Background Previous research has demonstrated that active commuting significantly and independently predicts objectively measured body weight and composition and self-reported diagnosed cardiovascular disease (CVD) in the UK general population. This study aims to further illuminate the relationship between AC and CVD-risk using blood analyte data from the UK Household Longitudinal Study (UKHLS – Understanding Society). The objectives are to test for independent associations between active commuting and three blood biomarkers for CVD.

Methods The UKHLS is a large, longitudinal panel study, collecting data from 40,000 households annually. Nurse-collected data from the UKHLS and BHPS health assessment subsamples (UKHLS waves 2 and 3 respectively) were combined with relevant mainstage individual interview data (n = 13,238). A complete-case analysis approach yielded an analytic sample size of 4210. Active commuting was operationalised as a 3-category variable (private, public and active transportation modes). Three CVD-related blood analyte outcomes were identified: (1) Total cholesterol (dichotomised using ≤5 mmol/l cut-point); (2) High-density lipoprotein (HDL) cholesterol (dichotomised at > 1 mmol/l); (3) triglycerides (dichotomised at <2 mmol/l). Hypothesised socioeconomic, behavioural, health-related and demographic confounders were identified. A series of gender-stratified nested multivariate logistic regression models were fitted for each outcome. A range of sensitivity analyses were undertaken.

Results Compared to their private transport using counterparts, male active commuters were 2.8 times more likely to have protective levels of HDL cholesterol (OR 2.97, CI 1.58–4.95, fully adjusted). Women who commuted via active modes were significantly less likely to have elevated triglyceride levels (OR 0.70, 95% CI 0.50 to 0.98, fully adjusted). These associations were not attenuated by adjustment for hypothesised confounders including age, general health, educational attainment, occupational social class, occupational or leisure physical activity and relevant medications. No significant, independent association were found between active commuting and total cholesterol for either men or women.

Conclusion This study contributes important new findings to the evidence base on active commuting and cardiovascular health, supporting policies to increase functional physical activity but suggesting that important gender differences exist. The use of blood biomarker outcomes is novel in this research area, and adds weight to evidence from previous studies which have used self-reported and anthropometric CVD risk outcomes, by illuminating a further section of the CVD causal chain. The main limitation of this study is the cross-sectional design. A repeat of the UKHLS nurse assessment would provide longitudinal data and important potential to strengthen causal inference.