Why 85% of Sarcopenia Exercise Trial Participants Stick With It

Exercise trials in older adults with sarcopenia report an encouraging number: participants completed 85% of prescribed exercise, on average. That figure comes from a 2025 meta-analysis of 17 randomized controlled trials involving 2,975 community-dwelling older adults with sarcopenia. The pooled estimate was 85% (95% CI, 79% to 89%), and individual trial estimates ranged from 71% to 100%.

Those are excellent attendance numbers. They do not tell us that 85% adherence is what a typical clinic, community center, or family can expect. The trials enrolled people who consented, completed baseline assessments, and could participate in structured programs. In other words, they captured older adults who had already cleared several hurdles: mobility, transportation, confidence, scheduling, and willingness to exercise.

The useful question is not whether exercise works for sarcopenia. It does. The harder question is which program designs keep people participating long enough to matter.

Shorter programs had the best adherence signal

In Wu et al.'s 2025 systematic review and meta-analysis, program duration was the only design characteristic that significantly predicted adherence. Programs lasting less than 24 weeks had higher adherence than programs lasting 24 weeks or longer (P=0.01). Exercise type, supervision modality, and the behavioral-support components coded in the trials did not significantly predict adherence.

That result gives program designers a practical starting point. A three- to five-month program may be easier to complete than one that runs for six months or more, especially when the program has ordinary staffing and support. But it does not establish a biological six-month expiration date on older adults' motivation.

The meta-analysis also found no significant association between a trial's adherence rate and its exercise efficacy outcomes. That null result needs restraint. Adherence across the included trials clustered between 71% and 100%, leaving little variation for a meta-regression to examine. The studies also used different definitions of adherence, including session attendance, compliance reports, dose intensity, and protocol fidelity. Counting sessions and completing the intended training dose are related, but they are not the same measurement.

A participant can attend nearly every session in a program that provides too little resistance stimulus to change muscle-related outcomes. Conversely, a lower attendance percentage may still include enough well-designed resistance training to improve function. The review cannot separate those possibilities cleanly.

Longer programs can work when support is built in

The duration finding should change how long programs are supported, not automatically how long they last. The 2025 EXERT randomized trial provides the clearest counterexample. It tested a supervised aerobic exercise program in older adults with mild cognitive impairment across multiple sites and maintained participation over a period longer than the 24-week threshold identified by Wu and colleagues.

EXERT studied a different population and a different exercise prescription, so it cannot erase the sarcopenia-specific duration result. Adults with mild cognitive impairment have their own barriers, care networks, and reasons for joining a study. Still, the trial shows that sustained participation beyond six months is possible when a program provides substantial monitoring, contact, and motivational support.

That distinction matters. "Shorter is better" may describe lower-resource programs. It does not prove that older adults with sarcopenia cannot maintain exercise over the long term. For a chronic condition, a 12-week completion rate is useful, but it is not the finish line. Muscle loss does not politely stop after the final study visit.

Choose exercise for the outcome, then design it for attendance

Resistance training has the strongest sarcopenia-specific rationale among the evidence discussed here. A 2024 meta-analysis by Xue et al. of randomized trials in Asian patients with sarcopenia found that resistance training significantly changed inflammatory biomarkers including IL-6 and TNF-α. Those biomarkers are not patient-centered outcomes, and they do not prove a complete mechanism for muscle gain or function. They do support the case for resistance training as more than generic activity advice.

Wu et al. did not find that exercise type significantly changed adherence. Given the narrow range of adherence and the small number of trials, that is not evidence that every format retains participants equally well. It means the available sarcopenia trials could not show a reliable difference.

Format may still matter. A 2026 three-level meta-analysis of dance interventions found improvements in physical function among healthy older adults. Dance asks participants to coordinate movement, remember sequences, attend scheduled sessions, and interact with a group. Its success argues against the assumption that a more complicated program automatically produces worse attendance. Enjoyment and social connection can make logistical complexity worth the trouble.

Healthy older adults are not interchangeable with people who have clinically diagnosed sarcopenia. Mobility limitation, fatigue, falls risk, and coexisting illness may all change what is feasible. The dance evidence is a design clue, not a sarcopenia treatment prescription.

The 85% figure likely excludes people facing the biggest barriers

Facility-based exercise programs select for access before they measure adherence. Older adults who cannot travel reliably, tolerate the trip, or arrange support may never enroll. A 2025 randomized trial of a virtual community-based group exercise program for adults with mobility limitations reflects this problem: virtual delivery exists because in-person programs do not reach everyone who could benefit.

Adjacent clinical populations point in the same direction, though they cannot quantify the gap for sarcopenia itself. A 2022 systematic review in intermittent claudication reported that adherence to supervised exercise programs remained poor despite guideline recommendations. A 2024 randomized trial in the same condition described supervised exercise as underutilized and tested home-based circuit training and community walking as alternatives.

Advanced liver disease offers another cautionary comparison. In a 2022 systematic review of randomized trials, exercise interventions improved physical-fitness and sarcopenia-related outcomes, while frailty and comorbidity burden complicated participation. These populations differ from community-dwelling older adults with sarcopenia, but the barriers are familiar: illness, travel, symptoms, and the practical cost of showing up repeatedly.

Delivery choices can also change participation patterns and costs. A 2023 randomized trial in men receiving androgen-deprivation therapy for prostate cancer compared home-based exercise with supervised group exercise and found systematic differences in adoption, participation, and cost-effectiveness. Treatment-related muscle loss is not age-related sarcopenia, yet the study makes an important operational point: one attendance number from one delivery model cannot describe the whole population.

Behavior-change techniques remain an open question

Wu et al. identified only nine behavior change techniques across the 17 sarcopenia trials. The meta-analysis found no significant adherence advantage from the behavioral components it coded, but the evidence base was too thin and inconsistent to conclude that reminders, planning, feedback, social support, or other techniques fail in this population.

The same caution applies to technology. A 2021 systematic review and meta-analysis of mobile-health exercise interventions in peripheral artery disease found improvements in walking performance, functional status, and quality of life. It does not show that an app will solve sarcopenia adherence. It does show that remote support can be a distinct route to participation for people whose symptoms or travel barriers make facility-based care difficult.

What to do with this evidence

For a new sarcopenia exercise program, start with a resistance-training component and a realistic first block shorter than 24 weeks. Build in a continuation plan before the initial block ends. Longer programs need more than a longer calendar: regular contact, transportation or remote options, social structure, and a way to adapt training when health or mobility changes.

Do not sell the 85% trial average as a community-care forecast. It describes selected trial participants under structured conditions. It is valuable because it shows what can happen when people receive a program they can attend. It is incomplete because the people least able to attend are often missing before the first session begins.

The next meaningful sarcopenia studies should measure enrollment barriers, access, training dose, attendance, and post-program activity in broader community samples. Until then, the best evidence supports exercise—and warns that retention depends as much on program infrastructure as on the workout itself.