INTRODUCTION INSOMNIA IS A PREVALENT HEALTH COMPLAINT, AND THE INCIDENCE INCREASES WITH AGING. Normal age-related changes in sleep patterns, combined with the increased prevalence of health problems and medication use, place older adults at greater risk for devel-

INSOMNIA IS A PREVALENT HEALTH COMPLAINT, AND THE INCIDENCE INCREASES WITH AGING. Normal age-related changes in sleep patterns, combined with the increased prevalence of health problems and medication use, place older adults at greater risk for developing sleep disturbances. Survey data indicate that between 12% and 25 % of healthy seniors—compared to 9% to 12% in the general population—experience chronic sleep difficulties.1 Despite the availability of effective behavioral treatments for insomnia, pharmacotherapy remains the most widely used treatment method.2 Benzodiazepine-receptor agonists are the most frequently prescribed drugs for insomnia, and, despite the advent of new nonbenzodiazepine agents (zolpidem, zaleplon, zopiclone), traditional benzodiazepines (BZ) are still widely prescribed for sleep. Although BZ can be indicated and effective for the short-term management of insomnia,3 most patients for whom a hypnotic is prescribed, particularly older adults, continue using it for prolonged periods.4,5 Unfortunately, the effects of the chronic use of BZ on sleep continuity, architecture, and quality remain poorly documented. Previous findings reported by our group6,7 have shown that the sleep electroencephalogram (EEG) recordings of older adults with subjective complaints of insomnia (medicated or not) was fairly similar to those of self-defined good sleepers (GS). Although a higher number of microarousals were found in BZ users (INSBZ) compared to drug-free insomnia sufferers (INS), sleep architecture and continuity parameters were similar between the two groups of insomnia sufferers, as well as in an age-matched control group composed of self-defined GS. Conversely, subjective (sleep diary) measures revealed significant differences on most sleep-continuity parameters, with both groups of insomnia sufferers reporting more difficulties initiating and maintaining sleep and less total sleep time, relative to GS. Moreover, INSBZ reported worse sleep quality than did INS. Such discrepancies between subjective and objective sleep measures, which have been reported by other researchers,8,9 may be due to the lack of precision in traditional sleep analyses, which may not be sensitive enough to detect subtle EEG changes contributing to the subjective sleep disturbances. Power spectral analysis (PSA) may be a sensitive and powerful tool to evaluate the influence of such factors as age, medication use, and insomnia complaints on the EEG during sleep and to further circumscribe discrepancies between subjective and objective measures of sleep. For example, a higher rate of high-frequency activity (beta) and less slow-wave activity (delta and theta) have been found in young adults with insomnia relative to GS.10-13 Although not exclusive, the increased rate of high-frequency activity in the EEG of insomnia sufferers would be consistent with hyperarousal, a classic feature of this sleep disorder.11,14, 15 Acute and short-term BZ usage is associated with subjective improvements of quality and depth of sleep and freshness upon arising,8 as well as with improved polysomnography (PSG)-defined sleep continuity and total sleep time.8,16 In addition, analyses of the sleep microstructure of short-term BZ users usually indicate an increase of beta and sigma activity and a decrease of delta activity.17-19 To our knowledge, no study has yet documented the EEG power spectra in INSBZ. Furthermore, the effects of chronic use of BZ remain poorly documented, especially in older adults, a segment of the population particularly at risk for prolonged BZ usage. The aim of the present study is to further document the sleep microstructure of older adults with insomnia and using BZ on a chronic basis, in comparison with unmedicated insomniacs and GS.