Associations of residential greenness, ambient air pollution, biological sex, and glucocorticoids levels in rural China

Aims: To evaluate the associations between residential greenness and glucocorticoid levels and whether air pollutants and sex modify the relationship between greenness and glucocorticoid level in Chinese rural adults. Methods: We collected cross-sectional survey data from 6055 participants from the Henan Rural cohort. The three-year average residential greenness for participants was assessed using normalized difference vegetation index (NDVI) values from a satellite platform. Liquid chromatography-tandem mass spectrometry was employed to quantify the concentrations of glucocorticoids, which were measured by morning blood draw after at least 8 hr of fasting. A random forest model was employed to obtain the average concentrations of PM1, PM2.5, and PM10. A general linear regression model was performed to estimate the associations of NDVI500−m values with cortisol, 11-deoxycortisol, and cortisone. Furthermore, interaction plots were used to present the interaction effects of particulate matter, sex, and green space on glucocorticoid levels. Results: After adjusting for multiple variables, an elevated average NDVI500−m value in the total population was associated with a decrease in cortisol levels (β = −0.063, 95 % confidence interval (CI): − 0.118, − 0.008), and 11-deoxycortisol levels (β = −0.118, 95 % CI: −0.190, −0.047), as well as an increase in cortisone levels (β = 0.130, 95 % CI: 0.079, 0.181). By adding the interaction terms of air pollutants and residential greenness into the regression model, interaction effects between air pollutants and residential greenness were found (cortisol, PM2.5: P interaction=:0.018; PM10: P interaction=0.016; 11-deoxycortisol, all pollutants: P interaction< 0.001), suggesting that the protective effect of residential greenness on serum glucocorticoids disappeared accompanying with increased concentrations of particulate matter. Moreover, trends towards modification in the association between green space and glucocorticoid levels were also evident by sex, but these did not reach statistical significance (for all glucocorticoids: P interaction> 0.05). Conclusion: Long-term exposure to green space was negatively correlated with cortisol and 11-deoxycortisol levels, and positively correlated with cortisone levels. There may be sex differences in these associations. Moreover, the protective effect of residential greenness on serum glucocorticoids was altered by high levels of particulate matter.