GLP-1 and exercise: preserving strength beyond weight loss

GLP-1 weight loss changes more than the scale. Read the evidence on lean mass, everyday strength, exercise and what happens after treatment ends.

日本語で読む（Japanese version）

The number on the bathroom scale is falling, and clothes that once felt tight now fit comfortably. Yet the stairs at the station still leave you out of breath. It is tempting to ask whether exercise has finished its job now that medication is helping your weight come down. A lighter body and a body that can do more are different outcomes. The scale reports only one of them.

Shopping bags still need carrying home. Chairs still need getting out of, and the journey through the station still involves moving your own body. These ordinary tasks give the question a practical shape: what do you want to keep as weight changes? Feeling reassured by a smaller number does not automatically answer whether strength, fitness and everyday movement are being maintained.

This feature examines exercise during medication-assisted weight loss and beyond, rather than how to use a drug. For a closer look at measurement, see our feature on body-fat measurements; the wider discussion of muscle and nutrition in middle age belongs in our sarcopenia and protein feature. Here, the focus is what the lost weight contains, what trials combining exercise and medication found, and how to read those results without turning them into promises.

When weight falls, what has actually been lost?

GLP-1 receptor agonists are a group of medicines that act on the receptor through which the substance GLP-1 exerts its effects. Related therapies can act on other receptors as well. A review of muscle health describes reduced energy intake and slower emptying of the stomach among the main routes through which GLP-1 receptor agonists produce weight loss. That describes the treatment’s effect on weight; it does not tell us, by itself, what happens to muscle strength or fitness. [10]

Body weight includes everything on the scale. It has no setting that labels the change as fat alone. A systematic review and network meta-analysis combined 22 randomized controlled trials involving 2,258 adults with diabetes and/or overweight or obesity. Compared with the control treatments, the average difference was −3.55 kg for body weight, −2.95 kg for fat mass and −0.86 kg for lean mass. The findings show a reduction in fat alongside a reduction in tissue classified as lean. Effects differed by drug: liraglutide reduced weight without a significant reduction in lean mass in this analysis. [12]

The 95% confidence intervals, which describe uncertainty around the estimates, were −4.81 to −2.29 kg for body weight, −4.11 to −1.79 kg for fat mass and −1.30 to −0.42 kg for lean mass. These are pooled comparisons for different outcomes. We cannot add and subtract them to reconstruct an individual person’s weight loss. [12]

The next distinction matters just as much: lean mass is not another name for skeletal muscle. A 2024 review explains that the measure includes organs, bone and fluids as well as muscle, with water in fat tissue also relevant. A measured reduction in lean mass therefore cannot be read as an identical weight of skeletal muscle being lost. A headline can sound much more alarming when it quietly replaces one term with the other. [4]

This is also why the word “muscle” on a body-composition display needs context. Before deciding what a number means, ask what the device or study actually measured. Reviews call for assessment of muscle quantity together with composition, strength, mobility and physical function. A weight-based measure answers a question about quantity; it does not settle all those other questions. [4][10]

A useful reading order is to separate body weight, fat mass and the amount classified as lean, then look for evidence about force and movement. Can the person still produce strength? Has physical performance changed? This makes room to recognize weight loss while taking the body’s capabilities seriously. The results need not fit into a single verdict of success or failure. The studies discussed below are useful precisely because some of them measured more than the weight of the participants. [4][17]

A share of lost weight is not a share of your original muscle

A 2026 meta-analysis of 20 randomized controlled trials included 15,782 adults with overweight or obesity. It reported that lean mass represented 25–39% of the weight lost with incretin-based therapies. Incretins are hormones derived from the gut and released in response to nutrients. The finding concerns the composition of weight loss, and understanding its denominator is essential before interpreting what it says about muscle. [1][2]

The denominator in this result is the weight that was lost. It is not the total amount of muscle present before treatment. Saying that lean mass accounted for roughly a third of lost weight is different from saying that roughly a third of someone’s original muscle disappeared. The latter changes both the denominator and the type of tissue being discussed. [2][4]

In the same meta-analysis, intensive lifestyle interventions produced a lean-mass share of 26.2%, with a 95% confidence interval of 24.1–28.3%. The comparison with medication-based treatment was not statistically significant. Lean mass forming part of lost weight therefore cannot be presented as a phenomenon unique to the drugs. It is an issue to consider during substantial weight reduction across methods, while still paying attention to the details of the intervention and the people studied. [2]

Lifestyle intervention combined with resistance training had a more favorable profile: 17.5% of lost weight was lean mass, with a confidence interval of 14.2–20.8%. Resistance training means producing force against resistance, commonly called strength training. This result provides a reason to consider that form of activity during weight loss. It does not guarantee that every person taking an incretin-based medicine will achieve the same proportion simply by adding some strength exercise. [2]

There is also considerable variation in the literature. Another review discussed studies in which lean mass accounted for 40–60% of total weight loss, alongside studies reporting approximately 15% or less. Differences in the populations, the drugs and coexisting illnesses may contribute to the variation. Choosing only the largest figure creates one misleading picture; choosing only the smallest creates another. [4]

For readers, the practical question is “a percentage of what?” It may refer to lost weight, starting weight or the composition of the body at a particular time. Those are not interchangeable. A change in the proportion occupied by muscle also differs from an increase in its absolute amount. Before comparing headlines, keep the starting quantity, the lost quantity and the remaining proportion separate. [2][4]

Does less muscle necessarily mean less strength?

Some reviews emphasize the concern about lean-mass loss. A 2024 narrative review discusses the rapid decline that can accompany substantial weight loss with these treatments and proposes individualized resistance training to help preserve muscle. A narrative review brings together and interprets research. Its recommendation should not be mistaken for a new randomized trial showing exactly what that recommendation achieved in people taking medication. [1]

Other reviews argue that a decline in muscle quantity need not mean worse health or function. Drawing on imaging research, a 2024 review describes muscle changes that may be proportionate to the weight lost. Less fat infiltrating muscle may also contribute to improved muscle quality. This interpretation asks us to examine what the muscle is like and what it can do, alongside how much of it is measured. It does not erase the change in quantity; it places it within a broader assessment. [4]

A 2025 review similarly reports that fat-mass reductions exceed lean-mass reductions and that consistent deterioration in strength or physical function has not been observed. It also distinguishes improvements in muscle quality from muscle enlargement. These are different outcomes. A reader should not turn the discussion of quality into a claim that medication-assisted weight loss makes muscles grow, just as a fall in lean mass should not automatically be turned into a diagnosis of declining function. [10]

Evidence from a specific patient population adds another perspective. A systematic review included 12 studies involving 810 adults with metabolic dysfunction-associated steatotic liver disease, a fatty liver condition associated with metabolic dysfunction. Changes in muscle mass were generally proportionate to weight loss. Among the 3 studies assessing strength, involving 477 participants, there was no evidence that handgrip strength or sit-to-stand performance deteriorated. [15]

That review also had clear limitations. The population was restricted, intervention and observational studies were mixed, and assessment methods varied. Few studies examined muscle quality, while small samples and inconsistent assessments limited interpretation. “No evidence of deterioration” consequently cannot become “deterioration never happens.” [15]

The apparently conflicting messages become easier to reconcile when we separate quantity from function. A reduction in measured lean mass and preservation of strength can occur together. Reviews concerned about loss and reviews discussing adaptation are not necessarily reporting mutually exclusive events. They direct attention to different parts of muscle health. [4][10][15]

Getting out of a work chair, carrying bags and using station stairs can provide concrete examples to remember when discussing changes with a doctor. They are everyday observations, not a diagnosis of muscle loss. Research uses defined assessments, and a personal impression does not reproduce them. Still, describing an actual task that has become harder or easier gives the conversation information that a weight entry alone cannot supply. The distinction is between useful observations and standardized functional testing. [4][15]

Exercise has outcomes beyond the scale

A randomized trial gives a more direct view of exercise alongside medication. A secondary analysis examined 193 adults with obesity, aged 18–65 years, who did not have diabetes. After losing weight on a low-energy diet, participants entered a maintenance phase comparing exercise, liraglutide, their combination, and placebo with usual activity. The question here was what happened to physical fitness during maintenance, rather than whether exercise could replace a prescription or produce the same effect in every reader. [17]

The exercise program combined group sessions involving indoor cycling and circuit training with individual moderate-to-vigorous activity. Participants in the exercise groups completed a median of 2.65 sessions and 116 minutes per week. These numbers describe what participants actually did in a structured program. A median is not a universal training prescription. [17]

Compared with liraglutide alone, the combined group improved its stair-climb test time by 1.2 seconds, with a 95% confidence interval of 0.6–1.9 seconds. Peak oxygen uptake normalized to fat-free mass, a measure of the capacity to take in oxygen during exercise, also improved. Exercise alone produced similar benefits, whereas medication alone did not improve physical fitness. That unfavorable finding for medication alone is central to the question: an effect on weight cannot simply be assumed to deliver an effect on fitness. [17]

Strength measurements require their own distinction. Absolute strength was preserved while body weight fell, changing strength relative to body weight. Relative-strength changes were −7.8% in the placebo group, −0.4% with exercise, +1.0% with liraglutide alone and +3.3% with the combination. The latter groups all had higher relative strength than placebo. A ratio can improve because the body being supported is lighter, even when the absolute force produced has not increased. [17]

The scale cannot write the result of the stair test on the study’s behalf. Researchers saw the fitness benefit because they measured it separately. Thinking of exercise only as a way to remove more weight misses this aspect of the trial. For someone whose weight is already falling, the relevant question may be how their body performs while carrying that lower weight. That makes exercise a distinct part of the discussion during treatment, rather than an activity whose purpose disappears once weight loss begins. [17]

Numbers: GLP-1 weight loss and exercise

22 trials, 2,258 participants: the systematic review and network meta-analysis examining body weight, fat mass and lean mass. [12]

−0.86 kg: the average lean-mass difference compared with control treatments, not a measurement of muscle alone. [12][4]

26.2%: the share of lost weight classified as lean mass with lifestyle intervention. [2]

17.5%: the corresponding share when lifestyle intervention included resistance training. [2]

1.2 seconds: the difference in stair-test time changes between combination treatment and medication alone. [17]

109 participants: the number attending the post-treatment assessment, which makes follow-up participation part of the interpretation. [13]