Abstract P205: Neighborhood Socioeconomic Status and Long-term Incidence of Cancer After Myocardial Infarction

Abstract

Background: Higher cancer incidence rates in patients with cardiovascular disease compared with the general population were recently documented in several European and North American settings. While different mechanisms were proposed to account for the association, the possibility of differential detection cannot be ruled out. We evaluated cancer incidence and case-fatality according to neighborhood socioeconomic status (NSES) in a cohort of myocardial infarction (MI) survivors. Methods: Patients aged ≤65 years admitted to hospital in central Israel with first MI in 1992-1993 were followed up for cancer (via the National Cancer Registry) and death (via the Population Registry) through 2011. NSES was estimated at baseline through a composite census-derived index (measured on a 1-20 scale) developed by the Israel Central Bureau of Statistics, along with other socio-demographic and clinical variables. Fine and Gray subdistribution hazard regression models were used to assess the hazard ratio (HR) for cancer associated with NSES, with death considered a competing event; Cox models were then used to assess the association with case fatality among incident cancer cases. Results: Among 1380 cancer-free patients at baseline (mean age, 54 years; 81% men), 230 developed cancer during a mean (SD) follow-up of 14 (6) years. The age-adjusted HR for cancer was 1.24 (95% CI: 1.08-1.41) per 1 SD increase in NSES. Further adjustment for individual SES measures and clinical variables (including smoking) did not weaken the association (HR=1.26, 95% CI: 1.09-1.46). Conversely, among patients diagnosed with cancer, NSES was associated with reduced mortality risk (age-adjusted HR=0.78, 95% CI: 0.65-0.93; multivariable-adjusted HR=0.80, 95% CI: 0.64-0.99). Conclusions: Residing in an affluent neighborhood was associated with higher incidence and lower case fatality rate of cancer post-MI, which, to some extent, may reflect detection bias due to increased medical surveillance.