Background: The recent cohort study by Nilsson et al. provides the first registry-based comparison of neurological and psychiatric morbidity between dermatitis herpetiformis (DH) and celiac disease (CD) without DH, reporting divergent risk profiles. Methods: A critical appraisal was conducted focusing on methodological design, particularly the reliance on ICD-coded administrative data and its implications for phenotypic ascertainment in neurological and neuropsychiatric outcomes. Results: The study reports a modest neurological burden in DH (HR 1.27, 95% CI 0.94-1.71) compared with significantly increased risks in CD for headache, seizures, depression, and anxiety (HR 1.31, 95% CI 1.09-1.56). However, two clinically relevant phenotypes are likely under-ascertained. First, cerebral calcifications-central to the Celiac Disease-Epilepsy-Occipital Calcification syndrome-are detectable only through neuroimaging and thus absent from registry data, potentially explaining the non-significant epilepsy risk despite meta-analytic evidence suggesting a 1.4-2-fold increase in CD. Second, neurosensorial manifestations (e.g., sensorineural hearing loss, vestibular dysfunction, neuro-ophthalmic deficits) are typically coded outside neurological classifications, limiting their capture despite their documented contribution to anxiety and cognitive impairment. Conclusion: Future registry-based studies in this field should move beyond conventional neurological ICD code extraction by integrating radiology registry linkage to capture structural brain involvement, expanding ascertainment to neurosensorial diagnostic domains, and stratifying analyses by age at diagnosis and gluten-free diet adherence. Adopting these extensions as methodological standards would substantially deepen phenotypic resolution across the CD-DH spectrum and bring population-level evidence closer to the full clinical reality of gluten-related neurological disease.
Rahmoune et al. (Mon,) studied this question.