Di Marco · Sleep 2024 · pooled post hoc analysis of randomized controlled trials · n=?

Effect of daridorexant on sleep architecture in patients with chronic insomnia disorder: a pooled post hoc analysis of two randomized phase 3 clinical studies.

Cited 15 times in the scientific literature.

Level 2 - randomized trial

Post hoc secondary analysis of pooled data from two randomized controlled trials.

PubMed 38644625 · doi:10.1093/sleep/zsae098 · record verified 2026-08-31

What was done

This study was a pooled post hoc analysis of two randomized phase 3 clinical trials (NCT03545191 and NCT03575104) assessing 3 months of daily daridorexant (placebo, 25 mg, or 50 mg) in adults with chronic insomnia disorder. Investigators evaluated sleep-wake transition probabilities, EEG spectral power across sleep stages, sleep spindle properties (density, dispersion, slow oscillation phase coupling), and a machine-learning-derived wake EEG similarity index (WESI).

What was found

At month 3, daridorexant 50 mg significantly reduced wake-to-wake transition probabilities (p < .05) and increased transitions from wake to N1 (p < .05), N2 (p < .05), and REM (p < .05), as well as N1-to-N2 transitions (p < .05) compared to baseline and placebo. During wake, 50 mg daridorexant decreased relative beta power (p = .011) and relative alpha power (p < .001), while increasing relative delta power (p < .001) compared to placebo. Relative beta power in N1 also decreased (p < .001). Daridorexant reduced WESI scores during wake versus baseline (p = .004). No significant changes were observed in N2, N3, or REM spectral bands, or in sleep spindle characteristics. Effects were consistent across months 1 and 3, and were less pronounced with the 25 mg dose. Exact numerical estimates were not reported in the abstract.

Why it matters

These findings suggest that dual orexin receptor antagonism with daridorexant reduces neurophysiological hyperarousal during wakefulness and sleep onset without disrupting deeper non-REM sleep architecture or physiological sleep spindle dynamics.

Limits

The analysis was post hoc, raising the potential for exploratory bias. The abstract does not provide exact participant counts, point estimates, or confidence intervals, providing only p-values.

Cited by