Darryl J. Holman

Declining fecundity and ovarian ageing in natural fertility populations

Worldwide, human fertility declines with increasing maternal age, after contraceptive-use patterns and behavioral factors are taken into consideration. Here, we summarize some of our theoretical and empirical work examining the biological factors contributing to this age pattern of fertility. We undertook an 11 month prospective endocrinological study in a natural fertility (non-contracepting) population (rural Bangladesh) to estimate the contributions of fetal loss and fecundability (the probability of conception) to declining fecundity with age. Prospective interviews and urine samples for pregnancy tests were collected twice weekly from up to 700 women. These data were used to test mathematical models of the underlying biological processes contributing to changing fecundability and fetal loss risk with maternal age. The results indicate that much of the decline in fecundity can be attributed to an increasing risk of fetal loss with maternal age. Much of this fetal loss is due to chromosomal abnormalities--a result of ageing oocytes. Fecundability, on the other hand, does not begin to decline until the early 40s. We hypothesize that this is also a result of ageing at the ovarian level, namely follicular atresia, in the years just prior to menopause. The irregularity of menstrual cycles--longer cycles and increasingly variable hormonal patterns--at these ages may be a direct result of the small and rapidly dwindling remaining pool of follicles. We present a simple mathematical model of this process, and some preliminary laboratory results that support the model.

Urinary Estrone Conjugate and Pregnanediol 3-Glucuronide Enzyme Immunoassays for Population Research

BACKGROUND Monitoring of reproductive steroid hormones at the population level requires frequent measurements, hormones or metabolites that remain stable under less than ideal collection and storage conditions, a long-term supply of antibodies, and assays useful for a range of populations. We developed enzyme immunoassays for urinary pregnanediol 3-glucuronide (PDG) and estrone conjugates (E1Cs) that meet these criteria. METHODS Enzyme immunoassays based on monoclonal antibodies were evaluated for specificity, detection limit, parallelism, recovery, and imprecision. Paired urine and serum specimens were analyzed throughout menstrual cycles of 30 US women. Assay application in different populations was examined with 23 US and 42 Bangladeshi specimens. Metabolite stability in urine was evaluated for 0-8 days at room temperature and for 0-10 freeze-thaw cycles. RESULTS Recoveries were 108% for the PDG assay and 105% for the E1C assay. Serially diluted specimens exhibited parallelism with calibration curves in both assays. Inter- and intraassay CVs were <11%. Urinary and serum concentrations were highly correlated: r = 0.93 for E1C-estradiol; r = 0.98 for PDG-progesterone. All Bangladeshi and US specimens were above detection limits (PDG, 21 nmol/L; E1C, 0.27 nmol/L). Bangladeshi women had lower follicular phase PDG and lower luteal phase PDG and E1Cs than US women. Stability experiments showed a maximum decrease in concentration for each metabolite of <4% per day at room temperature and no significant decrease associated with number of freeze-thaw cycles. CONCLUSIONS These enzyme immunoassays can be used for the field conditions and population variation in hormone metabolite concentrations encountered in cross-cultural research.

Progesterone and ovulation across stages of the transition to menopause

Objective: Detailed characterization of progesterone and ovulation across the menopausal transition provides insight into conception risk and mechanisms of reproductive aging. Methods: Participants (n = 108, aged 25-58 y) collected daily urine specimens for 6-month intervals in each of 5 consecutive years. Specimens were assayed for pregnanediol glucuronide (PDG), luteinizing hormone (LH), follicle-stimulating hormone (FSH), and estrone glucuronide (E1G). Reproductive stage was determined using cycle length variance. A hierarchical algorithm was used to identify ovulation. Linear mixed-effects models estimated (1) the frequency and day of ovulation by age and stage; (2) differences in FSH, LH, and E1G levels between ovulatory (O) and anovulatory (AO) cycles; and (3) total PDG levels and PDG levels in O cycles by age and stage. Results: The probability of AO cycles increased across the perimenopause (P < 0.0001); reproductive stage was a stronger predictor than age of the probability of anovulation. Most cycles in late perimenopause were AO (>60%), but one quarter of cycles longer than 60 days were O. Average day of ovulation was later in the late perimenopause (mean [SD] cycle day, 27 [25] d) compared with the premenopause. FSH and LH levels were higher and E1G levels were lower in AO than O cycles (P < 0.0001 for each). Total PDG decreased in the late perimenopause, but 95th percentile PDG in O cycles declined steadily across the transition. Conclusions: Exposure to the risk of conception in women experiencing cycles long enough to classify them as late perimenopausal is far from negligible. Reproductive stage is more informative than age about PDG levels and the likelihood of anovulation.

A Multistate Model of Fecundability and Sterility

This paper develops a multistate hazards model for estimating fecundability and sterility from data on waiting times to conception. Important features of the model include separate sterile and nonsterile states, a distinction between preexisting sterility and sterility that begins after initiation of exposure, and log-normally distributed fecundability among nonsterile couples. Application of the model to data on first birth intervals from Taiwan, Sri Lanka, and the Amish shows that heterogeneity in fecundability is statistically significant at most ages, but that preexisting sterility and new sterility are unimportant before age 40. These results suggest that sterility may not be an important determinant of natural fertility until later reproductive ages.

Longitudinal analysis of deciduous tooth emergence: II. Parametric survival analysis in Bangladeshi, Guatemalan, Japanese, and Javanese children

We present a form of parametric survival analysis that incorporates exact, interval-censored, and right-censored times to deciduous tooth emergence. The method is an extension of common cross-sectional procedures such as logit and probit analysis, so that data arising from mixed longitudinal and cross-sectional studies can be properly combined. We extended the method to incorporate and estimate a proportion of agenic teeth. While we concentrate on deciduous tooth emergence, the method is relevant to studies of permanent tooth emergence and other developmental events. Deciduous tooth emergence data were analyzed from four longitudinal studies. The samples are 1,271 rural Guatemalan children examined every three months up to age two and every six months thereafter as part of the INCAP study; 397 rural Bangladeshi children examined monthly to age one and quarterly thereafter as part of the Meheran Growth and Development Study; 468 rural Indonesian children examined monthly as part of the Ngaglik study; and 114 urban Japanese children examined monthly in studies from 1910 and 1920. Although all four studies were longitudinal, many observations from the Guatemala and Bangladesh studies were effectively cross-sectionally observed. Three different parametric forms were used to model the eruption process: a normal distribution, a lognormal distribution, and a lognormal distribution with age shifted to shortly after conception. All three distributions produced reliable estimates of central tendencies, but the shifted lognormal distribution produced the best overall estimates of shape (variance) parameters. Estimates of emergence were compared to other studies that used similar methods. Japanese children showed relatively fast emergence times for all teeth. Bangladeshi and Javanese children showed emergence times that were slower than are found in most previous studies. Estimates of agenesis were not significantly different from zero for most teeth. One or two central incisors showed significant agenesis that ranged from 0.1 to 0.8% in three of the samples; even so, failure to model the agenic proportion did not seriously bias the estimates.

Total and Unopposed Estrogen Exposure Across Stages of the Transition to Menopause

Detailed characterization of estrogen dynamics during the transition to menopause is an important step toward understanding its potential implications for reproductive cancers developing in the transition years. We conducted a 5-year prospective study of endogenous levels of total and unopposed estrogen. Participants (n = 108; ages 25-58 years) collected daily urine specimens for 6 months in each of 5 consecutive years. Specimens were assayed for estrone-3-glucuronide (E1G) and pregnanediol-3-glucuronide. Linear mixed-effects models were used to estimate exposure to total and unopposed estrogen by age and reproductive stage. Reproductive stage was estimated using menstrual cycle length variance. E1G mean area under the curve and mean E1G 5th and 95th percentiles represented total estrogen exposure. An algorithm identifying days of above-baseline E1G that coincided with the days of baseline pregnanediol-3-glucuronide was used to identify days of unopposed estrogen. Mean E1G area under the curve increased with age in the pretransition and early transition and decreased in the late transition. Ninety-fifth percentile E1G levels did not decline until after menopause, whereas 5th percentile levels declined from the early transition to the postmenopause. The number of days of unopposed estrogen was significantly higher during the transition compared with the pretransition. Given the length of time women spend in the transition, they are exposed to more total and unopposed estrogen than has been previously appreciated. Coupled with epidemiologic evidence on lifetime exposure to estrogen, these results suggest that variation in the amount of time spent in the transition may be an important risk factor for reproductive cancers. (Cancer Epidemiol Biomarkers Prev 2009;18(3):828–36)

Age-independent increases in male salivary testosterone during horticultural activity among Tsimane forager-farmers

Testosterone plays an important role in mediating male reproductive trade-offs in many vertebrate species, augmenting muscle and influencing behavior necessary for male-male competition and mating-effort. Among humans, testosterone may also play a key role in facilitating male provisioning of offspring as muscular and neuromuscular performance are deeply influenced by acute changes in testosterone. This study examines acute changes in salivary testosterone among 63 Tsimane men ranging in age from 16-80 (mean 38.2) years during one-hour bouts of tree-chopping while clearing horticultural plots. The Tsimane forager-horticulturalists living in the Bolivian Amazon experience high energy expenditure associated with food production, have high levels of parasites and pathogens, and display significantly lower baseline salivary testosterone than age-matched US males. Mixed-effects models controlling for BMI and time of specimen collection reveal increased salivary testosterone (p<0.001) equivalent to a 48.6% rise, after one hour of tree chopping. Age had no effect on baseline (p=0.656) or change in testosterone (p=0.530); self-reported illness did not modify testosterone change (p=0.488). A comparison of these results to the relative change in testosterone during a competitive soccer tournament in the same population reveals larger relative changes in testosterone following resource production (tree chopping), compared to competition (soccer). These findings highlight the importance of moving beyond a unidimensional focus on changes in testosterone and male-male aggression to investigate the importance of testosterone-behavior interactions across additional male fitness-related activities. Acutely increased testosterone during muscularly intensive horticultural food production may facilitate male productivity and provisioning.

Longitudinal analysis of deciduous tooth emergence in Indonesian children. I. Life table methodology

Deciduous tooth emergence was investigated in a sample of 468 children from central Java. The subjects were examined every 35 days for up to 2 years. Statistics on the timing of emergence for each tooth were derived from actuarial life table estimates, and male and female survivorship curves of tooth emergence were compared for statistical sex differences. Javanese children exhibited delayed emergence compared to other populations. No overall pattern of sexual dimorphism was detected, although upper first molars emerged significantly earlier in females. A comparison of emergence sequence polymorphisms between Javanese, Finnish, and Japanese children revealed that the i1‐i2‐m1‐c‐m2 sequence is most common in all three populations, but each shows a uniquely higher proportion of one or more less common polymorphisms.

Colostrum feeding behaviour and initiation of breast-feeding in rural Bangladesh.

Human breast milk is primarily colostrum immediately following birth. Colostrum gradually changes to mature milk over the next several days. The role of colostrum in fighting infections and promoting growth and development of the newborn is widely acknowledged. This role is mediated by differences across cultures in the acceptability of colostrum and the prevalence of colostrum feeding. This study examined the prevalence of colostrum feeding and time to initiation of breast-feeding in 143 rural Bangladeshi women in Matlab thana. Structured interviews were collected during a 9-month prospective study conducted in 1993. Women were usually interviewed within 4 days of giving birth and were asked about whether or not they fed their child colostrum and the number of hours until they began breast-feeding the baby. Ninety per cent of the mothers reported feeding their newborn colostrum. A logistic regression found no effect on the prevalence of colostrum feeding from the following covariates: mothers age, parity, history of pregnancy loss, childs sex, mothers self-report of delivery complications, and the time from birth to interview. Fifty-nine per cent of mothers initiated breast-feeding within 4 h, and 88% within 12 h of parturition. Survival analysis was used to estimate the effects of covariates on the time from delivery to initial breast-feeding. Time to initial breast-feeding was delayed slightly, but significantly, for older mothers, for male infants, and by mothers who did not report delivery complications. The percentage of mothers who fed their child colostrum was higher, and times to initial breast-feeding were shorter, than almost all previous reports from South Asia. These findings might be explained, in part, by methodological differences among studies, but it is suggested that recent changes towards earlier initiation of breast-feeding have taken place in rural Bangladesh.

Monitoring reproductive aging in a 5-year prospective study: aggregate and individual changes in luteinizing hormone and follicle-stimulating hormone with age

Objective: This study describes age-related changes in luteinizing hormone (LH) and follicle-stimulating hormone (FSH) in a 5-year prospective study of reproductive aging. Design: Participants (n = 156 college-educated, white, US women; 25 to 58 y) were recruited from the TREMIN Research Program on Womens Health. They collected daily urine specimens for 6 months in each of 5 consecutive years. Specimens were assayed for LH and FSH. Aggregate changes were calculated in LH and FSH with age, and multilevel models were used to estimate individual hormone trajectories and within-woman and between-woman variances by age. Results: Aggregate LH levels increased beginning after age 45; FSH increased at all ages, accelerating after age 45. Individual-level patterns with age included the following: reproductive-age LH and FSH levels, with increasing FSH and increasing or decreasing LH (ages 20 to 49); rapidly increasing LH and FSH (ages 40 to 59); and increasing or steady postmenopausal LH and FSH (ages 46 to 62). FSH levels were consistently high in the latter category, but LH levels overlapped with levels found in younger women (<45 y). Individual LH patterns showed more variability (5% to 35% of total variance) than FSH (3% to 22% of total variance). Both hormones had relatively low variation within individuals compared with between-woman differences (65% to 97% of total variance). Conclusions: Aggregate-level data do not reflect differences across women and oversimplify the age-related increases and variability in LH and FSH. Individual FSH levels are not distinguishable from reproductive-age levels until after rapid perimenopausal increases in FSH occur; individuals vary in whether their postmenopausal LH levels are distinguishable from reproductive-age levels.