Examining the association between physical activity, deprivation and risk of mortality: a UK Biobank Study

Talk Code: 
Hamish Foster
Dr C. Celis, Dr B. Nicholl, Dr. J. Gill , Prof C. O’Donnell, Prof F.S. Mair
Author institutions: 
University of Glasgow: General Practice and Primary Care; Institute of Cardiovascular and Medical Sciences


Every year physical inactivity (PA) contributes to >5 million deaths worldwide. Only 33% of UK adults meet PA recommendations and even fewer do so from deprived areas. However, it is unclear how deprivation influences the risk of mortality associated with physical inactivity. Here, we report the effect of deprivation on PA-associated mortality.


UK Biobank cohort study: data from 502,682 adults 40-65 years old, including measures of PA. Outcome: all-cause mortality. Exposure variables: Townsend deprivation index, total physical activity and sedentary time. Covariates: sociodemographics (age, sex, ethnicity), smoking status, BMI, diagnosis of diabetes, cancer, heart disease (high blood pressure, heart attack, angina, stroke) and long standing illness. All chronic disease data was self-report. Mortality data were obtained from National Registers. Total physical activity (walking, moderate and vigorous physical activity) was based on self-report using a validated questionnaire and computed in MET-h/week. Sedentary time was the sum of time spent driving, using computer and watching television. Cox-proportional hazard models used to investigate associations. Age and sex specific quintiles for physical activity (and sedentary time) and age and sex specific tertiles of deprivation were derived and hazard ratios calculated. Individuals in the most active quintile and least deprived tertile comprised the reference category.


Complete data were available for 495,303 participants: 54.3% women; 94.4% white ethnic background; mean Townsend score -1.30 (SD 3.1). Over a median follow up of 4.9 years 8591 participants died and there were 3433 CVD events of which 2787 were fatal. After adjusting for the covariates listed above there was a trend for increasing risk of mortality with lower total physical activity levels and higher levels of deprivation with the least active quintile and most deprived tertile having the second highest risk of mortality (HR 1.34 (CI 1.18-1.51)). There was a similar trend for sedentary time with the most sedentary quintile and most deprived tertile having the highest risk of mortality (HR 1.56 (CI 1.40 – 1.74).


There was a clear trend of those in more deprived groups, with lower levels of activity, having a higher risk of mortality than their more affluent and inactive counterparts after adjusting for a range of potential confounding variables. These results provide evidence that those in areas of deprivation would benefit from more support to increase physical activity. There is a need to further investigate issues that may influence engagement with PA in those from deprived areas in order to increase our understanding of what kind of supports might prove beneficial.

Submitted by: 
Hamish Foster
Funding acknowledgement: