RESUMO
BACKGROUND AND AIM: Studies have shown that increased maternal cortisol level is associated with child adverse health outcomes. Hair cortisol (HC) is suitable for assessing long-term circulating cortisol concentration. Only two previous studies reported beneficial associations between cortisol and residential greenness during pregnancy and no study focused on the first trimester. Our aim was to evaluate the association between residential greenness and first trimester HC levels among pregnant women in Israel. METHODS: Women were recruited during second and third trimesters. Hair samples were collected from the scalp and retrospective HC levels during the first trimester were quantified for 217 women. HC levels were natural log transformed and outliers were excluded. Based on geocoded birth address, small area sociodemographic status (SES) and mean residential surrounding greenness were calculated using high-resolution satellite-based Normalised Difference Vegetation Index (NDVI) data at 100, 300 and 500-m buffers in a cross-sectional approach. In addition, longitudinal exposure to mean greenness during a week preconception and during the first trimester were calculated. Missing covariates were imputed and linearity of the associations were evaluated. Generalized linear models were used to estimate the crude and adjusted associations controlled for the relevant covariates. RESULTS: After exclusion of outliers, for 211 women, crude and adjusted beneficial associations between exposure to higher mean NDVI and HC levels were observed for all the exposure measures. An increase in 1 interquartile range of greenness (100 m buffer) was associated with a statistically significant lower estimated natural log mean HC level (-0.27 95% CI: -0.44; -0.11). The associations were robust to adjustment for covariates. The findings were consistent for different buffers, for the longitudinal approach, when all observations were included in the analysis and slightly stronger associations were observed for women with addresses geocoded at the home or street level. For most of the exposure measures, stronger associations were observed among those of lower sociodemographic status. CONCLUSION: Our findings that more greenness associated with reduced maternal cortisol levels measured in the hair during the first trimester, could have substantial implications for urban planners and public health professional. If our observations will be replicated, it may present a useful avenue for public-health intervention to promote health through the provision of greenness exposure during early pregnancy, specifically to disadvantage populations.