New Study: Tirzepatide Causes More Muscle Loss Than Semaglutide in 8,000-Patient Analysis

A real-world analysis conducted by the data analytics firm nference, published as a preprint on medRxiv in April 2026, compared lean body mass changes in approximately 8,000 patients taking either tirzepatide (~1,800 patients) or semaglutide (~6,200 patients).

The headline finding: tirzepatide users lost more lean body mass than semaglutide users at every time point measured. At 3 months, tirzepatide users lost 1.1% more lean mass. At 6 months, the gap was 1.5%. At 9 months, 1.3%. At 12 months, 2.0%.

The high-loss group: among patients who achieved the most dramatic weight loss (20% or more of total body weight), approximately 1 in 10 tirzepatide users lost more than 5% of their lean body mass — a threshold considered clinically meaningful because it can impair physical function and metabolic health. Fewer than 7% of semaglutide users who achieved similar total weight loss crossed this threshold.

How they measured it. Patients were tracked using either low-radiation body composition scans (such as DEXA) or bioelectrical impedance "smart" scales that estimate body fat, muscle mass, bone mass, and other components. The data came from real-world clinical records across a federated health network, giving the study scale but also introducing measurement variability.

Lean body mass — primarily muscle, but also organ tissue and bone — is not just cosmetically important. It is metabolically active tissue that drives resting energy expenditure, insulin sensitivity, physical function, and long-term health outcomes.

The metabolic rate problem. When you lose muscle, your basal metabolic rate decreases. This means you burn fewer calories at rest, which makes weight regain more likely if you ever discontinue the drug or reduce the dose. The more lean mass lost during treatment, the lower the metabolic "floor" — and the harder it becomes to maintain weight loss independently.

The functional concern. For older adults (the population most likely to be prescribed GLP-1 drugs for metabolic disease), muscle loss can directly impair physical function — grip strength, gait speed, balance, and fall risk. Sarcopenia (age-related muscle loss) is already a major cause of disability in older adults. Adding drug-induced muscle loss to age-related muscle loss compounds the problem.

The body composition paradox. Tirzepatide produces 50% more total weight loss than semaglutide (20.2% vs 13.7% in SURMOUNT-5). But if a larger proportion of that additional weight loss is coming from lean tissue rather than fat, the net metabolic benefit may be smaller than the total weight loss number suggests. The question is not just "how much weight did you lose?" but "what kind of weight did you lose?"

Who is most at risk. The study found that decreased exercise tolerance during treatment was linked with greater lean mass loss — and this effect was more pronounced in tirzepatide patients. This suggests that patients who become less physically active during treatment (perhaps due to GI side effects, fatigue, or reduced caloric intake) face compounded muscle loss. Higher doses, longer treatment duration, and pre-existing musculoskeletal pain were also risk factors.

The study identified the difference but could not explain it mechanistically. Several hypotheses exist, none proven.

Hypothesis 1: Greater caloric deficit. The simplest explanation is that tirzepatide produces more total weight loss because it causes a larger caloric deficit — and any caloric deficit, regardless of mechanism, will cause some lean mass loss. Greater deficit = greater lean mass loss. Under this theory, the difference is not specific to tirzepatide's pharmacology but simply a consequence of more aggressive appetite suppression.

Hypothesis 2: GIP receptor effects. Tirzepatide activates both GLP-1 and GIP receptors, while semaglutide activates only GLP-1. The GIP receptor is expressed in adipose tissue and may have effects on fat-muscle metabolic cross-talk that are not yet well understood. It is possible that dual receptor activation changes the body's preferential fuel partitioning in ways that lead to proportionally more muscle catabolism. This is speculative.

Hypothesis 3: Differential exercise tolerance. The study found that reduced exercise tolerance was more strongly associated with lean mass loss in tirzepatide users. If tirzepatide's greater appetite suppression leads to lower energy availability for physical activity, patients may be less able to maintain the resistance training needed to preserve muscle. This is an indirect pharmacological effect rather than a direct one.

Hypothesis 4: Study design artifacts. The tirzepatide group was smaller (~1,800 vs ~6,200), and the body composition measurements were heterogeneous (DEXA scans vs smart scales). Selection bias — physicians may prescribe tirzepatide to patients with higher BMI or more metabolic disease — could also confound the comparison. Randomized controlled trials with standardized DEXA measurements would be needed to rule out these artifacts.

This study does not mean tirzepatide is inferior to semaglutide. The overall evidence profile still strongly favors tirzepatide for total efficacy.

SURMOUNT-5 head-to-head results remain definitive. In the randomized Phase 3 trial, tirzepatide produced 20.2% weight loss vs 13.7% for semaglutide, with greater waist circumference reduction and lower GI discontinuation rates. No observational preprint overrides these results.

All significant weight loss causes lean mass loss. This is true for surgery, caloric restriction, exercise-only programs, and every pharmacological intervention. The SURMOUNT-1 body composition substudy showed that approximately 25–40% of total weight lost on tirzepatide was lean mass — a ratio that is broadly consistent with other weight loss interventions. The question raised by the nference study is whether this ratio is slightly worse for tirzepatide than for semaglutide, not whether lean mass loss occurs.

Net metabolic benefit still favors tirzepatide. Even accounting for greater lean mass loss, tirzepatide users still achieve larger improvements in waist circumference, HbA1c, blood pressure, and lipid profiles compared to semaglutide users. The metabolic benefits of losing 20% body weight outweigh the metabolic costs of the accompanying lean mass reduction — as long as patients are actively working to preserve muscle.

The clinical takeaway. This study does not change prescribing recommendations, but it reinforces that lean mass preservation should be an active treatment goal — not an afterthought — for every patient on GLP-1 therapy, and especially for those on tirzepatide at higher doses.

The evidence for muscle preservation during GLP-1-mediated weight loss converges on three strategies.

Resistance training. Progressive resistance training 2–3 times per week is the single most effective intervention for preserving lean mass during caloric deficit. A meta-analysis of exercise interventions during pharmacological weight loss shows that resistance training can reduce lean mass loss by 25–50% compared to no exercise. This should be structured training with progressive overload — not walking or light cardio, which do not provide sufficient stimulus for muscle protein synthesis.

Protein intake. The evidence supports 1.2–1.6 grams of protein per kilogram of body weight daily during GLP-1 therapy. Because these drugs reduce appetite and total food intake, patients need to prioritize protein at every meal to hit these targets. A 180 lb (82 kg) person on tirzepatide should aim for approximately 100–130 grams of protein daily. Protein timing matters less than total daily intake, but distributing protein across 3–4 meals optimizes muscle protein synthesis.

Monitoring body composition. Body weight alone does not distinguish fat loss from muscle loss. Patients on tirzepatide — especially those losing weight rapidly — should consider periodic body composition measurements (DEXA scan or bioelectrical impedance) to track lean mass changes. If lean mass is declining disproportionately, the response should be to increase resistance training intensity and protein intake, not necessarily to change medications.

Dose optimization. Some clinicians are reconsidering whether all patients need the maximum 15mg tirzepatide dose. In SURMOUNT-1, the 10mg dose produced 19.5% weight loss vs 20.9% at 15mg — a modest difference that might come with meaningfully less lean mass loss. For patients who have achieved satisfactory weight loss at 10mg, the additional 1.4 percentage points of weight loss at 15mg may not justify the potential for greater muscle impact.

This is an important signal, but it is not definitive evidence. Several limitations should temper interpretation.

Not peer-reviewed. The study was published as a preprint on medRxiv, meaning it has not undergone formal peer review. Preprints are released to share findings rapidly, but they have not been vetted by independent experts for methodological rigor, statistical errors, or interpretive overreach. The findings may change during the review process.

Observational design. Patients were not randomized to tirzepatide vs semaglutide. Physicians chose which drug to prescribe based on patient characteristics, insurance coverage, and clinical judgment. Patients prescribed tirzepatide may have had higher baseline BMI, more metabolic disease, or other factors that independently affect lean mass. Without randomization, confounders cannot be fully controlled.

Heterogeneous measurements. Body composition was measured using a mix of DEXA scans and consumer-grade smart scales. Smart scales estimate lean mass using bioelectrical impedance, which is significantly less accurate than DEXA — particularly in patients with changing fluid status (common during GLP-1 therapy due to reduced food volume and changes in hydration). Measurement error could systematically differ between the two drug groups if one group had more DEXA measurements.

Missing lifestyle data. The study did not control for diet quality, protein intake, or structured exercise — the three most important determinants of lean mass preservation during weight loss. If tirzepatide patients happened to exercise less or eat less protein (perhaps because of greater appetite suppression), that alone could explain the lean mass difference without any pharmacological mechanism.

What would be definitive. A randomized controlled trial comparing tirzepatide vs semaglutide with standardized DEXA measurements at multiple time points, controlled protein intake, and structured exercise programs would definitively answer whether tirzepatide causes more lean mass loss when lifestyle factors are equalized. Several such trials are currently in progress.