# Network meta-analysis and dose–response analysis of exercise on sleep quality and BMI in obese populations

## TL;DR

This study compares different types of exercise to see which are best for improving sleep and reducing BMI in obese people, finding that resistance and combined training improve sleep, while aerobic exercise is best for weight loss within a specific dosage range.

## Contribution

The study provides novel evidence on the relative efficacy and optimal dosage of seven exercise interventions for sleep quality and BMI in obese populations using network meta-analysis and dose-response analysis.

## Key findings

- Combined aerobic and resistance training and resistance training are most effective for improving sleep quality.
- Aerobic exercise is most effective for reducing BMI within a 60–70 hour optimal dosage window.
- BMI improvement shows a non-linear 'U-shaped' dose–response relationship, while sleep quality does not show significant linear dose-dependency.

## Abstract

Exercise is key to ameliorating sleep disorders in obese populations; however, the relative benefits of different exercise modes and the optimal dosage remain unclear. This study aims to systematically evaluate the relative efficacy and dose–response characteristics of seven exercise interventions on sleep quality (Pittsburgh Sleep Quality Index [PSQI]) and body mass index (BMI) in overweight and obese populations via network meta-analysis (NMA).

This study strictly adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and systematically searched nine databases, including PubMed, Embase, and China National Knowledge Infrastructure (CNKI), from inception to 15 October 2025. Randomized controlled trials (RCTs) evaluating the effects of the aforementioned exercise interventions on sleep quality (primarily PSQI) or body mass index (BMI) in overweight or obese individuals were included. Network meta-analysis was conducted using Stata 18.0 software. The standardized mean difference (SMD) was used as the effect size, and interventions were ranked using the Surface Under the Cumulative RAnking curve (SUCRA).

The network meta-analysis revealed that combined aerobic and resistance training (ART, SUCRA ≈ 77.1%) and resistance training (RT, SUCRA ≈ 75.2%) were significantly superior to other interventions in improving sleep quality. Conversely, erobic Exercise (AE, SUCRA ≈ 74.0%) was most effective in reducing BMI. Dose-effect analysis indicated that improvements in sleep quality did not exhibit significant linear dose-dependency (p > 0.05). However, BMI improvement showed a significant non-linear “U-shaped” dose–response relationship (p_quadratic = 0.009). The fitted curve suggests that a cumulative intervention duration of 60–70 h represents the optimal dosage window for weight loss, with diminishing marginal returns observed beyond this range.

Exercise interventions demonstrate significant specificity in their benefits for health outcomes in obese populations. Resistance and combined training are recommended as the preferred strategies for improving sleep, with benefits depending more on the mode than on high cumulative dosage. Conversely, aerobic exercise is the optimal protocol for weight loss, provided the exercise volume is controlled within the optimal dosage window to maximize returns. These findings offer evidence-based grounds for clinical practice to formulate differentiated and precise exercise prescriptions for obesity and sleep management.

www.crd.york.ac.uk/prospero, identifier CRD420251251401.