ARTICLE OPEN ACCESS

Clinical risk factors in SUDEPA nationwide population-based case-control study

Olafur Sveinsson MD MSc Tomas Andersson BSc Peter Mattsson MD PhD Sofia Carlsson PhD and

Torbjorn Tomson MD PhD

Neurologyreg 202094e419-e429 doi101212WNL0000000000008741

Correspondence

Dr Sveinsson

olafursveinssonsllse

AbstractObjectiveWe conducted a nationwide case-control study in Sweden to test the hypothesis that specificclinical characteristics are associated with increased risk of sudden unexpected death in epilepsy(SUDEP)

MethodsThe study included 255 SUDEP cases (definite and probable) and 1148 matched controlsClinical information was obtained frommedical records and the National Patient Register Theassociation between SUDEP and potential risk factors was assessed by odds ratios (ORs) and95 confidence intervals (CIs) and interaction assessed by attributable proportion due tointeraction (AP)

ResultsExperiencing generalized tonic-clonic seizures (GTCS) during the preceding year was asso-ciated with a 27-fold increased risk (OR 2681 95 CI 1486ndash4838) whereas no excess riskwas seen in those with exclusively non-GTCS seizures (OR 115 95 CI 054ndash4838) Thepresence of nocturnal GTCS during the last year of observation was associated with a 15-foldrisk (OR 1531 95 CI 957ndash2447) Living alone was associated with a 5-fold increased risk ofSUDEP (OR 501 95 CI 293ndash857) and interaction analysis showed that the combination ofnot sharing a bedroom and having GTCS conferred an OR of 6710 (95 CI 2966ndash15188)with AP estimated at 069 (CI 053ndash085) Among comorbid diseases a previous diagnosis ofsubstance abuse or alcohol dependence was associated with excess risk of SUDEP

ConclusionsIndividuals with GTCS who sleep alone have a dramatically increased SUDEP risk Our resultsindicate that 69 of SUDEP cases in patients who have GTCS and live alone could beprevented if the patients were not unattended at night or were free from GTCS

RELATED ARTICLE

Patient PageSudden unexpected deathin epilepsy Assessing therisk factors

Page e436

MORE ONLINE

CME CourseNPuborgcmelist

From the Department of Neurology (OS TT) Karolinska University Hospital Department of Clinical Neuroscience (OS TT) and Institute of Environmental Medicine (TA SC)Karolinska Institutet Center for Occupational and Environmental Medicine (TA) Stockholm County Council and Department of Neuroscience (PM) University of Uppsala Sweden

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 Charge was funded by CURE foundation

This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 40 (CC BY-NC-ND) which permits downloadingand sharing the work provided it is properly cited The work cannot be changed in any way or used commercially without permission from the journal

Copyright copy 2019 The Author(s) Published by Wolters Kluwer Health Inc on behalf of the American Academy of Neurology e419

Sudden unexpected death in epilepsy (SUDEP) is the mostimportant epilepsy-related cause of death ranking secondonly to stroke among neurologic diseases in terms of po-tential years of life lost1 Several case-control studies haveattempted to identify risk factors for SUDEP2ndash5 to providea basis for an individualized risk assessment By pooling datafrom 4 such studies frequency of generalized tonic-clonicseizures (GTCS) in particular but also the duration of ep-ilepsy young age at epilepsy onset and male sex wereidentified as risk factors6 However a recent systematic re-view concluded that the frequency of GTCS was the onlyrisk factor identified with a high level of confidencewhereas eg lack of nighttime supervision and absence ofnocturnal listening device were risk factors with moderateconfidence7 Other risk factors including young age at ep-ilepsy onset long duration of epilepsy focal epilepsy andintellectual disability have been proposed in individualstudies8 but the evidence was considered low in the sys-tematic review7 The uncertainty can be attributed tomethodologic limitations such as small numbers and se-lected study populations affecting generalizability2ndash5 Dif-ferences in definitions of potential risk factors have alsohampered pooling of data67 To guide patient counselingand for the development of effective SUDEP preventionsthere is still need for large high-quality studies to elucidateSUDEP risk factors7 Therefore we analyzed the risk ofSUDEP in relation to a range of potential risk factors ina large nationwide population-based case-control study inSweden utilizing data from individual medical records andnational registries

MethodsSUDEP definition and classificationSUDEP is defined as sudden unexpected witnessed orunwitnessed nontraumatic and nondrowning death ofpatients with epilepsy with or without evidence of a seizureexcluding documented status epilepticus and in whompostmortem examination does not reveal a structural ortoxicologic cause for death9 In the present study we clas-sified SUDEP cases according to Anneger10 criteria Thisclassification was selected to facilitate comparison since ithas been used in most previous studies2ndash5 SUDEP caseswere divided into 3 subgroups based on the certainty of thediagnosis (1) definite SUDEP when all clinical criteria aremet and an autopsy is performed that reveals no alternativecause of death (2) probable SUDEP when all clinical criteriaare met but no autopsy is performed and (3) possible

SUDEP when SUDEP cannot be ruled out but there isinsufficient evidence regarding the circumstances of thedeath and no autopsy is performed10

Study populationThe Swedish National Patient Register (SNPR) contains allpatients hospitalized (starting in 1968 with total nationalcoverage from 1987) or managed in hospital-based ambula-tory care (since 2001) in Sweden11 Each individualrsquos out-patient visit or hospital discharge diagnosis (ICD code) islinked with a unique personal identification number Weidentified all persons who at some point during 1998ndash2005were registered in the SNPRwith an ICD-10 code for epilepsy(G40) (n = 78424) and alive on June 30 2006 (n = 60952)This constituted our study population

CasesDuring follow-up from July 1 2006 to December 31 20119605 deaths were identified by linkage to the NationalCause of Death Registry (ICD-10 classified since 1994)12

Eligible SUDEP cases were all deaths with epilepsy men-tioned on the death certificate (n = 1276) together will allindividuals who died during 2008 (n = 1890) (figure 1) Wepreviously conducted a study of the incidence of SUDEPduring 200813 which is why all deaths in the study pop-ulation were reviewed that particular year All death certifi-cates were reviewed by one neurologist (OS) Obviousnon-SUDEP deaths such as cancer terminal illness post-mortem confirmed pneumonia stroke or myocardial in-farction were excluded from further analysis based on theinformation in the death certificates (figure 1) This processconsidered all information on the death certificate post-mortem results and whether the patient died in the hospitalFor the remaining cases where SUDEP could potentially bethe cause of death (n = 1373) patient records from familyphysicians hospital records nursing homes or other insti-tutions police records and autopsy records were reviewed(OS) and all information was extracted using a standard-ized protocol Emphasis was on attaining the doctorrsquos orpolice report regarding circumstances surrounding thedeath including documented interviews with eyewitnessescaregivers and relatives All information was reviewed by 2neurologists (OS and TT) and classification of the caseswas made through consensus From patient records wedetermined if the patients met the criteria for a diagnosis ofepilepsy according to the definition of the InternationalLeague Against Epilepsy14 In the end 255 definite (n = 167)and probable (n = 88) cases according to the Annegerclassification were found and served as cases for this study

GlossaryAED = antiepileptic drug AP = proportion attributable to interaction CI = confidence interval GTCS = generalized tonic-clonic seizures ICD = International Classification of Diseases LISA = longitudinal integration database for health insuranceand labor market studiesOR = odds ratio SNPR = Swedish National Patient Register SUDEP = sudden unexpected death inepilepsy VNS = vagus nerve stimulation

e420 Neurology | Volume 94 Number 4 | January 28 2020 NeurologyorgN

(figure 1) Possible SUDEP cases (n = 73) were not used inthis study

ControlsFrom the study population the National Board of Health andWelfare randomly selected 5 epilepsy controls (n = 1275) foreach person with SUDEP of the same sex who were alive atthe casersquos time of death which served as an index date for thecontrols For these controls we requested patient records

from caregivers across the country and attained records for1232 (97) individuals Of these 84 (68) were judged notto have epilepsy This left 1148 individuals who served ascontrols in the present study (figure 1)

Information from patient recordsFor all cases and controls we used patient records to collectinformation on age sex and living condition (living alone orwith others including parents partners children and siblings

Figure 1 Flow chart describing the selection process

SUDEP = sudden unexpected death in epilepsy

NeurologyorgN Neurology | Volume 94 Number 4 | January 28 2020 e421

and if sharing a bedroom) If cases or controls were married orhad a partner they were classified as sharing a bedroom if nototherwise explicitly stated Further information was collectedon epilepsy onset duration of epilepsy type of epilepsy eti-ology15 history of tonic-clonic seizures (in this context in-cluding both generalized tonic-clonic seizures and focal tobilateral tonic-clonic seizures in accordance with most previouscase-control studies of SUDEP)14 presence and frequency oftonic-clonic nocturnal seizures during the last year of obser-vation presence of other seizures during the last year of

observation history of nocturnal seizures history of tonic-clonic nocturnal seizures presence of tonic-clonic nocturnalseizures during the last year of observation intellectual dis-ability antiepileptic drug (AED) treatment and whether thepatient had undergone epilepsy surgery or had ongoing treat-ment with vagus nerve stimulation (VNS)

Information from national registriesInformation on psychiatric comorbidity pulmonary diseaseand cardiovascular disease was obtained from ICD codes in

Table 1 Demographic and clinical characteristics of cases and controls

All Men Women

Cases Controls Cases Controls Cases Controls

No () 255 1148 154 (604) 680 (592) 101 (396) 468 (408)

Age at death yindex mean(range)

47 (4ndash92) 39 (3ndash94) 48 (4ndash92) 39 (3ndash93) 45 (5ndash88) 40 (3ndash94)

Age at epilepsy diagnosis y mean(range)

224 (0ndash86) 200 (0ndash86) 237 (0ndash86) 200 (0ndash84) 206 (0ndash84) 200 (0ndash86)

Duration of epilepsy y mean under(range)

24 (1ndash81) 20 (1ndash78) 24 (2ndash70) 19 (1ndash76) 24 (1ndash81) 21 (2ndash78)

Type of epilepsy n ()

Generalized 37 (145) 267 (233) 15 (97) 146 (213) 22 (218) 121 (261)

Focal 186 (730) 794 (693) 117 (760) 478 (700) 68 (673) 316 (681)

Focal and generalized 10 (40) 31 (27) 6 (39) 20 (30) 4 (40) 11 (24)

Unknown 22 (86) 56 (49) 15 (97) 40 (58) 7 (69) 16 (34)

Causes of epilepsy n ()

Genetic 48 (188) 303 (264) 21 (136) 164 (240) 26 (256) 139 (300)

Structural 129 (506) 444 (387) 85 (552) 279 (408) 26 (256) 165 (356)

Infectious 12 (47) 42 (37) 8 (52) 28 (41) 4 (40) 14 (30)

Metabolic 2 (08) 9 (08) 1 (06) 7 (10) 1 (10) 2 (04)

Autoimmune 2 (08) 10 (09) 1 (06) 4 (06) 1 (10) 6 (129)

Unknown 66 (259) 359 (313) 39 (253) 214 (313) 27 (267) 145 (312)

Living conditions n ()

Sharing household and bedroom 32 (125) 391 (341) 19 (123) 210 (309) 13 (129) 181 (387)

Sharing household but not bedroom 49 (192) 398 (347) 27 (175) 252 (371) 22 (218) 146 (312)

Not sharing household 174 (682) 304 (265) 108 (701) 177 (260) 66 (653) 127 (271)

Unknown 0 55 (48) 0 41 (60) 0 14 (30)

Highest education n ()

Postsecondary education 26 (102) 168 (146) 17 (110) 96 (141) 9 (89) 72 (154)

High schoolsecondaryeducation

86 (337) 359 (313) 53 (344) 201 (296) 33 (327) 158 (338)

Primary education 86 (337) 297 (258) 56 (364) 171 (251) 30 (297) 126 (269)

Missing educationa 57 (224) 324 (282) 28 (182) 212 (312) 29 (287) 112 (239)

a Younger than 16 and those who did not attend regular school

e422 Neurology | Volume 94 Number 4 | January 28 2020 NeurologyorgN

the national patient registry (from 1997 to death or index date)From the longitudinal integration database for health insuranceand labor market studies (LISA) which holds annual registerssince 1990 and includes all individuals 16ndash74 years of ageinformation on highest educational level was attained16 In theLISA registry this information is recorded as missing forindividuals below 16 years and for those who did not attendregular school due to intellectual disability

StatisticsCharacteristics were expressed as mean (range) or proportionThe association between SUDEP and potential risk factors wasestimated by odds ratios (ORs) with 95 confidence intervals(CIs) calculated by conditional logistic regression to accountfor matching by sex and calendar time As the control partic-ipants were sampled with an incidence density method theORs can be interpreted as incidence rate ratios17 In model 1

Table 2 Sudden unexpected death in epilepsy in relation to clinical characteristics living conditions and education

Cases Controls Model 1a Model 2b Model 3c

Age at onset y

lt18 140 724 115 (081ndash163) 063 (039ndash101) 060 (034ndash105)

18ndash65 (ref) 96 344 1 1 1

Over 65 14 62 061 (027ndash138) 055 (020ndash156) 065 (021ndash205)

Duration of epilepsy y

le15 (ref) 102 586 1 1 1

gt15 153 548 122 (089ndash167) 071 (046ndash108) 081 (050ndash131)

Type of epilepsy

Generalized (ref) 37 267 1 1 1

Focal 186 794 148 (100ndash220) 162 (098ndash266) 134 (077ndash233)

Focal and generalized 10 31 351 (155ndash796) 205 (077ndash550) 142 (049ndash415)

Unknown 22 56 243 (129ndash457) 306 (136ndash690) 351 (144ndash855)

Cause of epilepsyd

Genetic 48 303 084 (035ndash200) 084 (033ndash219) 083 (029ndash241)

Structural 129 444 136 (056ndash327) 138 (052ndash364) 120 (041ndash352)

Infectious 12 42 137 (046ndash402) 089 (027ndash291) 111 (030ndash413)

Metabolic 2 9 139 (028ndash691) 124 (017ndash891) 209 (031ndash1406)

Autoimmune 2 10 100 (018ndash576) 289 (039ndash2168) 241 (021ndash2751)

Unknown 66 359 089 (036ndash222) 117 (043ndash321) 107 (035ndash325)

Living conditions

Sharing household and bedroom (ref) 32 391 1 1 1

Sharing household but not bedroom 49 398 243 (136ndash432) 167 (08722) 228 (114ndash458)

Not sharing household 174 359e 611 (404ndash922) 409 (249ndash673) 501 (293ndash857)

Highest education

Postsecondary education (ref) 26 168 1 1 1

High school educationsecondary education 86 359 167 (103ndash272) 129 (068ndash242) 159 (078ndash327)

Primary education 86 297 206 (125ndash339) 112 (059ndash215) 121 (058ndash256)

Values are no or odds ratio (95 confidence interval)a Adjusted for age and sex (matching variable)b Adjusted for age sex and generalized tonic-clonic seizures frequencyc Adjusted for age sex generalized tonic-clonic seizures frequency and nocturnal generalized tonic-clonic seizures last year of observation living conditions(except in the analysis of living conditions) and antiepileptic drugsd Categories are not mutually exclusivee Includes 55 individuals with unknown living conditions

NeurologyorgN Neurology | Volume 94 Number 4 | January 28 2020 e423

OR was adjusted for age and sex (matching variable) Model 2included additional adjustments for GTCS frequency andmodel 3 included the same covariates as model 2 together withnocturnal GTCS last year of observation living conditions andAEDs In the Results all results are presented from model 3unless stated otherwise Interaction between GTCS during lastyear of observation (yesno) and sharing a bedroom (yesno)

defined as departure from additivity of effects was assessedwith the proportion attributable to interaction (AP)18 Theformula for AP is (OR11 minus OR10 minus OR01 + 1)OR11 whereOR11 indicates doubly exposed (having GTCS and sleepingalone) and OR01 or OR10 indicate either exposure (sleepingalone or having GTCS) The reference group is those withneither exposure and the ORs were adjusted for age and sex

Table 3 Sudden unexpected death in epilepsy in relation to type and frequency of seizures and treatment

Cases Controls Model 1a Model 2b Model 3c

History of GTCS

No (ref) 4 174 1 1

Yes 251 943 1056 (386ndash2886) 960 (344ndash2682)

Seizures during preceding year

No (ref) 26 577 1 1

Yes but not GTCS 12 290 097 (048ndash196) 115 (054ndash246)

GTCS 217 280 2270 (1372ndash3755) 2681 (1486ndash4838)

GTCS frequency during preceding year

0 (ref) 38 865 1 1

1ndash3 106 150 1951 (1194ndash3188) 2214 (1274ndash3846)

4ndash10 50 42 2824 (1536ndash5192) 3187 (1595ndash6367)

gt10 61 88 2638 (1462ndash4761) 2970 (1504ndash5863)

History of nocturnal seizures

No (ref) 63 711 1 1

Yes non-GTCS 2 102 023 (006ndash098) 027 (006ndash115)

Yes GTCS 190 335 844 (591ndash1204) 904 (608ndash1345)

Nocturnal GTCS during preceding year

No (ref) 145 1049 1 1

Yes 110 99 1298 (861ndash1956) 1531 (957ndash2447)

AED treatment

No (ref) 19 144 1 1 1

Monotherapy 120 546 127 (074ndash217) 039 (020ndash077) 047 (023ndash094)

Polytherapy 115 458 167 (098ndash284) 028 (014ndash057) 031 (015ndash066)

Epilepsy surgery

No (ref) 242 1098 1 1 1

Yes 13 50 127 (066ndash244) 089 (039ndash200) 077 (031ndash192)

VNS

No (ref) 244 1098 1 1 1

Yes 11 50 129 (065ndash257) 050 (022ndash111) 041 (017ndash098)

Abbreviations AED = antiepileptic drug GTCS = generalized tonic-clonic seizures VNS = vagus nerve stimulationValues are no or odds ratio (95 confidence interval)a Adjusted for age and sex (matching variable)b Adjusted for age sex and GTCS frequencyc Adjusted for age sex GTCS frequency and nocturnal GTCS last year of observation (except in the analyses of seizures) living conditions and AEDs (except inthe analysis of AED treatment)

e424 Neurology | Volume 94 Number 4 | January 28 2020 NeurologyorgN

(matching variable) Statistical Analysis Software (SAS) 94(SAS Institute Cary NC) was used for all analyses

Standard protocol approvals registrationsand patient consentsThe study was approved by the Ethics Committee of Kar-olinska Institutet which granted that individual informedconsent was not needed

Data availabilityAnonymized data will be shared by request from qualifiedinvestigators

ResultsCharacteristics of cases and controls are summarized in table1 Among the 255 SUDEP decedents 604 were men anddue to matching a similar male predominance was seenamong controls Mean age at diagnosis was 224 years for theSUDEP decedents and 20 years for controls and the dece-dents tended to have a slightly longer duration of epilepsy (24vs 20 years) The majority of decedents had focal epilepsy(730) and of structural origin (506) Comparing casesand controls indicated small differences in the type and causesof epilepsy but low education was slightly more commonamong cases (table 1) Decedents with SUDEP lived alone toa larger extent than controls 682 vs 265 and even if they

shared their household they were less likely than controls toshare a bedroom Generalized and genetic epilepsy was lesscommon among men with SUDEP compared to women withSUDEP and male and female controls In a similar fashionmen with SUDEP had a slightly higher age at epilepsy onsetand more often had focal and structural epilepsy

Clinical characteristics living conditionseducation and risk of SUDEPPreviously proposed risk factors such as young age at epilepsyonset longer duration of epilepsy and structural etiology werenot associated with SUDEP after adjustment for GTCS fre-quency (table 2) As for the type of epilepsy no excess risk wasseen in individuals with focal or focal and generalized epilepsycompared to generalized epilepsy after adjustment for GTCSfrequency but epilepsy of unknown type remained associatedwith SUDEP Compared with sharing a bedroom sharinghousehold but not bedroom was associated with a twofoldincreased risk and living alone was associated with a fivefoldincreased risk of SUDEP (OR 501 95 CI 293ndash857) evenafter adjustment for GTCS frequency and other covariates(table 2) No association between level of education andSUDEP was seen after adjustment for GTCS frequency

Seizures treatment and risk of SUDEPA history of GTCSwas associated with a tenfold increased riskof SUDEP (OR 960 95 CI 344ndash2682) (table 3) Only 4(16) SUDEP cases did not have a history of GTCS

Table 4 Sudden unexpected death in epilepsy in relation to comorbidity (yesno)

All no cases No controls Model 1a Model 2b Model 3c

Mental health disorder 128 470 169 (128ndash225) 085 (059ndash123) 080 (054ndash119)

Substance abuse 34 53 257 (163ndash405) 201 (110ndash366) 207 (107ndash401)

Alcohol dependence 26 34 299 (174ndash512) 242 (117ndash501) 230 (102ndash521)

Depression 20 74 102 (061ndash172) 123 (064ndash236) 099 (049ndash201)

Mood (affective disorders) 23 82 107 (066ndash175) 130 (069ndash245) 109 (055ndash217)

Anxiety disorder 28 81 144 (091ndash229) 142 (079ndash252) 142 (076ndash267)

Intellectual disabilityd 97 323 248 (179ndash342) 107 (069ndash166) 090 (054ndash151)

Diseases of the nervous system excluding epilepsy 91 379 115 (086ndash153) 073 (050ndash106) 075 (050ndash111)

Diseases of the circulatory system 88 327 094 (066ndash134) 072 (046ndash112) 076 (046ndash127)

Cerebrovascular disease 45 145 113 (075ndash170) 109 (064ndash185) 106 (059ndash191)

Ischemic heart disease 16 78 065 (035ndash120) 059 (027ndash127) 071 (030ndash170)

Heart failure 10 29 135 (061ndash299) 105 (036ndash310) 123 (039ndash392)

Myocarditis cardiomyopathy arrhythmias 25 78 119 (071ndash200) 108 (055ndash211) 125 (061ndash254)

Chronic lower respiratory diseases 30 106 151 (097ndash236) 095 (054ndash168) 104 (055ndash198)

Values are no or odds ratio (95 confidence interval)a Adjusted for age and sex (matching variable)b Adjusted for age sex and generalized tonic-clonic seizures frequencyc Adjusted for age sex generalized tonic-clonic seizures frequency and nocturnal generalized tonic-clonic seizures last year of observation living conditionsand antiepileptic drugsd Information extracted from patient records

NeurologyorgN Neurology | Volume 94 Number 4 | January 28 2020 e425

compared to 151 among the controls In those experiencingGTCS during the last year of observation the risk was in-creased 27-fold (OR 2681 95CI 1486ndash4838) Having 1ndash3GTCS in the previous year was associated with a 22-fold risk(OR 2214 95 CI 1274ndash3846) and having 4ndash10 GTCSincreased the risk to 32-fold (OR 3187 95 CI1595ndash6367) while we did not see a further risk increasewhen the GTCS exceeded 10 during the preceding year

History of nocturnal GTCSwas associated with a ninefold risk(OR 904 95CI 608ndash1345) of SUDEP and the presence ofnocturnal GTCS during last year of observation with a 15-fold risk (OR 1531 95 CI 957ndash2447) In individuals ex-periencing exclusively non-GTCS during the preceding yearno excess risk of SUDEP was seen (OR 115 95 CI054ndash246) Both monotherapy and polytherapy were asso-ciated with a reduced risk of SUDEP after adjusting for GTCSfrequency and other covariates (table 3) Previous epilepsysurgery was not associated with SUDEP while vagus nervestimulation was associated with a 59 reduced SUDEP riskafter adjustment for covariates

Comorbidity and risk of SUDEPAmong comorbid diseases a twofold increased risk of SUDEPwas seen in individuals with a previous diagnosis of substanceabuse or alcohol dependence (table 4) Mental health dis-orders and intellectual disability was not associated with in-creased SUDEP risk once we adjusted for frequency of GTCS

Interaction between living conditionsand GTCSTable 5 displays the risk of SUDEP in relation to the com-bination of living conditions and GTC seizure frequencyIndividuals who experienced ge4 GTCS had 20 times in-creased SUDEP risk if they shared a bedroom with someone34 times increased risk if they shared household but notbedroom and an 82 times increased risk if they lived alone(table 5) Interaction analysis indicated that the combinationof having at least one GTCS and not sharing a bedroom with

someone conferred a 67-fold increased risk of SUDEP com-pared to not having GTCS and sharing a bedroom AP wasestimated at 069 (053ndash085) (figure 2)

DiscussionOur results confirm the conclusion from previous case-control studies2ndash6 and the recent systematic review7 that thepresence and frequency of GTCS is by far the most importantrisk factor for SUDEP Importantly we could demonstratethat having seizures other than GTCS even at night did notincrease the risk for SUDEP Living alone especially notsharing a bedroom with anyone was associated with a sub-stantially increased risk of SUDEP and moreover the com-bination of frequent GTCS and sleeping alone dramaticallyincreased the risk of SUDEP Taking AEDs as monotherapyor polytherapy and treatment with VNS was associated withsignificantly reduced risk of SUDEP whereas substance abuseand alcohol dependence appeared to increase the risk Anumber of previously proposed risks were not associated withSUDEP once we adjusted for GTCS frequency

We saw an incremental risk increase from no seizures up to4ndash10 GTCS (table 3) largely in line with the previous pooledanalysis of case-control studies6 and the systematic review7

although with somewhat higher risk estimates in our analysisOne explanation why having more than 10 GTCS per year didnot increase the risk further could be that the recording ofseizure counts in the medical records may be less precise inpatients with a high frequency of seizures

Interestingly we did not observe an increased risk of SUDEPin patients with only non-GTCS To our knowledge this hasnot been specifically analyzed before2ndash7 It was possible toextract this information from the extensive records we had onboth cases and controls This novel finding is important in-formation when counseling the individual patient and insetting treatment goals For example improved treatmentwhere GTCS are converted into non-GTCS could reduce the

Table 5 Sudden unexpected death in epilepsy in relation to the combination of generalized tonic-clonic seizures (GTCS)and living conditions

Living conditions

GTCS frequency during preceding year

No GTCS 1ndash3 GTCS ge4 GTCS

No casescontrols OR (95 CI)

No casescontrols OR (95 CI)

No casescontrols OR (95 CI)

Sharing household andbedroom

8318 1 (ref) 1650 1589(605ndash4178)

821 1985(637ndash6184)

Sharing household but notbedroom

4287 110(030ndash402)

1850 3134(1122ndash8753)

2761 3355(1221ndash9218)

Not sharing household 26260 392(169ndash913)

7250 6590(2772ndash15665)

7648 8181(3360ndash19915)

Abbreviations CI = confidence interval OR = odds ratioAdjusted for age and sex (matching variable)

e426 Neurology | Volume 94 Number 4 | January 28 2020 NeurologyorgN

SUDEP risk for the individual patient Even though there area few reports of witnessed SUDEP without a preceding sei-zure or following a non-GTCS this seems to be rare1920 Inthe MORTEMUS study of SUDEP during video-EEG mon-itoring all cases followed in the aftermath of a GTCS21

Nocturnal GTCS were associated with an increased risk ofSUDEP This fits with previous observations22 includinga recent study on institutionalized individuals with epilepsycompared to controls living in the same institution23 Onenovelty in our study was to analyze separately nocturnal non-GTCS demonstrating that such seizures were not associatedwith SUDEP

As in previous studies3ndash6 there was a trend towards increasedrisk in focal epilepsy which however disappeared afteradjusting for other risk factors especially frequency of GTCSThe group focal and generalized epilepsy was a risk factorbefore adjusting for GTCS likely reflecting the severity of theepilepsy in this group Interestingly the unknown type of ep-ilepsy remained a risk factor in all models We have no clearexplanation for this except that there could be similarities withthis group and the focal and generalized group where it is oftendifficult to classify the epilepsy due to its complex nature It isalso possible that failure to classify the type of epilepsy may bea reflection of suboptimal epilepsy management which in itselfcan contribute to an increased SUDEP risk

We observed a substantial increase in SUDEP risk for thoseliving alone especially those not sharing a bedroom Ourobservations are in line with a previous report of a protectiveeffect of nighttime supervision regular checks throughout thenight or use of listening devices to detect seizures5 Fur-thermore a recent study from 2 epilepsy residential carehomes reported that SUDEP was more common in the centerwith less supervision at night23 The greatest novelty in our

findings shown with interaction analysis is the supra-additiveincrease in SUDEP risk for individuals having at least oneGTCS during the last year of observation and sleeping aloneThis demonstrates again that unattended GTCS are the mostimportant risk factor in SUDEP24 More than two-thirds of allcases exposed to both GTCS and not sharing a bedroom wouldbe prevented by removal of one of these risk factors Thissuggests that a patient with epilepsy with GTCS should sharea room with someone else whenever possible This can bedifficult to organize but hopefully there will be an improvementin different types of seizure monitoring devices that could alertfamily members or caretakers when a seizure is detected Noprospective studies regarding the effectiveness of seizure mon-itoring devices in preventing SUDEP have been conducted

Other risk factors could be hidden and sleeping alone could bea marker for fewer social connectionsnetworks We foundsubstance abuse to be a risk factor that can be connected toa reduced social network This field needs further research

Early case-control studies identified polytherapy with AEDs asa risk factor for SUDEP246 However with pooled data from4 case-control studies polytherapy was no longer a risk factorafter adjustment for GTCS frequency25 We did find excessrisk in individuals with polytherapy however once we ad-justed for GTCS both monotherapy and polytherapy wasassociated with a reduced risk of SUDEP These observationsare in line with the meta-analysis of placebo-controlled ran-domized add-on trials in refractory epilepsy which showeda substantially lower SUDEP risk among those randomized toadjunctive active treatment compared with placebo26 Amajorlimitation of this meta-analysis however was that adjustmentfor GTCS frequency was not possible Our findings indicatethat AEDs may have a protective effect beyond the seizure-controlling properties These potential mechanisms remain tobe explored

Figure 2 Odds ratio (OR) (95 confidence interval [CI]) of sudden unexpected death in epilepsy by combinations ofgeneralized tonic-clonic seizures (GTCS) and living conditions

AP = attributable proportion due to interaction

NeurologyorgN Neurology | Volume 94 Number 4 | January 28 2020 e427

Several studies have observed a reduced SUDEP risk aftersuccessful epilepsy surgery2728 We could not confirm thesefindings but our analyses were hampered by small numbersTreatment with VNS was associated with a reduced risk ofSUDEP A possible protective effect of VNS has been dis-cussed before29 but our data should be interpreted withcaution given the small numbers

Comorbid mental health disorders have previously been as-sociated with excess risk of SUDEP13 but we did not observean association once GTCS frequency was taken into accountIn line with the pooled analysis6 of previous case-controlstudies substance abuse including alcohol abuse was asso-ciated with an increased risk for SUDEP This should beconsidered when counseling individual patients We detectedno increased risk associated with a medical history of ischemicheart disease heart failure myocarditis cardiomyopathy orarrhythmias Neither was there an increased risk in individualswith a history of other neurologic disorders or those witha history of chronic lower respiratory diseases It is conceiv-able that patients with epilepsy with comorbid cardiovascularand respiratory diseases are more likely to be classified aspossible SUDEP which was not included in our analysis

The strengths of this study are its size the population-basednationwide nature and the fact that the controls came fromthe same population as the cases and furthermore that wewere able to attain records for 97 of the 1275 potentialcontrols In addition the validity of the epilepsy diagnosis wasascertained with chart review and those not meeting theepilepsy criteria were excluded Among the weaknesses arethat patient records have their inherent limitations which canhave an effect on eg the possibility to classify epilepsysyndromes even though we had extensive records for mostcases and controls In addition the authors extracting in-formation were not blinded to the outcome and were awareof previous reports on SUDEP risk factors which may in-troduce bias The information was collected identically usinga standardized protocol for both cases and controls It ispossible that information on living conditions was betterdocumented among cases due to the more extensive recordsin connection with their death However information onliving conditions was missing in only a small fraction of thecontrols (48 n = 55) compared to in none of the SUDEPcases and it is unlikely that this had a major effect on ourresults

Having GTCS nocturnal GTCS and living alone are asso-ciated with markedly increased risk of SUDEP Combininghigh frequency of GTCS and living alone is associated witha dramatically increased SUDEP risk suggesting that un-attended GTCS play a major role The data suggest that bettersupervision is needed for high-risk patients with uncontrolledGTCS However such efforts to reduce SUDEP risks must bebalanced against each patientrsquos right to independence andintegrity which can only be done on an individual basisLately there has been an increasing interest in the use of

seizure detection devices but it remains to be shown if thesecan reduce the SUDEP risk3031 The currently most impor-tant preventive method is to prescribe more effective treat-ments that reduce the occurrence of GTCS Our data suggestthat even a treatment that does not reduce the overall seizurefrequency but that prevents focal seizures from evolving tobilateral tonic-clonic seizures may be beneficial In a sub-sequent analysis we intend to focus in more detail on the roleof drug treatment utilizing data from the Swedish Drug Pre-scription Registry using the same study population

Study fundingThe study was supported by funding from Stockholm CountyCouncil GlaxoSmithKline and Citizens United for Researchin Epilepsy The sponsors had no influence on the conduct ofthe study analysis interpretation writing of the manuscriptor the decision to publish the results

DisclosureO Sveinsson has received grants fromGSK personal fees fromBiogen and honoraria to his institution from Biogen and UCBfor lectures and advisory board outside the submitted work TAndersson and S Carlsson report no disclosures relevant to themanuscript P Mattsson received research support from theUppsala County Council Epilepsifonden and SelanderFoundation T Tomson is an employee of Karolinska Insti-tutet is associate editor of Epileptic Disorders has receivedspeakerrsquos honoraria to his institution from Eisai Sanofi SunPharma UCB and Sandoz and received research support fromStockholmCounty Council EU CURE GSK UCB Eisai andBial Go to NeurologyorgN for full disclosures

Publication historyReceived by Neurology May 4 2019 Accepted in final formAugust 5 2019

Appendix Authors

Name Location Role Contribution

OlafurSveinssonMD MSc

KarolinskaInstitutet

Author Major role in design of study andacquisition of data drafted themanuscript for intellectualcontent

TomasAnderssonBSc

KarolinskaInstitutet

Author Statistical analysis interpretedthe data revised the manuscriptfor intellectual content

PeterMattssonMD PhD

Universityof Uppsala

Author Interpreted the data revised themanuscript for intellectualcontent

SofiaCarlssonPhD

KarolinskaInstitutet

Author Design of study interpreted thedata revised the manuscript forintellectual content

TorbjornTomsonMD PhD

KarolinskaInstitutet

Author Major role in design of studyinterpreted the data revised themanuscript for intellectualcontent

e428 Neurology | Volume 94 Number 4 | January 28 2020 NeurologyorgN

References1 Thurman DJ Hesdorffer DC French JA Sudden unexpected death in epilepsy

assessing the public health burden Epilepsia 2014551479ndash14852 Walczak TS Leppik IE DrsquoAmelio M et al Incidence and risk factors in sudden

unexpected death in epilepsy a prospective cohort study Neurology 200156519ndash525

3 Hitiris N Suratman S Kelly K Stephen LJ Sills GJ Brodie MJ Sudden unexpecteddeath in epilepsy a search for risk factors Epilepsy Behav 200710138ndash141

4 Nilsson L Farahmand BY Persson PG Thiblin I Tomson T Risk factors for suddenunexpected death in epilepsy a casendashcontrol study Lancet 1999353888ndash893

5 Langan Y Nashef L Sander JW Casendashcontrol study of SUDEP Neurology 2005641131ndash1133

6 Hesdorffer DC Tomson T Benn E et al Combined analysis of risk factors forSUDEP Epilepsia 2011521150ndash1159

7 Harden C Tomson T Gloss D et al Practice guideline summary sudden un-expected death in epilepsy incidence rates and risk factors report of the guidelinedevelopment dissemination and implementation Subcommittee of the AmericanAcademy of Neurology and the American Epilepsy Society Neurology 2017881674ndash1680

8 Tomson T Surges R Delamont R Haywood S Hesdorffer DC Who to target insudden unexpected death in epilepsy prevention and how Risk factors biomarkersand intervention study designs Epilepsia 201657(suppl 1)4ndash16

9 Nashef L Sudden unexpected death in epilepsy terminology and definitions Epi-lepsia 199738(suppl 11)6ndash8

10 Annegers IF United States perspective on definitions and classifications Epilepsia199738(suppl 11)9ndash12

11 Ludvigsson JF Andersson E Ekbom A et al External review and validation of theSwedish National Inpatient Register BMC Public Health 201111450

12 Johansson LA Bjorkenstam C Westerling R Unexplained differences betweenhospital and mortality data indicated mistakes in death certification an in-vestigation of 1094 deaths in Sweden during 1995 J Clin Epidemiol 2009621202ndash1209

13 Sveinsson O Andersson T Carlsson S Tomson T The incidence of SUDEP a na-tionwide population-based cohort study Neurology 201789170ndash177

14 Fisher RS Cross JH French JA et al Operational classification of seizure types by theInternational League Against Epilepsy position paper of the ILAE Commission forClassification and Terminology Epilepsia 201758522ndash530

15 Scheffer IE Berkovic S Capovilla G et al ILAE classification of the epilepsiesposition paper of the ILAE Commission for Classification and Terminology Epilepsia201758512ndash521

16 Ludvigsson JF Svedberg P Olen O Bruze G Neovius M The longitudinal integrateddatabase for health insurance and labour market studies (LISA) and its use in medicalresearch Eur J Epidemiol 201934423ndash437

17 Vandenbroucke JP Pearce N Case-control studies basic concepts Int J Epidemiol2012411480ndash1489

18 Andersson T Alfredsson L Kallberg H Zdravkovic S Ahlbom A Calculatingmeasures of biological interaction Eur J Epidemiol 200520575ndash579

19 Sveinsson O Andersson T Carlsson S Tomson T Circumstances of SUDEP a na-tionwide population-based case-series Epilepsia 2018591074ndash1082

20 Lhatoo SD Nei M Raghavan M et al Nonseizure SUDEP sudden unexpected deathin epilepsy without preceding epileptic seizures Epilepsia 2016571161ndash1168

21 Ryvlin P Nashef L Lhatoo SD et al Incidence and mechanisms of cardiorespiratoryarrests in epilepsy monitoring units (MORTEMUS) a retrospective study LancetNeurol 201312966ndash977

22 Lamberts RJ Thijs RD Laffan A Langan Y Sander JW Sudden unexpected death inepilepsy people with nocturnal seizures may be at highest risk Epilepsia 201253253ndash257

23 van der Lende M Hesdorffer DC Sander JW Thijs RD Nocturnal supervision andSUDEP risk at different epilepsy care settings Neurology 201891e1508ndashe1518

24 Devinsky O Hesdorffer DC Thurman DJ Lhatoo S Richerson G Sudden un-expected death in epilepsy epidemiology mechanisms and prevention LancetNeurol 2016151075ndash1078

25 Hesdorffer DC Tomson T Benn E et al ILAE Commission on Epidemiology(Subcommission on Mortality) Do antiepileptic drugs or generalized tonic-clonicseizure frequency increase SUDEP risk A combined analysis Epilepsia 201253249ndash252

26 Ryvlin P Cucherat M Rheims S Risk of sudden unexpected death in epilepsy inpatients given adjunctive antiepileptic treatment for refractory seizures a meta-analysis of placebo-controlled randomised trials Lancet Neurol 201110961ndash968

27 Hennessy MJ Langan Y Elwes RD et al A study of mortality after temporal lobeepilepsy surgery Neurology 1999531276ndash1283

28 Sperling MR Barshow S Nei M Asadi-Pooya AA A reappraisal of mortality afterepilepsy surgery Neurology 2016861938ndash1944

29 Ryvlin P So EL Gordon CM et al Long-term surveillance of SUDEP in drug-resistant epilepsy patients treated with VNS therapy Epilepsia 201859562ndash572

30 Ryvlin P Ciumas C Wisniewski I Beniczky S Wearable devices for sudden un-expected death in epilepsy prevention Epilepsia 201859(suppl 1)61ndash66

31 Rugg-Gunn F Duncan J Hjalgrim H Seyal M Bateman L From unwitnessed fatalityto witnessed rescue nonpharmacologic interventions in sudden unexpected death inepilepsy Epilepsia 201657(suppl 1)26ndash34

NeurologyorgN Neurology | Volume 94 Number 4 | January 28 2020 e429

DOI 101212WNL0000000000008741202094e419-e429 Published Online before print December 12 2019Neurology Olafur Sveinsson Tomas Andersson Peter Mattsson et al

Clinical risk factors in SUDEP A nationwide population-based case-control study

This information is current as of December 12 2019

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 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ServicesUpdated Information amp

httpnneurologyorgcontent944e419fullincluding high resolution figures can be found at

References httpnneurologyorgcontent944e419fullref-list-1

This article cites 31 articles 6 of which you can access for free at

Citations httpnneurologyorgcontent944e419fullotherarticles

This article has been cited by 3 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectiongeneralized_seizuresGeneralized seizures

y_cohort_case_controlhttpnneurologyorgcgicollectionclinical_trials_observational_studClinical trials Observational study (Cohort Case control)

httpnneurologyorgcgicollectioncase_control_studiesCase control studies

httpnneurologyorgcgicollectionall_epilepsy_seizuresAll EpilepsySeizuresfollowing collection(s) This article along with others on similar topics appears in the

Errata

content94104592fullpdf or page

nextAn erratum has been published regarding this article Please see

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 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

ServicesUpdated Information amp

httpnneurologyorgcontent944e419fullincluding high resolution figures can be found at

References httpnneurologyorgcontent944e419fullref-list-1

This article cites 31 articles 6 of which you can access for free at

Citations httpnneurologyorgcontent944e419fullotherarticles

This article has been cited by 3 HighWire-hosted articles

Subspecialty Collections

httpnneurologyorgcgicollectiongeneralized_seizuresGeneralized seizures

y_cohort_case_controlhttpnneurologyorgcgicollectionclinical_trials_observational_studClinical trials Observational study (Cohort Case control)

httpnneurologyorgcgicollectioncase_control_studiesCase control studies

httpnneurologyorgcgicollectionall_epilepsy_seizuresAll EpilepsySeizuresfollowing collection(s) This article along with others on similar topics appears in the

Errata

content94104592fullpdf or page

nextAn erratum has been published regarding this article Please see

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 2019 The Author(s) Published by

reg is the official journal of the American Academy of Neurology Published continuously sinceNeurology

Disputes amp Debates Editorsrsquo ChoiceSteven Galetta MD FAAN Section Editor

Reader response Menarche pregnancies and breastfeeding do notmodify long-term prognosis in multiple sclerosisVilija G Jokubaitis (Melbourne) and Ruth Dobson (London)

Neurologyreg 202094455ndash456 doi101212WNL0000000000009063

We read with interest the article by Zuluaga et al1 which used the uniquely valuable BarcelonaCIS (clinically isolated syndrome) cohort2 However evolving multiple sclerosis (MS) di-agnostic and treatment landscapes must be taken into account when using this cohort to informcurrent practice

Of those included in the analysis1 47did not have a second clinical attack 39did notmeet theMcDonald 2010 criteria and 32 did not meet the Barkhof criteria for the diagnosis ofMS This

Editorsrsquo note Menarche pregnancies and breastfeeding do notmodify long-term prognosis in multiple sclerosisIn the article ldquoMenarche pregnancies and breastfeeding do not modify long-term prog-nosis in multiple sclerosisrdquo Zuluaga et al reported that age at menarche pregnancy beforeor after the diagnosis of clinically isolated syndrome (CIS) and breastfeeding did notsubstantially modify the risk of progressing to clinically definite multiple sclerosis (CDMS)or disability accrual per the Expanded Disability Status Scale (EDSS) in a cohort of 501female participants with CIS In response Drs Jokubaitis and Dobson argued that thepatients with CDMS should be examined separately for the EDSS outcomes becausea substantial proportion of the overall cohort did not have a second clinical attack and didnotmeet either theMcDonald 2010 or Barkhof criteria forMS They seek additional detailsregarding the propensity scorendashmatched score analysis because a smaller number ofmatched pairs could limit the generalizability of the results In addition they noted that theanalyses for the association of pregnancy and breastfeeding on time to EDSS 30 were notadjusted for relapse and that the differences between exclusive breastfeeding and mixedfeeding strategies merit further exploration in prospective studies They also argue that theharmful effects of suspending disease-modifying treatments (DMTs) in those with ag-gressive disease who become pregnant should be considered Responding to these com-ments Drs Tintore et al noted that they built the model for time to EDSS 30 over theCDMS subcohort in addition to providing further details of the propensity scorendashmatchedanalyses They reported additional analyses for the adjusted hazard ratio for pregnancy (butnot for breastfeeding) on considering the annualized relapse rate over the first 3 and 5 yearsof disease and acknowledged that additional details of breastfeeding were unavailableRegarding the problem of suspending DMTs in pregnant patients they noted that they areanalyzing a subgroup of women treated with natalizumab or fingolimod As greaternumbers of young women become eligible for DMTs with more inclusive revisions of theMcDonald criteria neurologists are likely to encounter challenging questions about theassociation of pregnancies and breastfeeding with MS disease activity and the attendantDMT-related dilemmas in their practice

Aravind Ganesh MD DPhil and Steven Galetta MD

Neurologyreg 202094455 doi101212WNL0000000000009064

NeurologyorgN Neurology | Volume 94 Number 10 | March 10 2020 455

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

raises questions about cohort baseline heterogeneity because 2 of the primary outcomemeasuresare confirmed Expanded Disability Status Scale (EDSS) 30 or 60 There is an argument in favorof examining the clinically definite multiple sclerosis (CDMS) cohort separately to the non-CDMS cohort

Regarding the propensity score-matched analyses we are interested to know the matchingstrategy used how many matched pairs were included in this analysis the matching ratio themedian follow-up duration and censoring strategy Only 142 respondents had pregnancies aftera CIS1 it is thus possible that fewer than 142 matched pairs were included limiting the gen-eralizability of these results

It appears that the analyses of the impact of pregnancy and breastfeeding on time to EDSS 30were not adjusted for relapse Relapse particularly early in the disease phase and relapse recoveryare among the strongest predictors of future disability accumulation34

Breastfeeding was studied as both a dichotomous variable (breastfeeding vs not) and a time-dependent event1 However exclusive breastfeeding may be protective in a way that mixedfeeding is not5 A truly prospective design is required to address the subtleties of this question

The authors concluded that MS prognosis is not significantly affected by pregnancy once allother variables are considered1 However in the current era of highly active disease-modifyingtreatment (DMT) use pregnancy does not occur in isolation The potentially harmful effects ofsuspending DMT in those with aggressive disease must be taken into account when discussingfamily planning in MS We look forward to future studies to help answer the questions that thisstudy raises which is of prime importance to women with MS

1 Zuluaga MI Otero-Romero S Rovira A et al Menarche pregnancies and breastfeeding do not modify long-term prognosis in multiplesclerosis Neurology 201992e1507ndashe1516

2 TintoreM Rovira A Rıo J et al Defining high medium and low impact prognostic factors for developing multiple sclerosis Brain 20151381863ndash1874

3 Bermel RA You X Foulds P et al Predictors of long-term outcome in multiple sclerosis patients treated with interferon β Ann Neurol20137395ndash103

4 Jokubaitis VG Spelman T Kalincik T et al Predictors of long-term disability accrual in relapse-onset multiple sclerosis Ann Neurol20168089ndash100

5 Hellwig K Rockhoff M Herbstritt S et al Exclusive breastfeeding and the effect on postpartum multiple sclerosis relapses JAMANeurol 2015721132ndash1138

Copyright copy 2020 American Academy of Neurology

Author response Menarche pregnancies and breastfeeding do notmodify long-term prognosis in multiple sclerosisMar Tintore (Barcelona Spain) Santiago Perez-Hoyos (Barcelona Spain) and Susana Otero-Romero

(Barcelona Spain)

Neurologyreg 202094456ndash457 doi101212WNL0000000000009065

We thank Drs Jokubaitis and Dobson for the comment on our article1

We built the model for the time to Expanded Disability Status Scale (EDSS) 30 over theclinically definite multiple sclerosis (CDMS) subcohort The adjusted hazard ratio (aHR [CI95]) associated to pregnancy is aHR = 126 CI 95 (062 259)

Regarding the propensity scorendashmatched analyses we decided to perform inverse probability(IP) weighting to create the new pseudocohort to minimize the association between covariatesand pregnancy status Thus no matching was performed but we assigned IP weights to each ofthe patients in the cohort The probability of being pregnant at any time given the set of

456 Neurology | Volume 94 Number 10 | March 10 2020 NeurologyorgN

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

covariates was estimated via a logistic regression adjusted for age at clinically isolated syndrome(CIS) topography of the CIS oligoclonal bands (OB) number of T2 baseline lesions treat-ment status (as time dependent) number of T2 lesions at first year and CDMS (as timedependent)

We totally agree with the issue noted about not adjusting for relapse in the analyses of impact ofpregnancy and breastfeeding on time to EDSS 30 Incorporating relapses in the adjusted modelis key to predict disability The adjusted hazard ratio for pregnancy considering the annualizedrelapse rate over the first 3 years of disease is aHR = 115 CI 95 (056 236) Whencomputing the annualized relapse rate within the first 5 years of disease we obtain an aHR =145 CI 95 (070 302) A further step that we are exploring for this analysis is to includerelapses as a time-varying event with the aim of approaching in a more realistic way the dynamicnature of the disease We also agree that future research must focus on more precise modalitiesof breastfeeding such as mixed or exclusive breastfeeding Unfortunately this information wasmissing in our study1

In the era of high-efficacy drugs suspending disease-modifying treatments may be harmful forpatients with aggressive multiple sclerosis To answer the questions our study raised we are inthe process of independently analyzing a subgroup of pregnant women treated with natalizu-mab or fingolimod

1 Zuluaga MI Otero-Romero S Rovira A et al Menarche pregnancies and breastfeeding do not modify long-term prognosis in multiplesclerosis Neurology 201992e1507ndashe1516

Copyright copy 2020 American Academy of Neurology

Editorsrsquo note Teaching NeuroImages A rare case of Jacobsensyndrome with global diffuse hypomyelination of brainIn the article ldquoTeaching NeuroImages A rare case of Jacobsen syndrome with global diffusehypomyelination of brainrdquo Patel et al presented MRI fluid-attenuated inversion recovery(FLAIR) images at 18months and 3 years of age in a boywith Jacobsen syndrome due to an11q23-11q24 deletion The images showed improvement in white matter abnormalitieswhich were termed hypomyelination by the authors In response Wolf et al argued thathypomyelination is a permanentmyelin deficit and is associated with a less hyperintense T2white matter signal than is seen in this patient They noted that the patientrsquos deletionencompasses HEPACAM a gene for which haploinsufficiency is associated with leuko-dystrophy that improves with time They noted that the case is representative of limitationsin extant classifications of leukodystrophies as either hypomyelinating or demyelinatingResponding to these comments Patel et al agreed that HEPACAM loss of function mayaccount for some of the imaging abnormalities in Jacobsen syndrome but noted thatmacrocephaly and cysts (classical findings with HEPACAM mutations) are not typicallyseen in this syndrome They noted that the original neuroradiologist interpretation termedthe findings as global diffuse hypomyelination This exchange highlights current uncer-tainties in the terminology surrounding the white matter abnormalities particularly in thepediatric population

Aravind Ganesh MD DPhil and Steven Galetta MD

Neurologyreg 202094457 doi101212WNL0000000000009066

NeurologyorgN Neurology | Volume 94 Number 10 | March 10 2020 457

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

Reader response Teaching NeuroImages A rare case of Jacobsensyndrome with global diffuse hypomyelination of brainNicole I Wolf (Amsterdam) and Marjo S van der Knaap (Amsterdam)

Neurologyreg 202094458 doi101212WNL0000000000009070

With interest we read the report by Patel et al1 concerning a patient with Jacobsen syndromedue to an 11q23ndash11q24 deletion and MRI evidence for leukodystrophy with improvement ata follow-up substantiated by FLAIR images The authors claimed that these abnormalitiesrepresent hypomyelination Hypomyelination is defined as a significant and permanent myelindeficit2 Its MRI appearance is characterized by a diffusely hyperintense T2 white matter (WM)signal which is less high than the signal in other leukodystrophies23 and certainly less high thanthe WM signal in the patient discussed here1 who has strongly T2-hyperintense WM signalabnormalities

The chromosomal deletion encompasses HEPACAM Heterozygous and biallelic mutations inthis gene cause megalencephalic leukodystrophy with subcortical cysts (MLC) a vacuolatingleukodystrophy with macrocephaly In dominantHEPACAMmutations the leukodystrophyimproves over time4 In Jacobsen syndromeHEPACAM haploinsufficiency was earlier assumedto cause leukodystrophy5

Why did the authors classify their case as hypomyelination Many neurologists still categorizeleukodystrophies in hypomyelinating and demyelinating disorders3 Perhaps the MRI im-provement not compatible with a demyelinating (progressive) disorder prompted them tolabel this leukodystrophy hypomyelination This case nicely illustrated that not all leukodys-trophies are progressive and that there are more leukodystrophy categories beyond hypo-myelination and demyelination3

1 Patel H Kumar A Raymond G Mainali G Teaching NeuroImages a rare case of Jacobsen syndrome with global diffuse hypo-myelination of brain Neurology 201992e1665ndashe1666

2 Pouwels PJ Vanderver A Bernard G et al Hypomyelinating leukodystrophies translational research progress and prospects AnnNeurol 2014765ndash19

3 van der KnaapMS Schiffmann RMochel F Wolf NI Diagnosis prognosis and treatment of the leukodystrophies Lancet Neurol 201918962ndash972

4 van der KnaapMS Boor I Estevez R Megalencephalic leukoencephalopathy with subcortical cysts chronic white matter oedema due toa defect in brain ion and water homoeostasis Lancet Neurol 201211973ndash985

5 Yamamoto T Shimada S Shimojima K et al Leukoencephalopathy associated with 11q24 deletion involving the gene encoding hepaticand glial cell adhesion molecule in two patients Eur J Med Genet 201558492ndash496

Copyright copy 2020 American Academy of Neurology

Author response Teaching NeuroImages A rare case of Jacobsensyndrome with global diffuse hypomyelination of brainHimadri Patel (Hershey PA) Ashutosh Kumar (Hershey PA) Gerald Raymond (Hershey PA)

and Gayatra Mainali (Hershey PA)

Neurologyreg 202094458ndash459 doi101212WNL0000000000009069

We thank Drs Wolfe and Van der Knaap for their insightful comment on our TeachingNeuroImages study1 and clarification of their precise definition of hypomyelinatingdisorders We agree that HEPACAM loss of function may account for some of the issuein imaging in Jacobsen syndrome but it does not appear to be the entire explanationgiven the lack of macrocephaly or cysts in most patients reported Regarding the hypo-myelination classification this was derived from the original radiology report interpreted

458 Neurology | Volume 94 Number 10 | March 10 2020 NeurologyorgN

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited

by the neuroradiologist as a global diffuse hypomyelination with mild diffuse brain atrophyFurther longitudinal studies would certainly be of interest

1 Patel H Kumar A Raymond G Mainali G Teaching NeuroImages a rare case of Jacobsen syndrome with global diffuse hypo-myelination of brain Neurology 201992e1665ndashe1666

Copyright copy 2020 American Academy of Neurology

CORRECTIONS

Clinical and neural responses to cognitive behavioral therapy forfunctional tremorNeurologyreg 202094459 doi101212WNL0000000000008714

In the article ldquoClinical and neural responses to cognitive behavioral therapy for functional tremorrdquoby Espay et al1 the full authorrsquos name should have appeared throughout as W Curt LaFrance JrThe authors regret the error

Reference1 Espay AJ Ries S Maloney T et al Clinical and neural responses to cognitive behavioral therapy for functional tremor Neurology 2019

93e1787ndashe1798

Clinical risk factors in SUDEPAnationwide population-based case-control studyNeurologyreg 202094459 doi101212WNL0000000000009154

In the article ldquoClinical risk factors in SUDEP A nationwide population-based case-controlstudyrdquo by Sveinsson et al1 the bottom box in figure 1 should read ldquon = 255rdquo and the fifth boxdown on the right should read ldquoControlsrdquo The editorial staff regret the errors

Reference1 Sveinsson O Andersson T Mattsson P Carlsson S Tomson T Clinical risk factors in SUDEP a nationwide population-based case-

control study Neurology 202094e419ndashe429

Genetic determinants of disease severity in the myotonic dystrophytype 1 OPTIMISTIC cohortNeurologyreg 202094459 doi101212WNL0000000000008715

In the article ldquoGenetic determinants of disease severity in the myotonic dystrophy type 1OPTIMISTIC cohortrdquo by Cumming et al1 the study funding section should have read ldquoStudyfunded by European Unionrsquos Seventh Framework Programme (FP72007ndash2013) under grantagreement number 305697 (the OPTIMISTIC project) the Wellcome Centre for Mito-chondrial Research (ref 203105Z16Z)) and donations to the DGM group from theMyotonic Dystrophy Support Group The funders had no role in the study design datacollection analysis interpretation of data writing the report or decisions regarding when tosubmit publicationsrdquo The authors regret the error

Reference1 Cumming SA Jimenez-Moreno C Okkersen K et al Genetic determinants of disease severity in the myotonic dystrophy type 1

OPTIMISTIC cohort Neurology 201993e995ndashe1009

NeurologyorgN Neurology | Volume 94 Number 10 | March 10 2020 459

Author disclosures are available upon request (journalneurologyorg)

Copyright copy 2020 American Academy of Neurology Unauthorized reproduction of this article is prohibited