The article by Ford and Kamerow1 in this issue of The Journal has implications for clinicians, epidemiologists, and policymakers in the area of health care utilization and planning. Using a household probability sample,2 a conservative definition of sleep complaints, structured psychiatric interviewing, systematic follow-up, and logistic regression modeling, this is probably one of the most scientifically rigorous epidemiologic investigations of sleep disturbance and psychopathology ever reported.
A surprising result is the high prevalence of psychiatric disorders among respondents complaining of hypersomnia or excessive daytime sleepiness, variously 45.6% to 64%. Over the past quarter-century, most studies of excessive daytime sleepiness, which have been based on treatment-prevalence data (rather than on a household probability sample, as in the article by Ford and Kamerow), have noted the strong association of excessive daytime sleepiness with sleep apnea and narcolepsy-cataplexy but not with psychiatric disorders (see, for example, Coleman et al3). Thus,