Children's Lung Function Was Lower in 2017-2018 than in 1995-1996: The Roles of Air Quality Change and Individual-Level Risk Factors

Abstract

Air quality has improved while the lifestyle of children has changed substantially over the past 2 decades in four Chinese cities. It is unknown how these changes affected the lung function of children. We analyzed data collected in 1995-1996 and 2017-2018. In each period, >2000 children 6-13 years were measured for lung function and surveyed for behavioral, residential, and health conditions. Monitored and modeled data for ambient air pollution were obtained. Age- and covariate-adjusted FVC and FEV₁ values were lower, and the proportions of children with low FVC and FEV₁ were greater in the later period than in the earlier period, while PM_2.5, PM₁₀, SO₂, NO _x , and O₃ levels changed by -42.8 μg/m³ (95% CI, -67.3, -18.6 mg/m³), -65.1 μg/m³ (-113.5, -16.7 mg/m³), -34.6 ppb (-69.2, 0.0 ppb), -27.5 ppb (-60.7, 5.7 ppb), and 1.5 ppb (-4.2, 7.1 ppb), respectively. A 4 ppb O₃ increase was associated with lower FVC by 105 mL (95% CI: 30, 180 mL) in boys and 76 mL (17, 135 mL) in girls. A generational decline in the lung function of children in the 4 Chinese cities has significant public health ramifications. Our findings imply that the reduction of ambient air pollutant concentrations from the 1995-1996 levels to the 2017-2018 levels were not sufficient to make lung function improvements. Risk factors that were absent or not as prominent in the earlier period may be responsible for lowered lung function during the later period.

Keywords: Air Quality; Lifestyle; Lung Function; Ozone; School Children.

1

Annual average concentrations of air pollutants in the eight school districts. Colors represent four cities. Circles represent urban districts, and triangles represent suburban districts. Dotted lines connect data for the same school district. Gray shadings represent years in Period 1 (1995, 1996) and Period 2 (2016, 2017). Horizontal red lines represent the average air pollution levels in Period 1 and Period 2. Units are μg/m³ for particulate matters (PM_2.5 and PM₁₀) and ppb for gaseous pollutants (SO₂, NO _x , and O₃).

2

Measured (unadjusted for covariates) lung function by height group (A) and age group (B). Dots represent measured lung function for individual children, and lines of the box represent the 95th, 75th, 50th, 25th, and 5th percentiles, respectively. Asterisks indicate height and age groups in which lung function measurements were significantly different between Period 1 and Period 2 (by t test). **P < 0.05; *P < 0.1.

3

Changes in adjusted school-specific FVC and FEV₁ associated with an interquartile range (IQR) change in air pollutant concentration from Period 1 to Period 2. The IQR values for O₃, PM_2.5, PM₁₀, NO _x , and SO₂ were 4.0 ppb, −20.76 μg/m³, −42.55 μg/m³, −27.8 ppb, and −30.4 ppb, respectively. Solid circles show the point estimates of associations, and horizontal lines show the 95% confidence intervals. Vertical dashed lines show the reference of no change in lung function. Associations were estimated for all children (squares), boys (circles), and girls (triangles).