Virtual Lifestyle Program Shows Promise and Limits for People with Schizophrenia

People living with schizophrenia die years, sometimes decades, earlier than the general population, and the overwhelming majority of those premature deaths are not caused by the illness itself. They are caused by heart disease, diabetes, and other cardiometabolic conditions that accumulate quietly over years of sedentary living, poor diet, and the metabolic side effects of antipsychotic medications. A new pilot study published in the Community Mental Health Journal offers a candid look at whether a fully virtual lifestyle intervention — combining exercise and nutrition content delivered remotely — can help close that deadly gap, and the results are a mixture of genuine promise and sobering practical challenges that researchers say must be addressed before the approach can scale.

The study, led by Julia Browne of the Wellness and Recovery After Psychosis Program at Boston Medical Center Health System and colleagues at Boston University, Massachusetts General Hospital, and Harvard Medical School, enrolled nineteen adults with schizophrenia in a two-period randomized cross-over pilot trial. Participants were assigned to either an intervention phase or a control phase for the first eleven weeks, then crossed over for the second eleven-week period. The intervention consisted of virtually delivered lifestyle programming built around structured exercise sessions and nutrition education, designed to be accessible from home — a delivery model motivated by the well-documented barriers that keep people with serious mental illness out of traditional in-person programs, including transportation difficulties, social anxiety, stigma, scheduling conflicts, and the fatigue and motivational deficits that often accompany psychotic disorders.

The rationale for targeting physical health in this population is unambiguous in the epidemiological literature. Meta-analyses have consistently shown that cardiovascular disease is the leading cause of death among people with schizophrenia, with prevalence and incidence rates far exceeding those of matched controls. Large-scale studies of first-episode psychosis already reveal elevated cardiometabolic risk at the very beginning of treatment, and national register studies have documented increased cardiovascular mortality persisting across decades. People with psychosis are also among the most sedentary groups studied anywhere in medicine, and their diets tend to be poorer in quality than those of the general population. Lifestyle interventions — behavioral weight-loss programs, health promotion coaching, and structured exercise — have repeatedly demonstrated effectiveness in controlled settings, which makes the real question one of implementation: how do you get people to participate, and stay participating, in programs that demand sustained behavioral change from a population facing cognitive, motivational, and logistical obstacles?

On the feasibility side, the pilot’s numbers tell a nuanced story. During the first eleven-week period, retention was 58 percent and intervention engagement reached 67 percent — modest figures, but not dramatically out of line with what has been reported for physical activity interventions in schizophrenia generally, where dropout is a persistent and well-documented problem. During the second period, however, the picture deteriorated sharply. Retention fell to 37 percent and intervention engagement to 50 percent, rates low enough that the investigators were forced to abandon their planned cross-over analysis entirely. With so few participants contributing data in the second half of the study, neither a formal cross-over comparison nor an examination of outcome changes during that period was statistically possible. The attrition itself became one of the study’s most important findings.

The secondary outcomes, analyzed for the first period only, were considerably more encouraging. Comparing changes between the intervention and control groups across weeks one through eleven, the researchers found effects favoring the intervention across a striking range of measures: nutritional knowledge, physical activity levels, mental health, physical health, exercise motivation, self-esteem, and weight and body mass index. The standardized effect sizes, expressed as Cohen’s d values, ranged from 0.35 to 2.09 — a spread that runs from modest to very large by conventional standards. While the small sample and sparse data mean these estimates carry wide uncertainty and cannot be treated as definitive evidence of efficacy, the breadth and direction of the changes suggest that the intervention content itself, when participants actually engaged with it, was doing something meaningful across multiple domains of health and wellbeing simultaneously.

Satisfaction data, though available for only six participants, added a further note of optimism. Nearly all of those who completed the satisfaction assessment reported that the intervention was helpful and easy to follow. For a virtual format targeting a population that is often assumed to struggle with technology-mediated care, that acceptability signal matters. It aligns with a growing body of work from the same research group and others showing that people with serious mental illness are willing and often eager to use remote health tools, including virtual walking groups and mobile-health-supported peer programs, provided the content is designed with their needs in mind.

Why did engagement collapse in the second period? The authors point to attrition and the sparsity of data as the proximate obstacles to analysis, but the pattern itself raises questions that the pilot was not powered to answer definitively. Cross-over designs are elegant on paper — every participant serves as their own control, reducing between-person variability and requiring smaller samples — but they impose a double burden: participants must commit to a program lasting more than four months, and those assigned to control first must sustain engagement through a waiting period before receiving anything of perceived value. For people managing a chronic psychotic illness, with its attendant motivational challenges and symptom fluctuations, that structural demand may itself have driven the steep second-period losses. The study’s design, in other words, may have been as much a test of endurance as of the intervention.

The authors are explicit about the implication: intervention development research is needed to determine whether modifications to this virtual lifestyle intervention would enhance feasibility and acceptability for people with schizophrenia. That is the language of honest pilot science — the study was designed to generate lessons, not conclusions, and the lessons here are twofold. First, the content works well enough for those who stay engaged to produce measurable improvements in knowledge, behavior, mood, self-esteem, and weight. Second, the delivery model, or the study structure wrapped around it, loses people at rates that would undermine any real-world deployment. Candidate refinements might include shorter program durations, intensified early engagement strategies, simplified technology, peer support integration, or hybrid models that pair virtual sessions with periodic in-person contact — each of which has precedent in the broader lifestyle intervention literature for serious mental illness.

The stakes of solving this problem are difficult to overstate. The longevity gap between people with schizophrenia and the general population has been widening rather than narrowing in recent decades, a trend documented across multiple countries and health systems. Pharmacological and non-pharmacological interventions to improve physical health in this population are supported by an increasingly robust evidence base, including meta-reviews of hundreds of randomized trials. What remains scarce are delivery models that survive contact with the realities of patients’ lives. This pilot does not settle whether virtual delivery can be that model, but by pairing encouraging satisfaction and outcome signals with unflinching retention data, it gives the field something arguably more valuable than a positive result: a precise map of where the approach breaks down, and a foundation on which the next iteration can be built. For a population whose lives literally depend on accessible, sustainable health programs, that map is a contribution worth taking seriously.

Subject of Research: A pilot feasibility study of a virtual exercise and nutrition lifestyle intervention for adults with schizophrenia.

Article Title: A Pilot Feasibility Study of a Virtual Lifestyle Intervention for Individuals with Schizophrenia

Article References: Browne, J., Gouse, B. M., Weinberg, J., Blanton, A., Simons, G., Thomas, O., Camacho, L., LeFeber, L., Agarwal, N., Donovan, A. L., Cather, C., & Brown, H. E. (2026). A Pilot Feasibility Study of a Virtual Lifestyle Intervention for Individuals with Schizophrenia. Community Mental Health Journal. https://doi.org/10.1007/s10597-026-01722-6

Image Credits: AI Generated

DOI: 10.1007/s10597-026-01722-6

Keywords: schizophrenia, virtual lifestyle intervention, telehealth, exercise, nutrition, cardiometabolic health, psychosis, feasibility study, mental illness, physical activity, retention, pilot trial