Background and purpose: The diagnosis of Parkinson disease (PD) is made typically on the basis of motor abnormalities. PD is now recognized to have both motor and non-motor manifestations, indicating a need for the development of reliable non-motor diagnostic tests for PD. The aim of the present study was to compare the accuracy of various clinical motor and non-motor tests for the diagnosis of PD.
Methods: Forty-five PD patients (Hoehn and Yahr stages 1–3; mean age 59.5 ± 10.0 years) and 45 healthy controls matched for gender and age completed a clinimetric motor test battery to assess limb bradykinesia, tremor and balance. Non-motor tests consisted of depression, anxiety and smell identification ratings. Area under the receiver operator characteristic curve (AUC) analysis was used.
Results: We found that smell identification was the most accurate predictor of the presence of PD within the overall group of patients and matched control subjects (AUC = 0.886) and also in the subgroups of mild severity (Hoehn and Yahr stages 1–1.5; AUC = 0.923), young-onset (AUC = 0.888) and female PD patients (AUC = 0.797). The second best diagnostic test was the grooved pegboard test for the clinically most affected body side.
Conclusions: We conclude that olfactory function is the most accurate diagnostic predictor within a heterogeneous sample of patients with PD.