Depressive and stress symptoms in insomnia patients predict group cognitive-behavioral therapy for insomnia long-term effectiveness: A data-driven analysis.

Background: Insomnia Disorder is characterized by high degree of phenotypic heterogeneity, that might influence treatment response.

Methods: 123 of 294 insomnia patients initially recruited (66.7% females, age=40.59±11.89) were assessed before and after group Cognitive-Behavioral Therapy for Insomnia (CBT-I), as well as at follow-up (7.8±1.6 years after the end-of-treatment). By use of latent class analysis (LCA) we identified insomnia subtypes according with baseline scores of insomnia severity and features, anxiety, depression, stress and sleepiness symptoms, circadian rhythm, and treatment effectiveness (Delta score of Insomnia Severity Index-ISI between baseline and end-of-treatment).

Results: By LCA we revealed three classes: "PURE INSOMNIA", "INSOMNIA+ANXIETY+DEPRESSION+STRESS", and "INSOMNIA+ANXIETY". The improvements in insomnia severity was maintained up to 10 years after the end-of-treatment, but with differences between classes (p<0.05). Class "INSOMNIA+ANXIETY+DEPRESSION+STRESS" showed at the end-of-treatment the largest percentage of responders (63.5% = Insomnia Severity Index decrease ≥8). However, at follow up the effect was less and 48.1% had a moderate or severe insomnia (Insomnia Severity Index >14).

Limitations: The lack of a control group and the absence of a complete clinical assessment at the follow-up limit the interpretability of our results.

Conclusions: Our data driven analysis suggest insomnia heterogeneity can be categorized into sub-classes by depression, anxiety, and stress symptoms. In addition, insomnia patients with stress and depression symptoms maintained highest percentage of clinical depression at the end-of-treatment and insomnia at follow-up, in comparison with others classes. Stress and depression symptoms should be considered risk factors that play an important role in the long-term outcome of CBT-I.