AVC Xu2024

PLOS ONE
RESEARCH ARTICLE
Effectiveness of acupuncture in the governor

vessel and Yangming meridian for the
treatment of acute ischemic stroke: A
systematic review and network meta-analysis
Yingqi Xu1, Xiuzhen Xie2, Pingping Su3, Jiashan Wang3, Xinxin Luo1, Jianli Niu ID4,
Zhuqing Jin ID5*
1 The Second School of Clinical Medicine, Zhejiang Chinese Medical University, Hangzhou, China, 2 The
First School of Clinical Medicine, Zhejiang Chinese Medical University, Hangzhou, China, 3 The Third School
a1111111111 of Clinical Medicine, Zhejiang Chinese Medical University, Hangzhou, China, 4 Office of Human Research,
a1111111111 Memorial Healthcare System, Hollywood, Florida, United States of America, 5 School of Basic Medical
a1111111111 Sciences, Zhejiang Chinese Medical University, Hangzhou, China
a1111111111
* [email protected]
a1111111111
Abstract
OPEN ACCESS
Background
Citation: Xu Y, Xie X, Su P, Wang J, Luo X, Niu J,
et al. (2024) Effectiveness of acupuncture in the Acupuncture of the governor vessel and Yangming meridian are widely used in the treat-
governor vessel and Yangming meridian for the ment of acute ischemic stroke (AIS). However, the optimal meridian for acupuncture in the
treatment of acute ischemic stroke: A systematic
treatment of AIS remains uncertain.
review and network meta-analysis. PLoS ONE
19(4): e0300242. https://doi.org/10.1371/journal.
pone.0300242 Purpose
Editor: Yung-Hsiang Chen, China Medical This network meta-analysis study aimed to compare the clinical effectiveness of acupunc-
University, TAIWAN ture at governor vessel and Yangming meridian in the treatment of AIS.
Received: November 21, 2023
Accepted: February 22, 2024

Methods
All relevant studies published in CNKI, WANFANG, VIP, Sinomed, Cochrane Library, Web
Published: April 16, 2024
of Science, Pub Med, and Embase before January 13, 2024 were systematically retrieved.
Copyright: © 2024 Xu et al. This is an open access
The two researchers independently screened the studies and extracted the data. Cochrane
article distributed under the terms of the Creative
Commons Attribution License, which permits ROB tool was used to evaluate the quality of the studies, and Stata 14.0 software was used
unrestricted use, distribution, and reproduction in to conduct a network meta-analysis of neurological deficit score, activities of daily living
any medium, provided the original author and (ADL), clinical effective rate and Fugl-meyer motor function evaluation (FMA).
source are credited.
Data Availability Statement: All relevant data are Results

within the paper and its Supporting Information
files. A total of 401 studies were obtained, and 17 studies met the inclusion criteria. The surface
under the cumulative ranking curve (SUCRA) values of the four outcome indexes were all
Funding: This research was funded by National
Natural Science Foundation of China, grant number ranked by “Governor vessel acupuncture + Conventional neurology treatment(GVAc+CT) >
81774010. Yangming meridian acupuncture + Conventional neurology treatment(YMAc+CT) > Con-
Competing interests: The authors have declared ventional neurology treatment (CT)”. Compared to YMAc+CT and CT, GVAc+CT had the
that no competing interests exist. best effect in reducing the degree of neurological deficit score (SMD = -0.72, 95%CI =
PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 1 / 19

PLOS ONE Effectiveness of the governor vessel and Yangming meridian acupuncture for the acute ischemic stroke treatment
Abbreviations: AIS, Acute ischemic stroke; [-1.22,-0.21] and SMD = -1.07,95%CI = [-1.45,-0.69], respectively) and promoting the recov-
SUCRA, surface under the cumulative ranking ery of ADL((SMD = 0.59,95%CI = [0.31,0.88] and SMD = 0.96,95%CI = [0.70,1.21], respec-
curve; GV1, Changqiang; GV28, Yinjiao; LI1,
Shangyang; LI20, Yingxiang; ST1, Chengqi; ST12,
tively). Compared to CT, GVAc+CT also had a better clinical effective rate in the treatment
Quepen; ST30, Qichong; ST44, Lidui; ADL, of AIS (RR = 1.14,95%CI = [1.04,1.25]).
Activities of daily living; FMA, Fugl-meyer motor
function evaluation; RCTs, randomized controlled Conclusions
trials; SMD, standardized mean difference; RR,
relative risk; h, hour; d, day; w, week. Governor vessel acupuncture combined with conventional neurology treatment has the best
effect in reducing the degree of neurological deficit score and promoting the recovery of
ADL in AIS patientscompared to YMAc+CT and CT. Governor Vessel acupuncture is the
most preferable acupoint scheme for clinical acupuncture treatment of AIS.
Introduction
Acute ischemic stroke (AIS) is a medical emergency in which the arteries supplying blood to a
specific region of the brain are narrowed or blocked, resulting in the interruption of local cere-
bral microcirculation in the cerebral area and the initiation of an ischemic cascade response
that eventually leads to brain cell damage [1, 2]. In recent years, a large number of studies have
found that acupuncture treatment for AIS can enhance nerve regeneration, improve cerebral
blood flow, reduce inflammatory damage and neuroexcitotoxicity and apoptosis caused by
cerebral ischemia, thus promoting the recovery of brain function after AIS and effectively
improving the prognosis of patients with sudden cerebrovascular accidents [3, 4]. At the same
time, the earlier application of acupuncture treatment is associated with greater benefits for
the recovery of neurological function in patients with AIS [5]. Compared with drugs, surgery
and other treatment methods, acupuncture is more convenient, cost-effective, and has fewer
side effects. The World Health Organization has recommended acupuncture as a viable alter-
native and complementary strategy for stroke treatment [6].
In animal experiments, the governor vessel electroacupuncture has been shown to promote
the cortical somatosensory evoked potential (SEP) recovery and reduce the volume of cerebral
infarction after cerebral ischemia, which is stronger than body acupuncture [7]. Zhou F [8]
demonstrated by animal experiments that acupuncture at various meridians improved cere-
bral blood flow and recovery of patients with varying degrees of damage by cerebral ischemia,
indicating the importance of the selection of different meridians and acupoints in the treat-
ment of AIS with acupuncture. WangSL found that both acupuncture on the governor vessel
acupuncture and yangming meridian could up-regulate the mRNA expression of MAP-2 and
NF-L in the brain tissue of rats, thereby improving brain plasticity and promoting neuronal
regeneration, with the governor vessel acupuncture group exhibiting a more significant effect
[9]. In a study by Pan Jiang, electroacupuncture at the governor vessel was found to enhance
the expression of NGF in the cerebral infarction area, leading to a reduction in the volume of
cerebral infarction and resisting nerve damage [10]. Furthermore, the governor vessel group
showed a higher NGF expression in the infarct area compared to the Yangming meridian
group. These animal experiments shed light on crucial mechanisms underlying the difference
in clinical efficacy between the two acupuncture approaches.
The main trunk of the governor vessel starts from Changqiang (GV1, the midpoint of the
line between the end of the coccyx and the anus) along the posterior midline of the body and
ends at the Yinjiao (GV28, the junction of the upper lip frenulum and gingiva), and is closely
related to the brain and spinal cord (Fig 1). The Yangming meridian includes Yangming Large

Fig 1. Main trunk of governor vessel acupoints. Note: In order to show more clearly the main pathways of the three
meridians running through the muscles on the body surface, the red solid lines in only show the main pathways
connected by each meridian point.
https://doi.org/10.1371/journal.pone.0300242.g001
Intestine Meridian of Hand and Yangming Stomach Meridian of Foot: the Yangming Large
Intestine Meridian of Hand (Fig 2) mainly follows the radial side of the upper limb from
Shangyang (LI1) to the neck and face, finally stops at Yingxiang (LI20); The straight trunk of
the Yangming Stomach Meridian of Foot (Fig 3) starts from the Chengqi (ST1) and runs along
the neck to Quepen (ST12), then it goes down to the Qichong (ST30) located at the inguinal
artery along the front of the thigh and the outside of the calf until it reaches the Lidui (ST44).
Unilateral upper and lower limb dysfunctions are the most common symptoms of AIS [11].
Therefore, Yangming meridian acupuncture is also commonly used in the treatment of AIS.
According to the current basic and clinical research progress of acupuncture in the treat-
ment of cerebral ischemia, we planned to conduct a systematic review by integrating the exist-
ing clinical studies and using the network Meta-analysis method to compare the therapeutic
effect of acupuncture at governor vessel and Yangming meridian in the treatment of AIS, so as
to provide evidence-based medical basis for the selection of acupoints of meridians in the
treatment of AIS.
Materials and methods

This network meta-analysis was not registered on the website.
Search strategy
This study was conducted according to the Preferred Reporting Items for Systematic Review
and Network Meta-analysis (PRISMA NMA) checklist (S1 File). Literature published in

Fig 2. Main trunk of Yangming Large Intestine Meridian of Hand acupoints. Note: In order to show more clearly
the main pathways of the three meridians running through the muscles on the body surface, the red solid lines in only
show the main pathways connected by each meridian point.
CNKI, WANFANG, VIP, sinomed, Cochrane Library, Web of Science, Pub Med, and Embase
was systematically searched using the following search terms (S2 File): (Acute ischemic stroke
OR Acute brain Infarction OR Acute Ischemia Apoplexy OR Acute cerebral infarction OR
Acute cerebral embolism OR Acute Ischemic Apoplexy OR Acute brain Ischemia OR AIS OR


Fig 3. Main trunk of Yangming Stomach Meridian of Foot acupoints. Note: In order to show more clearly the main pathways of the
three meridians running through the muscles on the body surface, the red solid lines in only show the main pathways connected by each
meridian point.
ACI) AND (governor vessel OR governor meridian OR du meridian OR yangming meridian)

AND (acupuncture OR electropuncture). The range of publication dates were from their
establishment of the database to January 13, 2024. All published clinical randomized controlled
trials (RCTs) of acupuncture at acupoints of governor vessel or Yangming meridian in the
treatment of AIS, including conference and academic literature were included without lan-
guage restrictions.
Eligibility criteria
Inclusion criteria.
1. Study types: clinical RCT studies, published language unlimited, including conference
literature;
2. Study population: (a) Patients diagnosed with AIS following the “Guidelines for diagnosis
and treatment of Acute Ischemic Stroke in China 2014” [12] and were confirmed by imag-
ing (including CT, MRI, etc.). (b) The onset of the AIS was less than 14 days;
3. Intervention measures: (a) The control group was treated with conventional neurology
treatment (CT), including thrombolysis, antiplatelet, anticoagulation, defibrination, volume
expansion, blood lipid regulation, neuroprotection, and rehabilitation training. (b) In addi-
tion to the treatments in the control group, the treatment group added acupuncture at the
main acupoints of governor vessel or Yangming Meridian as a means of treatment;
4. Outcome measures: Neurological deficit score, Activities of daily living (ADL), Clinical
effective rate, and Fugl-meyer motor function evaluation (FMA).
Exclusion criteria
1. Study types: animal experiments, reviews, case reports, etc.;
2. Study population: The onset time is not clear;
3. Intervention measures: (a) The control group was treated with sham acupuncture, tradi-
tional Chinese medicine, etc.; (b) In the experimental group, the main acupuncture points
were selected by other meridians or with massage, Chinese medicine as a means of
treatment;
4. Insufficient original data, data duplication, omission, etc. in clinical research literature.
Study screening and data extraction

Two independent reviewers (XYQ and XXZ) used the NoteExpress software to strictly search,
screen, and manage the literature. The following data were extracted and stored in a spread-
sheet database: (1) the first author of the study and year of publication; (2) patient information
(Disease duration, number, gender and mean age); (3) intervention information (Intervention
and control measures and treatment duration); (4) outcome indicators; (5)adverse events. The
results of literature screening and data extraction were cross-checked. Discrepancies in the
data extraction were resolved by the discussion with the third reviewer (JZQ).

Certainty of evidence assessment

Another two independent reviewers (SPP and WJS) evaluated the quality of the studies based
on the quality assessment scale in the Cochrane Systematic Assessment Manual [13] using
RevMan(cochrane website, 5.3.5) software in seven aspects: randomization method, allocation
concealment, blinding (patient, operator), data integrity, selective data, and other biases, and
finally "high risk", "low risk", and "unknown risk" were obtained. The quality of four outcome
indicators was evaluated according to the results of network Meta-analysis from five aspects:
research limitations, inconsistency, indirectness, precision and publication bias, based on the
GRADE framework (Grading of Recommendation, Assessment, Development, and Evalua-
tion) in the Cochrane handbook [13]. Finally, four grades of high, medium, low and very low
were obtained, and a summary table was made.
Statistical analysis
A random-effect model was used for the network meta-analysis.Each outcome measure was
assessed using validated scales: Neurological deficit score was evaluated by Composite Score
Scale (CSS) and National Institute of Health stroke scale (NIHSS). ADL was assessed by Modi-
fied Barthel Index (MBI) and Barthel Index (BI). FMA was assessed by Fugl-Meyer assessment
scale. The number of points reduction to be considered effective varies per study, and clinical
effective rate was obtained according to the criteria of each study. The time point of the out-
come indicators that were investigated was the earliest result following the completion of the
study treatment course, and the difference before and after the intervention of each outcome
index was selected for analysis. Neurological deficit score, ADL and FMA were continuous vari-
ables, due to different measurement methods, the standardized mean difference (SMD) was
used to eliminate the effect. Clinical effective rate was a dichotomous variable and the effect size
was expressed as a relative risk (RR). If a closed loop was formed, global inconsistency test was
carried out to determine whether the consistency model was used, P > 0.05 indicates that there
was no global inconsistency, and the consistency model can be used. At the same time, the loop
inconsistency test was carried out, node-splitting analysis and inconsistency factors were used
to test inconsistency for closed-loop indirect comparison. If the lower limit of the 95% confi-
dence interval contained or near 0, it was considered that there was no loop inconsistency, and
the direct and indirect comparative evidence were considered consistent. The STATA (Stata-
Corp LLC, 14.0, "mvmeta" and "network" packages) was used to draw evidence network and to
more intuitively observe the relationship between interventions. League tables were used to
compare different interventions. The SUCRA was used to evaluate the ranking probability value
of each intervention and to summarize the best choice of meridian acupuncture for AIS. In
addition, subgroup analyses were performed post hoc with the use of RevMan 5.3 software. The
95% confidence interval indicates the size of the statistical effect. If there was no significant het-
erogeneity among the studies (P�0.1, I2<50%), a fixed-effects model was used, otherwise using
a random-effects model (P<0.1, I2�50%). In order to evaluate the stability of the results, the
source of heterogeneity was explored by subgroup analysis. A correction-compare funnel dia-
gram was used to evaluate the publication bias of each outcome index of the studies.
Results
Study inclusion
The search yielded 401 articles and 274 were obtained after subtracting duplicates. After
reviewing the title, abstract and full text, 17 papers were finally included in meta-analysis. Fig 4
shows the retrieval process.

Fig 4. PRISMA flowchart.

Study characteristics
A total of 1269 patients with AIS were included in the 17 RCTs, with 15 [14–22, 24–27, 29, 30]
two-arm studies and 2 [23, 28] three-arm studies (Table 1). Most of the included treatment
groups retained the needle for 20–30 min after deqi, and the course of treatment was 10–40
days. Among them, 5 [17, 24, 27, 29, 30] studies used electroacupuncture at a frequency of 1-
20HZ, and the intensity was based on patient tolerance.
Analysis of methodological quality evaluation

The quality of the included trials was assessed according to the bias risk tool of the Cochrane
Collaboration. ① In terms of random methods, 13 [14–19, 21–23, 25, 27, 29, 30] studies
applied random number table method, which could be rated as low risk, and 4 [20, 24, 26, 28]
studies only mentioned randomness and were rated as unknown risks; ② In terms of

Table 1. Baseline characteristics of the included studies.

Study Course of acupuncture No. of participants Therapy Conventional neurological Treatment Outcome
disease manipulation; (male/female);age T1 T2 C treatment drugs time
stimulation frequency (years)
Wei et al. <4.5h non-EA 40(23/17) GVAc CT rt-PA 4w ①②③④
(2023) [14] NA (52.13±6.70) +CT
Li et al. 0.8~4.1h non-EA 100(58/42) YMAc CT N 4w ①②③④
(2022) [15] 180–200 times/min (63.95±6.19) +CT
Ma et al. <2d non-EA 83(47/36) GVAc CT N 4w ①②④
(2021) [16] NA (61.58±5.57) +CT
Wang et al. 1~14d EA 73(40/33) YMAc CT Aspirin, Clopidogrel, Atorvastatin 14d ①②④
(2021) [17] 20HZ (50.52±13.49) +CT calcium, Troxerutin and
Cerebroprotein Hydrolysate
Injection
Liu(2020) 6~33h non-EA 126(71/55) GVAc YMAc N 28d ①②③
[18] NA (70.6±7.3) +CT +CT
Hua et al. 24h~2w non-EA 100(50/50) GVAc CT Aspirin 14d ③
(2019) [19] NA (63±10.6) +CT
Li(2018) <24 h non-EA 37(21/16) GVAc CT N 4w ①②
[20] NA (57.95±8.5) +CT
ZhangL <48h non-EA 37(22/15) GVAc CT N 4w ①③
et al.(2017) NA (65.73±9.01) +CT
[21]
Cai et al. 4-5h non-EA 60(31/29) GVAc CT Ozagrel Sodium for Injection 2w ①③
(2016) [22] NA (53.60±10.91) +CT
Li et al. <2w non-EA 90(58/32) GVAc YMAc CT Edaravone, Aspirin 10d ①②③
(2016) [23] NA (62.54±10.27) +CT +CT
Ma et al. 5~45h EA 78(40/38) GVAc YMAc N 24d ①②③
(2016) [24] NA (61.48±4.41) +CT +CT
Yu(2014) 1.79d~6.39d non-EA 65(44/21) GVAc CT N 14d ②③④
[25] �200 times/min (61.82±4.89) +CT
Tan et al. <72h non-EA 94(51/43) GVAc CT N 10d ①③
(2013) [26] NA (61.47±12.09) +CT
Huang et al. 4h~6d EA 58(28/30) YMAc CT N 14d ①②
(2012) [27] 20HZ (67.64±10.48) +CT
Ge(2009) 7h~3d non-EA 90(53/37) GVAc YMAc CT Ozagrel Sodium and Sodium 21d ①②③
[28] NA (60.16±10.88) +CT +CT Chloride Injection, Cerebroprotein
Hydrolysate for Injection
ZhangX 4h~6d EA 58(30/28) YMAc CT Cinepazide Maleate Injection, 2w ①②④
et al.(2008) 20HZ (67.64±10.48) +CT Aspirin, Piracetam
[29]
ZhangSJ <5d EA 80(56/24) GVAc CT N 15d ①③
et al.(2007) 1HZ (61.85±11.10) +CT
[30]
Note: T1, treatment group 1; T2, treatment group 2; C, control treatment group; GVAc, governor vessel acupuncture; YMAc, Yangming meridian acupuncture; CT,
conventional neurology treatment; EA,electroacupuncture; non-EA, non-electroacupuncture; N, Only the use of conventional neurological treatment was mentioned,
including thrombolysis, antiplatelet, anticoagulation, fibrin reduction, volume expansion, lipid regulation, neuroprotection, and rehabilitation. No specific medications
were mentioned. ①: Neurological deficit score; ②: Activities of daily living (ADL); ③: Clinical effective rate;④: Fugl-meyer motor function evaluation(FMA).
https://doi.org/10.1371/journal.pone.0300242.t001
allocation hiding,3 [25, 27, 29] articles with random envelope allocation were rated as low risk,
and the rest that were not mentioned were rated as unknown risk; ③ In terms of blinded pro-
tocol (patient and operator), 1 [15] study was clearly non-blinded and rated as high risk, the

Fig 5. Methodologic quality assessment of the risk of bias. Note: Low risk, unclear risk and high risk are indicated by the symbols “+”,“?”, and “-”
respectively. (A) Risk of bias graph. (B)The bias risk and the quality evaluation results of each included study by the Cochrane risk of bias tool.
rest that did not mention the blinded protocol were rated as unknown risk; ④ In terms of end-
ing data integrity, 4 [16, 20, 21, 25] studies reported the number of data dislodgements and the
reasons for them and were rated as low risk, while the rest that did not have incomplete data
were rated as low risk; ⑤ In terms of selective reporting and other bias, all the studies that did
not mention the relevant information and were rated as unknown risks (Fig 5).
Evidence network
Fig 6 shows the network diagram of interventions under different outcome measures. In the
network map, three interventions(GVAc+CT, YMAc+CT and CT) were considered as nodes.
Pairwise comparisons between interventions served as lines of the network map. The thickness
of the black line is the number of comparative studies, the blue dots are the interventions and
the sizes of the blue dots are the number of samples participating in the intervention. The
thicker of the black line and the larger of the blue circle correspond to the comparison study

Fig 6. Evidence network. Note: GVAc = governor vessel acupuncture; YMAc = Yangming meridian acupuncture;
CT = conventional neurology treatment; (A)Neurological deficit score; (B): ADL; (C): Clinical effective rate; (D): FMA.
and sample size. 15 [14–18, 20–24, 26–30] studies reported neurological deficit score, with
closed loop formation; 12 [14–18, 20, 23–25, 27–29] studies reported ADL, with closed loop
formation; 12 [14–15, 18, 19, 21–26, 28, 30] studies reported clinical effective rate, with closed
loop formation and 6 [14–17, 25, 29] studies reported FMA, all of which were indirect evi-
dence without closed loop formation, and no inconsistency detection was required.
Inconsistency test
There was a closed loop that was formed in the evidence network of three outcome indicators
(ADL, Clinical effective rate and Neurological deficit score), namely “GVAc+CT—YMAc+CT
—CT”. P values of neurological deficit score (P = 0.7567), ADL (P = 0.4483) and clinical effec-
tive rate (P = 0.8067) were all > 0.05, suggesting that there is a good overall consistency
between the comparisons. The results of loop inconsistency test showed that the inconsistency
factor (IF) values of neurological deficit score closed-loop (IF = 0.274,95%CI = [0.00,1.30]),
clinical effective rate (IF = 0.051,95%CI = [0.00,0.34]) closed-loop and ADL (IF = 0.286,95%CI
= [0.00,0.86]) closed-loo were close to 0, and the lower limit of 95% CI were 0, suggesting that
there is a good consistency between direct comparison and indirect comparison among the
intervention measures and the result of network meta-analysis is reliable. FMA did not form a
closed loop and no inconsistency test was performed.
Network meta-analysis
Neurological deficit score. As shown in Table 2, GVAc+CT (SMD = -0.72, 95%CI =
[-1.22,-0.21], low quality evidence / SMD = -1.07,95%CI = [-1.45,-0.69], low quality evidence)

Table 2. Network meta-analysis of neurological deficit score (SMD and 95%CI).

GVAc+CT
LL
-0.72(-1.22,-0.21) ��1,2,4 YMAc+CT
LL L
-1.07(-1.45,-0.69) ��1,2 ��1,2,4
-0.35(-0.84,0.13) CT
Note: 1: limitations of the study due to blinding or lack of allocation concealment (-1), 2. Indirect comparison in
evidence (-1), 3. Global or ring inconsistency leads to (-1), 4. Imprecision due to too small sample size (< 300) or too
L LL
wide confidence (-1). ��/�� = very low quality evidence, �� = low quality evidence. (Same
comment for the rest of the Tables 3–5.). Comparisons should be read from left to right. Estimates of statistical
significances are shown in bold (P < 0.05), and 95% CI must not overlap zero. Negative values indicate better effect
of the intervention on the left compared to the right.
was more effective than YMAc + CT and CT in reducing neurological deficit score. The differ-
ence were statistically significant (p < 0.05). Other comparative differences were not statisti-
cally significant.
ADL. As shown in Table 3, GVAc+CT (SMD = 0.59,95%CI = [0.31,0.88], low quality evi-
dence) had a better effect in enhancing ADL of patients than YMAc+CT. GVAc+CT
(SMD = 0.96,95%CI = [0.70,1.21], low quality evidence) and YMAc+CT (SMD = 0.36,95%CI
= [0.08,0.64], low quality evidence) had better effect in enhancing ADL, compared with CT.
All differences were statistically significant (p <0.05).
Clinical effective rate. As shown in Table 4, GVAc+CT(RR = 1.14,95%CI = [1.04,1.25],
low quality evidence) had a better effect in improving clinical effective rate, compared with
CT. Other comparative differences were not statistically.
FMA. As shown in Table 5, there was no statistical significance between the comparisons
(p < 0.05), and the evidence related to this outcome indicator was very low quality evidence.
SUCRA
Neurological deficit score. According to the result of ranking graph based on SUCRA,
GVAc+CT was the most effective intervention to improve neurological deficit score in
patients, with a SUCRA value of 99.9%, followed by 46.2% for YMAc+CT, and 3.9% for CT
(Fig 7A).
ADL. According to the result of ranking graph based on SUCRA, GVAc+CT was the
most effective intervention to improve ADL in patients. As shown in Fig 7B, the highest
SUCRA value was 100% for GVAc+CT, followed by 49.7% for YMAc+CT, and 0.3% for CT
(Fig 7B).
Clinical effective rate. According to the result of ranking graph based on SUCRA, GVAc
+CT was the most effective intervention to improve clinical effective rate in patients, with a
SUCRA value of 98.1%, followed by 31.5% for YM+C, and 20.3% for C (Fig 7C).
Table 3. Network meta-analysis of ADL (SMD and 95%CI).

GVAc+CT
LL
0.59(0.31,0.88) ��1,2 YMAc+CT
LL LL
0.96(0.70,1.21) ��1,2 0.36(0.08,0.64) ��1,2 CT
Note: Comparisons should be read from left to right. Estimates of statistical significances are shown in bold
(P < 0.05), and 95% CI must not overlap zero. Positive values indicate better effect of the intervention on the left
compared to the right.(Tables 3,5)

Table 4. Network meta-analysis of clinical effective rate(RR and 95%CI).

GVAc+CT
L
1.12(0.98,1.27) ��1,2,4 YMAc+CT
LL L
1.14(1.04,1.25) ��1,2 1.02(0.88,1.18)��1,2,4 CT
Note: Comparisons should be read from left to right. Estimates of statistical significances are shown in bold
(P < 0.05), and 95% CI must not overlap 1. The RR > 1 indicate better effect of the intervention on the left compared
to the right.(Table 4)
FMA. According to the result of ranking graph based on SUCRA, GVAc+CT was the
most effective intervention to improve FMA in patients, with a SUCRA value of 76.6%, fol-
lowed by 45% for YMAc+CT, and 28.5% for CT (Fig 7D).
Heterogeneity analysis
When observing the basic characteristics of the included literature, there were differences in
the use of electroacupuncture and non-electroacupuncture in different literatures, and the
stimulation frequency used was also different, which might affect the stability of the results.
Studies have shown [31] that the neurological deficit score is an objective basis for evaluating
the overall stroke condition and prognosis of patients, which can be the most representative
reflecting the clinical effect of acupuncture treatment. Therefore, we chose the neurological
deficit score as the representative index.
We divided the studies into the non-EA group and the EA group according to whether elec-
tricity was applied after acupuncture, and there was statistical heterogeneity among the studies
(I2 = 94%, P<0.00001). The random-effects model was then used for meta-analysis. The com-
bined effect was statistically significant (MD = -3.28, 95% = [-5.98, -0.58], P = 0.02). The results
confirmed that compared with Yangming meridian acupuncture, governor vessel acupuncture
could better reduce the neurological deficit score. Further subgroup analyses showed that: (1)
Two studies described the use of non-EA, but the pooled effect size was not statistically signifi-
cant (MD = -1.66, 95% = [-4.15, 0.83], P = 0.19), suggesting that there was no statistically sig-
nificant difference in the therapeutic effect between governor vessel Acupuncture and
Yangming meridian acupuncture. (2) Two studies reported the use of EA, and the results
showed that acupuncture at governor vessel was significantly better than acupuncture at
Yangming meridian in improving the degree of neurological deficit (MD = -5.17, 95% =
[-6.59, -3.75], P<0.00001). (Fig 8). Based on the above results, we speculate that the use of EA
may be a potential source of heterogeneity.
In addition, two studies described the stimulation frequency, and there was statistical het-
erogeneity between studies (I2 = 69%, P = 0.07). The results suggested that the governor vessel
therapy could reduce the neurological deficit score more effectively than the conventional
treatment, and the difference was statistically significant (MD = -2.92, 95% = [-5.22, -0.62],
P = 0.01). (Fig 9). Therefore, we speculate that different stimulation frequencies may be a
potential factor causing heterogeneity.
Table 5. Network meta-analysis of FMA(SMD and 95%CI).

GVAc+CT
0.31 (-0.97,1.59) ��1,2,3,4 YMAc+CT
0.41 (-0.33,1.51) ��1,2,3,4 0.10 (-0.95,1.15) ��1,2,3,4 CT

Fig 7. Comparative effectiveness between intervention categories: surface under the cumulative ranking curves
(SUCRA) for (A): Neurological deficit score; (B): ADL; (C): Clinical effective rate; (D): FMA. Note: The x-axis is the
possible level of each treatment (from the first to the worst according to different outcome indicators). The cumulative
probability that each treatment is the best, second, and third best intervention is represented by the y axis.
Consequently, each treatment method’s ranking is represented by the area under the curve. The likelihood that a
treatment approach falls into the highest level or one of the highest levels increases as the value increases.
Correction-compare funnel diagram

We plotted comparative-corrected funnel diagram of the neurological deficit score, ADL,
clinical effective rate and FMA using Stata 14.0 (Fig 10). The result showed that the studies
were roughly in the central, and had good symmetry, suggesting that publication bias is
unlikely.
Fig 8. Forest plot of subgroup analysis with acupuncture manipulation as the criterion.

Fig 9. Forest plots depicting stimulus frequencies.

Adverse events
Adverse events were reported in 4 of the included studies. Because the occurrence of adverse
reactions was inconsistent among the studies, only descriptive analysis was performed. 3 [16,
20, 26] of the included studies mentioned adverse events, including subcutaneous hematoma
and stagnant needles, during treatment, all of which occasionally occurred after acupuncture
treatment, and they were improved rapidly after active management. One study [19] referred
to the follow-up of adverse events after treatment, and the results showed that the incidence of
adverse events was much lower in the governor vessel acupuncture group than in the
Fig 10. Correction-compare funnel diagram. Note: (A) Neurological deficit score, (B) ADL, (C) Clinical effective rate, (D) FMA.

conventional treatment group, suggesting that the governor vessel acupuncture group has a
better effect in improving the prognosis of patients with AIS. No other adverse events were
observed in the included studies.
Discussion
AIS is the second leading cause of death worldwide and the primary cause of permanent dis-
ability [32]. Extensive studies have proven that acupuncture intervention is an effective
method to promoting nerve repair and improving functional dysfunction in patients with AIS
[33]. There have not been any network meta-analysis studies comparing the effects of AIS acu-
puncture treatment based on meridians. In this network meta-analysis consisting of 17 RCTs
we concluded that the efficacy of governor vessel acupuncture was the most significant in the
treatment of AIS.
Our systematic review and network meta-analysis focused on the neurological deficit score
ADL, clinical effective rate and FMA, The results showed that: (1) Regarding neurological defi-
cit score, GVAc+CT exhibited superiority over YMAc+CT and CT (P < 0.05); (2) In terms of
ADL, GVAc+CT dedmonstrated superiority to YMAc+CT and CT (P < 0.05). (3) In terms of
the clinical efficacy, GVAc+CT showed superiority to CT (P < 0.05). The above results proved
that governor vessel acupuncture has advantages in the treatment of AIS. The results of
SUCRA ranking chart of each outcome indicator were as follows: "GVAc+CT" > "YMAc+CT"
> "CT". In terms of improving ADL, the probability of GVAc+CT being the best therapy was
100%, showing an absolute advantage among the three therapies, suggesting that GVAc+CT
was the most likely to be the most effective intervention for AIS. However, this result is still
affected by the quality of evidence and the risk of bias.
Our evidence certainty assessment found that the included studies had some risk of bias,
while the overall quality of evidence was generally low and very low. Due to the use of network
meta-analysis, the evidence obtained from this study was indirect. Among them, the quality of
evidence supports that GVAc+CT is superior to YMAc+CT and CT in reducing the degree of
neurological deficit and improving ADL, even though there were some defects in the design of
clinical RCTs included in the study. In terms of the clinical effective rate, the quality of evi-
dence is very low in supporting that GVAc+CT is superior to YMAc+CT in improving the
clinical effective rate. This may be related to the imprecision of research data and the small
sample size. The quality of evidence between different interventions in enhancing FMA was
very low, mainly due to the inconclusive evidence for this outcome measure, and there was lit-
tle or no direct comparison evidence, suggesting that future relevant clinical RCT studies can
further explore the efficacy of each therapy in enhancing FMA in patients with AIS.
We also conducted the heterogeneity analysis around the differences between the use of EA
and non-EA, as well as the differences in stimulation frequency of acupuncture: (1) Four stud-
ies [18, 23, 24, 28] were divided into two groups according to whether EA was used or not. The
results showed that the performance of the EA group was relatively homogeneous (I2 = 0%,
P<0.00001), although the heterogeneity was decreased in the non-EA group, the difference
between the two groups was not statistically significant (I2 = 90%, P = 0.19). Therefore, we
speculated that the use of EA may be a potential source of heterogeneity. (2) The Meta-analysis
showed that there was some heterogeneity between the two studies describing the stimulation
frequency (I2 = 69%, P = 0.01), which may affect the heterogeneity of the results. We can spec-
ulate that the use of EA and stimulation frequency may affect the stability of the results, but
whether it is the main factor affecting the heterogeneity needs to be discussed in more studies.
There are some limitations of this study: (1) The assessment of FMA outcome was limited
by the absence of a direct comparison between GVAc + CT and YMAc + CT. Additionally, the

small sample size available for analysis of FMA outcome might contribute to the uncertainty
regarding the efficacy superiority of Governor Vessel acupuncture therapy to Yangming
meridian acupuncture in improving FMA (P > 0.05). Which suggests that in the future, when
treating AIS patients with severe limb motor deficits, we can add Yangming meridian acu-
puncture treatment, or carry out further research in this aspect. (2) The quality of evidence in
this study was downgraded mainly due to limitations in study design, including issues related
to blinding and allocation concealment as well as the imprecision of the data. (3) All eligible
studies included in the analysis were from China and the results were based on the Chinese
population, but may not generalize well to other races.
Conclusion
This study confirmed that the efficacy of governor vessel acupuncture is better than Yangming
meridian acupuncture for the treatment of AIS. However, due to the lack of definitive evidence
on FMA outcome indicators and the low quality of evidence on improving FMA in patients
with AIS, we were unable to conclude the true effect of governor vessel acupuncture in
improving FMA in patients with AIS. Further investigations are needed to demonstrate the
efficacy of the governor vessel acupuncture in the treatment of AIS by carrying out different
multi-ethnic, large-sample, clinical randomized controlled studies in the future.
Supporting information
S1 File. PRISMA NMA checklist.
(PDF)
S2 File. Search formulas.
(PDF)
Acknowledgments
We would like to express our appreciation to all authors of eligible studies which were included
in the current meta-analysis.
Author Contributions
Conceptualization: Zhuqing Jin.
Data curation: Yingqi Xu, Xiuzhen Xie.
Formal analysis: Yingqi Xu, Pingping Su, Jiashan Wang.
Funding acquisition: Zhuqing Jin.
Investigation: Xinxin Luo.
Software: Yingqi Xu, Pingping Su.
Validation: Xiuzhen Xie.
Writing – original draft: Yingqi Xu.
Writing – review & editing: Jianli Niu, Zhuqing Jin.
References
1. Walter K. What Is Acute Ischemic Stroke? JAMA. 2022 Mar; 327(9):885. https://doi.org/10.1001/jama.
2022.1420 PMID: 35230392

2. Dirnagl U. Pathobiology of injury after stroke: the neurovascular unit and beyond. Ann N Y Acad Sci.
2012 Sep; 1268:21–5. https://doi.org/10.1111/j.1749-6632.2012.06691.x PMID: 22994217
3. Mu JD, Ma LX, Zhang Z, Qian X, Zhang QY, Ma LH, et al. The factors affecting neurogenesis after
stroke and the role of acupuncture. Front Neurol. 2023 Jan; 14:1082625. https://doi.org/10.3389/fneur.
2023.1082625 PMID: 36741282
4. Chavez LM, Huang SS, MacDonald I, Lin JG, Lee YC, Chen YH. Mechanisms of Acupuncture Therapy
in Ischemic Stroke Rehabilitation: A Literature Review of Basic Studies. Int J Mol Sci. 2017 Oct; 18
(11):2270. https://doi.org/10.3390/ijms18112270 PMID: 29143805
5. Kou YF, Cai EL, Zhang W, Wu YM. Study on the protective effect of electroacupuncture intervention
time on brain after cerebral ischemia reperfusion in rats. Journal of Guizhou University of Traditional
Chinese Medicine. 2014; 36(04):39–41. https://doi.org/10.3969/j.issn.1002-1108.2014.04.016
6. World Health Organization. Acupuncture: review and analysis of reports on controlled clinical trials. Par-
kinsonism & Related Disorders. 2003. https://doi.org/10.1016/S1353-8020(11)70706-4
7. Jin ZQ, Cheng JS. Effects of acupuncture at different acupoints on cortical somatosensory evoked
potential and infarct volume in rats with transient focal cerebral ischemia. Shanghai Journal of Acupunc-
ture and Moxibustion. 1998;39–41. https://doi.org/10.13460 j.issn.1005
8. Zhou F, Guo JC, Cheng J, Wu G, Sun J, Xia Y. Electroacupuncture and Brain Protection against Cere-
bral Ischemia: Specific Effects of Acupoints. Evid Based Complement Alternat Med. 2013;
2013:804397. https://doi.org/10.1155/2013/804397 PMID: 23737846
9. Wang SL. Effect of acupuncture on the mRNA expression of MAP-2 and NF-L in the ischemic penum-
bra of rats undergoing cerebral ischemia and reperfusion. M.Sc. Thesis, Nanjing University of Chinese
Medicine. 2012. Available from: http://dx.doi.org/10.7666/d.y2124025
10. Pan J, Chen WS, Chen C, Zhang W. Effects of Electric-Acupuncture at Du-Meridian on MCAO Rats
Infarct Volume and NGF in Brain Tissue. Chinese Archives of Traditional Chinese Medicine. 2017;
(03),541–543. https://doi.org/10.13193/j.issn.1673-7717.2017.03.007
11. Chen GM, Huang CY, Liu YY, Zhang ZP, Qi XQ, Shi PY, et al. Efficacy and safety of grain moxibustion
in hemiplegia: A systematic review and meta-analysis protocol. Medicine (Baltimore). 2019 Apr; 98(17):
e15215. https://doi.org/10.1097/MD.0000000000015215 PMID: 31027068
12. Neurology Branch of Chinese Medical Association, Cerebrovascular Disease Group of Neurology
Branch of Chinese Medical Association. Guidelines for the diagnosis and treatment of acute ischemic
stroke in China 2014.Chinese Journal of Neurology. 2015, 48(4): 246–257. https://doi.org/10.3760/
cma.j.issn.1006-7876.2015.04.002
13. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al. The Cochrane Collabora-
tion’s tool for assessing risk of bias in randomised trials. BMJ. 2011, 34:d5928. https://doi.org/10.1136/
bmj.d5928 PMID: 22008217
14. Wei CY, Ma YB, Li XH. Application effect of Tongdu Tiaoshen acupuncture in patients with acute cere-
bral infarction hemiplegia. Lingnan Journal of Emergency Medicine,2023, 28(01):53–55.
15. Li LY, Qiao J, Ding HY. Effect of acupuncture method in the treatment of patients with acute cerebral
infarction. Medical Journal of Chinese People’s Health. 2022; 34(11):111–113. https://doi.org/10.3969/
j.issn.1672-0369.2022.11.035
16. Ma C, Li J, Wang XN, Chang YX. Clinical study on effect of Shengyang Tongdu acupuncture on electro-
encephalography topography, neurovascular unit function and heart ratevariability in patients with cere-
bral infarction. Shanghai Journal of Traditional Chinese Medicine. 2021; 55(09):50–54. https://doi.org/
10.16305/j.1007-1334.2021.1912090
17. Wang XW, Tan F, Li DD, Zhan J, Fang MF, Li M, et al. Influence of Electroacupuncture on Motor Dys-
function and Corticospinal TractInjury After Acute lschemic Stroke. Shanghai Journal of Acupuncture
and Moxibustion. 2021; 40(04):426–430. https://doi.org/10.13460/j.issn.1005-0957.2021.04.0426
18. Liu ZY. Analysis of the effect of acute cerebral infarction with hyperhomocysteinemia by transfer-
through-governor injection. Smart Healthcare. 2020; 6:153–4,7. https://doi.org/10.19335/j.cnki.2096-
1219.2020.05.066
19. Hua HB, Qian FW, Zheng WZ. Effect of acupuncture and moxibustion combined with tongdu therapy on
ischemic stroke and its influence on prognosis. Shanghai Journal of Acupuncture and Moxibustion.
2019; 38:1087–92. https://doi.org/10.13460/j.issn.1005-0957.2019.10.1087
20. Li YW. Study on the effect and curative effect of Tongdu Shenregulating acupuncture on ischemic pen-
umbra in patients with acute ischemic stroke. M.Sc. Thesis, Anhui University of Traditional Chinese
Medicine. 2018.
21. Zhang L, Zhang GQ, Zhu LL, Han W. Clinical observation on the treatment of wind-phlegm-Stasis acute
cerebral infarction by Tongdu-Tiaoshen acupuncture. Journal of Anhui Traditional Chinese Medical Col-
lege. 2017; 36:59–62. https://doi.org/10.3969/j.issn.2095-7246.2017.04.018

22. Cai J, Tang CZ, Zhu GQ. Clinical study on acupuncture Dumai point in the treatment of acute cerebral
infarction. Asia-Pacific Traditional Medicine. 2016; 12:116–7. https://doi.org/10.11954/ytctyy.
201603054
23. Li MX, Sha YP. The effect of ’ Xingnao Kaiqiao ’ acupuncture on acute ischemic stroke and its effect on
MMP-9. Journal of Clinical Acupuncture and Moxibustion. 2016; 32(12):17–19.
24. Ma XM, Li JJ, Yan B, Liu YF,Zou P, Huang RC. Clinical study on the treatment of acute cerebral infarc-
tion complicated with hyperhomocysteine by transfer-general acupuncture. Journal of Shaanxi College
of Traditional Chinese Medicine. 2016; 39:50–3,102. https://doi.org/10.13424/j.cnki.jsctcm.2016.04.
018
25. Yu FF, Clinical study of combining self-rehabilitation training for the treatment of acute stroke hemiple-
gia. M.Sc. Thesis, Guangzhou University of Chinese Medicine. 2014; 49.
26. Tan JC, Ding WT, Ge L. Effect of early acupuncture on 46 cases of acute cerebral infarction compli-
cated with type 2 diabetes mellitus. Hebei Traditional Chinese Medicine,2013, 35(5):722–724.
27. Huang T, Li GX.The effect of electroacupuncture on the expression of CD62p and D-Dimer and the
scores of ADL and NIHSS in patients with acute cerebral infarction. Lishizhen Medicine and Materia
Medica Research. 2012; 23(10):2665–2667. https://doi.org/10.3969/j.issn.1008-0805.2012.10.128
28. Ge CH. The effect of acupuncture on CRP and D-dimer in patients with acute cerebral infarction and its
curative effect. M.Sc. Thesis, Guizhou: Guiyang College of Traditional Chinese Medicine. 2009. Avail-
able from: http://dx.doi.org/10.7666/d.y1561180
29. Zhang X. Effect of electroacupuncture of Yangming meridian acupoint on the expression of PAC-1 and
CD62p in patients with acute cerebral infarction. M.Sc. Thesis, Guangzhou University of Chinese Medi-
cine. 2008.
30. Zhang SJ, Chen DF, Lai Z. Clinical observation of acupuncture Dumai point in the treatment of acute
cerebral infarction. Journal of Changchun University of Chinese Medicine. 2007; 23:57–8. https://doi.
org/10.3969/j.issn.1007-4813.2007.05.037
31. Dong Q, Wu DC, Lv CZ. Compare Analysis of the Neurological Functional Deficit Scales in 193 Patients
with Acute Stroke. Chinese Journal of Clinical Neurosciences. 2000. https://doi.org/10.3969/j.issn.
1008-0678.2000.03.011
32. Farina M, Vieira LE, Buttari B, Profumo E, Saso L. The Nrf2 Pathway in Ischemic Stroke: A Review.
Molecules (Basel, Switzerland). 2021; 26(16),5001. https://doi.org/10.3390/molecules26165001 PMID:
34443584
33. Chang QY, Lin YW, Hsieh CL. Acupuncture and neuroregeneration in ischemic stroke. Neural regener-
ation research. 2018; 13,573–583. https://doi.org/10.4103/1673-5374.230272 PMID: 29722298

AVC Xu2024

Uploaded by

Copyright:

Available Formats

AVC Xu2024

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AVC Xu2024

Uploaded by

Copyright:

Available Formats

PLOS ONE

Effectiveness of acupuncture in the governor

Accepted: February 22, 2024

Data Availability Statement: All relevant data are Results

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 1 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 2 / 19

Materials and methods

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 3 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 4 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 5 / 19

ACI) AND (governor vessel OR governor meridian OR du meridian OR yangming meridian)

Study screening and data extraction

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 6 / 19

Certainty of evidence assessment

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 7 / 19

Fig 4. PRISMA flowchart.

Analysis of methodological quality evaluation

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 8 / 19

Table 1. Baseline characteristics of the included studies.

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 9 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 10 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 11 / 19

Table 2. Network meta-analysis of neurological deficit score (SMD and 95%CI).

Table 3. Network meta-analysis of ADL (SMD and 95%CI).

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 12 / 19

Table 4. Network meta-analysis of clinical effective rate(RR and 95%CI).

Table 5. Network meta-analysis of FMA(SMD and 95%CI).

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 13 / 19

Correction-compare funnel diagram

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 14 / 19

Fig 9. Forest plots depicting stimulus frequencies.

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 15 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 16 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 17 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 18 / 19

PLOS ONE | https://doi.org/10.1371/journal.pone.0300242 April 16, 2024 19 / 19

You might also like