High preoperative gait variability is a prognostic predictor of gait and balance in Parkinson disease patients with deep brain stimulation

The objective biomarker for prediction of gait and balance in the long-term follow-up of Parkinson's disease(PD) patients with subthalamic nucleus deep brain stimulation(STN-DBS) has not yet been elucidated. We investigated the value of pre-operative quantitative gait parameters for the prediction of long-term prognosis of gait in PD patients with DBS.We retrospectively collected gait videos(both medication ON/OFF states) of PD patients recorded as preoperative evaluation before STN DBS. We enrolled patients who were followed-up for more than 5 years after the surgery from 2006 to 2014. We derived objective gait parameters from video-based gait analysis algorithm. We defined the clinical milestones of frequent falling, impaired walking, and loss of autonomy based on the Unified Parkinson's disease rating scale and Hoehn and Yahr stage, which were regularly followed up to 156 months after surgery. We calculated hazard ratios(HRs) of baseline gait parameters for predicting the clinical milestones.A total of 96 gait videos from 63 PD patients were analyzed. The mean follow-up duration(standard deviation) was 88.0(34.2) months after surgery. Relatively high (>mean + 1 standard deviation) variability for step length, step time and stride time (HR = 2.92[1.02-8.33], 3.91[1.38-11.11] and 7.16[2.09-24.52],respectively) in medication-ON state significantly predicted reaching any of the three clinical milestones of frequent falling, impaired walking and loss of autonomy. Gait parameters from the medication-OFF state did not predict any clinical milestone.High preoperative gait variability from the medication-ON state predicts long-term outcomes for gait and balance in PD patients with STN DBS.