---

---

Nocturnal Awakenings From Sleep In Older Adults: Prevalence, Correlates And Relationship With Global Sleep Dissatisfaction

Michael V. Vitiello, PhD and Maurice M. Ohayon
University of Washington, Seattle, WA
Stanford University, California, CA

The aging process interacts with aspects of sleep in complex ways. For example, while the prevalence of sleep-related complaints, per se, increases significantly with age, the prevalence of diagnosed insomnia does not (Ohayon, Sleep Medicine Reviews 2002). This suggests that the experience of nocturnal awakenings from sleep in older adults may not be the same as that reported by younger adults. Here we explore this possibility.

The study sample was composed of 8,937 community-dwelling individuals, aged > 18, who formed a representative sample of the populations of the states of California, New York and Texas. 1593 individuals in this sample were aged > 65. Subjects were interviewed by telephone using SleepEVAL. The interviews included sleeping habits, health and psychiatric and sleep disorders. The associations among nocturnal awakenings (NA), difficulties returning to sleep (DRS), daytime consequences (DC), global sleep dissatisfaction (GSD), medical and psychiatric illness and sleep disorders were examined.

The prevalence of nightly NA was significantly greater (p < .0001) in older (34%) than in middle-aged (26%) and younger (17%) adults. DRS peaked in middle age (45-65 yrs.) while GSD declined significantly with age. Reports of less frequent NA occurred with relatively low frequency and no age-related differences. Chronic and frequent NA associated with DRS and Daytime Consequences showed no age effect. The age-related increase in nightly NA was not explained by psychiatric illness, OSAS or RLS, although poor physical health and, in particular, pain, may be contributors. Older adults reporting GSD were much more likely to report frequent and chronic NA associated with DRS and CD than were middle-aged or younger adults.

These findings indicate that older adults, despite reporting more nightly NA than younger and middle-aged adults, report considerably less GSD and somewhat less DRS compared to middle-aged adults. Further they suggest either that older adults tend to evaluate their GSD based on NA relative to younger adults, or that younger adults use other factors in addition to NA to evaluate their GSD. This age-related difference in the subjective evaluation of sleep quality is consistent with a previously observed age-related difference in the experience of non-restorative sleep.

Do The Effects Of CBT-I Vary With Insomnia Subtype?

Michael L. Perlis, PhD
Sleep Research Laboratory
University of Rochester

Introduction: The present analysis examines whether treatment outcome with CBT-I varies with insomnia subtype (i.e. initial, middle, or mixed). The data used to address this issue were from two case series data sets drawn from a Behavioral Sleep Medicine Service located at our local sleep disorders center. Services were provided by two experienced CBT-I therapists, both of whom were clinical psychologists.

Methods: 74 patients were classified as having Initial, Middle, or Mixed Insomnia. The criteria for Initial, Middle and Mixed Insomnia were as follows: Initial Insomnia = sleep latency (SL) >30 minutes and wake after sleep onset (WASO) <30 minutes; Middle Insomnia = SL <30 minutes and WASO >30 minutes; Mixed Insomnia = SL >30 minutes and WASO >30 minutes. All patients completed baseline sleep diaries for 1-2 weeks and then underwent CBT-I. Post Treatment effects were estimated based on the sleep diaries for last treatment week. Effect sizes estimates were calculated for each insomnia subtype for SL and WASO.

Results : The mean age of the group was 43.9 yrs +/- 16 and 59% were women. The groups did not differ with respect to age or gender. The average effect size without respect to group was SL = 1.22 and WASO = 1.39. The Initial Insomnia group exhibited a SL effect of 1.17 and 0.55 for WASO. The Middle Insomnia group exhibited a reverse trend where the SL effect was 0.66 and the WASO effect was 1.95. The Mixed Insomnia group exhibited a 1.83 for SL and 1.66 for WASO. Overall, the Mixed Insomnia exhibited - predictably - the largest average effect size (1.75), followed by the Middle Insomnia group (1.30) and then the Initial Insomnia group (0.86).

Conclusion: These results suggest that CBT-I 1) overall, produces comparable effects for SL and WASO, 2) produces its largest clinical effects in association with the presenting subtype of insomnia, and 3) produces significant improvement in sleep continuity within areas which are not related to the presenting subtype of insomnia. In addition, it would appear that CBT-I, while in general is very effective, is most effective for patients with Mixed Insomnia and least effective for patients with Initial Insomnia. This last finding suggests that it may be useful to develop and evaluate strategies which may be used to augment SL effects.