Exercise interventions and dosage for weight management in overweight or obese middle-aged and older adults: A dose-response Bayesian network meta-analysis

Optimal exercise prescription for weight management in overweight and obese middle-aged and older adults remains unclear. This study aimed to determine optimal exercise types and doses through dose-response Bayesian network meta-analysis. We searched PubMed, Embase, Web of Science, and Cochrane Library (January 2000–March 2025) for randomized controlled trials examining exercise interventions in adults aged ≥50 years with BMI ≥24 kg/m2. Exercise dose was quantified as METs-min/day. Dose-response relationships were modeled using restricted cubic splines. Ninety-two studies (4,634 participants) were included. An L-shaped dose-response relationship was identified, with 150 METs-min/day as the inflection point. AT+SE demonstrated notable dose-efficiency at only 100 METs-min/day. Aerobic training achieved highest SUCRA rankings across outcomes (84.30 for body weight, 79.5 for BMI, 82.0 for body fat percentage). Combined interventions produced largest effect sizes: AT+RT for body weight (SMD = −0.41), RT+BE+AT for BMI (SMD = −0.42), and AT+SE for body fat percentage (SMD = −0.97). Evidence quality ranged from moderate to very low. This network meta-analysis provides preliminary evidence for exercise prescription in this population. The L-shaped dose-response relationship suggests diminishing returns beyond 150 METs-min/day, while AT+SE achieves equivalent benefits at lower volumes. These findings should be confirmed in future high-quality trials.Registration: PROSPERO CRD420251142557