Educational differentials in domain specific physical activity by ethnicity, age, and gender: findings from over 44,000 participants in The UK Household Longitudinal Study (2013-2015).

Abstract

Background: The prevalence of overall physical inactivity remains high, particularly amongst socioeconomically disadvantaged groups. It is unclear however if such inequalities vary systematically by age, sex, or ethnicity, and if there are differing effects across physical activity (PA) domains. \nMethods: We used data from a nationally representative survey of the UK, Understanding Society, with information on educational attainment (our indicator of socioeconomic position), PA and demographics collected in 2013-2015 (N= 44,903). Logistic regression analyses were conducted to test associations of education with three different PA domains (active travel, occupational and leisure time). To examine modification of the associations between education and physical activity in each domain by sex, age and ethnicity, we tested two-way interaction terms (education x ethnicity; education x sex; education x age). \nResults: Lower educational attainment was associated with higher active transportation and occupational physical activity, but lower weekly leisure-time activity. These associations were modified by sex, ethnicity, and age. For example, education-related differences in active travel were larger for females (difference in predicted probability of activity between highest and lowest educational groups: -10% in females, (95% CI: -11.9, 7.9) -3% in males (-4.8, -0.4). The education-related differences in occupational activity were larger among males -35% (-36.9, -32.4) than females -17% (-19.4, -15.0). Finally, education related differences in moderate to vigorous leisure time activity varied most substantially by ethnicity; for example, differences were 17% (16.2, 18.7) for White individuals compared with 6% (0.6, 11.6) for Black individuals. \nConclusions: Educational differences in PA vary by domain, and are modified by age, sex, and ethnicity. A better understanding of physically inactive sub-groups may aid development of tailored interventions to increase activity levels and reduce health inequalities.