Treatment expectations and depressive symptoms in an internet-based intervention for depression. A secondary analysis

Background

Treatment expectations are known to influence therapy outcomes, but their role in internet-based interventions (IBIs) for depression remains unclear. While previous research has primarily focused on expectations as a pre-treatment predictor (PTP), emerging evidence suggests that early process predictors (EPPs), including evolving expectations during treatment, may provide more relevant insights into therapeutic outcomes.

Objective

This secondary analysis of a factorial trial (Bur et al., 2022) investigates the role of treatment expectations as both a pre-treatment predictor and early process predictor in an internet-based intervention for mild to moderate depression. It also explores the temporal relationship between expectations and depressive symptoms, assessing whether earlier expectations predict later symptom severity and whether depressive symptoms influence subsequent expectations.

Methods

Treatment expectancy was measured using the Credibility and Expectancy Questionnaire (CEQ-8; Devilly & Borkovec, 2000; German version: Walach et al. 2008) at baseline (T0), two weeks (T1), and four weeks (T2), while depressive symptoms were assessed using the Patient Health Questionnaire-9 (PHQ-9; Kroenke et al., 2001) at the same time points as well as post-treatment (T3). To analyze the relationship between treatment expectations (CEQ-8) and depressive symptoms (PHQ-9 post-treatment), simple regression models were conducted while controlling for baseline PHQ-9 scores. Multiple regression analyses were then used to examine whether CEQ-8 predicted PHQ-9 or vice versa. In addition, as a sensitivity analysis, a cross-lagged panel model (CLPM) was estimated to account for the repeated-measures structure of the data.

Results

Baseline treatment expectations did not significantly predict depressive symptoms at post-treatment. However, expectations measured at two weeks (T1) and four weeks (T2) significantly predicted depressive symptoms at T3. The results of the multiple regression analyses indicate that treatment expectations can predict changes in depressive symptoms, whereas the reverse relationship was not observed. The CLPM yielded results that were consistent with the regression analyses, supporting the robustness of the findings.

Conclusions

Treatment expectations evolve throughout therapy and appear to function as an independent predictor of symptom improvement rather than merely reflecting symptom severity. Monitoring and addressing patient expectations early in treatment may enhance intervention outcomes. These findings support the inclusion of expectation-based strategies in IBIs to optimize engagement and effectiveness.