Increasing in mean temperature affects mortality by stroke in a Tropical environment

BACKGROUND In 2008 more than 9 million people died prematurely from cardiovascular disease (CVD). About 8 million of these deaths occurred in peripheral countries (World Atlas of Cardiovascular Diseases, 2011). In Brazil, non-communicable chronic diseases are responsible for the largest percentage of all deaths among men and women. In 2010, the leading cause of death in the country were diseases of the circulatory system, which mainly affected the age group above 60 years, with 320,000 deaths, accounting for 29% of total deaths. Among cardiovascular diseases, stroke is the leading cause of death, accounting for 10% of total deaths (Ministry of Health, 2012). Stroke is a serious public health issue as it is a major cause of death worldwide and is the leading cause of disability and second in deaths after coronary heart disease (WANG et al, 2009). Several studies have shown the influence of climatic variables on hospital admissions and deaths from stroke. However, this effect is not fully understood. In temperate locations in the United States and Taiwan, relations between hospital admissions and deaths from stroke and atmospheric variables (minimum or maximum temperature, atmospheric pressure, relative humidity) were found (EBI et al, 2004; LEE et al, 2008). However, other studies conducted in temperate locations, as Brussels, UK and Canada, found no association between atmospheric variables on the incidence of stroke (CAPON et al, 1992; FIELD, 2000; COWPERTHWAITE; BURNETT, 2011). In subtropical and tropical climates, where people are acclimated to the hot weather, few data are available on the relationship between temperature and stroke (WANG et al., 2009). In São Paulo, with a population of more than 11 million people in 2013, located at the latitude 23º32'52" S, where normal mean, minimum and maximum temperatures are 19,4º, 15,2º, 25,4ºC, these relationships were not investigated so far. Thus, the aim of this study was to investigate the influence of temperature on stroke mortality in São Paulo. METHODS Daily mortality for stroke, meteorological variables and air pollution data were obtained for the period of January 2002 to December 2011. The percentage increase of mortality for total stroke, hemorrhagic, and ischemic were estimated with endpoint-specific generalized additive Poisson regression models. Nonparametric smooth function was used to control seasonality and an indicator of holidays was adopted to control short-time trend. Effects of mean temperature, humidity, thermal amplitude, barometric pressure and air pollutants were assessed using specific third-degree polynomial distributed lag models for a time window of eight days. The models were controlled for humidity, atmospheric pressure, thermal amplitude and pollutants. The use of Generalized Additive Models (GAM) is an alternative that allows, in a simple way, the control of these confounding factors, since this type of model eliminates the need to specify a parametric form for the association between predictors and response. The R (R Development Core Team 2014) software was used for all analyses. RESULTS From GAM results we observed an estimated increase of 6.2% in total stroke mortality for each interquartile range increase in mean temperature in the first two days. For women, the risk is 5.7%, and 4.8% for males. For people over 65 years, the percentage increases 1.3, 7.8 and 1.1% for women and males, respectively(Table 1). Ischemic stroke (IS) mortality also increased 12% for each interquartile range. For female gender the increase was 5% on the second day of exposure of the mean temperature (Table 2). For men no significant results were found. For the higher age of 65 the percentage increases to 14.1% and remains similar to the female gender (4.7%) For hemorrhagic stroke no significative association was observed with mean temperature. CONCLUSION The increase of mean temperature affects total stroke mortality and, in especial, the ischemic subgroup in this city in a Tropical area. This reinforces the concept that the subgroups of stroke, ischemic and hemorrhagic, may present different pathophysiologic mechanisms and, hence, different risk factors.