Optimal dose and type of exercise across cardiometabolic outcomes in adults with overweight or obesity: protocol for a Bayesian model-based dose–response network meta-analysis of randomised controlled trials

INTRODUCTION: Overweight and obesity are major global public health challenges and increase the risk of type 2 diabetes, dyslipidaemia, hypertension and cardiovascular disease. Exercise is a safe and cost-effective non-pharmacological strategy to improve cardiometabolic health, yet the optimal combinations of exercise modality and dose for key cardiometabolic outcomes remain uncertain. This protocol aims to synthesise evidence on the joint effects of exercise modality and dose in adults with overweight or obesity and to identify modality-dose combinations associated with the most favourable cardiometabolic profiles. METHODS AND ANALYSIS: We will search PubMed, Embase, Web of Science, the Cochrane Library, Scopus, SPORTDiscus and China National Knowledge Infrastructure (CNKI) from inception to June 2026 for randomised controlled trials of exercise interventions in adults (≥18 years) with overweight or obesity. Risk of bias will be assessed using the Cochrane Risk of Bias 2 tool. We will conduct a Bayesian model-based dose-response network meta-analysis to estimate modality-specific dose-response relationships across cardiometabolic outcomes, with exercise dose standardised as metabolic equivalent of task minutes per week. Non-linear dose-response curves will be fitted to estimate minimum effective doses and optimal dose ranges. Meta-classification and regression tree analyses will be used to explore potential effect modifiers. ETHICS AND DISSEMINATION: Ethical approval is not required because no primary data will be collected. Findings will be submitted to a peer-reviewed journal. PROSPERO REGISTRATION NUMBER: CRD420251229131.