Background Conducting epidemiological studies on the short-term effect of low-concentration carbon monoxide (CO) exposure on cardiovascular and cerebrovascular disease mortality among populations in different regions can provide reference for making administrative decisions to control ambient CO pollution, and provide clues for toxicological and physiological research of CO, which has certain practical significance.

Objective This study quantitatively evaluates the short-term effect of low-concentration CO on cardiovascular and cerebrovascular disease mortality.

Methods Daily cardiovascular and cerebrovascular disease, coronary heart disease, and stroke mortalities, air pollutantCO, sulfur dioxide (SO₂), nitrogen dioxide (NO₂), fine particulate matters (PM_2.5), and inhalable particulate matters (PM₁₀) concentrations, and meteorological data (average temperature and relative humidity) of Shanghai from January 1, 2013 to December 31, 2018 were collected. Associations between low-concentration CO and daily mortality due to cardiovascular and cerebrovascular disease, coronary heart disease, and stroke were estimated by generalized additive models, controlling for time trends, day-of-the-week effects, and meteorological variables. Different lag patterns were also included in the models including single-day lag effect (lag0-lag5) and cumulative lag effect (lag01, lag03, and lag05). Two-pollutant models were used after controlling for other ambient pollutants. Excess risk (ER) and 95% confidence interval (CI) were used to indicate the increase of mortality risk.

Results Over the study period, the daily mean CO concentration was (0.74±0.28) mg·m^-3, which was far lower than the current ambient air quality standard in China (4 mg·m^-3). The CO concentration showed a downward trend year by year with obvious seasonal periodicity. The cardiovascular and cerebrovascular disease mortality in Minhang District was 19544. In the single-pollutant models, CO concentration at lag05 showed the strongest associations with cardiovascular and cerebrovascular disease, coronary heart disease, and stroke mortalities. For a 0.1mg·m^-3 increase in average CO concentration, we observed significant increments in mortality of 2.23% (95% CI:1.34%-3.12%) due to cardiovascular and cerebrovascular disease, 1.53% (95% CI:0.19%-2.89%) due to coronary heart disease, and 2.76% (95% CI:1.57%-3.97%) due to stroke. In the two-pollutant models, no significant associations were shown after PM_2.5, PM₁₀, SO₂, and NO₂ were adjusted. The exposure-response curves of CO concentration with the three causes of death at lag05 (max effect) showed a linear and slow increase with no obvious threshold.

Conclusion Low-concentration CO exposure has an acute impact on the mortality risks of cardiovascular and cerebrovascular disease, coronary heart disease, and stroke, while the independent effects after controlling for other air pollutants still need to be studied.