Swallowing disorders are one of the common complications after stroke, with a prevalence of approximately 37 to 78% in stroke patients (1). Patients with swallowing disorders may suffer from slow food intake, choking on water, and dysphagia, which seriously affect their quality of life (1). Due to a decline in swallowing function, patients are more prone to developing aspiration pneumonia (OR = 4.08, 95% CI = 2.13–7.79) (2) and malnutrition (OR = 0.91, 95% CI = 0.83–0.99) (3) or even death (OR = 4.07, 95% CI = 2.17–7.63) (2). In addition, PSD has also brought a huge economic burden to patients, families, and society. In the United States, the mean hospital stay of patients with dysphagia is 3.8 days longer than that of patients without dysphagia, which will incur an extra mean hospitalization expense of USD 6,243 (4). Therefore, it is necessary to find an effective way to treat dysphagia.

Currently, there are many clinical methods to treat dysphagia, such as swallowing training (5), neuromuscular electrical stimulation (6), non-invasive transcranial direct current stimulation (7), and balloon dilation therapy (8). Although these treatments have been employed to treat dysphagia (9), the optimal treatment protocols for dysphagia remain to be explored (10). It is generally believed that dysphagia in stroke patients is induced by a focal lesion that interrupts the connection between the nucleus tractus solitarius (NTS) and nucleus ambiguus (NA) on the same side of the lesion (11). Studies have revealed that electroacupuncture (EA) is a potential intervention strategy for dysphagia, which can improve the swallowing function through the neural circuit of the “primary motor cortex (M1) - parabrachial nucleus (PBN) - NTS” (12). However, the application of EA in the treatment of poststroke dysphagia (PSD) is still controversial. Some studies have shown that acupuncture, including EA, helps patients recover from post-stroke sequelae, like dysphagia (13, 14). However, other studies have shown that acupuncture has no effect on the recovery of patients’ functions after stroke (15, 16). A previous meta-analysis (17) only included articles comparing the efficacy of EA combined with swallowing training and single swallowing training, and the authors pointed out that the quality of the included literature was relatively poor. Therefore, these findings are inconclusive. More clinical studies on EA for the treatment of PSD have emerged afterward and shown positive therapeutic outcomes.

Therefore, this study included randomized controlled trials (RCTs) on both EA alone and EA combined with conventional interventions. Relevant articles were screened for inclusion or exclusion using the modified Jadad scale. The aim of this study was to conduct a meta-analysis of RCTs on EA and EA combined with other interventions for the treatment of poststroke dysphagia (PSD), in order to provide higher quality evidence-based basis for the clinical rehabilitation of patients with PSD.

This meta-analysis included 12 RCTs involving 1,358 patients with PSD. The results showed that EA could benefit patients with PSD in terms of the clinical response rate and VFSS score after the exclusion of low-quality literature and the inclusion of the latest studies.

Two meta-analyses (17, 32) previously published on the similar topic were also searched. In the study conducted by Huang JK et al. (17), although beneficial results were concluded, the authors pointed out that the quality of the included articles was relatively low. Accordingly, such results were not conclusive enough. Another meta-analysis published in a Chinese journal was a similar case (32), in which all the 18 articles included had a high risk of bias. Furthermore, some results of the two meta-analyses were significantly heterogeneous, but the reasons for heterogeneity were not discussed in either of them. These meta-analyses only included the articles on the comparison between EA combined with swallowing training and swallowing training alone, and excluded the studies on EA combined with other treatments. There is no study assessing whether the addition of EA will still be conducive to the rehabilitation of dysphagia when multiple interventions are combined. In our study, therefore, the studies on EA alone and EA combined with other interventions were included, while low-quality RCTs were excluded using the modified Jadad scale, which made our results more reliable and convincing. The assessment of evidence level also confirmed our judgment. The evidence level was high for six outcome indicators in this study, proving that the overall quality of this paper is high.

Based on our study results, there was no significant difference between both groups in the SSA score, PAS score, SWAL-QOL score, and incidence of adverse events. In terms of the clinical response rate and VFSS score, the improvement of the EA group was significantly superior to that of the control group. Our findings in the clinical response rate were the same as those concluded from previous meta-analyses (17, 32). However, the clinical response rate was not assessed using a unified measurement tool in previous meta-analyses. In our study, a subgroup analysis was conducted according to different measurement tools. Among the included studies, the clinical response rate was weighed according to changes in the VFSS score in three studies, while it was weighed based on the WST score in five studies. The results of the subgroup analysis still revealed that EA benefited patients and had relatively low heterogeneity. Therefore, our study was more credible from the perspective of the clinical response rate. Moreover, VFSS is the “gold standard” for the evaluation of dysphagia (33, 34). Based on our study findings, both the clinical response rate obtained from VFSS and the VFSS score showed that EA improved PSD. The efficacy of EA based on VFSS has also been demonstrated in the study conducted by Zhang Shengyu et al. (35), which is consistent with our observations.

Notably, EA did not improve the SSA score, PAS score, or SWAL-QOL score, which was inconsistent with the results of the clinical response rate and VFSS score. This may be related to different acupoint selections. It was found that in studies where the VFSS score was involved, the acupoint of Lianquan (CV23) was selected in all trials. In studies where the SSA score was involved, CV23 was selected in only one trial (21). In studies where PAS and SWAL-QOL scores were involved, it was not selected in any trial. The study conducted by Yuan S et al. (36) showed that administering EA on CV23 could regulate the swallowing function through excitatory neurons in the paraventricular hypothalamus. By comparing the effects of administering EA on CV23 and Neiguan (PC6), Ye QP et al. found that administering EA on CV23 could regulate the swallowing function by activating swallowing-related interneurons in the ventrolateral medulla oblongata (37). In addition, in a recent study, Ye QP et al. also demonstrated through optogenetics and chemogenetics that the NTS is involved in the regulation of EA on CV23 for PSD, revealing the role of the M1-NTS pathway in EA on CV23 (38). The latest research shows that in addition to the central nervous system, EA on CV23 can also have an impact on the peripheral nervous system and the local swallowing muscles. The study conducted by Yuan S et al. (39) suggests that TRPV1 at the CV23 can regulate local blood perfusion, thereby promoting the recovery of swallowing function. In combination with our study results, we believe that this may suggest a certain value of administering EA on CV23 in the improvement of PSD.

Additionally, we also recognize that the studies on EA in modern medicine have been increasingly getting advanced with the progress of science and technology. A growing number of articles are emerging to study and demonstrate the efficacy of EA with the aid of advanced equipment and technology. Wu Wenbao et al. (40) combined cerebral diffusion tensor imaging (DTI) techniques and VFSS to explore the clinical significance of acupuncture intervention in PSD. They administered EA on the acupoints of Sishencong, Baihui, Temple, Fengchi, and Tongue Triple Acupuncture and found that the fractional anisotropy (FA) values in the infarcted areas of the cerebral hemispheres were improved significantly, and that FA values were positively correlated with the integrity of white matter fiber tracts (41), suggesting that EA on relevant acupoints may have restored cortical function. Zhu Runjia et al. (42) observed the clinical efficacy of EA in the treatment of ischemic stroke through transcranial Doppler (TCD), finding that EA stimulation could improve the intracranial vascular function of patients. The main action principle is to reduce vascular resistance by changing the elasticity and compliance of the corresponding blood vessel, so as to increase cerebral blood supply and promote rehabilitation after stroke, thus indirectly improving the swallowing function. Lan Chunwei et al. (43) observed the effect of EA on PSD through magnetic resonance spectroscopy (MRS) and surface electromyography (sEMG), finding that EA could decrease the contents of rNAA and Lac/Cr in the brain to reduce the mean amplitude of sEMG in the pharynx and shorten the swallowing time course. Their findings indicate that EA improves the swallowing function of patients with PSD by influencing brain metabolism and ameliorating the contraction of peripheral swallowing muscles. Most of the outcome indicators included in our study were scale scores. In fact, we also wanted to analyze the efficacy of EA in combination with microcosmic study data. However, it was a pity that among our included studies, EEG was performed in only one study to detect the degree of cerebral cortex inhibition (24); a transcranial magnetic stimulator was used in one study to detect the effect of EA on motor-evoked potential (MEP) (27), and sEMG was performed in one study to quantitatively assess swallowing muscle groups (29). In this context, we cannot provide further evidence to explore the efficacy of EA. More high-quality studies are desired to solve this problem in the future.

We also acknowledge that this systematic review has some limitations. Firstly, based on the Cochrane Collaboration’s tool for assessing the risk of bias, nine of our included studies (22, 24–31) had unclear risk of bias. For example, Shao XZ et al. (28) only expressed random allocation but did not describe the specific method for random allocation, which reduced the quality of this meta-analysis. Secondly, some outcome indicators were only involved in two to three studies. In this regard, the insufficient sample size was also one of our regrets. Finally, it was impossible to continue the efficacy analysis of EA from the perspective of new technology since there were few studies reporting microcosmic study data.

In spite of some limitations, the findings of this systematic review are still of great clinical and basic scientific significance. In combination with previous meta-analyses (17, 32), we are surer that EA is a safe and reliable treatment approach for patients with PSD. Although further studies are still required to demonstrate the efficacy of EA (acupoint selection in EA or with the aid of new technology), this paper still provides a new evidence-based basis for the clinical rehabilitation of patients with PSD and offers the orientation and theoretical support for future studies. Therefore, our study is indeed of great value.

In conclusion, EA is effective and safe in treating PSD. EA combined with conventional treatment or other interventions can significantly improve the clinical response rate and VFSS score in patients with PSD, without increasing the incidence of adverse reactions. There is no statistical difference between the two groups in terms of the improvement of SSA score, PAS score, and SWAL-QOL score. More high-quality RCT studies are still needed in the future to further explore the efficacy of EA in the treatment of PSD.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

XL: Formal analysis, Investigation, Writing – original draft, Writing – review & editing. HL: Formal analysis, Investigation, Writing – original draft. WY: Methodology, Writing – original draft. HG: Formal analysis, Investigation, Writing – review & editing. KG: Methodology, Writing – review & editing. LL: Writing – review & editing. XF: Writing – review & editing. ZH: Conceptualization, Writing – review & editing.