Towns with extremely low mortality due to ischemic heart disease in Spain

BACKGROUND The cause of coronary disease inframortality in Spain is unknown. The aim of this study is to identify Spanish towns with very low ischemic heart disease mortality, describe their health and social characteristics, and analyze the relationship with a series of contextual factors. METHODS We obtained the number of deaths registered for each of 8,122 Spanish towns in the periods 1989-1998 and 1999-2003. Expected deaths, standardized mortality ratio (SMR), smoothed Relative Risk (RR), and Posterior Probability (PP) of RR > 1 were calculated using Bayesian hierarchical models. Inframortality was defined as any town that displayed an RR below the 10th percentile, an SMR of under 1 for both sexes, and a PP of RR > 1 less than or equal to 0.002 for male and 0.005 for female mortality, during the two periods covered. All the remaining towns, except for those with high mortality classified as "tourist towns", were selected as controls. The association among socioeconomic, health, dietary, lifestyle and vascular risk factors was analyzed using sequential mixed logistic regression models, with province as the random-effects variable. RESULTS We identified 32 towns in which ischemic heart disease mortality was half the national rate and four times lower than the European Union rate, situated in lightly populated provinces spread across the northern half of Spain, and revealed a surprising pattern of geographic aggegation for 23 of the 32 towns. Variables related with inframortality were: a less aged population (OR 0.93, 95% CI 0.89-0.99); a contextual dietary pattern marked by a high fish content (OR 2.13, 95% CI 1.38-3.28) and wine consumption (OR 1.50, 95% CI 1.08-2.07); and a low prevalence of obesity (OR 0.47, 95% CI 0.22-1.01); and, in the case of towns of over 1000 inhabitants, a higher physician-population ratio (OR 3.80, 95% CI 1.17-12.3). CONCLUSIONS Results indicate that dietary and health care factors have an influence on inframortality. The geographical aggregation suggests that other factors with a spatial pattern, e.g., genetic or environmental might also be implicated. These results will have to be confirmed by studies in situ, with objective measurements at an individual level.