Introduction: Gluten ataxia refers to the association of idiopathic ataxia despite exhaustive investigations with gluten sensitivity defined by anti-gliadin antibodies (AGA) presence in blood. This is a controversial concept.
Patients and methods: We screened 33 patients, who were hospitalized in 2003 and had subacute or chronic ataxia for presence of circulating AGA. Twelve patients were positive and their clinical and biological features were studied.
Results: Among the twelve patients, we concluded that gluten ataxia was present in only eight, including one case of celiac disease. Among these eight patients, five had the usual features of gluten ataxia (progressive cerebellar ataxia affecting mainly lower limbs), but one patient presented unusual left cerebellar hemisyndrome and the two others displayed polyneuropathy with proprioceptive ataxia. Cerebellar atrophy was confirmed with magnetic resonance imaging in five cases and association with other antibodies was found in six cases. Among the four other patients positive for AGA, investigations revealed one case of multiple sclerosis, one case of late-onset Friedreich ataxia, one case of basilar tuberculous meningitis and one case of type 2 diabetes.
Conclusion: Screening for AGA presence should be systematically performed at presentation of patients with unknown etiology ataxia; in the event AGA are present without any other etiology, treatment with gluten-free diet must be discussed. However, the responsibility of AGA in the pathogenesis of neurological signs is highly debatable and further experimental work is required. Two pathophysiological hypotheses are suggested: (1) overexpression of cerebellar epitopes, in case of primary cerebellar pathology, leading to excessive immune response directed against these epitopes; and (2) molecular mimicry with cross-reactivity of antigens usually directed against gliadin, but also recognizing Purkinje cells epitopes.