Objective To assess the extent to which smoking, moderate-to-vigorous physical activity (MVPA), and screen-based entertainment (SBE) explain the association between socio-economic position (SEP) and CVD risk.

Design Cross-sectional health examination survey linked to mortality data.

Setting The Scottish Health Survey 2003 is a survey of a random sample of the general population living in private households in Scotland.

Participants The cross-sectional component of this study included 2782 adults aged 16 and over who had complete information on all socio-economic and clinical measures used to calculate the SEP score and the cardio-metabolic risk score. The longitudinal component of this study considered 4621 respondents, aged 35 and over who consented to having their records linked to National Health Service administrative data.

Main outcome measures We calculated the percentage of the association between lower SEP and CVD risk that smoking, MVPA, and SBE explain in two ways: a) cross-sectionally using a cardio-metabolic risk score (based on total cholesterol, HDL cholesterol, HbA1c, C-reactive protein, BMI, waist, hypertension) dichotomized as three or more / less than three risk factors as the main outcome, and b) longitudinally with CVD (fatal/non-fatal) events as the main outcome. The main exposure variable in both sets of analyses was a composite SEP score (based on social class, income, and education). A total of 179 incident cardiovascular events including deaths, which occurred over 19 864 person years, an average of 4.3 years, was used in the analysis.

Results In both sets of analyses, SBE explained a larger percentage of the association between SEP and CVD risk than either smoking or MVPA. In the cross-sectional analysis, SBE accounted for 30.0% of the association between lowest SEP and having a cardio-metabolic risk score of three or more, followed by MVPA (16.4%) and smoking (10.9%). A similar pattern emerged from the longitudinal analysis, where SBE emerged as the largest contributor, accounting for 30.4%, to explaining the association between lowest SEP and increased risk of having a CVD event. Smoking explained the next highest percentage (26.7%) and MVPA the least (14.6%). The fully adjusted model with all three variables explained 52.5% of the relationship.

Conclusion Since SBE explains a larger proportion of the association between SEP and CVD risk than smoking or MVPA, public health policies aimed at reducing inequalities in health should include guidance on reductions of sedentary behaviour in addition to guidance already available on smoking cessation and the promotion of physical activity.