Preconception Phthalate Exposure and Women's Reproductive Health: Pregnancy, Pregnancy Loss, and Underlying Mechanisms.

BACKGROUND: Phthalates are endocrine-disrupting chemicals linked to adverse pregnancy outcomes. Despite the sensitivity of female reproductive processes to oxidation-reduction reaction stress and endocrine disruption, evidence for the impact of women's phthalate exposure on the ability to establish and maintain pregnancy has been inconclusive. OBJECTIVES: We aimed to determine the relationship of preconception phthalate metabolite exposure with a) fecundability and pregnancy loss and b) markers of potential biological mechanisms, including reproductive hormones, inflammation, and oxidative stress. METHODS: Data were collected from the Effects of Aspirin in Gestation and Reproduction (EAGeR) trial, a preconception study following 1,228 women who were attempting pregnancy, for up to six menstrual cycles and throughout pregnancy if they became pregnant. Twenty phthalate metabolites were measured in a consecutive 3-d pooled urine sample at enrollment. Pregnancy was determined through urinary human chorionic gonadotropin (hCG) at the expected date of menses during each cycle and pregnancy loss as an observed loss following positive hCG. Highly sensitive C-reactive protein (hsCRP) and isoprostanes were measured at enrollment, and reproductive hormones were measured during the follicular phase, ovulation, and luteal phase. Discrete-time Cox proportional hazards models evaluated the relationship of phthalate metabolites with fecundability and weighted Poisson models with robust variance evaluated the risk of pregnancy loss. RESULTS: An interquartile range (IQR) higher mono-(2-ethylhexyl) phthalate [fecundability odds ratio (FOR)=0.88; 95% confidence interval (CI): 0.78, 1.00], mono-butyl phthalate (FOR=0.82; 95% CI: 0.70, 0.96), and mono-benzyl phthalate (FOR=0.85; 95% CI: 0.74, 0.98) was associated with lower fecundability. No consistent associations were observed with pregnancy loss. Preconception phthalates were consistently associated with higher hsCRP and isoprostanes, as well as lower estradiol and higher follicle-stimulating hormone across the menstrual cycle. DISCUSSION: Women's preconception exposure to phthalates was associated with lower fecundability, changes in reproductive hormones, and increased inflammation and oxidative stress. The pre- and periconception periods may represent sensitive windows for intervening to limit the reproductive toxicity of phthalate exposure. https://doi.org/10.1289/EHP12287.

Pregnancy Complications and Long-Term Mortality in a Diverse Cohort.

BACKGROUND: Pregnancy complications are associated with increased risk of development of cardiometabolic diseases and earlier mortality. However, much of the previous research has been limited to White pregnant participants. We aimed to investigate pregnancy complications in association with total and cause-specific mortality in a racially diverse cohort and evaluate whether associations differ between Black and White pregnant participants. METHODS: The Collaborative Perinatal Project was a prospective cohort study of 48 197 pregnant participants at 12 US clinical centers (1959-1966). The Collaborative Perinatal Project Mortality Linkage Study ascertained participants' vital status through 2016 with linkage to the National Death Index and Social Security Death Master File. Adjusted hazard ratios (aHRs) for underlying all-cause and cause-specific mortality were estimated for preterm delivery (PTD), hypertensive disorders of pregnancy, and gestational diabetes/impaired glucose tolerance (GDM/IGT) using Cox models adjusted for age, prepregnancy body mass index, smoking, race and ethnicity, previous pregnancies, marital status, income, education, previous medical conditions, site, and year. RESULTS: Among 46 551 participants, 45% (21 107 of 46 551) were Black, and 46% (21 502 of 46 551) were White. The median time between the index pregnancy and death/censoring was 52 years (interquartile range, 45-54). Mortality was higher among Black (8714 of 21 107 [41%]) compared with White (8019 of 21 502 [37%]) participants. Overall, 15% (6753 of 43 969) of participants had PTD, 5% (2155 of 45 897) had hypertensive disorders of pregnancy, and 1% (540 of 45 890) had GDM/IGT. PTD incidence was higher in Black (4145 of 20 288 [20%]) compared with White (1941 of 19 963 [10%]) participants. The following were associated with all-cause mortality: preterm spontaneous labor (aHR, 1.07 [95% CI, 1.03-1.1]); preterm premature rupture of membranes (aHR, 1.23 [1.05-1.44]); preterm induced labor (aHR, 1.31 [1.03-1.66]); preterm prelabor cesarean delivery (aHR, 2.09 [1.75-2.48]) compared with full-term delivery; gestational hypertension (aHR, 1.09 [0.97-1.22]); preeclampsia or eclampsia (aHR, 1.14 [0.99-1.32]) and superimposed preeclampsia or eclampsia (aHR, 1.32 [1.20-1.46]) compared with normotensive; and GDM/IGT (aHR, 1.14 [1.00-1.30]) compared with normoglycemic. P values for effect modification between Black and White participants for PTD, hypertensive disorders of pregnancy, and GDM/IGT were 0.009, 0.05, and 0.92, respectively. Preterm induced labor was associated with greater mortality risk among Black (aHR, 1.64 [1.10-2.46]) compared with White (aHR, 1.29 [0.97-1.73]) participants, while preterm prelabor cesarean delivery was higher in White (aHR, 2.34 [1.90-2.90]) compared with Black (aHR, 1.40 [1.00-1.96]) participants. CONCLUSIONS: In this large, diverse US cohort, pregnancy complications were associated with higher mortality nearly 50 years later. Higher incidence of some complications in Black individuals and differential associations with mortality risk suggest that disparities in pregnancy health may have life-long implications for earlier mortality.

Preconception hemoglobin A1c concentration in healthy women is not associated with fecundability or pregnancy loss.

Objective: To examine the relationship of preconception hemoglobin A1c, a marker of cumulative exposure to glucose over the preceding 2-3 months, with time to pregnancy, pregnancy loss, and live birth among fecund women without diagnosed diabetes or other medical diseases. Design: A secondary analysis of a prospective cohort of women participating in the Effects of Aspirin in Gestation and Reproduction (EAGeR) trial. Setting: Four US academic medical centers. Patients: A total of 1,194 healthy women aged 18-40 years with a history of one or two pregnancy losses attempting spontaneous conception were observed for up to six cycles while attempting pregnancy and throughout pregnancy if they conceived. Interventions: Not applicable. Main Outcome Measures: Time to pregnancy, human chorionic gonadotropin pregnancy, clinical pregnancy, pregnancy loss, and live birth. Results: Although increasing preconception A1c level was associated with reduced fecundability (fecundability odds ratio [FOR] per unit increase in A1c 0.74; 95% confidence interval [CI] 0.57, 0.96) in unadjusted models and models adjusted for age, race, smoking and treatment arm (FOR 0.79; 95% CI 0.60, 1.04), results were attenuated after further adjustment for body mass index (FOR 0.91; 95% CI 0.68, 1.21). Preconception A1c levels among women without diagnosed diabetes were not associated with live birth or pregnancy loss. Conclusionss: Among healthy women without diagnosed diabetes, we observed no association of A1c with live birth or pregnancy loss. The association between A1c and fecundability was influenced by body mass index, a strong risk factor for both diabetes and infertility. These data support current recommendations that preconception A1c screening should be reserved for patients with risk factors for diabetes. Clinical Trial Registration Number: ClinicalTrials.gov: NCT00467363.

Circulating Vascular Endothelial Growth Factor and Soluble fms-Like Tyrosine Kinase-1 as Biomarkers for Endometrial Remodeling Across the Menstrual Cycle.

OBJECTIVE: To characterize variation in circulating vascular endothelial growth factor (VEGF) and its receptor, soluble fms-like tyrosine kinase-1 (sFLT-1), across the menstrual cycle in normal ovulating women in relation to reproductive hormones to identify the utility of VEGF and sFLT-1 as peripheral biomarkers of endometrial remodeling. METHODS: Ninety-six healthy, regularly menstruating ovulatory women, aged 18-44 years, enrolled in the BioCycle Study, a prospective cohort study at a U.S. academic research center. Vascular endothelial growth factor and sFLT-1 were measured in concurrently collected plasma, serum, and urine up to eight times across a single cycle. Reproductive hormones were measured in serum. Mean concentrations of VEGF and sFLT-1 were compared across phases of the cycle, and correlations between specimen types were calculated. Harmonic models estimated associations between VEGF and sFLT-1 and characteristics of hormonal patterns. RESULTS: No variation in VEGF or sFLT-1 levels were detected over the menstrual cycle. Median (25th percentile, 75th percentile) concentrations of VEGF during the menstrual cycle were 31.2 pg/mL (24.1, 56.9) in plasma, 194.1 pg/mL (125.4, 350.2) in serum, and 101.7 pg/mL (64.2, 165.8) in urine. Plasma and serum measures were consistently correlated, whereas urinary measures were not. Vascular endothelial growth factor was not consistently associated with reproductive hormone concentrations, although sFLT-1 was associated with higher mean and amplitude of estradiol. CONCLUSION: Circulating VEGF and sFLT-1 did not vary across the menstrual cycle and therefore are unlikely to be useful peripheral biomarkers of endometrial changes across the menstrual cycle. For studies measuring circulating VEGF for other reasons, plasma may be the preferred medium and timing to menstrual cycle phase need not be considered for reproductive-age women.

Is opioid use safe in women trying to conceive?

Background: Opioids are commonly prescribed to women of reproductive age, including after delivery and miscarriage. However, to our knowledge, opioid use has not been frequently studied in relation to the common reproductive complications of impaired fecundability and pregnancy. We examined the association of opioid use during the critical window of pregnancy establishment with fecundability and pregnancy loss. Methods: We measured opioid use by urine screening and self-report at multiple time points during preconception and early pregnancy in a prospective cohort of women attempting conception (n=1228). The main outcomes included time to hCG-detected pregnancy and incidence of live birth and pregnancy loss. We estimated fecundability odds ratios (FOR) and risk ratios (RR) with 95% confidence intervals (CI) adjusting for sociodemographic characteristics, reproductive characteristics, and use of antidepressants, tobacco, alcohol, and marijuana. Results: Prevalence of preconception opioid use was 18% (n=226 of 1228), and in early pregnancy was 5% (n=33 of 685). Opioid use while attempting pregnancy was associated with reduced fecundability (FOR: 0.71; 95% CI: 0.50, 1.0). Risk of pregnancy loss increased as opioid exposure was detected later in gestation, from the beginning of the cycle of conception (RR: 1.5; 95% CI 0.85, 2.6), to week 4 of pregnancy (RR: 2.1; 95% CI: 1.1, 4.1), and to week 4 and 8 of pregnancy (RR: 2.5; 95% CI: 1.3, 5.0). Conclusions: Our results are consistent with the hypothesis that opioid exposure while trying to conceive may be harmful, even among healthy, non-opioid-dependent women. Possible risks to fecundability and pregnancy viability are relevant to patients and providers when evaluating pain management approaches.ClinicalTrials.gov registration number: #NCT00467363.

Recalled maternal lifestyle behaviors associated with anti-müllerian hormone of adult female offspring.

Anti-müllerian hormone (AMH) is an established marker of ovarian reserve that decreases with age. Though the pool of ovarian follicles is established during fetal development, impacts of in utero exposures on AMH are uncertain. Thus, we sought to evaluate associations of in utero exposures with AMH of adult daughters with a prospective cohort study of adult daughters at university medical centers. Women noted their mother's reported use of diethylstilbestrol (DES), vitamins, tobacco, alcohol, and caffeine during pregnancy, and their mother's occupation during pregnancy. All participants were reproductive age women (18-40 years) enrolled in the Effects of Aspirin in Gestation and Reproduction (EAGeR) trial. Serum AMH concentrations were measured at baseline prior to conception and categorized using clinical guidelines. Multinomial regression models estimated associations between each exposure and high (>3.5 ng/mL) and low (<1.0 ng/mL) versus normal AMH (1.0-3.5 ng/mL), adjusting for participant's age, mother's age, mother's history of fertility treatment, and mother's use of vitamins. In 1202 women with available data, maternal caffeine use was associated with an increased risk of low AMH, compared to normal (relative risk [RR] 1.90, 95 % confidence interval [CI] 1.09, 3.30). Vitamins were associated with an increased risk of high AMH compared to normal (RR 1.93, 95 % CI 1.24, 3.00). Other exposures were not associated with AMH concentrations in offspring. Maternal caffeine and vitamin use during pregnancy may be associated with ovarian reserve in adult offspring, highlighting the potential importance of pregnancy lifestyle on the reproductive health of daughters.

Platelet activation and placenta-mediated adverse pregnancy outcomes: an ancillary study to the Effects of Aspirin in Gestation and Reproduction trial.

BACKGROUND: Platelet activation may play a role in the pathophysiology of placenta-mediated obstetrical complications, as evidenced by the efficacy of aspirin in preventing preeclampsia, but published data regarding the relationship between biomarkers for platelet activation and adverse obstetrical outcomes are sparse. In particular, it is unknown whether prepregnancy biomarkers of platelet activation are associated with adverse pregnancy outcomes. OBJECTIVE: This study aimed to determine the following: (1) whether maternal plasma concentrations of platelet factor 4 are associated with risk of placenta-mediated adverse obstetrical outcomes, and (2) whether these associations are modified by low-dose aspirin. STUDY DESIGN: This ancillary study included measurement of platelet factor 4 among 1185 of 1228 women of reproductive age enrolled in the Effects of Aspirin in Gestation and Reproduction trial with available plasma samples, with relevant outcomes assessed among 584 women with pregnancies lasting at least 20 weeks' gestation. We measured platelet factor 4 in plasma samples obtained at the prepregnancy study visit (before randomization to low-dose aspirin or placebo), 12 weeks' gestation, and 28 weeks' gestation. The primary outcome was a composite of hypertensive disorders of pregnancy, placental abruption, and small-for-gestational-age infant. We estimated the relative risks (RRs) and 95% confidence intervals (CIs) for the association between platelet factor 4 and the composite and individual outcomes at each time point using log-binomial regression that was weighted to account for potential selection bias and adjusted for age, body mass index, education, income, and smoking. To evaluate the potential effect modification of aspirin, we stratified the analyses by aspirin treatment assignment. RESULTS: During follow-up, 95 women experienced the composite adverse obstetrical outcome, with 57 cases of hypertensive disorders of pregnancy, 35 of small for gestational age, and 6 of placental abruption. Overall, prepregnancy platelet factor 4 was positively associated with the composite outcome (third tertile vs first tertile; relative risk, 2.36; 95% confidence interval, 1.38-4.03) and with hypertensive disorders of pregnancy (third tertile vs first tertile; relative risk, 2.14; 95% confidence interval, 1.08-4.23). In analyses stratified by treatment group, associations were stronger in the placebo group (third tertile vs first tertile; relative risk, 3.36; 95% confidence interval, 1.42-7.93) than in the aspirin group (third tertile vs first tertile; relative risk, 1.78; 95% confidence interval, 0.90-3.50). CONCLUSION: High concentrations of platelet factor 4 before pregnancy are associated with increased risk of placenta-mediated adverse pregnancy outcomes, particularly for hypertensive disorders of pregnancy. Aspirin may mitigate the increased risk of these outcomes among women with higher plasma concentrations of preconception platelet factor 4, but low-dose aspirin nonresponders may require higher doses of aspirin or alternate therapies to achieve obstetrical risk reduction.

Associations Between Preconception Plasma Fatty Acids and Pregnancy Outcomes.

BACKGROUND: Although fatty acids are involved in critical reproductive processes, the relationship between specific fatty acids and fertility is uncertain. We investigated the relationship between preconception plasma fatty acids and pregnancy outcomes. METHODS: We included 1,228 women attempting pregnancy with one to two previous pregnancy losses from the EAGeR trial (2007-2011). Plasma fatty acids were measured at baseline. We used log-binomial regression to assess associations between fatty acids and pregnancy, pregnancy loss, and live birth, adjusting for age, race, smoking, BMI, physical activity, income, parity, treatment arm, and cholesterol. RESULTS: Although total saturated fatty acids (SFAs) were not associated with pregnancy outcomes, 14:0 (myristic acid; relative risk [RR] = 1.10, 95% confidence interval [CI] = 1.02, 1.19, per 0.1% increase) and 20:0 (arachidic acid; RR = 1.05, 95% CI = 1.01, 1.08, per 0.1% increase) were positively associated with live birth. Findings suggested a positive association between total monounsaturated fatty acids (MUFAs) and pregnancy and live birth and an inverse association with loss. Total polyunsaturated fatty acids (PUFAs) were associated with lower probability of pregnancy (RR = 0.97, 95% CI = 0.95, 1.00) and live birth (RR = 0.96, 95% CI = 0.94, 0.99), and increased risk of loss (RR = 1.10, 95% CI = 1.00, 1.20), per 1% increase. Trans fatty acids and n-3 fatty acids were not associated with pregnancy outcomes. CONCLUSIONS: Preconception total plasma MUFAs were positively associated with pregnancy and live birth. PUFAs were inversely associated with pregnancy outcomes. Specific SFAs were associated with a higher probability of live birth. Our results suggest that fatty acids may influence pregnancy outcomes.

Preconception Perceived Stress Is Associated with Reproductive Hormone Levels and Longer Time to Pregnancy.

BACKGROUND: Women who experience pregnancy loss are especially prone to high stress, though the effects of stress on reproductive outcomes in this vulnerable population are unknown. We assessed relationships between perceived stress and hormones, anovulation, and fecundability among women with prior loss. METHODS: One thousand two hundred fourteen women with 1-2 prior losses were followed for ≤6 cycles while attempting pregnancy and completed end-of-cycle stress assessments. For cycles 1 and 2, women also collected daily urine and completed daily perceived stress assessments. We assessed anovulation via. an algorithm based on human chorionic gonadotropin (hCG), pregnanediol-3-glucuronide (PdG), luteinizing hormone (LH), and fertility monitor readings. Pregnancy was determined via. hCG. Adjusted weighted linear mixed models estimated the effect of prospective phase-varying (menses, follicular, periovulatory, and luteal) perceived stress quartiles on estrone-1-glucuronide (E1G), PdG, and LH concentrations. Marginal structural models accounted for time-varying confounding by hormones and lifestyle factors affected by prior stress. Poisson and Cox regression estimated risk ratios and fecundability odds ratios of cycle-varying stress quartiles on anovulation and fecundability. Models were adjusted for age, race, body mass index (BMI), parity, and time-varying caffeine, alcohol, smoking, intercourse, and pelvic pain. RESULTS: Women in the highest versus lowest stress quartile had lower E1G and PdG concentrations, a marginally higher risk of anovulation [1.28; 95% confidence interval (CI) = 1.00, 1.63], and lower fecundability (0.71; 95% CI = 0.55, 0.90). CONCLUSION: Preconception perceived stress appears to adversely affect sex steroid synthesis and time to pregnancy. Mechanisms likely include the effects of stress on ovulatory function, but additional mechanisms, potentially during implantation, may also exist.

Exposure to bisphenol A, chlorophenols, benzophenones, and parabens in relation to reproductive hormones in healthy women: A chemical mixture approach.

BACKGROUND: Little is known about the associations of bisphenol A, chlorophenols, benzophenones, and parabens with reproductive hormone levels in women. Our goal was to evaluate the associations between repeated measures of these chemicals and their mixtures with reproductive hormones in women. METHODS: Longitudinal urine samples from healthy, premenopausal women (n = 143 with 3-5 urine samples each) were measured for bisphenol A, five chlorophenols (2,4-dichlorophenol (2,4-DCP), 2,5-dichlorophenol, 2,4,5-trichlorophenol, 2,4,6-trichlorophenol, triclosan), two ultraviolet (UV) filters (benzophenone-1, benzophenone-3), and eight parabens and their metabolites (benzyl, butyl, ethyl, heptyl, methyl, propyl, 4-hydroxybenzoic acid (4-HB), 3,4-dihydroxybenzoic acid (3,4-DHB)) over two menstrual cycles. Estradiol, progesterone, luteinizing hormone (LH), and follicle stimulating hormone (FSH) were measured in blood up to 8 times each menstrual cycle. Linear mixed models were used for both single and multi-chemical exposures estimated using principal component analysis. Four factors were identified including: paraben; paraben metabolites and BPA, phenols, and UV filters. Models were adjusted for creatinine, age, race, and body mass index and weighted with inverse probability of exposure weights to account for time varying confounding. RESULTS: In single-chemical models, 3,4-DHB was associated with estradiol (0.06 (95% confidence interval (CI): 0.001, 0.12)), 2-4-DCP with increased progesterone 0.14 (0.06, 0.21) and decreased FSH -0.08 (-0.11, -0.04), and 4-HB was associated with increased FSH 0.07 (0.01, 0.13). In multi-chemical models, all factors were associated with increased progesterone (beta coefficient range: 0.15 for UV filter factor to 0.32 for paraben factor). The paraben factor and the paraben metabolite and BPA factor were associated with increased estradiol [0.21 (0.15, 0.28); 0.12 (0.07, 0.18)]. The phenol and UV filter factors were associated with decreased estradiol, FSH, and LH. The UV filter factor showed the strongest inverse association with estradiol -0.16 (-0.22, -0.10), FSH -0.12 (-0.17, -0.07), and LH -0.17 (-0.23, -0.10). CONCLUSION: Mixtures of phenols were associated with changes in reproductive hormones. Such changes could contribute to adverse health in women but additional research is necessary.

Shorter Time to Pregnancy With Increasing Preconception Carotene Concentrations Among Women With 1-2 Previous Pregnancy Losses.

Although maternal nutrition may affect fecundity, associations between preconception micronutrient levels and time to pregnancy (TTP) have not been examined. We assessed the relationship between preconception fat-soluble micronutrient concentrations and TTP among women with 1-2 prior pregnancy losses. This was a prospective cohort study of 1,228 women set within the Effects of Aspirin in Gestation and Reproduction (EAGeR) Trial (United States, 2007-2011), which assessed the association of preconception-initiated daily low-dose aspirin with reproductive outcomes. We measured preconception levels of zeaxanthin, cryptoxanthin, lycopene, α- and ß-carotene, and α- and γ-tocopherol in serum. We used discrete Cox regression models, accounting for left-truncation and right-censoring, to calculate fecundability odds ratios and 95% confidence intervals. The models adjusted for age, body mass index, race, smoking, alcohol, physical activity, income, vitamin use, cholesterol, treatment arm, and study site. Serum α-carotene levels (per log unit (µg/dL) increase, fecundability odds ratio (FOR) = 1.17, 95% confidence interval (CI): 1.00, 1.36) and serum α-carotene concentrations at or above the US average (2.92 µg/dL) versus below the average (FOR = 1.21, 95% CI: 1.02, 1.44) were associated with shorter TTP. Compared with levels below the US average (187 µg/dL), γ-tocopherol concentrations at or above the average were associated with longer TTP (FOR = 0.83, 95% CI: 0.69, 1.00). The potential for these nutrients to influence fecundability deserves further exploration.

Preconception Blood Pressure Levels and Reproductive Outcomes in a Prospective Cohort of Women Attempting Pregnancy.

Elevated blood pressure in young adulthood is an early risk marker for cardiovascular disease. Despite a strong biological rationale, little research has evaluated whether incremental increases in preconception blood pressure have early consequences for reproductive health. We evaluated preconception blood pressure and fecundability, pregnancy loss, and live birth in the EAGeR trial (Effects of Aspirin on Gestational and Reproduction; 2007-2011), a randomized clinical trial of aspirin and reproductive outcomes among 1228 women attempting pregnancy with a history of pregnancy loss. Systolic and diastolic blood pressure were measured during preconception in the first observed menstrual cycle and in early pregnancy and used to derive mean arterial pressure. Fecundability was assessed as number of menstrual cycles until pregnancy, determined through human chorionic gonadotropin testing. Pregnancy loss included both human chorionic gonadotropin-detected and clinical losses. Analyses adjusted for treatment assignment, age, body mass index, race, marital status, smoking, parity, and time since last loss. Mean preconception systolic and diastolic blood pressure were 111.6 mm Hg (SD, 12.1) and 72.5 (SD, 9.4) mm Hg. Risk of pregnancy loss increased 18% per 10 mm Hg increase in diastolic blood pressure (95% confidence interval, 1.03-1.36) and 17% per 10 mm Hg increase in mean arterial pressure (95% confidence interval, 1.02-1.35) in adjusted analyses. Findings were similar for early pregnancy blood pressure. Preconception blood pressure was not related to fecundability or live birth in adjusted analyses. Findings suggest that preconception blood pressure among healthy women is associated with pregnancy loss, and lifestyle interventions targeting blood pressure among young women may favorably impact reproductive health. CLINICAL TRIAL REGISTRATION: URL: http://www.clinicaltrials.gov. Unique identifier: NCT00467363.

Vitamin D is associated with bioavailability of androgens in eumenorrheic women with prior pregnancy loss.

BACKGROUND: Prior studies have reported mixed results regarding relationships between vitamin D, androgens, and sex hormone-binding globulin in patients with polycystic ovary syndrome. However, less is known regarding these associations in eumenorrheic, premenopausal women. OBJECTIVE: Our objective was to study the relationships between serum vitamin D and androgen biomarkers in eumenorrheic women with a history of pregnancy loss who were attempting pregnancy. STUDY DESIGN: This was an analysis of a cohort of 1191 participants from the Effects of Aspirin in Gestation and Reproduction trial (2006-2012). Participants were attempting to conceive, aged 18-40 years, with 1-2 documented prior pregnancy losses and no history of infertility, and recruited from 4 academic medical centers in the United States. Serum vitamin D (25-hydroxyvitamin D) and hormone concentrations were measured at baseline. RESULTS: Vitamin D concentration was negatively associated with free androgen index (percentage change [95% confidence interval, -5% (-8% to -2%)] per 10 ng/mL increase) and positively associated with sex hormone-binding globulin (95% confidence interval, 4% [2-7%]), although not with total testosterone, free testosterone, or dehydroepiandrosterone sulfate after adjusting for age, body mass index, smoking status, race, income, education, physical activity, and season of blood draw. CONCLUSION: Overall, vitamin D was associated with sex hormone-binding globulin and free androgen index in eumenorrheic women with prior pregnancy loss, suggesting that vitamin D may play a role in the bioavailability of androgens in eumenorrheic women. We are limited in making assessments regarding directionality, given the cross-sectional nature of our study.

Multiple Imputation for Incomplete Data in Epidemiologic Studies.

Epidemiologic studies are frequently susceptible to missing information. Omitting observations with missing variables remains a common strategy in epidemiologic studies, yet this simple approach can often severely bias parameter estimates of interest if the values are not missing completely at random. Even when missingness is completely random, complete-case analysis can reduce the efficiency of estimated parameters, because large amounts of available data are simply tossed out with the incomplete observations. Alternative methods for mitigating the influence of missing information, such as multiple imputation, are becoming an increasing popular strategy in order to retain all available information, reduce potential bias, and improve efficiency in parameter estimation. In this paper, we describe the theoretical underpinnings of multiple imputation, and we illustrate application of this method as part of a collaborative challenge to assess the performance of various techniques for dealing with missing data (Am J Epidemiol. 2018;187(3):568-575). We detail the steps necessary to perform multiple imputation on a subset of data from the Collaborative Perinatal Project (1959-1974), where the goal is to estimate the odds of spontaneous abortion associated with smoking during pregnancy.

Inverse-Probability-Weighted Estimation for Monotone and Nonmonotone Missing Data.

Missing data is a common occurrence in epidemiologic research. In this paper, 3 data sets with induced missing values from the Collaborative Perinatal Project, a multisite US study conducted from 1959 to 1974, are provided as examples of prototypical epidemiologic studies with missing data. Our goal was to estimate the association of maternal smoking behavior with spontaneous abortion while adjusting for numerous confounders. At the same time, we did not necessarily wish to evaluate the joint distribution among potentially unobserved covariates, which is seldom the subject of substantive scientific interest. The inverse probability weighting (IPW) approach preserves the semiparametric structure of the underlying model of substantive interest and clearly separates the model of substantive interest from the model used to account for the missing data. However, IPW often will not result in valid inference if the missing-data pattern is nonmonotone, even if the data are missing at random. We describe a recently proposed approach to modeling nonmonotone missing-data mechanisms under missingness at random to use in constructing the weights in IPW complete-case estimation, and we illustrate the approach using 3 data sets described in a companion article (Am J Epidemiol. 2018;187(3):568-575).

Principled Approaches to Missing Data in Epidemiologic Studies.

Principled methods with which to appropriately analyze missing data have long existed; however, broad implementation of these methods remains challenging. In this and 2 companion papers (Am J Epidemiol. 2018;187(3):576-584 and Am J Epidemiol. 2018;187(3):585-591), we discuss issues pertaining to missing data in the epidemiologic literature. We provide details regarding missing-data mechanisms and nomenclature and encourage the conduct of principled analyses through a detailed comparison of multiple imputation and inverse probability weighting. Data from the Collaborative Perinatal Project, a multisite US study conducted from 1959 to 1974, are used to create a masked data-analytical challenge with missing data induced by known mechanisms. We illustrate the deleterious effects of missing data with naive methods and show how principled methods can sometimes mitigate such effects. For example, when data were missing at random, naive methods showed a spurious protective effect of smoking on the risk of spontaneous abortion (odds ratio (OR) = 0.43, 95% confidence interval (CI): 0.19, 0.93), while implementation of principled methods multiple imputation (OR = 1.30, 95% CI: 0.95, 1.77) or augmented inverse probability weighting (OR = 1.40, 95% CI: 1.00, 1.97) provided estimates closer to the "true" full-data effect (OR = 1.31, 95% CI: 1.05, 1.64). We call for greater acknowledgement of and attention to missing data and for the broad use of principled missing-data methods in epidemiologic research.

Blood lead, cadmium and mercury in relation to homocysteine and C-reactive protein in women of reproductive age: a panel study.

BACKGROUND: To examine the relationship between cadmium, lead, and mercury concentrations with high-sensitivity C-reactive protein (hs-CRP) and homocysteine in women. METHODS: Metals were measured at enrollment in whole blood. Homocysteine and hs-CRP were measured in one (N = 9) or two (N = 250) menstrual cycles up to 3 and 8 times per cycle, respectively. Linear mixed models with inverse probability of exposure weights to account for time varying confounding were used and models were stratified by dietary and serum vitamin status (dietary: vitamin B6, B12, folate; serum: folate). RESULTS: Geometric mean (95% confidence interval (CI)) concentrations for cadmium, lead, and mercury were 0.29 (0.26-0.31) µg/L, 0.91 (0.86-0.96) µg/dL, and 1.05 (0.93-1.18) µg/L, respectively. Lead was associated with increased homocysteine (0.08; 95% CI: 0.01, 0.15) and this persisted among those in the lower three quartiles of consumption of vitamin B6, B12, folate, and serum folate but was not significant among those in the upper quartile. No associations were observed between metals and hs-CRP. CONCLUSIONS: Blood lead was associated with increased homocysteine in a cohort of healthy, premenopausal women but these associations did not persist among those consuming ≥75th percentile of essential micronutrients. Cadmium, lead, and mercury were not associated with hs-CRP concentrations.

Collinearity and Causal Diagrams: A Lesson on the Importance of Model Specification.

BACKGROUND: Correlated data are ubiquitous in epidemiologic research, particularly in nutritional and environmental epidemiology where mixtures of factors are often studied. Our objectives are to demonstrate how highly correlated data arise in epidemiologic research and provide guidance, using a directed acyclic graph approach, on how to proceed analytically when faced with highly correlated data. METHODS: We identified three fundamental structural scenarios in which high correlation between a given variable and the exposure can arise: intermediates, confounders, and colliders. For each of these scenarios, we evaluated the consequences of increasing correlation between the given variable and the exposure on the bias and variance for the total effect of the exposure on the outcome using unadjusted and adjusted models. We derived closed-form solutions for continuous outcomes using linear regression and empirically present our findings for binary outcomes using logistic regression. RESULTS: For models properly specified, total effect estimates remained unbiased even when there was almost perfect correlation between the exposure and a given intermediate, confounder, or collider. In general, as the correlation increased, the variance of the parameter estimate for the exposure in the adjusted models increased, while in the unadjusted models, the variance increased to a lesser extent or decreased. CONCLUSION: Our findings highlight the importance of considering the causal framework under study when specifying regression models. Strategies that do not take into consideration the causal structure may lead to biased effect estimation for the original question of interest, even under high correlation.

Dietary fat intake and reproductive hormone concentrations and ovulation in regularly menstruating women.

BACKGROUND: Emerging evidence suggests potential links between some dietary fatty acids and improved fertility, because specific fatty acids may affect prostaglandin synthesis and steroidogenesis. OBJECTIVE: The objective of this exploratory study was to evaluate associations between total and specific types of dietary fat intake and 1) hormone concentrations and 2) the risk of sporadic anovulation in a cohort of 259 regularly menstruating women in the BioCycle Study. DESIGN: Endogenous reproductive hormones were measured up to 8 times/cycle for up to 2 cycles, with visits scheduled with the use of fertility monitors. Dietary intake was assessed with up to four 24-h recalls/cycle. Linear mixed models and generalized linear models were used to evaluate the associations between dietary fatty acids and both reproductive hormone concentrations and ovulatory status. All models were adjusted for total energy intake, age, body mass index, and race. RESULTS: Relative to the lowest levels of percentage of energy from total fat, the highest tertile was associated with increased total and free testosterone concentrations (total: percentage change of 4.0%; 95% CI: 0.7%, 7.3%; free: percentage change of 4.1%; 95% CI: 0.5%, 7.7%). In particular, the percentage of energy from polyunsaturated fatty acids (PUFAs) in the highest tertile was associated with increases in total and free testosterone (total: percentage change of 3.7%; 95% CI: 0.6%, 6.8%; free: percentage change of 4.0%; 95% CI: 0.5%, 7.5%). The PUFA docosapentaenoic acid (22:5n-3) was not significantly associated with testosterone concentrations (P-trend = 0.86 in energy substitution models) but was associated with increased progesterone and a reduced risk of anovulation (highest tertile compared with the lowest tertile: RR: 0.42; 95% CI: 0.18, 0.95). Fat intakes were not associated with other reproductive hormone concentrations. CONCLUSIONS: These results indicate that total fat intake, and PUFA intake in particular, is associated with very small increases in testosterone concentrations in healthy women and that increased docosapentaenoic acid was associated with a lower risk of anovulation.

Serum Antioxidants Are Associated with Serum Reproductive Hormones and Ovulation among Healthy Women.

BACKGROUND: Evidence is growing that the equilibrium between reactive oxygen species and antioxidants plays a vital role in women's reproductive health. OBJECTIVE: The objective of this study was to evaluate variations in serum antioxidant concentrations across the menstrual cycle and associations between antioxidants and reproductive hormones and anovulation among healthy women. METHODS: The BioCycle Study, a prospective cohort, followed 259 women aged 18-44 y for up to 2 menstrual cycles. Serum fat-soluble vitamin and micronutrient (α-tocopherol, γ-tocopherol, retinol, lutein, lycopene, and ß-carotene), ascorbic acid, and reproductive hormone concentrations were measured 5-8 times/cycle. We used weighted linear mixed models to assess associations between antioxidants and hormone concentrations, after adjustment for age, race, body mass index, parity, sleep, pain medication use, total energy intake, concurrent hormones, serum cholesterol, F2-isoprostanes, and other antioxidants. Generalized linear models were used to identify associations with anovulation. RESULTS: Serum antioxidant concentrations varied across the menstrual cycle. Retinol and α-tocopherol were associated with higher estradiol [RR: 1.00 pg/mL (95% CI: 0.67, 1.34 pg/mL); RR: 0.02 pg/mL (95% CI: 0.003, 0.03 pg/mL), respectively] and testosterone [RR: 0.61 ng/dL (95% CI: 0.44, 0.78 ng/dL); RR: 0.01 ng/dL (95% CI: 0.001, 0.01 ng/dL), respectively]. Ascorbic acid was associated with higher progesterone (RR: 0.15 ng/mL; 95% CI: 0.05, 0.25 ng/mL) and with lower follicle-stimulating hormone (RR: -0.06 mIU/mL; 95% CI: -0.09, -0.03 mIU/mL). The ratio of α- to γ-tocopherol was associated with an increased risk of anovulation (RR: 1.03; 95% CI: 1.01, 1.06). CONCLUSIONS: These findings shed new light on the intricate associations between serum antioxidants and endogenous hormones in healthy premenopausal women and support the hypothesis that concentrations of serum vitamins affect steroidogenesis even after adjustment for oxidative stress.