Fig. 2

Income inequalities in risks of myocardial infarction incidence, death without myocardial infarction, and death after myocardial infarction by sex (Reference: low income, period 1 (2006–2008)). Hazard Ratios were estimated from multistate survival models including the interaction income group*period; 95%-CI bootstrapped (with replacement) using 1000 replications; * significant interaction term (income group*period) based on 95% CIs; all analyses are controlled for age in single-year age groups (as second-degree polynomial); MI myocardial infarction