The Forensic Mental Health in Community Health (FMHiCH) program was established across five Victorian sites as part of a ten-year mental health reform process. It provides mental and general health care for adults referred from Community Correctional Services who are subject to community-based orders, have a moderate-severity mental health disorder, and present a medium-to-high risk of re-offending. At one of the five sites, exercise physiology and speech pathology were woven into the multidisciplinary team — a deliberate experiment in whether accessible, non-stigmatising services could draw people into care who might otherwise disengage.

The rationale is rooted in the realities of this population. Many adults completing community-based orders experience co-occurring physical health conditions, poor health literacy, and weak connections to mainstream health providers. Recidivism rates in some cohorts reach 43 percent. Engagement in treatment is notoriously fragile, constrained by low motivation, stigma, distrust and structural barriers. The program’s designers reasoned that a gym-based exercise session might feel far less confrontational than a therapy room, particularly for a predominantly male client group, while speech pathology could address the communication difficulties that are disproportionately common among justice-involved people and can quietly undermine their ability to understand and participate in rehabilitation.

To test whether these disciplines were associated with better engagement, the research team led by Frances Shawyer of Monash University conducted a retrospective cohort study using de-identified service data from December 2018 to June 2023. Because the program’s database contained no direct measure of engagement — no ratings of attendance quality, session participation or attitude — the researchers operationalised engagement through two behavioural indicators: whether treatment was completed, and whether treatment goals were achieved or partially achieved. These outcomes reflect sustained involvement in and progress through care, and are consistent with common practice in forensic and mandated-treatment research.

The statistical approach was appropriately rigorous for real-world service data. The team used mixed-effect logistic regression on 1,111 care episodes, accounting for the clustering of episodes within individual clients, and adjusted for a wide range of demographic and socio-economic factors including age, sex, country of birth, marital status, employment, living arrangement, health conditions and area-level socio-economic disadvantage measured by the Index of Relative Socio-Economic Disadvantage. Missing data were handled through multiple imputation by chained equations, generating ten imputed datasets as a sensitivity analysis.

The results were striking. In minimally adjusted models, exercise physiology involvement was associated with roughly five to six times higher odds of treatment completion (odds ratio 5.65, 95 percent confidence interval 2.41 to 13.23) and goal achievement (odds ratio 6.67, 95 percent confidence interval 2.42 to 18.42). Speech pathology showed positive associations as well, with odds ratios of 2.50 for completion and 4.05 for goal achievement. The findings held up in fully adjusted models and in the imputed sensitivity analyses: exercise physiology odds ratios of 5.68 for completion and 7.26 for goal achievement, and speech pathology odds ratios of 2.44 and 3.75 respectively. In plain terms, people who saw an exercise physiologist or speech pathologist were several times more likely to stay in treatment and make progress.

The study also explored dose — whether more contact meant better outcomes. It did. For exercise physiology, a doubling of contact time was associated with approximately 55 percent higher odds of both treatment completion and goal achievement. For speech pathology, doubling contact time increased the odds of completion by roughly 30 to 40 percent, with a 56 percent increase in the odds of goal achievement per unit increase in log-transformed time. The number of exercise physiology contacts was also positively associated with treatment completion. This dose-response pattern suggests that greater therapeutic exposure may itself confer benefit, although the authors caution that clients who accumulated more hours may simply have been more motivated or more stable to begin with.

The broader descriptive findings add important context. Employment and secure housing were consistently linked to better engagement outcomes, while homelessness and schizophrenia spectrum disorders were associated with poorer outcomes. Age mattered too: clients over 41 were more likely to complete treatment, a pattern consistent with developmental criminology research showing that offending typically declines with age, possibly reflecting a growing motivation to close the chapter on justice involvement. Intriguingly, clients who travelled longer to reach the service were more likely to complete treatment — perhaps because those willing to travel represent a more motivated or better-supported subgroup, or because they came from more advantaged suburbs, though the area-level disadvantage measure did not support the latter explanation.

The authors are careful about what the study can and cannot claim. The observational design cannot establish causation; participation in exercise physiology and speech pathology was not random but shaped by clinician judgement, client readiness and capacity to engage, raising the possibility of selection bias. The sample sizes for the adjusted models were modest, particularly for speech pathology, producing wide confidence intervals. Services were delivered largely by single clinicians, so outcomes may partly reflect individual therapeutic style rather than the intervention models themselves. And the two outcome indicators capture observable behaviour, not the attitudinal or relational dimensions of engagement.

Nevertheless, the convergence of the main effects and the dose-response findings across both disciplines strengthens the case that these allied health interventions add value beyond general service contact. The researchers argue that the Victorian program is ideally positioned for a definitive test: four other FMHiCH sites currently operate without exercise physiology and speech pathology, making a stepped-wedge cluster randomised controlled trial — in which sites sequentially adopt the interventions over time — a pragmatic next step. A similar design was used successfully in the PULSAR study of recovery-oriented practice training across eighteen Victorian mental health services. If controlled research confirms these associations, the humble gym session and the communication assessment could become standard fixtures of community forensic mental health care, offering a route into treatment for people whom traditional services too often lose.

Subject of Research: Effects of exercise physiology and speech pathology on treatment engagement among adults on community-based forensic mental health orders

Article Title: Exercise physiology and speech pathology support treatment engagement in forensic mental health in community health: evidence from a retrospective cohort study

Article References: Shawyer, F., Wilson, K., Dhungana, R. R., Atkinson, B., Hipgrave, L., Dawadi, S., Enticott, J., & Meadows, G. (2026). Exercise physiology and speech pathology support treatment engagement in forensic mental health in community health: evidence from a retrospective cohort study. Frontiers in Psychiatry, 17, Article 1964444. https://doi.org/10.3389/fpsyt.2026.1964444

Image Credits: AI Generated

DOI: 10.3389/fpsyt.2026.1964444

Keywords: forensic mental health, exercise physiology, speech pathology, community-based orders, treatment engagement, retrospective cohort study, community health services, odds ratio, dose-response, social determinants of health, Victoria Australia, rehabilitation