The role of treatment expectancy in lifestyle interventions for depression

BACKGROUND: Treatment expectancy is a key mechanism underlying placebo. While well-documented in psychopharmacology and psychotherapy, its role in lifestyle-based interventions for mental disorders is uncertain. Given the emerging role of lifestyle-based mental health care, a comprehensive evaluation of the role of treatment expectancy has not yet been explored, especially when compared to other more established therapies. The potential for genuine treatment efficacy to be masked by placebo response may bring into question the magnitude of effects of this treatment. METHODS: We sought to: 1) investigate the role of treatment expectancy as a moderator of depression treatment outcomes in a completed randomised controlled trial of psychotherapy versus lifestyle therapy (the CALM trial); and 2) provide guidance on measuring and reporting expectancy effects in future trials. CALM was a two-arm, parallel-group, non-inferiority randomised controlled trial, comparing lifestyle therapy (delivered by accredited dietitians and exercise physiologists) and psychotherapy (facilitated by psychologists). A total of 182 participants (91 per arm) experiencing psychological distress (equivalent to 'indicative depression') during the COVID-19 pandemic were randomised. We assessed associations of expectancy (measured prior to randomisation using the Credibility-Expectancy Questionnaire) with reductions in depressive symptoms (as measured by the PHQ-9). RESULTS: Expectancy scores did not differ between treatment groups at baseline (β = 1.14; 95%CI = -1.93, 4.22). Both arms reduced PHQ-9 between baseline and 8-weeks (lifestyle mean difference = -30.25%, SD = 58.23%; psychotherapy mean difference = -22.46%, SD = 66.87%). However, higher expectancy was associated with a greater reduction in depression only in the psychotherapy arm: for every 1-unit increase in expectancy score depressive symptoms reduced by an additional 3.21% for psychotherapy compared to lifestyle therapy (95%CI = -5.77, -0.65). CONCLUSIONS: In this study we showed that participants' expectations about treatment at enrolment were only associated with treatment response in psychotherapy. These findings highlight the need to account for modality-specific drivers of expectancy when evaluating all treatments for depression, and support the inclusion of expectancy in reporting guidelines for future trials. REGISTRATION: The trial was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12621000387820; https://www.anzctr.org.au/Trial/Registration/TrialReview.aspx?id=380897).