Abstract
Type-2 diabetes approximately doubles the risk of Alzheimer’s disease. This association is well established epidemiologically, but the specific metabolic pathways that mediate the link — the molecular “why” — remain poorly characterized. Insulin resistance, chronic inflammation, lipid dysregulation, and altered amino acid metabolism have all been proposed as candidate mechanisms, but disentangling their individual contributions in observational data is challenging, particularly when the patients in question are taking medications (metformin, statins) that directly alter the very metabolites under investigation.
This study, conducted at Harvard Medical School and Brigham and Women’s Hospital, takes advantage of the MGB Biobank’s Nightingale Health NMR metabolomics platform — which quantifies approximately 249 biomarkers from a single blood sample — to identify the metabolomic signatures that differentiate T2D patients who go on to develop AD from those who do not. The study employs a 2×3 factorial observational design that separates the contributions of diabetes status and cognitive outcome, with careful covariate adjustment for age, sex, BMI, and critically, for metformin and statin use. This pharmacological adjustment is essential: without it, any observed metabolomic differences could reflect medication effects rather than disease biology.
The analytical pipeline uses a multi-stage approach. Differential analysis (limma) identifies individual metabolites that differ between groups. Multivariate discrimination (PLS-DA) assesses whether the overall metabolomic profile can separate the groups, and pathway analysis maps significant metabolites onto known biological pathways to generate mechanistic hypotheses.
The ultimate goal is twofold: to generate mechanistic insight into the T2D–AD biological axis, and to identify metabolomic features that could serve as biomarkers for risk stratification — measurable from a routine blood draw and actionable in a clinical setting. If specific metabolomic signatures reliably flag T2D patients at elevated AD risk, this would enable targeted monitoring and preventive intervention in a large, identifiable population.
This study is in progress.