In populations that approximate conditions of natural fertility, the period of lactational infecundability is one of the most important and the most variable components of birth intervals, lasting from a few months to several years. Breastfeeding acts to suppress hypothalamic function which, in turn, suppresses the reproductive axis. A debate has emerged in the last decade over the mechanism by which breastfeeding suppresses hypothalamic activity. One hypothesis proposes a direct neurological pathway in response to nursing intensity. The alternative hypothesis is that hypothalamic function is suppressed in response to the metabolic load of breastfeeding. For this project, we develop and evaluate two classes of biodemographic tools for investigating the mechanisms of lactational amenorrhea. First, we will develop new endocrine tools for assessing metabolic load for large population-based field projects. We will develop and validate assays for urinary leptin, C-peptide, T3, and IGF-I. Second, we develop two new models of lactational amenorrhea. We extend the mixture model of Ford and Kim (1987) to include time-varying covariates modeled on the hazard of resumption of menses in each subgroup and the probability of subgroup membership. We also propose a new three-state stochastic model that may represent more realistically the underlying mechanisms of lactational amenorrhea. These tools will be applied to three existing longitudinal data sets in order to evaluate their potential for a more targeted study. The statistical tools will be used to examine the relative effects of nutrition (assessed by anthropometric measures), nursing behavior, maternal workload and other health and socioeconomic variables on the time to resumption of menses in two of the studies (the Ngaglik study from Java and the INCAP study in Guatemala). A third 9-month study from Bangladesh includes urine specimens collected from breastfeeding women. Specimens will be assayed for urinary leptin, C-peptide, T3, IGF-I, and cortisol (using an existing assay). The effects of endocrine markers and nursing behavior on the maintenance of PPA will be examined by statistically fitting the new models to these observations. Relevance: In many developing settings the contraceptive effects of nursing contribute significantly to birth spacing. This project will develop tools and provide preliminary analyses to disentangle how the metabolic load of nursing and the frequency and duration of nursing suppress reproduction. The results will be important for understanding and projecting effects of improvements in women's health, nutrition, and work patterns on the duration to return of full fecundity. The resulting endocrine tools will also be useful for other large-scale (field) studies that require measurement of relative metabolic load.
