Quantifying the Predictive Performance of Objectively Measured Physical Activity on Mortality in the UK Biobank

Abstract

Background Objective measures of physical activity (PA) derived from wrist-worn accelerometers are compared with traditional risk factors in terms of mortality prediction performance in the UK Biobank. Methods A subset of participants in the UK Biobank study wore a tri-axial wrist-worn accelerometer in a free-living environment for up to 7 days. A total of 82,304 individuals over the age of 50 (439,707 person-years of follow-up, 1,959 deaths) had both accelerometry data that met specified quality criteria and complete data on a set of traditional mortality risk factors. Predictive performance was assessed using cross-validated Concordance (C) for Cox regression models. Forward selection was used to obtain a set of best predictors of mortality. Results In univariate Cox regression, age was the best predictor of all-cause mortality (C=0.681) followed by twelve PA predictors, led by minutes of moderate to vigorous PA (C=0.661) and total acceleration (C=0.661). Overall, 16 of the top 20 predictors were objective PA measures (C from 0.578 to 0.661). Using a threshold of 0.001 improvement in Concordance, the Concordance for the best model that did not include PA measures was 0.735 (9 covariates) compared with 0.748 (12 covariates) for the best model with PA variables (p-value<0.001). Conclusions Objective measures of PA derived from accelerometry outperform traditional predictors of all-cause mortality in the UK Biobank except age and substantially improve the prediction performance of mortality models based on traditional risk factors. Results confirm and complement previous findings in the National Health and Nutrition Examination Survey (NHANES).

Diurnal Physical Activity Patterns across Ages in a Large UK Based Cohort: The UK Biobank Study

“…We follow a recent study from Leroux et al [ 9 ] to aggregate the 5-second level data into 1-minute epochs. This aggregation involves averaging the 5-second level data combined with a step to impute the small proportion of missing data in the sample.…”

“…This aggregation involves averaging the 5-second level data combined with a step to impute the small proportion of missing data in the sample. A detailed description of this procedure is provided in the supplemental material of Leroux et al [ 9 ]. We apply the same exclusion criteria as [ 9 ] to determine what constitutes a “good” day of accelerometry data.…”

“…A detailed description of this procedure is provided in the supplemental material of Leroux et al [ 9 ]. We apply the same exclusion criteria as [ 9 ] to determine what constitutes a “good” day of accelerometry data. Specifically, we determine a good day of acceleration data to be one with sufficient estimated wear time (≥95% of the day, or 1368 min), as well meeting the three data quality criteria provided by the UK Biobank, which relate to data calibration and sufficient device wear time.…”

“…In addition, 16.1%, 66.0%, and 17.9% of participants had 3, 4, and 5 good days of weekday accelerometry data, respectively. There is evidence [ 9 ] that individuals who participated in the UK Biobank accelerometry study were younger and healthier than the UK Biobank study as a whole, which has been reported to be healthier than the general UK population [ 27 ]. As a result, our analytic sample is likely subject to multiple sources of healthy participant bias.…”

“…Because physical activity (PA) levels are modifiable for most adults, PA is an attractive target for interventions aimed at improving the quality of life in older adults. In addition, features of PA are known to be highly correlated with prevalence of various health conditions [ 4 ] and risk of mortality [ 5 , 6 , 7 , 8 , 9 ]. Taken together, these observations suggest the power of monitoring PA in a free-living environment to both inform the epidemiology of healthy aging and facilitate safe, independent, home living for aging individuals if incorporated into, for example, a clinical monitoring program through individuals’ primary care provider.…”