Aging and sex are associated with multiple sleep latency test findings and their relationship with self-reported sleepiness.

The aim of this study was to assess age- and sex-related differences in multiple sleep latency test (MSLT) results and in the performance of the Epworth Sleepiness Scale (ESS) at classifying objective hypersomnia (mean sleep latency (MSL) ≤ 8 min). We studied 480 consecutive adults (39.3 ± 15.3 years old [18-93], 67.7% female) who underwent hypersomnia evaluation. We fit linear regression models to investigate associations between age and sex and sleep latencies (mean and for every nap), after adjusting for total sleep time and sleep efficiency (on polysomnography), and REM-suppressing antidepressant effect. A logistic regression was performed to assess whether age and sex were associated with sleep-onset REM period (SOREMP) occurrence. ROC analysis assessed the diagnostic performance of ESS scores to identify a MSL ≤ 8 min in different age/sex groups. For every 10 years of age, there was 0.41 (95% CI 0.11-0.72,  = 0.008) min reduction in MSL. Objectively (MSL ≤ 8 min) sleepy patients had shortening of latencies in naps 4 and 5 with aging. Female sex was associated with a higher MSL only in patients with MSL > 8 min. A 2.4% reduction in the odds of SOREMP occurrence was observed for every year of age in objectively sleepy patients ( = 0.045). ESS scores had a better diagnostic performance in older (≥ 50 years old) men than younger (< 50 years old) women ( < 0.05). Older patients with objectively confirmed hypersomnia may be sleepier in later naps, possibly due to less restorative naps and/or circadian rhythm factors. Self-reported sleepiness is more predictive of objective sleepiness in older men than younger women.