Abstract
Dystonia is one of the most prevalent movement disorders, but its neural substrates have remained enigmatic despite decades of research. Brain lesions leading to dystonia offer unique causal inference that could lend insight into the neural networks driving the disorder. Here, we studied the published cases of lesions causing dystonia to a variety of different body parts (n=179) and used lesion network mapping to test whether lesion locations mapped to a common brain network. Specificity was investigated by comparing the findings to 216 lesions causing other movement disorders, and 499 lesions causing non-specific neurological symptoms. While lesions causing dystonia occurred in heterogeneous locations, they localised to a common and specific cortico-basal ganglia-cerebellar network. Stronger connectivity to this dystonia network from a causal brain lesion was associated with a greater extent of dystonia throughout the body. Further, there were significant differences in the brain networks depending on the body parts affected by dystonia, which also demonstrated different somatotopic representations in the cerebellum. Our results suggest that dystonia stems from a cortico-basal ganglia-cerebellar network, with the manifestation in the body depending on the preferential involvement of the hubs of the network, and its somatotopy.