Portal de Periódicos da CAPES

Circulating maternal testosterone concentrations at 18 weeks of gestation predict circulating levels of antimüllerian hormone in adolescence: a prospective cohort study

2010; Elsevier BV; Volume: 94; Issue: 4 Linguagem: Inglês

10.1016/j.fertnstert.2009.12.060

1556-5653

Roger Hart, Deborah M. Sloboda, Dorota A. Doherty, Robert J. Norman, Helen C. Atkinson, John P. Newnham, Jan E. Dickinson, Martha Hickey,

Sexual Differentiation and Disorders

This prospective study was established to determine the impact of maternal circulating androgen levels during normal pregnancy on ovarian function, as determined by early follicular phase antimüllerian hormone (AMH) levels, inhibin B levels, and antral follicle count (AFC) in 244 female offspring in adolescence. Maternal circulating total testosterone levels at 18 weeks' gestation were statistically significantly correlated with early follicular-phase circulating AMH levels in female adolescent offspring, but no other statistically significant correlations were determined among the maternal androgens at 18 or 34 weeks of gestation and the markers of adolescent ovarian function. This prospective study was established to determine the impact of maternal circulating androgen levels during normal pregnancy on ovarian function, as determined by early follicular phase antimüllerian hormone (AMH) levels, inhibin B levels, and antral follicle count (AFC) in 244 female offspring in adolescence. Maternal circulating total testosterone levels at 18 weeks' gestation were statistically significantly correlated with early follicular-phase circulating AMH levels in female adolescent offspring, but no other statistically significant correlations were determined among the maternal androgens at 18 or 34 weeks of gestation and the markers of adolescent ovarian function. There is evidence from both human and animal studies that supraphysiologic maternal androgen levels may lead to disordered folliculogenesis in female offspring with a polycystic ovary syndrome (PCOS) phenotype (1Robinson J.E. Birch R.A. Foster D.L. Padmanabhan V. Prenatal exposure of the ovine fetus to androgens sexually differentiates the steroid feedback mechanisms that control gonadotropin releasing hormone secretion and disrupts ovarian cycles.Arch Sex Behav. 2002; 31: 35-41Crossref PubMed Scopus (37) Google Scholar, 2Forsdike R.A. Hardy K. Bull L. Stark J. Webber L.J. Stubbs S. et al.Disordered follicle development in ovaries of prenatally androgenized ewes.J Endocrinol. 2007; 192: 421-428Crossref PubMed Scopus (71) Google Scholar, 3Abbott D.H. Barnett D.K. Bruns C.M. Dumesic D.A. Androgen excess fetal programming of female reproduction: a developmental aetiology for polycystic ovary syndrome?.Hum Reprod Update. 2005; 11: 357-374Crossref PubMed Scopus (394) Google Scholar). However, the impact of variations in maternal androgens within the normal physiologic range is less well understood. We have recently reported that higher maternal testosterone levels within the normal range do not predict PCOS in adolescence (4Hickey M. Sloboda D.M. Atkinson H.C. Doherty D.A. Franks S. Norman R.J. et al.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: 3714-3720Crossref PubMed Scopus (111) Google Scholar). During normal pregnancy, the fetus is protected from maternal androgens by placental aromatase. It is possible, however, that placental dysfunction may expose the fetus to higher concentrations of androgens, although it has yet to be proven empirically. The aim of our study was to determine the impact of variation in maternal circulating androgens within the normal physiologic range on ovarian function in adolescent offspring by use of a large, unselected, prospective birth cohort. To evaluate ovarian function during adolescence, we measured antimüllerian hormone (AMH) and inhibin B levels and antral follicle count (AFC) in the early follicular phase of the menstrual cycle. The established West Australian Pregnancy Cohort (Raine) study (http://www.rainestudy.org.au) was designed to measure the relationships between early life events and subsequent health and behavior. Participants were randomized to a control or an intensive investigation arm where maternal blood samples were collected at 18 and 34/36 weeks. Umbilical cord blood (mixed arterial and venous) was collected at 870 singleton deliveries. The current cohort includes 1800 adolescents aged 16 to 18 years. This study was approved by the Raine Executive Committee and the ethics committee of King Edward Memorial Hospital. As previously reported elsewhere, adolescent girls within the cohort were invited to participate in the study to assess early life influences upon reproductive function (4Hickey M. Sloboda D.M. Atkinson H.C. Doherty D.A. Franks S. Norman R.J. et al.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: 3714-3720Crossref PubMed Scopus (111) Google Scholar, 5Hart R. Sloboda D.M. Doherty D.A. Norman R.J. Atkinson H.C. Newnham J.P. et al.Prenatal determinants of uterine volume and ovarian reserve in adolescence.J Clin Endocrinol Metab. 2009; 94: 4931-4937Crossref PubMed Scopus (43) Google Scholar). Plasma was collected on the day of the ultrasound visit on days 2 to 5 of their menstrual cycle. Samples were stored at –80°C and have not been thawed since collection. All ultrasound imaging studies were performed via the transabdominal route by one of two experienced gynecologic ultrasonographers. On imaging, follicles between 2 and 9 mm in diameter were counted in both ovaries to determine the AFC. All images were reviewed by one expert (J.D.). If any follicle >10 mm was seen, the study was repeated in the early follicular phase of the next cycle. Levels of sex-hormone–binding globulin (SHBG) (68562; Orion Diagnostica, Espoo, Finland), dehydroepiandrosterone-sulfate (DHEAS, DSL-2700; Beckman Coulter, Gladesville, New South Wales, Australia), androstenedione (A4) (DSL-4200; Beckman Coulter), and total testosterone (TT) (DSL-4100; Beckman Coulter) were determined in maternal, umbilical, and adolescent samples by commercially available immunoassays. These results have previously been published by our group elsewhere (4Hickey M. Sloboda D.M. Atkinson H.C. Doherty D.A. Franks S. Norman R.J. et al.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: 3714-3720Crossref PubMed Scopus (111) Google Scholar, 5Hart R. Sloboda D.M. Doherty D.A. Norman R.J. Atkinson H.C. Newnham J.P. et al.Prenatal determinants of uterine volume and ovarian reserve in adolescence.J Clin Endocrinol Metab. 2009; 94: 4931-4937Crossref PubMed Scopus (43) Google Scholar). The lower limit of sensitivity of the total testosterone assay was 347 pmol/L, and the interassay and interpatient coefficients of variation (CV) were 6% and 15%, respectively, at 1 nmol/L concentration. Free testosterone concentrations were calculated (cFT) from the measured TT and SHBG, as we previously reported elsewhere (4Hickey M. Sloboda D.M. Atkinson H.C. Doherty D.A. Franks S. Norman R.J. et al.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: 3714-3720Crossref PubMed Scopus (111) Google Scholar). The free androgen index (FAI) is the TT in nM/SHBG in nM × 100. Concentrations of AMH in adolescent samples were determined by enzyme-linked immunosorbent assay (ELISA, A16507; Immunotech/Beckman Coulter). Plasma pools with means of 12 and 77 pmol/L were used to compare the assay characteristics among the seven assays used for this study. The interassay coefficient of variation was <5%, and the intra-assay coefficient of variation was <9%. The assay sensitivity was 1 pmol/L. Inhibin B concentrations in adolescent samples were measured by ELISA (Active DSL-10-84100; Beckman Coulter). Inhibin B sensitivity was 7 pg/mL, and the interassay and intra-assay coefficients of variation were 9.2% and 7.0%, respectively. Continuous data were summarized using mean and standard deviation or median and interquartile range, according to data normality. Categorical data were summarized using frequency distributions. Linear regression was used to examine associations between prenatal androgens and AMH, inhibin B, and AFC; and bivariate correlations and correlations adjusted for pubertal timing (r), and partial correlations (partial-r) obtained after removing the effect of pubertal timing on the relationship between circulating androgens and AMH, inhibin B, and AFC were reported to summarize what proportion of the variability was explained by the statistically significant predictors. Regression analyses were adjusted for time since menarche and current age, as appropriate, to account for the differences in pubertal timing. Log and square root data transformations were performed to achieve data normality. All hypothesis tests were two-sided, and P<.05 was considered statistically significant. SPSS statistical software (version 15.0; SPSS Inc., Chicago, IL) was used for data analysis. A total of 244 girls (median age 15.1 years, interquartile range [IQR] 14.5–17.6 years) consented to participate in the study and were able to attend the clinic on days 2 to 5 of their menstrual cycle (for a flowchart of girls recruited see [5Hart R. Sloboda D.M. Doherty D.A. Norman R.J. Atkinson H.C. Newnham J.P. et al.Prenatal determinants of uterine volume and ovarian reserve in adolescence.J Clin Endocrinol Metab. 2009; 94: 4931-4937Crossref PubMed Scopus (43) Google Scholar]). Twelve girls were currently using the combined oral contraceptive pill. The great majority (221 girls, 90.6%) were Caucasian. Of the girls who attended for assessment, maternal serum was available from 122, of whom 109 girls had AMH concentrations available (44.7%) at 18 weeks and 107 girls (43.9%) at 34 weeks of gestation. Cord blood was available for 77 (31.6%) of the girls. Participant characteristics including time since menarche and antenatal androgen concentrations at 18 and 34 weeks in maternal and in cord blood have previously been reported elsewhere (4Hickey M. Sloboda D.M. Atkinson H.C. Doherty D.A. Franks S. Norman R.J. et al.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: 3714-3720Crossref PubMed Scopus (111) Google Scholar, 5Hart R. Sloboda D.M. Doherty D.A. Norman R.J. Atkinson H.C. Newnham J.P. et al.Prenatal determinants of uterine volume and ovarian reserve in adolescence.J Clin Endocrinol Metab. 2009; 94: 4931-4937Crossref PubMed Scopus (43) Google Scholar, 6Sloboda D.M. Hart R. Doherty D.A. Pennell C.E. Hickey M. Age at menarche: influences of prenatal and postnatal growth.J Clin Endocrinol Metab. 2007; 92: 46-50Crossref PubMed Scopus (209) Google Scholar). The median (IQR) of plasma inhibin B levels, AMH levels, and AFC were 11 (7Dunn J.F. Nisula B.C. Rodbard D. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma.J Clin Endocrinol Metab. 1981; 53: 58-68Crossref PubMed Scopus (995) Google Scholar, 8Stanczyk F.Z. Cho M.M. Endres D.B. Morrison J.L. Patel S. Paulson R.J. Limitations of direct estradiol and testosterone immunoassay kits.Steroids. 2003; 68: 1173-1178Crossref PubMed Scopus (152) Google Scholar, 9Sir-Petermann T. Maliqueo M. Angel B. Lara H.E. Perez-Bravo F. Recabarren S.E. Maternal serum androgens in pregnant women with polycystic ovarian syndrome: possible implications in prenatal androgenization.Hum Reprod. 2002; 17: 2573-2579Crossref PubMed Scopus (297) Google Scholar, 10Sir-Petermann T. Codner E. Maliqueo M. Echiburu B. Hitschfeld C. Crisosto N. et al.Increased anti-Müllerian hormone serum concentrations in prepubertal daughters of women with polycystic ovary syndrome.J Clin Endocrinol Metab. 2006; 91: 3105-3109Crossref PubMed Scopus (111) Google Scholar, 11Zawadzki J. Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach.in: Polycystic ovarian syndrome Dunaif A. Givens J. Haseltine F. Merriam G. Blackwell, Boston1992: 377-384Google Scholar, 12ESHRE/ASRM Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome.Fertil Steril. 2004; 81: 19-25Abstract Full Text Full Text PDF Scopus (4508) Google Scholar, 13Hart R. Doherty D.A. Norman R.J. Franks S. Dickinson J.E. Hickey M. Sloboda D.M. Serum antimullerian hormone (AMH) levels are elevated in adolescent girls with polycystic ovaries and the polycystic ovarian syndrome (PCOS).Fertil Steril. 2010 Jan 6; ([Epub ahead of print.])Abstract Full Text Full Text PDF Scopus (90) Google Scholar, 14Bammann B.L. Coulam C.B. Jiang N.S. Total and free testosterone during pregnancy.Am J Obstet Gynecol. 1980; 137: 293-298Abstract Full Text PDF PubMed Scopus (96) Google Scholar), 25 (18–36) pmol/L, and 50 (34–66) pg/mL, respectively. Correlations with circulating androgen levels in pregnancy are shown in Table 1.Table 1Correlations between AMH, Inhibin B, antral follicle count and circulating androgen concentrations during pregnancy.AMHInhibin BAFCrA-rP-rrA-rP-rrA-rP-r18 pregnancy wk A4 (nmol/L)0.1100.3360.1670.0820.0960.089–0.0330.116–0.018 DHEAS (μmol/L)0.0830.3090.0930.0760.0840.077–0.1110.159–0.111 SHBG (nmol/L)0.0950.3450.134–0.0980.103–0.095–0.1320.177–0.137 TT (pmol/L)0.193∗Statistically significant associations indicated by P < 0.05.0.351∗Statistically significant associations indicated by P < 0.05.0.209∗Statistically significant associations indicated by P < 0.05.0.1070.1140.108–0.0060.115–0.004 cFT (pmol/L)0.0920.3090.0940.0610.0700.060–0.0370.121–0.039 FAI0.0850.3070.0890.0550.0650.055–0.0370.121–0.038 N10510511334/36 pregnancy wk A4 (nmol/L)–0.0300.321–0.003–0.0310.049–0.027–0.0420.119–0.033 DHEAS (μmol/L)0.0810.3340.0980.0240.0490.026–0.0930.143–0.088 SHBG (nmol/L)0.0950.3450.134–0.0980.103–0.0950.1320.1770.137 TT (pmol/L)–0.0030.312–0.016–0.0100.043–0.0120.1260.1670.123 cFT (pmol/L)–0.0510.330–0.0820.0500.0620.0470.0340.1180.030 FAI–0.0460.330–0.0800.0420.0560.0390.0310.1170.027 N103103109Cord blood A4 (nmol/L)–0.0540.270–0.0290.0020.0600.008–0.0990.286–0.082 DHEAS (μmol/L)0.0050.2680.002–0.0470.076–0.047–0.1110.280–0.061 SHBG (nmol/L)0.0410.2710.043–0.0730.095–0.074–0.0740.287–0.087 TT (nmol/L)–0.0150.268–0.0090.0410.0760.047–0.0650.278–0.046 cFT (pmol/L)–0.0440.269–0.0180.0740.1000.080–0.0320.274–0.005 FAI–0.0500.269–0.0280.1020.1230.1070.0320.2740.029 N747479Note: Univariate (r), correlations adjusted for pubertal timing (A-r) and partial correlations (P-r) after removing the effect of pubertal timing are shown. Except for Inhibin B, all other analyte concentrations were either log-transformed or square-root transformed to achieve data normality. A4, androstenedione; AFC, antral follicle count; AMH, antimüllerian hormone; cFT, free testosterone concentration; DHEAS, dehydroepiandrosterone-sulfate; FAI, free androgen index; SHBG, sex-hormone–binding globulin; TT, total testosterone.∗ Statistically significant associations indicated by P < 0.05. Open table in a new tab Note: Univariate (r), correlations adjusted for pubertal timing (A-r) and partial correlations (P-r) after removing the effect of pubertal timing are shown. Except for Inhibin B, all other analyte concentrations were either log-transformed or square-root transformed to achieve data normality. A4, androstenedione; AFC, antral follicle count; AMH, antimüllerian hormone; cFT, free testosterone concentration; DHEAS, dehydroepiandrosterone-sulfate; FAI, free androgen index; SHBG, sex-hormone–binding globulin; TT, total testosterone. Statistically significant positive correlations were found between AMH and inhibin B levels and AFC (r = 0.293, P<.001, between AMH and inhibin B; r = 0.396, P<.001, between AMH and AFC; and r = 0.230, P=.001 between inhibin B and AFC). We observed a statistically significant positive association between maternal serum TT concentrations at 18 weeks of gestation and AMH concentrations in adolescence (r = 0.193, P=.044) (see Table 1) and after adjustment for time since menarche (r = 0.345, partial-r = 0.209, P=.030), with TT explaining 4% variability of the AMH concentrations in the adjusted model. When girls currently taking oral contraceptive pills were excluded from the analysis, the relationship between maternal TT and AMH concentrations in adolescence remained statistically significant with the adjustment for time since menarche (r = 0.351, partial-r = 0.198, P=.044). No other statistically significant associations between maternal and cord androgens and AMH in adolescence were found. No associations between inhibin B in adolescence or AFC and prenatal androgens were found in univariate regression analyses and with adjustments for pubertal timing (see Table 1). On grouping adolescent AMH levels in tertiles, the positive associations were with TT at 18 weeks, and DHEAS at 18 and 34 weeks of gestation (P=.037, P=.014, and P=.037, respectively). This is the first prospective study to examine the relationship between circulating maternal androgen concentrations in normal pregnancy and markers of ovarian function in adolescence. Our study showed a statistically significant relationship between maternal total testosterone concentrations in early in pregnancy and circulating AMH concentrations during adolescence. This is the first evidence to suggest that during an otherwise unremarkable pregnancy, maternal androgens may influence ovarian function in offspring. During pregnancy, the majority of circulating testosterone is protein bound. It is thought that only the free fraction is biologically active (7Dunn J.F. Nisula B.C. Rodbard D. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma.J Clin Endocrinol Metab. 1981; 53: 58-68Crossref PubMed Scopus (995) Google Scholar). However, current direct immunoassays for measuring free testosterone have significant limitations (8Stanczyk F.Z. Cho M.M. Endres D.B. Morrison J.L. Patel S. Paulson R.J. Limitations of direct estradiol and testosterone immunoassay kits.Steroids. 2003; 68: 1173-1178Crossref PubMed Scopus (152) Google Scholar). Further, since both albumin and SHBG levels are markedly increased in pregnancy, it is not clear whether standardized methods for calculating free testosterone are valid during pregnancy. During normal pregnancy, the fetus is protected from maternal androgens by placental aromatase rapidly converting androgens to estrogens. It is believed that pregnant women with PCOS have elevated androgen levels compared with pregnant women without PCOS, and this has been postulated as a potential cause of prenatal androgenization (9Sir-Petermann T. Maliqueo M. Angel B. Lara H.E. Perez-Bravo F. Recabarren S.E. Maternal serum androgens in pregnant women with polycystic ovarian syndrome: possible implications in prenatal androgenization.Hum Reprod. 2002; 17: 2573-2579Crossref PubMed Scopus (297) Google Scholar), which may perturb ovarian function in their daughters. Indeed, the prepubertal daughters of women with PCOS have been reported to have elevated serum concentrations of AMH (10Sir-Petermann T. Codner E. Maliqueo M. Echiburu B. Hitschfeld C. Crisosto N. et al.Increased anti-Müllerian hormone serum concentrations in prepubertal daughters of women with polycystic ovary syndrome.J Clin Endocrinol Metab. 2006; 91: 3105-3109Crossref PubMed Scopus (111) Google Scholar), testosterone, 17-hydroxyprogesterone, and DHEAS, and an increased ovarian volume. Our findings suggest that raised maternal androgens within the normal range may impact ovarian function in female offspring. In the same cohort we have failed to confirm that raised maternal androgen levels in normal pregnancy increase the incidence of PCOS in adolescence (5Hart R. Sloboda D.M. Doherty D.A. Norman R.J. Atkinson H.C. Newnham J.P. et al.Prenatal determinants of uterine volume and ovarian reserve in adolescence.J Clin Endocrinol Metab. 2009; 94: 4931-4937Crossref PubMed Scopus (43) Google Scholar), but there may be several reasons for this apparent anomaly. The androgen levels that we measured in the maternal blood may not be representative of fetal exposure due to placental aromatization, our cohort may have been too small to determine an effect, or potentially antenatal androgen exposure may not lead to all the features of PCOS as described by the Rotterdam or U.S. National Institutes of Health criteria (11Zawadzki J. Dunaif A. Diagnostic criteria for polycystic ovary syndrome: towards a rational approach.in: Polycystic ovarian syndrome Dunaif A. Givens J. Haseltine F. Merriam G. Blackwell, Boston1992: 377-384Google Scholar, 12ESHRE/ASRM Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome.Fertil Steril. 2004; 81: 19-25Abstract Full Text Full Text PDF Scopus (4508) Google Scholar). We have reported in this cohort that serum AMH is elevated in girls with PCO and PCOS, defined by both the Rotterdam and U.S. National Institutes of Health criteria; however, its ability to discriminate the presence of the condition from controls was poor (13Hart R. Doherty D.A. Norman R.J. Franks S. Dickinson J.E. Hickey M. Sloboda D.M. Serum antimullerian hormone (AMH) levels are elevated in adolescent girls with polycystic ovaries and the polycystic ovarian syndrome (PCOS).Fertil Steril. 2010 Jan 6; ([Epub ahead of print.])Abstract Full Text Full Text PDF Scopus (90) Google Scholar). The normal changes in androgen levels during pregnancy follow a pattern of an increase in testosterone concentrations during the first trimester with further increments through to term (14Bammann B.L. Coulam C.B. Jiang N.S. Total and free testosterone during pregnancy.Am J Obstet Gynecol. 1980; 137: 293-298Abstract Full Text PDF PubMed Scopus (96) Google Scholar). Placental aromatase activity was not measured in our study, and it is possible that variations in the conversion of androgens to estrogens may have altered fetal androgen exposure without changing circulating maternal androgen concentrations (15Jones M.E. Boon W.C. McInnes K. Maffei L. Carani C. Simpson E.R. Recognizing rare disorders: aromatase deficiency.Nat Clin Pract Endocrinol Metab. 2007; 3: 414-421Crossref PubMed Scopus (114) Google Scholar). Similarly, we have no measurements of fetal androgen responsiveness or androgen receptor affinity. Our is the first report of a prospective association between maternal testosterone levels and ovarian function in an unselected population. Our findings suggest that elevated maternal testosterone levels within the normal range may impact follicular development of female offspring. The authors thank all the families who took part in this study and the whole Raine Study team including data collectors, cohort managers, data managers, clerical staff, research scientists, and volunteers; and the Raine Medical Research Foundation of UWA and the Telethon Institute of Child Health Research for financial support and general support over the years. The collection of maternal data and samples was funded by the Women and Infants' Research Foundation (WIRF) ; the collection of adolescent data and samples was funded by NHMRC project grant number 403968 and by a UWA Ada Bartholomew grant . M. Hickey is funded by an NHMRC Clinical Career Development Award .