Abstract
INTRODUCTION: The study aimed to explore characteristics of responders to a sleep robot intervention for adults with insomnia, and the likelihood that participants responded to the intervention. METHODS: Data from the intervention and the control group in a randomized waitlist-controlled trial (n = 44) were pooled together after both had undergone the intervention. A repeated measures ANOVA and Friedman tests were used to explore changes over time. Differences in baseline characteristics between responders (n = 13), defined as a reduction of -5 on the Insomnia Severity Index from pre- to post-intervention, and non-responders (n = 31) were analyzed with t-tests and chi-square tests. Finally, logistic regression models were estimated. RESULTS: Baseline anxiety was the only statistically significant difference between responders and non-responders (p = 0.03). A logistic regression model with anxiety and sleep quality as predictors was statistically significant, correctly classifying 83.3% of cases. DISCUSSION: The results imply that people with lower anxiety and higher sleep quality at baseline are more likely to report clinically significant improvements in insomnia from the sleep robot intervention.