An Integrated Quantitative Methodology to Longitudinally Characterize Complex Dynamic Processes Associated with...

Area of Science:

• * Biostatistics and Bioinformatics
• * Reproductive Endocrinology and Aging
• * Epidemiological Methods

Background:

• * Dynamic biological processes, such as female ovarian aging and menopausal transition, exhibit complex and highly variable patterns.
• * Accurate characterization of these changes is crucial for understanding reproductive health and aging.
• * Existing statistical methods may not fully capture the intricacies of these dynamic processes and associated model uncertainties.

Purpose of the Study:

• * To develop and validate an integrative statistical methodology for characterizing complex dynamic biological processes.
• * To estimate population mean profiles, change points, and time-window lengths within these processes.
• * To apply the methodology to understand the stages of female ovarian aging and menopausal transition using hormone data.

Main Methods:

• * Utilized a semi-/non-parametric stochastic mixed-effect model to estimate population mean profiles and change points.
• * Employed a Bayesian Model Averaging (BMA) approach to address and quantify model uncertainty.
• * Fitted piecewise linear mixed-effect models to estimate the mean rate of change for sub-processes.

Main Results:

• * The developed methodology successfully characterized the complex patterns of change in female ovarian aging.
• * Identified multiple significant change points and defined time-windows related to the menopausal transition.
• * Provided robust estimates of hormone profiles (estradiol [E2] and follicle stimulating hormone [FSH]) and their rates of change.

Conclusions:

• * The integrative methodology offers a powerful tool for analyzing highly variable dynamic biological processes.
• * It provides a robust framework for understanding the stages of female reproductive aging and menopausal transition.
• * The approach accounts for model uncertainty, leading to more reliable characterization of biological aging patterns.