Exercise Plus a GLP-1: What Each One Actually Changes

A randomized trial split people who had lost weight by dieting into four groups: exercise, a GLP-1 drug, both, or neither. On physical fitness the answer was unambiguous and not the one the drug’s reputation implies. The drug alone did not improve fitness at all. Exercise did. Both together did most.^[1]

What changed, and in whom

Compared with liraglutide alone, the combined group completed a stair-climb test 1.2 seconds faster — an 8.6% improvement — and improved peak oxygen consumption by 3.0 mL/min per kg of fat-free mass. Exercise alone produced similar benefits. Liraglutide alone produced none.^[1]

The exercise was real: a median of 2.65 sessions a week, 116 minutes weekly, at 79% of maximum heart rate. That is structured training rather than walking more, and the trial found no difference in how much of it people did whether they were on the drug or on placebo — which answers a common worry about whether these drugs sap the capacity to train.

The strength result, read carefully

Relative muscle strength — strength normalized to body weight — changed by −7.8% on placebo, −0.4% with exercise, +1.0% with liraglutide and +3.3% with both.^[1] It is tempting to read the drug column as the drug making people stronger. It did not.

Relative strength rose because weight fell and absolute strength held. The numerator did not improve; the denominator shrank.

The trial says exactly this: the improvement came from lower weight with preserved absolute strength. That is a genuinely good outcome — carrying the same force through a lighter body is why stairs get easier — and it is not the same claim as building muscle. The distinction is the same one we drew in muscle volume versus muscle function.

Why this is the most useful trial in the batch

Because it is directly actionable and the action is unglamorous. There is no supplement, no peptide, no dose adjustment. The finding is that the drug handles weight and exercise handles fitness, and that they do not substitute for each other.

• If fitness matters to you, the drug will not deliver it. Not at any dose, on this evidence.
• Training on a GLP-1 is achievable. Adherence in the drug arm matched the placebo arm at nearly three sessions a week.
• The combination beat either alone on the functional measures, which is the outcome people actually experience.
• Doing neither, after weight loss, cost 7.8% of relative strength. That is the price of the passive option.

What it does not settle

The trial used liraglutide, a daily older drug, after diet-induced weight loss. Whether semaglutide or tirzepatide behave the same on fitness is untested here, and their larger weight effects could plausibly change the balance in either direction.

It also does not tell you what kind of training is best. Our exercise pairing tool and the strength training cheat sheet cover the practical side, and neither is contradicted by anything here.