light
Impaired night vision
Sensitive to light
Tinnitus
(6) Have you ever been diagnosed with migraine or have you hada headache of moderate or severe intensity in the past (Please answer yesor no)
(7) Have you ever taken any illicit drugs in the past
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e567
population and more frequent in the VSS population Thepresence of migraine aura was not routinely investigated becausewe considered this diagnosis unreliable for a dichotomousquestionnaire It was spontaneously reported in 37 (p= 005) ofparticipants in the VSS group and in all these cases the diagnosiswas confirmed with thorough follow-up questions
Clinical characterization of VS StaticWe collected information on the type of static that patients ex-perienced and the associated visual symptoms that form the VSSTable 4 shows the frequencies of these characteristics comparingthem across the 3 groups and with the 2014 study cohort
Each of the 4 static types was reported more frequently in theVSS group This is probably accounted for by the fact that thisgroup had a significantly higher proportion of participantsreporting all 4 types of static The most common type of staticdiffered between the VSS and VS groups compared to theHPPD group When only 1 type of static was present this wasmost commonly black and white in the VSS and VS groupsand transparent in the HPPD group When 2 types of staticwere present the most frequent combination was black andwhite and transparent for the VSS and VS groups and coloredand flashing for the HPPD group When 3 types of static werepresent the combination of black and white flashing and
Table 3 Demographics comorbid conditions and characteristics of symptom onset
VSS VS HPPD Schankin et al1
Patients n ( of cohort) 1061 (90) 43 (4) 70 (6) 78
Age (mean plusmn SD) y 295 plusmn 103 294 plusmn 123 275 plusmn 72 30 plusmn 10
Femalemale within group n 521539 (p = 06) 1825 (p = 03) 2050b 3741
Region of origin n ()
Europe 497 (48) 15 (35) 41 (61)
North America 429 (41) 23 (54) 21 (31)
Central and South America 23 (2) 0 1 (2)
Central Asia and Middle East 28 (3) 1 (23) 0
Southeast Asia 17 (2) 1 (23) 0
Oceania 49 (5) 3 (7) 4 (6)
Africa 3 (03) 0 0
Age at symptom onset (mean plusmn SD) y 128 plusmn 132 87 plusmn 102 205 plusmn 67b 21 plusmn 9
Data available n 823 28 60
Symptoms present since childhood n () 326 (40) 13 (46) 1 (001)b 19 (24)
Years of disease (mean plusmn SD) 172 plusmn 136 202 plusmn 169 71 plusmn 81b mdash
Stepwise worsening of symptoms n () 99 (9) 0 4 (6) 10 (13)
Sudden onset of symptoms nN () 176691 (26) 416 (25) 3847 (81)b
Abbreviations HPPD = hallucinogen persisting perception disorder VS = visual snow VSS = visual snow syndromea p lt 005 measured accordingly by analysis of variance or χ2 testsb p lt 0001 measured accordingly by analysis of variance or χ2 tests
e568 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
or never replied after having been redirected to the patientwebsite these individuals were excluded from further datacollection Two individuals had an insufficient English level 6had a serious underlying ophthalmic condition and 8 did notfulfill criterion A meaning the static they reported either wasof episodic nature or was present in only 1 part of the visualfield These patients were all excluded from the analysis
Demographic characteristics of the remaining participants (n= 1174) are presented in table 3 which also shows details ofsymptom onset and associated comorbid conditions Themajority of the cohort (n = 1061 90) had complete VSSForty-three participants in the cohort were considered to haveVS alone because they provided a very clear description of thedynamic continuous pan-field tiny dots described in criterion
A These participants however did not present at least 2 visualsymptoms from the additional categories and were grouped inthe VS category Of these 7 participants had no additionalsymptoms 20 had only 1 symptom and 16 had between 2 and4 symptoms of the same category eg palinopsia for stationaryobjects and trailing or several entoptic phenomena
Seventy participants were grouped as possible HPPD followingthe criteria previously described1 With the exception of expo-sure to recreational drugs in the 12 months before the onset ofsymptoms they all had the remaining criteria for VSS diagnosis
Features of the disorder in the cohorts andcomparison to HPPDIn table 3 the 3 groups are compared among themselves andwith the cohort from the 2014 study when the required datawere available The 4 groups did not differ with regard to ageMale and female ratios were similar for participants with VS(VSS and VS) however in the HPPD population mostpatients (71 p lt 0001) were male The large majority ofparticipants came from North America (n = 429 41 for VSSn = 23 54 for VS n = 21 31 for HPPD) and Europe (n =497 48 for VSS n = 15 35 for VS n = 41 61 for HPPD)
Participants with HPPD had a significantly later onset ofsymptoms compared to patients with VS both with and with-out the syndrome (p lt 0001) As a consequence the averageyears with disease were lower in theHPPDgroup (plt 0001) Asmall number (n = 99) of participants in the VSS groupreported a clear stepwise worsening of symptoms at some pointof the condition however this feature was not routinelyscreened for so it is not possible to infer its actual prevalenceForty percent of patients with VSS for whom data on onset agewere available reported the presence of symptoms since child-hood meaning for as long as they could recall This was higherthan the proportion found in the 2014 study About one-quarterof the participants with VS (VSS and VS) reported a suddenonset of their symptoms however the real frequency of thisform of onset might be different because participants were notinterrogated about it directly Of these spontaneous reports ofsudden onset some were related to specific conditions in-dicated in the table 3 a migraine attack was the most frequentIn the majority of cases however the participants could notrecall any specific associated event A sudden onset of symp-toms was significantly more frequent in the HPPD group (81p lt 0001) By definition all of these patients had the start ofsymptoms within a year after using recreational drugs Four ofthese participants could also recall other specific events (ie amigraine attack a new medication and a mild head trauma) instrict temporal relation to the beginning of their symptoms
Tinnitus and migraineThe presence of tinnitus in the VSS population was the highestbut overall similar to that of the HPPD and 2014 cohorts Thefrequency of this symptom was however significantly lower inthe VS group compared to the others This was also the case formigraine which was significantly less frequent in the VS
Table 2 Online survey (available on eyeonvisionorg)
Name
Address
Date of birth (daymonthyear)
Telephone number
(1) Please make a brief statement that you are willing to be contacted forresearch This is a European data protection issue Example ldquoYes pleasekeep my contact details and you may contact me for research purposesrdquo
(2) Brief description of all symptoms you relate to visual snow syndrome
(3) Date or age when your symptoms started
(4) Visual snow what type
Black and white (ie only black dots on white background white dots onblack background)
Clear (ie color of the background)
Flashing (ie always white brighter than background)
Colored
All of these
(5) Other symptoms (please only answer yes or no)
After images
Trailing of images in the vision
Blue field entoptic phenomenon (ie white squiggly lines movingpulsating on the blue sky)
Floaters in vision
Colored clouds or waves with eyes closed
Flashes of light
Impaired night vision
Sensitive to light
Tinnitus
(6) Have you ever been diagnosed with migraine or have you hada headache of moderate or severe intensity in the past (Please answer yesor no)
(7) Have you ever taken any illicit drugs in the past
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e567
population and more frequent in the VSS population Thepresence of migraine aura was not routinely investigated becausewe considered this diagnosis unreliable for a dichotomousquestionnaire It was spontaneously reported in 37 (p= 005) ofparticipants in the VSS group and in all these cases the diagnosiswas confirmed with thorough follow-up questions
Clinical characterization of VS StaticWe collected information on the type of static that patients ex-perienced and the associated visual symptoms that form the VSSTable 4 shows the frequencies of these characteristics comparingthem across the 3 groups and with the 2014 study cohort
Each of the 4 static types was reported more frequently in theVSS group This is probably accounted for by the fact that thisgroup had a significantly higher proportion of participantsreporting all 4 types of static The most common type of staticdiffered between the VSS and VS groups compared to theHPPD group When only 1 type of static was present this wasmost commonly black and white in the VSS and VS groupsand transparent in the HPPD group When 2 types of staticwere present the most frequent combination was black andwhite and transparent for the VSS and VS groups and coloredand flashing for the HPPD group When 3 types of static werepresent the combination of black and white flashing and
Table 3 Demographics comorbid conditions and characteristics of symptom onset
VSS VS HPPD Schankin et al1
Patients n ( of cohort) 1061 (90) 43 (4) 70 (6) 78
Age (mean plusmn SD) y 295 plusmn 103 294 plusmn 123 275 plusmn 72 30 plusmn 10
Femalemale within group n 521539 (p = 06) 1825 (p = 03) 2050b 3741
Region of origin n ()
Europe 497 (48) 15 (35) 41 (61)
North America 429 (41) 23 (54) 21 (31)
Central and South America 23 (2) 0 1 (2)
Central Asia and Middle East 28 (3) 1 (23) 0
Southeast Asia 17 (2) 1 (23) 0
Oceania 49 (5) 3 (7) 4 (6)
Africa 3 (03) 0 0
Age at symptom onset (mean plusmn SD) y 128 plusmn 132 87 plusmn 102 205 plusmn 67b 21 plusmn 9
Data available n 823 28 60
Symptoms present since childhood n () 326 (40) 13 (46) 1 (001)b 19 (24)
Years of disease (mean plusmn SD) 172 plusmn 136 202 plusmn 169 71 plusmn 81b mdash
Stepwise worsening of symptoms n () 99 (9) 0 4 (6) 10 (13)
Sudden onset of symptoms nN () 176691 (26) 416 (25) 3847 (81)b
Abbreviations HPPD = hallucinogen persisting perception disorder VS = visual snow VSS = visual snow syndromea p lt 005 measured accordingly by analysis of variance or χ2 testsb p lt 0001 measured accordingly by analysis of variance or χ2 tests
e568 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
population and more frequent in the VSS population Thepresence of migraine aura was not routinely investigated becausewe considered this diagnosis unreliable for a dichotomousquestionnaire It was spontaneously reported in 37 (p= 005) ofparticipants in the VSS group and in all these cases the diagnosiswas confirmed with thorough follow-up questions
Clinical characterization of VS StaticWe collected information on the type of static that patients ex-perienced and the associated visual symptoms that form the VSSTable 4 shows the frequencies of these characteristics comparingthem across the 3 groups and with the 2014 study cohort
Each of the 4 static types was reported more frequently in theVSS group This is probably accounted for by the fact that thisgroup had a significantly higher proportion of participantsreporting all 4 types of static The most common type of staticdiffered between the VSS and VS groups compared to theHPPD group When only 1 type of static was present this wasmost commonly black and white in the VSS and VS groupsand transparent in the HPPD group When 2 types of staticwere present the most frequent combination was black andwhite and transparent for the VSS and VS groups and coloredand flashing for the HPPD group When 3 types of static werepresent the combination of black and white flashing and
Table 3 Demographics comorbid conditions and characteristics of symptom onset
VSS VS HPPD Schankin et al1
Patients n ( of cohort) 1061 (90) 43 (4) 70 (6) 78
Age (mean plusmn SD) y 295 plusmn 103 294 plusmn 123 275 plusmn 72 30 plusmn 10
Femalemale within group n 521539 (p = 06) 1825 (p = 03) 2050b 3741
Region of origin n ()
Europe 497 (48) 15 (35) 41 (61)
North America 429 (41) 23 (54) 21 (31)
Central and South America 23 (2) 0 1 (2)
Central Asia and Middle East 28 (3) 1 (23) 0
Southeast Asia 17 (2) 1 (23) 0
Oceania 49 (5) 3 (7) 4 (6)
Africa 3 (03) 0 0
Age at symptom onset (mean plusmn SD) y 128 plusmn 132 87 plusmn 102 205 plusmn 67b 21 plusmn 9
Data available n 823 28 60
Symptoms present since childhood n () 326 (40) 13 (46) 1 (001)b 19 (24)
Years of disease (mean plusmn SD) 172 plusmn 136 202 plusmn 169 71 plusmn 81b mdash
Stepwise worsening of symptoms n () 99 (9) 0 4 (6) 10 (13)
Sudden onset of symptoms nN () 176691 (26) 416 (25) 3847 (81)b
Abbreviations HPPD = hallucinogen persisting perception disorder VS = visual snow VSS = visual snow syndromea p lt 005 measured accordingly by analysis of variance or χ2 testsb p lt 0001 measured accordingly by analysis of variance or χ2 tests
e568 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
transparent was the most common in the VSS and VS groupswhereas the combination of colored flashing and transparentwas the most common in the HPPD group The 2014 cohortshowed different results from all groups in the present studywith no patients reporting a combination of gt2 types of staticThis could be due to the fact that the current questionnairewas more flexible on static types accounting for all types asdefined by criterion A
Clinical characterization of VS-associatedvisual symptomsWith regard to the associated visual symptoms the VSS andHPPD groups did not significantly differ with regard to themean number of associated visual symptoms reported byeach patient The 3 most common symptoms in the VSS
group were in order floaters afterimages and photophobiaIn the VS cohort the most frequent symptoms were all ofthe entoptic phenomena group floaters self-light of the eyeand flashes In the HPPD group the 4 most commonlyreported symptoms were afterimages photophobia floatersand nyctalopia Taking each symptom separately the VSSand HPPD groups did not differ among themselves or withthe 2014 cohort with regard to the frequencies of eachsymptom
Predicting severity based on symptomsOrdinal logistic regressionThe results are shown in table 5We first performed a completecase analysis in which any case with a missing value on eitherthe outcome or the covariates was omitted from the analysis
Table 4 VS characteristics and frequencies of associated symptoms
VSS (n = 1061) n ()[95 CI]
VS (n = 43) n () [95CI]
HPPD (n = 70) n ()[95 CI]
Schankin et al1 (n = 78) n ()[95 CI]
Types of static
Black and white static 609 (58) (054ndash060)a 17 (40) (026ndash054) 30 (44) (032ndash055) 34 (44) (033ndash055)
Abbreviations CI = confidence interval HPPD = hallucinogen persisting perception disorder IQR = interquartile range VS = visual snow VSS = visual snowsyndromeValues are number (percentages within groups) (95 CIs of proportions)a p lt 005 measured accordingly by analysis of variance or χ2b p lt 0001 measured accordingly by analysis of variance or χ2
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e569
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
We then performed the same analysis using 10 imputeddatasets in which a chained equations approach was usedand the imputations were stratified by sex The parameterestimates were similar across both models This is un-surprising because there were no missing data for thesymptoms and the level of missing data was low for mostcovariates (lt2) except for disease duration and presenceof migraine These 2 variables were added later to the datacollection procedure and thus likely to be missing at ran-dom One participant for whom sex was not specified wasexcluded from the analysis
The results from the multiply imputed analysis indicatedthat female participants were approximately one-third morelikely to experience a higher number of symptoms than maleparticipants Those reporting migraine were gt25 timesmore likely to experience a higher number of symptomsand those with tinnitus were about twice as likely to expe-rience a higher number of symptoms A test for an in-teraction between migraine and tinnitus was not significantindicating that these 2 concomitant conditions exert in-dependent and additive effects on the number of VSsymptoms experienced Age disease duration and type ofonset were not related to the number of symptomsexperienced
Latent class analysisThe results of the latent class analysis are shown in table 6 andfigure 2 A and B This was performed first only on partic-ipants with VSS (table 6a) with the exclusion of 1 participantfor whom sex was not defined (n = 1060) Model fit criteriasuggested that a 2-class solution provided the most parsi-monious explanation of the data where classes 1 and 2accounted for x and y of the sample respectively The classes
that were obtained separated the patients into groups basedon additional visual symptom frequency Logistic regressionindicated that the same variables as the ordinal logistic re-gression for symptom frequency were related to latent classmembership (table 6b)
We then performed the same analysis on participants withVSS and VS (n = 1104 table 6c) excluding only those withHPPD The analysis showed that a third latent class wasobtained with this further step however the model stillshowed a group separation based solely on additional visualsymptom frequency
DiscussionWe here describe and analyze a large cohort of patients withVS The data support VSS as a well-delineated clinicallyrecognizable disorder Most patients have VS with other visualsymptoms There is an association with migraine and tinnitusthat is independent and the condition is clearly not simplydue to hallucinogenic intake The size of our patient sampleand its provenance from different parts of the world make itrepresentative of the real population and allow some inferenceon several aspects of this condition that should facilitate andguide further study of the problem
The results from this study indicate that participants withVSS are usually young and most commonly present withblack and white or transparent static as well as a highnumber of additional visual symptoms Even if there is nospecific sex prevalence identifying with female sex is sig-nificantly associated with reporting increased severity of thecondition The visual static can occur in different
Table 5 Ordinal logistic regression of frequency of additional visual symptoms
Abbreviations CI = confidence interval OR = odds ratioa p lt 005b p lt 0001
e570 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
combinations of color Visual disturbances can also presentin several combinations However floaters afterimages andphotophobia are almost invariably present and might in factconstitute a hallmark of the syndrome The disorder usuallystarts in early life and in many cases the patients have itsince childhood and can never recall seeing differently Inthese cases the affected person can find out almost seren-dipitously about the anomaly in seeing VS usually bycomparison with unaffected family members or friends Ina significant number of patients VS can start abruptly andspontaneously however this is not necessarily related toa higher number of symptoms in the condition
This study has shown that once specific criteria are definedand followed1 VS is a recognizable disorder with a very ho-mogeneous clinical presentation The description of the pri-mary symptom (ie the static) was highly reproducible acrossour cohort with only a few participants actually presentinga visual disturbance not attributable to VS The overall clinicalpresentation was also quite similar across our participants (seeprevious paragraph) and with the second largest cohort in theliterature1 albeit with some variations perhaps attributable todifferent sample sizes or different methodology For the 2014study patients were in fact interviewed in detail via telephonewhereas the present study was questionnaire based
Table 6 Latent class analysis performed on patients with VSS only (a b n = 1060) and on patients with VSS and VS (c n =1104)
1 2 3 4 5 6
(a) Model fit criteria of latent class analysis on patients with VSSsuggested that a 2-class solution provided the most parsimoniousexplanation of the data The classes obtained separated patientsinto groups based solely on frequency of additional visualsymptoms
(b) Logistic regression of the latent class analysis for patients with VSS The same variables were used as inthe ordinal logistic regression in table 5 reinforcing the relationship between frequency of additionalvisual symptoms and latent class membership
Age 099 0009 minus085 097ndash101
Sex 13 0202 17 096ndash177
Disease-years 099 0010 minus034 098ndash102
Migraine 236 0390 518 17ndash323
Tinnitus 19 0330 37 135ndash267
Onset type
1 11 0288 031 064ndash183
2 09 0284 minus04 0466ndash165
1 2 3 4 5 6
(c) Latent class analysis with patients with VSS and VS (n = 1104)Model fit criteria of latent class analysis suggested that a 3-classsolution provided the most parsimonious explanation of the dataAn extra class is recovered in the analysis with respect to table 6ahowever themodel still separated patients into groups based solelyon frequency of additional visual symptoms
Abbreviations AIC = Akaike information criterion BIC = bayesian information criterion CI = confidence interval OR = odds ratio VS = visual snow VSS = visualsnow syndromeNumber of observations n = 753 likelihood ratio χ2 = 5084 probability gt χ2 = 00000 pseudo R2 = 00491 log likelihood = minus49278a Two-class solutionb Three-class solution
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e571
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
The current criteria for the syndrome usefully eliminate false-positive participants In fact only a small minority of self-reporting patients consecutively recruited in the present studydid not fit the full syndrome definition It is nonetheless im-portant to recognize the presence of VS in these patients evenin the absence of additional visual disturbances which bothcharacterize a higher severity and define the syndrome but arenot a sine qua non for VS itself In fact VS is likely to con-stitute a spectrum type of disorder with patients ranging inseverity In this context a severe end of the spectrum could berepresented by those patients who have the static with all thevisual disturbances and are highly affected by them and a mildend could be represented by those who have only static andare not bothered by it possibly even considering it normal formost of their lives This theory is reinforced by the fact thatVSS and VS did not differ in their key clinical features such asaverage age sex distribution mode and age at onset Theydid however differ when it came to comorbid conditionswhich were more likely to be found within the syndrome inpatients with a higher number of associated symptoms (ie amore severe condition) and were less frequent in patients withVS but no syndrome compared to patients with VSS
This is emphasized by the latent class analysis itself showingthat VS does not present with specific clinical endopheno-types and is classified predominantly on its severity (measuredwith the burden of additional symptoms) Further studieswith objective measures on the levels of static such as severityand disability scales as perceived by patients would be neededto confirm this theory
VS has 2 main comorbid conditions migraine andtinnitus6ndash12 This strongly reported association suggests thatthese 2 conditions might share some common pathophysio-logic mechanism with VS This hypothesis is substantiated bya study that investigated brain metabolism in 17 patients withVS with the use of [18F]-fluorodeoxyglucose-PET6 andshowed a hypermetabolism of the right lingual gyrus inpatients with VS The area of the lingual gyrus corresponds toBrodmann area 19 in the supplementary visual cortex and is pivotalin processing complex downstream visual inputs This area is alsoinvolved in photophobia in migraine13 which further corroboratesthe concept of a shared pathophysiology betweenmigraine andVSpossibly on the basis of a dysfunctional cortical mechanism com-mon to both conditions
Figure 2 Latent class analysis
(A) Latent class analysis performed on n = 1060 patients with complete visual snow (VS) syndrome (VSS) Model fit criteria (table 6a) suggested that a 2-classsolution best explained the data The latent classes which separated the patients into groups based on additional visual symptom frequency are shownbelow (B) Latent class analysis performed on n = 1104 patients with complete VSS and VS without the syndrome Patients with hallucinogen persistingperception disorder were excluded With the addition of patients with VS an extra class was recovered (classes 1ndash3 shown below) However the model stillseparated the patients into groups based on additional visual symptom frequency only BFEP = blue field entoptic phenomenon
e572 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
Here we confirmed the presence of these comorbid con-ditions in a larger sample of patients with VS and demon-strated that both conditions are associated with a worsepresentation of VS defined by having more additional visualsymptoms These comorbid conditions independently pre-dicted the affinity to a severity class in the latent class modeland the number of additional symptoms in the ordinal lo-gistic regression This confirms the clinical and pathophys-iologic importance of interaction between migraine tinnitusand VS
Tinnitus is a common disorder in the general population witha prevalence ranging between 5 and 2514ndash17 In the presentcohort three-quarters of patients with VSS also had tinnitussuggesting a more than chance association between the 2conditions On a theoretical basis VS and tinnitus represent 2different manifestations of a similar disorder which is theperception of a sensory stimulus that is not present or is sub-threshold This neurobiological dysfunction would probablypoint to a central neuronal mechanism which could involveaberrant sensory processing at the level of association corticesor the thalamo-cortical network
Given that tinnitus not only is more frequently present but alsopredicts the severity of VS it is possible that both disordersshare a common pathophysiologic mechanism which if suffi-ciently active can manifest clinically with both conditions Thisis what seems to happen in the majority of the patients in thepresent study This hypothesized mechanisms could involvethalamo-cortical dysrhythmia1819 and cortical disexcitabilityboth of which have also been largely implicated in the contextof migraine pathophysiology2021 A possible confirmation ofsuch hypothesis comes from a very recent neurophysiologicstudy that showed late visual evoked potential alterations in VShighly suggestive of a visual association cortex dysfunction22
Therefore VS could represent an abnormality of sensory per-ception potentially involving multiple senses simultaneously
A dishabituation mechanism common to migraine and VSwould explain the worsening of the VS condition when mi-graine is present which was found here and in previousstudies6 as well as the strong comorbidity between them Thepresence of associated visual symptoms enhanced entopticphenomena in particular also potentially points to a disorderof habituation and sensory processing which allows the per-ception of stimuli that are normally ignored by the brain Amigrainous pathophysiology alone however is not sufficientto explain the VS biology primarily because of the chronicnature of this disorder as opposed to the ictal features ofmigraine but also because most preventive migraine medi-cation used empirically shows very little effect on VS23
An important issue with VS is the assumption that it could bedue to HPPD HPPD is a condition codified in DSM-V24 andcharacterized by the re-experiencing of perceptual symptoms(flashbacks) typically of the visual type that follow the ces-sation of the use of a hallucinogen and were experienced
during the intoxication25 VS and HPPD indeed share someclinical aspects mostly characterized by the possibility of thelatter to manifest with visual static palinopsia flashes andother types of visual dysperceptions52526 Recent literatureseems to suggest that HPPD can be distinguished into 2 mainentities In type 2 HPPD the visual symptoms are constant ornearly constant27 consistent with the group of participantsfrom our cohort In this study to avoid any possible con-founding overlap between HPPD and VS we applied strictcriteria to identify VSS and VS We considered 12 monthsfrom the intake of any recreational drugs as an appropriatetime to exhaust possible effects of psychotropic substances onthe visual system Any participants exposed within this timeframe were excluded from the VS groups We believe thisallowed us to confirm that VS pathophysiology does not havea connection with the use of recreational substances Fur-thermore we were able to analyze a third group of patientspresenting with the VSS phenotype but for whom HPPDcould not be excluded These participants were mostly malewhich might be due to substance use being generally morecommon in men than women28 and exhibited a later onset ofVS symptoms in most cases with an abrupt start Howeverthey fulfilled all remaining criteria for the syndrome and didnot differ from the main VSS group with regard to clinical VScharacteristics Studies with confirmed HPPD would benecessary to shed more light on the interesting overlap be-tween VSS and drug intake which may indeed representdifferent aspects of a same disorder or 2 distinct conditionswith shared pathophysiologic mechanism Our data do sug-gest that VSS itself is not part of HPPD but rather that HPPDcan manifest in the VSS clinical spectrum
There are important limitations to this work largely centeredaround recruitment bias First recruited patients had con-tacted the group directly to be involved in research this islikely to have selected participants at the more severe end ofthe clinical spectrum Nonetheless most participants werenot seeking medical help when they contacted us statingthat their primary reason for contact was simple curiosityabout their disorder The fact that access to the study wassolely through the internet might have also biased towarda younger population The absence of an objective measureof clinical severity is another limitation to this study suchmeasures are yet to be developed Finally the methodologywas based on questionnaires completed remotely and reliedheavily on patient participation The absence of a structuredinterview conducted by a physician might have hindered theclinical description in some cases However using a web-based survey allowed broader geography that simply is notfeasible with in-person approaches Moreover the web-based approach guaranteed the largest possible participationin the study which again would be very challenging if allcases were to have had telephone interviews The combi-nation of detailed phenotyping in person and hypothesistesting using broad internet-based samples offers a powerfultool particularly when studying a relatively poorly charac-terized condition
NeurologyorgN Neurology | Volume 94 Number 6 | February 11 2020 e573
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
VSS is a geographically widely distributed neurologic disorderthat can be consistently defined by the current diagnosticcriteria VSS does not manifest with specific endophenotypesand likely represents a clinical continuum with patientsranging in different degrees of severity On the severe end ofthe spectrum VS is more likely to present with its commoncomorbid conditions migraine and tinnitus VSS is in-dependent of the use of hallucinogenic substances althoughHPPD can manifest in the VS spectrum In the future furtherstudies are needed to enhance our understanding of the un-derlying neurobiology of VS and consequently to move to-ward an era of better management of the condition
AcknowledgmentThe authors thank the thousands of patients who have takenpart in the survey and who have volunteered to share theirsymptoms and experience This study would not have beenpossible without their participation and support
Study fundingThe study was supported by crowd funding from the self-helpgroup for VS Eye On Vision Foundation and funding from theVisual Snow Initiative and the SLaMBiomedical ResearchCentre
DisclosureThe authors report no disclosures relevant to the manuscriptGo to NeurologyorgN for full disclosures
Publication historyReceived by Neurology February 11 2019 Accepted in final formAugust 28 2019
References1 Schankin CJ Maniyar FH Digre KB Goadsby PJ ldquoVisual snowrdquo a disorder distinct
from persistent migraine aura Brain 20141371419ndash14282 Puledda F Schankin C Digre K Goadsby PJ Visual snow syndrome what we know
so far Curr Opin Neurol 20173152ndash583 Liu GT Schatz NJ Galetta SL Volpe NJ Skobieranda F Kosmorsky GS Persistent
positive visual phenomena in migraine Neurology 199545664ndash6684 Headache Classification Committee of the International Headache Society The In-
ternational Classification of Headache Disorders 3rd edition Cephalalgia 2018381ndash211
5 Abraham HD Visual phenomenology of the LSD flashback Arch Gen Psychiatry198340884ndash889
6 Schankin CJ Maniyar FH Sprenger T Chou DE Eller M Goadsby PJ The relationbetween migraine typical migraine aura and ldquovisual snowrdquo Headache 201454957ndash966
7 Simpson JC Goadsby PJ Prabhakar P Positive persistent visual symptoms (visualsnow) presenting as a migraine variant in a 12-year-old girl Pediatr Neurol 201349361ndash363
8 Bessero AC Plant GT Should ldquovisual snowrdquo and persistence of after-images berecognised as a new visual syndrome J Neurol Neurosurg Psychiatry 2014851057ndash1058
9 Unal-Cevik I Yildiz FG Visual snow in migraine with aura further characterization bybrain imaging electrophysiology and treatment case report Headache 2015551436ndash1441
10 Lauschke JL Plant GT Fraser CL Visual snow a thalamocortical dysrhythmia of thevisual pathway J Clin Neurosci 201628123ndash127
11 Rastogi RG VanderPluym J Lewis KS Migrainous aura visual snow and ldquoAlice inWonderlandrdquo syndrome in childhood Semin Pediatr Neurol 20162314ndash17
12 Fraser CL White OB Therersquos something in the air Surv Ophthalmol 201864729ndash733
13 Denuelle M Boulloche N Payoux P Fabre N Trotter Y Geraud G A PET study ofphotophobia during spontaneous migraine attacks Neurology 201176213ndash218
14 Shargorodsky J Curhan GC Farwell WR Prevalence and characteristics of tinnitusamong US adults Am J Med 2010123711ndash718
15 Wu BP Searchfield G Exeter DJ Lee A Tinnitus prevalence in New Zealand N ZMed J 201512824ndash34
16 Kim HJ Lee HJ An SY et al Analysis of the prevalence and associated risk factors oftinnitus in adults PLoS One 201510e0127578
17 van den Berge MJC Free RH Arnold R et al Cluster analysis to identify possiblesubgroups in tinnitus patients Front Neurol 20178115
18 Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dys-rhythmia a neurological and neuropsychiatric syndrome characterized by magneto-encephalography Proc Natl Acad Sci USA 19999615222ndash15227
19 De Ridder D Vanneste S Langguth B Llinas R Thalamocortical dysrhythmiaa theoretical update in tinnitus Front Neurol 20156124
20 Coppola G Ambrosini A Di Clemente L et al Interictal abnormalities of gammaband activity in visual evoked responses in migraine an indication of thalamocorticaldysrhythmia Cephalalgia 2007271360ndash1367
21 Coppola G Pierelli F Schoenen J Is the cerebral cortex hyperexcitable or hyperre-sponsive in migraine Cephalalgia 2007271427ndash1439
22 Eren O Rauschel V Ruscheweyh R Straube A Schankin CJ Evidence of dysfunctionin the visual association cortex in visual snow syndrome Ann Neurol 201884946ndash949
23 Puledda F Ffytche D OrsquoDaly O Goadsby PJ Imaging the Visual Network in theMigraine Spectrum Frontiers in Neurology 2019101325
24 American Psychiatric Association Diagnostic and Statistical Manual of Mental Dis-orders 5th ed ArlingtonAmerican Psychiatric Association 2013
25 Halpern JH Pope HG Jr Hallucinogen persisting perception disorder what do weknow after 50 years Drug Alcohol Depend 200369109ndash119
26 Martinotti G Santacroce R Pettorruso M et al Hallucinogen persisting perceptiondisorder etiology clinical features and therapeutic perspectives Brain Sci 20188E47
27 Halpern JH Lerner AG Passie T A review of hallucinogen persisting perceptiondisorder (HPPD) and an exploratory study of subjects claiming symptoms of HPPDCurr Top Behav Neurosci 201836333ndash360
28 Smith K Gender differences in primary substance of abuse across age groups In TheCBHSQ Report Rockville Substance Abuse and Mental Health Services Adminis-tration 20131ndash8
Appendix Authors
Name Location Role Contribution
FrancescaPuleddaMD
KingrsquosCollegeLondon UK
Author Design and conceptualizedstudy acquired and analyzedthe data performed statisticalanalysis drafted themanuscript for intellectualcontent
ChristophSchankinMD
University ofBernSwitzerland
Author Interpreted the data revisedthe manuscript for intellectualcontent
Peter JGoadsbyMD PhD
KingrsquosCollegeLondon UK
Author Conceptualized studyinterpreted the data revisedthe manuscript for intellectualcontent
e574 Neurology | Volume 94 Number 6 | February 11 2020 NeurologyorgN
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
DOI 101212WNL0000000000008909202094e564-e574 Published Online before print January 15 2020Neurology
Francesca Puledda Christoph Schankin and Peter J GoadsbyVisual snow syndrome A clinical and phenotypical description of 1100 cases
This information is current as of January 15 2020
ServicesUpdated Information amp
httpnneurologyorgcontent946e564fullincluding high resolution figures can be found at
This article has been cited by 1 HighWire-hosted articles
Subspecialty Collections
httpnneurologyorgcgicollectiontinnitusTinnitus
httpnneurologyorgcgicollectionmigraineMigrainefollowing collection(s) This article along with others on similar topics appears in the
Permissions amp Licensing
httpwwwneurologyorgaboutabout_the_journalpermissionsits entirety can be found online atInformation about reproducing this article in parts (figurestables) or in
Reprints
httpnneurologyorgsubscribersadvertiseInformation about ordering reprints can be found online
ISSN 0028-3878 Online ISSN 1526-632XWolters Kluwer Health Inc on behalf of the American Academy of Neurology All rights reserved Print1951 it is now a weekly with 48 issues per year Copyright Copyright copy 2020 The Author(s) Published by
reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology
