Lean mass loss with incretin drugs vs lifestyle change: a 20-trial meta-analysis

A 2026 systematic review and meta-analysis in Diabetes, Obesity and Metabolism examined lean mass loss in adults with overweight or obesity treated with incretin-based drugs or with intensive lifestyle intervention ^[1]. It asked whether lean mass makes up a different share of total weight lost under each approach ^[1]. GLP-1 receptor agonists, also called incretin mimetics, mimic glucagon-like peptide-1, a hormone made by the small intestine that triggers insulin release, slows stomach emptying and increases fullness after eating ^[4]. The review covered two GLP-1 receptor agonists, semaglutide and liraglutide, and tirzepatide, a dual GLP-1/GIP receptor agonist ^[1][4].

What was studied

The authors searched PubMed, Embase, Cochrane CENTRAL and Web of Science from inception through 20 January 2026 ^[1]. Eligible studies were randomised controlled trials of semaglutide, tirzepatide, liraglutide or lifestyle interventions that reported body composition measured by dual-energy X-ray absorptiometry (DXA) or magnetic resonance imaging (MRI) ^[1]. The two co-primary outcomes were the absolute change in lean mass in kilograms and the proportion of total weight loss attributable to lean mass ^[1]. Results were pooled with random-effects models, risk of bias was assessed with the Cochrane RoB 2 tool, and certainty of evidence was rated with GRADE ^[1].

Lean mass is not the same quantity as skeletal muscle ^[2]. A separate 2026 systematic review in Annals of Internal Medicine expected fat-free mass or lean soft tissue measured by DXA or bioelectrical impedance to account for about 25% of total weight lost, and skeletal muscle measured by computed tomography or MRI to account for about 15% ^[2].

What was found

Twenty randomised controlled trials comprising 15 782 participants met the inclusion criteria ^[1]. The pooled proportions of total weight loss accounted for by lean mass were:

• Semaglutide: 35.2% (95% CI 31.5 to 38.9) ^[1].
• Tirzepatide: 25.4% (95% CI 22.8 to 28.0) ^[1].
• Liraglutide: 26.8% (95% CI 23.1 to 30.5) ^[1].
• Lifestyle interventions: 26.2% (95% CI 24.1 to 28.3), with p = 0.42 reported for the comparison ^[1].
• Lifestyle intervention plus resistance training: 17.5% (95% CI 14.2 to 20.8), the most favourable profile in the analysis ^[1].

Across the three drugs, lean mass accounted for 25% to 39% of weight lost ^[1]. Heterogeneity between trials was moderate (I² = 68%), and Egger's test did not detect publication bias (p = 0.23) ^[1].

The authors concluded that lean mass loss during significant weight reduction is substantial and that the lean share of weight lost is broadly comparable between incretin pharmacotherapy and lifestyle intervention ^[1]. They also stated that muscle mass can be significantly preserved by adding resistance training, adequate protein intake and body composition monitoring to weight-loss programmes ^[1].

How it sits with other evidence

The Annals review included 35 randomised trials, with a median duration of 26 weeks and a median of 78 participants, and did not pool results because body composition methods and reporting varied too much ^[2]. Within incretin groups, the median share of weight loss from muscle-related indices was 28.3% (IQR 15.9% to 39.9%), and 65% of studies exceeded the benchmark of about 25% ^[2]. In the 13 studies reporting weight loss in lifestyle or placebo comparator groups, median weight loss was −2.5%, and 38% exceeded the relevant benchmark ^[2]. None of the included studies reported objective physical function outcomes ^[2].

A 2025 meta-analysis restricted to people with diabetes reported mainly favourable effects of GLP-1 receptor agonists on muscle in animal models, with human evidence limited to six studies measuring lean mass and four trials assessing skeletal muscle mass index; its authors called for large randomised trials in humans ^[3].

Limitations

• The abstract does not report the pooled absolute lean mass change in kilograms, although this was a co-primary outcome ^[1].
• The abstract does not report the RoB 2 results or GRADE certainty ratings, so the strength of evidence behind each estimate cannot be judged from the abstract alone ^[1].
• The abstract does not state whether the drug and lifestyle estimates come from head-to-head trials or from separate arms pooled by intervention type, nor whether differences among the three drugs were formally tested ^[1].
• Heterogeneity was moderate (I² = 68%), so each pooled proportion averages over trials whose results differed ^[1].
• DXA-derived and MRI-derived measures were combined, and the Annals review treats these methods as capturing compartments with different expected shares of weight loss ^[1][2].
• A proportion is not an absolute amount; because weight loss was consistently larger with incretin therapy than with lifestyle or placebo comparators in the Annals review, a similar lean share can correspond to more kilograms of lean mass lost ^[1][2].
• The resistance-training estimate comes from lifestyle interventions, and the abstract gives no pooled estimate for incretin therapy combined with resistance training and no quantified data on protein intake or body composition monitoring ^[1].
• The abstract reports no strength or physical function outcomes, and no trial in the Annals review measured objective physical function, so the clinical meaning of these lean mass changes remains uncertain ^[1][2].