Objective: Before the deep brain stimulation of the subthalamic nucleus (STN-DBS) an individual patient clinical predictor would be useful to inform the patient about freezing of the gait outcome.

Background: The response of gait freezing to STN-DBS is controversial and obviously depending very much on factors which are poorly controlled.

Method: A total number of 124 patients undergoing STN-DBS were included in the study. With regard to its predictive value we compared the freezing of gait severity based on item 14 of the Unified Parkinson’s disease rating scale, the UPDRS total motor score, and a freezing of gait scale rated during 4 freezing of gait provoking circumstances based on the video-documented levodopa test at baseline and the outcome in the worst and best condition 1 year postoperatively.

Results: We found ‘freezing during the turning task’ to be the best predictor with a ROC-value of 0.857 compared to 0.603 for the UPDRS Item 14 and 0.583 for the total UPDRS III. An improvement of 1 or 2 grades of the turning item during the preoperative levodopa test predicts an improvement during the worst condition postoperatively of 1 grade or more with an 80% probability.

Conclusion: This freezing prediction test is simple and clinically useful. The test needs to be studied in a prospective study.