BackgroundWhile there is evolving knowledge on change processes of digital cognitive behavioral therapy (CBT) in the treatment of depression, little is known about how these interventions produce therapeutic change in the comorbid constellation of chronic back pain (CBP). Here, we examined whether the effects of a digital intervention to treat depression in patients with CBP are mediated by three pain-related variables (i.e., pain self-efficacy, pain-related disability, pain intensity). MethodsThis study is a secondary analysis of a randomized clinical trial conducted in routine care at 82 orthopedic clinics across Germany. In total, 209 adults with CBP and diagnosed depression (SCID interview) were randomly assigned to the intervention (n = 104) or treatment-as-usual (n = 105). Cross-lagged mediation models were estimated to investigate longitudinal mediation effects of putative mediators with depression symptom severity (PHQ-9) as primary outcome at post-treatment. ResultsLongitudinal mediation effects were observed for pain self-efficacy (ß = −0.094, 95%-CI [−0.174, −0.014], p = 0.021) and pain-related disability (ß = −0.068, 95%-CI [−0.130, −0.001], p = 0.047). Furthermore, the hypothesized direction of the mediation effects was supported, reversed causation did not occur. Pain intensity did not reveal a mediation effect. ConclusionsThe results suggest a relevant role of pain self-efficacy and pain-related disability as change processes in the treatment of depression for patients with CBP in routine care. However, further research is needed to disclose potential reciprocal relationships of mediators, and to extend and specify our knowledge of the mechanisms of change in digital CBT for depression.