The effects of changing meteorological parameters on fatal aortic catastrophes

Abstract

Abstract Background Over the span of the last decade, medical research has been increasingly putting greater emphasis on the study of meteorological parameters due to their connection to cardiovascular diseases. The main goal of this study was to explore the relationship between fatal aortic catastrophes and changes in atmospheric pressure and temperature. Methods We used a Cox process model to quantify the effects of environmental factors on sudden deaths resulting from aortic catastrophes. We used transfer entropy to draw conclusion about the causal connection between mortality and meteorological parameters. Our main tool was a computer program which we developed earlier in order to evaluate the relationship between pulmonary embolism mortality and weather on data sets comprised of aortic aneurysm (AA) and acute aortic dissection (AAD) cases, where one of these two medical conditions had led to fatal rupture of the aorta. Our source for these cases were the autopsy databases of Semmelweis University, from the time period of 1994 to 2014. We have examined 160 aneurysm and 130 dissection cases in relation to changes in meteorological parameters. The algorythm implemented in our program is based on a non-parametric a Cox process model. It is capable of splitting slowly varying unknown global trends from fluctuations potentially caused by weather. Furthermore, it allows us to explore complex non-linear interactions between meteorological parameters and mortality. Results Model measures the relative growth of the expected number of events on the nth day caused by the deviation of environmental parameters from its mean value. The connection between ruptured aortic aneurysms (rAA) and changes in atmospheric pressure is more significant than their connection with mean daily temperatures. With an increase in atmospheric pressure, the rate of rAA mortality also increased. The effects of meteorological parameters were weaker for deaths resulting from acute aortic dissections (AAD), although low mean daily temperatures increased the intensity of occurrence for AAD-related deaths. Conclusion The occurrence rate of fatal aortic catastrophes showed a slight dependence on the two examined parameters within our groups.

Keywords

• Acute aortic dissection
• Ruptured aortic aneurysm
• Cox process model
• Non-parametric approach
• Statistical learning