Association between accelerometer-measured amplitude of rest-activity rhythm and future health risk : a prospective cohort study of the UK Biobank

BACKGROUND: The health effects of rest-activity rhythm are of major interest to public health, but its associations with health outcomes remain elusive. We aimed to examine the associations between accelerometer-measured rest-activity rhythm amplitude and health risks among the general UK population. METHODS: We did a prospective cohort analysis of UK Biobank participants aged 43-79 years with valid wrist-worn accelerometer data. Low rest-activity rhythm amplitude was defined as the first quintile of relative amplitude; all other quintiles were classified as high rest-activity rhythm amplitude. Outcomes of interest were defined using International Classification of Diseases 10th Revision codes and consisted of incident cancer and cardiovascular, infectious, respiratory, and digestive diseases, and all-cause and disease-specific (cardiovascular, cancer, and respiratory) mortality. Participants with a current diagnosis of any outcome of interest were excluded. We assessed the associations between decreased rest-activity rhythm amplitude and outcomes using Cox proportional hazards models. FINDINGS: Between June 1, 2013, and Dec 23, 2015, 103 682 participants with available raw accelerometer data were enrolled. 92 614 participants (52 219 [56·4%] women and 40 395 [42·6%] men) with a median age of 64 years (IQR 56-69) were recruited. Median follow-up was 6·4 years (IQR 5·8-6·9). Decreased rest-activity rhythm amplitude was significantly associated with increased incidence of cardiovascular diseases (adjusted hazard ratio 1·11 [95% CI 1·05-1·16]), cancer (1·08 [1·01-1·16]), infectious diseases (1·31 [1·22-1·41]), respiratory diseases (1·26 [1·19-1·34]), and digestive diseases (1·08 [1·03-1·14]), as well as all-cause mortality (1·54 [1·40-1·70]) and disease-specific mortality (1·73 [1·34-2·22] for cardiovascular diseases, 1·32 [1·13-1·55] for cancer, and 1·62 [1·25-2·09] for respiratory diseases). Most of these associations were not modified by age older than 65 years or s