Changes in sleep following internet-delivered cognitive-behavioral therapy for insomnia in women treated for breast cancer: a 3-year follow-up assessment

BACKGROUND: Sleep disturbances are common in women treated for breast cancer. We have previously shown that internet-delivered cognitive-behavioral therapy for insomnia (e-CBT-I) is an efficacious, low-cost treatment approach. Furthermore, research has shown that e-CBT-I can result in sustained improvements at 12 months post-treatment. However, given the complexity and long duration of post-treatment symptomatology in breast cancer patients, as well as the recommended use of antihormonal therapy for up to 10 years, it is relevant to investigate long-term (>12 months) changes in sleep following e-CBT-I in this population. In the present study, we report data from a 3-year long-term follow-up assessment after e-CBT-I.

METHODS: Women treated for breast cancer with sleep disturbances (Pittsburg Sleep Quality Index [PSQI] global score >5) who had previously been enrolled in a randomized-controlled trial investigating the efficacy of e-CBT-I (n = 255), were invited to participate in a 3-year follow-up study. All women in the initial control group had also been granted access to e-CBT-I. Assessment included self-reported sleep quality (PSQI), insomnia severity (Insomnia Severity Index, ISI), cancer-related fatigue and symptoms of depression. Within-group changes in these outcomes from baseline to the 3-year long-term follow-up assessment were analyzed.

RESULTS: A total of 131 women (51%) participated in the 3-year follow-up study of which 77 (59%) were from the initial intervention group and 54 (41%) from the initial control group. For the pooled sample, within-group improvements from baseline to the 3-year follow-up assessment corresponding to large effect sizes were observed in sleep quality (Cohen's d = 1.0 95% CI [0.78, 1.21]) and insomnia severity (Cohen's d = 1.36 CI 95% [1.12, 1.59]). Similar changes were observed in cancer-related fatigue (Cohen's d = 0.48 CI 95% [0.30, 0.66]) and symptoms of depression (Cohen's d = 0.80 CI 95%. [0.60, 0.99]). The proportion of patients with scores above established cut-offs on the PSQI and the ISI were 56.1% and 29.8%, respectively. Within the initial intervention group, 15.6% evidenced relapse at the 3-year assessment.

CONCLUSION: Overall, these results indicate that long-term sleep quality and insomnia severity following the use of e-CBT-in women treated for breast cancer is significantly lower than the pre-treatment levels. However, a substantial proportion of participants still evidence sleep disturbances.