Advertisement
JOGC

The Role of Serum Testosterone in Early Pregnancy Outcome: A Comparison in Women With and Without Polycystic Ovary Syndrome

      Abstract

      Objective

      Hyperandrogenic conditions in women are associated with increased rates of miscarriage. However, the specific role of maternal testosterone in early pregnancy and its association with pregnancy outcome is unknown. The purpose of this study was to compare serum testosterone levels during early pregnancy in women with and without polycystic ovary syndrome (PCOS) who either had successful pregnancies or miscarried.

      Method

      We collected serum samples from women attending a university-based fertility centre at the time of their first positive serum beta human chorionic gonadotropin pregnancy test. The samples were subsequently assayed for total testosterone level. We used logistical regression modelling to control for PCOS diagnosis, BMI, and age.

      Results

      Total testosterone levels were available for 346 pregnancies, including 286 successful pregnancies and 78 first trimester miscarriages. We found no difference in total testosterone levels between women who subsequently had an ongoing pregnancy (mean concentration 3.6±2.6 nmol/L) and women with a miscarriage (mean 3.6±2.4 nmol/L). Using the Rotterdam criteria to identify women with PCOS, we also found no differences in serum testosterone between women who had ongoing pregnancies or miscarriages, either with PCOS (P=0.176) or without PCOS (P=0.561).

      Conclusions

      Our findings show that early pregnancy testosterone levels do not predict pregnancy outcome, and they call into question the role of testosterone in causing miscarriage in populations of women with PCOS. Further research is needed to elucidate the normal progression of testosterone levels during pregnancy and to investigate further the relationship between PCOS and miscarriage.

      Résumé

      Objectif

      Chez les femmes, les troubles liés à l’hyperandrogénie sont associés à des taux accrus de fausse couche. Toutefois, le rôle particulier que joue la testostérone maternelle aux débuts de la grossesse et l’influence qu’elle exerce sur l’issue de la grossesse restent inconnus. L’objectif de la présente étude était de comparer les taux sériques de testostérone aux débuts de la grossesse chez des femmes qui, en présence ou non d’un syndrome d’ovaires polykystiques (SOPK), avaient connu soit une grossesse réussie, soit une fausse couche.

      Méthode

      Nous avons prélevé des échantillons sériques chez des femmes qui fréquentaient un centre de fertilité universitaire, au moment de l’obtention de leur premier résultat positif au test de grossesse fondé sur le taux sérique de bêta-gonadotropine chorionique humaine. Les échantillons ont ensuite été analysés en vue d’établir le taux total de testostérone. Nous avons utilisé un modèle de régression logistique pour neutraliser l’effet du diagnostic de SOPK, de l’IMC et de l’âge.

      Résultats

      Des taux totaux de testostérone étaient disponibles pour 346 grossesses (286 grossesses réussies et 78 fausses couches au premier trimestre). Nous n’avons constaté aucune différence en matière de taux total de testostérone entre les femmes qui ont été en mesure de poursuivre leur grossesse (concentration moyenne de 3,6±2,6 nmol/l) et les femmes qui ont connu une fausse couche (moyenne de 3,6±2,4 nmol/l). En ayant recours aux critères de Rotterdam pour identifier les femmes présentant un SOPK, nous avons constaté qu’il n’existait également aucune différence en matière de taux sérique de testostérone entre les femmes qui ont pu poursuivre leur grossesse et celles qui ont connu une fausse couche, qu’il y ait eu présence d’un SOPK (P=0,176) ou non (P=0,561).

      Conclusions

      Les résultats que nous avons obtenus démontrent que les taux de testostérone présents aux débuts de la grossesse ne permettent pas de prévoir l’issue de la grossesse et remettent en question l’influence qu’exerce la testostérone sur la survenue d’une fausse couche chez les femmes qui présentent un SOPK. Des recherches plus poussées sont nécessaires pour élucider l’évolution normale des taux de testostérone pendant la grossesse, ainsi que pour explorer plus à fond le lien qui existe entre le SOPK et la fausse couche.

      Key Words

      ABBREVIATIONS

      β-hCG
      beta human chorionic gonadotropin
      PCOS
      polycystic ovary syndrome
      SHBG
      sex hormone binding globulin

      INTRODUCTION

      The establishment of a successful pregnancy requires both a healthy endometrial environment and a competent embryo. Whereas fetal causes of pregnancy loss (e.g. fetal aneuploidy) are well known, factors affecting endometrial competence are poorly understood. Under normal circumstances, the endometrium goes through a series of changes under steroid hormone regulation to become receptive to implantation by an embryo. For example, studies in mouse models and donor oocyte populations have shown that estrogen priming before progesterone treatment is necessary for endometrial receptivity.
      • Carson D.D.
      • Bagchi I.
      • Dey S.K.
      • Enders A.C.
      • Fazleabas A.T.
      • Lessey B.A.
      • et al.
      Embryo implantation.
      • Couse J.F.
      • Curtis S.W.
      • Washburn T.F.
      • Lindzey J.
      • Golding T.S.
      • Lubahn D.B.
      • et al.
      Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene.
      • Curtis Hewitt S.
      • Goulding E.H.
      • Eddy E.M.
      • Korach K.S.
      Studies using the estrogen receptor alpha knockout uterus demonstrate that implantation but not decidualization-associated signaling is estrogen dependent.
      Androgens and estrogens demonstrate an intricately balanced relationship in the normal control of endometrial cell growth and gene expression. Testosterone has been shown to inhibit estrogen-related gene expression in the endometrium, and elevated androgen levels have been proposed as a mechanism for miscarriage related to implantation defects rather than aneuploidy.
      • Kowalski A.A.
      • Vale-Cruz D.S.
      • Simmen F.A.
      • Simmen R.C.
      Uterine androgen receptors: roles in estrogen-mediated gene expression and DNA synthesis.
      Conditions associated with elevated testosterone levels, such as polycystic ovary syndrome and obesity, have been shown to have higher than expected miscarriage rates.
      • Okon M.A.
      • Laird S.M.
      • Tuckerman E.M.
      • Li T.C.
      Serum androgen levels in women who have recurrent miscarriages and their correlation with markers of endometrial function.
      • Wang J.X.
      • Davies M.J.
      • Norman R.J.
      Polycystic ovarian syndrome and the risk of spontaneous abortion following assisted reproductive technology treatment.
      • Ludwig M.
      • Finas D.F.
      • al-Hasani S.
      • Diedrich K.
      • Ortmann O.
      Oocyte quality and treatment outcome in intracytoplasmic sperm injection cycles of polycystic ovarian syndrome patients.
      • Cohen J.
      Laparoscopic procedures for treatment of infertility related to polycystic ovarian syndrome.
      • Bussen S.
      • Sutterlin M.
      • Steck T.
      Endocrine abnormalities during the follicular phase in women with recurrent spontaneous abortion.
      To date there is no evidence of higher aneuploidy rates in the embryos of women with PCOS who miscarry,
      • Weghofer A.
      • Munne S.
      • Chen S.
      • Barad D.
      • Gleicher N.
      Lack of association between polycystic ovary syndrome and embryonic aneuploidy.
      and no increase in rates of miscarriage has been observed when women with PCOS act as oocyte donors.
      • Ashkenazi J.
      • Farhi J.
      • Orvieto R.
      • Homburg R.
      • Dekel A.
      • Feldberg D.
      • et al.
      Polycystic ovary syndrome patients as oocyte donors: the effect of ovarian stimulation protocol on the implantation rate of the recipient.
      These findings suggest that the effects of elevated androgen levels on the endometrial environment may be the cause of miscarriage in women with PCOS.
      • Okon M.A.
      • Laird S.M.
      • Tuckerman E.M.
      • Li T.C.
      Serum androgen levels in women who have recurrent miscarriages and their correlation with markers of endometrial function.
      • Tulppala M.
      • Stenman U.H.
      • Cacciatore B.
      • Ylikorkala O.
      Polycystic ovaries and levels of gonadotrophins and androgens in recurrent miscarriage: prospective study in 50 women.
      Yet human studies have not shown a significant correlation between pre-pregnancy testosterone levels and pregnancy outcomes.
      • Nardo L.G.
      • Rai R.
      • Backos M.
      • El-Gaddal S.
      • Regan L.
      High serum luteinizing hormone and testosterone concentrations do not predict pregnancy outcome in women with recurrent miscarriage.
      However, testosterone levels have not been carefully examined during early pregnancy and implantation. In a small study of 53 women undergoing IVF, Takeuchi et al. did not find testosterone levels at three days and at 14 days after embryo transfer to be predictive of pregnancy rates using logistic regression analysis, but they did find a higher mean testosterone level at day 14 in women who became pregnant.
      • Takeuchi T.
      • Nishii O.
      • Okamura T.
      • Yaginuma T.
      • Kawana T.
      Free testosterone and abortion in early pregnancy.
      However, it is unclear in that study how testosterone was measured and whether birth outcome was recorded.
      The objective of this study was to examine serum testosterone levels at four to six weeks’ gestational age in both successful pregnancies and those ending in miscarriage, and also in women with and without PCOS. It was hypothesized that women who miscarried would demonstrate significantly higher total testosterone levels in early pregnancy than women with ongoing pregnancies.

      METHODS

      Women attending our Northern California university-based fertility clinic for management of infertility who had a possible pregnancy were approached for participation in this study. In addition to having blood drawn for quantitative β-hCG assessment at this time, women who gave informed consent had an additional aliquot of serum drawn and frozen at −20 °C. In this study, an initial serum sample was collected from these women at the time of first positive serum pregnancy test, at between four and six weeks’ gestation. For most participants (287/346, 83%), we collected a follow-up serum sample on average two days after the initial sample (minimum 2 days, maximum 6 days). In this study, samples from women with single ongoing pregnancies or first trimester miscarriages during a three-year period were later analyzed for total testosterone.
      All subjects, regardless of method of conception, were excluded from the analysis if they had thyroid dysfunction (TSH2.5 mU/L), hyperprolactinemia on two samples (serum prolactin above the upper limit of normal for the assay used), ovarian or adrenal tumours, congenital adrenal hyperplasia, or hypothalamic amenorrhea. All subjects underwent hysteroscopy before treatment. Women with submucous fibroids, endometrial polyps, or intrauterine synechiae had these surgically corrected before treatment was initiated. Women with uterine anomalies or intramural fibroids>3 cm in diameter were excluded from the evaluation.
      During the study period, a total of 445 infertile women who conceived with or without ovarian stimulation were considered for inclusion. Forty-two women who conceived after oocyte donation or frozen embryo transfer and 61 women with multiple gestations were excluded from the analysis. Four women underwent oocyte donation and conceived a multiple gestation. In total, 99 women were excluded using these criteria; the remaining 346 women were included in the final analysis.
      Study participants with increasing serum β-hCG levels were scheduled for ultrasound examination to confirm pregnancy viability at six to seven weeks’ gestation, with a repeat ultrasound one to two weeks later. If no cardiac activity was detected at transvaginal ultrasound on two occasions, missed abortion was diagnosed, and dilatation and curettage was offered. All women who chose to undergo suction curettage for a missed abortion were offered cytogenetic testing of the products of conception. Only clinical miscarriages or pregnancies lost after the confirmation of an intrauterine gestational sac on ultrasound were included in this study. Pregnancies continuing beyond the first trimester were referred for obstetrical care.
      Demographic information and mode of conception were obtained through review of medical records. PCOS status was determined using the Rotterdam criteria.
      • Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group
      Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS).
      Serum total testosterone levels in women with miscarriages and ongoing pregnancies were compared. Samples were assayed for total testosterone using the Immulite 2500 analyzer (Siemens Healthcare Diagnostics, Malvern, PA). The kit uses a solid-phase, competitive chemiluminescent enzyme immunoassay with an analytical sensitivity of 0.52 nmol/L and a reportable assay range of 0.69 to 55.52 nmol/L. Intra-assay coefficients of variation were 10.2% and 6.3% for testosterone concentrations of 3.85 and 24.12 nmol/L, respectively. Inter-assay coefficients of variation were 12.4%, 12.0%, and 9.4% for testosterone concentrations of 6.97, 22.42, and 45.01 nmol/L, respectively. For participants with more than one blood draw (n=239), total testosterone values were averaged; this was felt to provide a more accurate picture of total testosterone levels during this gestational window because of the potentially cyclical nature of testosterone secretion, although studies on the pattern of testosterone secretion in pregnancy are scarce.
      • Takeuchi T.
      • Nishii O.
      • Okamura T.
      • Yaginuma T.
      • Kawana T.
      Free testosterone and abortion in early pregnancy.
      All statistical analyses were performed using SPSS 11.0 (IBM Corp., Armonk NY). Student t test, and analysis of variance was used for comparison of continuous variables. Levene’s test for equality of variances and tests for homogeneity of variances (in the case of analysis of variance) were applied to the data and the corresponding t test and P values were accepted depending on whether the variances were equal. Chi-square analysis was used to compare categorical data. Pearson scores were used to assess the correlation of total testosterone with demographic characteristics. Logistic regression models were used to control for PCOS, BMI, and age. Statistical significance was accepted as a two-sided P value0.05. A post-hoc statistical power analysis was performed with the following assumptions: independent samples, a 5% alpha error in a two-sided test, and 80% power to detect a difference, with mean serum total testosterone levels of 3.6 nmol/L with a standard deviation of 2.67 and a difference between groups of 25% (conservative), required a total of 298 subjects to detect a difference between groups.
      The Stanford University Institutional Review Board for Human Subjects Research reviewed and approved this study.

      RESULTS

      Demographic characteristics and serum testosterone levels in women with an ongoing pregnancy (n=268) and women who experienced a miscarriage (n=78) are shown in Table 1. There were no significant differences between groups in mean age, BMI, number of prior pregnancies and miscarriages, or total testosterone levels two to four weeks after conception. The first (earlier in pregnancy) total testosterone levels ranged from 0.66 to 23.46 nmol/L (95% of subjects had a level<7.6 nmol/L) and the second (later in pregnancy) levels ranged from 0.66 to 20.79 nmol/L (95% of subjects had a level<10.24 nmol/L). Rates of PCOS were similar in the two groups: of the women with ongoing pregnancies, 18% (n=46) had PCOS according to the Rotterdam criteria,
      • Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group
      Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS).
      compared to 13% (n=10) of the women who miscarried (P=0.24). The lack of significant differences held true even when PCOS status was controlled for using logistic regression models (data not shown).
      Table 1Demographics of women with ongoing pregnancy versus miscarriage
      Ongoing (n=268)Miscarriage (n=78)P
      Age, years35.6±4.1636.1±4.490.354
      Prior pregnancies1.14±1.281.32±1.460.329
      Prior full-term deliveries0.30±0.500.30±0.460.974
      Prior miscarriages0.64±1.020.86±1.280.125
      BMI, kg/m
      • Couse J.F.
      • Curtis S.W.
      • Washburn T.F.
      • Lindzey J.
      • Golding T.S.
      • Lubahn D.B.
      • et al.
      Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene.
      23.2±4.7523.5±4.700.558
      Serum progesterone, nmol/L106.8±163.8108.1±168.20.949
      Serum FSH, IU/L7.34±3.527.57±4.470.656
      Mean total testosterone, nmol/L3.6±2.663.6±2.370.916
      Polycystic ovary syndrome, n (%)49 (18)10 (13)0.259
      Natural cycle conceptions, n (%)75 (28)16 (21)0.244
      Women with PCOS were on average significantly younger, had a higher BMI, and had fewer previous pregnancies, full-term deliveries, and miscarriages than women without PCOS (Table 2). Mean serum total testosterone was significantly higher in the PCOS group (P=0.002). Pearson correlation analyses did not indicate a significant correlation between age and total testosterone levels (P=0.741), but did show a significant positive correlation between total testosterone and BMI (Pearson correlation=0.164, P=0.003).
      Table 2Demographics of women with polycystic ovary syndrome versus non-PCOS
      Non-PCOS (n=287)PCOS (n=59)P
      Age, years36.3±4.1732.5±2.99<0.001
      Prior pregnancies1.28±1.350.71±1.04<0.001
      Prior full-term deliveries0.33±0.510.16±0.3650.003
      Prior miscarriages0.75±1.130.41±0.7950.008
      BMI (kg/m2)22.7±3.9126.0±6.990.001
      Mean serum progesterone, nmol/L109.1±170.197.3±133.90.62
      Mean serum FSH, IU/L7.76±3.935.61±1.60<0.001
      Mean serum total testosterone, nmol/L3.4±2.44.8±3.20.002
      A comparison of total testosterone levels in women with ongoing pregnancy and in those who miscarried, stratified by PCOS status, is shown in Table 3. While total testosterone levels in women with PCOS and an ongoing pregnancy or a miscarriage were higher than the corresponding levels in women without PCOS (4.58 vs. 3.44 nmol/L and 6.07 vs. 3.26 nmol/L, respectively), the differences in total testosterone levels within the PCOS and non-PCOS populations were not significant when stratifying for miscarriage versus ongoing pregnancy outcome. Among women without PCOS who miscarried, there was no difference in total testosterone levels between those with aneuploid products of conception and those whose products of conception had a normal karyotype. All of the women with PCOS who miscarried had an aneuploid pregnancy.
      Table 3Serum total testosterone (nmol/L) stratified by PCOS status
      Non-PCOS (n=287)PCOS (n=59)P
      Ongoing pregnancy3.45±2.52 (n=219)4.57±3.10 (n=49)0.021
      Miscarriage3.26±1.91 (n=68)6.07±3.63 (n=10)0.037
      P=0.561P=0.176
      Normal karyotype2.54±1.05 (n=11)No casesN/A
      Abnormal karyotype3.49±1.99 (n=26)5.35±2.87 (n=5)0.086
      No karyotype performed3.31±2.06 (n=31)6.83±4.48 (n=5)0.154
      P=0.373P=0.549

      DISCUSSION

      There are minimal data comparing total testosterone levels at four to six weeks’ gestation in women with ongoing pregnancies and women who miscarry. Although implantation of the blastocyst occurs six to seven days after fertilization, the four- to six-week gestational period is marked by placental invasion and establishment of maternal–placental blood flow, in addition to critical organogenesis in the fetus.
      • Sadler T.W.
      • Langman J.
      It has been suggested that elevated maternal total testosterone levels during early pregnancy could disrupt implantation and lead to pregnancy loss.
      • Giudice L.C.
      Endometrium in PCOS: implantation and predisposition to endocrine CA.
      However, we found no difference in total testosterone levels between women with an early miscarriage and women with ongoing pregnancies.
      Small studies have assessed serum testosterone levels (measured as total testosterone and/or free androgen index) in women during the period from the fourth to the twelfth week of gestation; the findings were mixed, with some studies reporting a higher mean testosterone level in women who miscarried
      • Takeuchi T.
      • Nishii O.
      • Okamura T.
      • Yaginuma T.
      • Kawana T.
      Free testosterone and abortion in early pregnancy.
      • Cocksedge K.A.
      • Saravelos S.H.
      • Wang Q.
      • Tuckerman E.
      • Laird S.M.
      • Li T.C.
      Does free androgen index predict subsequent pregnancy outcome in women with recurrent miscarriage?.
      and other studies showing no difference according to pregnancy outcome.
      • Nardo L.G.
      • Rai R.
      • Backos M.
      • El-Gaddal S.
      • Regan L.
      High serum luteinizing hormone and testosterone concentrations do not predict pregnancy outcome in women with recurrent miscarriage.
      • Aksoy S.
      • Celikkanat H.
      • Senoz S.
      • Gokmen O.
      The prognostic value of serum estradiol, progesterone, testosterone and free testosterone levels in detecting early abortions.
      The optimal methods for measuring serum testosterone and the timing of sample acquisition are controversial, making it difficult to compare studies. It is also unclear whether and how much testosterone levels change during this significant gestational interval, as the normal testosterone levels during early pregnancy have not been adequately defined.
      • Regan L.
      • Owen E.J.
      • Jacobs H.S.
      Hypersecretion of luteinising hormone, infertility, and miscarriage.
      • Tarlatzis B.C.
      • Grimbizis G.
      • Pournaropoulos F.
      • Bontis J.
      • Lagos S.
      • Spanos E.
      • et al.
      The prognostic value of basal luteinizing hormone: follicle-stimulating hormone ratio in the treatment of patients with polycystic ovarian syndrome by assisted reproduction techniques.
      • Sugiura-Ogasawara M.
      • Sato T.
      • Suzumori N.
      • Kitaori T.
      • Kumagai K.
      • Ozaki Y.
      The polycystic ovary syndrome does not predict further miscarriage in Japanese couples experiencing recurrent miscarriages.
      • Cocksedge K.A.
      • Li T.C.
      • Saravelos S.H.
      • Metwally M.
      A reappraisal of the role of polycystic ovary syndrome in recurrent miscarriage.
      • Legro R.S.
      Pregnancy considerations in women with polycystic ovary syndrome.
      Our study benefits from having a larger sample size and a relatively narrow gestational age interval compared with other studies.
      • Takeuchi T.
      • Nishii O.
      • Okamura T.
      • Yaginuma T.
      • Kawana T.
      Free testosterone and abortion in early pregnancy.
      • Aksoy S.
      • Celikkanat H.
      • Senoz S.
      • Gokmen O.
      The prognostic value of serum estradiol, progesterone, testosterone and free testosterone levels in detecting early abortions.
      • Legro R.S.
      Pregnancy considerations in women with polycystic ovary syndrome.
      • Bammann B.L.
      • Coulam C.B.
      • Jiang N.S.
      Total and free testosterone during pregnancy.
      • Carlsen S.M.
      • Jacobsen G.
      • Romundstad P.
      Maternal testosterone levels during pregnancy are associated with offspring size at birth.
      • Shaheen S.O.
      • Hines M.
      • Newson R.B.
      • Wheeler M.
      • Herrick D.R.
      • Strachan D.P.
      • et al.
      Maternal testosterone in pregnancy and atopic outcomes in childhood.
      One potential limitation of this study was the measurement of total testosterone rather than free testosterone, free androgen index, or “bioavailable” testosterone, which are possibly more reliable and/or specific in their assessment of hyperandrogenemia. It has been shown that total testosterone levels are positively correlated with free testosterone, possibly through a negative effect of testosterone on sex hormone binding globulin production
      • Boomsma C.M.
      • Fauser B.C.
      • Macklon N.S.
      Pregnancy complications in women with polycystic ovary syndrome.
      • Hardy K.
      • Robinson F.M.
      • Paraschos T.
      • Wicks R.
      • Franks S.
      • Winston R.M.
      Normal development and metabolic activity of preimplantation embryos in vitro from patients with polycystic ovaries.
      • Sahu B.
      • Ozturk O.
      • Ranierri M.
      • Serhal P.
      Comparison of oocyte quality and intracytoplasmic sperm injection outcome in women with isolated polycystic ovaries or polycystic ovarian syndrome.
      ; however, these studies did not address populations of pregnant women. It is hypothesized, for example, that estrogenic effects in pregnant women can cause increases in SHBG levels.
      • Ruutiainen K.
      • Erkkola R.
      • Kaihola H.L.
      • Santti R.
      • Irjala K.
      The grade of hirsutism correlated to serum androgen levels and hormonal indices.
      In studies of women with PCOS during pregnancy, blood samples taken between five and 12 weeks’ gestation showed evidence of elevated SHBG levels
      • Belgorosky A.
      • Rivarola M.A.
      Dynamics of SHBG response to testosterone. Implications upon the immediate biological effect of sex hormones.
      ; however, what happens to levels of total, free, and bioavailable testosterone and of SHBG throughout pregnancy, and the optimal methods of assessing free and bioavailable testosterone, continue to be debated.
      • Boomsma C.M.
      • Fauser B.C.
      • Macklon N.S.
      Pregnancy complications in women with polycystic ovary syndrome.
      • Hossein Rashidi B.
      • Hormoz B.
      • Shahrokh Tehraninejad E.
      • Shariat M.
      • Mahdavi A.
      Testosterone and dehydroepiandrosterone sulphate levels and IVF/ICSI results.
      In pregnancy there are significant changes in hormone levels, including estrogens, progesterone, prolactin, thyroid stimulating hormone, adrenocorticotropin hormone, cortisol, and insulin, which exert both positive and negative effects on serum SHBG levels.
      • Boomsma C.M.
      • Fauser B.C.
      • Macklon N.S.
      Pregnancy complications in women with polycystic ovary syndrome.
      • Hardy K.
      • Robinson F.M.
      • Paraschos T.
      • Wicks R.
      • Franks S.
      • Winston R.M.
      Normal development and metabolic activity of preimplantation embryos in vitro from patients with polycystic ovaries.
      • Sahu B.
      • Ozturk O.
      • Ranierri M.
      • Serhal P.
      Comparison of oocyte quality and intracytoplasmic sperm injection outcome in women with isolated polycystic ovaries or polycystic ovarian syndrome.
      • Ruutiainen K.
      • Erkkola R.
      • Kaihola H.L.
      • Santti R.
      • Irjala K.
      The grade of hirsutism correlated to serum androgen levels and hormonal indices.
      • Hossein Rashidi B.
      • Hormoz B.
      • Shahrokh Tehraninejad E.
      • Shariat M.
      • Mahdavi A.
      Testosterone and dehydroepiandrosterone sulphate levels and IVF/ICSI results.
      Because of the strong correlation between total and free testosterone levels, the changing hormonal effects on SHBG levels throughout pregnancy, and the controversy surrounding the optimal method to determine bioavailable testosterone, we selected total testosterone for evaluation. Serum free testosterone levels could be calculated by measuring SHBG levels in frozen samples, but the cost of the SHBG assay precluded our measuring these levels. Clearly, an evaluation of pre-pregnancy hormone levels including testosterone would not have been useful in this study. The values would have changed significantly in pregnancy.
      Other limitations of the study include the fact that the study population consisted of women presenting to an infertility clinic in a relatively affluent, suburban area. Correspondingly, it should be noted that the average BMI overall and in the PCOS subgroup is much lower than is found in most studies of the PCOS population.
      • Shaheen S.O.
      • Hines M.
      • Newson R.B.
      • Wheeler M.
      • Herrick D.R.
      • Strachan D.P.
      • et al.
      Maternal testosterone in pregnancy and atopic outcomes in childhood.
      • Sagle M.
      • Bishop K.
      • Ridley N.
      • Alexander F.M.
      • Michel M.
      • Bonney R.C.
      • et al.
      Recurrent early miscarriage and polycystic ovaries.
      These observations are important to bear in mind when generalizing the results of our study to other pregnant populations. A more detailed analysis of this PCOS population, including the specific distribution of Rotterdam criteria
      • Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group
      Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS).
      used to assess the syndrome, is outside the scope of this study but is an area deserving of future investigation. Additionally, we did not identify reasons for infertility, including information on what proportion was due to male factor, which could potentially play a role in determining pregnancy outcome.
      Differences in demographic characteristics and total testosterone levels between women with and without PCOS followed expected patterns that have been established in previous reports.
      • Belgorosky A.
      • Rivarola M.A.
      Dynamics of SHBG response to testosterone. Implications upon the immediate biological effect of sex hormones.
      For example, women with PCOS had a higher mean BMI, lower rates of prior pregnancy and live birth, and higher total testosterone levels. Although early reports suggested an increased risk of miscarriage in women with PCOS (potentially related to high LH levels),
      • Regan L.
      • Owen E.J.
      • Jacobs H.S.
      Hypersecretion of luteinising hormone, infertility, and miscarriage.
      • Tarlatzis B.C.
      • Grimbizis G.
      • Pournaropoulos F.
      • Bontis J.
      • Lagos S.
      • Spanos E.
      • et al.
      The prognostic value of basal luteinizing hormone: follicle-stimulating hormone ratio in the treatment of patients with polycystic ovarian syndrome by assisted reproduction techniques.
      • Sagle M.
      • Bishop K.
      • Ridley N.
      • Alexander F.M.
      • Michel M.
      • Bonney R.C.
      • et al.
      Recurrent early miscarriage and polycystic ovaries.
      • Balen A.H.
      • Tan S.L.
      • MacDougall J.
      • Jacobs H.S.
      Miscarriage rates following in-vitro fertilization are increased in women with polycystic ovaries and reduced by pituitary desensitization with buserelin.
      little prospective evidence using current PCOS diagnostic criteria is available to assess this association, and recent studies have failed to corroborate these findings.
      • Sugiura-Ogasawara M.
      • Sato T.
      • Suzumori N.
      • Kitaori T.
      • Kumagai K.
      • Ozaki Y.
      The polycystic ovary syndrome does not predict further miscarriage in Japanese couples experiencing recurrent miscarriages.
      • Cocksedge K.A.
      • Li T.C.
      • Saravelos S.H.
      • Metwally M.
      A reappraisal of the role of polycystic ovary syndrome in recurrent miscarriage.
      • Legro R.S.
      Pregnancy considerations in women with polycystic ovary syndrome.
      • Boomsma C.M.
      • Fauser B.C.
      • Macklon N.S.
      Pregnancy complications in women with polycystic ovary syndrome.
      Since studies on oocytes in women with PCOS do not indicate quality inferior to that of oocytes in women without PCOS,
      • Hardy K.
      • Robinson F.M.
      • Paraschos T.
      • Wicks R.
      • Franks S.
      • Winston R.M.
      Normal development and metabolic activity of preimplantation embryos in vitro from patients with polycystic ovaries.
      • Sahu B.
      • Ozturk O.
      • Ranierri M.
      • Serhal P.
      Comparison of oocyte quality and intracytoplasmic sperm injection outcome in women with isolated polycystic ovaries or polycystic ovarian syndrome.
      researchers continue to search for the mechanisms of a potential link between PCOS and miscarriage. It should be noted that although the results were not statistically significant, the rate of miscarriage among women with PCOS was lower than among normal control subjects (17% vs. 24%).
      This study showed a significantly higher total testosterone level in the pregnancies of women with PCOS, whether or not miscarriage occurred. Interestingly, the mean total testosterone level in the women with PCOS who miscarried was 34% higher than in those with ongoing pregnancies. Unfortunately, our study did not have adequate statistical power to assess this association. Although there was no clear difference in the women without PCOS, we believe that the relationship between serum testosterone and pregnancy outcomes in women with PCOS deserves further study. Other hormones for future evaluation include luteinizing hormone and insulin-like growth factor 1.
      Our study provides important data on total testosterone measured during a consistent window in early pregnancy, and is notable for its large sample size. Future studies are needed to elucidate the impact of maternal serum testosterone on the endometrium during the window of implantation and early placentation.

      CONCLUSION

      These findings suggest that early pregnancy testosterone levels do not predict pregnancy outcome, and call into question the role of testosterone in causing miscarriage in PCOS populations. Future research is needed to better elucidate the normal progression of testosterone levels during pregnancy and to investigate further the reported relationship between PCOS and miscarriage.

      ACKNOWLEDGEMENTS

      The authors are grateful to DSL Inc. for donating reagents for testosterone assays.

      REFERENCES

        • Carson D.D.
        • Bagchi I.
        • Dey S.K.
        • Enders A.C.
        • Fazleabas A.T.
        • Lessey B.A.
        • et al.
        Embryo implantation.
        Dev Biol. 2000; 223: 217-237
        • Couse J.F.
        • Curtis S.W.
        • Washburn T.F.
        • Lindzey J.
        • Golding T.S.
        • Lubahn D.B.
        • et al.
        Analysis of transcription and estrogen insensitivity in the female mouse after targeted disruption of the estrogen receptor gene.
        Mol Endocrinol. 1995; 9: 1441-1454
        • Curtis Hewitt S.
        • Goulding E.H.
        • Eddy E.M.
        • Korach K.S.
        Studies using the estrogen receptor alpha knockout uterus demonstrate that implantation but not decidualization-associated signaling is estrogen dependent.
        Biol Reprod. 2002; 67: 1268-1277
        • Kowalski A.A.
        • Vale-Cruz D.S.
        • Simmen F.A.
        • Simmen R.C.
        Uterine androgen receptors: roles in estrogen-mediated gene expression and DNA synthesis.
        Biol Reprod. 2004; 70: 1349-1357
        • Okon M.A.
        • Laird S.M.
        • Tuckerman E.M.
        • Li T.C.
        Serum androgen levels in women who have recurrent miscarriages and their correlation with markers of endometrial function.
        Fertil Steril. 1998; 69: 682-690
        • Wang J.X.
        • Davies M.J.
        • Norman R.J.
        Polycystic ovarian syndrome and the risk of spontaneous abortion following assisted reproductive technology treatment.
        Hum Reprod. 2001; 16: 2606-2609
        • Ludwig M.
        • Finas D.F.
        • al-Hasani S.
        • Diedrich K.
        • Ortmann O.
        Oocyte quality and treatment outcome in intracytoplasmic sperm injection cycles of polycystic ovarian syndrome patients.
        Hum Reprod. 1999; 14: 354-358
        • Cohen J.
        Laparoscopic procedures for treatment of infertility related to polycystic ovarian syndrome.
        Hum Reprod Update. 1996; 2: 337-344
        • Bussen S.
        • Sutterlin M.
        • Steck T.
        Endocrine abnormalities during the follicular phase in women with recurrent spontaneous abortion.
        Hum Reprod. 1999; 14: 18-20
        • Weghofer A.
        • Munne S.
        • Chen S.
        • Barad D.
        • Gleicher N.
        Lack of association between polycystic ovary syndrome and embryonic aneuploidy.
        Fertil Steril. 2007; 88: 900-905
        • Ashkenazi J.
        • Farhi J.
        • Orvieto R.
        • Homburg R.
        • Dekel A.
        • Feldberg D.
        • et al.
        Polycystic ovary syndrome patients as oocyte donors: the effect of ovarian stimulation protocol on the implantation rate of the recipient.
        Fertil Steril. 1995; 64: 564-567
        • Tulppala M.
        • Stenman U.H.
        • Cacciatore B.
        • Ylikorkala O.
        Polycystic ovaries and levels of gonadotrophins and androgens in recurrent miscarriage: prospective study in 50 women.
        Br J Obstet Gynaecol. 1993; 100: 348-352
        • Nardo L.G.
        • Rai R.
        • Backos M.
        • El-Gaddal S.
        • Regan L.
        High serum luteinizing hormone and testosterone concentrations do not predict pregnancy outcome in women with recurrent miscarriage.
        Fertil Steril. 2002; 77: 348-352
        • Takeuchi T.
        • Nishii O.
        • Okamura T.
        • Yaginuma T.
        • Kawana T.
        Free testosterone and abortion in early pregnancy.
        Int J Gynaecol Obstet. 1993; 43: 151-156
        • Rotterdam ESHRE/ASRM-sponsored PCOS Consensus Workshop Group
        Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome (PCOS).
        Hum Reprod. 2004; 19: 41-47
        • Sadler T.W.
        • Langman J.
        Langman’s medical embryology. Lippincott Williams & Wilkins, Philadelphia2006
        • Giudice L.C.
        Endometrium in PCOS: implantation and predisposition to endocrine CA.
        Best Pract Res Clin Endocrinol Metab. 2006; 20: 235-244
        • Cocksedge K.A.
        • Saravelos S.H.
        • Wang Q.
        • Tuckerman E.
        • Laird S.M.
        • Li T.C.
        Does free androgen index predict subsequent pregnancy outcome in women with recurrent miscarriage?.
        Hum Reprod. 2008; 23: 797-802
        • Aksoy S.
        • Celikkanat H.
        • Senoz S.
        • Gokmen O.
        The prognostic value of serum estradiol, progesterone, testosterone and free testosterone levels in detecting early abortions.
        Eur J Obstet Gynecol Reprod Biol. 1996; 67: 5-8
        • Regan L.
        • Owen E.J.
        • Jacobs H.S.
        Hypersecretion of luteinising hormone, infertility, and miscarriage.
        Lancet. 1990; 336: 1141-1144
        • Tarlatzis B.C.
        • Grimbizis G.
        • Pournaropoulos F.
        • Bontis J.
        • Lagos S.
        • Spanos E.
        • et al.
        The prognostic value of basal luteinizing hormone: follicle-stimulating hormone ratio in the treatment of patients with polycystic ovarian syndrome by assisted reproduction techniques.
        Hum Reprod. 1995; 10: 2545-2549
        • Sugiura-Ogasawara M.
        • Sato T.
        • Suzumori N.
        • Kitaori T.
        • Kumagai K.
        • Ozaki Y.
        The polycystic ovary syndrome does not predict further miscarriage in Japanese couples experiencing recurrent miscarriages.
        Am J Reprod Immunol. 2009; 61: 62-67
        • Cocksedge K.A.
        • Li T.C.
        • Saravelos S.H.
        • Metwally M.
        A reappraisal of the role of polycystic ovary syndrome in recurrent miscarriage.
        Reprod Biomed Online. 2008; 17: 151-160
        • Legro R.S.
        Pregnancy considerations in women with polycystic ovary syndrome.
        Clin Obstet Gynecol. 2007; 50: 295-304
        • Bammann B.L.
        • Coulam C.B.
        • Jiang N.S.
        Total and free testosterone during pregnancy.
        Am J Obstet Gynecol. 1980; 137: 293-298
        • Carlsen S.M.
        • Jacobsen G.
        • Romundstad P.
        Maternal testosterone levels during pregnancy are associated with offspring size at birth.
        Eur J Endocrinol. 2006; 155: 365-370
        • Shaheen S.O.
        • Hines M.
        • Newson R.B.
        • Wheeler M.
        • Herrick D.R.
        • Strachan D.P.
        • et al.
        Maternal testosterone in pregnancy and atopic outcomes in childhood.
        Allergy. 2007; 62: 25-32
        • Boomsma C.M.
        • Fauser B.C.
        • Macklon N.S.
        Pregnancy complications in women with polycystic ovary syndrome.
        Semin Reprod Med. 2008; 26: 72-84
        • Hardy K.
        • Robinson F.M.
        • Paraschos T.
        • Wicks R.
        • Franks S.
        • Winston R.M.
        Normal development and metabolic activity of preimplantation embryos in vitro from patients with polycystic ovaries.
        Hum Reprod. 1995; 10: 2125-2135
        • Sahu B.
        • Ozturk O.
        • Ranierri M.
        • Serhal P.
        Comparison of oocyte quality and intracytoplasmic sperm injection outcome in women with isolated polycystic ovaries or polycystic ovarian syndrome.
        Arch Gynecol Obstet. 2008; 277: 239-244
        • Ruutiainen K.
        • Erkkola R.
        • Kaihola H.L.
        • Santti R.
        • Irjala K.
        The grade of hirsutism correlated to serum androgen levels and hormonal indices.
        Acta Obstet Gynecol Scand. 1985; 64: 629-633
        • Belgorosky A.
        • Rivarola M.A.
        Dynamics of SHBG response to testosterone. Implications upon the immediate biological effect of sex hormones.
        J Steroid Biochem. 1983; 18: 783-787
        • Hossein Rashidi B.
        • Hormoz B.
        • Shahrokh Tehraninejad E.
        • Shariat M.
        • Mahdavi A.
        Testosterone and dehydroepiandrosterone sulphate levels and IVF/ICSI results.
        Gynecol Endocrin. 2009; 25: 194-198
        • Sagle M.
        • Bishop K.
        • Ridley N.
        • Alexander F.M.
        • Michel M.
        • Bonney R.C.
        • et al.
        Recurrent early miscarriage and polycystic ovaries.
        BMJ. 1988; 297: 1027-1028
        • Balen A.H.
        • Tan S.L.
        • MacDougall J.
        • Jacobs H.S.
        Miscarriage rates following in-vitro fertilization are increased in women with polycystic ovaries and reduced by pituitary desensitization with buserelin.
        Hum Reprod. 1993; 8: 959-964