Alexander C. McLain

Methodology for establishing a population-based birth cohort focusing on couple fertility and children's development, the Upstate KIDS Study.

BACKGROUND Critical data gaps remain regarding infertility treatment and child development. We assessed the utility of a birth certificate registry for developing a population cohort aimed at answering such questions. METHODS We utilised the Upstate New York livebirth registry (n = 201,063) to select births conceived with (n = 4024) infertility treatment or exposed infants, who were then frequency-matched by residence to a random sample of infants conceived without (n = 14,455) treatment or unexposed infants, 2008-10. Mothers were recruited at 2-4 months postpartum and queried about their reproductive histories, including infertility treatment for comparison with birth certificate data. Overall, 1297 (32%) mothers of exposed and 3692 of unexposed (26%) infants enrolled. RESULTS Twins represented 22% of each infant group. The percentage of infants conceived with/without infertility treatment was similar whether derived from the birth registry or maternal report: 71% none, 16% drugs or intrauterine insemination, and 14% assisted reproductive technologies (ART). Concordant reporting between the two data sources was 93% for no treatment, 88% for ART, and 83% for fertility drugs, but differed by plurality. Exposed infants had slightly (P < 0.01) earlier gestations than unexposed infants (38.3 ± 2.8 and 38.7 ± 2.7 weeks, respectively) based upon birth certificates but not maternal report (38.7 ± 2.7 and 38.7 ± 2.9, respectively). Conversely, mean birthweight was comparable using birth certificates (3157 ± 704 and 3194 ± 679 g, respectively), but differed using maternal report (3167 ± 692 and 3224 ± 661, respectively P < 0.05). CONCLUSIONS The birth certificate registry is a suitable sampling framework as measured by concordance with maternally reported infertility treatment. Future efforts should address the impact of factors associated with discordant reporting on research findings.

Newborn Adipokines and Birth Outcomes

BACKGROUND Adipokines can serve as a measure of adipose tissue activity. Although birthweight correlates with neonatal adiposity, findings for cord blood levels of adipokines and birth outcomes have been conflicted. Therefore, we determined the cross-sectional associations between adipokines measured in newborn dried blood spots (DBS) and birth outcomes. METHODS The Upstate KIDS study enrolled mothers and infants from 2008 to 2010. Among infants whose parents consented to the use of residual DBS from newborn screening, 2397 singletons and 1240 twins had adipokine measurements from the Human Obesity Panel (R&D Systems) by Luminex. Odds ratios were estimated by multivariable logistic regression for risk of birth outcomes of preterm delivery (<37 weeks for singletons, <32 for twins) and small-for-gestational age (SGA <10th for singletons and <3rd for twins age- and sex-specific percentiles) by adipokine quintiles. Generalised estimating equations were applied to account for correlations between twins. RESULTS Singletons in the lowest compared with the highest quintile of adiponectin were more likely preterm (adjusted odds ratio 3.26; 95% confidence interval [CI] 1.99, 5.34) and SGA (1.81; [95% CI 1.18, 2.77]). Similar associations were observed among twins. Resistin was associated with preterm birth (Q1 vs. Q5: 2.08; [95% CI 1.20, 3.62]) only among singletons. Adipsin had inconsistent associations after adjustment. CONCLUSIONS This large population-based study demonstrates that newborn DBS-measured adipokines are associated with birth outcomes, particularly preterm birth and SGA among those with lower adiponectin levels regardless of plurality.

Multiple Testing of Composite Null Hypotheses in Heteroscedastic Models

In large-scale studies, the true effect sizes often range continuously from zero to small to large, and are observed with heteroscedastic errors. In practical situations where the failure to reject small deviations from the null is inconsequential, specifying an indifference region (or forming composite null hypotheses) can greatly reduce the number of unimportant discoveries in multiple testing. The heteroscedasticity issue poses new challenges for multiple testing with composite nulls. In particular, the conventional framework in multiple testing, which involves rescaling or standardization, is likely to distort the scientific question. We propose the concept of a composite null distribution for heteroscedastic models and develop an optimal testing procedure that minimizes the false nondiscovery rate, subject to a constraint on the false discovery rate. The proposed approach is different from conventional methods in that the effect size, statistical significance, and multiplicity issues are addressed integrally. The external information of heteroscedastic errors is incorporated for optimal simultaneous inference. The new features and advantages of our approach are demonstrated using both simulated and real data. The numerical studies demonstrate that our new procedure enjoys superior performance with greater accuracy and better interpretability of results.

Persistent organochlorine pollutants and menstrual cycle characteristics

An evolving body of evidence suggests an adverse relation between persistent organochlorine pollutants (POPs) and menstruation, though prospective longitudinal measurement of menses is limited and served as the impetus for study. We prospectively assessed the relation between a mixture of persistent organochlorine compounds and menstrual cycle length and duration of bleeding in a cohort of women attempting to become pregnant. Eighty-three (83%) women contributing 447 cycles for analysis provided a blood specimen for the quantification of 76 polychlorinated biphenyls and seven organochlorine pesticides, and completed daily diaries on menstruation until a human chorionic gonadotropin confirmed pregnancy or 12 menstrual cycles without conception. Gas chromatography with electron capture detection was used to quantify concentrations (ng g(-1)serum); enzymatic methods were used to quantify serum lipids (mg dL(-1)). A linear regression model with a mixture distribution was used to identify chemicals grouped by purported biologic activity that significantly affected menstrual cycle length and duration of bleeding adjusting for age at menarche and enrollment, body mass index, and cigarette smoking. A significant 3-d increase in cycle length was observed for women in the highest tertile of estrogenic PCB congeners relative to the lowest tertile (β=3.20; 95% CI 0.36, 6.04). A significant reduction in bleeding (<1 d) was observed among women in the highest versus lowest tertile of aromatic fungicide exposure (γ=-0.15; 95% CI -0.29, -0.00). Select POPs were associated with changes in menstruation underscoring the importance of assessing chemical mixtures for female fecundity.

A Joint Mixed Effects Dispersion Model for Menstrual Cycle Length and Time-to-Pregnancy

Menstrual cycle patterns are often used as indicators of female fecundity and are associated with hormonally dependent diseases such as breast cancer. A question of considerable interest is in identifying menstrual cycle patterns, and their association with fecundity. A source of data for addressing this question is prospective pregnancy studies that collect detailed information on reproductive aged women. However, methodological challenges exist in ascertaining the association between these two processes as the number of longitudinally measured menstrual cycles is relatively small and informatively censored by time to pregnancy (TTP), as well as the cycle length distribution being highly skewed. We propose a joint modeling approach with a mixed effects dispersion model for the menstrual cycle lengths and a discrete survival model for TTP to address this question. This allows us to assess the effect of important characteristics of menstrual cycle that are associated with fecundity. We are also able to assess the effect of fecundity predictors such as age at menarche, age, and parity on both these processes. An advantage of the proposed approach is the prediction of the TTP, thus allowing us to study the efficacy of menstrual cycle characteristics in predicting fecundity. We analyze two prospective pregnancy studies to illustrate our proposed method by building a model based on the Oxford Conception Study, and predicting for the New York State Angler Cohort Prospective Pregnancy Study. Our analysis has relevant findings for assessing fecundity.

Preconception stress and the secondary sex ratio: a prospective cohort study

OBJECTIVE To study the association between salivary stress biomarkers and the secondary sex ratio. DESIGN Prospective, longitudinal cohort study. SETTING Community setting in the United Kingdom. PATIENT(S) On discontinuation of contraception for purposes of becoming pregnant, 338 women aged 18-40 years with complete data (90%) were followed until pregnant or up to six menstrual cycles. INTERVENTION(S) None. MAIN OUTCOME MEASURE(S) Secondary sex ratio. RESULT(S) Human chorionic gonadotropin pregnancies were detected in 207 (61%) women of whom 130 (63%) delivered singleton infants with available gender data. The adjusted odds ratio for a male birth was decreased for women in the highest quartile (AOR = 0.26; 95% confidence interval = 0.09, 0.74) of salivary cortisol relative to women in the lowest quartile during cycle 1. An inverse relation was observed between α-amylase and the 2° sex ratio, though not statistically significant. CONCLUSION(S) Our findings are consistent with a reversal in the 2° sex ratio with increasing preconception salivary cortisol concentrations. This relation suggests that activation of the hypothalamus-pituitary-adrenal axis may have implications in sex allocation and requires further study.

Maternal Lipid Change in Relation to Length of Gestation: A Prospective Cohort Study with Preconception Enrollment of Women

Background/Aims: We sought to investigate the association between preconception serum lipids and their daily rate of change in relation to length of gestation. Methods: In a cohort of 70 women, 61 (87%) became pregnant, resulting in 48 (69%) live births. Serum lipid measurements (in milligrams per deciliter) included total cholesterol, free cholesterol, triglycerides and phospholipids at preconception, upon human chorionic gonadotropin-confirmed pregnancy and following pregnancy loss (<14 weeks) or post partum. Pregnancy outcome (loss, preterm and term delivery) and gestational length were modeled relative to daily rate of change in lipids using multinomial regression and Cox proportional hazards models, respectively, adjusting for body mass index and smoking. Results: A rate of triglyceride change below the median was associated with an increased risk for pregnancy loss compared with term birth (adjusted odds ratio: 9.02; 95% CI: 1.62-50.30). A rate of triglyceride change of ≤0.01 mg/dl per day versus above the median was associated with a trend for increased risk of pregnancy loss or preterm (<37 weeks) birth (adjusted hazard ratio: 1.77; 95% CI: 0.94-3.33). Conclusion: A low rate of triglyceride change during early pregnancy may be a signal of risk of pregnancy loss or preterm birth. Lipids offer promise for identifying pregnancies at risk for adverse outcomes.

Use of assisted reproductive technology treatment as reported by mothers in comparison with registry data: the Upstate KIDS Study.

OBJECTIVE To assess the validity of maternally reported assisted reproductive technologies (ART) use and to identify predictors of reporting errors. DESIGN Linkage study. SETTING Not applicable. PATIENT(S) A total of 5,034 (27%) mothers enrolled, from whom 4,886 (97%) self-reported information about use of infertility treatment, including ART, for the index birth. INTERVENTION(S) None. MAIN OUTCOME MEASURE(S) Four measures of validity (sensitivity, specificity, positive and negative predictive values) and use of net reclassification improvement (NRI) methods to identify predictors associated with concordant/discordant maternal reporting. RESULT(S) The Upstate New York Infant Development Screening Program (Update KIDS Study) was linked with the Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS) using a defined algorithm for 2008-2010. The sensitivity, specificity, positive and negative predictive values were high (0.93, 0.99, 0.80, and 1.00, respectively). The validity of maternal report was high, reflecting few differences by participant characteristics except for maternal age dichotomized at 29 years as identified with NRI methods. CONCLUSION(S) Maternally reported ART is valid, with little variation across various characteristics. No strong predictors of discordant reporting were found, supporting the utility of population-based research with SART CORS linkage.

Successive time to pregnancy among women experiencing pregnancy loss

STUDY QUESTION Is time to pregnancy (TTP) similar across successive pregnancy attempts among women experiencing pregnancy loss? SUMMARY ANSWER TTP after a loss may be longer compared with TTP before a loss. WHAT IS KNOWN ALREADY Two pregnancy cohort studies have reported that TTP is similar across pregnancy attempts in fertile women. However, this has not been investigated among women experiencing pregnancy losses. STUDY DESIGN, SIZE, DURATION Data for this analysis come from the Longitudinal Investigation of Fertility and the Environment Study, a population-based, preconception cohort of couples attempting pregnancy. During 2005-2009, recruitment was targeted to 16 counties in Michigan and Texas with reported exposures to persistent environmental chemicals. A total of 501 couples were recruited and followed for up to 12 months of pregnancy attempts allowing for continued participation of women with pregnancy losses until censoring. PARTICIPANTS, SETTING, METHODS We assessed TTP among 70 couples recruited upon discontinuing contraception for purposes of becoming pregnant and experiencing ≥1 prospectively observed pregnancy losses during 12 months of trying. There were 61 couples who contributed two pregnancy attempts and 9 who contributed three. Women were instructed in the use of urine-based home fertility monitors to time intercourse relative to ovulation and recorded their bleeding patterns in daily journals. TTP was defined as the number of menstrual cycles taken to achieve pregnancy. Women were also instructed in the use of home digital pregnancy tests and asked to begin pregnancy testing on the day of expected menses. Women recorded the results of their pregnancy tests in a daily journal with a single positive pregnancy test result indicating an hCG-confirmed pregnancy. Pregnancy losses were ascertained from a subsequent recorded negative pregnancy test or clinically confirmed loss. We estimated fecundability odds ratios (FORs) comparing subsequent to first TTP using discrete Cox models with robust standard errors, accounting for cycles off contraception before study entry and adjusting for maternal age, body mass index, reproductive history and time-varying cigarette, alcohol and caffeine usage while trying. MAIN RESULTS AND THE ROLE OF CHANCE The mean female age was 30.3 ± 4.3 years; 21% had a prior pregnancy loss before study entry. Of the second and third attempts, 59 and 43%, respectively, were longer compared with the first attempt. FORs <1 suggest reduced fecundability or a longer TTP when comparing the second with the first attempt (0.42, 95% confidence interval (CI): 0.28, 0.65), and similarly for the third relative to the first attempt (0.64, 95% CI: 0.18, 2.36). TTP in the second attempt was a median of 1 cycle longer (interquartile range: 0, 3 cycles) compared with TTP in the first attempt. LIMITATIONS, REASONS FOR CAUTION As this is the first study to investigate successive TTP exclusively among women experiencing pregnancy loss, our findings await corroboration since most losses occurred early in gestation. As such, the generalizability of our findings for all pregnancy losses awaits further research. We also had limited power to detect a reduction in fecundability for the third compared with first pregnancy attempt. WIDER IMPLICATIONS OF THE FINDINGS Unlike fertile women, TTP in women experiencing early pregnancy losses may trend towards longer subsequent attempts. If the findings are corroborated, women experiencing losses may benefit from counselling regarding trying times. STUDY FUNDING/COMPETING INTERESTS This research was supported by the Intramural Research Program of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (contracts N01-HD-3-3355, N01-HD-3-3356 and NOH-HD-3-3358). K.J.S. was supported by an Intramural Research Training Award from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, Division of Intramural Population Health Research. The authors have no conflicts of interest to declare.

Nonparametric estimation of the conditional mean residual life function with censored data

The conditional mean residual life (MRL) function is the expected remaining lifetime of a system given survival past a particular time point and the values of a set of predictor variables. This function is a valuable tool in reliability and actuarial studies when the right tail of the distribution is of interest, and can be more informative than the survivor function. In this paper, we identify theoretical limitations of some semi-parametric conditional MRL models, and propose two nonparametric methods of estimating the conditional MRL function. Asymptotic properties such as consistency and normality of our proposed estimators are established. We investigate via simulation study the empirical properties of the proposed estimators, including bootstrap pointwise confidence intervals. Using Monte Carlo simulations we compare the proposed nonparametric estimators to two popular semi-parametric methods of analysis, for varying types of data. The proposed estimators are demonstrated on the Veteran’s Administration lung cancer trial.