We aimed to identify the prevalence of circadian phase and phase angle abnormalities in patients with insomnia.

We conducted a cross-sectional, multicenter study at three sleep laboratories in the United States and Australia. Patients with insomnia and healthy control participants completed a sleep log for 7 days. Circadian phase was assessed from salivary dim light melatonin onset (DLMO) time during a 12-hour laboratory visit.

Seventy-nine patients meeting the Research Diagnostic Criteria for Primary, Psychophysiological, Paradoxical, and/or Idiopathic Childhood Insomnia (46 females, 35.5 ± 12.3 years [M ± SD]) and 21 controls (14 females, 34.4 ± 11.8 years). As compared to controls, patients with insomnia tried to initiate sleep on average at the same clock time (24:17 ± 1:17 hours vs. 24:13 ± 1:30 hours, respectively; p = .84) but had a later average DLMO times (20:56 ± 1:55 hours, 18:17–01:21 vs. 22:02 ± 2:02 hours, 17:11–04:52, respectively; p = .04). Consequently, patients with insomnia slept at an earlier circadian phase than controls (phase angle, bedtime-DLMO 2:13 hours (± 1:43) vs. 3:10 hours (± 1:08), respectively; p = .008), of whom 10% tried to sleep at or before DLMO (compared to 0 controls), and 22% tried to sleep before or within 1 hour after DLMO (compared to 6% of controls).

A substantial proportion (10%–22%) of patients with insomnia initiate sleep at too early a circadian phase, implicating a circadian etiology for their insomnia. Outpatient circadian phase assessments should be considered to improve differential diagnoses in insomnia and to inform the development of appropriately timed circadian-based treatments.