Fig. 2: An overview of the percentage of patients being divided into the balance-improvement group according to the BBS and PT criterion.

a The percentage of patients being divided into the balance-improvement group according to the BBS criteria in both conditions of off- and on-medication in the 1-month follow-up and 12-month follow-up. The orange and the blue circles indicated dichotomies in the explorative set in off- and on-medication conditions, respectively. The orange dotted line and the blue solid line indicated dichotomies in the validation set in off- and on-medication conditions, respectively. b The percentage of patients being divided into the balance-improvement group according to the PT criteria in both conditions of off- and on-medication in the 1-month follow-up and 12-month follow-up. c Heatmap showing explorative outcomes of balance-improvement predictors. x-axis showed 4 different conditions ((medication-on & off) × (criterion-BBS & PT)) based on which balance-improvement dichotomy was made. y-axis listed 17 potential predictors. The color of the blocks demonstrated the p value in the comparisons of each potential predictor between the balance-improvement group and the non-improvement group, with darker color indicating smaller p values. After Bonferroni correction, a p < 0.0026 (0.05/19) was considered significant. Blocks with red dotted border were the factors that remained statistically significant in the multivariate regression model. BBS, Berg balance scale; PT, pull test; FG, a history of festinating gait; AOO, age of onset; DD, disease duration; H-Y, Hoehn-Yahr stage; FOG-Q, freezing of gait questionnaire; PDQ-39, Parkinson’s disease questionnaire; MMSE, mini-mental state examination; LEDD, levodopa equivalent daily dose; U-II, UPDRS-II; U-III_res, UPDRS-III response to levodopa; berg_res, Berg balance scale response to levodopa; PW, pulse width; freq, frequency; TEED, total electrical energy delivered; Off_BBS, the dichotomy of off-medication BBS improvement; Off_PT, the dichotomy of off-medication PT improvement; On_BBS, the dichotomy of on-medication BBS improvement; On_PT, the dichotomy of on-medication PT improvement.