Determinants of treatment response to cognitive behavioral therapy in veterans presenting with comorbid insomnia and sleep apnea

Purpose: Although cognitive behavioral therapy for insomnia (CBT-I) is considered the preferred treatment for insomnia in patients with comorbid insomnia and obstructive sleep apnea (COMISA), the remission rate with CBT-I is generally considered lower than in insomnia-only populations. There is also a sizable variability in individual treatment responses. Due to the limited availability of CBT-I, we sought to identify specific clinical attributes that predict benefit from Brief Behavioral Therapy for Insomnia (BBTI)-an adaptation of CBT-I-in patients with COMISA. Methods: We conducted a retrospective analysis of the National Veterans Health Administration (VHA) electronic medical records covering veterans diagnosed with COMISA between January 2021 and December 2023. Insomnia Severity Index (ISI) scores were recorded at baseline and after 12±1 weeks after BBTI. A positive response to BBTI was defined as a reduction in ISI score of ≥ 8 from baseline. A multivariate generalized linear model analysis was performed to delineate predictive factors of BBTI responsiveness. Results: 131 eligible cases received BBTI over 6 weeks, 56 (43%) of whom did not respond. Non-whites (OR 3.5, 95% CI [1.4, 8.8]) and shorter sleep time (OR 0.98, 95% CI [0.98, 0.99] were independent predictors of blunted response to BBTI. These findings remained true even when depression and AHI were forced into the regression model. Patients with a total sleep duration of < 4.1 h were at greatest risk of being nonresponsive to BBTI. Conclusion: These findings indicate that identifying insomnia phenotypes in patients with COMISA would help deliver personalized care while maximizing BBTI treatment resources.