Celiac disease has been associated with gastrointestinal (GI) cancers in small studies; risks have not been estimated from large populations or based on histopathology analyses.

We examined the risk of GI cancers by using data from cohorts of patients with celiac disease (villous atrophy, Marsh score of 3; n = 28,882) or inflammation (Marsh score of 1–2; n = 12,860); biopsy samples were evaluated at 28 pathology centers. A third cohort included 3705 individuals with latent celiac disease (normal mucosa, but positive serology results). Data were compared with those from an age- and sex-matched population.

Of patients with celiac disease, 372 developed incident GI cancers; 347 patients with inflammation and 38 with latent celiac disease developed GI cancers. In the first year after diagnosis and initial biopsy, celiac disease was associated with 5.95-fold increase in risk of incident GI cancer (95% confidence interval [CI], 4.64–7.64); the hazard ratio [HR] for inflammation was 9.13 (95% CI, 7.19–11.6) and for latent celiac disease was 8.10 (95% CI, 4.69–14.0). After the first year, patients were at no significant increase in risk for GI cancers; the HR for celiac disease was 1.07 (95% CI, 0.93–1.23), for inflammation it was 1.16 (95% CI, 0.98–1.37), and for latent celiac disease it was 0.96 (95% CI, 0.56–1.66). The absolute risk for any GI cancer in patients with celiac disease was 101/100,000 person-years, with an excess risk of 2/100,000 person-years.

Although celiac disease, inflammation, and latent disease all increase risk for GI cancers in the first year after diagnosis, there is no increase in risk thereafter.