Long-term exposure to air pollution, greenness and temperature and survival after a nonfatal myocardial infarction

Abstract

Background: Little is known about the impact of environmental exposures on mortality risk after a myocardial infarction (MI).

Objective: The goal of this study was to evaluate associations of long-term temperature, air pollution and greenness exposures with mortality among survivors of an MI.

Methods: We used data from the US-based Nurses' Health Study to construct an open cohort of survivors of a nonfatal MI 1990-2017. Participants entered the cohort when they had a nonfatal MI, and were followed until death, loss to follow-up, end of follow-up, or they reached 80 years old, whichever came earliest. We assessed residential 12-month moving average fine particulate matter (PM_2.5) and nitrogen dioxide (NO₂), satellite-based annual average greenness (in a circular 1230 m buffer), summer average temperature and winter average temperature. We used Cox proportional hazard models adjusted for potential confounders to assess hazard ratios (HR and 95% confidence intervals). We also assessed potential effect modification.

Results: Among 2262 survivors of a nonfatal MI, we observed 892 deaths during 19,216 person years of follow-up. In single-exposure models, we observed a HR (95%CI) of 1.20 (1.04, 1.37) per 10 ppb NO₂ increase and suggestive positive associations were observed for PM_2.5, lower greenness, warmer summer average temperature and colder winter average temperature. In multi-exposure models, associations of summer and winter average temperature remained stable, while associations of NO₂, PM_2.5 and greenness attenuated. The strength of some associations was modified by other exposures. For example, associations of greenness (HR = 0.88 (0.78, 0.98) per 0.1) were more pronounced for participants in areas with a lower winter average temperature.

Conclusion: We observed associations of air pollution, greenness and temperature with mortality among MI survivors. Some associations were confounded or modified by other exposures, indicating that it is important to explore the combined impact of environmental exposures.

Keywords: Air pollution; Built environment; Exposome; MI survival; Temperature.

Figure 1.. Results of effect modification analyses between PM_2.5 NO₂, NDVI 1230m, summer and winter average temperature and mortality before age 80 among survivors of a nonfatal MI in the Nurses’ Health Study followed 1990–2017 (n = 2,262, 892 deaths, 19,215.7 person-years). ^{a, b, c}

a Associations are expressed per 5 μg/m3 increase for PM_2.5, per 10 ppb increase for NO₂, per 0.1 increase in NDVI 1230m, per 5°C increase for summer temperature and per 5°C increase for winter temperature. b The model was stratified by age in months and calendar year, and was adjusted for race, marital status, living alone, retirement status, occupation of both of the nurses’ parents when she was 16, educational status of her spouse/partner, smoking status, pack years, diet, alcohol consumption, BMI, physical activity, cancer and nSES. The models included PM_2.5, NO₂, NDVI separately and summer and winter average temperature simultaneously. c The following values (median) were used to create low and high exposure indicators: PM_2.5 low < 11.66 μg/m3 =< PM_2.5 high; NO₂ low < 8.83 ppb =< NO₂ high; NDVI low < 0.36 =< NDVI high; summer temperature low < 21.89°C =< summer temperature high; winter temperature low < 0.07°C =< winter temperature high.