Efficacy of an Internet-delivered self-compassion intervention in preventing depressive relapse and recurrence: A randomized controlled trial in a clinical sample

Abstract

Depression is a prevalent and burdensome mental health condition. Unfortunately, depression is chronic and recurrent, with a lifetime average of four to five episodes. Although evidence supports the efficacy of self-compassion interventions for a range of psychological disorders (Ferrari et al., 2019), research on scalable forms of compassion-based interventions for recurrent depression remains scarce. The primary purpose of this two-arm randomized controlled trial (RCT) was to examine the efficacy of a 6-week internet-delivered, self-directed self-compassion intervention in preventing depressive relapse and recurrence over a 12-month period in a clinical sample of adults with a history of depression. The efficacy of the intervention was tested against a self-reflection active control condition. In addition to examining differential relapse/recurrence rates, the RCT compared changes in the primary outcome of depressive symptoms (PHQ-9), and secondary outcomes of self-compassion (SCS), mindfulness (FFMQ-15), decentering (EQ-D), and fear of self-compassion (FCSS) across conditions and time. A total of 158 participants meeting eligibility criteria (87% identifying female; M number of prior depressive episodes = 2.88 [SD = 1.82]) were randomized into the self-compassion (SC; n = 80) intervention condition or self-reflection (SR; n = 78) active control condition. All participants completed primary and secondary outcome measures at baseline, weekly throughout the 6-week intervention period, at post-treatment, and at each of four followups (at 3-, 6-, 9-, and 12-month post-randomization). Participants also completed a phone interview at each follow-up to determine relapse/recurrence status. Depressive episodes were assessed by a structured interview. Kaplan-Meier survival analyses were conducted to compare rates of depressive relapse/recurrence between the two conditions across the trial period. Cox proportional hazards regression analyses were conducted to examine the effects of predictors on time until relapse/recurrence. Multilevel modeling and mixed model analysis of variance were conducted to examine differential changes in primary and secondary outcomes between the two conditions. Contrary to primary hypotheses, survival analysis revealed no significant differences in relapse/recurrence rates between SC participants (relapse rate = 33%) and SR participants (relapse rate = 43%) across the 12-months trial period, χ2(1) = .35, p = .56. Number of past depressive episodes and baseline fear of self-compassion predicted time until relapse/recurrence in an exploratory survival analysis; greater number of past episodes and higher levels of fear of self-compassion were associated with a higher likelihood of relapse. Further, SC participants did not report significantly greater reductions in depressive symptoms, nor significantly greater increases in mindfulness and decentering compared to SR participants. However, SC participants reported significantly greater increases in self-compassion and significantly greater decreases in fear of self-compassion compared to SR participants across the trial period. Results revealed a main effect of time on all outcome measures; participants in both conditions reported significantly lower distress across time. The current findings support self-directed self-compassion training as a feasible approach to cultivating protective and reducing risk factors associated with recurrent depression. Self-compassion training through an internet-delivered self-directed program may be appropriate for use as an adjunct to more established evidence-based treatments specifically targeting relapse in depression. Keywords: recurrent depression, relapse/recurrence, self-compassion, mindfulness, decentering, fear of self-compassion, compassion-based intervention, RCT