Associations between residential greenness, land cover and risk of celiac disease in genetically at‐risk children: Celiac Prediction in Skåne study

Objectives: Our aim was to study the association between residential land cover and greenness during childhood and risk of developing celiac disease (CeD). Methods: The Celiac Prediction in Skåne study prospectively followed 5969 human leukocyte antigen (HLA)-genotyped children born 2001–2004 who were screened for CeD at ages 3, 9, and 15 years in Skåne, Sweden. Among these, 138 (2.3%) children only in the HLA at-risk group were identified and diagnosed with CeD through screening. Children diagnosed with CeD outside the screening program were excluded. For the present study, 2535 children at HLA-risk were included. Residential addresses at birth and screening time points were transformed into spatial coordinates. Coordination of Information on the Environment Land Cover data was collected the corresponding years. Normalized Difference Vegetation Index and Leaf Area Index (LAI) were calculated for the summer months. Results: Higher residential LAI, an indicator of forest and agricultural land cover within a 500-meter buffer, was associated with increased odds of CeD in adjusted logistic regression models at age 3 (OR 1.52, 95% CI: 1.14–1.99) and age 9 (OR 1.62, 95% CI: 1.13–2.27). After false discovery rate adjustment, several associations with LAI remained statistically significant, whereas no land cover classes were associated with CeD. Conclusions: Residential greenness was associated with increased CeD risk in the Skåne province. This exploratory observational finding warrants replication in independent populations before conclusions can be drawn regarding potential environmental drivers in the etiology of CeD.