*3.1. Characteristics of the Included Studies and Network Geometry*

A total of 2505 articles were screened from eight databases. After careful review of the title and abstract, 17 articles were finally included (Figure 1). In 17 RCTs, 1071 participants were included. A list of the 28 studies excluded after reviewing the full text is provided in Supplementary Materials Digital Content 2. Eight articles were written in English [36–43], one article was written in Japanese [44], two in Korean [15,45], and six articles were written in Chinese [14,46–50]. Detailed characteristics of the included studies including publication year, first author, country, sample size (initial and final), age, sex, disease severity, disease duration, CM dosage (mg/day), treatment and control group intervention, and treatment period are described in Table 1.

Five types of arms were identified: (1) manual acupuncture + conventional drug (MANU), (2) electroacupuncture + conventional drug (ELEC), (3) BVA + conventional drug (BEEV), (4) sham acupuncture + conventional drug (SHAM), and (5) conventional drug alone (CONV). Fifteen RCTs had two armed designs, and only two RCTs had three armed designs (one RCT included MANU vs. SHAM vs. CONV [38], and one RCT included BEEV vs. MANU vs. CONV [41]). Seven trials included the ELEC arm, nine trials included the MANU arm, and two included the BEEV arm. Therefore, a total of 18 comparisons (36 treatment arms) were included in the 17 RCTs. Six RCTs compared ELEC vs. CONV [14,40,42,46,48,49]; five RCTs compared MANU vs. CONV [38,41,44,47,50]; one RCT compared ELEV vs. SHAM [36]; five RCTs compared MANU vs. SHAM [15,37,38,43,45]; and one RCT compared BEEV vs.

1

SHAM [39]. Detailed descriptions of each intervention, including the acupuncture point, needle stimulation, retention time, treatment duration, and frequency, are described in Table 2.

**Figure 1.** PRISMA flow diagram.
