Abstract
Context
Growing preclinical evidence suggests that hormonal programming by androgens in utero may contribute to cardiovascular disease risk in adult offspring. However, the effect of prenatal androgens on cardiometabolic outcomes in the human population, especially their potential differential impact on male vs female offspring, has not been well studied.
Design
Adult offspring (n = 274) of mothers enrolled in the New England birth cohorts of the Collaborative Perinatal Project were assessed at ages 39 to 50. Androgen bioactivity was measured in maternal serum during the third trimester using a receptor-mediated luciferase expression bioassay. Metabolic syndrome (MetS) using Adult Treatment Panel III criteria was assessed in adult offspring. Bioactive androgens were analyzed as quartiles, with the lowest quartile (Q1) defined as the reference. Generalized estimating equations were used to evaluate the relationship of maternal bioactive androgens on offspring MetS risk overall and by sex, controlling for potential confounders and intrafamilial correlation.
Results
Mean age and body mass index of adult offspring were 44.7 ± 2.6 years and 29.7 ± 6.7 kg/m2, respectively. Participants born to mothers with the highest quartile (Q4) compared with Q1 of bioactive androgens had higher risk for MetS [adjusted odds ratio (aOR): 2.53(1.07 to 6.02)]. Stratified by sex, this association was found to be significant among women [Q4 vs Q1; aOR: 4.06 (1.10 to 14.93)] but not men [Q4 vs Q1; aOR: 1.67 (0.53 to 5.26)]. Women born to mothers with the highest levels of maternal bioactive androgens also demonstrated a 4.84-fold increased odds for having hypertension [Q4 vs Q1; aOR: 4.84 (1.12 to 20.85)].
Conclusion
Higher levels of maternal androgens were associated with increased risk for incident MetS in adult offspring, an effect that was significant in women but not men.
We studied the sex differences in prenatal programming of adult metabolic syndrome (MetS) by maternal androgens. Higher maternal androgen levels predicted MetS risk in adult female offspring.
Cardiovascular disease (CVD) is a major cause of mortality worldwide, with a prevalence that is higher overall in men than women (1), a sex difference that is attenuated after menopause (2, 3). A key factor underlying CVD is the metabolic syndrome (MetS), a constellation of five metabolic risk factors [hypertension, low high-density lipoprotein (HDL), hypertriglyceridemia, abdominal obesity, and impaired glucose metabolism], which is associated with a twofold increased risk of CVD (4). There is a growing body of evidence that the endocrine, nutritional, and metabolic milieu during fetal life programs lifelong effects on cardiometabolic risk in adulthood (5, 6). The Barker hypothesis (7) recognizes the importance of the environmental milieu in the developmental origins of adult disease. Although this hypothesis was originally based on nutritional deficits in utero, its concept has also been extended to intrauterine exposure to sex steroid hormones and metabolic disruption in adulthood.
Growing evidence suggests that prenatal exposure to androgens (either of maternal or fetal origin) could be a source of early programming, leading to the development of CVD and metabolic diseases later in adult life (8, 9). Several female animal and nonhuman primate models, including rhesus monkeys, rodents, and sheep exposed to prenatal androgen excess, develop adverse reproductive and cardiometabolic traits (hypertension, adiposity, insulin resistance) in adulthood, resembling women with polycystic ovary syndrome (PCOS; i.e., a common disorder of androgen excess in reproductive age women) (9–13). Likewise, prenatal androgen treatment is associated with more body fat, higher blood pressure, and the development of insulin resistance in male rats and rhesus monkeys (14, 15), similar to prenatally exposed females. Some (11, 16, 17) but not all (18, 19) studies have shown that women with PCOS have higher serum testosterone (T) levels during pregnancy compared with controls. Furthermore, sons and daughters born to women with PCOS tend to exhibit worse metabolic parameters than those born to healthy mothers (20, 21). These findings suggest the possibility that early exposure to high androgen levels during critical developmental periods of fetal life may contribute to increased risk for adverse cardiometabolic phenotypes in adulthood.
Concerns about increasing human exposure to endocrine-disrupting environmental factors that interact with the androgen receptor and/or modulate androgen receptor signaling have focused attention on the potential clinical and public health impact of developmental androgen exposure (22). Despite evidence from animal studies and specific patient populations (i.e., PCOS), there is a paucity of human studies investigating the role of prenatal androgens in the developmental programming of adult cardiometabolic disease. The sex differences in the associations between circulating androgen levels and the risk for MetS and CVD are well recognized (23). High androgen levels in postmenopausal women are associated with insulin resistance, MetS, visceral fat accumulation, and coronary heart disease (24). In contrast, epidemiological studies show that men with low total T levels have increased risk for insulin resistance, adiposity, and CVD (25), suggesting that the association between T levels and cardiometabolic phenotypes differs by sex. However, there have been few investigations evaluating associations of developmental prenatal exposure to androgens, from endogenous or exogenous sources, with sex-dependent cardiometabolic abnormalities in adult human offspring. The paucity of longitudinal offspring datasets followed into adulthood has rendered it difficult to test hypotheses about the effects of prenatal androgen exposure on adult metabolic phenotypes.
Thus, the objective of this study was to examine prospectively the sex-dependent associations between prenatal androgens and the risk of MetS in the adult offspring of mothers participating in the New England Family Study (NEFS). The current study took advantage of two unique resources: (1) a unique community-based population of mothers, who during and after their pregnancy, were part of the Collaborative Perinatal Project (CPP), which followed individuals in utero (born from 1959 to 1966), and whose offspring have been evaluated in adulthood ∼45 years later as part of the NEFS; and (2) use of a highly sensitive bioassay to measure total circulating androgen bioactivity in maternal serum. Bioactive androgen bioassays offer the advantage of measuring the relative activity of all androgens, including chemically unidentified environmental androgenic compounds (26). We hypothesized that higher levels of prenatal bioactive androgens assessed in maternal serum during the third trimester would be associated with increased risk of the development of the MetS in adulthood. We further hypothesized that the association of higher prenatal bioactive androgens with incident adult offspring MetS would be more pronounced among women than men.
Methods
Study sample
Study participants were adult offspring born to pregnant women between 1959 and 1966 as part of the New England cohorts of the CPP. The CPP prospectively followed participating mothers during their pregnancies, collected and stored their sera, and followed their offspring from delivery up to age 7 years (27, 28). There were 17,921 pregnancies enrolled at the Providence, Rhode Island, and Boston, Massachusetts, sites of the CPP, also known as the NEFS. Approximately 4000 NEFS offspring were located, consented, and were reassessed as adults (29). The participants in the current study were selected from the NEFS substudy on Shared Fetal Antecedents to Major Depression and Risk for CVD (National Institute of Mental Health–National Heart, Lung, and Blood Institute Grant R01MH074679 to J.M.G.), which included 318 offspring from live, singleton births (28). Participants in this substudy were term infants (38 to 43 weeks), originally selected as same-sex sibling sets discordant on fetal growth restriction (i.e., birth weight, adjusted for gestational age) and/or preeclampsia to create variability in levels of maternal steroid and immune responses during pregnancy and their association with adult outcomes. Selection and sampling for these NEF participants have been previously described (27, 28). Out of these 318 participants, 274 adult offspring had bioactive androgen data already available from assayed maternal prenatal serum, as well as data on MetS criteria, constituting our analytic sample. At the time of data collection for the current study, adult offspring were ∼39 to 50 years of age (mean = 45 years). All NEFS participants provided written, informed consent. There were no significant differences among maternal and pregnancy factors between the 318 and the 274 subset sample (data not shown).
Maternal bioactive androgens
From 1959 to 1966, maternal blood samples were collected approximately every 2 months during pregnancy and stored at the National Institutes of Health repositories at −20°C. Maternal serum samples for this study were obtained from the CPP central repository at the National Institutes of Health and assayed for androgen bioactivity at the beginning of the third trimester. Sera were drawn from the beginning of third trimester, given that it encompasses a critical, “sensitive” period of the sexual differentiation of the brain, i.e., organizational effects of gonadal steroids on brain development. In fact, the third trimester has also been found to be a critical period for maturation of other fetal physiological organ systems (30).
Bioactive androgens were measured using an androgen receptor-mediated assay developed from human embryonic kidney 293 cells stably cotransfected with a human androgen receptor plasmid and a luciferase reporter gene under the control of the mouse mammary tumor virus promoter (2933Y) (26). Luciferase activity was measured by a luminometer in cell lysates and expressed as relative light units. The amount of bioactive androgen in the sample was directly related to the relative light units detected. Dihydrotestosterone was the most potent steroid tested, followed by total T; 4-androstene-3, 17β-diol; 4-androstene-3, 17-dione; 5α-androstane-3, 17β-diol; and 5-androstene-3, 17β-diol. No significant androgenic activity was induced by dehydroepiandrosterone up to a concentration of 10−6 M. The assay was not responsive to estradiol or progesterone up to 10−9 M and to cortisol up to 10−8 M. No significant androgenic activity was detected for dehydroepiandrosterone sulfate at 10−4 M. The bioassay has demonstrated a linear response to T up to 1.0 nM. The analytical sensitivity of the bioactive androgen assay was 7.8 pM/L. The intra- and interassay coefficients of variation were 7.4% and 7.5% at a T level of 0.25 nM and 4.9% and 6.4% at a T level of 0.03 nM, respectively. This androgen bioassay has been previously validated in premenopausal women in whom androgen concentrations are near the lower limit of sensitivity in many clinical immunoassays (26, 31). Although there are no established reference ranges for pregnant women, we have previously reported levels of circulating bioactive androgens with this bioassay in a small sample of pregnant women (range 0.40 to 2.12 nM) (26). However, differences in maternal levels by trimester or fetal sex have not been previously studied.
Adult offspring outcomes
The adult offspring data were collected as part of a two-part (telephone and in-clinic) interview at the Brigham and Women’s Hospital Center for Clinical Investigation in Boston, Massachusetts, and Brown University in Providence, Rhode Island. Standardized measurements (e.g., anthropometric measurement and blood pressure) were performed, and questionnaires were used to obtain health information (e.g., sociodemographic, medical, and family health histories). Fasting blood samples were also collected.
Anthropometry
Weight and height were measured and used to calculate body mass index (BMI). Waist circumference was measured in a horizontal plane at the level of the iliac crests.
Blood pressure
After resting for 5 minutes, blood pressure was measured five times, 1 minute apart. Multiple measurements were averaged for analysis.
Biomarker assays
Lipid profile
The determination of total cholesterol, triglycerides, and HDL-cholesterol (HDL-C) concentrations was performed on the Hitachi 911 analyzer using reagents and calibrators from Roche Diagnostics (Indianapolis, IN).
Glucose and HbA1c
Glucose was measured on the Hitachi 911 analyzer using Roche Diagnostics reagents. Hemoglobin (Hb)A1c was measured using the Roche P Modular system based on turbidimetric immunoinhibition using hemolyzed whole blood or packed red cells (Roche Diagnostics).
MetS
MetS was defined using modified Adult Treatment Panel III (ATP III) criteria, requiring the presence of more than or equal to three of the following criteria: waist circumference >40 inches in men or >35 inches in women; HDL-C <40 mg/dL in men and <50 mg/dL in women or drug treatment of low HDL; triglycerides ≥150 mg/dL or drug treatment of elevated triglycerides; high blood pressure: systolic blood pressure ≥130 mmHg or diastolic blood pressure ≥85 mmHg or on antihypertensive treatment; and fasting glucose ≥100 mg/dL or on diabetes treatment (32).
Statistical analyses
Demographic characteristics and clinical measures of the analytic sample were assessed and compared by sex using χ2 and t test statistics for categorical and continuous data, respectively. Bioactive androgen levels were categorized into quartiles with the lowest quartile (Q1) defined as the reference level. The associations between the higher quartiles compared with Q1 of maternal prenatal bioactive androgen levels and MetS and each of the five MetS components were estimated as odds ratios and 95% confidence intervals using generalized estimating equations modeling to adjust for intrafamilial correlation. Analytic models were further adjusted for relevant demographic and clinical factors associated with both the outcome (Wald P < 0.2 in final model) and the exposure of interest (accounting for ≥10% change in the exposure estimate when added to the model). Models were run for the overall sample, as well as stratified by sex. Effect modification was examined using interaction terms for maternal bioactive androgens and offspring sex. Final models were adjusted for maternal age, maternal race, prepregnancy weight, and trimester at registration. Tests of our hypotheses were two-sided with an α = 0.05. All analyses were performed using SAS, version 9.4 (SAS Inc., Cary, NC).
Results
Baseline characteristics
The demographic and clinical characteristics of the study population are displayed in Table 1. Mothers of female vs male offspring did not differ significantly in maternal age, race, trimester at registration, socioeconomic status, smoking status, education, or prepregnancy weight. The mothers of male offspring had over twice the mean level of third-trimester bioactive androgen levels than those of female offspring, although the difference did not quite reach statistical significance (males: 4.2 ± 14.0 nM, females: 2.0 ± 1.1 nM, P = 0.09).
Table 1.
Participant Characteristics
| Total Sample (n = 274) |
Male (n = 116) | Female (n = 158) | P a | |
|---|---|---|---|---|
| Maternal characteristics | ||||
| Maternal age, y | 24.7 (5.3) | 24.2 (4.9) | 25.0 (5.6) | 0.20 |
| Trimester at registration | 1.8 (0.6) | 1.7 (0.6) | 1.9 (0.6) | 0.08 |
| White race, n (%) | 254 (92.7) | 252 (92.6) | 148 (93.7) | 0.47 |
| Parental socioeconomic index | 5.9 (1.8) | 5.9 (1.8) | 5.8 (1.8) | 0.52 |
| Maternal smoking, n (%) | 146 (53.7) | 63 (54.8) | 83 (52.9) | 0.75 |
| Maternal education, y | 11.3 (2.0) | 11.3 (2.2) | 11.3 (1.9) | 0.90 |
| Prepregnancy weight, lb | 133.8 (25.2) | 131.8 (25.3) | 135.3 (25.2) | 0.26 |
| Pregnancy factors | ||||
| Offspring weeks gestation, wk | 39.7 (1.4) | 39.6 (1.5) | 39.9 (1.3) | 0.14 |
| Weeks gestation at serum draw, wk | 32.8 (3.2) | 33.2 (3.4) | 32.5 (3.0) | 0.05 |
| Bioactive androgens, nM | 2.9 (9.2) | 4.2 (14.0) | 2.0 (1.1) | 0.09 |
| Fetal growth restriction, n (%) | 112 (40.9) | 51 (44.0) | 61 (38.6) | 0.37 |
| Preeclampsia, n (%) | 106 (39.3) | 45 (39.8) | 61 (38.8) | 0.87 |
| Adult characteristics | ||||
| Adult age, y | 44.7 (2.6) | 44.3 (2.7) | 45.0 (2.5) | 0.04 |
| Current smoking, n (%) | 59 (21.8) | 23 (20.0) | 36 (23.1) | 0.54 |
| Anthropometry | ||||
| BMI, kg/m2 | 29.7 (6.7) | 30.7 (5.8) | 29.0 (7.2) | 0.04 |
| Waist circumference, cm | 96.4 (18.3) | 103.4 (17.2) | 91.2 (17.3) | <0.01 |
| Blood pressure | ||||
| Systolic blood pressure, mmHg | 118 (15) | 126 (12.4) | 112 (14) | <0.01 |
| Diastolic blood pressure, mmHg | 72.8 (9.7) | 77.6 (9.3) | 69.3 (8.4) | <0.01 |
| Serum metabolic profile | ||||
| Total cholesterol, mg/dL | 191.8 (39.2) | 194.8 (39.2) | 189.6 (39.2) | 0.29 |
| LDL-C, mg/dL | 117.6 (35.5) | 123.4 (37.8) | 113.3 (33.3) | 0.02 |
| HDL-C, mg/dL | 51.4 (15.5) | 43.7 (11.1) | 57.0 (15.9) | <0.01 |
| Triglycerides, mg/dL | 119.4 (88.2) | 142.1 (103.8) | 102.8 (70.7) | <0.01 |
| Fasting glucose, mg/dL | 94.1 (29.9) | 101.0 (37) | 89.0 (21.1) | <0.01 |
| HbA1c, % | 5.9 (1.0) | 6.1 (1.3) | 5.8 (0.7) | 0.04 |
| MetS, n (%) | 66 (24.3) | 41 (35.3) | 25 (16.0) | <0.01 |
| Central adiposity | 140 (51.2) | 60 (52.2) | 80 (51.3) | 0.88 |
| High blood pressure | 70 (25.5) | 45 (38.8) | 25 (15.8) | <0.01 |
| Hypertriglyceridemia | 67 (24.4) | 39 (33.6) | 28 (17.7) | <0.01 |
| Low HDL | 107 (39.0) | 50 (43.1) | 57 (36.1) | 0.24 |
| Hyperglycemia | 51 (18.6) | 33 (28.4) | 18 (11.4) | <0.01 |
Values are expressed as means ± SD and/or percentages (%).
Abbreviation: LDL-C, low-density lipoprotein-cholesterol; HbA1c, hemoglobin A1c.
Differences by adult offspring sex compared using χ2 test (categorical variables) or t test (continuous variables). Parental socioeconomic index constructed as a composite index using education, occupation, and income (range: 0 to 9.3).
The mean age of the adult offspring in this study sample was 44.7 ± 2.6 years, with an average BMI of 29.7 ± 6.7 kg/m2. Overall, the participants were predominantly white (92.7%) and female (57.7%), and the prevalence of MetS was 24.3%. Compared with the female offspring, the male offspring had a significantly worse cardiometabolic profile across a wide range of parameters (P < 0.05 across most sex comparisons). The prevalence of MetS in the study population also differed significantly by sex, with 35.3% of the male offspring and 16.0% of the female offspring having the condition, respectively (P < 0.01). The adult male offspring also had a significantly higher prevalence of the individual MetS components of hypertension, hypertriglyceridemia, and hyperglycemia compared with the female offspring (P < 0.01).
As a result of overselection by study design of fetal growth restriction and/or preeclampsia pregnancies in this cohort, 69% of the offspring were from fetal growth-restricted and/or preeclampsia-complicated pregnancies. The rates of fetal growth restriction/preeclampsia did not differ by offspring sex.
Maternal prenatal bioactive androgen levels and MetS
The adjusted odds ratios (aORs) for the association between maternal bioactive androgens and MetS in the adult offspring are displayed in Table 2. In the total study sample, we found that participants with the highest quartile (Q4) of maternal bioactive androgens had significantly greater odds of developing the MetS compared with those in Q1 [aOR: 2.53 (1.07 to 6.02)]. No significant associations were found at the lower quartiles (Q2 or Q3) of maternal bioactive androgens compared with the reference (Q1).
Table 2.
Association Between Prenatal Maternal Bioactive Androgens and MetS
|
|
|
Odds Ratio (95% CI) for MetS (ATP III Criteria) |
|
||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Quartile | Overall | P | n | Males | P | n | Females | P |
| 272 | 116 | 156 | |||||||
| 68 | Q1 | Ref. | – | 29 | Ref. | – | 39 | Ref. | – |
| 68 | Q2 | 1.51 (0.65–3.50) | 0.34 | 29 | 0.58 (0.20–1.70) | 0.32 | 39 | 3.14 (0.79–12.51) | 0.32 |
| 68 | Q3 | 1.53 (0.61–3.80) | 0.36 | 29 | 0.37 (0.11–1.23) | 0.10 | 39 | 3.18 (0.78–12.88) | 0.10 |
| 68 | Q4 | 2.53 (1.07–6.02) | 0.03 | 29 | 1.67 (0.53–5.26) | 0.38 | 39 | 4.06 (1.10–14.93) | 0.03 |
Analyses by generalized estimating equations adjusted for maternal age, maternal race, prepregnancy weight, trimester at registration, and intrafamilial correlation. P values represent comparisons among the second (Q2), third (Q3), and fourth (Q4) quartiles vs Q1 [reference (Ref.)] of bioactive androgens at the 0.05 significance level.
When stratified by sex, the association between Q4 of bioactive androgens and Q1 was found to be significant among women [Q4 vs Q1, aOR: 4.06 (1.10 to 14.93)] but not among men [Q4 vs Q1, aOR: 1.67 (0.53 to 5.26)]. Specifically, the female offspring, whose mothers were in Q4 of maternal bioactive androgen levels, had nearly 4 times the odds of having MetS compared with those in Q1. Although the results were statistically significant only in Q4, the female offspring in all quartiles had greater odds of MetS compared with the male offspring. Additional tests for interaction by sex were nonsignificant given the lack of power.
Maternal prenatal bioactive androgen levels and MetS components
The aORs for the association between maternal bioactive androgens and the individual components of the MetS are displayed in Table 3. In the overall sample, there were no significant associations between Q4 and Q1 of maternal bioactive androgens among the five MetS components. Central adiposity was associated with maternal bioactive androgen levels. However, this finding was limited to the third (Q3)-highest quartile compared with Q1 [Q3 vs Q1, aOR: 2.31 (1.01 to 5.23)].
Table 3.
Association of Prenatal Maternal Bioactive Androgens and MetS Criteria
| Quartile | Odds Ratio (95% CI) for MetS (ATP III Criteria) |
||||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Central Adiposity | Overall | P | n | Males | P | n | Females | P |
| 68 | Q1 | Ref. | – | 29 | Ref. | – | 39 | Ref. | – |
| 68 | Q2 | 1.28 (0.62–2.62) | 0.50 | 29 | 1.48 (0.51–4.28) | 0.46 | 39 | 1.83 (0.69–4.83) | 0.22 |
| 68 | Q3 | 2.31 (1.01–5.23) | 0.04 | 28 | 1.34 (0.38–4.75) | 0.65 | 39 | 1.75 (0.65–4.68) | 0.26 |
| 67 | Q4 | 1.52 (0.68–3.39) | 0.30 | 29 | 1.56 (0.44–5.49) | 0.48 | 39 | 2.00 (0.66–6.07) | 0.22 |
| n | Hypertension | Overall | P | n | Males | P | n | Females | P |
| 68 | Q1 | Ref. | – | 39 | Ref. | – | 39 | Ref. | – |
| 69 | Q2 | 1.19 (0.52–2.73) | 0.67 | 40 | 1.07 (0.35–3.26) | 0.91 | 40 | 1.88 (0.57–6.21) | 0.30 |
| 69 | Q3 | 1.73 (0.71–4.19) | 0.22 | 40 | 0.78 (0.22–2.73) | 0.70 | 40 | 2.50 (0.64–9.76) | 0.19 |
| 68 | Q4 | 2.31 (0.94–5.69) | 0.07 | 39 | 1.37 (0.39–4.86) | 0.62 | 39 | 4.84 (1.12–20.85) | 0.03 |
| n | Low HDL | Overall | P | n | Males | P | n | Females | P |
| 68 | Q1 | Ref. | – | 29 | Ref. | – | 39 | Ref. | – |
| 69 | Q2 | 1.32 (0.63–2.76) | 0.46 | 29 | 0.28 (0.09–0.94) | 0.04 | 40 | 2.94 (1.02–8.49) | 0.04 |
| 69 | Q3 | 1.05 (0.48–2.32) | 0.90 | 29 | 0.25 (0.08–0.76) | 0.01 | 40 | 1.74 (0.61–4.92) | 0.29 |
| 68 | Q4 | 1.06 (0.49–2.27) | 0.88 | 29 | 0.57 (0.25–1.33) | 0.19 | 39 | 1.39 (0.45–4.27) | 0.57 |
| n | High Triglycerides | Overall | P | n | Males | P | n | Females | P |
| 68 | Q1 | Ref. | – | 29 | Ref. | – | 39 | Ref. | – |
| 69 | Q2 | 1.08 (0.50–2.33) | 0.83 | 29 | 0.80 (0.26–2.53) | 0.71 | 40 | 1.32 (0.39–4.49) | 0.66 |
| 64 | Q3 | 1.05 (0.46–2.38) | 0.91 | 29 | 0.48 (0.13–1.81) | 0.28 | 40 | 1.82 (0.53–6.22) | 0.34 |
| 68 | Q4 | 1.78 (0.81–3.93) | 0.15 | 29 | 0.87 (0.29–2.64) | 0.81 | 39 | 2.43 (0.73–8.08) | 0.15 |
| n | High Glucose | Overall | P | n | Males | P | n | Females | P |
| 68 | Q1 | Ref. | – | 29 | Ref. | – | 39 | Ref. | – |
| 69 | Q2 | 1.29 (0.50–3.31) | 0.60 | 29 | 0.71 (0.19–2.68) | 0.62 | 40 | 1.61 (0.33–7.75) | 0.55 |
| 69 | Q3 | 0.78 (0.28–2.19) | 0.64 | 29 | 0.84 (0.22–3.13) | 0.79 | 40 | 1.59 (0.40–6.39) | 0.51 |
| 68 | Q4 | 1.63 (0.63–4.23) | 0.31 | 29 | 0.64 (0.24–1.75) | 0.39 | 39 | 1.64 (0.39–6.86) | 0.50 |
Analyses by generalized estimating equations adjusted for maternal age, maternal race, prepregnancy weight, trimester at registration, and intrafamilial correlation. P values represent comparisons among Q2, Q3, and Q4 vs Q1 [reference (Ref.)] of bioactive androgens at the 0.05 significance level.
When stratified by sex, there was a significant association between the MetS component of hypertension and maternal bioactive androgen levels among the female offspring only. Specifically, the female offspring of mothers in Q4 of maternal bioactive androgen levels had 4.84 times the odds of having hypertension compared with the female offspring of mothers in Q1 [Q4 vs Q1, aOR: 4.84 (1.12 to 20.85)]. Although only significant at Q4, the female offspring in all quartiles had numerically greater odds of MetS compared with male offspring. For the low HDL criteria of MetS, we found that women had higher odds only at the second-highest quartile (Q2) compared with Q1 [Q2 vs Q1, aOR: 2.94 (1.02 to 8.49)]. In men, we found lower odds of low HDL at Q2 and Q3 compared with Q1 [Q2 vs Q1, aOR: 0.28 (0.09 to 0.94); Q3 vs Q1, aOR: 0.25 (0.08 to 0.76)]. Tests for interaction by sex were nonsignificant given inadequate power.
Additional analyses controlling for exposure to fetal growth restriction or preeclampsia in the final models did not significantly alter the results reported above.
Discussion
In this prospective birth cohort study of male and female offspring followed for up to 50 years after birth, we found that higher prenatal levels of bioactive androgens in the mothers were associated with an increased risk of MetS in the adult offspring. Further sex-stratified analyses showed that this association was present in the female offspring but not in the male offspring. Among the female offspring, higher prenatal levels of maternal bioactive androgens were also positively associated with increased risk for the hypertension component of MetS, suggesting a potential sex-dependent link between prenatal androgen exposure and adult CVD risk. To our knowledge, this is the first population-based prospective cohort study to demonstrate sex differences in the association of circulating maternal androgen levels with the development of MetS in later adult life of the offspring. These findings support the hypothesis that the programming of adult CVD risk by androgens may have fetal origins, particularly among the female offspring.
Maternal sex hormones play an important role in supporting and regulating fetal growth and development. Specifically, T levels increase throughout normal pregnancy and reach their peak at birth (33). Our findings raise the possibility that relative androgen excess in the mother could have potential detrimental effects on pregnancy outcomes. Although T levels of fetuses and mothers are believed to be independent of each other, higher levels of maternal T levels have been associated with higher rates of offspring morbidity and mortality (34, 35). Several prenatal insults associated with gestational hyperandrogenic states, such as obesity (36), nicotine exposure (37), preeclampsia (38), and exposure to environmental-disrupting chemicals (39), have been linked to metabolic disturbances in the adult offspring (9). These observations are supported by evidence in animal models demonstrating that prenatal androgen excess can lead to several cardiometabolic impairments in the adult offspring (9, 10, 13, 40). Pregnant women with PCOS have higher T levels during pregnancy, and their adult offspring develop worse metabolic parameters (20, 21) compared with those born to healthy mothers. Consistent with studies in these preclinical models and in women with PCOS, we found that higher circulating maternal androgens in the third trimester were positively associated with MetS risk in the adult offspring in a general human population cohort. Our findings provide longitudinal evidence in a human model that alterations in the gestational androgen milieu during critical developmental periods can have life-long programming effects on the metabolic health of the offspring.
Several studies have reported sex differences in the relation between circulating androgen levels and MetS, with higher androgen levels associated with increased risk in women and lower levels associated with increased risk in men (23). In sex-stratified analyses, we showed that higher prenatal levels of maternal bioactive androgens were positively associated with incident MetS among the female offspring in our cohort, whereas no substantial relationship was observed in the male offspring. Out of the five MetS criteria, we also found that higher circulating maternal androgens specifically associated with increased risk for hypertension among women but not men. This dose-response relationship was not observed for the other MetS criteria. Mechanistic studies in animal models have demonstrated that prenatal T excess leads to hypertension in the female adult offspring (41, 42). Our findings suggest that adverse programming of the vasculature could be a primary mechanism linking maternal androgens to CVD risk that is sex dependent. Unfortunately, the sample size for this study did not allow adequate power for a significant test for interaction with sex. However, the effect size in the association of higher prenatal androgens and adult offspring MetS was substantially higher in the female offspring compared with the male offspring [aOR: 4.06 (1.10 to 14.93) in women vs aOR: 1.67 (0.53 to 5.26) in men], underscoring the validity of the sex-related effect. Furthermore, the higher MetS risk in females was specific for hypertension, which is typically more prevalent in males at early ages (43). Our results suggest that whereas circulating maternal androgen levels may be an important predictor of adult MetS, the programming effects of the gestational androgen milieu on offspring health may depend on whether the mother is carrying a male or female fetus. Thus, our findings are important, given that evidence in human models examining sex differences in the relationship between prenatal androgens and adverse cardiometabolic programming is sparse.
The potential mechanisms of prenatal androgen programming of metabolic dysfunction and CVD in humans are not well understood. Based on evidence from studies conducted in animal and nonhuman primates, some potential androgen-mediated mechanisms of cardiometabolic reprogramming include reduction in insulin-like growth factor bioavailability, resulting in intrauterine growth restriction and postnatal catch-up growth [known prenatal risk factors for CVD (44)], epigenetic modulation of blood pressure regulators, such as nitric oxide and the renin-angiotensin system [leading to hypertension and endothelial dysfunction (41, 45)], downregulation of insulin signaling pathways in peripheral tissues (46), and downregulation of amino acid nutrient delivery to the fetus (47). Thus, it is possible that androgen exposure during the prenatal period may induce life-long epigenetic effects related to alterations in DNA methylation, histone modifications, or microRNA expression that may alter the lifetime risk of CVD in the offspring. Future research is needed to establish the mechanisms and epigenetic markers/pathways involved in the regulation of metabolic and cardiovascular reprogramming by androgens.
Our study has notable strengths and some limitations. We had the ability to take a lifecourse approach to prospectively examine the impact of prenatal androgen levels in utero on adult offspring MetS risk after 50 years of follow-up in a general human population. Fetal programming of CVD risk has not been well studied in humans as a result of lack of longitudinal data that can address these developmental pathways across the lifespan. We used a highly sensitive reporter gene bioassay for the measurement of maternal bioactive androgens. Although bioactive androgen assays do not provide the breakdown of different androgen subtypes, they measure the activity of all known and chemically unidentified environmental androgenic compounds. We recognize that reporter gene bioassays do not distinguish genomic from nongenomic action of androgens or downstream effects of androgen receptor activation. Although this would contribute to the understanding of the sources of bioactive androgen production, it does not diminish from our overall findings that higher levels of bioactive androgens can affect long-term offspring metabolic outcomes. Further information on the diagnosis of PCOS in the CPP women, although not available in the CPP, would also contribute to the understanding of potential sources of the higher level of maternal androgen exposure.
The long-term stability of sex hormones in maternal prenatal serum from the CPP after over 40 years of storage has been validated (48). As a result of unavailability of cord serum and amniotic fluid, we used maternal serum as an indirect biomarker of fetal androgen exposure. Studies have reported associations between hormone concentrations in the maternal and fetal circulations (49), as well as evidence that neonates tend to have more health-related problems when their mothers have elevated T levels (35). Given our analyses were limited to third-trimester maternal androgen levels, our study did not address the timing of androgen exposure as a factor in the development of offspring MetS. However, we demonstrated, as hypothesized, that this particular timing of exposure, i.e., a critical period of the organizational effects of gonadal steroids on development, had long-term sex-dependent effects in the offspring. Future human studies are needed to determine the effects of timing of prenatal androgen exposure throughout all trimesters of pregnancy on MetS in the adult offspring.
We recognize that there are several different definitions of MetS, each with its own assets and liabilities. We selected the ATP III definition of MetS, which has been widely used in the United States. In addition, we examined sex differences in the fetal androgen programming of CVD risk, a topic that has not been sufficiently addressed in previous studies. Although our important findings were primarily limited to women, the sample size of men in our cohort was relatively smaller compared with women, potentially restricting the power to detect small but significant effects in men. We acknowledge that the age range (39 to 50 years) in the adult offspring may not be representative of older populations who are at higher CVD risk. Despite our younger adult population, we still found an important relationship between circulating maternal androgens and risk for MetS, particularly among women. The female effect was striking given that the component of MetS primarily affected was hypertension, a domain usually reserved for men in early midlife more than women. These findings suggest that younger premenopausal women born to mothers with higher circulating prenatal androgens may be at risk for CVD and target a key period for early screening and therapeutic intervention.
In summary, higher prenatal levels of maternal androgens were associated with increased risk for development of the MetS in adulthood, an effect that was observed in female offspring and not in males, and specifically, impacting hypertension. Thus, maternal androgen levels could potentially serve as an early biomarker for predicting long-term cardiometabolic health in the offspring. Future research is needed to elucidate the mechanisms that mediate the sex-dependent developmental programming of CVD risk by androgens. These mechanistic insights may help identify therapeutic targets for intervention and prevention strategies.
Acknowledgments
We thank Anne Remington for help with project management. We also thank Nancy Gee for performing the bioactive androgen assays for this study.
Financial Support: Support for this work was provided by National Heart, Lung, and Blood Institute (NHLBI) Grant K08 HL132122-02 (to G.H.); State of Arizona Arizona Biomedical Research Commission (ABRC) Grant ADHS14-00003606 (to R.J.H. and J.M.G.); National Institute of Mental Health–NHLBI Grant R01MH074679 (to J.M.G.); and National Institute on Aging Grants RC2AG036666 (to E.B.L. and S.L.B.) and R01AG023397 (to S.L.B.).
Disclosure Summary: S.B. has received research grant support from AbbVie Pharmaceuticals, Transition Therapeutics, and Metro International Biotechnology, LLC, for investigator-initiated research unrelated to this study. S.B. has served as a consultant to AbbVie and Novartis. S.B. has a financial interest in Function Promoting Therapies, LLC, a company aiming to develop innovative solutions that enhance precision and accuracy in clinical decision making and facilitate personalized therapeutic choices in reproductive health. S.B.’s interests were reviewed and are managed by Brigham and Women’s Hospital and Partners HealthCare in accordance with their conflict of interest policies. The remaining authors have nothing to disclose.
Glossary
Abbreviations:
- aOR
adjusted odds ratio
- ATP III
Adult Treatment Panel III
- BMI
body mass index
- CPP
Collaborative Perinatal Project
- CVD
cardiovascular disease
- Hb
hemoglobin
- HDL-C
high-density lipoprotein-cholesterol
- MetS
metabolic syndrome
- NEFS
New England Family Study
- PCOS
polycystic ovary syndrome
- Q1–Q4
lowest to highest quartile
- T
testosterone
References
- 1. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Després JP, Fullerton HJ, Howard VJ, Huffman MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Matchar DB, McGuire DK, Mohler ER III, Moy CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW, Turner MB; American Heart Association Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics—2015 update: a report from the American Heart Association [published corrections appear in Circulation 2015;131(24):e535 and 2016;133(8):e417]. Circulation. 2015;131(4):e29–e322. [DOI] [PubMed] [Google Scholar]
- 2. Lerner DJ, Kannel WB. Patterns of coronary heart disease morbidity and mortality in the sexes: a 26-year follow-up of the Framingham population. Am Heart J. 1986;111(2):383–390. [DOI] [PubMed] [Google Scholar]
- 3. Humphries KH, Izadnegahdar M, Sedlak T, Saw J, Johnston N, Schenck-Gustafsson K, Shah RU, Regitz-Zagrosek V, Grewal J, Vaccarino V, Wei J, Bairey Merz CN. Sex differences in cardiovascular disease - Impact on care and outcomes. Front Neuroendocrinol. 2017;46:46–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ford ES. Risks for all-cause mortality, cardiovascular disease, and diabetes associated with the metabolic syndrome: a summary of the evidence. Diabetes Care. 2005;28(7):1769–1778. [DOI] [PubMed] [Google Scholar]
- 5. Rinaudo P, Wang E. Fetal programming and metabolic syndrome. Annu Rev Physiol. 2012;74(1):107–130. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Osmond C, Barker DJ. Ischaemic heart disease in England and Wales around the year 2000. J Epidemiol Community Health. 1991;45(1):71–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Barker DJ. The developmental origins of chronic adult disease. Acta Paediatr Suppl. 2004;93(446):26–33. [DOI] [PubMed] [Google Scholar]
- 8. Recabarren SE, Sir-Petermann T, Maliqueo M, Lobos A, Rojas-García P. [Prenatal exposure to androgens as a factor of fetal programming]. Rev Med Chil. 2006;134(1):101–108. [DOI] [PubMed] [Google Scholar]
- 9. Hakim C, Padmanabhan V, Vyas AK. Gestational hyperandrogenism in developmental programming. Endocrinology. 2017;158(2):199–212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Xita N, Tsatsoulis A. Review: Fetal programming of polycystic ovary syndrome by androgen excess: evidence from experimental, clinical, and genetic association studies. J Clin Endocrinol Metab. 2006;91(5):1660–1666. [DOI] [PubMed] [Google Scholar]
- 11. Sir-Petermann T, Maliqueo M, Angel B, Lara HE, Pérez-Bravo F, Recabarren SE. Maternal serum androgens in pregnant women with polycystic ovarian syndrome: possible implications in prenatal androgenization. Hum Reprod. 2002;17(10):2573–2579. [DOI] [PubMed] [Google Scholar]
- 12. Dumesic DA, Goodarzi MO, Chazenbalk GD, Abbott DH. Intrauterine environment and polycystic ovary syndrome. Semin Reprod Med. 2014;32(3):159–165. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Padmanabhan V, Manikkam M, Recabarren S, Foster D. Prenatal testosterone excess programs reproductive and metabolic dysfunction in the female. Mol Cell Endocrinol. 2006;246(1-2):165–174. [DOI] [PubMed] [Google Scholar]
- 14. Bruns CM, Baum ST, Colman RJ, Eisner JR, Kemnitz JW, Weindruch R, Abbott DH. Insulin resistance and impaired insulin secretion in prenatally androgenized male rhesus monkeys. J Clin Endocrinol Metab. 2004;89(12):6218–6223. [DOI] [PubMed] [Google Scholar]
- 15. Lazic M, Aird F, Levine JE, Dunaif A. Prenatal androgen treatment alters body composition and glucose homeostasis in male rats. J Endocrinol. 2011;208(3):293–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Homburg R, Gudi A, Shah A, M Layton A. A novel method to demonstrate that pregnant women with polycystic ovary syndrome hyper-expose their fetus to androgens as a possible stepping stone for the developmental theory of PCOS. A pilot study. Reprod Biol Endocrinol. 2017;15(1):61. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Caanen MR, Kuijper EA, Hompes PG, Kushnir MM, Rockwood AL, Meikle WA, Homburg R, Lambalk CB. Mass spectrometry methods measured androgen and estrogen concentrations during pregnancy and in newborns of mothers with polycystic ovary syndrome. Eur J Endocrinol. 2016;174(1):25–32. [DOI] [PubMed] [Google Scholar]
- 18. Hickey M, Sloboda DM, Atkinson HC, Doherty DA, Franks S, Norman RJ, Newnham JP, Hart R. The relationship between maternal and umbilical cord androgen levels and polycystic ovary syndrome in adolescence: a prospective cohort study. J Clin Endocrinol Metab. 2009;94(10):3714–3720. [DOI] [PubMed] [Google Scholar]
- 19. Lathi RB, Dahan MH, Reynolds-May MF, Milki AA, Behr B, Westphal LM. The role of serum testosterone in early pregnancy outcome: a comparison in women with and without polycystic ovary syndrome. J Obstet Gynaecol Can. 2014;36(9):811–816. [DOI] [PubMed] [Google Scholar]
- 20. Recabarren SE, Smith R, Rios R, Maliqueo M, Echiburú B, Codner E, Cassorla F, Rojas P, Sir-Petermann T. Metabolic profile in sons of women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2008;93(5):1820–1826. [DOI] [PubMed] [Google Scholar]
- 21. Yildiz BO, Yarali H, Oguz H, Bayraktar M. Glucose intolerance, insulin resistance, and hyperandrogenemia in first degree relatives of women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2003;88(5):2031–2036. [DOI] [PubMed] [Google Scholar]
- 22. Stavreva DA, George AA, Klausmeyer P, Varticovski L, Sack D, Voss TC, Schiltz RL, Blazer VS, Iwanowicz LR, Hager GL. Prevalent glucocorticoid and androgen activity in US water sources. Sci Rep. 2012;2(1):937. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Brand JS, van der Tweel I, Grobbee DE, Emmelot-Vonk MH, van der Schouw YT. Testosterone, sex hormone-binding globulin and the metabolic syndrome: a systematic review and meta-analysis of observational studies. Int J Epidemiol. 2011;40(1):189–207. [DOI] [PubMed] [Google Scholar]
- 24. Maturana MA, Breda V, Lhullier F, Spritzer PM. Relationship between endogenous testosterone and cardiovascular risk in early postmenopausal women. Metabolism. 2008;57(7):961–965. [DOI] [PubMed] [Google Scholar]
- 25. Liu PY, Death AK, Handelsman DJ. Androgens and cardiovascular disease. Endocr Rev. 2003;24(3):313–340. [DOI] [PubMed] [Google Scholar]
- 26. Chen J, Sowers MR, Moran FM, McConnell DS, Gee NA, Greendale GA, Whitehead C, Kasim-Karakas SE, Lasley BL. Circulating bioactive androgens in midlife women. J Clin Endocrinol Metab. 2006;91(11):4387–4394. [DOI] [PubMed] [Google Scholar]
- 27. Susser E, Buka S, Schaefer CA, Andrews H, Cirillo PM, Factor-Litvak P, Gillman M, Goldstein JM, Henry PI, Lumey LH, McKeague IW, Michels KB, Terry MB, Cohn BA; for the EDAH Team . The Early Determinants of Adult Health study. J Dev Orig Health Dis. 2011;2(6):311–321. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Goldstein JM, Cherkerzian S, Buka SL, Fitzmaurice G, Hornig M, Gillman M, O’Toole S, Sloan RP. Sex-specific impact of maternal-fetal risk factors on depression and cardiovascular risk 40 years later. J Dev Orig Health Dis. 2011;2(6):353–364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Goldstein JM, Cherkerzian S, Seidman LJ, Donatelli J-A, Remington AG, Tsuang MT, Hornig M, Buka SL. Prenatal maternal immune disruption and sex-dependent risk for psychoses. Psychol Med. 2014;44(15):3249–3261. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Ellman LM, Schetter CD, Hobel CJ, Chicz-Demet A, Glynn LM, Sandman CA. Timing of fetal exposure to stress hormones: effects on newborn physical and neuromuscular maturation. Dev Psychobiol. 2008;50(3):232–241. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Raivio T, Palvimo JJ, Dunkel L, Wickman S, Jänne OA. Novel assay for determination of androgen bioactivity in human serum. J Clin Endocrinol Metab. 2001;86(4):1539–1544. [DOI] [PubMed] [Google Scholar]
- 32. Grundy SM, Cleeman JI, Daniels SR, Donato KA, Eckel RH, Franklin BA, Gordon DJ, Krauss RM, Savage PJ, Smith SC Jr, Spertus JA, Costa F; American Heart Association; National Heart, Lung, and Blood Institue . Diagnosis and management of the metabolic syndrome. An American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Executive summary. Cardiol Rev. 2005;13(6):322–327. [PubMed] [Google Scholar]
- 33. Makieva S, Saunders PT, Norman JE. Androgens in pregnancy: roles in parturition. Hum Reprod Update. 2014;20(4):542–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Grant VJ, Irwin RJ. A simple model for adaptive variation in the sex ratios of mammalian offspring. J Theor Biol. 2009;258(1):38–42. [DOI] [PubMed] [Google Scholar]
- 35. Furui T, Imai A, Ohno T. Pre-term labour in cases with high maternal testosterone levels. J Obstet Gynaecol. 2007;27(2):155–156. [DOI] [PubMed] [Google Scholar]
- 36. Li J, Daly E, Campioli E, Wabitsch M, Papadopoulos V. De novo synthesis of steroids and oxysterols in adipocytes. J Biol Chem. 2014;289(2):747–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Kandel DB, Udry JR. Prenatal effects of maternal smoking on daughters’ smoking: nicotine or testosterone exposure? Am J Public Health. 1999;89(9):1377–1383. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Salamalekis E, Bakas P, Vitoratos N, Eleptheriadis M, Creatsas G. Androgen levels in the third trimester of pregnancy in patients with preeclampsia. Eur J Obstet Gynecol Reprod Biol. 2006;126(1):16–19. [DOI] [PubMed] [Google Scholar]
- 39. Palioura E, Diamanti-Kandarakis E. Polycystic ovary syndrome (PCOS) and endocrine disrupting chemicals (EDCs). Rev Endocr Metab Disord. 2015;16(4):365–371. [DOI] [PubMed] [Google Scholar]
- 40. Demissie M, Lazic M, Foecking EM, Aird F, Dunaif A, Levine JE. Transient prenatal androgen exposure produces metabolic syndrome in adult female rats. Am J Physiol Endocrinol Metab. 2008;295(2):E262–E268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Sathishkumar K, Elkins R, Yallampalli U, Balakrishnan M, Yallampalli C. Fetal programming of adult hypertension in female rat offspring exposed to androgens in utero. Early Hum Dev. 2011;87(6):407–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42. King AJ, Olivier NB, Mohankumar PS, Lee JS, Padmanabhan V, Fink GD. Hypertension caused by prenatal testosterone excess in female sheep. Am J Physiol Endocrinol Metab. 2007;292(6):E1837–E1841. [DOI] [PubMed] [Google Scholar]
- 43. Colafella KMM, Denton KM. Sex-specific differences in hypertension and associated cardiovascular disease. Nat Rev Nephrol. 2018;14(3):185–201. [DOI] [PubMed] [Google Scholar]
- 44. Manikkam M, Crespi EJ, Doop DD, Herkimer C, Lee JS, Yu S, Brown MB, Foster DL, Padmanabhan V. Fetal programming: prenatal testosterone excess leads to fetal growth retardation and postnatal catch-up growth in sheep. Endocrinology. 2004;145(2):790–798. [DOI] [PubMed] [Google Scholar]
- 45. Grigore D, Ojeda NB, Alexander BT. Sex differences in the fetal programming of hypertension. Gen Med. 2008;5(Suppl A):S121–S132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Nada SE, Thompson RC, Padmanabhan V. Developmental programming: differential effects of prenatal testosterone excess on insulin target tissues. Endocrinology. 2010;151(11):5165–5173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. Sathishkumar K, Elkins R, Chinnathambi V, Gao H, Hankins GD, Yallampalli C. Prenatal testosterone-induced fetal growth restriction is associated with down-regulation of rat placental amino acid transport. Reprod Biol Endocrinol. 2011;9(1):110. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48. Stroud LR, Solomon C, Shenassa E, Papandonatos G, Niaura R, Lipsitt LP, Lewinn K, Buka SL. Long-term stability of maternal prenatal steroid hormones from the National Collaborative Perinatal Project: still valid after all these years. Psychoneuroendocrinology. 2007;32(2):140–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49. Troisi R, Potischman N, Roberts JM, Harger G, Markovic N, Cole B, Lykins D, Siiteri P, Hoover RN. Correlation of serum hormone concentrations in maternal and umbilical cord samples. Cancer Epidemiol Biomarkers Prev. 2003;12(5):452–456. [PubMed] [Google Scholar]