Clustering Accelerometer Activity Patterns from the UK Biobank Cohort

Abstract

Many researchers are beginning to adopt the use of wrist-worn accelerometers to objectively measure personal activity levels. Data from these devices are often used to summarise such activity in terms of averages, variances, exceedances, and patterns within a profile. In this study, we report the development of a clustering utilising the whole activity profile. This was achieved using the robust clustering technique of k-medoids applied to an extensive data set of over 90,000 activity profiles, collected as part of the UK Biobank study. We identified nine distinct activity profiles in these data, which captured both the pattern of activity throughout a week and the intensity of the activity: “Active 9 to 5”, “Active”, “Morning Movers”, “Get up and Active”, “Live for the Weekend”, “Moderates”, “Leisurely 9 to 5”, “Sedate” and “Inactive”. These patterns are differentiated by sociodemographic, socioeconomic, and health and circadian rhythm data collected by UK Biobank. The utility of these findings are that they sit alongside existing summary measures of physical activity to provide a way to typify distinct activity patterns that may help to explain other health and morbidity outcomes, e.g., BMI or COVID-19. This research will be returned to the UK Biobank for other researchers to use.

Weekday and weekend physical activity patterns and their correlates among young adults

“…This finding corroborates prior studies showing that women have more total accelerometer-measured daily physical activity compared to men. 4,8,37 Some of this sex difference may also be explained by engagement in household chores. Although time use statistics from Finland suggest that the gap between women and men in time used for household chores has diminished during the last decades, women still use more time in household chores compared to men across all age groups 38,39 and these activities are likely well captured with wrist-worn accelerometers.…”

“…Previous studies based on the National Health and Nutrition Examination survey from the US and the UK Biobank study among adult populations have identified the following groups: consistently inactive, consistently active, active only on weekdays/weekends, and morning/ evening movers. 8,9 The main limitation in these large studies is that they have examined physical activity patterns over the entire week, which does not take into account possible heterogeneity in daily patterns between weekdays and weekends. Moreover, heterogeneity of physical activity patterns have so far been examined in working-age populations including predominantly middle-aged and older adults.…”

“…3,4,22 Previous studies applying cluster and latent class analysis have shown that the consistently least active individuals were more often unemployed and had lower education compared with the higher activity groups. [8][9][10] Accelerometer-based physical activity data have also indicated that married individuals are more likely to belong to the more active groups compared to the lowest activity group, 10 and that individuals with children have higher levels of total daily physical activity compared to nonparents. 23 Collectively, these observations lend support to the notion that these individual-level characteristics may be correlates of the weekday and weekend physical activity patterns also among young adults.…”

“…This information is important for the development of targeted physical activity interventions. In previous literature heterogeneity in physical activity patterns across the week has been identified using a cluster analysis 8 and latent class analysis, 9–11 which both enable identification groups of individuals who share a similar weekly physical activity pattern. Previous studies based on the National Health and Nutrition Examination survey from the US and the UK Biobank study among adult populations have identified the following groups: consistently inactive, consistently active, active only on weekdays/weekends, and morning/evening movers 8,9 .…”

Association between circadian physical activity patterns and mortality in the UK Biobank

“…Such methods provide data on PA type, dose, and timing but they suffer from PA measurement error (2). Accelerometers are increasingly preferred as PA measure (5) because they show high validity (6) and quantify temporal nuances in movement behaviors (7). However, raw accelerometry data show considerable within and between subject heterogeneity and the data volume and complexity are challenging to analyze (8).…”