Multimodal data analysis reveals asynchronous aging dynamics across female reproductive organs
Authors:
- Oleksandra SoldatkinaOSOleksandra Soldatkina
- Laura Ventura-San PedroLPLaura Ventura-San Pedro
- Allal El HommadAHAllal El Hommad
- Jose Miguel RamirezJRJose Miguel Ramirez
- Aida Ripoll-CladellasARAida Ripoll-Cladellas
- Maria Sopena-RiosMSMaria Sopena-Rios
- Jaume OrdiJaume Ordi
- Natàlia Pujol-GualdoNPNatàlia Pujol-Gualdo
- Marta MeléMMMarta Melé
Preprint
Abstract
Female reproductive aging is a complex process with profound systemic health implications, yet the molecular and structural dynamics of aging across reproductive organs and tissues remain largely unexplored. Here, we integrate deep learning-based analysis of 1,112 histological images with RNA-seq data from 659 samples across seven reproductive organs in 304 female donors aged 20-70 years. We show that female reproductive organs and tissues have asynchronous aging dynamics: while the ovary and vagina age gradually, the uterus undergoes an abrupt transcriptional and cellular transition around the age of menopause. Tissue segmentation highlights that the myometrium in the uterine wall is the most age-affected tissue, marked by extracellular matrix remodeling and immune activation. Across reproductive organs, the epithelial tissue is also strongly affected by age, with the vaginal epithelium showing a unique sharp menopausal transition. Integration via multi-omics factor analysis links these tissue-specific histological transformations to specific molecular shifts, many with nonlinear expression trajectories and enriched in heritable reproductive traits such as pelvic organ prolapse and age at menarche. These findings position menopause as a key inflection point in female aging and provide insights with tissue-specific focus to support healthier menopausal transitions and reduce age-related disease risk.Keywords
All Related Versions
- Published version: Nature Aging, 6 (5), 1177.
This publication has 122 references indexed in Scilit:
- Integration of multi-modal datasets to estimate human agingMachine Learning, 2024
- A machine learning approach for quantifying age-related histological changes in the mouse kidneyGeroScience, 2023
- Histopathologic brain age estimation via multiple instance learningActa Neuropathologica, 2023
- Bacterial vaginosis and its association with infertility, endometritis, and pelvic inflammatory diseaseAmerican Journal of Obstetrics and Gynecology, 2021
- Anatomy and Physiology of the Breast during Pregnancy and Lactation2020
- Artificial Intelligence and Machine Learning for Digital Pathology2020
- Weighted Pivot Coordinates for Compositional Data and Their Application to Geochemical MappingMathematical Geosciences, 2017
- Genitourinary syndrome of menopause: an overview of clinical manifestations, pathophysiology, etiology, evaluation, and managementAmerican Journal of Obstetrics and Gynecology, 2016
- Compositional Data Analysis2011
- On Early Stopping in Gradient Descent LearningConstructive Approximation, 2007