When researchers study drugs that lower blood sugar, they often focus on a single number: glucose. But those same compounds also shift how the body stores fat and maintains muscle, and those shifts can matter a great deal for long-term metabolic health. A 2026 systematic review and network meta-analysis published in Diabetes, Obesity and Metabolism set out to compare several classes of antidiabetic agents side by side, specifically looking at changes in fat mass and lean body mass rather than blood sugar alone.
The analysis pooled data from 41 randomized controlled trials covering 2,906 participants. Because the trials rarely compared every drug against every other drug head-to-head, the researchers used a network meta-analysis design, a statistical method that links indirect comparisons to produce a ranking across all agents at once. The result is one of the most comprehensive looks yet at how these compounds reshape body composition.
What the researchers measured
The two primary outcomes were fat mass, meaning the total weight of adipose tissue in the body, and lean body mass, meaning everything else: muscle, bone, organs, and water. Both were reported as mean differences in kilograms compared with a reference condition, usually placebo or a comparator drug.
Trials were identified by searching PubMed, Web of Science, and Scopus from the earliest available records through March 2025. Only randomized controlled trials made the cut, giving the analysis a relatively high-quality evidence base. The statistical framework used a frequentist random-effects model, which accounts for the fact that individual trials vary in their populations and study designs.
Fat mass reductions across drug classes
Among all the agents studied, the dual GLP-1 and GIP receptor agonist tirzepatide produced the largest reduction in fat mass, with a mean difference of roughly 10.7 kilograms less fat compared with placebo (95% confidence interval: 13.4 to 8.0 kilograms). To put that in perspective, losing roughly 10 kilograms of pure fat tissue is a substantial shift in body composition by any measure in clinical research.
The GLP-1 receptor agonist liraglutide combined with structured exercise came in second. Semaglutide and liraglutide used alone produced moderate fat mass reductions, also statistically significant compared with placebo. These results are consistent with the known mechanism of GLP-1 receptor agonists, which slow gastric emptying and reduce appetite signaling in the brain, leading to lower caloric intake over time.
On the other end of the spectrum, insulin glargine and the DPP-4 inhibitor alogliptin were actually associated with fat mass gain relative to metformin and exenatide. SGLT-2 inhibitors, which work by causing the kidneys to excrete glucose in urine, showed modest fat mass reductions. Metformin, traditional insulin regimens, other DPP-4 inhibitors, and sulfonylureas had minimal impact on fat mass in either direction.
Lean body mass: the trade-off
The picture becomes more nuanced when lean body mass enters the frame. Tirzepatide, which produced the greatest fat loss, also showed the largest reduction in lean mass, a mean difference of roughly 4.4 kilograms less lean tissue compared with placebo (95% confidence interval: 7.6 to 1.2 kilograms). Liraglutide also reduced lean mass significantly, by approximately 1.5 kilograms on average.
This trade-off is not unusual in research on caloric restriction. When total body weight falls quickly, some of that loss comes from lean tissue rather than fat alone. The ratio of fat loss to lean loss is sometimes described as body composition quality of weight loss, and the data here suggest that GLP-1 class agents, while effective at reducing fat, do not fully spare muscle.
SGLT-2 inhibitors caused minor lean mass losses but considerably smaller ones than the GLP-1 agents. Metformin, insulin regimens, and DPP-4 inhibitors showed neutral effects on lean body mass, meaning their use was not associated with meaningful muscle loss in the trials examined.
The role of exercise
One of the more practically relevant findings in the analysis involves the liraglutide plus exercise arm. When structured physical activity was added to liraglutide treatment, fat mass reduction remained strong while the loss of lean body mass was mitigated. This aligns with a large body of exercise science research showing that resistance and aerobic training provide a protective signal for muscle tissue during periods of caloric deficit.
The authors note this finding as a reason to examine lifestyle co-interventions more carefully in future trials. At present, most randomized controlled trials of antidiabetic agents do not standardize physical activity, which means the lean mass losses seen in drug-only arms might look different in real-world settings where participants exercise regularly.
Context and limitations
Network meta-analyses are powerful tools, but they carry assumptions worth understanding. When two drugs have never been tested head-to-head, the method borrows strength from shared comparators, such as placebo or metformin. That indirect evidence is generally less reliable than a direct randomized comparison. The authors used a random-effects framework specifically to acknowledge the variability across trials in populations, doses, and follow-up durations.
The 41 included trials covered 2,906 participants in total, a relatively modest number given how many people worldwide use these drug classes. Trial durations and doses also varied, and the analysis was not able to fully disentangle those variables. The authors explicitly call for further research on dose, duration, and lifestyle influences, acknowledging that the current picture is informative but incomplete.
The literature also does not yet offer long-term data on whether the lean mass reductions observed with GLP-1 and dual GIP/GLP-1 agents translate into functional consequences such as changes in strength, mobility, or metabolic rate. Those outcomes were outside the scope of this particular analysis.
What this means for peptide research
This analysis is useful for anyone following the science of peptides that act on metabolic pathways. GLP-1 receptor agonists and dual GIP/GLP-1 agonists are peptide-derived compounds, meaning they are short chains of amino acids engineered to mimic or enhance naturally occurring gut hormones. Their mechanism, binding to specific receptors in the gut, brain, and pancreas, is fundamentally a peptide biology story.
For researchers interested in the intersection of peptide pharmacology and body composition, the findings highlight two questions worth tracking in future literature. First, can adjunct strategies, whether exercise, nutrition protocols, or co-administered peptides, preserve lean mass during GLP-1-driven fat loss? Second, do the lean mass reductions carry functional significance, or does the net improvement in fat-to-lean ratio still produce favorable metabolic outcomes over longer time horizons? The current systematic review sets a useful baseline for those questions without answering them definitively.
Early data from related research areas points at compounds that support muscle protein signaling, connective tissue remodeling, and growth hormone secretion as potential areas of interest for researchers studying lean mass preservation alongside metabolic intervention. That remains a hypothesis-generating observation rather than a clinical recommendation, and the literature has not yet directly tested such combinations in rigorous controlled settings.
