LPS and its relationship with subjective–objective discrepancies of sleep onset latency in patients with psychiatric disorders

Subjective–objective discrepancies in sleep onset latency (SOL), which is often observed among psychiatric patients, is attributed partly to the definition of sleep onset. Recently, instead of SOL, latency to persistent sleep (LPS), which is defined as the duration from turning out the light to the first consecutive minutes of non-wakefulness, has been utilized in pharmacological studies. This study aimed to determine the non-awake time in LPS that is most consistent with subjective sleep onset among patients with psychiatric disorders. We calculated the length of non-awake time in 30-s segments from lights-out to 0.5–60 min. The root mean square error was then calculated to determine the most appropriate length. The analysis of 149 patients with psychiatric disorders showed that the optimal non-awake time in LPS was 12 min. On the other hands, when comorbid with moderate or severe obstructive sleep apnea (OSA), the optimal length was 19.5 min. This study indicates that 12 min should be the best fit for the LPS non-awake time in patients with psychiatric disorders. When there is comorbidity with OSA, however, a longer duration should be considered. Measuring LPS minimizes discrepancies in SOL and provides important clinical information.