Results of binomial regressions of prevalence of obesity and overweight by SEP indicator and survey period.

1Estimated from binomial regression model for each prevalence measure and including SEP indicator, age as polynomial terms and four survey periods (1993/96, 1997/00, 2001/04, 2005/08): a positive value indicates that the prevalence is higher in the lower SEP subgroup.
2Estimated from binomial regression model for each prevalence measure and including SEP indicator, age as polynomial terms, survey periods and interaction between four survey periods and SEP indicator: a positive value indicates that the difference in prevalence between SEP subgroups has widened over time.
3Generalised obesity and overweight based on categories of body mass index.
4Abdominal obesity and overweight based on categories of waist circumference.
5Waist circumferences were not collected on the HSE core samples in 1995-96, 1999-00 & 2004.
6Shorter vs longer time in education.
7Manual vs non-manual social class.