performance (Ulfberg et al., 1996), accidents and injuries, and deterioration of the quality of life (see later sections).

A meta-analysis of the case-control studies found that the magnitude of the cognitive disturbance was greatest in individuals with severe OSA. Cognitive domains most affected were attention and executive function (the capacity to plan and organize complex tasks) with only milder effects on memory (Engleman et al., 2000). The meta-analysis also found some cognitive benefit associated with continuous positive airway pressure (CPAP) treatment. In a series of randomized, placebo-controlled crossover trials, people with mild OSA exhibited a trend toward better performance. The failure to detect a robust effect may have been due to the fact that the patients had mild disease, were nonadherent to therapy, or that they had a possibly irreversible component to the cognitive impairment. The cognitive deficits with sleep-disordered breathing are thought to be related to both sleep fragmentation and hypoxemia (Weaver and George, 2005). However, one study showed no clear threshold level between level of hypoxia and performance deficits (Adams et al., 2001). Animal models of chronic episodic hypoxia have led to the hypothesis that cognitive deficits in humans result from injury of nerve cells in the pre-frontal cortex (Beebe and Gozal, 2002), the area of the brain responsible for problem solving, emotion, and complex thought.

Sleepiness is a significant, and possibly growing, contributor to serious motor vehicle injuries. Almost 20 percent of all serious car crash injuries in the general population are associated with driver sleepiness, independent of alcohol effects (Connor et al., 2002). Driver sleepiness is most frequently a manifestation of sleep loss, as discussed below, but other sleep disorders, which have lower prevalence, contribute to the problem, including sleep-disordered breathing, restless legs syndrome, and narcolepsy.

The 20 percent figure, cited above, is the population-attributable risk, which is a key public health measure indicating what percentage of car crash injuries, including fatal injuries of passengers, could be avoided by eliminating driver sleepiness. The finding was based on a population-based case-control study in a region of New Zealand in which 571 car drivers and a matched control sample were asked detailed questions about measures of acute sleepiness while driving (Connor et al., 2002). The study adjusted for potential confounding factors, including alcohol. Crashes examined in this study involved a hospitalization or death. The greatest risk factor for the crashes was sleep loss and time of day (driving between 2:00 a.m. to 5:00 a.m.), but sleep apnea symptoms were not risk factors.