Background: Cerebral palsy is the most frequent motor disability in childhood, but little is known about its etiology. It has been suggested that cerebral palsy risk may be increased by prenatal thyroid hormone disturbances. The objective of this study was to investigate whether maternal thyroid disorder is associated with increased risk of cerebral palsy. Methods: A population-based cohort study using two study populations. 1) 1,270,079 children born in Denmark 1979–2007 identified in nationwide registers, and 2) 192,918 children born 1996–2009 recruited into the Danish National Birth Cohort and The Norwegian Mother and Child Cohort study, combined in the MOthers and BAbies in Norway and Denmark (MOBAND) collaboration cohort. Register-based and self-reported information on maternal thyroid disorder was studied in relation to risk of cerebral palsy and its unilateral and bilateral spastic subtypes using multiple logistic regression. Children were followed from theage of 1yeartothe ageof6years, andcerebral palsy was identified in nationwide registers with verified diagnoses. Results: In register data, hypothyroidism was recognized in 12,929 (1.0%), hyperthyroidism in 9943 (0.8%), and unclassifiable thyroid disorder in 753 (< 0.1%) of the mothers. The odds ratio for an association between maternal thyroid disorder and bilateral spastic cerebral palsy was 1.0 (95% CI: 0.7–1.5). Maternal thyroid disorder identified during pregnancy was associated with elevated risk of unilateral spastic cerebral palsy (odds ratio 3.1 (95% CI: 1.2–8.4)). In MOBAND, 3042 (1.6%) of the mothers reported a thyroid disorder in pregnancy, which was not associated with cerebral palsy overall (odds ratio 1.2 (95% CI: 0.6–2.4)). Conclusions: Maternal thyroid disorder overall was not related to bilateral spastic cerebral palsy, but maternal thyroid disorder identified in pregnancy was associated with increased risk of unilateral spastic cerebral palsy. These findings should be replicated in studies making use of maternal blood samples. Keywords: Maternal thyroid disorder, Pregnancy, Prenatal exposure, Cerebral palsy, The Danish National Birth Cohort, The Norwegian mother and child cohort study, Register-based cohort Background operate prenatally . Several studies have demonstrated Cerebral palsy (CP) is the most prevalent severe motor that elevated maternal serum levels of thyrotropin and disability in childhood affecting approximately 2 per low thyroid hormone in pregnancy may affect child neu- 1000 live-born children . Recent studies have sug- rodevelopment, including motor function [3–6]. A num- gested that birth complications constitute only a small ber of studies have linked thyroid hormone disturbances part of the factors contributing to the multifactorial eti- of mothers or newborns to CP [7–11], but not every ology of CP, and that most CP risk factors probably study finds the association . Endogenous fetal thyroid hormone production begins around 10-18th week of gestation. The fetus, therefore, * Correspondence: firstname.lastname@example.org depends on maternal thyroid hormone entirely in early Section of Social Medicine, Department of Public Health, University of Copenhagen, Gothersgade 160, 1123 Copenhagen, Denmark pregnancy, and from mid-gestation fetal thyroid hormone Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Petersen et al. BMC Pediatrics (2018) 18:181 Page 2 of 10 production acts in concert with the maternal Individual-level data from nationwide registers on the hypothalamic-pituitary-thyroid axis [8, 13, 14]. Thyroid children and their mothers were linked by use of the hormone is required for many aspects of brain develop- unique personal identifier to which all live-born children ment, including myelination of nerve cells, and insuffi- in Denmark are assigned. We excluded infant deaths (n = cient myelination is often present in individuals with CP 7022), leaving 1,270,079 children for the final analyses. [8, 15]. Moreover, hypo- and hyperthyroidism are corre- The other study population is derived from the lated with coagulation abnormalities that can lead to is- MOthers and BAbies in Norway and Denmark chemia or bleeding, which may underlie unilateral spastic (MOBAND) collaboration cohort , which consists of CP [16, 17]. Thyroid diseases are difficult to stabilize with pooled data from the Danish National Birth Cohort treatment, especially in pregnancy , and it is not un- (DNBC) and The Norwegian Mother and Child Cohort usual for overtreatment to lead women diagnosed with study (MoBa) [21, 22]. MOBAND includes 210,400 hypothyroidism to have elevated thyroid hormone levels live-born children born in 1996–2009 on whom detailed and women with hyperthyroidism to have reduced thyroid information on prenatal exposures was obtained around hormone levels. Thus, we aimed to investigate the associ- gestational week 16 and 31 by telephone interviews in ation between maternal thyroid disorder in pregnancy and DNBC, and around gestational week 17 and 30 by risk of cerebral palsy in the child in two study populations self-administered questionnaires in MoBa, as described in Denmark and Norway, each one with a distinct re- in more detail elsewhere . We excluded children search advantage. One study population permitted exam- without any prenatal information from the earliest data ination of different thyroid disorder and CP subtypes in a collection (n = 16,691), infant deaths (n = 634), and chil- very large study sample, while the other provided oppor- dren with incomplete information on maternal thyroid tunities for performing analyses that controlled for life- disorder (n = 157), leaving 192,918 children for the final style factors. analyses. Methods Register-based data Study design and populations (Fig. 1) Diagnoses of thyroid disorder in the Danish mothers One study population is register-based and comprises were derived from the Danish National Patient Register children born in the eastern part of Denmark in 1979– , which keeps information on hospital admissions 1996, and children born in all of Denmark in 1997– since 1977 and outpatient visits since 1995. Until 1994, 2007. In total, 1,277,101 live births were identified all diagnoses were coded using the eighth version of the through the Danish Civil Registration System . Internal Classification of Diseases (ICD-8), and since Fig. 1 Flowchart of the register-based study population and MOBAND study population. Abbreviations: MOBAND, MOthers and BAbies in Norway and Denmark (MOBAND). DNBC, Danish National Birth Cohort. MoBa, Norwegian Mother and Child Cohort study Petersen et al. BMC Pediatrics (2018) 18:181 Page 3 of 10 1995, the tenth version (ICD-10). The Danish National Cases of cerebral palsy Prescription Registry , established in 1995, provided Danish CP cases were identified through the Danish Na- the Anatomical Therapeutic Chemical (ATC) codes indi- tional Cerebral Palsy Registry that includes all children cating redeemed prescriptions dispensed from Danish surviving the first year of life with a neuro-pediatrician pharmacies. The Danish Medical Birth Registry  and validated diagnosis of CP at age five-six years [31, 32]. In population registers in Statistics Denmark  provided 1979–1996 the register covered the eastern part of information on characteristics of the Danish partici- Denmark, and in 1997–2007 it was nationwide. Ap- pants, while the Medical Birth Registry of Norway  proximately 80% of the Norwegian cases were identified provided information on the Norwegian participants. though the Cerebral Palsy Registry of Norway . The remaining cases were identified in record linkage with Exposure to maternal thyroid disorder the Norwegian Patient Registry and verified by Maternal hypothyroidism and hyperthyroidism in the neuro-pediatricians’ examinations of medical records register-based study population were defined by a . We assessed all CP subtypes combined and the two hospital diagnosis and at least one redeemed prescrip- major subtypes: unilateral and bilateral spastic CP. tion of the appropriate medication, i.e. thyroid hor- mone (ACT-code: H03A) for hypothyroidism and Covariates anti-thyroid medication (ACT-code: H03B) for hyper- We used register-based data to form the variables: the thyroidism, to enhance the validity of the identified year of child’s birth, maternal age (< 25, 25–29, 30–34, disorders. Hypothyroidism was identified by ICD-8 ≥35 years), child’s sex (boy, girl), gestational age (≥37, < codes 243.99 and 244.00–244.09, and by ICD-10 37 weeks of gestation), and maternal diabetes (no, type codes E00, E03.0-E03.9 and E89.0, excluding 244.02, 1, type 2) for both study populations. For the E03.0A, E03.1B, and E03.4. Hyperthyroidism was register-based study population, we also obtained infor- identified by ICD-8 as 242.00–242.29 and by ICD-10 mation from administrative registers about maternal as E05.0-E05.9, excluding E05.4, E05.8A, and E05.9A. educational level (basic, intermediate, higher). In There were some exceptions; when a first-time diag- MOBAND, we used self-reported information on mater- nosis was recorded before the establishment of the nal occupational status (employed, unemployed, student, Danish National Prescription Registry in 1995, the receiving benefits or pension), maternal alcohol con- diagnosis code exclusively determined exposure status. sumption per week (0, 0.5, 1–2.5, ≥3 units), and number Further, as the Danish National Patient Register does of cigarettes smoked per day (0, 1–9, ≥10); all reported not keep information on diagnoses made by general in the earliest data collection. practitioners, we also defined hypothyroidism by at least two redemptions of thyroid hormone prescrip- Statistical analyses tions and no redemptions of anti-thyroid medication We used logistic regression to estimate odds ratios (OR) prescription, and we defined hyperthyroidism by at with 95% confidence intervals (95% CIs) for the relation- least two redemptions of anti-thyroid prescriptions; ship of CP with maternal thyroid disorder. Robust stand- regardless of thyroid hormone prescriptions, in case of no ard errors were used to take into account the potential record of thyroid diagnosis [28–30] (see Additional file 1: dependency between siblings. To guide our decision about eMethod 1 for information on additional coding). We in- which potential confounders we should adjust for, we used cluded all thyroid disorders recorded until 5 years subse- Directed Acyclic Graphs . The adjusted models in- quent to pregnancy until 2010 inclusive. Time of cluded the child’s birth year, maternal diabetes, maternal identification of thyroid disorder was defined as the day of age, and maternal socioeconomic position (education/oc- the first diagnosis code or redeemed prescription, which- cupation), and in addition smoking and alcohol consump- ever was recorded first, and categorized into; ‘before preg- tion in pregnancy in the MOBAND study population. We nancy’, ‘during pregnancy’,and ‘within 5 years after imputed missing values of covariates by use of multiple pregnancy’. imputations (see Additional file 1: eMethod 2). In MOBAND, information on maternal thyroid dis- In the register-based study population, we stratified by order in pregnancy was based on self-reports from the child’s sex and gestational age, respectively, to explore earliest data collection in pregnancy, which was around whether exposure to maternal thyroid disorder has a 16–17 weeks of gestation. The data collected in MoBa greater impact on risk of CP in boys and children born only distinguished thyroid disorder overall and could not at term. Potential misclassification of exposure was separate hypo- and hyperthyroidism. We used informa- assessed by examining the agreement of self-reported tion from the second pregnancy interview of DNBC and and register-based information on maternal thyroid dis- the earliest questionnaire of MoBa to define the use of order in the 90,088 Danish children included in both the thyroid medication in pregnancy. register-based and MOBAND study population. We Petersen et al. BMC Pediatrics (2018) 18:181 Page 4 of 10 subsequently applied the calculated positive and negative hypo- and hyperthyroidism separately. In the stratified agreement to adjust estimates for non-differential mis- analyses, estimates were similar across strata of sex and classification error by using a probabilistic approach  gestational ages, respectively, and no interaction was (for more details see Additional file 1: eMethod 3). suggested (p-values for interaction > 0.2, Additional file To assess the sensitivity of our findings to diagnostic er- 1: eTable 1–2), though the estimates were imprecise. rors or incomplete data, we ran several secondary In MOBAND, 402 children were diagnosed with CP of analyses. In the register-based study population, we re- whom 47.8% had bilateral, and 37.1% had unilateral spas- stricted caseness to hospital-diagnosed maternal thyroid tic CP. Six of the children with bilateral spastic CP were disorders, and also to diagnoses and redeemed prescrip- exposed to maternal thyroid disorder, while only one child tions recorded from 10 years before to 5 years after preg- with unilateral spastic CP was exposed, making it unfeas- nancy. In both study populations, we also restricted ible to confirm or deny the register findings for unilateral analyses to children with complete data on covariates. Fi- spastic CP. The estimates for maternal thyroid disorder nally, we included children who died within the first year and CP overall did not suggest an association (adjusted of life in the study population and assessed the relation- OR 1.2 (95% CI: 0.6–2.4), Table 2), although an high, but ship of maternal thyroid disorder to infant death in case imprecise estimate of risk for bilateral spastic CP (adjusted infant death linked to maternal thyroid disorder (perhaps OR 1.9 (95% CI: 0.8–4.3)) was noted. A higher estimate of with brain damage) precluded the possibility of a CP overall CP risk (adjusted OR 2.5 (95% CI: 0.8–7.9)) was diagnosis. All analyses were performed using StataSE seen in children exposed to an untreated thyroid disorder 14 (64-bit). than to a treated disorder, but these estimates were also imprecise. Adjustment for maternal lifestyle factors did Results not alter the results (Table 2). Characteristics of the study populations In the register-based study population, 23,625 (1.9%) of Assessment of bias the mothers had a thyroid disorder. Hypothyroidism was The proportion of positive agreement between recorded in 12,929 (1.0%) and hyperthyroidism in 9943 self-reported and register-based information on maternal (0.8%) of the mothers; we were unable to classify the thyroid disorder was 60%, while the proportion of negative condition in the remaining 753 (< 0.1%) mothers with agreement was 99%. Estimates adjusted for systematic bias thyroid disorders. In MOBAND, 3042 (1.6%) of the were unstable but suggested that non-differential misclassi- mothers reported a thyroid disorder in pregnancy, of fication of exposure biased towards the null (Table 3). whom 2229 (73.3%) reported use of thyroid medication. Further, changing the categorization of maternal thyroid More than twice as many mothers in MoBa (2.1%) than disorder and restricting the analyses to complete cases did in DNBC (1.0%) reported a thyroid disorder. Mothers not alter the results. Finally, we found an increased risk of with thyroid disorders were more likely to be older, have infant death in children exposed to both maternal diabetes, and deliver preterm than mothers without the hypothyroidism (unadjusted OR 1.4 (95% CI: 1.1–1.9)) and disorder. Also, mothers with thyroid disorders were hyperthyroidism (unadjusted OR 1.9 (95% CI. 1.4–2.4)). more likely to have intermediate or higher education in the register-based study population, were more likely to Discussion be unemployed or receive benefits/pension, and to con- Main findings sume less alcohol and smoke less in pregnancy in Results from the register part of this study show that MOBAND (Table 1). thyroid disorder in pregnancy is not related to bilateral spastic CP, but may possibly be related to unilateral CP. Maternal thyroid disorder and risk of cerebral palsy Though the statistical power of MOBAND data was lim- CP was diagnosed in 2798 children in the register-based ited, information on lifestyle during pregnancy enabled study population. Bilateral spastic CP was the most com- us to perform more thorough control for potential con- mon subtype with 1490 cases, while 912 children had founders, which did not influence the results. unilateral spastic CP. Maternal thyroid disorder diag- nosed or treated for the first time before pregnancy until Potential mechanisms 5 years subsequent to pregnancy was not associated with We studied unilateral and bilateral spastic CP separately, CP overall or either of the subtypes: unilateral or bilat- because they may have distinct etiological profiles. It has eral spastic CP (Table 2). Maternal thyroid disorder been hypothesized that thyroid hormone deficiency can identified during pregnancy was associated with in- cause CP by altering myelination, differentiation, and creased risk of unilateral spastic CP (adjusted OR 3.1 migration of nerve cells , which would likely be (95% CI: 1.2–8.4) (Table 2). Sufficient statistical power reflected in bilateral damage to the brain. An increased was unavailable to address the timing of identification of risk of bilateral spastic CP was suggested in MOBAND Petersen et al. BMC Pediatrics (2018) 18:181 Page 5 of 10 Table 1 Characteristics of exposed versus unexposed in the register-based and MOBAND study population, respectively. Contains data prior to multiple imputation of missing values Danish register-based study population No = 1,270,079 MOBAND study population No = 192,918 Characteristics Exposed to maternal Unexposed Exposed to maternal Unexposed thyroid disorder No = 1,246,454 thyroid disorder No = 189,876 No = 23,625 No = 3042 No (%) No (%) No (%) No (%) Maternal age ≥ 30 years 13,921 (58.9) 544,852 (43.7) 1997 (65.6) 103,084 (54.3) Missing 0 (−)0(−)0(−)0(−) Maternal occupational status Unemployed/ receiving benefits or pension–– 310 (10.2) 6816 (3.6) Missing –– 50 (1.6) 2055 (1.1) Maternal educational level Basic 4781 (20.2) 280,337 (22.5) –– Missing 694 (2.9) 93,414 (7.5) –– Maternal diabetes Diabetes type 1 or 2 518 (2.2) 6362 (<1) 61 (2.0) 855 (<1) Missing 0 (−)0(−)0(−)0(−) Smoking in pregnancy ≥ 1 cigarettes/day –– 293 (9.6) 23,009 (12.1) Missing –– 18 (<1) 900 (<1) Alcohol consumption in pregnancy ≥ 0.5 units/week –– 529 (17.4) 50,510 (26.6) Missing –– 206 (6.8) 10,129 (5.3) Child’s sex Male 11,690 (49.5) 606,514 (48.7) 1572 (51.7) 97,265 (51.2) Missing 5 (<1) 864 (<1) 0 (−) 2 (< 1) Gestational age < 37 weeks (preterm) 1796 (7.6) 72,204 (5.8) 233 (7.7) 11,691 (6.2) Missing 0 (−) 5 (< 1) 12 (< 1) 423 (<1) Abbreviations: MOBAND MOthers and BAbies in Norway and Denmark, No number in complete case data Socioeconomic measure in the MOBAND study population Socioeconomic measure in the register-based study population data only, but the estimate was unstable, and the finding thyroid hormone levels may affect boys differently than was not replicated in the larger register-based study. The girls , but we were unable to find any sex differences most striking finding in this study, a three-fold increase in the association between maternal thyroid disorder in risk of unilateral CP in association with thyroid dis- and the risk of CP. Further, we hypothesized that the order identified in pregnancy, is biological plausible. Ma- risk of CP in association with maternal thyroid disorder ternal thyroid disorder may affect the coagulation would be elevated mainly in children born at term, as system and increase the risk of thrombosis (leading to is- we expect prenatal factors to play a greater role in the chemia) and bleeding [16, 17], and such vascular events etiology of CP in children born at term than in children most likely cause unilateral spastic CP . In support born preterm . The estimates were unstable after of this line of reasoning, markers of coagulation abnor- stratification by gestational age, and there was no indica- malities including Factor V Leiden mutations, which im- tion of any differences in risk. plying an increased risk of thrombosis, have been linked to spastic CP ; especially, the unilateral subtype, Previous findings though the evidence is sparse . The syndrome of neurological cretinism provides a con- The male excess of CP indicates perhaps a heightened vincing indication of a link between maternal thyroid vulnerability to brain injury in boys , and abnormal disturbances and CP. Children born to women with Petersen et al. BMC Pediatrics (2018) 18:181 Page 6 of 10 Table 2 Maternal thyroid disorder and risk of cerebral palsy Danish register-based study population Maternal thyroid disorder All CP (No = 2798) Unilateral Spastic CP (No = 912) Bilateral Spastic CP (No = 1490) b c b c b c No No OR (95% CI) OR (95% CI) No OR (95% CI) OR (95% CI) No OR (95% CI) OR (95% CI) No thyroid disorder 1,246,454 2748 1 (ref.) 1 (ref.) 894 1 (ref.) 1 (ref.) 1465 1 (ref.) 1 (ref.) Thyroid disorder 23,625 50 0.96 (0.73–1.27) 0.99 (0.74–1.31) 18 1.06 (0.67–1.69) 0.93 (0.58–1.48) 25 0.90 (0.61–1.34) 1.03 (0.69–1.53) Hypothyroidism 12,929 26 0.91 (0.62–1.34) 0.95 (0.64–1.39) 9 0.97 (0.50–1.87) 0.83 (0.43–1.61) 13 0.86 (0.50–1.48) 1.01 (0.58–1.74) Hyperthyroidism 9943 24 1.10 (0.73–1.64) 1.11 (0.94–1.66) 9 1.26 (0.65–2.43) 1.13 (0.59–2.18) 12 1.03 (0.58–1.81) 1.13 (0.64–1.99) Identification of thyroid disorder No 1,246,454 2748 1 (ref.) 1 (ref.) 894 1 (ref.) 1 (ref.) 1465 1 (ref.) 1 (ref.) Before pregnancy 9622 20 0.94 (0.61–1.46) 0.97 (0.62–1.50) 7 1.01 (0.48–2.13) 0.86 (0.41–1.81) 11 0.97 (0.54–1.76) 1.13 (0.62–2.04) In pregnancy 1551 7 2.05 (0.98–4.32) 2.00 (0.95–4.21) 4 3.60 (1.35–9.63) 3.14 (1.17–8.42) <4 NE NE ≤ 5 years after pregnancy 12,452 23 0.84 (0.56–1.26) 0.87 (0.58–1.31) 7 0.78 (0.37–1.65) 0.70 (0.33–1.48) 12 0.82 (0.46–1.45) 0.94 (0.53–1.66) MOBAND study population Maternal thyroid disorder in pregnancy All CP (No = 402) b c e No No OR (95% CI) OR (95% CI) OR (95% CI) No thyroid disorder 189,876 394 1 (ref.) 1 (Ref.) 1 (ref.) Thyroid disorder 3042 8 1.27 (0.63–2.55) 1.18 (0.58–2.39) 1.17 (0.58–2.38) No medication 534 3 2.71 (0.87–8.46) 2.54 (0.81–7.94) 2.54 (0.81–7.93) Use of medication 2229 4 0.86 (0.32–2.30) 0.82 (0.30–2.19) 0.82 (0.30–2.19) Multiple logistic regressions were used to calculate odds ratios (ORs) with 95% confidence intervals (95% CIs) in data with multiple imputation of missing values Abbreviations: CP cerebral palsy, No number in complete case data, NE no estimate, MOBAND MOthers and BAbies in Norway and Denmark Maternal thyroid disorder identified before pregnancy until 5 years subsequent to pregnancy Unadjusted Adjusted for birth year, maternal age, maternal diabetes, and maternal socioeconomic status Unclassifiable thyroid disorder is included in the overall measure of thyroid disorder in addition to hypothyroidism and hyperthyroidism Adjusted for birth year, maternal age, maternal diabetes, maternal socioeconomic status, and smoking and alcohol consumption in pregnancy The governmental organization Statistics Denmark responsible for the register-based data, do not allow data extraction of figures below four Information on maternal use of thyroid medication was available for 2763 children exposed to maternal thyroid disorder in the MOBAND study, missing information has been imputed Petersen et al. BMC Pediatrics (2018) 18:181 Page 7 of 10 Table 3 Assessment of misclassification of exposure to maternal thyroid disorder Agreement between sources of maternal thyroid disorder in 90,088 danish children included in both the danish register-based study population and MOBAND study population Register-based information Self-report Unexposed Hypothyroidism Hyperthyroidism Unclassifiable Total Unexposed 88,841 110 48 158 89,157 Hypothyroidism 164 240 17 68 489 Hyperthyroidism 213 21 147 21 402 Unclassifiable 19 14 4 3 40 Total 89,237 385 216 250 90,088 Observed proportionate agreement Proportion of positive agreement 0.60 Proportion of negative agreement 0.99 Kappa 0.60, p < 0.001 Bias-adjusted estimates of thyroid disorder and risk of cerebral palsy All CP Unilateral spastic CP Bilateral spastic CP b c Bias-adjusted OR (study error) in register-based study population 0.89 (0.40–1.21) 1.20 (0.72–3.83) 0.77 (0.24–1.22) Abbreviations: CP cerebral palsy, OR odds ratio The exposure window of the register-based measure was changed to an identified thyroid disorder before pregnancy until week 18 of gestation in order to correspond with the measure based on maternal self-report in MOBAND Raw estimates adjusted for misclassification using a probabilistic approach assuming non-differential misclassification of exposure. Assumptions regarding level of sensitivity and specificity were guided by the calculated positive and negative agreement (for more details see Additional file 1: eMethod 3) Study error includes both systematic error (interval encompassing 95% of the corrected estimates) and random error (95% confidence interval) severe iodine deficiency have a substantial risk of im- Strengths and limitations paired cognitive and motor function, and clinical find- The large scale register data enabled examination of dif- ings and brain imaging compatible with CP has been ferent subtypes of thyroid disorder and CP, but lacked observed in children with the neurologic form of en- information on maternal lifestyle factors. Although the demic cretinism . Maternal thyroid diseases have statistical power of MOBAND was limited, among pro- only been investigated in relation to risk of CP in few spective cohort studies, MOBAND holds by far the lar- studies. Nelson et al. found in a cohort study of gest sample of CP cases with detailed information on 45,559 children, of whom 189 had CP, an increased lifestyle collected during pregnancy, which allowed us to risk of CP in infants with a birth weight ≥ 2500 g who address such potential confounders. were born to women with hyperthyroidism and in in- CP was verified by neuropediatricians based on clinical fants exposed to maternal thyroid hormone and estro- presentation when the children were five-six years old, gen supplementation in pregnancy . In another unaware of maternal illness during pregnancy, which en- study of 183 children with CP and 549 controls with- hances the validity of the CP diagnoses. Likewise, mea- out CP, more cases than controls were born to sures of maternal thyroid disorder were not affected by women who were treated with thyroid hormone; how- knowledge of whether the child had CP, because of the ever, the difference was not statistically significant register recording or self-reporting during pregnancy. . Recent register-based studies from Denmark by We restricted our analyses to children surviving to age 1 Andersen et al. [28, 43] have indicated that thyroid year as CP cannot reliably be diagnosed before this age. disorders identified subsequent to pregnancy, espe- We saw an increased risk of infant death in children pre- cially within 5 years after pregnancy, are correlated natally exposed to maternal thyroid disorders. If children with increased risk of attention deficit hyperactivity with brain damage compatible with CP also are more disorder, autism spectrum disorder, and seizures. We likely to die before CP can be diagnosed, this restriction found a tendency to increased risk of unilateral spas- might have biased our results towards the null. tic CP in children born to women with thyroid disor- We did not have information on causes of thyroid dis- ders identified during pregnancy. These children will orders, which may be important for understanding the probably have been exposed to abnormal thyroid hor- mechanism by which unregulated thyroid disorder is as- mone levels in utero, as abnormal levels may be sociated with risk of CP. Moreover, it is possible that present for a period before the disorder is diagnosed autoimmune conditions confound the association. Thy- and treated for the first time. roid disorders in reproductive-age women are most Petersen et al. BMC Pediatrics (2018) 18:181 Page 8 of 10 often autoimmune in origin, and may be associated with identified before pregnancy, since abnormal thyroid other autoimmune manifestations . Autoimmune hormone levels may be present for a period before the disorders might lead to CP either because autoantibodies disorder is diagnosed and treated for the first time. We are themselves pathogenic, or because of the presence of cannot exclude the possibility that the observed associ- inflammation in autoimmunity, which is an established ation is due to chance or unmeasured confounding. risk factor for spastic CP . We adjusted for maternal Replication of our findings in studies that test maternal diabetes, but it is possible that confounding by other thyroid hormone level in early pregnancy is therefore autoimmune diseases and other abnormalities that coex- needed. ist with thyroid disorder, for which we did not have in- formation, had occurred. Additional file The comparison of measures of thyroid disorder from Additional file 1: The file contain supplementary methods and tables. different sources revealed that some non-differential (DOCX 32 kb) misclassification had occurred, which may have led to an underestimation of the association of thyroid disorder Abbreviations and CP. In the register data, we were unable to identify 95% CI: 95% confidence interval; CP: Cerebral palsy; DNBC: The Danish women who had a thyroid disorder diagnosed outside National Birth Cohort; MoBa: The Norwegian Mother and Child Cohort study; MOBAND: The MOthers and BAbies in Norway and Denmark; OR: Odds ratio hospital settings and who did not redeem any thyroid medication before the establishment of the Danish Na- Acknowledgements tional Prescription in 1995. Moreover, we may have cate- We thank the participating families in Denmark and Norway who took part gorized some women as exposed to maternal thyroid in the DNBC and MoBa. The MOBAND-CP project was initiated in 2011 by senior investigator Allen Wilcox from the National Institute of Environmental disorder in pregnancy, even though the mother may Health Sciences in the United States and has been established by him in have recovered from the disease before pregnancy, e.g. collaboration with the University of Copenhagen, Aarhus University, the postpartum thyroiditis is often transient. In MOBAND, Danish Cerebral Palsy Registry, the Norwegian Institute of Public Health, the University of Bergen, and the Cerebral Palsy Registry of Norway. Data some women may have been unaware of their diagnoses, harmonization was led by Dag Moster and conducted by Ingeborg Forthun which also may have led to misclassification. However, and Mette C. Tollånes at the University of Bergen, and by Katrine Strandberg- in the register-based study population, we used thyroid Larsen and Tanja Gram Petersen at the University of Copenhagen, under the supervision of Dr. Wilcox.We would like to express our thanks to the disorder diagnosed within 5 years after pregnancy as a MOBAND scientific advisory board: Eve Blair from Telethon Kids Institute at proxy for subclinical and asymptomatic thyroid diseases the University of Western Australia and Karin B. Nelson from the National in pregnancy, since recent findings from DNBC based Institute of Neurological Disorders and Stroke at the National Institute of Health, U.S. for constructive comments to the manuscript. We would also like on blood samples drawn in early pregnancy have shown to express our thanks to Diana Haggerty from Department of Epidemiology that abnormal thyroid function may be present for a and Biostatistics at Michigan State University and Steven Korzeniewski from period before the disorder is identified and that asymp- Department of Obstetrics & Gynecology at Wayne State University School of Medicine for inputs to analyses and discussion. tomatic thyroid diseases are common . We antici- pate that much initially asymptomatic thyroid disorders Funding would have come to medical attention within 5 years, This work was supported by The Faculty of Health and Medical Sciences at University of Copenhagen [grant no 620–0032/14–3000 to T.G.P]; Bevica and that the milder and most subclinical forms of thy- Foundation; Ludvig and Sara Elsass Foundation [grant no. 5269 to T.G.P]. The roid disorders would be less likely to have an impact on research salary of UFR was sponsored by an unrestricted research grant from risk of CP. Moreover, as we did not have access to infor- NovoNordic Foundation. Funding was provided to MOBAND-CP by the Intra- mural Research Program at the National Institute of Environmental Health mation on the actual thyroid hormone level, we cannot Sciences, National Institutes of Health, the University of Copenhagen, and know whether women with an identified thyroid dis- Norwegian Institute of Public Health. The Norwegian Mother and Child order actually had abnormal thyroid hormone values Cohort Study are supported by the Norwegian Ministry of Health and Care Services and the Ministry of Education and Research, NIH/NIEHS (contract no during pregnancy or whether the women had a hypo- or N01-ES-75558), NIH/NINDS (grant no.1 UO1 NS 047537–01 and grant no.2 hyperthyroid condition due to overtreatment. The nat- UO1 NS 047537-06A1). The funding agencies had no role in the design or ural next step is to make use of maternal blood samples conduct of the study; data collection, analyses, or interpretation; preparation, review, or approval of the manuscript; or in decision about submitting the collected during pregnancy to study the link between manuscript for publication. maternal thyroid disorder and CP. Availability of data and materials The datasets generated and analyzed during the current study are not Conclusion publicly available because it contains personally sensitive data. Data is stored It is reassuring that maternal thyroid disorders do not on secure servers at Statistic Denmark and University of Bergen and can be seem to be related to the predominant CP subtype, bilat- accessed by applying the appropriate Danish and Norwegian agencies and the steering committees of DNBC, MoBa, and MOBAND. eral spastic CP. However, our findings hint that risk of unilateral spastic CP may be higher in children of Authors’ contributions mothers with thyroid disorder identified in pregnancy, TGP conceptualized and designed the study, carried out all analyses and which presumably is more unregulated than disorders takes responsibility for the integrity of the data and the accuracy of the data Petersen et al. BMC Pediatrics (2018) 18:181 Page 9 of 10 analyses, contributed substantially to interpretation of the work and drafted 12. Blair E, Stanley F. When can cerebral palsy be prevented? The generation of the manuscript. KS-L conceptualized and designed the study, contributed causal hypotheses by multivariate analysis of a case-control study. Paediatr substantially to interpretation of the work, and drafted the manuscript. Perinat Epidemiol. 1993;7:272–301. A-MNA, PU, NP, UF-R and MCT conceptualized and designed the study, 13. Jain V, Agarwal R, Deorari AK, Paul VK. Congenital hypothyroidism. Indian J contributed substantially to interpretation of the work, and reviewed and Pediatr. 2008;75:363–7. revised the manuscript. All authors approved the final manuscript as 14. Burrow GN, Fisher DA, Larsen PR. Maternal and fetal thyroid function. N submitted. Engl J Med. 1994;331:1072–8. 15. Hung PL, Huang CC, Huang HM, Tu DG, Chang YC. Thyroxin treatment protects against white matter injury in the immature brain via brain-derived Ethics approval and consent to participate neurotrophic factor. Stroke. 2013;44:2275–83. Informed consent of all participants in DNBC and MoBa was obtained. DNBC 16. Stuijver DJ, van Zaane B, Romualdi E, Brandjes DP, Gerdes VE, Squizzato A. and the register-linkage were according to Danish legislation approved by The effect of hyperthyroidism on procoagulant, anticoagulant and the Danish Data Protection Agency. The establishment of data collection fibrinolytic factors: a systematic review and meta-analysis. Thromb Haemost. in Moba has obtained licence from the Norwegian Data Inspectorate and 2012;108:1077–88. approval from the Regional Committee of Medical Research Ethics. All 17. Squizzato A, Romualdi E, Buller HR, Gerdes VE. Clinical review: thyroid consents were written. dysfunction and effects on coagulation and fibrinolysis: a systematic review. J Clin Endocrinol Metab. 2007;92:2415–20. Competing interests 18. Laurberg P, Andersen SL, Hindersson P, Nohr EA, Olsen J. Dynamics and The authors declare that they have no competing interests. predictors of serum TSH and fT4 reference limits in early pregnancy: a study within the Danish National Birth Cohort. J Clin Endocrinol Metab. 2016;101: 2484–92. Publisher’sNote 19. Pedersen CB, Gotzsche H, Moller JO, Mortensen PB. The Danish Civil Springer Nature remains neutral with regard to jurisdictional claims in Registration System. A cohort of eight million persons. Dan Med Bull. 2006; published maps and institutional affiliations. 53:441–9. 20. Tollanes MC, Strandberg-Larsen K, Forthun I, Petersen TG, Moster D, Author details Andersen AM, Stoltenberg C, Olsen J, Wilcox AJ. Cohort profile: cerebral Section of Social Medicine, Department of Public Health, University of palsy in the Norwegian and Danish birth cohorts (MOBAND-CP). BMJ Open. Copenhagen, Gothersgade 160, 1123 Copenhagen, Denmark. Pediatric 2016;6:e012777. Department at Rigshospitalet, Juliane Maries Vej 8, 2100 Copenhagen, 21. Olsen J, Melbye M, Olsen SF, Sorensen TI, Aaby P, Andersen AM, Taxbol D, Denmark. Department of Epidemiology & Biostatistics and Department of Hansen KD, Juhl M, Schow TB, et al. The Danish National Birth Cohort–its Pediatrics and Human Development, Michigan State University, 909 Fee background, structure and aim. Scand J Public Health. 2001;29:300–7. Road, East Lansing, MI 48824, USA. Department of Medical Endocrinology at 22. Magnus P, Birke C, Vejrup K, Haugan A, Alsaker E, Daltveit AK, Handal M, Rigshospitalet, Ole Maaloees Vej 26, 2200 Copenhagen, Denmark. Domain Haugen M, Hoiseth G, Knudsen GP, et al. Cohort profile update: the Norwegian for Mental and Physical Health at Norwegian Institute of Public Health, mother and child cohort study (MoBa). Int J Epidemiol. 2016;45:382–8. Kalfarveien 31, 5020 Bergen, Norway. 23. Lynge E, Sandegaard JL, Rebolj M. The Danish National Patient Register. Scand J Public Health. 2011;39:30–3. Received: 14 December 2017 Accepted: 18 May 2018 24. Kildemoes HW, Sorensen HT, Hallas J. The Danish National Prescription Registry. Scand J Public Health. 2011;39:38–41. 25. Knudsen LB, Olsen J. The Danish medical birth registry. Dan Med Bull. 1998; References 45:320–3. 1. Oskoui M, Coutinho F, Dykeman J, Jette N, Pringsheim T. An update on the 26. Mortensen LH, Helweg-Larsen K, Andersen AM. Socioeconomic differences prevalence of cerebral palsy: a systematic review and meta-analysis. Dev in perinatal health and disease. Scand J Public Health. 2011;39:110–4. Med Child Neurol. 2013;55:509–19. 27. Irgens LM. The Medical Birth Registry of Norway. Epidemiological research 2. Nelson KB, Blair E. Prenatal factors in singletons with cerebral palsy born at and surveillance throughout 30 years. Acta Obstet Gynecol Scand. 2000;79: or near term. N Engl J Med. 2015;373:946–53. 435–9. 3. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J, O'Heir 28. Andersen S, Laurberg P, Wu C, Olsen J. Attention deficit hyperactivity disorder CE, Mitchell ML, Hermos RJ, Waisbren SE, et al. Maternal thyroid deficiency and autism spectrum disorder in children born to mothers with thyroid during pregnancy and subsequent neuropsychological development of the dysfunction: a Danish nationwide cohort study. Bjog. 2014;121:1365–74. child. N Engl J Med. 1999;341:549–55. 29. Andersen SL, Olsen J, Wu CS, Laurberg P. Smoking reduces the risk of 4. Klein RZ, Mitchell ML. Maternal hypothyroidism and cognitive development hypothyroidism and increases the risk of hyperthyroidism: evidence from of the offspring. Curr Opin Pediatr. 2002;14:443–6. 450,842 mothers giving birth in Denmark. Clin Endocrinol. 2014;80:307–14. 5. Pop VJ, Brouwers EP, Vader HL, Vulsma T, van Baar AL, de Vijlder JJ. Maternal 30. Andersen SL, Olsen J, Wu CS, Laurberg P. Psychiatric disease in late hypothyroxinaemia during early pregnancy and subsequent child adolescence and young adulthood. Foetal programming by maternal development: a 3-year follow-up study. Clin Endocrinol. 2003;59:282–8. hypothyroidism? Clin Endocrinol (Oxf). 2014;81:126–33. 6. Pop VJ, Kuijpens JL, van Baar AL, Verkerk G, van Son MM, de Vijlder 31. Topp M, Langhoff-Roos J, Uldall P. Validation of a cerebral palsy register. JJ, Vulsma T, Wiersinga WM, Drexhage HA, Vader HL. Low maternal J Clin Epidemiol. 1997;50:1017–23. free thyroxine concentrations during early pregnancy are associated 32. Uldall P, Michelsen SI, Topp M, Madsen M. The Danish Cerebral Palsy with impaired psychomotor development in infancy. Clin Registry. A registry on a specific impairment. Dan Med Bull. 2001;48:161–3. Endocrinol. 1999;50:149–55. 33. Andersen GL, Irgens LM, Haagaas I, Skranes JS, Meberg AE, Vik T. Cerebral 7. Nelson KB, Ellenberg JH. Antecedents of cerebral palsy. Multivariate analysis palsy in Norway: prevalence, subtypes and severity. Eur J Paediatr Neurol. of risk. N Engl J Med. 1986;315:81–6. 2008;12:4–13. 8. Hong T, Paneth N. Maternal and infant thyroid disorders and cerebral palsy. 34. Hollung SJ, Vik T, Wiik R, Bakken IJ, Andersen GL. Completeness and Semin Perinatol. 2008;32:438–45. correctness of cerebral palsy diagnoses in two health registers: implications 9. Leviton A, Paneth N, Reuss ML, Susser M, Allred EN, Dammann O, Kuban K, for estimating prevalence. Dev Med Child Neurol. 2016;59:402–6. Van Marter LJ, Pagano M. Hypothyroxinemia of prematurity and the risk of cerebral white matter damage. J Pediatr. 1999;134:706–11. 35. Hernan MA, Hernandez-Diaz S, Werler MM, Mitchell AA. Causal knowledge 10. Reuss ML, Paneth N, Pinto-Martin JA, Lorenz JM, Susser M. The relation of as a prerequisite for confounding evaluation: an application to birth defects transient hypothyroxinemia in preterm infants to neurologic development epidemiology. Am J Epidemiol. 2002;155:176–84. at two years of age. N Engl J Med. 1996;334:821–7. 36. Lash TL, Fox MP, Fink AK. A Guide to Implementing Quantitative Bias 11. Suzumura H, Nitta A, Tsuboi Y, Watabe Y, Kuribayashi R, Arisaka O. Thyroxine Analysis. In: Applying Quantitative Bias Analysis to Epidemiologic Data. 1st for transient hypothyroxinemia and cerebral palsy in extremely preterm ed. London, New York: Springer; 2009. p. 13–32. https://link.springer.com/ infants. Pediatr Int. 2011;53:463–7. content/pdf/10.1007%2F978-0-387-87959-8.pdf. Petersen et al. BMC Pediatrics (2018) 18:181 Page 10 of 10 37. Kirton A, deVeber G. Cerebral palsy secondary to perinatal ischemic stroke. Clin Perinatol. 2006;33:367–86. 38. Nelson KB, Dambrosia JM, Grether JK, Phillips TM. Neonatal cytokines and coagulation factors in children with cerebral palsy. Ann Neurol. 1998;44:665–75. 39. Thorarensen O, Ryan S, Hunter J, Younkin DP. Factor V Leiden mutation: an unrecognized cause of hemiplegic cerebral palsy, neonatal stroke, and placental thrombosis. Ann Neurol. 1997;42:372–5. 40. Himmelmann K, Hagberg G, Beckung E, Hagberg B, Uvebrant P. The changing panorama of cerebral palsy in Sweden. IX. Prevalence and origin in the birth-year period 1995-1998. Acta Paediatr. 2005;94:287–94. 41. Briet JM, van Wassenaer AG, van Baar A, Dekker FW, Kok JH. Evaluation of the effect of thyroxine supplementation on behavioural outcome in very preterm infants. Dev Med Child Neurol. 1999;41:87–93. 42. Ma T, Lian ZC, Qi SP, Heinz ER, DeLong GR. Magnetic resonance imaging of brain and the neuromotor disorder in endemic cretinism. Ann Neurol. 1993;34:91–4. 43. Andersen SL, Laurberg P, Wu CS, Olsen J. Maternal thyroid dysfunction and risk of seizure in the child: a Danish nationwide cohort study. J Pregnancy. 2013;2013:636705. 44. Lisovska N, Daribayev Z, Lisovskyy Y, Kussainova K, Austin L, Bulekbayeva S. Pathogenesis of cerebral palsy through the prism of immune regulation of nervous tissue homeostasis: literature review. Childs Nerv Syst. 2016;32: 2111–7. 45. Andersen SL, Olsen J. Early pregnancy thyroid function test abnormalities in biobank sera from women clinically diagnosed with thyroid dysfunction before or after pregnancy. Thyroid. 2017;27:451–9.
– Springer Journals
Published: May 31, 2018