Amplitude reduction of the circadian temperature and sleep rhythms in the elderly

This study examined the relationship between circadian temperature rhythm amplitude reduction and sleep consolidation parameters in St group of healthy elders experiencing a 6-h phase advance in routine. Twenty-five healthy old people (15 women, 10 men, 77-91 years old) lived in a time-isolation apartment. Throughout the study, subjects were instructed when to go to bed, get up, and take meals. The experiment started with 5 baseline days during which subjects were kept to a daily routine corresponding to their habitual sleep-wake cycle, The wake time of the 6th night was phase-advanced by 6 h and the routine for the remainder of the experiment was held constant at the new earlier phase position, Rectal temperature was recorded continuously and all sleep episodes recorded polygraphically. Time series of temperature data for each subject were analyzed by complex demodulation (CD). Five of the subjects were excluded from analysis because: the percentage of variance accounted for by the remodulate was less than 55% for the postshift days and one subject was excluded because he showed an average sleep efficiency of less than 55% during baseline. In the remaining 19 subjects, the phase shift produced a large decrease of the mean amplitude of the temperature circadian rhythm (from 0.45 degrees C to 0.25 degrees C). During the first 3 nights following the phase shift, sleep efficiency was decreased and amount of wakefulness in the first half of the night (WFirst) Was increased. No effect was found for the amount of wake in the second half of the night (WSecond). The change in amplitude of the temperature rhythm was significantly correlated with change in sleep efficiency (r = 0.5; p = 0.03) and with change in WFirst (r = -0.7; p < 0.001). There was no correlation between change in the amplitude of the temperature rhythm and WSecond. These results suggest that in older subjects, amplitude of the output of the circadian oscillator might indeed be involved in the sleep consolidation process but, in the first, rather than the second half of the night.