In 37,853 people with celiac disease, an eating-disorder diagnosis was more common than in matched peers

What the research found

A nationwide cohort study in Sweden was published on 12 November 2026. It included 37,853 people with celiac disease diagnosed from 2002 to 2023 and 185,159 matched reference people. Median age at the index date was 28.7 years (interquartile range 11.7 to 53.0), which crosses child, young-adult, and midlife ages, so the age tag is all ages. 23,704 patients (62.6 percent) were female. That is a majority, not a women-only cohort, so the gender tag is all. Median follow-up was 10.9 years (interquartile range 5.9 to 15.8). Any eating disorder was diagnosed in 485 patients and 1,578 reference people. The adjusted hazard ratio was 1.39 (95 percent CI 1.27 to 1.53), most apparent in the first 5 years. Over 20 years, the authors expect about 1 extra case per 222 patients. Biological relatives did not show the same elevation: parents 0.82 (0.61 to 1.12), siblings 1.11 (0.99 to 1.24), and offspring 1.05 (0.92 to 1.20). Absolute risk stayed low. The authors tie the pattern to the disease and to a gluten-free diet, and the relative finding argues against a simple inherited explanation. DOI 10.1001/jamanetworkopen.2026.43815. Not medical advice.

Evidence level: Human study. A study in people, but an association isn't proof of cause.

Read the original at JAMA Network Open