People with Parkinson's disease (PD) can be distinguished from controls, according to a model developed and then tested on 1192 patients with PD and 426 controls without neurological disease. Researchers found:

• The model correctly distinguished patients with PD from controls at an area under the curve (AUC) of 0.923, with high sensitivity (0.834) and specificity (0.903) at its optimum AUC threshold (0.655).

• Simulations showed that when parsed into random subgroups, subgroup data matched the overall cohort.

• External validation showed good classification of PD, with AUCs in close agreement for 5 separate cohorts.

• Four of 17 patients with scans without evidence of dopaminergic deficit (SWEDD) who were classified by the model as having PD converted to PD within 1 year.

• Only 1 of 38 patients with SWEDD who were not classified as having PD underwent conversion to PD.

Citation: Nalls MA, McLean CY, Risk J, et al. Diagnosis of Parkinson's disease on the basis of clinical and genetic classification: a population-based modelling study. [Published online ahead of print August 10, 2015]. Lancet Neurol. doi: 10.1016/S1474-4422(15)00178-7.