LAUSR

Cognitive behavioral therapy for insomnia (CBTI) is gold standard for insomnia care. However, recent data suggest Black CBTI patients may complete fewer sessions and experience poorer outcomes relative to white CBTI patients. The present study examined whether racial disparities in CBTI engagement mediated group differences in treatment outcomes on insomnia symptoms. We conducted a secondary analysis of pretreatment and posttreatment data from 569 Black and white patients with clinical insomnia symptoms (Insomnia Severity Index [ISI] ≥ 15) who were randomized to digital CBTI in a large-scale RCT. Study outcomes included number of sessions completed (≥4 sessions considered an adequate dose) and insomnia symptoms (ISI). Black CBTI patients completed fewer CBTI session than white patients (M±SD: 2.81±2.36 vs 3.79±2.45; t(561)=4.06, p<.001). Consequently, Black patients were 1.53 times more likely than white patients to prematurely discontinue CBTI before receiving an adequate dose (premature discontinuation rates: 62.5% vs 40.7%, Χ2=19.28, p<.001). A multivariate linear regression showed that Black patients reported higher ISI scores than white patients after CBTI (b=1.28, p=.009) while controlling for pretreatment ISI (p=<.001). However, race became non-significant in the multivariate model (b=.55, p=.248) when sessions completed was entered as a predictor (b=-.60, p<.001). The PRODLIN method estimated the indirect to be αβ=2.27 with a 95% Confident Interval of 1.59 to 2.96, thereby supporting a significant mediation. We estimated that 55.5% of racial disparities in posttreatment ISI outcomes were mediated through group differences in number of sessions completed. Notably, posttreatment insomnia remission (ISI≤9) rates did not differ between Black and white patients (36.9% vs 39.6% remission, p=.613), whereas insomnia remission rates were nearly twice as high among CBTI patients who received an adequate dose relative to those who discontinued prematurely (46.7% vs 24.9% remission, Χ2=23.46, p<.001). Black patients reported greater insomnia symptoms after CBTI relative to white patients, but these racial disparities in insomnia outcomes were largely attributable to lower rates of CBTI engagement among them. No racial disparities in CBTI outcomes were observed among those who completed ≥4 sessions. Enhancing treatment engagement among Black patients may improve dCBTI efficacy in this group. This RCT (NCT03322774) was supported by NIMH R01-MH122636.