Lopes Menezes · Diabetes, obesity & metabolism 2026 · systematic review and meta-analysis · n=19 studies

Body Composition Remodelling During GLP-1-Based Therapy: A Systematic Review and Meta-Analysis Using a Hierarchical Physiological Framework.

Cited 0 times in the scientific literature.

Level 1 - systematic review of randomized trials

Systematic review and meta-analysis of comparative and single-arm studies

PubMed 42608323 · doi:10.1111/dom.71220 · record verified 2026-08-28

What was done

Authors searched PubMed, Embase, and Cochrane CENTRAL through May 2026 for studies reporting longitudinal body composition outcomes (adiposity, muscle quantity, muscle quality, body weight) in adults receiving GLP-1 receptor agonists or dual incretins. Comparative evidence was prioritized and analyzed using random-effects meta-analyses with restricted maximum likelihood estimation and Hartung-Knapp adjustment within a predefined physiological framework.

What was found

Nineteen studies were included, spanning comparative and single-arm designs. Only three comparative studies were eligible for the primary adiposity meta-analysis and two for the muscle quantity meta-analysis. Comparative analyses suggested a direction toward greater adiposity reduction with GLP-1 therapy, but the pooled estimate had substantial heterogeneity and a wide confidence interval crossing the null (no numeric point estimates or intervals were provided in the abstract). Muscle quantity data were similarly heterogeneous and imprecise, while muscle quality could not be meta-analyzed due to incomplete reporting. Exploratory single-arm studies showed decreases in weight and adiposity, with variable effects on muscle quantity.

Why it matters

Despite widespread clinical use of incretin-based therapies, robust comparative evidence remains insufficient to establish preferential fat loss over lean mass reduction or to rule out clinically meaningful muscle loss.

Limits

Only two to three comparative studies contributed to primary pooled estimates, resulting in wide confidence intervals, substantial between-study heterogeneity, and low to very low GRADE certainty. Muscle quality metrics could not be quantitatively pooled, functional outcomes were not meta-analyzed, and total participant counts were omitted from the abstract.

Cited by