Abstract:Menopause fundamentally reshapes female physiology, yet current understanding is limited by small longitudinal cohorts that characterize it as a gradual transition. Large-scale biomedical datasets remain underutilized because the age of the final menstrual period (FMP) is rarely recorded. Here, we present a computational framework that leverages cross-sectional data to reconstruct systemic physiology as a function of time relative to FMP. We adapted a deconvolution framework from astronomy to recover systemic biological trajectories by deconvolving the population distribution of FMP age from chronological data. Applying this to two national cohorts with 300 million laboratory tests from 1.3 million females, we transformed cross-sectional measurements into high-resolution timelines anchored to the FMP. Our analysis reveals a step-like physiological cliff at the FMP across endocrine, skeletal, hepatic, renal, inflammatory, and lipid systems. These discontinuities are absent in males and highly concordant across independent populations. We demonstrate that systemic dysregulation begins over a decade prior to FMP, significantly expanding the window for preventive intervention. Furthermore, hormone replacement therapy (HRT) appears to markedly attenuate these abrupt physiological shifts. These findings further support a systemic and quantifiable view of the menopausal transition and provide a generalizable strategy for recovering hidden biological trajectories from human datasets, applicable to other life stages such as puberty or disease progression.
From: Glen Pridham [view email]
[v1]
Sat, 8 Nov 2025 08:10:29 UTC (5,962 KB)
[v2]
Thu, 13 Nov 2025 06:48:11 UTC (6,649 KB)
[v3]
Sun, 24 May 2026 16:19:26 UTC (7,027 KB)