Celiac disease and risk of neuropsychiatric disorders: A nationwide cohort study

Abstract Introduction Previous studies have indicated that patients with celiac disease (CD) may have an increased risk of developing neuropsychiatric disorders. However, large‐scale epidemiologic studies on the topic are still scarce. We aimed to examine the association between CD and development of neuropsychiatric disorders during an 18‐year follow‐up period. Methods We conducted a prospective cohort study. All Danish patients with an incident diagnosis of CD (ICD‐10 K90.0) from 2000 to 2018 were identified in nationwide registries and compared with birthdate‐ and sex‐matched controls (variable 1:10 ratio) for the development of a neuropsychiatric disease. Individual neuropsychiatric diseases were also examined. The absolute risk was calculated by the cumulative incidence, and the relative risk was estimated in Cox regression models. Results We identified a cohort of 6329 patients with CD diagnosed from 2000 to 2018 and 63,287 matches at risk for developing incident neuropsychiatric disorders. The cumulative incidence of development of any neuropsychiatric disorder was 3.9%, 14.9%, 24.8%, 35.9% after 1, 5, 10, and 15 years of follow‐up, respectively, in patients with CD compared with 1.8%, 9.3%, 18.3%, and 27.0% in controls. Gray's test for equality p < 0.001. The relative risk was HR = 1.58 (95% confidence interval: 1.49–1.68) in CD patients compared with matches. For the individual outcomes, CD was associated with an increased relative risk of developing anxiety, depression, eating disorders, epilepsy, migraine, and stress. We also found indications of an increased relative risk of ADHD, alcoholism, bipolar disorders, and drug abuse, although the associations were less clear. No associations were found between CD and dementia, Parkinson's disease, and schizophrenia. Conclusions In this nationwide study including more than 6000 patients with CD, we found an increased risk of development of a neuropsychiatric disorder compared with age‐ and sex‐matched controls. The causes and the clinical relevance of these associations remain to be elucidated.