A RANDOMIZED CONTROLLED TRIAL OF CBT-I AND CPAP FOR COMORBID INSOMNIA AND OSA: INITIAL FINDINGS FROM THE MATRICS STUDY

The purpose of this study was to compare the efficacy of three treatment sequences using CPAP with or without CBT-I for the treatment of comorbid insomnia and obstructive sleep apnea (COMISA). One-hundred and twenty-one adults (52.50% female, mean age=50.00) with COMISA were randomized to receive one of three treatment models using a partial factorial design: CBT-I followed by CPAP (Model A), CBT-I concurrent with CPAP (Model B), and CPAP only (Model C). CPAP was delivered following standard procedures for in-lab CPAP titration and home set-up with a third-party vendor. CBT-I was delivered in four individual sessions. Primary outcomes were CPAP adherence across the first 90 days of use and global measures of insomnia (Insomnia Severity Index; ISI) and sleep quality (Pittsburgh Sleep Quality Index; PSQI). Planned comparisons were conducted on those who received CBT-I+CPAP (Models A and B combined) versus CPAP only (Model C) using Wilcoxon signed ranks test and linear mixed models. No significant differences were found between the CPAP+CBT-I groups versus the CPAP only group on percent of nights used (Median=48.65% vs Median=71.10%, p=.29), minutes used per night (Median=232.87 vs Median=186.06, p=.21) or percentage of participants who were regular CPAP users (≥4 hours on ≥70% of nights during 30-day period; 37.50% vs 36.59%, p=.92). For mild OSA (AHI≥5 and <15), the CPAP+CBT-I groups used CPAP on a greater percentage of nights compared to the CPAP only group (Median=30.00% vs Median=5.00%), which approached significance (p=.05). A significant main effect was found for improvement in the ISI (p