Comparison of the actigraph versus patients' diary information in defining circadian time periods for analyzing ambulatory blood pressure monitoring data

Background Assessment of 24-h changes in blood pressure is one of the unique features that ambulatory blood pressure monitoring (ABPM) can provide, Most studies agree that sleep/wake periods should be based on patients' actual sleep and wake times. Actual wake and sleep time determinations are often based on patients' diary information. Several publications indicate that actigraphy is, at least, as accurate as activity diary in determining sleep/wake periods. Objectives To compare subjects' compliance with actigraphy and diary keeping and to compare actigraphy and diary data in determination of sleep and wake times, mean blood pressures, classification of hypertension, and assessment of nocturnal dipping status. Methods We evaluated ABPM studies of 62 subjects, Blood pressure data were obtained using Spacelabs monitors for 24 h. Sleep and wake times were determined by both the actigraph and patients' activity diary. Results In the 62 studies, 56 subjects had successful actigraphy (90%), 44 had activity diary completion (71%), and 38 subjects had both (61%), There was no statistically significant difference between the mean wake and sleep onset using the two methods, but up to 3 hours' difference in wake or sleep onset was noted in some studies, Although the two methods did not significantly affect the calculated systolic blood pressure (SBP) or diastolic blood pressure (DBP) loads in either awake or sleep periods, approximately 55% of the subjects' dipping status was changed when diary information on sleep time was used compared to actigraph. Conclusions Our data indicate that in children and young adults, compliance with the actigraph was superior to diary completion and use of the actigraphy method rather than diary information changed the interpretation of some ABPM data. Our study suggested that actigraphy is superior to diary keeping in providing the information needed for appropriate interpretation of some ABPM data. (C) 2001 Lippincott Williams & Wilkins.