Resistance training materially preserves lean mass during GLP-1 treatment. The Karakasis 2025 meta-analysis found GLP-1 monotherapy loses approximately 25% of total weight as lean tissue; exercise-combined arms cut that fraction to roughly 10–15%. On a 15 kg semaglutide weight-loss course, structured training can prevent 2–4 kg of lean tissue loss, a 60–70% reduction in lean mass attrition.
What Is the Measured Lean Mass Loss Fraction on GLP-1 Therapy Without Structured Exercise?
The Karakasis 2025 meta-analysis in Metabolism, pooling 22 randomised trials, found GLP-1 receptor agonists reduced lean mass by approximately 25% relative to total weight lost. The Santić 2026 MDPI Metabolites review reports semaglutide-specific DXA data showing 26% of weight lost as lean soft tissue. Both figures reflect uncontrolled real-world behaviour: no exercise mandate, no protein floor.
The 26–40% lean-loss range across published semaglutide trials reflects heterogeneity in protein intake and physical activity rather than a fixed pharmacological effect. STEP 1 body composition sub-analyses sit at the upper end of this range. The SEMALEAN study (Alissou and colleagues, PMC 2025) reported a 12-month lean mass decline of approximately 11% in a cohort where 59% of participants achieved 10% or greater total weight loss.
Handgrip strength in SEMALEAN increased by approximately four kilograms at 12 months, suggesting functional muscle quality improved even as lean mass volume declined. Tirzepatide's lean-loss fraction is lower than semaglutide's in DXA sub-studies at approximately 17–25% of total weight lost, attributable to GIP receptor co-agonism attenuating glucagon-driven protein catabolism. A 2026 medRxiv preprint found tirzepatide produced greater absolute lean-body-mass decline than semaglutide in routine care because its larger total weight loss amplifies the absolute lean-tissue cost.
How Large Is the Lean Mass Preservation Effect of Adding Resistance Training to GLP-1 Therapy?
The van den Hoek DMJ 2026 study found semaglutide alone reduced fat mass by 31% but also lean mass by 11%. Adding exercise further reduced fat mass to 45% while attenuating lean mass loss. Rossi and colleagues (PMC 2025) synthesised human evidence showing resistance training can reduce the lean-loss fraction from approximately 25% to below 10% of total weight lost.
In absolute terms, a 15 kg total weight loss on semaglutide without exercise produces approximately 4–6 kg of lean mass loss. Resistance training, based on the Rossi 2025 synthesis, reduces that to roughly 1–2 kg. At approximately 13 kcal per kilogram of lean mass per day, preserving 3 kg of lean tissue protects approximately 39 kcal/day of resting metabolic rate.
That resting metabolic rate buffer is a meaningful defence against post-therapy weight regain. The Tinsley 2025 PMC cohort study of GLP-1 users engaged in intentional resistance training 3–5 days per week found that some participants increased lean soft tissue despite ongoing total weight reduction. Two participants showed net lean mass gains, establishing that lean mass gain during GLP-1-driven weight loss is mechanistically achievable with sufficient training stimulus.
Does Resistance Training During GLP-1 Therapy Preserve Functional Strength, Not Just DXA Lean Mass?
Functional strength outcomes diverge from DXA lean mass in GLP-1 trials. The SEMALEAN study found handgrip strength rose by approximately four kilograms at 12 months and sarcopenic obesity prevalence fell from 49% to 33%, despite a measurable decline in absolute lean mass volume. Codella and colleagues (PMC 2025) confirm resistance training preserves physical function beyond what DXA numbers alone capture.
The discordance between declining lean mass volume and improving functional strength is explained by two mechanisms. First, fat infiltration of muscle decreases during GLP-1-driven fat loss, improving muscle quality per unit of lean mass. Second, resistance training increases neuromuscular efficiency and motor unit recruitment independently of hypertrophy. Strength gains can therefore occur even when lean mass volume is static or slightly declining.
The Linge 2024 Circulation primer explicitly addresses this distinction, noting that absolute lean mass loss during GLP-1 therapy does not necessarily indicate maladaptive muscle change when functional markers are tracked concurrently. For performance-oriented users, DXA lean mass alone is an incomplete outcome metric. Handgrip strength, leg press 1RM, and 6-minute walk test distance provide complementary functional data that DXA cannot capture.
What Does the LEAN-PREP Trial Design Reveal About the Current Evidence Gap?
The LEAN-PREP trial (Alawadhi and colleagues, BMJ Open 2026) is a 6-month, 232-participant RCT using a 2x2 factorial design: resistance exercise versus no exercise, crossed with protein supplementation versus no supplementation, in adults on semaglutide or tirzepatide. Its 2026 publication as a protocol confirms no completed factorial RCT with DXA lean mass as the primary endpoint yet exists.
The LEAN-PREP design isolates the independent and combined effects of resistance training and protein intake on lean mass during GLP-1 therapy. DXA is the primary endpoint; physical function is a secondary endpoint. Its 232-participant sample provides 80% power to detect a 1–2 kg lean mass difference between arms, a threshold representing approximately one-third of mean absolute lean mass loss in uncontrolled GLP-1 trials.
Until LEAN-PREP reports, the evidence base rests on one large meta-analysis (Karakasis 2025, 22 RCTs), one preclinical mechanistic study with exercise arms (van den Hoek DMJ 2026), and one observational cohort with exercise data (Tinsley 2025). The effect direction is consistent across all sources. The precise magnitude remains meta-analytic rather than factorial-trial confirmed.
What Training Volume and Frequency Produce Measurable Lean Mass Protection in GLP-1 Users?
Rossi and colleagues (PMC 2025) identify a threshold at 2 resistance training sessions per week targeting major muscle groups. Below this frequency, lean mass protection is not reliably observed. Three sessions per week is the inflection point where the lean-loss fraction drops below 15% of total weight lost, and intensity at 60% or more of one-repetition maximum is required.
The medrxiv 2025 interrupted time-series analysis hypothesised that resistance training's effect on lean mass over time would be non-inferior to the GLP-1 agonist's catabolic pressure. This framing is operationally useful: the training dose needed to preserve lean mass is not trivial, and low-frequency or low-intensity programmes are unlikely to offset GLP-1-driven lean tissue loss. Minimum effective dose requires at least 2–3 sessions per week at moderate-to-high intensity.
Compound multi-joint movements — squat, hip hinge, horizontal and vertical press, horizontal and vertical pull — recruit the largest muscle mass per session and generate the strongest systemic anabolic hormone response. For GLP-1 users whose appetite suppression limits total energy availability, compound-movement protocols maximise the mechanical mTORC1 stimulus per unit of training time. Isolation exercises contribute marginal additional lean mass protection beyond what compound training achieves in this context.
Does Resistance Training During GLP-1 Therapy Also Protect Bone Density and Connective Tissue?
Bone mineral density loss is a documented secondary consequence of GLP-1-driven rapid weight loss, independent of lean mass changes. Codella and colleagues (PMC 2025) confirm resistance training attenuates bone density decline through mechanical loading, a stimulus pharmacotherapy alone cannot replicate. This bone-protective effect is additive to lean mass preservation and represents a distinct, non-redundant benefit of structured training.
The mechanism is osteogenic loading: compressive forces on bone during resistance training stimulate osteoblast activity via Wnt/beta-catenin signalling and suppress osteoclast differentiation via OPG/RANKL ratio shifts. GLP-1 receptor agonists have a direct but modest osteogenic effect through receptor expression on osteoblasts. Resistance training's mechanical loading signal is substantially larger in magnitude and operates through a distinct pathway, making the two signals additive rather than redundant.
Connective tissue, including tendons and ligaments, also benefits from resistance training during GLP-1 therapy. Rapid weight loss without mechanical loading reduces tendon cross-sectional area and collagen synthesis rates. Resistance training maintains mechanical stress on tendons, preserving collagen turnover and tensile strength. This connective tissue benefit is not captured by DXA lean mass measurements and represents an additional functional preservation outcome that lean mass metrics systematically undercount.
How Should Practitioners Grade the 2026 Evidence Quality for Resistance Training During GLP-1 Therapy?
The 2026 evidence base is mechanistically strong and directionally consistent but structurally limited. The Karakasis 2025 meta-analysis provides the strongest quantitative estimate — lean-loss fraction reduced from approximately 25% to 10–15% — but included trials were not designed to isolate resistance training as the independent variable. Evidence grade: consistent observational signal, no factorial RCT confirmation.
The Santić 2026 MDPI review explicitly grades the resistance training evidence as promising but not yet definitive, noting that two independent 2025 meta-analyses confirm the directional benefit while acknowledging heterogeneous exercise protocols across included trials. The LEAN-PREP trial, when completed, will provide the first purpose-designed factorial evidence. Until then, practitioners should treat the 10–15% lean-loss fraction as a plausible target rather than a confirmed outcome.
For cross-site context: the network meta-analysis showing GLP-1 plus resistance training ranked first on weight, fat mass, and HOMA-IR outcomes is covered at Does Adding Resistance Training to GLP-1 Treatment Improve Fat Mass and Insulin Sensitivity More Than GLP-1 Alone in 2026? The mechanistic basis of mTORC1 activation via mechanical loading is detailed at Does Resistance Training Fundamentally Change How GLP-1 Therapies Like Semaglutide Affect Lean Mass in 2026? For clinical protocol design, the semaglutide body composition protocol on Peptides Plus provides practitioner-level structure. The mechanistic evidence on GLP-1 receptor agonist molecular pathways is reviewed at What Does the 2026 Systems Medicine View of Semaglutide Reveal About Its Inflammatory, Lipid, and ECM Pathways? Which GLP-1/GIP Combination Peptide Protocols Best Preserve Lean Mass While Improving Glycemic Control in 2026 Self-Experimentation? What Do 2026 Primary Studies Show About GLP-1/GIP Dual Agonists Versus GLP-1 Monotherapy for Body-Weight Loss and Cardiometabolic Outcomes? Does Retatrutide's Phase 3 Evidence in 2026 Show Clinically Meaningful Weight Loss Beyond Semaglutide and Tirzepatide?