Back to browse results
Estimating the exposure–response function between long-term ozone exposure and under-5 mortality in 55 low-income and middle-income countries: a retrospective, multicentre, epidemiological study
Authors: Tao Xue, Ruohan Wang, Mingkun Tong, Prof Frank J Kelly, Hengyi Liu, Jiajianghui Li, Pengfei Li, Xinghua Qiu, Jicheng Gong, Jing Shang, and Prof Tong Zhu
Source: Lancet, Volume 7, Issue 9; DOI:
Topic(s): Childhood mortality
Children under five
Country: More than one region
  Multiple Regions
Published: SEP 2023
Abstract: Background In 2021, WHO suggested new target concentration limits for long-term exposure to ambient ozone. However, the harmful effects of ozone on vulnerable children have not been sufficiently studied. We aimed to evaluate the association between long-term ozone exposure and mortality in children younger than 5 years (hereafter denoted under-5 mortality) in low-income and middle-income countries (LMICs) and to estimate this mortality burden for 97 LMICs. Methods By combining information from 128 Demographic and Health Surveys, we evaluated the association between the survival status of more than 1·2 million children younger than 5 years from 2457 sampling strata in 55 LMICs and the average peak-season ozone concentration during the life course, using a fixed-effects Cox model. A non-linear exposure–response function was developed by integrating the marginal effects of within-strata variation in exposure. We extrapolated the function obtained from the 55 LMICs to estimate the under-5 mortality burden attributable to ozone exposure in 97 LMICs, in which more than 95% of global deaths in this age group occur. Findings The fixed-effects model showed a robust association between ozone and under-5 mortality. According to the fully adjusted linear model, an increment of 10 ppb in the life-course average peak-season ozone concentration was associated with a 6·4% (95% CI 2·4–10·7) increase in the risk of under-5 mortality. The non-linear exposure–response function showed a sublinear curvature with a threshold, suggesting that the effect of ozone exposure was non-significant at concentrations lower than the first-stage interim target (100 µg/m3) recommended by WHO. Using this function, we estimate that, in 2010, long-term ozone exposure contributed to 153?361 (95% CI 17?077–276?768; 2·3% [0·3–4·1]) deaths of children younger than 5 years in 97 LMICs, which is equivalent to 56·8% of all ozone-related deaths in adults (269?785) in these countries. From 2003 to 2017, the ozone-related under-5 mortality burden decreased in most of the 97 LMICs. Interpretation Long-term exposure to ozone concentrations higher than the WHO first-stage interim target is a risk factor for under-5 mortality, and ozone exposure contributes substantially to mortality in this age group in LMICs. Increased efforts should be made to control ambient ozone pollution as this will lead to positive health benefits. Funding Ministry of Science and Technology of the People's Republic of China and China National Natural Science Foundation.