Longitudinal associations of air pollution and green space with cardiometabolic risk factor clustering among children in the Netherlands

Abstract

Background: This study examines longitudinal associations of air pollution and green space with cardiometabolic risk among children in the Netherlands.

Methods: Three Dutch prospective cohorts with a total of 13,822 participants aged 5 to 17 years were included: (1) the Amsterdam Born Children and their Development (ABCD) study from Amsterdam (n = 2,547), (2) the Generation R study from Rotterdam (n = 5,431), and (3) the Lifelines study from northern Netherlands (n = 5,844). Air pollution (PM_2.5, PM₁₀, NO₂, and elemental carbon (EC)) and green space exposures (density in multiple Euclidean buffer sizes) from 2006 to 2017 at home address level were used. Cardiometabolic risk factor clustering was assessed by a MetScore, which was derived from a confirmatory factor analysis of six cardiometabolic risk factors to assess the overall risk. Linear regression models with change in Metscore as the dependent variable, adjusted for multiple confounders, were conducted for each cohort separately. Meta-analyses were used to pool cohort-specific estimates.

Results: Exposure to higher levels of NO₂ and EC was significantly associated with increases in MetScore in Lifelines (per SD higher exposure: β_NO2 = 0.006, 95 % CI = 0.001 to 0.010; β_EC = 0.008, 95 % CI = 0.002 to 0.014). In the other two cohort studies, these associations were in the same direction but these were not significant. Higher green space density in 500-meter buffer zones around participants' residential addresses was not significantly associated with decreases of MetScore in all three cohorts. Higher green space density in 2000-meter buffer zones was significantly associated with decreases of MetScore in ABCD and Lifelines (per SD higher green space density: β_ABCD = -0.008, 95 % CI = -0.013 to -0.003; β_Lifelines = -0.002, 95 % CI = -0.003 to -0.00003). The pooled estimates were β_NO2 = 0.003 (95 % CI = -0.001 to 0.006) for NO₂, β_EC = 0.003 (95 % CI = -0.001, 0.007) for EC, and β_{500m buffer} = -0.0014 (95 % CI = -0.0026 to -0.0001) for green space.

Conclusions: More green space exposure at residence was associated with decreased cardiometabolic risk in children. Exposure to more NO₂ and EC was also associated with increased cardiometabolic risk.

Keywords: Cardiometabolic diseases; Cohort study; Environmental health; Pediatrics.