Introduction
Weight management is a cornerstone in type 2 diabetes care, as sustained weight loss can modify the disease course, improve glycaemic control, and decrease cardiovascular risk. However, significant weight reduction is often accompanied by loss of muscle mass, potentially impairing strength and physical function, a concern especially in older adults at risk of sarcopenia. Muscle quality—comprising muscle volume and myosteatosis (fat infiltration)—is a key determinant of functional outcomes. Understanding the impact of antidiabetic therapies on muscle composition is critical, particularly for novel agents inducing substantial weight loss.
Tirzepatide, a dual glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 receptor agonist (GLP-1 RA), has demonstrated marked weight and fat mass reductions in type 2 diabetes. The SURPASS-3 trial previously confirmed its superiority over insulin degludec for glycaemic control and body fat distribution improvements. This post-hoc analysis from the SURPASS-3 MRI substudy investigated tirzepatide’s association with skeletal muscle composition—specifically thigh muscle volume, muscle volume Z score (adjusted for sex, height, weight, and BMI), and muscle fat infiltration—over 52 weeks, contextualizing findings against population-based data from UK Biobank.
Study Design and Methods
The SURPASS-3 MRI substudy was an international, randomized, open-label, phase 3 trial conducted at 45 centers across eight countries. Insulin-naive adults (≥18 years) with type 2 diabetes on metformin ± SGLT-2 inhibitor, HbA1c 7.0–10.5% (53–91 mmol/mol), BMI ≥25 kg/m2, and fatty liver index ≥60 were enrolled. Participants (N=296) were randomized 1:1:1:1 to weekly tirzepatide (5 mg, 10 mg, 15 mg) or once-daily insulin degludec. The substudy focused on participants with valid MRI scans at baseline and week 52 (N=246).
MRI quantified thigh muscle parameters using validated automated multiatlas segmentation: muscle volume (fat-free muscle volume; voxels <50% fat fraction), muscle fat infiltration (mean fat fraction in viable muscle tissue), and muscle volume Z score reflecting deviations from matched normative data from UK Biobank. Longitudinal MRI data from 2942 UK Biobank participants with repeated imaging approximately two years apart were used to generate population-based models predicting muscle composition changes relative to bodyweight changes.
Statistical analyses included paired t-tests comparing baseline to week 52 changes within groups, ANCOVA adjusted for baseline variables comparing tirzepatide pooled dose groups versus insulin degludec, and paired t-tests comparing observed changes with population-based expected changes. Subgroup analyses examined treatment effects by sex, age, baseline muscle composition status, and SGLT-2 inhibitor use. Correlations between muscle composition changes, body weight, and biomarkers were also evaluated.
Key Findings
Of 246 analyzed participants, the mean age was 56.0 years, median diabetes duration 6.7 years, mean BMI 33.4 kg/m2, with 60% male. Baseline muscle fat infiltration, volume, and volume Z scores were similar between tirzepatide and insulin degludec groups.
Tirzepatide treatment (all doses pooled) induced significant reductions from baseline to week 52 in muscle fat infiltration (mean –0.36 percentage points, p<0.0001), muscle volume (–0.64 L, p<0.0001), and muscle volume Z score (–0.22 SD units, p<0.0001), concurrent with substantial weight loss (~10% reduction). Dose-specific analyses confirmed significant changes across 5 mg, 10 mg, and 15 mg doses. In contrast, insulin degludec was associated with modest weight gain (3.3%) and a slight increase in muscle volume (0.16 L, p=0.043), without significant changes in muscle fat or volume Z score.
Comparing tirzepatide with insulin degludec revealed statistically significant differences across all three muscle composition parameters (p<0.01). Correlative analyses showed muscle volume changes strongly associated with weight changes in tirzepatide-treated patients, whereas muscle fat infiltration and volume Z score exhibited weaker but significant correlations.
Contextualizing these findings with UK Biobank data demonstrated that tirzepatide-induced reductions in muscle volume aligned with population-level expectations relative to weight loss (mean difference –0.04 L, p=0.22), indicating an adaptive response rather than accelerated muscle loss. Notably, tirzepatide exerted significantly greater reductions in muscle fat infiltration compared to expected population-based changes (mean difference –0.42 percentage points, p<0.0001), suggesting a potentially direct beneficial effect on myosteatosis. The 15 mg tirzepatide dose also induced significantly larger reductions in muscle volume Z score than expected (mean difference –0.18 SD, p=0.0016).
Subgroup analyses revealed that females and older participants (≥60 years) experienced greater weight loss with tirzepatide, but no significant heterogeneity in muscle composition changes was noted across sex, baseline muscle composition status, or age groups, except a somewhat greater reduction in muscle volume Z score in older participants.
Muscle fat infiltration changes correlated with metabolic biomarkers, including triglycerides, HDL cholesterol, adiponectin, and free fatty acids, particularly in the tirzepatide group, reinforcing the metabolic relevance of muscle composition modulation.
Expert Commentary
This pioneering analysis employing gold-standard MRI techniques elucidates tirzepatide’s impact on muscle health in type 2 diabetes patients undergoing weight loss, a profile distinct from basal insulin therapy. The concurrent fat infiltration improvement and muscle volume decrease align with an adaptive physiological remodeling rather than pathological muscle loss.
Mechanistically, tirzepatide combines GIP and GLP-1 receptor agonism, enhancing glycaemic control, promoting weight and fat mass reduction, and possibly exerting direct muscle metabolic effects. Preclinical evidence suggests GLP-1 receptor agonists improve muscle glucose uptake and microvascular flow, potentially preserving muscle function. The distinct reductions in myosteatosis, exceeding expected weight-loss effects, hint at favorable muscle tissue remodeling beyond pure caloric deficit.
Despite significant methodological strengths, including robust MRI quantification and large multinational sampling, limitations include lack of functional muscle assessments, lifestyle interference control, and absence of direct comparison with selective GLP-1 agonists. Different MRI scanners and field strengths could introduce measurement variability, although prior validations mitigate this concern. The population was predominantly White with moderate obesity and hepatic steatosis risk, potentially limiting generalizability.
Clinical implications underscore tirzepatide as an effective agent for weight and metabolic control without disproportionate muscle loss risk. Future research integrating muscle function measures and longer-term outcomes will clarify the functional significance of these muscle composition changes.
Conclusion
In adults with type 2 diabetes and overweight or obesity, tirzepatide treatment for 52 weeks was associated with significant reductions in thigh muscle fat infiltration and muscle volume consistent with adaptive remodeling amidst substantial weight loss. Compared with insulin degludec, tirzepatide favorably modulated muscle composition, potentially enhancing muscle health. These findings enrich the therapeutic understanding of tirzepatide’s metabolic effects and support its role in integrated diabetes and obesity management. Further studies are warranted to link muscle compositional changes with clinical functional outcomes and long-term benefits.
References
Sattar N, Neeland IJ, Dahlqvist Leinhard O, Fernández Landó L, Bray R, Linge J, Rodriguez A. Tirzepatide and muscle composition changes in people with type 2 diabetes (SURPASS-3 MRI): a post-hoc analysis of a randomised, open-label, parallel-group, phase 3 trial. Lancet Diabetes Endocrinol. 2025 Jun;13(6):482-493. doi: 10.1016/S2213-8587(25)00027-0 IF: 41.8 Q1 . Epub 2025 Apr 30. PMID: 40318682 IF: 41.8 Q1 .