Thyroid Hormone Status in Umbilical Cord Serum Is Positively Associated with Male Anogenital Distance

Abstract

Context:

In human adults and adolescents, thyroid function affects sex hormones and male reproductive functions. Little is known about the thyroid function effects on the gonadal development in human infants.

Objective:

The aim was to examine the association between thyroid hormones (THs) and sexually dimorphic genital development or fetal growth.

Design:

This is a birth cohort study.

Participants:

A total of 616 mothers and newborns were analyzed from two local hospitals.

Main Outcome Measures:

TSH, free T₃ (FT3), and free T₄ (FT4) levels in cord blood serum, anogenital distance (AGD), birth weight, birth length, birth body mass index, and head circumference in neonates.

Results:

Longer AGD in male newborns was observed with higher cord serum FT3 (β, 1.36 mm [95% confidence interval (CI), 0.58–2.13] for 1 pmol/L FT3), FT4 (β, 0.12 mm [95% CI, 0.00–0.25] for 1 pmol/L FT4), and TSH (β, 3.14 mm [95% CI, 0.65–5.63] for a 10-fold TSH increase), and with a lower FT4/FT3 ratio (β, −0.11 mm [95% CI, −0.20 to −0.02] for doubling FT4/FT3 ratio). The relationships between TSH, birth weight, and birth length were different by secondhand smoke exposure. Secondhand smoke exposure had an effect modification, with interaction P value .039 and .010, respectively. Secondhand smoke exposure also had an effect modification on the relation between FT4 and head circumference with interaction P value .020.

Conclusions:

In the absence of overt thyroid dysfunction, THs are positively associated with AGD in male newborns. TH effects on body size and head circumference may be modified by maternal secondhand smoke exposure.

Abstract

THs are positively associated with anogenital distance in male newborns without overt thyroid dysfunction, suggesting an interplay between thyroid and androgen hormonal systems during development.

Thyroid hormones (THs), including TSH, T₃, T₄, free T₃ (FT3), and free T₄ (FT4), are detectable in the fetal circulation, beginning early in gestation, and have important developmental, metabolic, and maturational effects on the fetus. However, the influence of THs on fetal reproductive development in human infants remains unclear.

Anogenital distance (AGD) at birth is used as a noninvasive indicator of fetal reproductive development (1). It reflects perineal growth and caudal migration of the genital tubercle (2). Measurement of AGD at birth is an effective method of “looking back in time” to identify whether a hormonally disruptive action had occurred during gestation (3). AGD also reflects reproductive health and function over the life course. A number of studies have confirmed that reduced AGD is associated with hypospadias, cryptorchidism, and poorer semen quality in humans (2, 4). Most studies found a positive relationship between AGD and androgens (5, 6), but other studies didn't (7, 8). Moreover, Adibi et al (9) found that the first trimester human chorionic gonadotropin (hCG) levels were also correlated with the AGD in neonates.

The thyroid gland is the first endocrine gland to develop in the human embryo. This development starts with a thickening of the floor of the primitive pharynx between the diverging aorta at about day 22 after conception (10). The capacity to incorporate iodine into THs is delayed until weeks 10–12 (10). In the first trimester and early in the second trimester, the fetus relies entirely on the maternal supply of THs for normal development. From the middle of the second trimester onward, circulating THs of both maternal and fetal origin are present in the fetus (11). As is reported, one-third to one-half of the T₄ in cord blood came from the mother (12). Also, some T₃ crosses the placenta from the mother by monocarboxylate transporters 8 and 10 (13). TH action is primarily dependent upon intracellular concentrations of FT3 and thyroid hormone receptor-bound T₃. These intracellular T₃ levels are regulated by the supply of circulating T₄ and T₃ and the conversion of T₄ by deiodination to T₃.

Clinical research has shown that in human adults, THs affect sex hormone metabolism and synthesis, including but not limited to T metabolism, the free T/free estradiol ratio, the responses of LH and FSH to GnRH administration, resulting in gynecomastia, decreased semen quality, and sexuality (14). Few clinical research studies have been conducted to prove the relationship of fetal sexual development and THs because fetal sexual development information and TH levels are difficult to obtain noninvasively. In fact, some evidence shows that THs may affect fetal sexual development. Researchers found that TH receptors were present in human testes at birth (15) and T₄ and T₃ initiated an immediate rapid effect on Sertoli cells by nongenomic mechanisms (16). Reliable clinical research of large sample size on the relationship of fetal THs and sexual development needs to be carried out.

THs also regulate the development of body size. Maternal hypothyroidism is known to have serious adverse effects on the fetus; even subclinical hypothyroidism or euthyroidism may be associated with adverse outcomes including, but not limited to, premature birth, miscarriage, low birth weight, and stillbirth (17, 18). In adults, studies have characterized the relationship between THs and body mass index (BMI) or body weight, but the results are not consistent (19–21). Studies have shown that cigarette exposure may influence thyroid function (22). In newborns, some others found that cord FT4 is associated with birth weight and length (23).

Our aims in this study are to examine: 1) whether THs relate to fetal AGD, and if the relationship is sex dependent; and 2) whether THs relate to fetal body size, including head circumference, birth weight, birth length and birth BMI, and if secondhand smoke modifies the relationship.

Subjects and Methods

Study population

Subjects enrolled in the study were healthy pregnant women who delivered in a local hospital in Guiyu and Haojiang between May 2011 and May 2012. Both towns are located in Shantou, Guangdong, China. Guiyu is one of the largest e-waste recycling centers in the world, described in our previous studies (8, 24, 25). Before sampling, approval allowing us to collect and use human cord blood was obtained from each subject, and the study was approved by the Human Ethics Committee of Shantou University Medical College, China, and the University of Cincinnati Institutional Review Board. All participants completed a questionnaire administered by trained research staff. The questionnaire acquired information about sociodemographic characteristics (maternal age, education level, and occupation) and lifestyle and dietary habits (maternal alcohol and meat consumption during the pregnancy, smoking habits during pregnancy, and family member smoking habits). If family members smoked more than three times a week, pregnant women were defined as having secondhand smoke exposure. Neonatal birth length, birth weight, head circumference, and AGD were measured by trained nurses. For males in this study, AGD was the distance from the center of the anus to the base of the scrotum where the skin changes from rugated to smooth. For females, AGD was the distance from the center of the anus to the base of the posterior fourchette where the skinfolds fuse (8). A total of 620 pregnant women were recruited and provided umbilical cord blood. We excluded one case whose cord serum TSH was above 60 mIU/L, similar to reported congenital hypothyroidism morbidity of 1:2000 to 1:3000 in China (26, 27). After excluding three multiple pregnancy cases, data from 616 subjects were analyzed.

Measurement of THs

All cord serum samples were stored at −80°C until THs were measured by an XH-6020 automatic gamma-immuno counter in the Department of Nuclear Medicine, the First Affiliated Hospital of Shantou University Medical College. Serum FT3 and FT4 concentrations were quantified by Iodine [¹²⁵I] Free Triiodothyronine Radioimmunoassay kits and Iodine [¹²⁵I] Free Thyroxine Radioimmunoassay kits (3V, Biotechnology Corp), respectively. Serum TSH concentrations were measured by solid phase immunoradiometric assay (3V, Biotechnology Corp).

Statistical analysis

The distribution of cord serum TSH concentrations was strongly skewed to the right and was log₁₀-transformed for analysis. FT4 and FT3 were expressed on an arithmetic scale. The relationships between TSH, FT4, and FT3 were also explored.

First, we compared the difference in maternal sociodemographic, behavioral, and perinatal factors, fetal body size, and TH levels between males and females, with or without exposure to secondhand smoke.

Second, before conducting multivariable linear regression analyses, we examined a number of covariates potentially predictive of AGD and body size, including the infant's gestational age at birth. We first fit the data to univariate models to explore the relationship of THs and AGD; we then used sex as a stratification variable to conduct multivariable linear regression analyses and ran models to examine the possible effect modification. We also standardized (Z-score) the dependent variable of TSH, FT3, and FT4. Then we fit the data to univariate models to explore the relationships of AGD and TSH Z-score, AGD and FT3 Z-score, and AGD and FT4 Z-score in male newborns. Furthermore, we added another standardized variable to analyze the relationship of AGD and THs in male newborns, termed the anogenital index (AGI = AGD/weight [mm/kg]), as described in previous literature, because AGD is affected by weight (28).

Third, we fit the data to univariate models to explore the relationship of THs and body size, used secondhand smoke exposure as a stratification variable to conduct multivariable linear regression analyses, and ran models to examine the possible effect modification.

We considered P values of .10 for interaction terms to be indicative of effect modification, as previously reported (29). All analyses were performed using Empower(R) (www.empowerstats.com; X&Y Solutions, Inc) and R (http://www.R-project.org).

Results

In this birth cohort, the mean ± SD of TSH was 5.87 ± 4.06 mIU/L, with the 25th, 50th, and 75th percentiles being 3.52, 4.79, and 6.81 mIU/L, respectively. The mean FT3 was 1.63 ± 0.71 pmol/L, with the 25th, 50th, and 75th percentiles being 1.05, 1.60, and 2.11 pmol/L, respectively. The mean FT4 was 15.61 ± 4.92 pmol/L, with the 25th, 50th, and 75th percentiles being 12.50, 15.30, and 18.51 pmol/L, respectively. FT3 is positively correlated to log₁₀TSH (r = 0.13; P < .001) and FT4 (r = 0.38; P < .001). FT4 did not correlate to log₁₀TSH (r = 0.05; P = .193).

Table 1 shows the maternal and infant characteristics stratified by sex and exposure to secondhand smoke.

Table 2 displays the regression results of AGD and THs stratified by sex. In male neonates, AGD increased 1.36 mm with a FT3 increase of 1 pmol/L, increased 0.12 mm with a FT4 increase of 1 pmol/L, increased 3.14 mm with a 10-fold increase in TSH, and decreased 0.11 mm when the FT4/FT3 ratio doubled. In female neonates, the relationship between AGD and THs was not statistically significant. The results of the relationships of AGD and TSH Z-score, AGD and FT3 Z-score, and AGD and FT4 Z-score in male newborns are β, 1.4 (95% CI, 0.7–2.1), P < .001; β, 0.7 (95% CI, 0.0–1.4), P = .067; and β, 0.9 (95% CI, 0.2–1.6), P = .009, respectively.

We further analyzed the influence of THs on AGI in male neonates (Table 3). The AGI increased with FT3 and TSH and decreased with the FT4/FT3 ratio when they were categorized in quartiles; the P values for the trends were all less than .05.

Birth weight was positively associated with FT4 (β, 8.48 g [95% CI, 1.28–15.69] per 1 pmol/L FT4; P = .021), after adjustment for gestational age and sex. We did not find other relationships between THs and body size (Table 4) until sex and secondhand smoke exposure were stratified (Table 5).

Table 5 shows the regression results of fetal growth and TSH level stratified by secondhand smoke exposure. Without secondhand smoke exposure, birth weight decreased 250.61 g with a 10-fold TSH increase after adjustment for gestational weeks, delivery mode, and sex. With exposure to secondhand smoke, the negative association disappeared, and the interaction P value was .039. Without secondhand smoke exposure, the relationship of birth length and TSH was negative (β, −0.96 cm [95% CI, −1.94 to 0.02] per 10-fold TSH increase; P = .056) after adjustment, but with secondhand smoke exposure, the relationship of birth length and TSH was positive (β, 1.06 cm [95% CI, −0.14 to 2.26] per 10-fold TSH increase; P = .084) (interaction P value = .010). Moreover, secondhand smoke exposure modified the relationship of head circumference and FT4 levels. Without secondhand smoke exposure, the relationship of head circumference and FT4 was positive (β, 0.04 cm [95% CI, 0.00–0.09] per 1 pmol/L FT4; P = .059); with secondhand smoke exposure, the relationship of head circumference and FT4 was negative (β, −0.03 cm [95% CI, −0.07 to 0.01) per 1 pmol/L FT4; P = .179). The interaction P value was .020 after adjustment for gestational age and sex.

Discussion

Our study found that the AGD in male neonates is positively associated with the levels of FT3, FT4, and TSH and negatively associated with the FT4/FT3 ratio. This study provides the first epidemiological evidence implying that THs may be involved in male gonad development in utero. Also, maternal passive smoking affects the relationship of TSH and birth length or birth weight and the relationship of FT4 and head circumference.

AGD increased with FT3 and FT4 in male but not female newborns. The first possibility is that elevated T₃ is affecting androgen levels and actions, with ultimate consequences for AGD. Fetal Leydig cells begin to synthesize androgens at 8 weeks and undergo continuous proliferation and differentiation until birth (14). T₃ affects Leydig cell differentiation and steroidogenesis, as well as Sertoli cell proliferation and functional maturation (30). A lack of T₃ is associated with decreased serum total T (31, 32). Also, T₃ enhanced androgen receptor expression, affected androgen-related gene expression, increased LH and FSH responses to GnRH administration, and enhanced the activity of cytochrome P450 and other enzymes involved in synthesizing androgen (32).

The second possibility is that elevated androgen, indicated by increased AGD, affects TH status. The FT4/FT3 ratio shows a negative association with AGD and AGI, suggesting a higher conversion from FT4 to FT3 in long AGD male newborns. The conversion of T₄ by enzymatic outer (phenolic) ring deiodination to T₃ is performed by type 1 deiodinase (D1) in the liver and kidney and by type 2 deiodinase (D2) in the brain, testis, and placenta. Therefore, high D1 and D2 activity played an important role in more conversion from FT4 to FT3 (33). T has been confirmed to modulate D1 activity. T treatment normalizes reduced liver D1 activity and mRNA levels in male castrated rats (34). Conversely, inhibition of androgen activity resulted in reduced expression of dio1 mRNA and dio2 mRNA (32). Type 3 deiodinase (D3) is a critical factor in testis development. The absence of D3 leads to a systemic impairment in T₃ clearance, elevated serum T₃ three times higher than normal mice, an approximately 75% reduction in testis size, and a decreased serum T in neonatal mice (35). D3 may contribute to AGD development, although we cannot assess this without measuring D3 activity.

Although there is substantial evidence to support interactions between the hypothalamic-pituitary-thyroid axis and the hypothalamic-pituitary-testicular axis to influence male AGD, our results cannot exclude other mechanisms. One possible mechanism is that T₃ stimulates the skeletal muscle growth by increasing the number and diameter of muscle fibers between the anus and the external genitalia around the perineal central tendon (36). However, it may not be the main reason because there was no difference in the influences of THs on skeletal muscle between male and female. The other mechanism may be that the placenta affects both fetal THs and fetal AGD. Placental chorionic villi excrete hCG and transfer T₄ and T₃ to the fetus from the mother (13). Because reduced secretion and transport abilities by placental chorionic villi may lead to the decrease of hCG and be followed by decreased AGD in male newborns (9), and because the cord FT4 level was positively associated with placental weight (23), we surmise that placental chorionic villi changes may lead to decreased transfer of T₄ and T₃ to fetal circulation, as well as decreased AGD.

TH levels measured in cord blood at birth are more representative of fetal thyroid status during gestation than those measured a few days after birth. Physiological neonatal TSH surge occurs shortly after birth, and cord TSH may not be reflective of fetal thyroid function during the in utero period. Also, TSH surge levels in cord blood were affected by delivery mode, passive smoking, and some other potential confounders (37). In our data, FT3 is the most important of these three variables. When we examined the relative strength of TSH, FT3, and FT4 Z-scores on AGD, the regression coefficients of FT3, FT4, and TSH were 1.4, 0.7, and 0.9, suggesting that FT3 was more predictive than TSH. The relationship of TSH and AGD may be due to the relationship between TSH and FT3.

FT4 is positively associated with birth weight, consistent with another report (23). When considering passive smoking by the mother, the mechanisms behind the opposite correlations between FT4 with head circumference and TSH with birth weight or birth length are not yet known. On the one hand, thiocyanate, one toxicant of passive tobacco smoking components, stimulates thyroglobulin secretion and inhibits iodide transport (38). As a result, this effect may alter fetal serum TH levels and TH action on growth. TH levels have been related to birth weight, birth length, and head circumference (23). On the other hand, smoking is related to lower birth weight, shorter birth length, and smaller head circumference (39–41). Taken together, exposure to secondhand tobacco smoke may alter the relationship of THs and birth weight, birth length, and head circumference. We did not find the correlation between TSH and fetal BMI, consistent with the results of Manji et al (21) and inconsistent with the results of Nyrnes et al (20). We also did not find effect modification of passive smoking on the relationship between TSH and fetal BMI, inconsistent with the results of Asvold et al (42). We suppose the reason is that potential confounders are more complicated in newborns than in adults, and toxic components of active smoking are different from those of passive smoking. More prospective studies with a large sample are required to clarify their relationship further.

The strengths of this study include a cohort study with a relatively large sample size and careful adjustments for gestational age, sex, and smoking. These results provide another possible explanation for the correlation of AGD and endocrine disruptors; that is, endocrine disruptions may affect both the thyroid axis and androgen. A limitation of this study is the lack of total T₄ levels that could supply us with more information. Another limitation is not using a sensitive and specific immunofluorometric assay. A third limitation is that we cannot measure fetal hormone concentrations during the fetal masculinization programming window because clinical technologies are not feasible in the general population to get amniotic fluid or umbilical cord blood in utero.

Conclusion

In conclusion, our results show that AGD increases with increasing serum TH levels and decreases with increasing FT4/FT3 ratio in male newborns, and that newborn growth is possibly related to THs. These results suggest that: 1) TH status in newborns is associated with AGD, suggesting an interplay between thyroid and androgen hormonal systems during development; and 2) TH effect on the growth of the fetus is vulnerable to interference by confounding factors, of which passive smoking by the mother is an important modifying factor.

Acknowledgments

We are grateful to the participants for participating in this study. We thank Dr. Stanley Lin for his constructive comments and language editing.

This work was supported by the U.S. National Institute of Environment Health Sciences (NIH 1RC4ES019755-01) and the Project of International Cooperation and Innovation Platform in Guangdong Universities (2013gjhz0007).

Disclosure Summary: The authors have nothing to disclose.

Abbreviations

 
• AGD

  anogenital distance

 
• AGI

  anogenital index

 
• BMI

  body mass index

 
• CI

  confidence interval

 
• D1

  D2, and D3, type 1, 2, and 3 deiodinase

 
• FT3

  free T₃

 
• FT4

  free T₄

 
• hCG

  human chorionic gonadotropin.

References