Efficacy and safety of acupuncture treatment for stiff neck: A systematic review and meta-analysis

Abstract

Background:

Stiff neck is a common acute musculoskeletal condition that significantly affects the quality of life of patients. Acupuncture is recommended as an effective method for alleviating pain and restoring neck mobility in patients with stiff neck, but there is currently a lack of scientific evidence supporting its efficacy and safety. The purpose of this study was to investigate the efficacy and safety of acupuncture in the treatment of stiff neck.

Method:

This study searched 8 Chinese and English electronic medical databases, including China Biology Medicine disc, VIP database, Wanfang Data, China National Knowledge Infrastructure, Web of Science, PubMed, Embase, and the Cochrane Library, with a search period up to May 13, 2024. The focus was on clinical randomized controlled trials evaluating acupuncture treatment for stiff neck. The primary outcome measures were the total effective rate and visual analog scale scores. The quality of evidence and methodology of the included studies were assessed according to the GRADEpro guidelines. Meta-analysis was conducted to assess the results, with heterogeneity analysis, sensitivity analysis, subgroup analysis, trial sequential analysis, and publication bias analysis performed to verify the robustness of the combined results and explore potential sources of heterogeneity.

Result:

This study evaluated 10 clinical randomized controlled trials comparing acupuncture therapy with conventional treatment, involving 754 patients. The treatment group received acupuncture alone or in combination with conventional treatment, whereas the control group received only conventional treatment. The analysis results showed that the treatment group was significantly superior to the control group in improving the total effective rate (risk ratio = 1.12, 95% confidence interval [CI] [1.04, 1.21], P = .002), reducing visual analog scale scores (mean difference [MD] = −0.93, 95% CI [−1.29, −0.57], P < .001), reducing neck disability index scores (MD = −6.39, 95% CI [−6.79, −6.00], P < .001), and restoring cervical range of motion (cervical lateral flexion: MD = 4.29, 95% CI [3.15, 5.43], P < .001; cervical rotation: MD = 6.08, 95% CI [4.46, 7.70], P < .001).

Conclusion:

Acupuncture is an effective and safe method for treating stiff neck. However, to validate this conclusion, more rigorously designed and higher-quality studies are needed in the future.

1. Introduction

Stiff neck, a prevalent cervical malady, manifests predominantly as persistent unilateral muscle tension, pain, rigidity, and limited mobility.^[1] In the context of an accelerated pace of life and mounting occupational stress, the incidence of stiff neck has witnessed a gradual increase. Previous studies reported that the incidence rate in China was about 14.6%, with an alarming recurrence rate of up to 85% within 1 to 5 years post-initial onset.^[2,3] While minor symptoms may resolve spontaneously within 4 to 5 days, untreated severe cases can persist for weeks or even months, leading to recurrent episodes that severely compromise the patient’s quality of life. The pathogenic mechanisms of stiff neck are intricately linked to inappropriate sleeping postures or improper pillow use, resulting in prolonged abnormal positional strain on the cervicocranial juncture, sustained neck muscle tension, biomechanical imbalance in the cervical musculature, and ensuing static imbalance and injury.^[4,5] Trauma may also precipitate stress contraction of the neck muscles, and prolonged overextension or flexion of the head and neck can provoke a stiff neck. The treatment of a stiff neck aims to relieve neck pain and relax taut muscles to restore the normal range of motion of the cervical spine. The contemporary therapeutic mainstay encompasses physical therapy modalities such as joint mobilization, manual therapy, and shockwave therapy.^[6] Acupuncture has been substantiated as notably efficacious in analgesia and is widely employed for treating musculoskeletal injuries due to its minimal invasiveness and high safety profile.^[7] Previous studies indicated that acupuncture can activate specific brain regions to release endogenous opioid-like substances, effectively reducing the perception and response of the cerebral cortex to pain.^[8] Moreover, acupuncture alleviates muscular tension and contraction, reduces local inflammation and edema, and potentially recalibrates the function of affected small joints, thereby restoring musculoskeletal biomechanical equilibrium.^[9] This interventional mechanism not only ameliorates symptoms but also facilitates the body’s self-repair processes, enhancing the overall therapeutic impact. Hence, acupuncture is regarded as an efficacious complementary and alternative therapy for various pain and musculoskeletal conditions.

As the integration of acupuncture and its combination with standard medical treatments for stiff neck garners increasing utilization, its performance within clinical control trials has piqued the interest of medical researchers. Although individual trials have yielded preliminary evidence, to date, there exists a significant gap in the form of a comprehensive systematic review and meta-analysis that could consolidate these disparate data strands to bolster clinical efficacy with robust statistical support. The present study endeavors to address this lacuna by conducting a thorough analysis of extant literature, appraising the effectiveness of acupuncture alone or in conjunction with conventional treatments in alleviating pain and restoring cervical mobility in cases of stiff neck. The systematic review will be conducted within a rigorous methodological framework, encompassing literature screening, data extraction, quality assessment, and data synthesis. The meta-analysis will employ statistical techniques to amalgamate trial findings, assessing the overall efficacy and safety of acupuncture therapy. Through this study, we anticipate forging a more scientific and precise evaluation of the therapeutic efficacy of acupuncture for stiff neck, thereby informing clinicians and refining patient treatment protocols.

2. Materials and methods

2.1. Research registration

The present study has been prospectively registered with the PROSPERO database for systematic reviews (CRD42024497741). It is structured and reported in adherence to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines, ensuring a rigorous and transparent evaluation process.^[10]

2.2. Retrieval strategy

Two reviewers (FJL and YBB) conducted comprehensive searches using terms such as “acupuncture,” “acupuncture therapy,” “acupuncture treatment,” “stiff neck,” and “acute neck strain” as keywords or subject headings. The databases searched included the China National Knowledge Infrastructure, Wanfang Data, VIP, China Biology Medicine disc, PubMed, EMBASE, Web of Science, and the Cochrane Library. The search encompassed all relevant clinical randomized controlled trials on the treatment of stiff neck with acupuncture up until May 13, 2024. The specific search strategies are outlined in Table S1, Supplemental Digital Content, http://links.lww.com/MD/N867.

2.3. Inclusion criteria

The criteria for inclusion in this systematic review were as follows: (1) study type: clinical randomized controlled trials involving acupuncture treatment for stiff neck; (2) study subjects: patients with a definitive diagnosis of stiff neck based on authoritative diagnostic criteria; (3) intervention measures: the treatment approach for the experimental group should primarily involve acupuncture therapy, which may be combined with the treatment methods used in the control group (CG). The CG should receive conventional treatment modalities; (4) efficacy indicators: the primary outcome measures include the total effective rate and pain intensity, with pain intensity assessed by the visual analog scale (VAS) score.

2.4. Exclusion criteria

Exclusion criteria for the systematic review were meticulously defined as follows: (1) literature not primarily focused on acupuncture as a treatment modality; (2) narrative reviews, anecdotal expert experiences, case reports, theoretical discussions, mechanistic studies, animal experimental research, and patent-related literature; (3) in the event of multiple publications, only the most recent article would be retained; (4) clinical studies with fewer than 20 participants in the experimental group; (5) studies treating stiff neck in conjunction with other comorbid conditions; (6) literature not written in either Chinese or English.

2.5. Data extraction

Two specially trained evaluators (FJL and YBB) utilized a predesigned form to establish a literature database on “Acupuncture Treatment for stiff neck,” inputting the first author’s name, publication year, sample size, patient gender, age, disease duration, detailed treatment methods, treatment frequency, total course of treatment, outcome indicators, and literature quality scores of the articles that met the inclusion and exclusion criteria. The treatment methods and needle names in the literature meeting the inclusion and exclusion criteria were standardized according to the standards of secondary disciplines in Chinese and Western clinical medicine to ensure the consistency and accuracy of the information. Independently, based on predefined screening and review criteria, the standardized literature underwent further screening, browsing, verification, and cross-review. For literature with disputed inclusion, a third senior evaluator (WTT) decided whether to include it based on the preestablished dispute resolution criteria.

2.6. Deviation risk assessment

Two evaluators (RSJ and QYC) conducted a risk of bias assessment on the included literature according to the guidelines of the Cochrane Handbook version 5.4.0, which encompassed aspects such as random sequence generation, allocation concealment, blinding, integrity of outcome data, selective reporting, and other sources of bias. Each study was categorized based on the scoring criteria into low risk, unclear risk, or high risk of bias. In the event of any disagreements during the evaluation process, a third evaluator (YS) would be consulted to determine the final risk level.

2.7. Statistical analysis

Statistical analyses were conducted using Review Manager 5.4, StataSE 15, and trial sequential analysis (TSA) software. For dichotomous variables, the risk ratio (RR) with a 95% confidence interval (CI) was used, while for continuous variables, the mean difference (MD) with a 95% CI was employed, setting a threshold of P ≤ .05 for statistical significance.^[11] Heterogeneity was assessed using I², and when I² > 50%, a random-effects model was used to analyze the pooled effects; otherwise, a fixed-effects model was applied. Sensitivity analyses were employed to validate the robustness of the pooled results, and subgroup analyses were conducted to further investigate the sources of heterogeneity. Potential publication bias was evaluated through a funnel plot combined with the Egger test, setting a significance threshold for publication bias at P < .05.^[12] Lastly, TSA was utilized to control random errors and evaluate the robustness and reliability of the cumulative evidence, with the required information size determined based on an expected relative risk reduction of 20%, a Type I error rate of 5%, and a power of 80%.

2.8. Subgroup analysis

Subgroup analyses in this study were conducted based on the following criteria: (1) whether the treatment group’s (TG) acupuncture therapy was combined with conventional therapy; (2) the average age of patients, categorized as either < 40 years or ≥ 40 years; (3) the average duration of the patient’s condition, categorized as either < 2 days or ≥ 2 days; (4) the total duration of the treatment course, categorized as either ≤ 3 days or > 3 days.

2.9. Quality assessment

Two assessors (RSJ and QYC) independently utilized the GRADEpro Guideline Development Tool framework and the modified Jadad scale to systematically assess the quality of the evidence. The assessment criteria included the rigor of study design, potential risk of bias, consistency of results, direct applicability of evidence, and precision of data. Based on the assessment results, the quality of evidence was classified as “high,” “moderate,” “low,” or “very low.” The modified Jadad scale has a total score of 7, with scores ranging from 1 to 3 indicating low quality and 4 to 7 indicating high quality.

3. Result

3.1. Search result

A total of 2281 articles were retrieved for this study. After applying a literature deduplication tool, 1009 articles were excluded. Following a detailed review of titles and abstracts, 760 articles were further excluded. Subsequently, the remaining 512 articles were subjected to full-text reading and screened based on inclusion and exclusion criteria, resulting in the exclusion of 502 articles. Reasons for exclusion included inability to access the full text, insufficient sample size, or lack of clear diagnostic criteria, among others. In the end, 10 Chinese articles were included in the study. The detailed search methodology is depicted in Figure 1.

Figure 1.

Retrieval flowchart.

3.2. Summary of included studies

This study included 10 single-center randomized controlled trials conducted in China, involving 754 patients. In terms of diagnostic criteria, all studies utilized published authoritative diagnostic standards.^[13] Regarding the intervention measures, 2^[9,14] studies employed acupuncture therapy alone in the TG, while the remaining studies combined acupuncture therapy with conventional treatment. Specifically, the studies that used acupuncture therapy alone employed Millineed needle acupuncture. The studies that combined acupuncture therapy with conventional treatment included 6^[15–20] studies using Millineed needle acupuncture, 1^[21] study using warm acupuncture moxibustion, and 1^[22] study using electroacupuncture. The sample sizes ranged from 46 to 110 cases, with an average patient age ranging from 23 to 47 years. The average duration of the illness ranged from 1 day to 5 days, and the treatment duration ranged from 3 days to 7 days. For more detailed information, please refer Table 1.

Table 1.

Characteristics of the included trials.

Study ID	Sample size	(M/F)		Age (mean ± SD) (year)		Course of disease		Interventions		Frequency of treatment		Total treatment duration (day)	Outcome measures	Score of the modified Jadad scale
	TG/CG	TG	CG	TG	CG	TG	CG	TG	CG	TG	CG
Chen Hongping (2009)[15]	34/34	21/13	18/16	25.70	26.90	2 h–3 d	2 h–3 d	MNAC + MC	MC	Once daily	Once daily	3	①、④	3
Li Bo (2024)[16]	52/55	20/32	25/30	47 ± 13	43 ± 13	1.2 ± 0.4 (day)	1.1 ± 0.3 (day)	MNAC + MA	MA	Every other day	Every other day	7	④、⑥、⑦	5
Fang Xiupeng (2020)[21]	40/40	17/23	21/19	30.65 ± 5.78	31.73 ± 5.84	1.81 ± 0.54(day)	1.82 ± 0.61 (day)	WACM + MA	MA	Once daily	Once daily	3	①、④、⑦	3
Xiong Zhenkun (2017)[17]	30/30	16/14	15/15	33.30 ± 12.03	35.98 ± 13.82	30.74 ± 26.12 (hour)	32.23 ± 23.62 (hour)	MNAC + MA	MA	Once daily	Once daily	3	①、③、④	2
Cheng Sheng (2023)[9]	31/32	16/15	15/17	44. 23 ± 8.36	45.72 ± 7.24	2.42 ± 1.16 (day)	2.31 ± 1.28 (day)	MNAC	TM	Once daily	Every other day	3	①、②、③、④	4
Hu Chaoyao (2021)[22]	31/29	11/20	9/20	39.28 ± 5.71	39.86 ± 5.85	2.00 ± 0.71 (day)	2.09 ± 0.62 (day)	EA + MA	MA	Once daily	Once daily	4	①、④	1
Ni Weiping (2012)[14]	21/22	14/9	13/10	43 ± 13	44 ± 11	18.70 ± 15.38 (hour)	20.13 ± 19.66 (hour)	MNAC	MA	Once daily	Once daily	3	①、④	4
Wang Qikai (2022)[19]	52/51	26/26	25/26	37.21 ± 4.88	38.32 ± 5.14	19.12 ± 4.37 (hour)	18.47 ± 4.22 (hour)	MNAC + MA	MA	Once daily	Twice daily	3	①、②、③、④、⑤	3
Ling En (2019)[20]	55/55	31/24	28/27	45.5 ± 8.9	46.5 ± 7.9	4.2 ± 2.5 (day)	4.5 ± 2.3 (day)	MNAC + OM	OM	Once daily	Once daily	7	①、③、④	3
Tian Hongzhao (2015)[18]	30/30	15/15	17/13	25.00 ± 1.00	23.00 ± 2.00	NM	NM	MNAC + MA	MA	Once daily	Once daily	3	④、⑥	3

① Visual analog scale for pain (VAS); ② neck disability index (NDI); ③ cervical spine mobility; ④ total effective rate; ⑤ serum inflammatory factor levels; ⑥ McGill pain questionnaire short-form (SF-MPQ); ⑦ cervical spine mobility score.

CG = control group, EA = electroacupuncture, F = female, M = male, MA = massage, MC = microwave, MNAC = Millineed needle acupuncture, NM = not mentioned, OM = oral medication, TG = treatment group, TM = topical medication, WACM = warm acupuncture moxibustion.

3.3. Risk of trial bias

Regarding the bias in random sequence generation, 7^{[9,15,16,18–21]} studies used random number tables and 1^[14] study used simple randomization, all of which were classified as low risk. One^[17] study had a methodological description missing, which made it difficult to accurately assess the risk of random bias, therefore it was classified as unclear. Additionally, 1^[22] study was classified as high risk due to flaws in the randomization procedure. As for allocation concealment, 1^[16] study used the envelope method for allocation concealment and was assessed as low risk. The details of allocation concealment were not specified in the remaining studies, thus classified as unclear. Due to the nature of acupuncture therapy, achieving double-blinding was challenging. One^[16] study employed single-blinding design, hence it was classified as high risk for blinding of patients and operators, but low risk for blinding of outcome assessors. The reports of the remaining studies lacked specific details on blinding, making it difficult to thoroughly assess the risk of blinding bias for participants and outcome assessors, thus categorized as unclear risk. Regarding complete outcome data, all studies reported complete data, which was classified as low risk. In terms of reporting bias, 1^[14] study had missing follow-up data without providing reasons, resulting in a high-risk classification. One^[9] study had dropout cases with reasonable explanations, thus rated as low risk. The remaining studies did not have missing data and were classified as low risk. Finally, in assessing other potential sources of bias, as specific information about other factors that could affect the results was lacking, all studies were classified as unclear risk. For detailed information, refer Figure 2.

Figure 2.

Risk of bias graph.

3.4. Primary outcomes

3.4.1. Total effective rate

All studies reported the total effective rate. Heterogeneity tests suggested a high degree of heterogeneity among studies (I² = 68%, P < .001). Sensitivity analysis indicated a strong robustness of the combined effect size, with no significant changes in heterogeneity among studies. Analysis using a random-effects model demonstrated that the TG’s total effective rate was significantly superior to that of the CG, with a statistically significant difference (RR = 1.12, 95% CI [1.04, 1.21], P = .002). Detailed information can be found in Figures 3 and 4.

Figure 3.

Sensitivity analysis of total effective rate.

Figure 4.

Forest plot of meta-analysis on the total effective rate.

Subgroup analysis showed that there was no significant difference in the total effective rate between the TG and the CG when the average age of patients was ≥ 40 years (RR = 1.14, 95% CI [0.96, 1.35], P = .14) or when the total course of treatment was > 3 days (RR = 1.10, 95% CI [0.94, 1.30], P = .24). However, when the average age of patients was < 40 years (RR = 1.14, 95% CI [1.08, 1.22], P < .001) or when the total course of treatment was ≤ 3 days (RR = 1.16, 95% CI [1.09, 1.24], P < .001), the total effective rate of the TG was significantly better than that of the CG. Furthermore, the significant superiority of the TG over the CG was not affected by whether acupuncture therapy was combined with conventional treatment (combine: RR = 1.10, 95% CI [1.03, 1.18], P = .007; sole: RR = 1.28, 95% CI [1.06, 1.53], P = .009) or by the average course of disease in patients (<2 days: RR = 1.10, 95% CI [1.00, 1.22], P = .05; ≥2 days: RR = 1.22, 95% CI [1.08, 1.36], P < .001). Refer Table 2.

Table 2.

Subgroup analysis for the outcomes of included studies.

Subgroup	Eligible studies	Sample size (n)		Heterogeneity analysis		Meta-analysis		Effect model
		TG	CG	I² (%)	P	MD/SMD/RR (95% CI)	P
Total effective rate
Whether combined with conventional therapy
Combine	8	324	324	65	.005	1.10 [1.03, 1.18]	.007	Random
Sole	2	52	54	33	.22	1.28 [1.06, 1.53]	.009	Fixed
Course of treatment
>3 days	3	138	139	80	.006	1.10 [0.94, 1.30]	.24	Random
≤3 days	7	238	239	48	.08	1.16 [1.09, 1.24]	<.001	Fixed
Age
≥40 years	4	159	164	83	<.001	1.14 [0.96, 1.35]	.14	Random
<40 years	6	217	214	42	.12	1.14 [1.08, 1.22]	<.001	Fixed
Course of disease
<2 days	5	195	198	72	.007	1.10 [1.00, 1.22]	.05	Random
≥2 days	3	117	116	0	.38	1.22 [1.08, 1.36]	<.001	Fixed
VAS
Whether combined with conventional therapy
Combine	6	239	239	96	<.001	-0.82 [-1.19, -0.44]	<.001	Random
Sole	2	52	54	0	.89	-1.44 [-1.92, -0.96]	<.001	Fixed
Course of treatment
>3 days	2	83	84	97	<.001	-0.99 [-2.61, 0.62]	.23	Random
≤3 days	6	208	209	84	<.001	-0.92 [-1.27, -0.57]	<.001	Random
Age
≥40 years	3	104	109	92	<.001	-0.94 [-1.97, 0.10]	.08	Random
<40 years	5	187	184	90	<.001	-0.97 [-1.38, -0.55]	<.001	Random
Course of disease
<2 days	5	195	198	93	<.001	-0.48 [-0.78, -0.17]	.002	Random
≥2 days	2	62	61	6	.3	-1.64 [-2.01, -1.27]	<.001	Fixed

CG = control group, MD = mean difference, RR = risk ratio, TG = treatment group.

TSA indicated that the Z curve crossed the TSA-adjusted benefit boundary, demonstrating that the cumulative evidence has achieved statistical significance. Consequently, further trial results are unlikely to change this conclusion (see Fig. 5). This implies that acupuncture therapy has a robust and conclusive effect in enhancing the total effective rate in treating patients with a stiff neck.

Figure 5.

Trial sequential analysis of total effective rate.

3.4.2. VAS

Eight^{[9,14–17,19,21,22]} studies reported VAS scores. Heterogeneity analysis revealed an extremely high level of heterogeneity among studies (I² =  95%, P < .001). Sensitivity analysis indicated that the combined results were highly robust and the high heterogeneity among studies did not significantly change. Analysis using a random-effects model demonstrated that the TG had a more significant effect in reducing VAS scores compared to the CG (MD = −0.93, 95% CI [−1.29, −0.57], P < .001). Detailed results can be found in Figures 6 and 7.

Figure 6.

Sensitivity analysis of VAS score. VAS = visual analog scale.

Figure 7.

Forest plot and meta-analysis of VAS score. VAS = visual analog scale.

Subgroup analysis showed that when the average age of patients was ≥ 40 years (MD = −0.94, 95% CI [−1.97, 0.10], P = .08) or the total course of treatment was > 3 days (MD = −0.99, 95% CI [−2.61, 0.62], P = .23), there was no significant difference in the effect of reducing VAS scores between the TG and the CG. However, when the average age of patients was < 40 years (MD = −0.97, 95% CI [−1.38, −0.55], P < .001) or the total course of treatment was ≤ 3 days (MD = −0.92, 95% CI [−1.27, −0.57], P < .001), the effect of reducing VAS scores in the TG was significantly better than that in the CG. Additionally, the significant superiority of the TG over the CG was not affected by whether acupuncture therapy was combined with conventional treatment (combine: MD = −0.82, 95% CI [−1.19, −0.44], P < .001; sole: MD = −1.44, 95% CI [−1.92, −0.96], P < .001) or by the average course of disease in patients (<2 days: MD = −0.48, 95% CI [−0.78, −0.17], P = .002; ≥2 days: MD = −1.64, 95% CI [−2.01, −1.27], P < .001) (see Table 2).

TSA revealed that the Z curve crossed the TSA-adjusted benefit boundary, indicating that the cumulative evidence has reached a statistically significant level, and further trial results are unlikely to alter this conclusion (see Fig. 8). This suggests that acupuncture therapy has a robust and definitive effect in reducing VAS scores in patients with a stiff neck.

Figure 8.

Trial sequential analysis of VAS score. VAS = visual analog scale.

3.5. Secondary outcomes

3.5.1. Neck disability index

Two^[9,19] studies reported neck disability index (NDI). Given the homogeneity among studies (I² = 0%, P = .70), a fixed-effects model analysis was employed, which revealed that the TG could effectively reduce NDI scores compared to the CG (MD = −6.39, 95% CI (−6.79, −6.00), P < .001). These findings are represented in Figure 9.

Figure 9.

Forest plot and meta-analysis of NDI. NDI = neck disability index.

3.5.2. CSM

Four studies reported CSM. Due to differences in statistical methods, only 2 studies met the criteria for inclusion in the analysis, which evaluated both lateral flexion and rotation. Heterogeneity analysis showed homogeneity among studies (I² = 0%, P = .87; I² = 0%, P = 1.00), hence a fixed-effects model was used for analysis. The results indicated that the TG showed significantly better recovery in cervical lateral flexion and rotation functions compared to the CG, with statistically significant differences (cervical lateral flexion: MD = 4.29, 95% CI (3.15, 5.43), P < .001; cervical rotation: MD = 6.08, 95% CI (4.46, 7.70), P < .001). See Figures 10 and 11 for details.

Figure 10.

Forest plot and meta-analysis of cervical lateral flexion range of motion.

Figure 11.

Forest plot and meta-analysis of cervical rotation range of motion.

3.6. Adverse events

A meticulous review of the included literature revealed that only 1 study reported adverse reactions. This lack of reporting may suggest a high safety profile of the therapies under study but could also reflect inadequacies in the monitoring, collection, or reporting of adverse reactions. Therefore, while there is no current evidence of serious adverse effects during the treatment course, caution should be exercised in interpreting this finding.

3.7. Quality assessment

The GRADEpro assessment results indicate that the quality of evidence for all outcome measures is very low (see Table S2, Supplemental Digital Content, http://links.lww.com/MD/N867). The modified Jadad scale assessment shows that 1^[16] study scored 5 points, and 2^[9,14] studies scored 4 points, classifying them as high quality. The remaining studies were evaluated as low quality.

3.8. Publication bias

This study utilized funnel plots and Egger test to assess publication bias in the primary outcome measures. The funnel plot for total effective rate showed significant asymmetry, and combined with the results of Egger test (t = 6.98, P < .001), it suggests that conclusions regarding the total effective rate may be influenced by publication bias. Similarly, the asymmetry of the funnel plot and the results of Egger test (t = −4.50, P = .004) indicate a significant publication bias in VAS scores. Detailed information can be found in Figures 12 and 13.

Figure 12.

Publication bias of total effective rate.

Figure 13.

Publication bias of VAS score. VAS = visual analog scale.

4. Discussion

This study conducted a meta-analysis and systematic review of 10 studies, encompassing 754 patients. The meta-analysis results indicate that acupuncture therapy is significantly superior to the CG in reducing VAS scores, decreasing NDI scores, restoring cervical spine mobility, and improving total effective rate. Sensitivity analysis demonstrated the robustness of the combined results. Subgroup analysis showed that heterogeneity and combined results changed significantly in some subgroups after grouping, but overall robustness remained high. The TSA further validated the conclusiveness of the combined results for the primary outcome measures.

However, several issues indicate that the above conclusions should be interpreted cautiously. The quality assessment of the evidence shows very low-certainty evidence supporting that acupuncture therapy is superior to conventional therapy in reducing VAS, decreasing NDI, improving cervical spine mobility, and increasing total effective rate. This is related to the high heterogeneity among studies and flaws in the study designs. Additionally, the modified Jadad scale assessment reveals that most of the included studies are classified as “low quality,” mainly due to lack of allocation concealment, blinding, and information regarding the number and reasons for dropouts.

Heterogeneity analysis shows homogeneity among studies for secondary outcome measures such as NDI and cervical spine mobility. However, there is high heterogeneity for primary outcome measures such as total effective rate and VAS scores. This significant difference may stem from various factors, such as differing patient population characteristics, intervention measures, total treatment duration, and the subjectivity of outcome assessments. Further sensitivity analysis indicates that the heterogeneity of the combined results did not change significantly, suggesting that the sources of this high heterogeneity are likely multifaceted and complex. Publication bias analysis for the primary outcomes shows significant publication bias for VAS scores and total effective rate. Similarly, the subjectivity of outcome assessments is likely a potential source of publication bias.

Subgroup analysis showed that in terms of outcome robustness, when the patient’s age is ≥ 40 years and the total treatment duration is > 3 days, acupuncture therapy showed no significant difference compared with conventional therapy in improving the total effective rate and reducing VAS scores. This may be due to the decreased recovery ability of the body with age and the tendency of pain to become chronic as the treatment duration increases. Acupuncture therapy achieved better results than conventional therapy in all other subgroups. In terms of heterogeneity, when the TG used acupuncture therapy alone or the duration of the patient’s condition was ≥ 2 days, heterogeneity in the total effective rate and VAS score outcomes was significantly reduced. Similarly, when the patient’s age was < 40 years or the total treatment duration was ≤ 3 days, the high heterogeneity in the total effective rate outcomes was also significantly reduced. This may be related to the standardization of treatment methods reducing methodological errors, the significant immediate effect of acupuncture therapy, the better physiological recovery capacity and adaptability of younger individuals, and the gradual stabilization of symptoms in patients as the duration of the condition extends.

This study showcases several innovations in design and methodology. First, this research focuses on the effects of standalone acupuncture treatment and its combination with conventional therapy, eliminating the interference of other traditional Chinese medicine treatments and ensuring methodological rigor. Second, it covers large-sample randomized controlled trials of various types of acupuncture, enhancing the generalizability of the findings. Third, the robustness of the combined results is verified through sensitivity and subgroup analyses, which also delve into the sources of high heterogeneity. Additionally, the reliability of the findings is further confirmed through trial sequential analysis. Moreover, funnel plots and Egger test are combined to precisely assess potential publication bias from both visual and statistical perspectives, potentially enhancing the credibility of the results. Finally, by employing a unified pain scoring tool, the potential systematic bias caused by different scoring systems is minimized to the greatest extent.

This research preliminarily validates acupuncture’s effectiveness and safety in managing stiff neck, establishing a robust theoretical groundwork for its broad application in clinical settings in the future. Nonetheless, this systematic review and meta-analysis comes with certain constraints. To start with, the geographical uniformity of the studies: exclusively conducted in China: introduces clear geographic bias that may curb the broad applicability of these results. It is worth noting that all the indicators in this study are subjective, which limits the reliability of the evidence. This may be related to the general lack of objective indicators in the clinical evaluation of stiff neck treatment efficacy. This suggests that future studies should explore objective indicators related to the pathology and treatment mechanisms of stiff neck, such as inflammatory markers and electromyography. In addition, the absence of randomized controlled trials directly comparing traditional acupuncture with placebo treatments within this body of work detracts from the precise evaluation of its benefits. Furthermore, the prevalence of methodological issues, including incorrect randomization techniques and a lack of blinding, introduces significant variability across the research examined. The lack of information on adverse reactions and extended follow-up also casts doubt on the dependability of these findings. Moving forward, it is imperative that subsequent investigations align with global standards for clinical trial reporting, embracing more expansive, multicenter, double-blinded, and randomized controlled trials and comprehensively documenting adverse reactions and long-term treatment outcomes to build a stronger empirical foundation for acupuncture treatment of stiff neck, providing more reliable guidance for clinical practice.

5. Conclusion

This study, through a meta-analysis and systematic review of existing randomized controlled trials, concludes that acupuncture therapy is more effective than conventional treatments in alleviating pain, reducing functional impairments, and enhancing cervical range of motion in patients with stiff necks. However, due to design and methodological flaws in previous studies, further validation through high-quality, rigorous, and adequately powered clinical randomized controlled trials is necessary.

Author contributions

Data curation: Bingbing Yu, Yan Yang, Junlin Fang.

Formal analysis: Bingbing Yu, Yan Yang, Junlin Fang, Yun Guo, Yicheng Qiu, Shun Yang, Sijia Ran, Kai Zheng, Tingting Wang.

Methodology: Bingbing Yu, Tingting Wang, Yingru Huang.

Project administration: Tingting Wang, Yingru Huang.

Software: Bingbing Yu, Junlin Fang, Yicheng Qiu, Shun Yang, Sijia Ran.

Writing – original draft: Bingbing Yu.

Writing – review & editing: Bingbing Yu, Yun Guo, Tingting Wang, Yingru Huang.