Independent Associations Between Different Measures of Socioeconomic Position and Smoking Status: A Cross-Sectional Study of Adults in England

Abstract

Abstract Introduction To gain a better understanding of the complex and independent associations between different measures of socioeconomic position (SEP) and smoking in England. Aims and Methods Between March 2013 and January 2019 data were collected from 120 496 adults aged 16+ in England taking part in the Smoking Toolkit Study. Of these, 18.04% (n = 21 720) were current smokers. Six indicators of SEP were measured: social grade, employment status, educational qualifications, home and car ownership and income. Models were constructed using ridge regression to assess the contribution of each measure of SEP, taking account of high collinearity. Results The strongest predictor of smoking status was housing tenure. Those who did not own their own home had twice the odds of smoking compared with homeowners (odds ratio [OR] = 2.01). Social grade, educational qualification, and income were also good predictors. Those in social grades C1 (OR = 1.04), C2 (OR = 1.29), D (OR = 1.39), and E (OR = 1.78) had higher odds of smoking than those in social grade AB. Similarly, those with A-level/equivalent (OR = 1.15), GCSE/vocational (OR = 1.48), other/still studying (OR = 1.12), and no post-16 qualifications (OR = 1.48) had higher odds of smoking than those with university qualifications, as did those who earned in the lowest (OR = 1.23), third (OR = 1.18), and second quartiles (OR = 1.06) compared with those earning in the highest. Associations between smoking and employment (OR = 1.03) and car ownership (OR = 1.05) were much smaller. Conclusions Of a variety of socioeconomic measures, housing tenure appears to be the strongest independent predictor of smoking in England, followed by social grade, educational qualifications, and income. Employment status and car ownership have the lowest predictive power. Implications This study used ridge regression, a technique which takes into account high collinearity between variables, to gain a better understanding of the independent associations between different measures of SEP and smoking in England. The findings provide guidance as to which SEP measures one could use when trying to identifying individuals most at risk from smoking, with housing tenure identified as the strongest independent predictor.