The prevalence of sleep-disordered breathing has not been well studied in women, especially in terms of the effects of age, body mass index (BMI), and menopause. We evaluated this question using a two-phase random sample from the general population. In Phase I, 12,219 women and 4,364 men ranging in age from 20 to 100 yr were interviewed; and in Phase II, 1,000 women and 741 men of the Phase I subjects were selected for one night of sleep laboratory evaluation. The results of our study indicated that, for clinically defined sleep apnea (apnea/hypopnea index > or = 10 and daytime symptoms), men had a prevalence of 3.9% and women 1.2%, resulting in an overall ratio of sleep apnea for men to women of 3.3:1 (p = 0.0006). The prevalence of sleep apnea was quite low in premenopausal women (0.6%) as well as postmenopausal women with hormone replacement therapy (HRT) (0.5%). Further, in these women the presence of sleep apnea appeared to be associated exclusively with obesity (BMI > or = 32.3 kg/m2). Postmenopausal women without HRT had a prevalence of sleep apnea that was significantly higher than the prevalence in premenopausal women with HRT (2.7 versus 0.6%, p = 0.02) and was more similar to the prevalence in men (3.9%), although it remained significantly less when controlling for age and BMI (p = 0.001). These data combined indicate that menopause is a significant risk factor for sleep apnea in women and that hormone replacement appears to be associated with reduced risk.

The effects of age on the prevalence of sleep apnea in the general population remain unclear, because previous studies have focused on specific populations. The effects of age on the severity of apnea are unknown. This study was based on a two-stage general random sample of men (aged 20 to 100 yr), consisting of a telephone survey (n = 4,364) and a sleep laboratory evaluation of a survey subsample (n = 741). Obstructive sleep apnea (OSA), based on both sleep laboratory and clinical criteria (apnea/hypopnea index [AHI] > or = 10 and the presence of daytime symptoms) was found in 3.3% of the sample, with its maximum prevalence in the middle age group (45 to 64 yr). Also, based solely on laboratory criteria, the prevalence of OSA (obstructive AHI > or = 20) showed an age distribution similar to that of OSA diagnosed by laboratory and clinical criteria. The prevalence of any type of sleep apnea (central and obstructive) increased monotonically with age. However, central apnea appeared to account for this monotonic relationship with age. Severity of sleep apnea, as indicated by both number of events and minimum oxygen saturation, decreased with age when any sleep apnea criteria were used and when controlling for body mass index (BMI). The study shows that the prevalence of sleep apnea tends to increase with age but that the clinical significance (severity) of apnea decreases. On the basis of these findings, the sleep laboratory criteria used for diagnosis of sleep apnea should be adjusted for age.

Total sleep restriction in humans is associated with increased daytime sleepiness, decreased performance, and hormonal/metabolic disturbances. The effects of mild chronic sleep restriction that mimic real life are not known. To assess the effects of modest sleep restriction from 8 to 6 h/night for 1 wk, 25 young, healthy, normal sleepers (12 men and 13 women) were studied for 12 consecutive nights in the sleep laboratory. After 1 wk of sleep restriction, although subjects' nighttime sleep was deeper, subjects were significantly sleepier (multiple sleep latency test) and performed worse in four primary variables of psychomotor vigilance test (both P < 0.01). Furthermore, 24-h secretion of IL-6 was increased by 0.8 +/- 0.3 pg/ml (P < 0.05) in both sexes, whereas TNFalpha was increased only in men. Also, the peak cortisol secretion was lower after sleep restriction than at baseline, and this difference was stronger in men (55.18 +/- 24.83 nmol/liter; P < 0.05) than in women (35.87 +/- 24.83 nmol/liter; P < 0.16). We conclude that in young men and women, modest sleep loss is associated with significant sleepiness, impairment of psychomotor performance, and increased secretion of proinflammatory cytokines. Given the potential association of these behavioral and physical alterations with health, well-being, and public safety, the idea that sleep or parts of it are optional should be regarded with caution.

It appears that the presence of EDS is more strongly associated with depression and metabolic factors than with sleep-disordered breathing or sleep disruption per se. Our findings suggest that patients with a complaint of EDS should be thoroughly assessed for depression and obesity/diabetes independent of whether sleep-disordered breathing is present.