Clin Res Cardiol (2023). https://doi.org/10.1007/s00392-023-02180-w

Association of air pollution and mortality in individuals with high cardiovascular risk
R. Maitra1, D. Grün2, L. K. Elsner3, J. S. Wolter4, M. Weferling4, S. Kriechbaum4, C. Liebetrau5, C. W. Hamm3, B. von Jeinsen4, T. Keller1
1Franz-Groedel-Institut (FGI), Justus-Liebig-Universität Giessen, Bad Nauheim; 2Medizinische Klinik I, Kardiologie und Angiologie, Justus-Liebig-Universität Giessen, Gießen; 3Medizinische Klinik I - Kardiologie und Angiologie, Universitätsklinikum Gießen und Marburg GmbH, Gießen; 4Abteilung für Kardiologie, Kerckhoff Klinik GmbH, Bad Nauheim; 5CCB am AGAPLESION BETHANIEN KRANKENHAUS, Frankfurt am Main;

Background and Aim: Environmental factors have been associated with negative cardiovascular outcomes in the general population. There is only little knowledge about the impact of extrinsic risk factors such as air pollution on people with pre-existing cardiovascular disease.  So far, well-known risk evaluation measures such as the ESC (European Society of Cardiology) SCORE 2 consider only established patient intrinsic risk factors. This study aims to evaluate the prognostic impact of the extrinsic risk factor air pollution on a population with pre-existing cardiovascular conditions in Hesse and comparing it with established risk-factors.

Methods: A total of 4610 patients (median age 69.0 [Interquartile Range (IQR59.8-75.66], 32% female) scheduled for coronary angiography were enrolled in a prospective registry cohort between 2010 and 2019 and were used in the present analyses. Baseline characteristics were calculated for each patient. As primary outcome measure overall  mortality was used (1122 patients died). Publicly available air pollution data of the Hessian State Agency for Nature Conservation, Environment and Geology (HLNUG) was assigned to each patient through geocoding on a zip code level. We investigated the parameters carbon monoxide (CO), nitrogen oxide (NO), nitrogen dioxide (NO2), ozone (O3), particulate matter 10 (PM10-Feinstaub < 10 µm), particulate matter 2.5 (PM2.5- Feinstaub <2,5 µm) and sulphur dioxide (SO2) as air pollution markers. We performed receiver-operating characteristic (ROC) analyses for the time periode between the enrolment day and each day up to three years prior to study enrolment to investigate the time dependant association between each air pollution marker and mortality. ROC analyses were carried out for patients’ ESC SCORE 2 at study enrolment.

Results: Reflecting a relevant cardiovascular burden high rates of arterial hypertension (84.54%) and dyslipidemia (81.33%) were observed; other cardiovascular risk factors were less prevalent like diabetes mellitus (27.7%), smoking (19.47%), and obesity (BMI>30) (35.72%).  Presence or previous history of coronary artery disease (51.46%), percutaneous coronary intervention (34.59%), coronary artery bypass graft (12.17%) and myocardial infarction (20.15%) were also recorded among our cohort. The median ESC SCORE 2 was 17 [IQR 13-22]. 

Regarding investigation of the potential association between markers of air pollution and mortality, PM2.5 was the most promising marker with an AUC for PM2.5 to predict mortality  of 0.59 [Confidence Interval (CI): 0.56-0.61] 402 days prior to study enrollment. (Figure 1) In comparison, the ESC SCORE 2 yielded an AUC of 0.57 [CI: 0.55-0.59] to predict mortality. For the remaining air pollution markers the highest AUC results were 0.52 [CI: 0.5-0.54] (CO), 0.54 [CI: 0.52-0.56] (NO), 0.55 [CI: 0. 0.53-0.57] (NO2), 0.52 [CI: 0.5-0.54] (O3), 0.52 [CI: 0.5-0,54] (PM10) and 0.59 [CI: 0.57-0.6] (SO2) regarding the time period  3, 1077, 1014, 348, 232, 1058 days before study enrolment, respectively.

Conclusion: Time dependant exposure to air pollution was associated with mortality in patients with relevant cardiovascular burden. The predictive value for PM2.5 exposure was comparable to the predictive value of the ESC SCORE 2. Influence of extrinsic risk factors like air pollution on patients with high cardiovascular risk in addition to intrinsic risk factors might further improve individual risk estimation and needs further investigation.


https://dgk.org/kongress_programme/jt2023/aV629.html