How Obesity Genetics Shape Multimorbidity

The GEMINI study used a genetics-informed, pairwise multimorbidity framework to examine why long-term conditions co-occur. Across approximately 70 long-term conditions, diseases were analyzed as pairs to estimate shared inherited liability. A strong BMI-related genetic signal appeared across many pairs, indicating that obesity-related genetic variation contributes to a large share of common multimorbidity patterns.

Within this framework, BMI-associated genetic liability (used as a proxy for obesity risk) was identified as a statistically significant contributing factor for the majority of conditions analyzed. In a smaller subset of 10 condition pairs, adjustment for obesity-related genetics fully attenuated the estimated shared genetic correlation, meaning that—within the limits of the model—obesity genetics accounted for the observed overlap. Many other pairs showed different dominant drivers, underscoring that multimorbidity arises through heterogeneous pathways.

To translate these findings into population-level impact, the authors modeled a hypothetical 4.5-unit reduction in BMI. Under this counterfactual scenario, fewer people were estimated to have both conditions in selected pairs—for example, 17 fewer per 1,000 people with chronic kidney disease and osteoarthritis, and 9 fewer per 1,000 with type 2 diabetes and osteoarthritis. These are modeled risk shifts at the population level, not individual guarantees or effects of a specific intervention.