Association of air pollutants, weather variables, and myocardial infarction incidence in Berlin. A study of the Berlin Brandenburg Myocardial Infarction Registry (B2HIR) based on 17873 cases

Abstract

Abstract Background Harmful health effects of air pollutants have long been assumed. Respective associations with respiratory diseases have frequently been reported. Recently, studies also pointed at possible links between air pollution and cardiovascular disease. Currently, nitric oxide (NOx) and particulate matter are of concern in Germany. Purpose To investigate the association of NOx, particulate matter (PM10), and weather variables with the incidence of myocardial infarction (MI) in Berlin. Methods The Berlin Brandenburg myocardial infarction registry (B2HIR) is documenting symptom onset, patient characteristics, comorbidity, procedural data, and outcome of patients with acute MI (onset ≤24h before admission) in Berlin and Brandenburg. From 2008 to 2014 (study period) the stable number of enrolling cardiology departments enabled epidemiologic analyses. Daily (aggregate and variable-specific) MI incidence was assessed for the study period. The government of Berlin provides detailed regional air pollution data (BLUME network). Daily and local NOx and PM10 concentrations were assessed and grouped according to regions (downtown, main roads, suburbs). Preceding days (single and 3 day average) values were also assorted to every single day. Ambient temperature, precipitation, and sunshine duration were assessed from the Berlin-Tempelhof weather station. Bivariate correlations and Poisson regression were calculated to examine associations between air pollutants, weather parameters, and MI incidence. Results 17873 MI cases were included over 7 years. Bivariate correlations and the Poisson regression model revealed a highly significant independent association of the overall MI incidence with same day NOx concentrations (mean values across all measuring stations) and with average PM10 concentration over three preceding days. Based on regression coefficient and interquartile range of pollutant concentrations, MI incidence variation with same day NOx was 3.2% (p<0.001) and variation with three preceding days PM10 was 4.8% (p<0.001). Daily peak ambient temperature was inversely related to MI incidence (p<0.001), whereas for sunshine duration and precipitations no independent association could be identified. Interestingly, the MI incidence in current smokers was unaffected (p=0.145 and p=0.104, respectively) by NOx and PM10 pollution. Conclusion This study is the first to show significant associations between the air pollutants NOx (same day level) and PM10 (3 day preceding average level) and MI incidence in Berlin. MI incidence varied with these pollutants at single digit percentage points. A causal relationship appears likely, as continuously self-intoxicating smokers did not exhibit this association. Funding Acknowledgement Type of funding sources: None.