Virginia Tech leads $9.26 million NIH push to model hormones and women’s health

A multimillion-dollar effort to answer one of the most stubborn questions in women’s health — why the same hormonal milestones can produce wildly different outcomes in different women — will be led from Virginia Tech, backed by an award of up to $9.26 million from the National Institutes of Health. The project, funded through a growing initiative by the NIH Office of Research on Women’s Health, will bring together engineers, mathematicians, data scientists, reproductive biologists, and practicing clinicians from institutions across the United States. Their shared target is the structure and function of vaginal tissue, a material the body remodels continuously across pregnancy, childbirth, postpartum recovery, aging, and menopause, and whose mechanical behavior is deeply entangled with pelvic health, injury risk, and quality of life.

The team is led by Raffaella De Vita, professor in the Department of Mechanical Engineering at Virginia Tech, who has spent years studying the biomechanics of vaginal and pelvic tissue. Her group has built experimental and computational tools to characterize how soft tissues deform, recover, and fail, but she argues that the new award demands something more ambitious than applying existing methods. What excites her most, she says, is the chance to ask a much bigger question: if researchers had all the data in the world, could they build a model that predicts how hormonal changes reshape vaginal tissue over time? She has emphasized that she does not want to simply apply the computational tools already available; instead, she wants the underlying biological questions to push the team toward new mathematical, computational, and AI-enabled approaches, and then to use those approaches to generate knowledge about women’s health that experiments or data alone could not provide.

The biological problem at the heart of the project is deceptively simple to state. A woman’s hormonal environment changes substantially throughout life, and each transition — reproductive maturity, pregnancy, postpartum recovery, aging, and menopause — alters not only hormone levels but also the composition, structure, and mechanical behavior of vaginal tissue. Collagen networks reorganize, smooth muscle and connective tissue adapt, and the tissue’s stiffness and resilience shift accordingly. Those shifts matter clinically: the pelvic floor supports the bladder, uterus, and bowel, and changes in tissue mechanics are implicated in pelvic floor disorders, prolapse, and complications of childbirth. Yet the outcomes of these transitions vary enormously among women. Pregnancy and childbirth can leave very different long-term signatures, and the timing and experience of menopause differ from person to person. Understanding what drives that variability is one of the central unsolved challenges in the field.

De Vita argues that no single discipline can crack it. Tissue-level questions require engineering and mechanics; hormone dynamics require reproductive biology and endocrinology; population-level patterns require clinical medicine and epidemiology; and integrating data across all of these scales requires mathematics, statistics, and machine learning. The new project is deliberately constructed around that premise. Rather than a single laboratory pursuing a narrow hypothesis, the award funds a multi-institutional consortium whose members were chosen precisely because their expertise overlaps at the seams where traditional fields usually stop.

The technical core of the project is a new computational framework for predicting how vaginal tissue changes in response to hormones over time. The researchers will integrate existing experimental, clinical, and population-level data, then combine artificial intelligence and machine learning with mathematical and computational models to connect otherwise disparate data sets. The goal is to identify quantitative relationships between hormone dynamics and changes in tissue microstructure and mechanics — essentially, to learn the rules that map a woman’s hormonal trajectory onto the structural state of her tissue. Machine learning is expected to play a dual role: finding patterns in heterogeneous data that classical models would miss, and accelerating simulations of tissue remodeling that would otherwise be computationally prohibitive.

A distinctive feature of the plan is its treatment of uncertainty. Biological systems are noisy, and data on vaginal tissue — collected across different cohorts, measurement techniques, and life stages — carries limitations that could easily undermine naive predictions. To guard against that, the team will use uncertainty quantification to determine how biological variability and gaps in the available data affect the reliability of their predictions. Pinar Acar, associate professor of mechanical engineering at Virginia Tech, works in exactly this area, developing data-driven methods for predictive modeling under uncertainty. Justin Krometis, research associate professor at the Virginia Tech National Security Institute and affiliate faculty in mathematics, brings complementary expertise in parameter estimation, helping the team determine how variability and uncertainty in data and model parameters propagate into computational predictions. Once predictions are generated, the researchers will test and refine them against available experimental and clinical data, published findings, and other relevant observations — a closed loop of prediction, validation, and revision.

The mathematical side of the consortium is anchored by three Virginia Tech faculty members. Stanca Ciupe, professor of mathematics, develops models of biological systems with particular expertise in describing how complex biological processes evolve dynamically over time, a skill set well suited to tracking hormone-driven remodeling across decades. Traian Iliescu, also a professor of mathematics, brings decades of experience with mathematical and computational approaches for complex systems, including reduced-order methods that make large computational problems tractable — a critical capability when models must span molecular, tissue, and population scales. Sean Lawley, professor at the University of Utah, develops mathematical approaches to biological and physiological problems, including research related to reproductive aging and menopause, giving the team direct theoretical access to the hormonal transitions that most dramatically alter tissue behavior.

Clinical expertise is woven throughout the collaboration. Marianna Alperin, professor and vice chair for research in the Department of Obstetrics, Gynecology, and Reproductive Sciences at the University of California San Diego, is a physician-scientist specializing in urogynecology and reconstructive pelvic surgery whose research seeks to understand the mechanisms underlying pelvic floor disorders and translate scientific discoveries into improved strategies for women’s health. Kathleen Connell, professor at the University of Colorado Anschutz School of Medicine, is a board-certified specialist in urogynecology and reconstructive pelvic surgery whose clinical and research work connects fundamental research with women’s pelvic health. Alperin has framed the clinical stakes directly: for a clinician, the ability to better understand how and why tissues change across different stages of a woman’s life could fundamentally change how the field thinks about women’s health. The long-term goal, she says, is to turn that understanding into better ways to anticipate problems, prevent them when possible, and improve treatment.

Reproductive biology completes the picture. Joshua Johnson, associate professor at the University of Colorado Anschutz School of Medicine, studies female reproductive biology with expertise in ovarian function, reproductive aging, and the cellular and molecular processes that regulate reproductive health — the biological machinery that sets the hormonal signals the computational models must interpret. On the Virginia Tech engineering side, Justin Dubik, instructor and research collaborator in the Department of Mechanical Engineering, conducted research in De Vita’s laboratory before joining the instructional faculty and has contributed to experimental and computational studies of vaginal tissue biomechanics, providing continuity between the lab’s prior work and the new initiative. Together, the team’s members bring complementary perspectives to a challenge that no single discipline can address alone: building a predictive understanding of how hormones shape vaginal tissue health across a woman’s lifespan.

If the framework succeeds, the implications could extend well beyond the laboratory. A validated predictive model of hormone-driven tissue change would give clinicians a way to identify which women face elevated risk at particular life stages, opening the door to earlier screening, personalized prevention strategies, and more targeted treatments for pelvic floor disorders. It would also give researchers a principled way to design experiments, by highlighting where uncertainty is largest and where new data would most improve predictions. The award, numbered OT2OD042813, reflects the NIH Office of Research on Women’s Health’s growing investment in approaches that treat women’s health as a quantitative, predictive science rather than a descriptive one. For De Vita and her collaborators, the measure of success will be whether the models they build can do what experiments alone cannot: turn decades of fragmented observations about hormones and tissue into a coherent, testable theory of how women’s bodies change over time — and what can be done to keep those changes from becoming health problems.

Subject of Research: Hormonal regulation of vaginal tissue structure and mechanics across the female lifespan, studied through computational modeling

Article Title: Virginia Tech leads multimillion-dollar project for women's health

Article References: Virginia Tech leads multimillion-dollar project for women's health. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: women's health, vaginal tissue biomechanics, hormones, menopause, pregnancy, pelvic floor disorders, machine learning, computational modeling, uncertainty quantification, NIH funding, Virginia Tech, reproductive aging