Muscle Without Meat After 55: What the IMPACT Trial Found

Can you build muscle on a vegetarian diet after 55, or does the protein source quietly cap the gains? The IMPACT trial put the question in the only place it can be answered: a supervised weight room, older adults, twelve weeks, and a body scan that measures muscle rather than opinions. Only one group cleared the bar the researchers set for muscle mass, and the reason is worth reading slowly.

How the trial worked

Researchers at the Amsterdam University of Applied Sciences randomized 108 adults aged 55 and over to one of two diets, both paired with the same training program. Most were in their 60s, 58% were women, and the group skewed healthy, White and highly educated. The study ran from March 2024 to June 2025 and was registered in advance at ClinicalTrials.gov.

The training was serious: three sessions a week for 12 weeks, six exercises at a heavy prescribed load, one trainer for every two participants. Median attendance was 92% of sessions among vegetarians and 88% among omnivores.

Nobody was handed meals. Participants chose their own food with dietitian counseling. The vegetarian group dropped meat and fish but kept eggs and dairy, aiming for at least 60% of protein from plants; the omnivore group aimed for at least 60% from animal foods. Both were steered toward the same total protein target, a study target rather than a recommendation.

The result, in absolute numbers

Muscle in the arms and legs, measured by DXA scan, rose by 0.38 kg in the vegetarian group and by 0.66 kg in the omnivore group. Both gains were statistically clear. Against a starting point of about 22 kg of arm and leg muscle, that is roughly 1.7% versus 3%.

Strength told a friendlier story. Leg press maximum climbed by 34.7 kg among vegetarians and 39.5 kg among omnivores, from a baseline near 138 kg, and the gap between groups was not statistically significant. The five rise chair test got faster by 1.2 seconds in both groups. Handgrip did not move in either group.

The quiet number behind all of it is protein. At week 12, the vegetarians were eating 1.24 g of protein per kg of body weight per day, the omnivores 1.58 g. In plain grams, about 94 versus 119 a day. The vegetarian group did hit its plant share target, with 76% of protein from plants.

Read the wording like a scientist

IMPACT was a noninferiority trial: it asked whether the vegetarian diet was worse by no more than a preset margin, 0.25 kg for arm and leg muscle. The estimated shortfall for vegetarians was 0.33 kg, with a confidence interval running from 0.69 kg less to 0.03 kg more. Because the bad end of that range blows well past the margin, noninferiority was not shown. The same verdict came back for protein intake.

That is neither "vegetarians lost out" nor "no difference." The interval still includes zero, so the trial cannot prove the omnivores truly gained more; it could not rule out a gap big enough to matter. For strength, no margin was set at all, so "no detected difference" there is an absence of evidence, not proof of equality. If those terms feel slippery, our guide on how to read a health study in 5 minutes walks through the same traps.

The open question is how much of the gap is protein source, and how much is simply protein amount.

Source or quantity? The trial can't untangle it

Because people chose their own food, the vegetarian group ended up eating less protein and, notably, about 289 fewer calories a day than the omnivores. Source, quantity and energy all moved together. The design was deliberate, mirroring real kitchens, but it leaves the muscle gap with more than one suspect.

One clue points toward quantity. Within the vegetarian group, people who ate more protein gained more muscle: every extra 10 g a day lined up with about 153 g more arm and leg muscle. Among omnivores there was no such link, which the authors read as a ceiling already reached. That is a correlation inside a trial, not a randomized comparison, so treat it as a lead rather than a verdict. Quantity probably explains a good share of the gap, but this study cannot prove it.

The vegetarian diet also brought more fiber and less saturated fat, changes a muscle scan does not capture and this trial was not built to judge.

Limits worth keeping in mind

Twelve weeks is short for muscle. Nobody was blinded. There was no follow up after counseling stopped, so nobody knows whether the eating patterns stuck. DXA lean tissue is a good proxy for muscle, not a direct measure. The volunteers were healthier and better educated than average, so frailer older adults may respond differently. And the project was funded by a Dutch public and private grant in which Royal FrieslandCampina, a dairy company, took part as industrial partner; the authors report no personal payments and say they controlled the design and analysis.

What comes next

The trial that would settle the source question matches total protein and calories gram for gram and varies only where the protein comes from, over a longer stretch, with a check after coaching ends. Vegan diets, which drop eggs and dairy too, are a separate question this trial did not test. For a related angle on keeping muscle during big body changes, see the race to keep muscle on weight loss drugs.

Until then, the fair reading is twofold. Older adults who trained hard built muscle and strength without meat. They also built less muscle than the meat eaters beside them, while eating less protein, and nobody yet knows which of those two facts will hold up longer.