ACTIVITY-REST AND SKIN TEMPERATURE CIRCADIAN RHYTHMS AND SLEEP PATTERNS IN COMMUNITY-DWELLING WOMEN

Purpose: The purpose of the study was to explore activity-rest and temperature circadian rhythms and sleep patterns in ambulatory women of reproductive, perimenopausal, and postmenopausal stages. Study aims were to (1) correlate the patterns of activity-rest with skin temperature and determine (2) lead-lag relation between activity-rest and circadian temperature rhythm, (3) stability of activity-rest patterns during work and non-work days, (4) duration of actigraphy sampling sufficient to reliably capture activity-rest patterns, and (5) sleep parameters in participating women across reproductive stages. Design: Analysis of extant data sets. Sample and Setting: Seven community-dwelling women, 27-to-54-years old Methods: Participants wore an actigraph on the non-dominant wrist and a thermal sensor on the abdomen or under the breast for 13-52 consecutive days. Data were fit to the 3-parameter cosinor model to test for a statistically significant circadian rhythm; acceptable goodness-of-fit was defined as R2 ≥.10. Cosinor parameters from the activity-rest and temperature rhythms were correlated using the Spearman correlation coefficient to determine lead-lag relation. Stability of acrophase on work and non-work days was determined using the paired t-test. The standard error of the mean was co mputed for individual participants and the sample to determine minimum actigraphy monitoring duration to reliably assess activity-rest rhythms. Sleep parameters were scored by the Cole-Kripke method using Actiware™ software (Version 5.5, Mini Mitter). Results: All participants had statistically significant activity-rest and skin temperature circadian rhythms (p ≤ 0.001). Six of seven participants (86%) had a statistically significant correlation (p ≤ 0.05) between activity-rest and temperature; three participants had a strong biologically meaningful (R2 ≥.10) circadian rhythm for both activity-rest and temperature. A lead-lag relationship was estimated in 6 of the 7 participants (86%); one participant had a temperature rise before an increase in activity, indicating temperature led activity, and five participants had an activity rise before an increase in temperature, indicating activity led temperature. Differences in mesor, amplitude, and acrophase of the activity-rest rhythms between work and non-work days were not statistically significant in the majority of women. Weighted standard error of the mean decreased linearly as monitoring days increased after 3 days. Participants in the reproductive and menopausal stages experienced the most interrupted sleep, with a median of 22 awakenings during the sleep period. Participants in the perimenopausal stage averaged more sleep time, least number of awakenings during the sleep period, and least hours of wake after sleep onset than those in the reproductive or menopausal stage. The menopausal women slept the fewest number of hours a day (median, 4.6 hr). Conclusions: Activity-rest and temperature circadian rhythms were correlated in the majority of participants. Lead-lag relationship between activity-rest and temperature rhythms was inconsistent. Activity-rest rhythms were stable in work and non-work days in the majority of participants. A minimum of 4 monitoring days was needed to reliably determine activity-rest rhythms. In this small sample, sleep parameters differed by reproductive stage. The circadian rhythm findings are consistent with those from laboratory studies, but suggest that behaviors in the naturalistic setting may alter the lead-lag relationship between activity-rest and body temperature