Future studies of diagnostic strategies should address the many limitations of the literature noted above. The field could benefit from adoption of a common terminology for fundamental concepts such as apnea and hypopnea, and the relation between AI and AHI should be established in order to allow conversions and comparisons across studies. Researchers should seek to clarify the prevalence of apnea and hypopnea in general populations by gender and age. More naturalistic sleep studies (in the home) are still of interest, as it is possible that much of the uncertainty about the nature of SA and its pathophysiology, risk factors, and clinical consequences derives from the fact that the phenomenon called SA may be altered by the very fact of observing it via standard PSG. All sleep monitoring systems that are proposed as prequalifiers or replacements for PSG must be validated in the setting in which they are intended to be used.