Results

Characteristics of the study population are presented in Table 1. Among the 2568 participants included, 1257 (48.9%) were female. Fourteen percent of children were exposed to maternal smoking during pregnancy, and about 5% of mothers had been exposed to secondhand smoke during pregnancy. The percentage of cohort members who had developed allergic rhinitis during the follow-ups increased from 6.4% (up to six years of age) to 26.0% (up to 27 years of age).
Distribution of the mean values of cumulative exposure to NDVI and air pollution levels are presented in Tables E2 (Figure S1) and E3, respectively. The mean cumulative NDVI (SD) values within 300 m during pregnancy in spring were significantly lower among children with allergic rhinitis up to six years of age compared to children without allergic rhinitis [0.200 (0.158) vs. 0.244 (0.201), p =0.046]. Similarly, lower mean cumulative NDVI (SD) values within 300 m during pregnancy in summer were also observed among children with allergic rhinitis up to 27 years of age compared to children without allergic rhinitis [0.794 (0.428)vs. 0.834 (0.436), p =0.037] (Table S4). Although no statistically significant differences were found, the mean NDVI values during pregnancy and early-life for both seasons were lower among children with allergic rhinitis up to 12 years of age compared to children without allergic rhinitis (Table S4).
Figure 1 shows the adjusted associations between residential exposure to NDVI at specific time points and allergic rhinitis. There was no clear evidence of association between the cumulative exposure to NDVI during pregnancy and allergic rhinitis (Figure 1, Table S5). However, an increase in cumulative exposure to NDVI during spring season in early-life was associated with an increased risk of allergic rhinitis up to 12 years of age [HR (95% CI) = 1.726 (1.078; 2.765)] and up to 27 years of age [HR (95% CI) = 1.703 (1.139; 2.545)]. During the summer, an increase in the cumulative exposure to NDVI was associated with a decrease in the risk of allergic rhinitis up to 12 years of age [HR (95% CI) = 0.754 (0.585; 0.972)] and up to 27 years of age [HR (95% CI) = 0.801 (0.649; 0.989)]. Similar results were observed when considering the 500 m and 1000 m buffer sizes (Table S5).
Air pollution levels modified the association between early-life cumulative exposure to NDVI and allergic rhinitis. In children exposed to high levels of primary air pollutants (PM10, PM2.5, SO2, and NO2), an increase in NDVI during early-life in summer decreased the risk of allergic rhinitis (Table 2). In children exposed to low levels of primary air pollutants, an increase in NDVI values in early-life in spring increased the risk of allergic rhinitis up to 12 and 27 years of age. Similar results were observed between early-life cumulative exposure to NDVI during spring and summer within larger buffer sizes (500 m and 1000 m) and allergic rhinitis among children exposed to different levels of air pollution (Table S6). No significant associations were observed between the cumulative exposure to NDVI during pregnancy in both seasons and allergic rhinitis in children exposed to different levels of air pollution (Table 2). Additionally, the mediation analysis revealed that associations between NDVI and allergic rhinitis do not appear to be mediated by ambient air pollution (Table S7).