Association between early life exposure to agriculture, biodiversity, and green space and risk of inflammatory bowel disease: a population-based cohort study

Abstract

Background: Specific pollutants and environmental exposures are implicated in modulating inflammatory bowel disease (IBD) risk. However, the role of environmental exposures, particularly during the early life period, towards IBD risk, has not been systematically evaluated.

Methods: We conducted a nationwide population-based cohort study during the study period extending from January 1, 1995, to September 1, 2020, using cross-linked Danish registers, maps, and inventories to ascertain the impact of agricultural land use, biodiversity, green space, urban space, blue space, and normalized difference vegetation index during pregnancy and the first two years of life on IBD, Crohn's disease (CD), and ulcerative colitis (UC) risk, using adjusted Cox proportional hazards regression analyses. We adjusted for covariates sex, maternal age at delivery, calendar year of birth, municipal-level socioeconomic status, and first-degree relative with IBD.

Findings: Of 1,438,487 individuals included in the study who were followed from age 2 years until a median (IQR) age of 14 (8-20) years, 3768 individuals were diagnosed with IBD. Exposure to the second, third and highest quartiles of agriculture land use during early life, relative to the lowest quartile, were associated with increased CD risk (aHR 1.12, 95% CI 1.01, 1.26, 1.19, 95% CI 1.05, 1.34 and, 1.24 95% CI 1.06, 1.46, respectively). There was no association of agriculture land use with UC risk. Conversely, exposure to the third quartile of biodiversity in early life, compared to the lowest quartile, were associated with a lower CD risk (aHR 0.86, 95% CI 0.75, 0.98). A protective effect of greenspace was noted in the highest quartile for CD (aHR 0.87, 95% CI 0.78, 0.98).

Interpretation: In a nationwide cohort with long-term follow up data, early life environmental exposures were associated with modulation of CD risk, with a harmful effect of agriculture land use and protective effect of biodiversity and green space.

Funding: Danish National Research Foundation, the International Organization for the Study of Inflammatory Bowel Disease, the National Institute of Diabetes and Digestive and Kidney Diseases.

Keywords: Crohn's disease; Epidemiology; Land use; Natural environment; Ulcerative colitis; Urban space.

Fig. 1

A. Histograms of distributions of exposure variables by quartile and B. Correlation across variables. NDVI, normalized difference vegetation index.

Fig. 2

Forest plot demonstrating the unadjusted and adjusted hazards ratios for the outcomes inflammatory bowel disease across quartiles of early life environmental exposure variables. The adjusted model is adjusted for sex (female, male), maternal age at delivery (continuous), calendar year of birth, municipal-level socioeconomic status, clustering on municipality of residence, and first-degree relative with IBD prior to the birth of the child (yes, no). Bold numbers indicate number of IBD cases, person-years at risk, and HRs for statistically significant adjusted HRs, p<0.05. CI, confidence interval; HR, hazard ratio; IBD, inflammatory bowel disease; NDVI, normalized difference vegetation index.

Fig. 3

Forest plot demonstrating the unadjusted and adjusted hazards ratios for outcomes (A) Crohn's disease and (B) Ulcerative colitis across quartiles of early life environmental exposure variables. A. The adjusted model is adjusted for sex (female, male), maternal age at delivery (continuous), calendar year of birth, municipal-level socioeconomic status, clustering on municipality of residence, and first-degree relative with IBD prior to the birth of the child (yes, no). CI, confidence interval; CD, Crohn's disease; HR, hazard ratio; NDVI, normalized difference vegetation index. B The adjusted model is adjusted for sex (female, male), maternal age at delivery (continuous), calendar year of birth, municipal-level socioeconomic status, clustering on municipality of residence, and first-degree relative with IBD prior to the birth of the child (yes, no). CI, confidence interval; HR, hazard ratio; NDVI, normalized difference vegetation index; UC, Ulcerative colitis. Bold numbers indicate number of CD / UC cases, person-years at risk, and HRs for statistically significant adjusted HRs, p<0.05.