ABSTRACT
Thyroid diseases are common among women of reproductive age and may be related to childhood cancer among offspring, though the evidence is inconsistent. Using Danish registries, we conducted a population‐based case–control study to investigate the relationship between maternal thyroid diseases and medication use and childhood cancer among offspring. Cases (N = 2521) diagnosed between 1973 and 2016 were ascertained from the Danish Cancer Registry, and controls (N = 63,014) were randomly selected from the Central Population Registry and matched on sex and date of birth. Information on diagnoses and prescriptions was obtained from the National Patient Register and the National Prescription Register, respectively. Conditional logistic regression was applied to estimate odds ratios (ORs) and 95% confidence intervals (CIs). Goiter was associated with any cancer (OR = 2.08, 95% CI: 1.44, 3.00), leukemia (OR = 2.18, 95% CI: 1.13, 4.18), acute lymphoblastic leukemia (OR = 2.12, 95% CI: 1.02, 4.42), and lymphoma (OR = 3.06, 95% CI = 1.20, 7.82). There was no observed association between childhood cancer and maternal hypothyroidism or hyperthyroidism; however, these results were likely underpowered. No increased risk of cancer was observed for maternal thyroid medication use. Though the small sample size must be considered, the elevated risk of cancer with maternal thyroid diseases, especially goiter, was notable.
Keywords: childhood cancer, Danish population‐base study, maternal disease, thyroid disease, thyroid medication
What's New?
Thyroid diseases are common among women of reproductive age and may be related to childhood cancer, but the evidence is inconsistent. This study of childhood cancer with maternal thyroid diseases and medication use in a population‐based case‐control study using Danish registries found positive associations between maternal goiter and overall cancer, leukemia, and lymphoma in the offspring. No increased risk of childhood cancer was observed with maternal thyroid medication use. Furthermore, no association was observed between childhood cancer and maternal hypothyroidism or hyperthyroidism, probably due to limited statistical power.
Thyroid diseases are common among women of reproductive age and may be related to childhood cancer, but the evidence is inconsistent. This study of childhood cancer with maternal thyroid diseases and medication use in a population‐based case‐control study using Danish registries found positive associations between maternal goiter and overall cancer, leukemia, and lymphoma in the offspring. No increased risk of childhood cancer was observed with maternal thyroid medication use. Furthermore, no association was observed between childhood cancer and maternal hypothyroidism or hyperthyroidism, probably due to limited statistical power.
Abbreviations
- ALL
acute lymphoblastic leukemia
- ATC
anatomical therapeutic chemical
- BMI
body mass index
- CI
confidence interval
- ICD‐10: revision 10
international classification of diseases
- ICD‐O
international classification of diseases for oncology
- OR
odds ratio
- TH
thyroid hormone
1. Introduction
Cancer is the leading cause of disease‐related childhood death, second only to accidents in high income countries [1]. Emerging evidence suggests that in utero exposure to maternal chronic conditions and medication use might increase the risk of childhood cancer in offspring [2].
Between 1997 and 2003, 2.1% of women were diagnosed with thyroid disease before or during pregnancy in Denmark [3]. Even though antithyroid use is relatively low, as only 0.2% of Danish children were exposed to maternal antithyroid use in early pregnancy from 1997 to 2016, the necessity and safety of thyroid hormones and antithyroid medication use during pregnancy is an important issue to explore [4]. Maternal thyroid diseases and medications can impact pregnancy outcomes, as they have been shown to increase the risks of maternal gestational diabetes and hypertension, as well as adverse birth outcomes such as preterm birth and abnormal birth weight, which are also associated with elevated risks of childhood cancer in the offspring [2, 5, 6].
The evidence for potential effects of maternal thyroid diseases or medications on childhood cancer risk is limited and mixed. A French study reported associations between acute lymphoblastic leukemia (ALL) and self‐reported family history of overall and autoimmune thyroid diseases [7], whereas two other studies did not detect any association between record‐based maternal autoimmune thyroid diseases and ALL [8], or between self‐reported family history of thyroid diseases (hyperthyroidism or Graves' disease, thyroiditis or Hashimoto's disease, and hypothyroidism) and overall childhood cancer [9]. A registry‐based Finnish study found associations between maternal hypothyroidism and any cancer as well as maternal thyroid disease and pediatric lymphomas, but no association for other cancer types [6]. Maternal use of thyroid hormone was not associated with leukemia or neuroblastoma among offspring in North America [10, 11], but was associated with an increased risk of cancer in Finland [6]. Our study aims to investigate the association of childhood cancer risk with maternal thyroid diseases and medication use.
2. Materials and Methods
This population‐based matched case–control study relied on the linkage of multiple national registries in Denmark, as previously described [1]. Cases of childhood cancer (0–19 years) were diagnosed between 1973 and 2016 and ascertained from the Danish Cancer Registry based on the International Classification of Diseases for Oncology (ICD‐O) until 2003 and the International Classification of Diseases (ICD), Revision 10 (ICD‐10) after that; the subtypes were classified based on the International Classification of Childhood Cancer [12]. Controls were randomly selected from the Central Population Registry and individually matched to cases (25:1 matching rate) on sex and date of birth. Controls were selected separately for each cancer subtype. Eligible controls were alive and free of cancer on the date that their matched case was diagnosed. All participants were born in Denmark between 1995 and 2014, and cases were diagnosed between 1996 and 2016. Pregnancies with gestational age under 20 weeks or over 46 weeks, or a birth weight below 500 g or above 6800 g (n = 11), were excluded from the study. The final sample included 2521 cases and 63,014 controls.
We ascertained diagnosis of any thyroid disease [3], hyperthyroidism (ICD‐8 codes: 242.00–242.29; ICD‐10 codes: E05–E05.9, excluding E05.4, E05.8A and E05.9A), hypothyroidism (ICD‐8 codes: 243.99, 244.00–244.09, excluding 244.02; ICD‐10 codes: E03–E03.9 and E89.0, excluding E03.0A, E03.1B, and E03.4) [13], goiter and its subtypes from the National Patient Register using the Danish extended version of ICD‐8 and ICD‐10. Simple/unspecified goiter was identified using ICD‐8 codes 240.0 and 240.9 and ICD‐10 codes E01.0, E01.2, E04.0, E04.8, and E04.9; non‐toxic goiters were identified using ICD‐8 code 241 and ICD‐10 codes E01.1, E04.1, and E04.2 [14, 15]. We only included thyroid disease categories with prevalence comparable to prior literature. Although we attempted to derive autoimmune thyroid diseases, the prevalence in our study was substantially lower than what has been reported in other Danish studies. Therefore, we decided not to include autoimmune thyroid disease as a separate category. We included the results examining any thyroid diseases (all types combined) to enable comparison with findings from other studies (Table S1). For those diagnosed with both hyperthyroidism and hypothyroidism, we classified exposure based on the most recent diagnosis within the exposure period since patients with hyperthyroidism are sometimes prescribed both antithyroid medications and thyroid hormone, which is called block‐and‐replace therapy [16]. Information on prescriptions was obtained from the National Prescription Register (established 1994) which provides coverage for 97.5% of the Danish population [17]. Prescriptions of thyroid medications were identified by using Anatomical Therapeutic Chemical (ATC) codes (thyroid hormone for hypothyroidism: H03A; antithyroid drug for hyperthyroidism: H03B) [3]. The exposure period for maternal thyroid diseases was defined as mother's lifetime up until 6 months postpartum, whereas the exposure period for maternal thyroid medication use was defined as pregnancy through 6 months postpartum, in order to account for the breastfeeding time period and the high prevalence of breastfeeding in Denmark [18]. Information on covariates including parental age, birth weight, birth order, mother's place of birth, urbanicity of birth residence, pre‐pregnancy BMI, maternal smoking status, child's birth weight and gestational weeks was obtained from the Central Population Registry and the Medical Birth Registry.
We had a priori hypothesized that maternal thyroid diseases and related medication use would be associated with increased risks of childhood cancer. We performed conditional logistic regression to estimate odds ratios (ORs) and corresponding 95% confidence intervals (CIs) for the associations of maternal thyroid diseases and medication use with the risk of overall and different types of childhood cancer. Selection of covariates for potential confounding control was based on the previous literature [2]. We used a uniform adjustment set for all cancer subtypes. Maternal age (continuous) and birth order (> 1 vs. 1) were included in all models. Maternal thyroid disease and medication use (yes/no) were also mutually adjusted for in some models; specifically, antithyroid medication use with hyperthyroidism and separately thyroid hormone use with hypothyroidism. We tried adjustment for additional covariates, including maternal smoking (yes/no), maternal pre‐pregnancy BMI (continuous), maternal history of thyroid cancer (yes/no), and maternal diabetes (yes/no), but they were not included in the final models because they did not change the results substantially [19]. We included the results adjusted for maternal diabetes (Table S2) and the results excluding participants with a maternal history of cancer (Table S3). For maternal smoking and pre‐pregnancy BMI, we conducted sensitivity analyses restricted to participants with non‐missing data (N = 21,794). The results were underpowered and therefore not presented. We did not adjust for iodine deficiency since no epidemiologic studies have directly examined the association between iodine deficiency or iodine supplementation during pregnancy and the risk of pediatric cancers included in our study.
Given the link between birth weight and pediatric leukemia, and the role of thyroid hormones in growth and differentiation, we evaluated whether birth weight mediated the association between thyroid diseases and cancer [20]. Since birth weight explained less than 1% of the association between thyroid diseases and cancer, we did not present the mediation results. Only cancer types with at least five exposed cases were kept in our analyses. All statistical analyses were conducted using SAS, Version 9.4 (SAS Institute Inc., Cary, NC, USA).
3. Results
Table 1 presents the general characteristics of cases and matched controls. Compared to controls, cancer cases were more likely to be first‐born, reside in urban areas, have a higher birth weight, have been born preterm, and have mothers with a higher pre‐pregnancy BMI. There were 1051 women diagnosed with any type of thyroid disease (Table 2). Among them, 41.1% of women were prescribed any type of thyroid medication, with 11.9% and 30.6% prescribed antithyroid and thyroid hormone, respectively. Among 386 women with hyperthyroidism, 29.5% were prescribed antithyroid medication. Among women with hypothyroidism, 71.4% were prescribed thyroid hormone. Only 21.2% of women with goiter were prescribed any type of thyroid medication, likely because benign goiter is usually small and asymptomatic and does not require treatment during pregnancy [16]. A small number of patients presented with both goiter and hypothyroidism concurrently, or with goiter and hyperthyroidism concurrently. Specifically, among these 87 goiter patients who received thyroid hormone or antithyroid medication during pregnancy, 36 were also analyzed as hyperthyroidism and 28 as hypothyroidism based on the ICD codes.
TABLE 1.
Characteristics of participants by cancer case and control status, births 1995–2014 (N = 65,535).
| Case | Control | |
|---|---|---|
| Number | 2521 | 63,014 |
| Maternal age at time of index pregnancy (years), mean (SD) | 29.8 (4.8) | 29.9 (4.8) |
| Paternal age at time of index pregnancy (years), mean (SD) | 32.4 (5.7) | 32.5 (5.8) |
| First born child, n (%) | 1074 (42.6) | 25,398 (40.3) |
| Mother's place of birth, n (%) | ||
| Denmark | 2205 (88.0) | 55,434 (88.1) |
| Other Europe & North America | 111 (4.4) | 2735 (4.4) |
| Other | 189 (7.5) | 4746 (7.5) |
| Missing | 16 | 99 |
| Urbanicity of residence, n (%) | ||
| Urban | 896 (36.1) | 22,045 (35.0) |
| Small town | 677 (27.3) | 17,671 (28.0) |
| Rural | 910 (36.6) | 23,298 (37.0) |
| Missing | 38 | 0 |
| Pre‐pregnancy BMI* (kg/m2), n (%) | ||
| 18.4 or less | 37 (4.2) | 887 (4.2) |
| 18.5–25 | 528 (60.6) | 13,591 (63.7) |
| 25+ | 307 (35.2) | 6872 (32.2) |
| Missing | 186 | 5094 |
| Smoking status, n (%) | ||
| Yes | 487 (20.1) | 12,449 (20.6) |
| No | 1934 (79.9) | 48,079 (79.4) |
| Missing | 100 | 2486 |
| Child's birth weight (grams), n (%) | ||
| < 1500 | 21 (0.8) | 405 (0.7) |
| 1500–2500 | 111 (4.4) | 2718 (4.3) |
| 2501–4000 | 1811 (72.5) | 47,289 (75.6) |
| > 4000 | 556 (22.3) | 12,153 (19.4) |
| Missing | 22 | 449 |
| Gestational weeks (weeks), n (%) | ||
| Very preterm (< 33) | 36 (1.4) | 752 (1.2) |
| Preterm (33–36) | 149 (5.9) | 3293 (5.2) |
| Term (37–42) | 2333 (92.6) | 58,895 (93.5) |
| Late (43+) | < 5 (0.1) | 74 (0.1) |
| Maternal thyroid diseases, n (%) | ||
| Hyperthyroidism | 16 (0.6) | 370 (0.6) |
| Hypothyroidism | 7 (0.3) | 329 (0.5) |
| Goiter | 31 (1.2) | 379 (0.6) |
| Non‐toxic goiter | 17 (0.7) | 221 (0.4) |
| Simple/unspecified goiter | 16 (0.6) | 190 (0.3) |
| Maternal thyroid medication use, n (%) | 18 (0.7) | 611 (1.0) |
| Antithyroid | 7 (0.3) | 144 (0.2) |
| Thyroid hormone | 9 (0.4) | 395 (0.6) |
Pre‐pregnancy BMI was collected from 2003 to 2014.
TABLE 2.
Types of thyroid disease and medication use among mothers (N = 65,535).
| Any thyroid disease (N = 1051) | Hyperthyroidism (N = 386) | Hypothyroidism (N = 336) | Goiter (N = 410) | |
|---|---|---|---|---|
| N (%) | N (%) | N (%) | N (%) | |
| No thyroid medication | 619 (58.9) | 231 (59.8) | 93 (27.7) | 323 (78.8) |
| Any thyroid medication | 432 (41.1) | 155 (40.2) | 243 (72.3) | 87 (21.2) |
| Antithyroid | 125 (11.9) | 114 (29.5) | < 5 (1.2) | 26 (6.3) |
| Thyroid hormone | 322 (30.6) | 55 (14.3) | 240 (71.4) | 65 (15.9) |
Without adjustment for medications, goiter was strongly associated with more than a 2‐fold increase in the odds of any cancer (OR = 2.08, 95% CI = 1.44, 3.00), leukemia (OR = 2.18, 95% CI = 1.13, 4.18), and ALL (OR = 2.12, 95% CI = 1.02, 4.42), as well as a 3‐fold increase in the odds of lymphoma (OR = 3.06, 95% CI = 1.20, 7.82) (Table 3). An elevated odds of overall cancer was observed among offspring of women with non‐toxic goiter (OR = 1.96, 95% CI = 1.20, 3.23). About two‐fold higher odds of overall cancer (OR = 2.11, 95% CI = 1.27, 3.53) and leukemia (OR = 2.32, 95% CI = 1.00, 5.37) were found among offspring of women with simple/unspecified goiter. Use of any maternal thyroid medications was associated with a decreased risk of childhood cancer after adjusting for thyroid diseases (OR = 0.51, 95% CI = 0.30, 0.87). However, no association was found between subtypes of thyroid medications and overall cancer risk either before or after additional adjustment or maternal thyroid diseases.
TABLE 3.
Odds ratios (ORs) and 95% confidence intervals (CIs) for maternal thyroid diseases or medication use and risk of offspring cancers, births 1995–2014 (N = 65,535).
| Thyroid disease/medication | Crude model | Adjusted model a | Adjusted model | |
|---|---|---|---|---|
| N (%) | OR (95% CI) | OR (95% CI) | OR (95% CI) | |
| Hyperthyroidism* | Adjusted model b | |||
| Controls | 370 (0.6) | Ref | Ref | Ref |
| All cancers | 16 (0.6) | 1.09 (0.66, 1.80) | 1.10 (0.67, 1.82) | 1.04 (0.58, 1.86) |
| Leukemia | 6 (0.9) | 1.22 (0.54, 2.78) | 1.22 (0.53, 2.78) | 1.43 (0.60, 3.41) |
| Hypothyroidism** | Adjusted model c | |||
| Controls | 329 (0.5) | Ref | Ref | |
| All cancers | 7 (0.3) | 0.53 (0.25, 1.12) | 0.53 (0.25, 1.13) | 0.67 (0.27, 1.67) |
| Goiter | ||||
| Controls | 379 (0.6) | Ref | Ref | Ref |
| All cancers | 31 (1.2) | 2.06 (1.43, 2.98) | 2.08 (1.44, 3.00) | — |
| Leukemia | 10 (1.4) | 2.22 (1.16, 4.27) | 2.18 (1.13, 4.18) | — |
| Acute lymphoblastic leukemia | 8 (1.5) | 2.16 (1.04, 4.49) | 2.12 (1.02, 4.42) | — |
| Lymphoma | 5 (1.8) | 3.06 (1.20, 7.80) | 3.06 (1.20, 7.82) | — |
| Non‐toxic goiter | ||||
| Controls | 221 (0.4) | Ref | Ref | Ref |
| All cancers | 17 (0.7) | 1.95 (1.19, 3.20) | 1.96 (1.20, 3.23) | — |
| Simple/unspecified goiter | ||||
| Controls | 190 (0.3) | Ref | Ref | Ref |
| All cancers | 16 (0.6) | 2.10 (1.26, 3.51) | 2.11 (1.27, 3.53) | — |
| Leukemia | 6 (0.9) | 2.35 (1.02, 5.45) | 2.32 (1.00, 5.37) | — |
| Any thyroid medication | Adjusted model d | |||
| Controls | 611 (1.0) | Ref | Ref | |
| All cancers | 18 (0.7) | 0.73 (0.46, 1.18) | 0.74 (0.46, 1.18) | 0.51 (0.30, 0.87) |
| Antithyroid | Adjusted model e | |||
| Controls | 144 (0.2) | Ref | Ref | Ref |
| All cancers | 7 (0.3) | 1.22 (0.57, 2.61) | 1.23 (0.57, 2.62) | 1.19 (0.50, 2.86) |
| Thyroid hormone | Adjusted model f | |||
| Controls | 395 (0.6) | Ref | Ref | Ref |
| All cancers | 9 (0.4) | 0.57 (0.29, 1.11) | 0.57 (0.29, 1.11) | 0.71 (0.32, 1.58) |
We excluded individuals with thyroid diseases other than hyperthyroidism to ensure that the unexposed group = represented women without indications for antithyroid medications. Thus, only women with hyperthyroidism or with no thyroid disease were included in this analysis.
We excluded individuals with thyroid diseases other than hypothyroidism to ensure that the unexposed group = represented women without indications for thyroid hormones. Thus, only women with hypothyroidism or with no thyroid disease were included in this analysis.
Adjusted for maternal age (continuous) and firstborn (yes/no).
Adjusted for maternal age (continuous), firstborn (yes/no), and maternal antithyroid medication use (yes/no).
Adjusted for maternal age (continuous), firstborn (yes/no), and maternal thyroid hormone use (yes/no).
Adjusted for maternal age (continuous), firstborn (yes/no), and any maternal thyroid disease (yes/no).
Adjusted for maternal age (continuous), firstborn (yes/no), and maternal hyperthyroidism (yes/no).
Adjusted for maternal age (continuous), firstborn (yes/no), and maternal hypothyroidism (yes/no).
For any type of maternal thyroid disease, we found an increased odds of childhood cancer overall (OR = 1.32, 95% CI = 0.99, 1.75) and this association became stronger after additionally adjusting for thyroid medication use (OR = 1.65, 95% CI = 1.20, 2.28) (Table S1). However, the positive association with neuroblastoma (OR = 2.56, 95% CI = 1.08, 6.09) did not change substantially after further adjustment for medications (OR = 2.53, 95% CI = 0.91, 7.03). After adjusting for medication, any thyroid diseases appeared to be associated with leukemia (OR = 1.77, 95% CI = 1.01, 3.10), especially ALL (OR = 1.82, 95% CI = 0.97, 3.40).
4. Discussion
Our study of childhood cancer with maternal thyroid diseases and medication use in a population‐based case‐control study in Denmark found positive associations of maternal thyroid goiter with overall cancer, leukemia, and lymphoma. No association was observed between childhood cancer and maternal hypothyroidism or hyperthyroidism, probably due to limited statistical power. To our knowledge, this is the first study to link maternal goiter to childhood cancer. We are not aware of studies linking maternal iodine deficiency to childhood cancer.
In our study population, the prevalence of maternal thyroid diseases and medication use was close to what has been reported previously for Denmark [3], with a prevalence of goiter (0.6%) being also comparable to what was found for women in most European countries (0.5%–0.7%), but was lower than that in the United States (approximately 0.8%) [21]. The most common cause of goiter is iodine deficiency, and iodine deficiency has been observed in Danish pregnant women [22]. A voluntary iodine fortification program was initiated in 1998 and became mandatory in 2000 to decrease mild–moderate iodine deficiency in Denmark, after which the prevalence of goiter decreased in western Denmark (median thyroid volume from 13.6 mL to 11.6 mL) and eastern Denmark (from 11.9 to 11.2 mL) [22].
Discrepancies in our results compared to previous studies may be due to limited power, different ascertainment methods and definitions of exposure, study design, and inappropriate pooling of thyroid diseases. Given the heterogeneity of different thyroid diseases, inappropriate pooling in prior studies may have obscured thyroid disease‐specific associations and contributed to inconsistent findings. The studies that explored the relationship between maternal thyroid hormone use and childhood cancer were limited in power [10, 11]. Most studies used parental interview, which may have been subject to reporting/recall bias or low response rates [7, 10]. Variation may also be attributable to exposures being defined broadly, such as family history instead of maternal history of thyroid disease, limiting comparability with our results [7, 9], or by the use of hospital controls [7]. All studies, including ours, have constraints in defining exposures and outcomes in terms of statistical power or the ability to obtain accurate exposure information. For example, we did not investigate autoimmune thyroid diseases as a separate category due to poor ascertainment in the available data. Although we reported overall thyroid disease results for comparison purposes, those results should be interpreted with caution given the variation across thyroid disease types.
Concerning potential mechanistic links between maternal thyroid diseases and abnormal levels of thyroid hormone (TH) and development of neuroblastoma, leukemia, and lymphoma, an in vitro study found that TH inhibits cell proliferation and promotes differentiation of neuroblastoma N2a‐β cells [23]. Kruppel‐like factor 9 (KLF9) is critical for normal hematopoiesis and is directly regulated by TH via a TH receptor, and TH deficiency impairs T‐ and B‐lymphocyte development or hematopoietic system development in humans [24], which may explain the association between goiter and leukemia. Another possible mechanism is that excessive iodine levels can disrupt thyroid hormone synthesis and act as a potential endocrine‐disrupting chemical. This effect, along with other thyroid abnormalities, may occur partially through oxidative stress pathways which are also biologically linked to leukemia [25]. For the association between maternal goiter and lymphoma, it is possible that goiter induces inflammation resulting in lymphoid neogenesis, which is linked to development of lymphoma [26]. We observed no evidence that maternal thyroid medication increased risk of childhood cancer among offspring. However, according to the American Thyroid Association, antithyroid drugs should still be taken with caution in pregnancy considering their potential teratogenicity and hepatoxicity [16].
This is the first study to mutually adjust for both thyroid diseases and medication use to address potential confounding by indication and provide risk–benefit information in terms of treatment plans in pregnancy for clinicians. Our study was population‐based, and exposure was prospectively obtained from national registries, which minimized the risk of selection bias and reporting/recall bias. However, our study was subject to several limitations. First, we were unable to investigate all subtypes of thyroid disease and childhood cancer. Despite the clinical importance of autoimmune thyroid diseases, such as Graves', Hashimoto's, and postpartum thyroiditis, we were unable to investigate them separately due to poor ascertainment and limited statistical power. Second, information on thyroid‐stimulating hormone levels was unavailable, which might cause misclassification of exposures. However, this misclassification should be non‐differential and bias the results towards the null. We might also have unmeasured confounding due to lack of information, such as maternal iodine intake in pregnancy.
In summary, positive associations were observed between maternal thyroid disease and the risks of overall cancer and neuroblastoma among offspring. Among offspring of mothers with goiter, the risks of leukemia and lymphoma were elevated. We found no evidence that thyroid medication use during pregnancy increased the cancer risk in offspring; however, use of antithyroid drugs requires caution due to their potential adverse effects on other health outcomes among offspring. Future studies incorporating information on thyroid‐stimulating hormone and iodine levels in pregnancy are needed to better understand the role of thyroid and iodine metabolism in the development of childhood cancer and the underlying biological mechanisms.
Author Contributions
Chuanjie Deng: writing – original draft, writing – review and editing, methodology, formal analysis, investigation, validation. Kosuke Inoue: writing – review and editing, methodology. Chai Saechao: writing – review and editing, methodology. Johnni Hansen: data curation, resources, writing – review and editing, validation, project administration. Beate Ritz: writing – review and editing, supervision, validation, methodology. Julia E. Heck: conceptualization, funding acquisition, supervision, resources, project administration, writing – review and editing, methodology, validation, investigation.
Funding
This work was supported by the National Institutes of Health, R21CA175959.
Ethics Statement
Our study was approved by the Office of the Human Research Protection Program, University of California, Los Angeles (IRB#13–001904), the University of North Texas (IRB‐20‐255), and the Danish Data Protection Agency.
Conflicts of Interest
Chuanjie Deng reports part‐time employment at AbbVie. AbbVie had no role in the study design, data collection, analysis, interpretation, or manuscript preparation. All other authors report no conflicts of interest.
Supporting information
Table S1: Odds ratios (ORs) and 95% confidence intervals (CIs) for overall maternal thyroid diseases and risk of offspring cancers, births 1995–2014 (N = 65,535).
Table S2: Odds ratios (ORs) and 95% confidence intervals (CIs) for maternal thyroid diseases or medication use and risk of offspring cancers, additionally adjusting for maternal diabetes, births 1995–2014 (N = 65,535).
Table S3: Odds ratios (ORs) and 95% confidence intervals (CIs) for maternal thyroid diseases or medication use and risk of offspring, excluding participants with maternal history of cancer, births 1995–2014 (N = 64,883).
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request and permission according to General Data Protection Regulation (GDPR).
References
- 1. Orimoloye H. T., Nguyen N., Deng C., et al., “Maternal Autoimmune Disease and Its Association With Childhood Cancer: A Population‐Based Case‐Control Study in Denmark,” EJC Paediatric Oncology 3 (2024): 100145, 10.1016/j.ejcped.2024.100145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Ricci A. M., Emeny R. T., Bagley P. J., et al., “Causes of Childhood Cancer: A Review of the Recent Literature: Part I—Childhood Factors,” Cancers 16, no. 7 (2024): 1297, 10.3390/cancers16071297. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Andersen S. L., Olsen J., and Laurberg P., “Maternal Thyroid Disease in the Danish National Birth Cohort: Prevalence and Risk Factors,” European Journal of Endocrinology 174, no. 2 (2016): 203–212, 10.1530/EJE-15-0816. [DOI] [PubMed] [Google Scholar]
- 4. Andersen S. L., Knøsgaard L., Olsen J., Vestergaard P., and Andersen S., “Maternal Thyroid Function, Use of Antithyroid Drugs in Early Pregnancy, and Birth Defects,” Journal of Clinical Endocrinology and Metabolism 104, no. 12 (2019): 6040–6048, 10.1210/jc.2019-01343. [DOI] [PubMed] [Google Scholar]
- 5. Lee S. Y., Cabral H. J., Aschengrau A., and Pearce E. N., “Associations Between Maternal Thyroid Function in Pregnancy and Obstetric and Perinatal Outcomes,” Journal of Clinical Endocrinology and Metabolism 105, no. 5 (2019): e2015–e2023, 10.1210/clinem/dgz275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Seppälä L. K., Madanat‐Harjuoja L. M., Leinonen M. K., Lääperi M., and Vettenranta K., “Maternal Thyroid Disease and the Risk of Childhood Cancer in the Offspring,” Cancers 13, no. 21 (2021): 5409, 10.3390/cancers13215409. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Perillat‐Menegaux F., Clavel J., Auclerc M. F., et al., “Family History of Autoimmune Thyroid Disease and Childhood Acute Leukemia,” Cancer Epidemiology, Biomarkers & Prevention 12, no. 1 (2003): 60–63. [PubMed] [Google Scholar]
- 8. Westbom L., Aberg A., and Källén B., “Childhood Malignancy and Maternal Diabetes or Other Auto‐Immune Disease During Pregnancy,” British Journal of Cancer 86, no. 7 (2002): 1078–1080, 10.1038/sj.bjc.6600192. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Zierhut H., Linet M. S., Robison L. L., Severson R. K., and Spector L., “Family History of Cancer and Non‐Malignant Diseases and Risk of Childhood Acute Lymphoblastic Leukemia: A Children's Oncology Group Study,” Cancer Epidemiology 36, no. 1 (2012): 45–51, 10.1016/j.canep.2011.06.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cook M. N., Olshan A. F., Guess H. A., et al., “Maternal Medication Use and Neuroblastoma in Offspring,” American Journal of Epidemiology 159, no. 8 (2004): 721–731, 10.1093/aje/kwh108. [DOI] [PubMed] [Google Scholar]
- 11. Ross J. A., Xie Y., Davies S. M., ou S. X., Pendergrass T. W., and Robison L. L., “Prescription Medication Use During Pregnancy and Risk of Infant Leukemia (United States),” Cancer Causes & Control 14, no. 5 (2003): 447–451, 10.1023/a:1024953532355. [DOI] [PubMed] [Google Scholar]
- 12. Steliarova‐Foucher E., Stiller C., Lacour B., and Kaatsch P., “International Classification of Childhood Cancer, Third Edition,” Cancer 103, no. 7 (2005): 1457–1467, 10.1002/cncr.20910. [DOI] [PubMed] [Google Scholar]
- 13. Andersen S. L., Laurberg P., Wu C. S., and Olsen J., “Maternal Thyroid Dysfunction and Risk of Seizure in the Child: A Danish Nationwide Cohort Study,” Journal of Pregnancy 2013 (2013): 636705, 10.1155/2013/636705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Brandt F., Thvilum M., Almind D., et al., “Graves' Disease and Toxic Nodular Goiter Are Both Associated With Increased Mortality but Differ With Respect to the Cause of Death: A Danish Population‐Based Register Study,” Thyroid 23, no. 4 (2013): 408–413, 10.1089/thy.2012.0500. [DOI] [PubMed] [Google Scholar]
- 15. Kitahara C. M., Körmendiné Farkas D., Jørgensen J. O. L., Cronin‐Fenton D., and Sørensen H. T., “Benign Thyroid Diseases and Risk of Thyroid Cancer: A Nationwide Cohort Study,” Journal of Clinical Endocrinology and Metabolism 103, no. 6 (2018): 2216–2224, 10.1210/jc.2017-02599. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Alexander E. K., Pearce E. N., Brent G. A., et al., “2017 Guidelines of the American Thyroid Association for the Diagnosis and Management of Thyroid Disease During Pregnancy and the Postpartum,” Thyroid 27, no. 3 (2017): 315–389, 10.1089/thy.2016.0457. [DOI] [PubMed] [Google Scholar]
- 17. Bliddal M., Broe A., Pottegård A., Olsen J., and Langhoff‐Roos J., “The Danish Medical Birth Register,” European Journal of Epidemiology 33, no. 1 (2018): 27–36, 10.1007/s10654-018-0356-1. [DOI] [PubMed] [Google Scholar]
- 18. Hörnell A. and Lagström H., “Infant Feeding—A Scoping Review for Nordic Nutrition Recommendations 2023,” Food and Nutrition Research 68 (2024): 10–29219, 10.29219/fnr.v68.10456. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Greenland S., “Modeling and Variable Selection in Epidemiologic Analysis,” American Journal of Public Health 79, no. 3 (1989): 340–349, 10.2105/ajph.79.3.340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lei U., Wohlfahrt J., Hjalgrim H., Hjalgrim L. L., Simonsen H., and Melbye M., “Neonatal Level of Thyroid‐Stimulating Hormone and Acute Childhood Leukemia,” International Journal of Cancer 88, no. 3 (2000): 486–488. [PubMed] [Google Scholar]
- 21. Malboosbaf R., Hosseinpanah F., Mojarrad M., Jambarsang S., and Azizi F., “Relationship Between Goiter and Gender: A Systematic Review and Meta‐Analysis,” Endocrine 43, no. 3 (2013): 539–547, 10.1007/s12020-012-9831-8. [DOI] [PubMed] [Google Scholar]
- 22. Tang Møllehave L., Knudsen N., Linneberg A., et al., “The Danish Investigation on Iodine Intake and Thyroid Disease (DanThyr): History and Implications,” European Thyroid Journal 13, no. 3 (2024): e230230, 10.1530/ETJ-23-0230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Garcia‐Silva S., Perez‐Juste G., and Aranda A., “Cell Cycle Control by the Thyroid Hormone in Neuroblastoma Cells,” Toxicology 181‐182 (2002): 179–182, 10.1016/s0300-483x(02)00277-9. [DOI] [PubMed] [Google Scholar]
- 24. Zhang Y., Xue Y., Cao C., et al., “Thyroid Hormone Regulates Hematopoiesis via the TR‐KLF9 Axis,” Blood 130, no. 20 (2017): 2161–2170, 10.1182/blood-2017-05-783043. [DOI] [PubMed] [Google Scholar]
- 25. Karbownik‐Lewińska M., Stępniak J., Iwan P., and Lewiński A., “Iodine as a Potential Endocrine Disruptor—A Role of Oxidative Stress,” Endocrine 78, no. 2 (2022): 219–240, 10.1007/s12020-022-03107-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Bende R. J., van Maldegem F., and van Noesel C. J. M., “Chronic Inflammatory Disease, Lymphoid Tissue Neogenesis and Extranodal Marginal Zone B‐Cell Lymphomas,” Haematologica 94, no. 8 (2009): 1109–1123, 10.3324/haematol.2009.005983. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Odds ratios (ORs) and 95% confidence intervals (CIs) for overall maternal thyroid diseases and risk of offspring cancers, births 1995–2014 (N = 65,535).
Table S2: Odds ratios (ORs) and 95% confidence intervals (CIs) for maternal thyroid diseases or medication use and risk of offspring cancers, additionally adjusting for maternal diabetes, births 1995–2014 (N = 65,535).
Table S3: Odds ratios (ORs) and 95% confidence intervals (CIs) for maternal thyroid diseases or medication use and risk of offspring, excluding participants with maternal history of cancer, births 1995–2014 (N = 64,883).
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request and permission according to General Data Protection Regulation (GDPR).