PMSB-E Trial: Exercise Improved Function, Not Mood

Exercise improved objective physical function in older adults with depression and mobility limits. The same 12-month program did not improve their reported depression or anxiety symptoms. That split result is the useful finding from PMSB-E, a large randomized trial that tested a combined psychoeducation-and-exercise program against usual care.

The trial does not show that treating physical and mental health together improves both. It shows a program can move one domain while leaving the other unchanged—and that its design cannot tell us whether psychoeducation added anything beyond exercise.

What PMSB-E tested

Alegria et al.'s 2025 randomized trial enrolled 427 adults aged 60 and older with clinically significant depression or anxiety symptoms and minor-to-moderate mobility limitations. Participants were recruited through 51 community organizations and 17 clinics in Massachusetts, New York, Puerto Rico, and Maryland. The study ran across 68 sites, with data collection from November 2020 through October 2024.

The intervention paired 10 psychoeducational sessions with 36 exercise sessions over six months. Half of intervention participants were then offered a six-month group maintenance phase. The comparator was usual care, which included up to four calls. Researchers assessed outcomes at baseline and at months 3, 6, and 12. The prespecified outcomes most relevant here were objective physical function on the Short Physical Performance Battery (SPPB), mental-health symptoms on the Hopkins Symptom Checklist-25 (HSCL-25), and self-reported limitations on the Late-Life Function and Disability Instrument (LLFDI).

The exercise result held up. At six months, the SPPB standardized difference was 0.34 standard deviations (95% CI, 0.14 to 0.55); at 12 months, it was 0.28 standard deviations (95% CI, 0.07 to 0.49). Both estimates were statistically significant. These are small-to-moderate improvements, not a reversal of age-related decline, but they persisted through a year of follow-up.

The mental-health endpoint did not move

The HSCL-25 showed no statistically significant improvement at either six or 12 months. The LLFDI also showed no improvement at either time point. The study found no interaction by race or ethnicity for these outcomes or for physical function.

That matters because the intervention claimed two targets. An improvement in a performance battery does not establish improvement in mood, anxiety, or daily function. The SPPB is assessed by a blinded evaluator; the HSCL-25 and LLFDI are self-reported. The mismatch could reflect a real separation between physical capacity and psychological symptoms. It could also reflect the limits of remote self-report assessment, including response bias or floor and ceiling effects. The study included no objective mental-health or biological measures, such as actigraphy-based sleep measures, inflammatory markers, or cortisol, so it cannot distinguish those explanations.

More session completion predicted larger SPPB gains. The publication does not report whether completion predicted HSCL-25 outcomes. Higher attendance may therefore explain some physical-function variation, but it does not establish a minimum effective mental-health dose.

The physical result fits established exercise evidence

PMSB-E's physical-function effect is credible and unsurprising. A 2023 systematic review and network meta-analysis by Valenzuela et al. found that structured exercise improves physical function in older adults in residential care. A separate 2025 randomized trial by Kim et al. found that home-based multicomponent exercise with wearable activity trackers improved fall-related parameters and physical function in older adults.

Those studies support the practical point: older adults with mobility limitations can improve measurable performance through structured exercise, including home-based programs. PMSB-E adds scale and a year of follow-up to that familiar result.

They also expose the trial's largest interpretive gap. PMSB-E did not include an exercise-only arm. Since exercise-alone programs already improve objective function, the trial cannot show whether psychoeducation contributed to the 0.28-standard-deviation SPPB difference at 12 months. It cannot show whether psychoeducation was inert, helpful, or harmful for mental-health outcomes either.

Why the null mood result deserves attention

The null HSCL-25 finding is not what the wider literature would lead you to expect. A network meta-analysis of 35 randomized trials by Chen et al. concluded that non-pharmacological interventions reduce depressive symptoms in older adults. Exercise also improved psychological well-being in people with knee osteoarthritis in a 2021 network meta-analysis by Hall et al., a population that shares the burden of pain and limited mobility but not necessarily the same clinical depression profile.

Reviews of high-intensity exercise reach a similar conclusion in other populations. Martland et al. reported mental-health improvements with high-intensity interval training in general and physically ill populations, while Zeng et al. reported antidepressant effects of high-intensity exercise in patients with depression. Neither review proves that PMSB-E should have improved HSCL-25 scores. Its exercise intensity and dose cannot be benchmarked fully against those prescriptions from the reported information. But the broader evidence makes the null result a question to explain, not one to wave away.

Virtual delivery is one possible explanation, because the psychoeducation component was delivered remotely during a study period that began in the pandemic. It is not a sufficient explanation by itself. The HEAL-HOA randomized trial found psychosocial benefits from layperson-delivered telephone behavioral activation among low-income older adults during COVID-19. Behavioral activation is not the same intervention as psychoeducation, and the trials used different endpoints. Still, HEAL-HOA shows that remote psychological care can work for older adults under pandemic conditions.

The honest conclusion is narrower than the program's name: PMSB-E improved measured physical performance, but it did not demonstrate a mental-health benefit or an added benefit from psychoeducation.

What to do with this result

For an older adult with depression and mobility limitations, structured multicomponent exercise remains an evidence-supported choice for physical capacity. Track something objective when possible: walking speed, chair stands, balance, or a formal SPPB assessment. Wearables can help with activity tracking, though a step count is not a depression treatment.

For depression or anxiety, do not assume that a bundled virtual psychoeducation-and-exercise program improves symptoms simply because it targets both mind and body. The theory remains plausible; the PMSB-E trial did not confirm it. A more informative trial would compare exercise alone, psychoeducation alone, their combination, and usual care, while measuring exercise intensity, adherence, sleep, and objective mental-health-related outcomes.

Until then, the evidence supports exercise for physical function and supports established mental-health treatment for depression. The "enhanced" label does not substitute for a comparator.