Background Few studies have attempted to disentangle the complex relationship between socioeconomic status (SES) and ethnicity and its influence on body mass index (BMI) and on change in BMI over a decade. The present study highlights the influence of ethnicity (country of birth) and SES on BMI, adjusting for certain lifestyle factors. Methods The longitudinal self-reported data, based on a simple random sample, were obtained from 1972 women and 1871 men, aged 25-74, interviewed in 1980/81 and re-interviewed in 1988/89 using the same questionnaire. The influence of the independent variables on BMI was analysed by the Generalized Estimation Equation (GEE) method. The change in BMI for different states of the independent variables was analysed in a linear model with the difference in BMI between 1988/89 and 1980/81 as dependent variable, by sex. Results There was a graded relationship between male SES, defined as educational status, and BMI when adjusted for background variables, while for females, only a low educational level was associated with a higher BMI. Southern European men and Finnish women showed a higher increase in BMI than the reference group (Swedes and West Europeans) when adjusted for age, smoking, exercise, education, marital status, and health status. Not taking exercise was related to a higher BMI compared with those who were taking exercise. The BMI levels and smoking habits were little changed between 1980/81 and 1988/89, while exercise habits showed a more variable pattern. When the focus was on change in BMI, former smokers had a larger increase in BMI than never smokers. Further, men who ceased taking exercise had a larger increase in BMI than those who were taking regular exercise on both occasions. Conclusions When adjusted for smoking and the other background variables, SES and ethnicity are two separate independent factors influencing BMI in men and women. When the focus was on change in BMI, former smokers had a larger increase in BMI than never smokers. Further, men who ceased taking exercise had a larger increase in BMI than those who took regular exercise on both occasions.
