91³ÉÈ˰涶Òô

Despite rising life expectancy, Australian women bear a greater burden of chronic health conditions than men, often emerging during perimenopause and menopause stages. These stages, lasting 10 to 15 years, involve hormonal shifts affecting metabolism, potentially leading to higher blood sugar, increased belly fat, unhealthy cholesterol levels, inflammation markers, decreased muscle mass and worsen brain health.

Historically under-researched, this transition is challenging to study since it involves heterogenous factors, such as interacting hormonal fluctuations, genetics, lifestyle and overall health; it is a gradual process spanning over several years; and it has large individual variability. To address these challenges, this project aims to develop a comprehensive computational model integrating established scientific understanding of women's metabolic systems with real-world data analysis. This model will be used to create more effective and personalised strategies to help women prevent health issues and improve their overall well-being during and after menopause. 

Aims

This PhD project aims to investigate how transitions through menopause stages (pre-peri, and post) might influence a woman's metabolic system, and how individual women respond to different therapies. It will help guide the development of preventative health strategies to optimise long-term outcomes for women.

Objectives:

i) Develop a Hybrid Digital Twin of the female metabolic system. 

ii) Investigate longitudinal changes in metabolic health markers (e.g., body adiposity, anthropometrics, blood pressure, serum lipids, and blood glucose) across pre-menopause, perimenopause, and post-menopause stages in Australian women, and explore the mechanisms by which hormonal fluctuations during menopause may impact these markers. 

iii) Model the heterogeneity in women’s response to selected therapies as a function of their metabolic biomarkers and lifestyle factors. 

iv) Utilise the digital twin model to evaluate preventative approaches and identify optimal strategies for enhancing long-term health outcomes in women as they age.

Design

This project will develop a hybrid digital twin system integrating a mechanistic model based on Ordinary Differential Equations (ODEs) and advanced data-driven machine learning (ML) techniques. This project will utilise the extensive Australian Longitudinal Study on Women’s Health dataset, which includes health and wellbeing data from over 57,000 women across four cohorts, alongside the Medicine Insight dataset and potentially the Lumos dataset, a highly comprehensive data repository of general practice and hospital data in NSW.

Once established, the digital twin of the women's metabolic system will focus on optimising stratified preventative strategies for potential adverse health outcomes. Advanced analytical techniques, such as causal counterfactual analysis and post-hoc explainability, will be applied within this framework to deepen our understanding of the factors influencing metabolic health. This approach aims to facilitate the development of targeted and effective preventative strategies tailored to individual needs.