Subgroup-specific cardiovascular effects of intensive lifestyle intervention: a causal forest analysis in adults with type 2 diabetes and overweight/obesity

OBJECTIVE: To evaluate the heterogeneity of treatment effects (HTEs) of intensive lifestyle intervention (ILI) in adults with type 2 diabetes and overweight/obesity, and identify a subgroup with greater cardiovascular benefits from ILI. METHODS: In a post-hoc analysis of the Look AHEAD trial, causal forest modeling was used to identify HTEs of ILI. The study population was stratified into four subgroups, and the associations of ILI with cardiovascular outcomes were assessed using multivariable Cox modeling compared to diabetes support & education (DSE). RESULTS: Among 4710 participants (mean age 58.9 years, 58.5% women), 768 primary outcomes occurred over a median follow-up of 9.5 years. Key variables identified through causal forest modeling were SF-36 mental health, diabetes duration, and urine albumin-creatinine ratio (ACR). In Subgroup 4 (SF-36 mental health > 55.64 and ACR > 10 mg/g), which had more cardiovascular risk factors and comorbidities, ILI significantly reduced the primary outcome risk (HR: 0.65, 95% CI: 0.48-0.87, P = 0.004) and three secondary outcomes compared to DSE. No cardiovascular benefits were observed in participants with SF-36 mental health ≤ 55.64 or ACR < 10 mg/g. CONCLUSION: This post-hoc analysis of the Look AHEAD trial showed HTEs of ILI in adults with type 2 diabetes and overweight/obesity. Participants with better mental health, poorer renal function, and more cardiovascular risk factors were more likely to benefit from ILI.