*2.5. PMA*

In the PMA, we conducted a conventional direct comparison of the two study arms. Data synthesis was performed using the Review Manager (RevMan) ([Computer program]. Version 5.4, The Cochrane Collaboration, 2020). The random effect model was adopted because it was judged that there was heterogeneity due to differences in the study design, such as baseline characteristics, number of interventions, and methods among the included studies. The mean difference (MD) for the continuous variables and 95% confidence interval (CI) were used to assess the effect size of the intervention on UPDRS-III and II. Heterogeneity was determined by both the chi-square (χ2) test and Higgins' I 2 statistic. The heterogeneity interpretation based on the I 2 statistic is considered not to be important (0 to 40%), moderate heterogeneity (30% to 60%), substantial heterogeneity (50% to 90%), and considerable heterogeneity (75% to 100%) [27]. A *p*-value of ≤0.1 was considered to indicate significant heterogeneity [28].
