Research Article Maternal Thyroid Dysfunction and Risk of ...downloads.hindawi.com/journals/jp/2013/636705.pdfJournal of Pregnancy examined the impact of maternal thyroid dysfunction

Hindawi Publishing CorporationJournal of PregnancyVolume 2013 Article ID 636705 10 pageshttpdxdoiorg1011552013636705

Research ArticleMaternal Thyroid Dysfunction and Risk of Seizure in the ChildA Danish Nationwide Cohort Study

Stine Linding Andersen1 Peter Laurberg1 Chun Sen Wu2 and Joslashrn Olsen2

1 Department of Endocrinology Aalborg University Hospital Soslashndre Skovvej 15 9000 Aalborg Denmark2 Section for Epidemiology Department of Public Health Aarhus University 8000 Aarhus C Denmark

Correspondence should be addressed to Stine Linding Andersen stinearndk

Received 11 April 2013 Revised 3 June 2013 Accepted 27 June 2013

Academic Editor Antonio Farina

Copyright copy 2013 Stine Linding Andersen et alThis is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Thyroid hormones are essential for brain development andmaternal thyroid disease may affect child neurocognitive developmentSome types of seizures may also depend upon early exposure of the developing central nervous system and we hypothesizedthat maternal thyroid dysfunction could increase the risk of seizure in the child In a Danish population-based study we included1699693 liveborn singletons and from the Danish National Hospital Register we obtained information on maternal diagnosis ofhyper- or hypothyroidism and neonatal seizure febrile seizure and epilepsy in the child Maternal diagnosis of thyroid dysfunctionbefore or after birth of the childwas registered in two percent of the singleton births In adjusted analysesmaternal hyperthyroidismand hypothyroidism first time diagnosed after birth of the child were associated with a significant increased risk of epilepsy in thechild Moreover hypothyroidism diagnosed after birth of the child was associated with a significant increased risk of neonatal andfebrile seizures No significant association was seen for maternal diagnosis prior to birth of the child We speculate if some degreeof maternal thyroid dysfunction was already present during the pregnancy in mothers diagnosed after birth of the child and if thisuntreated condition may present a neurodevelopmental risk

1 Introduction

Thyroid hormones are essential for fetal brain development[1 2] and required for normal myelination the regulationof cell migration in the cerebral cortex hippocampus andcerebellum the differentiation of neurons oligodendrocytesastrocytes and microglia and synaptogenesis [3] Thyroidhormone receptors are expressed in the developing humanfetal brain during the first trimester of pregnancy [4 5]before the onset of fetal thyroid hormone production Thusmaternal thyroid hormones are the key source of thyroidhormone supply to the fetus in the first half of pregnancy andthey also have an important role in fetal brain developmentin the second half of pregnancy as illustrated by paucity ofneurological symptoms in congenital hypothyroidism withfailure of the fetal thyroid gland [6]

Thyroid diseases are common in women of reproductiveage either as overt or subclinical disease [7 8] and severalstudies addressed the impact ofmaternal thyroid dysfunction

on child neurocognitive development [9ndash12] Seizures are oneof the manifestations of neurological disorders and resultfrom an abnormal or excessive discharge of a set of neuronsin the brain [13] Seizures may depend upon early centralnervous system (CNS) alterations [14 15] and the highestincidence occurs during the first year of life [16] Seizuresin the neonatal period are important to recognize as theyoften relate to an underlying neurotoxic exposure that mightadversely affect the developing brain [16] After the neonatalperiod febrile seizures are the most common type of seizuresin childhood affecting 2ndash5 of children [17] Epilepsy ischaracterized by recurrent seizures and is the most commonserious neurologic disorder in childhood [15] The aetiologyof febrile seizures and epilepsy is only partly understood butenvironmental and genetic risk factors have been proposed[15 17]

We hypothesized that maternal thyroid dysfunctionhyper- and hypothyroidism would increase the risk of sei-zure in the child via subtle changes in brain structure We

2 Journal of Pregnancy

examined the impact of maternal thyroid dysfunction diag-nosed before during and after the pregnancy on the riskof neonatal seizure febrile seizure and epilepsy in thechild More specifically we hypothesized that an associationbetween maternal thyroid dysfunction and the risk of seizurein the child could be caused by one or more of the followingmechanisms (1) Fetal Programming Abnormal thyroid hor-mone levels or treatment of a thyroid disease beforeduringpregnancy could affect the fetal brain and increase suscep-tibility to seizure in the child An association with maternalthyroid disease diagnosed before birth of the childmight thenbe expected (2) Genetics A genetic component passed on tothe child from the parent suffering from thyroid disease couldincrease susceptibility to seizure in the child If genetic factorsplay a role an association with paternal thyroid disease orwith maternal thyroid disease diagnosed after birth of thechild might then also be expected (3) Subclinical ThyroidDisease The impact of maternal thyroid dysfunction presentbefore andor during pregnancy but clinically diagnosed andtreated after birth of the child could increase susceptibilityto seizure in the child An association with maternal thyroiddisease diagnosed after birth of the child might then beexpected

2 Materials and Methods

21 Study Population and Design We conducted a popu-lation-based cohort study using Danish nationwide registersA unique ten-digit personal identification number is used inall national registers in Denmark and in the Danish CivilRegistration System [18] we identified all liveborn singletonsinDenmark between January 1 1978 andNovember 30 2006

The Danish National Hospital Register (DNHR) [19]holds data on all admissions to any Danish Hospital since1977 and all hospital outpatient visits since 1995 For everyadmission the register contains date of admission and dis-charge and diagnoses classified according to the 8th revisionof the International Classification of Disease (ICD-8) from1977 to 1993 and the 10th revision (ICD-10) from 1994 andonwards We included in- and outpatient visits with a mainor additional first time diagnosis (see choice of ICD codesbelow) before January 1 2007 The ldquoonsetrdquo of disease wasdefined as the day of admission to hospital and onset ofmaternal disease was categorized as beforeafter birth ofthe child and during pregnancy The pregnancy period wasestimated by subtracting gestational age at birth from the datethe child was born

22 Diagnosis ofThyroid Dysfunction Diagnoses of maternalthyroid dysfunction were obtained from DNHR Hyperthy-roidism was defined as ICD-8 24200ndash24229 and ICD-10E05ndashE059 (excluding thyrotoxicosis factitia (E054) over-production of thyroid-stimulating hormone (E058A) andthyrotoxic heart disease (E059A)) Hypothyroidism wasdefined as ICD-8 24399 24400ndash24409 (excluding sec-ondary hypothyroidism 24402) and ICD-10 E03ndashE039 andE890 (excluding unspecified congenital goitre (E030A) andatrophy of the thyroid (congenital E031B acquired E034))

23 Diagnosis of Seizure Diagnoses of seizure were obtainedfrom DNHR using the following ICD codes febrile seizureICD-8 78021 and ICD-10 R560 and R560A epilepsy ICD-8 34509ndash34599 and ICD-10 G400ndashG419 and unspecifiedseizure ICD-8 78029 and ICD-10 R568 Diagnoses ofbreath-holding spells (ICD-8 78020 and ICD-10 R568C)were not included

Neonatal seizure was defined as ICD-10 P909 or any ofthe seizure diagnoses listed above obtained in the neonatalperiod (first 28 days of life) Febrile seizure was definedas a diagnosis of febrile seizure after the neonatal periodand before the age of 5 years with no previous diagnosis ofneonatal seizure or epilepsy Finally epilepsy was defined asa diagnosis of epilepsy after the neonatal period [13]

24 Covariates From theMedical Birth Registry [20] whichhas been computerized since 1973 we obtained informationon gender of the child gestational age at birth birth weight5 minutes Apgar score and maternal parity and age at birthof child

From Statistics Denmark we obtained information onmaternal cohabitation income origin and geographical res-idence at birth of child Information was only available from1980 and data from 1980 were used to substitute missing val-ues in 1978 and 1979 Formaternal cohabitation and origin wereplaced additional missing values by available informationin the preceding or following five (origin) or three (cohab-itation) years whichever came first Maternal residence atbirth of child was used as an indicator variable for iodineintake Iodine is required for thyroid hormone synthesisand Denmark was previously iodine deficient with regionaldifferences moderate iodine deficiency in West Denmarkand mild iodine deficiency in East Denmark (divided by theGreat Belt) The mandatory iodine fortification of salt wasintroduced in the year 2000 and had increased urinary iodineto lower recommended level in 2004-2005 [21]

From DNHR we obtained information on parental diag-nosis of epilepsy and febrile seizure maternal diagnosis ofpreeclampsiaeclampsia (ICD-8 63703ndash63719 and ICD-10O14-O150) and diabetes (ICD-8 24900ndash25009 and ICD-10E100ndashE149 and O24ndashO249) and child diagnosis of cerebralpalsy (ICD-8 34399 and 34499 and ICD-10 G80ndash809) andcongenital malformations (ICD-8 74099ndash75999 and ICD-10 Q00ndashQ99)

We excluded children with missing information onmaternal identification number 119899 = 13 or maternal covari-ates parity 119899 = 169 origin 119899 = 5015 cohabitation 119899 = 1257and income 119899 = 1335 thus leaving 1699693 singletons inthe final cohort for followup Singletons not alive after theneonatal period (119899 = 838) or who had emigrated (119899 = 109)during the neonatal period were not included in the followupstudy on the risk of febrile seizure and epilepsy

3 Statistical Analyses

Children were followed up from birth (neonatal seizure) orfrom day 29 after birth (febrile seizure and epilepsy) and untilthe date of admission for the first in- or outpatient visit witha diagnosis of seizure emigration or death or until day 28

Journal of Pregnancy 3

after birth (neonatal seizure) age of 5 years (febrile seizure)or December 31 2006 (epilepsy)

TheCox proportional hazardsmodel was used to estimatehazard ratio (HR) with 95 confidence interval (95 CI)for risk of neonatal seizure febrile seizure and epilepsyin children exposed to maternal hyper- or hypothyroidismdiagnosed before andor after birth of the child comparedto children unexposed The proportional hazards assump-tion was tested in plots of log cumulative hazards Robuststandard errors were used to adjust for dependence betweenmaternal multiple pregnancies Univariate and multivariateanalyses were performed and in the multivariate analyseswe adjusted for potential confounders including gender ofthe child (boygirl) birth year (lt1980 1980ndash1982 1983ndash1985 1986ndash1988 1989ndash1991 1992ndash1994 1995ndash1997 1998ndash2000 2001ndash2003 2004ndash2006) and the following maternalvariables obtained at the time of the childrsquos birth age (lt2020ndash24 25ndash29 30ndash34 35ndash39 ge40 years) parity includingindex pregnancy (1 2 3 ge4) cohabitation (marriednotmarried) income (1st 2nd 3rd 4th quartile) origin (bornin Denmarknot born in Denmark) residence (EastWestDenmark) and maternal diagnosis of febrile seizure andorepilepsy registered in the Danish National Hospital Register(DNHR) before January 1 2007 (yesno) In parallel analysesHR with 95 CI for risk of seizure in children exposed topaternal thyroid dysfunction was estimated

Apgar score birth weight and gestational age were a pri-ori considered possible intermediates and hence not includedin themainmodel Analyses were repeated after the exclusionof (a) children with a diagnosis of cerebral palsy congenitalmalformations andor 5-minute Apgar score below 7 (119899 =173440) (b) children born to mothers with a diagnosis ofpreeclampsiaeclampsia andor diabetes (119899 = 126812) (c)children with gestational age at birth lt37 or gt41 weeks orbirth weight not appropriate for gestational age (below 10thor above 90th percentile for gestational week and gender)(119899 = 505397) and (d) children with a diagnosis of thyroiddisease (119899 = 2391)

Potential confounding by maternal smoking was evalu-ated in the cohort of children born after December 31 1995when information on maternal smoking during pregnancywas available in DNHR Statistical analyses were performedusing STATA version 11 (Stata Corp College Station TexasUSA) and a 5 level of significance was chosen The studywas approved by the Danish Data Protection Agency

4 Results

41 Maternal Diagnosis of Thyroid Dysfunction in relationto Pregnancy Among 1699693 liveborn singletons in Den-mark between January 1 1978 and November 30 2006we identified 34582 singletons (20) whose mother hada diagnosis of hyperthyroidism (13) or hypothyroidism(07) registered in DNHR before January 1 2007 Table 1presents characteristics of children andmothers at birth of thechild according to maternal diagnosis of thyroid dysfunctionbefore or after birth of the child

The majority of exposed singletons were born to mothersdiagnosed with hyper- or hypothyroidism after birth of

0

500

1000

1500

2000

5th 4th 3rd 2nd 1st During 1st 2nd 3rd 4th 5th

Hyperthyroidism

Hypothyroidism

Time intervals around pregnancy (years)

Num

ber o

f sin

glet

on li

ve b

irths

a

AfterBefore

Figure 1 First-time diagnosis of maternal thyroid dysfunction inrelation to birth of child The figure illustrates number of singletonlive births per year with a first-time diagnosis of maternal hyper- orhypothyroidism registered in the Danish National Hospital Register(DNHR) during the pregnancy (during) and in 1-year intervalsbeforeafter the pregnancy (1st 2nd 3rd etc) aSingleton livebirths with missing information on gestational age or registrationof gestational age lt20 or gt45 weeks not included (119899 = 60352)Number of singleton live births with maternal diagnosis during thepregnancy estimated per year

the child (743) with a median time from birth of the childto maternal diagnosis of 107 years (range 0003ndash289 years)For maternal diagnosis before or at birth of the child mediantime from diagnosis to birth of the child was 30 years (0ndash275 years) for hyperthyroidism and 17 years (0ndash298 years)for hypothyroidism

In the group of singletons born to mothers with a diag-nosis of thyroid dysfunction before birth of the child andvalid information on gestational age 2498 singleton birthsfollowed a maternal first time diagnosis of thyroid diseaseduring the pregnancy Figures 1 and 2 illustrate the distribu-tion of singleton births with a diagnosis of maternal hyper-or hypothyroidism during pregnancy and in 1 year intervalsbefore and after pregnancy the relative frequency ofmaternaldiagnosis during pregnancy was higher for hypothyroidismin comparison to hyperthyroidism

42 Main Analyses Children were followed up for up to289 years and Figure 3 illustrates the cumulative percent-ages of children with no diagnosis of seizure according tofollow-up time and maternal thyroid dysfunction Table 2presents the number of children diagnosed with seizureduring followup Overall a diagnosis of seizure (neonatalseizure febrile seizure andor epilepsy) occurred more oftenamong children whose mother had a diagnosis of hyper-or hypothyroidism (53) than among unexposed children(46) 119875 lt 0001 Median age at diagnosis was 14 years forfebrile seizure and 53 years for epilepsy and in 182 of thecases a diagnosis of epilepsy occurred during the first year oflife


Table 1 Characteristics of singletons born in Denmark between January 1 1978 and November 30 2006 and their mothers at birth of thechild

No maternal thyroid dysfunctiona Maternal hyperthyroidismb Maternal hypothyroidismb

Before birth After birth Before birth After birth119899 119899 119899 119899 119899

Singletons 1665111 980 5738 03 17139 10 3155 02 8550 05Maternal characteristicsParityc

Primipara 759826 456 1827 318 7804 455 1068 338 4018 470Multipara 905285 544 3911 682 9335 545 2087 662 4532 530

Age (years)lt25 378535 227 518 90 4523 264 309 98 2076 24325ndash35 1155045 694 4241 739 11277 658 2325 737 5757 673gt35 131531 79 979 171 1339 78 521 165 717 84

OriginBorn in Denmark 1520213 913 5090 887 15556 908 2698 855 7497 877Not born in Denmark 144898 87 648 113 1583 92 457 145 1053 123

Residenced

West Denmark 939188 564 3329 580 10621 620 1610 510 4688 548East Denmark 725923 436 2409 420 6518 380 1545 490 3862 452

Child characteristicsGenderBoy 854831 513 2884 503 8665 506 1615 512 4429 518Girl 810280 487 2854 497 8474 494 1540 488 4121 482

Gestational age at birth (weeks)e

lt37 70984 44 363 65 818 51 193 62 425 5337ndash41 1392330 870 4870 867 13837 866 2713 875 6896 855gt41 138383 86 382 68 1328 83 196 63 740 92

Birth weight (grams)e

Mean (SD) 3488 (559) 3442 (591) 3425 (567) 3546 (596) 3492 (584)5 minutes Apgar scoref

7ndash10 1635867 992 5625 990 16854 993 3096 991 8415 9940ndash6 12554 08 55 10 118 07 27 09 49 06

aNo maternal diagnosis of hyper- or hypothyroidism registered in the Danish National Hospital Register (DNHR) before January 1 2007bMaternal first-time diagnosis of thyroid dysfunction (hyper- or hypothyroidism) before or after birth of the child and registered in DNHR before January 12007cNumber of births (live and stillbirths) including index pregnancydDivided into regions by the Great Belt West Denmark with previously moderate iodine deficiency and East Denmark with previously mild iodine deficiencyeSingletons with missing value on gestational age or birth weight registration of gestational age lt20 or gt45 weeks or birth weight lt500 or gt6000 grams notincluded (n = 65235)fSingletons with missing value on 5-minute Apgar score not included (n = 17033)

The crude and adjusted HR for risk of seizure in thechild differed according to seizure type and maternal thyroiddysfunction (Table 2) The adjusted model only changedthe estimates slightly whereas maternal hypothyroidismfirst time diagnosed before or after birth of the child wasassociated with increased risk of neonatal seizure febrileseizure and epilepsy maternal hyperthyroidism only showeda statistically significant association with epilepsy

43 Time of Maternal Diagnosis We subsequently evaluatedif the increased risk of epilepsy in children born to mothers

with a diagnosis of hyper- or hypothyroidismdiffered accord-ing to time of maternal diagnosis in relation to birth of thechild (Figure 4) Maternal hyper- and hypothyroidism firsttime diagnosed after birth of the childwere associatedwith anincreased risk of epilepsy in the child For maternal diagnosisprior to birth of the child the estimated HR revealed rathersimilar increased risk However the number of cases wassmaller and the confidence intervals were wider Thus theassociations were not statistically significant

Maternal hypothyroidism was associated with an incre-ased risk of febrile seizure and neonatal seizure in the child


0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

ffirstborn

singl

eton

live

birt

hsa

AfterBefore

Figure 2 First-time diagnosis of maternal thyroid dysfunction inrelation to birth of firstborn child The figure illustrates numberof firstborn singleton live births per year with a first-time diagnosisof maternal hyper- or hypothyroidism registered in the DanishNational-Hospital Register (DNHR) during the pregnancy (during)and in 1 year intervals beforeafter the pregnancy (1st 2nd 3rd etc)aFirstborn refers to primiparous mother Firstborn singleton livebirths with missing information on gestational age or registrationof gestational age lt20 or gt45 weeks not included (119899 = 30363)Number of singleton live births with maternal diagnosis during thepregnancy estimated per year

Figure 5 illustrates the stratified analyses for febrile seizurein the child according to time of maternal diagnosis ofhypothyroidism Only maternal diagnosis after birth of thechildwas associatedwith significantly increased risk of febrileseizure in the child and this applied to maternal diagnosis ofdisease both within 5 years and more than 5 years after birthof the child For neonatal seizure in the child both maternaldiagnosis prior to and after birth of the child revealed anincreased risk but results only reached statistical significancein children born to mothers diagnosed after birth of the child(diagnosis before birth of the child 165 (088ndash306) afterbirth of the child 183 (127ndash265))

Among the 2498 children whose mother had hyper- orhypothyroidismdiagnosed during the pregnancy a total of 111children developed seizure during followup (neonatal 119899 = 8febrile 119899 = 83 and epilepsy 119899 = 26) and the adjusted HRfor risk of seizure in the child did not indicate that maternaldiagnosis during pregnancy revealed a high risk neonatalseizure HR 164 (95CI 082ndash328) febrile seizure 110 (088ndash136) and epilepsy 125 (085ndash184) However numbers weresmall and did not permit stratification by maternal type ofthyroid dysfunction

44 PaternalThyroidDysfunction We identified 5742 single-tons (03) whose father had a diagnosis of hyperthyroidism(119899 = 3981) or hypothyroidism (119899 = 1761) registered inDNHRbefore January 1 2007 and among these children 288 (5)had a diagnosis of seizure (neonatal 119899 = 9 febrile 119899 = 190and epilepsy 119899 = 109) No significant increased risk of seizure

0 10 20 30

100

98

96

94

92

90

HypothyroidismHyperthyroidism

Followup time (years)

No hyper-or hypothyroidism

Cum

ulat

ive ldquo

seiz

ure-

freerdquo

surv

ival

()ab

Figure 3 Cumulative percentages of children with no diagnosisof seizure (neonatal seizure febrile seizure or epilepsy) by follow-up time from birth of the child and maternal diagnosis of thy-roid dysfunction (none hyperthyroidism hypothyroidism) aCoxregressionmodel Outcome a diagnosis of seizure (neonatal seizurefebrile seizure or epilepsy) Exposure maternal diagnosis of hyper-thyroidism hypothyroidism or no maternal thyroid dysfunctionDiagnoses were registered in the Danish National Hospital Registerbefore January 1 2007Model included gender of the child (boygirl)birth year (lt1980 1980ndash1982 1983ndash1985 1986ndash1988 1989ndash19911992ndash1994 1995ndash1997 1998ndash2000 2001ndash2003 2004ndash2006) and thefollowingmaternal variables obtained at the time of the childrsquos birthage (lt20 20ndash24 25ndash29 30ndash34 35ndash39 ge40 years) parity includingindex pregnancy (1 2 3 ge4) cohabitation (marriednot married)income (1st 2nd 3rd 4th quartile) origin (born in Denmarknotborn in Denmark) residence (EastWest Denmark) and maternaldiagnosis of febrile seizure andor epilepsy registered in the DanishNational Hospital Register (DNHR) before January 1 2007 (yesno)b317 cases were censored (febrile seizure 119899 = 89 epilepsy 119899 = 224and epilepsy and febrile seizure 119899 = 4) all because of emigrationDistribution of the censored cases by exposure group no hyper-or hypothyroidism (119899 = 306) hyperthyroidism (119899 = 8) andhypothyroidism (119899 = 3)

in children exposed to paternal thyroid disease was observed(adjustedHRneonatal seizure 087 (95CI 045ndash168) febrileseizure 099 (086ndash115)) The adjusted HR for risk of epilepsyin the child was on the borderline of statistical significance120 (95 CI 099ndash146)

45 Sensitivity Analyses The exclusion of children with adiagnosis of cerebral palsy congenital malformations andorApgar score below 7 revealed almost identical adjusted HRto the nonrestricted analyses However after such exclusionthe association between maternal hypothyroidism and therisk of neonatal seizure in the child was no longer statisti-cally significant (adjusted HR 134 (083ndash216)) Restrictingthe cohort to children born at term with a birth weightappropriate for gestational age did not change results formaternal hyperthyroidism but for hypothyroidism the risk of


Table 2 Hazard ratio (HR) with 95 confidence interval (95 CI) for seizure in singletons born in Denmark 1978ndash2006 according tomaternal diagnosis of hyper- or hypothyroidism registered in theDanishNationalHospital Register (DNHR) before January 1 2007 Referenceis singletons born to mothers with no diagnosis of thyroid dysfunction registered in DNHR before January 1 2007

Singletons (n) Crude HR (95 CI) Adjusted HR (95 CI)a Pb

Seizurec

No hyper- or hypothyroidism 77444 100 (reference) 100 (reference)Hyper- or hypothyroidism 1844 111 (106ndash116) 112 (107ndash118) lt0001Hyperthyroidism 1185 106 (100ndash112) 108 (102ndash115) 0009Hypothyroidism 659 120 (111ndash130) 120 (111ndash130) lt0001

Neonatal seizured

No hyper- or hypothyroidism 3109 100 (reference) 100 (reference)Hyper- or hypothyroidism 89 139 (113ndash171) 142 (115ndash175) 0001Hyperthyroidism 50 118 (089ndash156) 122 (092ndash162) 0161Hypothyroidism 39 180 (131ndash247) 178 (130ndash244) lt0001

Febrile seizuree


Epilepsyf

No hyper- or hypothyroidism 22173 100 (reference) 100 (reference)Hyper- or hypothyroidism 628 125 (115ndash135) 121 (111ndash131) lt0001Hyperthyroidism 426 124 (113ndash137) 120 (109ndash132) lt0001Hypothyroidism 202 126 (109ndash144) 122 (106ndash140) 0005

aAdjusted model included gender of the child (boygirl) birth year (lt1980 1980ndash1982 1983ndash1985 1986ndash1988 1989ndash1991 1992ndash1994 1995ndash1997 1998ndash20002001ndash2003 2004ndash2006) and the following maternal variables obtained at the time of the childrsquos birth age (lt20 20ndash24 25ndash29 30ndash34 35ndash39 ge40 years)parity including index pregnancy (1 2 3 ge4) cohabitation (marriednot married) income (1st 2nd 3rd 4th quartile) origin (born in Denmarknot bornin Denmark) residence (EastWest Denmark) and maternal diagnosis of febrile seizure andor epilepsy registered in the Danish National Hospital Register(DNHR) before January 1 2007 (yesno)bP value for results of the adjusted modelcDiagnosis of neonatal seizure febrile seizure andor epilepsydDiagnosis of seizure in the neonatal period (first 28 days of life)eDiagnosis of febrile seizure after the neonatal period and before the age of 5 years with no previous diagnosis of epilepsy or neonatal seizurefDiagnosis of epilepsy after the neonatal period and before January 1 2007 Epilepsy diagnosis following neonatal seizure (n = 654) following febrile seizure(n= 3104)

neonatal seizure (adjusted HR 142 (087ndash232)) and epilepsy(118 (098ndash143)) were no longer statistically significant

The exclusion of children born to mothers with a diag-nosis of preeclampsia andor diabetes did not change resultsA small group of children had a diagnosis of thyroid disease(hyperthyroidism 119899 = 960 hypothyroidism 119899 = 1431) witha median age at diagnosis of 146 years The frequency ofseizure was higher in this group (84) but the associationswere similar after the exclusion of these children Restrictinganalyses to firstborn child did not change results neitherdid the exclusion of children born to foreign-born mothersWe identified 627100 singletons with available informationon maternal smoking during the pregnancy Applying thesame main model to this restricted cohort revealed similarassociations regardless of the inclusion of maternal smokingin the model (data not shown)

5 Discussion

51 Principle Findings Based on Danish nationwide registerswe found an increased risk of seizure in children born to

mothers diagnosed with thyroid dysfunction in a Danishhospital after birth of the child Maternal hypothyroidismincreased the risk of neonatal seizure febrile seizure andepilepsy whereas maternal hyperthyroidism solely revealedan increased risk of epilepsy in the child No significantassociation with maternal thyroid dysfunction diagnosedprior to birth of the child orwith paternal thyroid dysfunctionwas observed

52 Brain Development Thyroid Hormones and Risk ofEpilepsy Thyroid hormones are essential for fetal braindevelopment [2] and involved in a number of developmentalevents including regulation of cell migration in the cerebralcortex hippocampus and cerebellum and the differentiationof neurons oligodendrocytes astrocytes and microglia [3]Disruptions in the normal process of cerebral cortex develop-ment and in glial function have been proposed as some of theunderlying causes of epilepsy [14] Malformations of corticaldevelopment are macro- or microscopic abnormalities of thecerebral cortex that may arise during cortical development inthe first or second trimester of pregnancy andmay predispose


Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20

Maternal diagnosis after birth of the child and risk of epilepsy in the child

Maternal diagnosis before birth of the child and risk of epilepsy in the child

Figure 4 Adjusted hazard ratio with 95 confidence intervalfor risk of epilepsy in singletons exposed to maternal hyper- orhypothyroidismfirst time diagnosed before or after birth of the childand registered in the Danish National Hospital Register (DNHR)before January 1 2007 Reference is children born tomothers with nodiagnosis of hyper- or hypothyroidism registered in DNHR beforeJanuary 1 2007 Model included gender of the child (boygirl)birth year (lt1980 1980ndash1982 1983ndash1985 1986ndash1988 1989ndash19911992ndash1994 1995ndash1997 1998ndash2000 2001ndash2003 2004ndash2006) and thefollowingmaternal variables obtained at the time of the childrsquos birthage (lt20 20ndash24 25ndash29 30ndash34 35ndash39 ge40 years) parity includingindex pregnancy (1 2 3 ge4) cohabitation (marriednot married)income (1st 2nd 3rd 4th quartile) origin (born in Denmarknotborn in Denmark) residence (EastWest Denmark) and maternaldiagnosis of febrile seizure andor epilepsy registered in the DanishNational Hospital Register (DNHR) before January 1 2007 (yesno)

to epileptic seizures [22] Glia cells are nonneuronal cellsinvolved in neuronal functions and the maintenance oftissue homeostasis and it has been proposed that especiallyastrocytes and microglia dysfunction may contribute to thepathophysiology of epilepsy [23]

It is intriguing how both lack and excess of maternalthyroid hormones can increase the risk of epilepsy in thechild but experimental studies in rats provide evidence thatboth maternal hyper- and hypothyroidism may interfereprofoundly with early fetal brain development [24]

53 Risk of Neonatal and Febrile Seizures Maternal hypothy-roidism in contrast to hyperthyroidism was associated withincreased risk of neonatal and febrile seizures in the childFebrile seizures are the most common seizures in childhoodand most often benign [25] The aetiology is only partlyunderstood although genetic and environmental risk factors

Reference

After

Before

075 10 125 15 175 20

Maternal diagnosis of hypothyroidism before birth of the childand risk of febrile seizure in the child

Maternal diagnosis of hypothyroidism after birth of the childand risk of febrile seizure in the child

le5 years after

gt5 years after

Figure 5 Adjusted hazard ratio with 95 confidence interval forrisk of febrile seizure in singletons exposed to maternal hypothy-roidism first time diagnosed before or after birth of the child andregistered in the Danish National Hospital Register (DNHR) beforeJanuary 1 2007 Reference is children born to mothers with nodiagnosis of hyper- or hypothyroidism registered in DNHR beforeJanuary 1 2007 Model included gender of the child (boygirl)birth year (lt1980 1980ndash1982 1983ndash1985 1986ndash1988 1989ndash19911992ndash1994 1995ndash1997 1998ndash2000 2001ndash2003 2004ndash2006) and thefollowingmaternal variables obtained at the time of the childrsquos birthage (lt20 20ndash24 25ndash29 30ndash34 35ndash39 ge40 years) parity includingindex pregnancy (1 2 3 ge4) cohabitation (marriednot married)income (1st 2nd 3rd 4th quartile) origin (born in Denmarknotborn in Denmark) residence (EastWest Denmark) and maternaldiagnosis of febrile seizure andor epilepsy registered in the DanishNational Hospital Register (DNHR) before January 1 2007 (yesno)

have been described [25] Concerning febrile seizures andbrain development controversies exist on the possible associ-ationwith hippocampal sclerosis and atrophy [17] It has beenbrought forward that structural changes in the hippocampusmight either preexist and predispose to febrile seizures orresult from prolonged febrile seizures and increase the long-term risk of epilepsy [17 26] Concerning a possible associ-ation with maternal hypothyroidism evidence suggests thatthe hypothyroid state during brain development may inducestructural and functional alterations in the hippocampus[27] and one study reported that children with congenitalhypothyroidism on regular Levothyroxine substitution wereless prone to develop febrile seizures [28]

Neonatal seizures often relate to an underlying causewhich is most often perinatal hypoxia-ischemia but alsostroke haemorrhage infections or metabolic disorders [16]


The association between maternal hypothyroidism and riskof neonatal seizure was no longer significant when excludingchildren with a diagnosis of cerebral palsy congenital malfor-mations or low Apgar score or when restricting the analysesto children born at term with a birth weight appropriate forgestational age These factors might constitute intermediateson the causal pathway [29ndash31] and hence adjustment couldinduce bias including collider bias [32] Alternatively the lackof a significant association after such exclusion could be dueto the reduced size of the study population

54 Time of Maternal Diagnosis in relation to Birth of theChild The increased risk of seizure in children born tomothers with a later diagnosis of thyroid dysfunction couldrelate to disturbances in fetal brain development secondaryto abnormal maternal thyroid hormone levels We onlyfound a significant association in children born to moth-ers diagnosed after birth of the child Low-grade thyroiddysfunction may be present for years before diagnosis [33]and these children could have been exposed to maternalsubclinical thyroid disease present during the pregnancyOn the other hand children born to mothers diagnosedbefore or during the pregnancy might not have been exposedto abnormal maternal thyroid hormone levels to the sameextent since proper treatment of maternal disease is expectedto have been initiated Another possible explanation of theassociation with maternal thyroid dysfunction diagnosedafter birth of child could be a genetic component linkingmaternal thyroid dysfunction and risk of seizure in the childhowever no significant association with paternal thyroiddisease was observed Finally we cannot exclude if the lack ofan association with maternal thyroid dysfunction diagnosedbefore birth of child and with paternal thyroid dysfunctionwas due to the relatively small number of children in thesegroups and consequently lack of power to detect an existingdifference with statistical significance

55 Strengths and Limitations The strength of our studyis the large study population and long follow-up periodwith virtually complete followup Thus bias due to selectionof study participants is unlikely We are aware of the riskof violating the principles of ldquonot conditioning on thefuturerdquo (exposure registered after outcome has occurred)when analysing the association withmaternal thyroid diseasediagnosed after birth of the childThe time frombirth of childto maternal diagnosis becomes ldquoimmortalrdquo which may causebias [34] However we estimate this bias to be small due tothe low mortality during followup

Thequality of the diagnoses inDNHRhas been evaluatedand the positive predictive value of a diagnosis of febrileseizure and epilepsy was 93 and 81 respectively [35 36]A diagnosis of thyroid disease not particularly in pregnancybut in the general population only revealed misclassificationin 2 of the cases [37] We were able to distinguish betweenmaternal hyper- and hypothyroidism but further classifica-tion of maternal disease andor epilepsy was not possible asthe validity decreases when the number of digits in the ICD-code increases [38]

Misclassification of exposure and outcome does exist inour study population however we believe this misclassifica-tion is nondifferential Maternal onset of disease was definedas the day of first admission to hospital We only includedwomen with a hospital diagnosis of thyroid disease andwe did not have results of thyroid function tests thyroidautoantibodies scintigraphy andor treatment modality Webelieve that pregnant women are in general referred to hospi-tal for diagnosis and therapy [39] but theymight present withsymptoms and even have therapy initiated in general practiceyears before referral to hospital On the other hand wecannot exclude that management of maternal thyroid diseasefirst time diagnosed after birth of child could take place ingeneral practice alone In our study population maternalhyperthyroidism was more frequent than hypothyroidismThis might be due to different referral pattern for hyper-and hypothyroidism In addition the Danish populationwas iodine deficient during part of the study period andin this period there were relatively more hyperthyroidismand less hypothyroidism in the population [40] A first-time diagnosis of maternal thyroid disease during pregnancyoccurred relatively more often in hypothyroidism comparedto hyperthyroidism This finding seems likely as hyperthy-roidism in women of reproductive age might present withobvious symptoms leading to diagnosis whereas hypothy-roidism might not be detected until symptoms exaggerateduring pregnancy or if thyroid function parameters arescreened

Although we adjusted for a number of potential con-founders unmeasured confounding might still exist Infor-mation on maternal infections and use of antibiotic duringpregnancy [41] was not included as these variables are notexpected to be related to our exposure

56 Perspective The potential benefits and harms of screen-ing pregnant women for thyroid dysfunction in early preg-nancy have been an ongoing discussion in recent years asdiscussed in detail in [42] The present study adds furtherevidence to this discussion and if future studies corroboratethat many pregnant women suffer from undiagnosed thyroiddysfunction and that this increases the risk of neurocognitivediseases in the child screeningmay be indicated as suggestedby several authors [43 44]

6 Conclusions

Maternal hyper- and hypothyroidism diagnosed after birth ofthe child may increase the risk of epilepsy in the child Inaddition maternal hypothyroidism diagnosed after birth ofchild was associated with increased risk of neonatal seizureand febrile seizureThe findings lead to speculations whetherundetected maternal thyroid hormone aberrations werealready present during the pregnancy in mothers diagnosedafter birth of the child

Conflict of Interests

The authors declare that they have no conflict of interests


Acknowledgments

The study was supported by a grant from ldquoMusikforlaeligggerneAgnes og Knut Moslashrks Fondrdquo Chun Sen Wu is supported bythe Individual Postdoctoral Grants from the Danish MedicalResearch Council (FSS 12-32232)

References

[1] S Chan andMDKilby ldquoThyroid hormone and central nervoussystem developmentrdquo Journal of Endocrinology vol 165 no 1pp 1ndash8 2000

[2] GM de EscobarM J Obregon and F Escobar del Rey ldquoMater-nal thyroid hormones early in prenancy and fetal brain devel-opmentrdquo Best Practice and Research vol 18 no 2 pp 225ndash2482004

[3] J Bernal ldquoThyroid hormones and brain developmentrdquoVitaminsand Hormones vol 71 pp 95ndash122 2005

[4] J Bernal and F Pekonen ldquoOntogenesis of the nuclear 3531015840-triiodothyronine receptor in the human fetal brainrdquoEndocrinol-ogy vol 114 no 2 pp 677ndash679 1984

[5] J Iskaros M Pickard I Evans A Sinha P Hardiman and REkins ldquoThyroid hormone receptor gene expression in first trim-ester human fetal brainrdquo Journal of Clinical Endocrinology andMetabolism vol 85 no 7 pp 2620ndash2623 2000

[6] P Santisteban and J Bernal ldquoThyroid development and effecton the nervous systemrdquo Reviews in Endocrine and MetabolicDisorders vol 6 no 3 pp 217ndash228 2005

[7] A Carle P Laurberg I B Pedersen et al ldquoEpidemiology ofsubtypes of hypothyroidism in Denmarkrdquo European Journal ofEndocrinology vol 154 no 1 pp 21ndash28 2006

[8] A Carle I B Pedersen N Knudsen et al ldquoEpidemiology ofsubtypes of hyperthyroidism in Denmark a population-basedstudyrdquo European Journal of Endocrinology vol 164 no 5 pp801ndash809 2011

[9] J E Haddow G E Palomaki W C Allan et al ldquoMaternalthyroid deficiency during pregnancy and subsequent neuropsy-chological development of the childrdquo New England Journal ofMedicine vol 341 no 8 pp 549ndash555 1999

[10] R Z Klein J D Sargent P R Larsen S E Waisbren J EHaddow and M L Mitchell ldquoRelation of severity of maternalhypothyroidism to cognitive development of offspringrdquo Journalof Medical Screening vol 8 no 1 pp 18ndash20 2001

[11] V J Pop E P Brouwers H L Vader T Vulsma A L Van Baarand J J De Vijlder ldquoMaternal hypothyroxinaemia during earlypregnancy and subsequent child development a 3-year follow-up studyrdquo Clinical Endocrinology vol 59 no 3 pp 282ndash2882003

[12] JH Lazarus J P Bestwick S Channon et al ldquoAntenatal thyroidscreening and childhood cognitive functionrdquo New EnglandJournal of Medicine vol 366 no 6 pp 493ndash501 2012

[13] ldquoGuidelines for epidemiologic studies on epilepsy Commissionon Epidemiology and Prognosis International League AgainstEpilepsyrdquo Epilepsia vol 34 no 4 pp 592ndash596 1993

[14] B S Chang and D H Lowenstein ldquoMechanisms of diseaseepilepsyrdquo New England Journal of Medicine vol 349 no 13 pp1257ndash1266 2003

[15] R Guerrini ldquoEpilepsy in childrenrdquo The Lancet vol 367 no9509 pp 499ndash524 2006

[16] M L Zupanc ldquoNeonatal seizuresrdquo Pediatric Clinics of NorthAmerica vol 51 no 4 pp 961ndash978 2004

[17] C Waruiru and R Appleton ldquoFebrile seizures an updaterdquoArchives of Disease in Childhood vol 89 no 8 pp 751ndash7562004

[18] C B Pedersen ldquoThe Danish civil registration systemrdquo Scandi-navian Journal of Public Health vol 39 no 7 pp 22ndash25 2011

[19] T F Andersen M Madsen J Joslashrgensen L Mellemkjaeligr and JH Olsen ldquoThe Danish National Hospital Register a valuablesource of data for modern health sciencesrdquo Danish MedicalBulletin vol 46 no 3 pp 263ndash268 1999

[20] L B Knudsen and J Olsen ldquoTheDanishmedical birth registryrdquoDanish Medical Bulletin vol 45 no 3 pp 320ndash323 1998

[21] L B Rasmussen A Carle T Joslashrgensen et al ldquoIodine intakebefore and aftermandatory iodization inDenmark results fromthe Danish Investigation of Iodine Intake andThyroid Diseases(DanThyr) studyrdquo British Journal of Nutrition vol 100 no 1 pp166ndash173 2008

[22] R J Leventer R Guerrini and W B Dobyns ldquoMalformationsof cortical development and epilepsyrdquo Dialogues in ClinicalNeuroscience vol 10 no 1 pp 47ndash62 2008

[23] M L Foresti G M Arisi and L A Shapiro ldquoRole of gliain epilepsy-associated neuropathology neuroinflammation andneurogenesisrdquo Brain Research Reviews vol 66 no 1-2 pp 115ndash122 2011

[24] O M Ahmed A W El-Gareib A M El-bakry S M AbdEl-Tawab and R G Ahmed ldquoThyroid hormones states andbrain development interactionsrdquo International Journal of Devel-opmental Neuroscience vol 26 no 2 pp 147ndash209 2008

[25] M Vestergaard O Basso T Brink Henriksen J R Oslashstergaardand J Olsen ldquoRisk factors for febrile convulsionsrdquo Epidemiol-ogy vol 13 no 3 pp 282ndash287 2002

[26] M Vestergaard C B Pedersen P Sidenius J Olsen and JChristensen ldquoThe long-term risk of epilepsy after febrile seiz-ures in susceptible subgroupsrdquo American Journal of Epidemiol-ogy vol 165 no 8 pp 911ndash918 2007

[27] C Koromilas C Liapi K H Schulpis K Kalafatakis A Zarrosand S Tsakiris ldquoStructural and functional alterations in thehippocampus due to hypothyroidismrdquoMetabolic Brain Diseasevol 25 no 3 pp 339ndash354 2010

[28] T Asami F Sasagawa S Kyo K Asami and M UchiyamaldquoIncidence of febrile convulsions in children with congenitalhypothyroidismrdquo Acta Paediatrica vol 87 no 6 pp 623ndash6261998

[29] Y Sun M Vestergaard C B Pedersen J Christensen and JOlsen ldquoApgar scores and long-term risk of epilepsyrdquo Epidemiol-ogy vol 17 no 3 pp 296ndash301 2006

[30] Y Sun M Vestergaard C B Pedersen J Christensen O Bassoand JOlsen ldquoGestational age birthweight intrauterine growthand the risk of epilepsyrdquo American Journal of Epidemiology vol167 no 3 pp 262ndash270 2008

[31] S L Andersen J Olsen C S Wu and P Laurberg ldquoLow birthweight in children born to mothers with hyperthyroidism andhigh birth weight in hypothyroidism whereas preterm birthis common in both conditions a Danish National HospitalRegister studyrdquo EuropeanThyroid Journal vol 2 no 2 pp 135ndash144 2013

[32] A J Wilcox C R Weinberg and O Basso ldquoOn the pitfallsof adjusting for gestational age at birthrdquo American Journal ofEpidemiology vol 174 no 9 pp 1062ndash1068 2011

[33] J Karmisholt S Andersen and P Laurberg ldquoVariation inthyroid function in subclinical hypothyroidism importance ofclinical follow-up and therapyrdquo European Journal of Endocrinol-ogy vol 164 no 3 pp 317ndash323 2011


[34] S Suissa ldquoImmortal time bias in pharmacoepidemiologyrdquoAmerican Journal of Epidemiology vol 167 no 4 pp 492ndash4992008

[35] M Vestergaard C Obel T B Henriksen et al ldquoThe Dan-ish National Hospital Register is a valuable study base forepidemiologic research in febrile seizuresrdquo Journal of ClinicalEpidemiology vol 59 no 1 pp 61ndash66 2006

[36] J Christensen M Vestergaard J Olsen and P Sidenius ldquoVal-idation of epilepsy diagnoses in the Danish National HospitalRegisterrdquo Epilepsy Research vol 75 no 2-3 pp 162ndash170 2007

[37] P Vestergaard and L Mosekilde ldquoFractures in patients withhyperthyroidism and hypothyroidism a nationwide follow-upstudy in 16249 patientsrdquoThyroid vol 12 no 5 pp 411ndash419 2002

[38] J Mosbech J Joslashrgensen M Madsen K Rostgaard K Thorn-berg and T D Poulsen ldquoThe national patient registry Evalu-ation of data qualityrdquo Ugeskrift for Laeger vol 157 no 26 pp3741ndash3745 1995

[39] A Carle P Laurberg I B Pedersen et al ldquoMainly the youngerhypothyroid patients are referred to hospitalmdashevidence forreferral biasrdquo Journal of Clinical Epidemiology vol 62 no 4 pp446ndash451 2009

[40] P Laurberg C Cerqueira L Ovesen et al ldquoIodine intake as adeterminant of thyroid disorders in populationsrdquo Best Practiceand Research vol 24 no 1 pp 13ndash27 2010

[41] J E Miller L H Pedersen Y Sun and J Olsen ldquoMaternaluse of cystitis medication and childhood epilepsy in a Danishpopulation-based cohortrdquo Paediatric and Perinatal Epidemiol-ogy vol 26 no 6 pp 589ndash595 2012

[42] P Laurberg S L Andersen I B Pedersen S Andersen and ACarle ldquoScreening for overt thyroid disease in early pregnancymay be preferable to searching for small aberrations in thyroidfunction testsrdquo Clinical Endocrinology 2013

[43] D Glinoer ldquoThe systematic screening and management ofhypothyroidism and hyperthyroidism during pregnancyrdquo Tre-nds in Endocrinology andMetabolism vol 9 no 10 pp 403ndash4111998

[44] G Morreale De Escobar M J Obregon and F Escobar DelRey ldquoClinical perspective is neuropsychological developmentrelated tomaternal hypothyroidism or tomaternal hypothyrox-inemiardquo Journal of Clinical Endocrinology andMetabolism vol85 no 11 pp 3975ndash3987 2000

Submit your manuscripts athttpwwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


examined the impact of maternal thyroid dysfunction diag-nosed before during and after the pregnancy on the riskof neonatal seizure febrile seizure and epilepsy in thechild More specifically we hypothesized that an associationbetween maternal thyroid dysfunction and the risk of seizurein the child could be caused by one or more of the followingmechanisms (1) Fetal Programming Abnormal thyroid hor-mone levels or treatment of a thyroid disease beforeduringpregnancy could affect the fetal brain and increase suscep-tibility to seizure in the child An association with maternalthyroid disease diagnosed before birth of the childmight thenbe expected (2) Genetics A genetic component passed on tothe child from the parent suffering from thyroid disease couldincrease susceptibility to seizure in the child If genetic factorsplay a role an association with paternal thyroid disease orwith maternal thyroid disease diagnosed after birth of thechild might then also be expected (3) Subclinical ThyroidDisease The impact of maternal thyroid dysfunction presentbefore andor during pregnancy but clinically diagnosed andtreated after birth of the child could increase susceptibilityto seizure in the child An association with maternal thyroiddisease diagnosed after birth of the child might then beexpected

2 Materials and Methods

21 Study Population and Design We conducted a popu-lation-based cohort study using Danish nationwide registersA unique ten-digit personal identification number is used inall national registers in Denmark and in the Danish CivilRegistration System [18] we identified all liveborn singletonsinDenmark between January 1 1978 andNovember 30 2006

The Danish National Hospital Register (DNHR) [19]holds data on all admissions to any Danish Hospital since1977 and all hospital outpatient visits since 1995 For everyadmission the register contains date of admission and dis-charge and diagnoses classified according to the 8th revisionof the International Classification of Disease (ICD-8) from1977 to 1993 and the 10th revision (ICD-10) from 1994 andonwards We included in- and outpatient visits with a mainor additional first time diagnosis (see choice of ICD codesbelow) before January 1 2007 The ldquoonsetrdquo of disease wasdefined as the day of admission to hospital and onset ofmaternal disease was categorized as beforeafter birth ofthe child and during pregnancy The pregnancy period wasestimated by subtracting gestational age at birth from the datethe child was born

22 Diagnosis ofThyroid Dysfunction Diagnoses of maternalthyroid dysfunction were obtained from DNHR Hyperthy-roidism was defined as ICD-8 24200ndash24229 and ICD-10E05ndashE059 (excluding thyrotoxicosis factitia (E054) over-production of thyroid-stimulating hormone (E058A) andthyrotoxic heart disease (E059A)) Hypothyroidism wasdefined as ICD-8 24399 24400ndash24409 (excluding sec-ondary hypothyroidism 24402) and ICD-10 E03ndashE039 andE890 (excluding unspecified congenital goitre (E030A) andatrophy of the thyroid (congenital E031B acquired E034))

23 Diagnosis of Seizure Diagnoses of seizure were obtainedfrom DNHR using the following ICD codes febrile seizureICD-8 78021 and ICD-10 R560 and R560A epilepsy ICD-8 34509ndash34599 and ICD-10 G400ndashG419 and unspecifiedseizure ICD-8 78029 and ICD-10 R568 Diagnoses ofbreath-holding spells (ICD-8 78020 and ICD-10 R568C)were not included

Neonatal seizure was defined as ICD-10 P909 or any ofthe seizure diagnoses listed above obtained in the neonatalperiod (first 28 days of life) Febrile seizure was definedas a diagnosis of febrile seizure after the neonatal periodand before the age of 5 years with no previous diagnosis ofneonatal seizure or epilepsy Finally epilepsy was defined asa diagnosis of epilepsy after the neonatal period [13]

24 Covariates From theMedical Birth Registry [20] whichhas been computerized since 1973 we obtained informationon gender of the child gestational age at birth birth weight5 minutes Apgar score and maternal parity and age at birthof child

From Statistics Denmark we obtained information onmaternal cohabitation income origin and geographical res-idence at birth of child Information was only available from1980 and data from 1980 were used to substitute missing val-ues in 1978 and 1979 Formaternal cohabitation and origin wereplaced additional missing values by available informationin the preceding or following five (origin) or three (cohab-itation) years whichever came first Maternal residence atbirth of child was used as an indicator variable for iodineintake Iodine is required for thyroid hormone synthesisand Denmark was previously iodine deficient with regionaldifferences moderate iodine deficiency in West Denmarkand mild iodine deficiency in East Denmark (divided by theGreat Belt) The mandatory iodine fortification of salt wasintroduced in the year 2000 and had increased urinary iodineto lower recommended level in 2004-2005 [21]

From DNHR we obtained information on parental diag-nosis of epilepsy and febrile seizure maternal diagnosis ofpreeclampsiaeclampsia (ICD-8 63703ndash63719 and ICD-10O14-O150) and diabetes (ICD-8 24900ndash25009 and ICD-10E100ndashE149 and O24ndashO249) and child diagnosis of cerebralpalsy (ICD-8 34399 and 34499 and ICD-10 G80ndash809) andcongenital malformations (ICD-8 74099ndash75999 and ICD-10 Q00ndashQ99)

We excluded children with missing information onmaternal identification number 119899 = 13 or maternal covari-ates parity 119899 = 169 origin 119899 = 5015 cohabitation 119899 = 1257and income 119899 = 1335 thus leaving 1699693 singletons inthe final cohort for followup Singletons not alive after theneonatal period (119899 = 838) or who had emigrated (119899 = 109)during the neonatal period were not included in the followupstudy on the risk of febrile seizure and epilepsy

3 Statistical Analyses

Children were followed up from birth (neonatal seizure) orfrom day 29 after birth (febrile seizure and epilepsy) and untilthe date of admission for the first in- or outpatient visit witha diagnosis of seizure emigration or death or until day 28






4 Results



0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

f sin

glet

on li

ve b

irths

a

AfterBefore













Residenced




lt37 70984 44 363 65 818 51 193 62 425 5337ndash41 1392330 870 4870 867 13837 866 2713 875 6896 855gt41 138383 86 382 68 1328 83 196 63 740 92



7ndash10 1635867 992 5625 990 16854 993 3096 991 8415 9940ndash6 12554 08 55 10 118 07 27 09 49 06







0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

ffirstborn

singl

eton

live

birt

hsa

AfterBefore





0 10 20 30

100

98

96

94

92

90




Cum

ulat

ive ldquo

seiz

ure-

freerdquo

surv

ival

()ab







Seizurec


Neonatal seizured


Febrile seizuree


Epilepsyf





5 Discussion





Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





















4 Results



0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

f sin

glet

on li

ve b

irths

a

AfterBefore













Residenced




lt37 70984 44 363 65 818 51 193 62 425 5337ndash41 1392330 870 4870 867 13837 866 2713 875 6896 855gt41 138383 86 382 68 1328 83 196 63 740 92



7ndash10 1635867 992 5625 990 16854 993 3096 991 8415 9940ndash6 12554 08 55 10 118 07 27 09 49 06







0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

ffirstborn

singl

eton

live

birt

hsa

AfterBefore





0 10 20 30

100

98

96

94

92

90




Cum

ulat

ive ldquo

seiz

ure-

freerdquo

surv

ival

()ab







Seizurec


Neonatal seizured


Febrile seizuree


Epilepsyf





5 Discussion





Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
























Residenced




lt37 70984 44 363 65 818 51 193 62 425 5337ndash41 1392330 870 4870 867 13837 866 2713 875 6896 855gt41 138383 86 382 68 1328 83 196 63 740 92



7ndash10 1635867 992 5625 990 16854 993 3096 991 8415 9940ndash6 12554 08 55 10 118 07 27 09 49 06







0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

ffirstborn

singl

eton

live

birt

hsa

AfterBefore





0 10 20 30

100

98

96

94

92

90




Cum

ulat

ive ldquo

seiz

ure-

freerdquo

surv

ival

()ab







Seizurec


Neonatal seizured


Febrile seizuree


Epilepsyf





5 Discussion





Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















0

500

1000

1500

2000


Hyperthyroidism

Hypothyroidism


Num

ber o

ffirstborn

singl

eton

live

birt

hsa

AfterBefore





0 10 20 30

100

98

96

94

92

90




Cum

ulat

ive ldquo

seiz

ure-

freerdquo

surv

ival

()ab







Seizurec


Neonatal seizured


Febrile seizuree


Epilepsyf





5 Discussion





Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of



















Seizurec


Neonatal seizured


Febrile seizuree


Epilepsyf





5 Discussion





Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















Reference

Hypothyroidism

Hyperthyroidism

Hypothyroidism

Hyperthyroidism

075 10 125 15 175 20







Reference

After

Before

075 10 125 15 175 20



le5 years after

gt5 years after












6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
























6 Conclusions





Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















Acknowledgments


References



















































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

































of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





















of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















Documents

Research Article Maternal Thyroid Dysfunction and Risk of ...downloads.hindawi.com/journals/jp/2013/636705.pdfJournal of Pregnancy examined the impact of maternal thyroid dysfunction