Strength Training on Ozempic: How to Preserve Muscle Mass

In 2024, Olof Linge and colleagues at AMRA Medical and Lund University ran an MRI sub-study on participants from the STEP 1 trial of semaglutide. They scanned 72 trial completers using full-body MRI at baseline and at the 68-week endpoint, then quantified what kinds of tissue had actually been lost. The headline number, published in The Lancet Diabetes & Endocrinology, was uncomfortable for a field that had been celebrating scale outcomes: approximately 40% of total weight loss in the semaglutide arm came from lean tissue, and approximately 60% from fat.

Forty percent is higher than the ratio typically seen in diet-and-exercise interventions that include structured resistance training and adequate protein, where lean-tissue loss usually runs 20% to 30% of total. It is not, by historical standards, catastrophic — but it is high enough that the obesity medicine field has spent the past two years building consensus around what to do about it. The answer that has consistently emerged is the answer the resistance-training literature has been giving for forty years: lift weights, eat protein, do it consistently.

Why muscle loss matters more on GLP-1 medications

The case for preserving muscle during any weight loss is well established. Lean tissue accounts for the large majority of resting metabolic rate. Muscle is the primary storage and processing site for glucose. Strength predicts functional independence in later life and is independently associated with all-cause mortality across multiple cohort studies. Losing lean tissue means losing metabolic capacity, glucose-handling capacity, and physical capacity simultaneously.

What is specific to GLP-1 treatment is the magnitude of weight loss involved. Patients losing 15% to 22% of body weight on semaglutide or tirzepatide are losing absolute amounts of tissue that previous interventions rarely produced outside bariatric surgery. A patient who starts at 110 kg and loses 22 kg may be losing 8 to 9 kg of lean tissue without focused intervention. That is not a marginal concern. It is a substantial change in body composition that affects future metabolic rate, physical function, and the risk of regaining weight as fat rather than as the original mix of tissues.

The Heymsfield group at Pennington Biomedical Research Center has argued, in commentary published alongside the Linge findings, that body-composition assessment should become standard in obesity-pharmacotherapy monitoring rather than an afterthought. The argument is essentially that the scale alone is no longer a sufficient endpoint when interventions are producing surgery-level losses.

What 40% lean-tissue loss actually means

Lean tissue is not synonymous with skeletal muscle. The MRI quantification includes water associated with glycogen, organ mass, and connective tissue alongside skeletal muscle. Some portion of the lean loss reflects normal adaptation to a smaller body — organs do not scale linearly with weight, but they do shrink modestly with sustained weight loss. The clinically concerning portion is skeletal muscle loss, which is the component that adequate protein and resistance training can directly address.

In trial participants who reported resistance training during semaglutide treatment, the lean-loss percentage was meaningfully lower in observational sub-analyses. The protective effect of training has not yet been definitively quantified in randomised GLP-1 trials, but the broader literature on weight-loss resistance training is unambiguous: in older adults losing weight, resistance training preserves substantially more lean mass than aerobic exercise alone, and adequate protein adds to that effect.

The protein target

Stuart Phillips at McMaster University has spent decades quantifying the protein thresholds for muscle preservation during energy restriction. His systematic review in the American Journal of Clinical Nutrition concluded that during energy-restricted weight loss, protein intake of approximately 1.6 g/kg of body weight per day is needed to limit lean-tissue loss, with diminishing returns above approximately 2.4 g/kg. For a 90 kg patient, this works out to roughly 144 g of protein per day. For a 110 kg patient, closer to 175 g.

The practical problem on GLP-1 medications is that reduced appetite makes hitting these targets harder. A patient eating 1,400 calories per day on semaglutide is in a different position from a non-medicated dieter eating the same amount, because the appetite reduction is intrinsic rather than effortful. The medication does not specifically suppress protein appetite; it suppresses appetite generally. Protein adequacy requires deliberate planning rather than passive consumption.

Most clinical recommendations on GLP-1 treatment converge on prioritising protein first at each meal — eating the protein component before carbohydrates and fats — to ensure intake before satiety signalling cuts the meal short. Protein-dense, lower-volume foods (chicken, fish, Greek yoghurt, eggs, cottage cheese, lean beef) tend to work better than higher-volume options. Protein supplementation through shakes or powders is reasonable when whole-food intake falls short.

Why 1.6 g/kg, not the standard RDA

The standard dietary reference intake for protein in adults — 0.8 g/kg — was set to prevent deficiency in sedentary adults at energy balance. It was not designed for adults in negative energy balance, and it was not designed for adults losing substantial body weight. The 1.6 g/kg target reflects the increased protein need created by energy restriction and the protective role of higher protein intake in preserving lean mass during weight loss. It is roughly double the RDA, which is part of why most non-athletic adults do not hit it without conscious planning.

The resistance training prescription

The training prescription for muscle preservation during weight loss is, by athletic-performance standards, modest. The literature does not require advanced periodisation or high training volumes. Two to three full-body resistance sessions per week, performed consistently across months and years, capture the large majority of the protective benefit.

The core movements that should be present in most protocols are simple: a squat or leg press variation, a hinge (deadlift, hip thrust, Romanian deadlift), a horizontal push (push-up, bench press, dumbbell press), a horizontal pull (row variation), a vertical push (overhead press), and a vertical pull (pulldown or assisted pull-up). Two to three sets of six to twelve repetitions per movement, taken close to but not all the way to failure, is the working range that consistently produces hypertrophy and strength gains across the literature.

What matters more than the specific protocol is sustained execution. Patients who train consistently at moderate intensity preserve more lean mass than patients who train intensively for a few weeks and then stop. The window for muscle preservation during weight loss is the entire weight-loss phase — typically 12 to 18 months on a GLP-1 — not a discrete training block.

Training around side effects

The first weeks of GLP-1 titration are often the worst for training quality. Nausea, fatigue, and reduced food intake can make demanding sessions feel disproportionately hard. The clinically reasonable response is to reduce training volume during titration rather than skipping training entirely. Maintaining the training habit at low intensity is more valuable than abandoning it and trying to restart at full intensity after the side effects settle.

Hydration also matters more than usual. Slowed gastric emptying and reduced spontaneous fluid intake on semaglutide create a real risk of subclinical dehydration that worsens both training performance and recovery. Drinking water through training sessions and across the day is a small change with large effects on how training feels.

Why this matters beyond aesthetics

The case for muscle preservation during GLP-1 treatment is not primarily about appearance. The most important downstream outcome is what happens when treatment is paused or discontinued. Weight regain after stopping semaglutide is documented in the STEP 4 data — Tom Wadden's group reported approximately two-thirds regain within a year. The composition of that regain matters. Patients who entered treatment with intact lean mass and preserved it through the loss phase tend to regain weight in a more favourable ratio than patients who lost substantial lean mass and then regain primarily as fat.

The same logic applies during indefinite treatment. Lean mass preserved across years on a GLP-1 contributes to maintained resting metabolic rate, maintained glucose handling, and maintained physical function. The patient at age 70 who has spent a decade on a GLP-1 will have substantially different functional capacity depending on whether resistance training was part of the treatment plan.

For patients building realistic expectations for treatment, the muscle-preservation work should be framed as a parallel intervention to the medication rather than an optional add-on. The medication shifts the hormonal environment; the training and protein work shift what kind of body the weight loss produces. Both inputs matter, and both are within the patient's control to a meaningful degree.

A starter protocol

For patients new to resistance training, the simplest entry is two full-body sessions per week with the six core movement patterns, two sets each, eight to twelve repetitions, three to five repetitions short of failure. Each session takes approximately 40 minutes once the patient is familiar with the movements. Adding a third session after several months — or substituting one session for a focused leg or upper-body emphasis — is reasonable as tolerance builds. Working with a qualified trainer for the first 4 to 8 weeks reduces injury risk and improves technique substantially.

Protein structure should follow eating opportunities. A protein-first sequence at each meal, with at least 30 g per meal across three to four meals daily, is a workable starting target for most adults. Adjust upward toward the 1.6 g/kg figure as appetite allows.

Key takeaways

• Linge's 2024 MRI sub-study of STEP 1 found approximately 40% of weight loss on semaglutide came from lean tissue, higher than the 20–30% typical of diet-and-resistance-training protocols.
• Phillips' protein threshold for limiting lean-tissue loss during energy restriction is approximately 1.6 g/kg/day — roughly double the standard RDA.
• Two to three full-body resistance training sessions per week, performed consistently across months and years, capture the large majority of the muscle-preservation benefit.
• Reduced appetite on GLP-1 medications makes protein targets harder to hit; protein-first sequencing and protein-dense foods are practical solutions.
• Training intensity should be reduced rather than abandoned during titration; maintaining the habit at low intensity outperforms stopping and restarting.
• The preserved lean mass matters most after treatment is paused or discontinued, when the composition of regained tissue is influenced by what was preserved during the loss phase.