Effect of depression treatment on subjective sleep components among primary care patients: Data from the eIMPACT trial

Abstract

Objective: Sleep disturbance is a multifaceted symptom of depression that disproportionately impacts marginalized groups. Depression treatment improves sleep disturbance in some individuals; however, the components of sleep disturbance improved remain unidentified. This secondary analysis of the eIMPACT randomized controlled trial examines effects of depression treatment on subjective sleep disturbance components.

Methods: 216 primary care patients with depression from a safety net healthcare system were randomized to 12-months of modernized collaborative care (internet cognitive-behavioral therapy [CBT], telephonic CBT, and/or antidepressants; n = 107) or usual primary care for depression (primary care providers supported by embedded behavioral health clinicians and psychiatrists; n = 109). Subjective sleep disturbance components were assessed by the Pittsburgh Sleep Quality Index (PSQI).

Results: Mixed-effect models examined intervention effects on changes in subjective sleep disturbance components. The intervention improved PSQI global scores, sleep onset latency, subjective sleep quality, and daytime dysfunction across 24-months, and these effects diminished after treatment termination. At post-treatment, intervention participants had greater improvements in PSQI global scores (p < 0.001, d=-0.62), sleep onset latency (p < 0.01, d=-0.43), daytime dysfunction (p < 0.01, d=-0.35), and sleep disturbances (p = 0.01, d=-0.26) compared to usual care, but no differences in subjective sleep quality, total sleep time, sleep efficiency, or sleep medication use. Intervention effects were not moderated by race, education, or income. Pre- to post-treatment improvements in depressive symptoms were associated with improvements in some sleep disturbance components.

Conclusions: Depression treatment improves some, not all, subjective sleep disturbance components, with benefits diminishing after termination. Adjunctive interventions are likely needed to address the lingering components of sleep disturbance.

Clinicaltrialsgov idenifier: NCT02458690.

Keywords: collaborative care; depression; health disparities; internet interventions; randomized controlled trial; sleep disturbance.

Fig. 1

Effect of the eIMPACT intervention, versus usual primary care for depression, on subjective sleep components across 24 months. Outcomes were assessed at pre-treatment (baseline), mid-treatment (6-months), post-treatment (12-months), and follow-up (24-months). The graphs were created based on quadratic mixed-effects models for all outcomes except for total sleep time, sleep efficiency, and sleep medication use score. For these three outcomes, linear mixed effects models were used. All models were adjusted for baseline imbalance in education, income, and sleep medication use. PSQI = Pittsburgh Sleep Quality Index; n.s. = not significant. N = 216.