Update evidence of effectiveness on pain relieving of cupping therapy: a systematic review and Meta-analysis of randomized controlled trials

Abstract

OBJECTIVE:

To update the current best evidence on the effectiveness and safety of cupping therapy in pain management.

METHODS:

The protocol of this systematic review was registered at PROSPERO (CRD42021261308). An updated literature searching in 7 databases was conducted from January 2014 to January 2023. Two authors extracted data and assessed the risk of bias independently. Statistical analysis was performed using RevMan 5.4.1 software (Cochrane Collaboration, London, UK). Meta-analysis with a random effect model was conducted when there was no serious statistical heterogeneity among trials (I²≤75%). Grading of Recommendations Assessment, Development, and Evaluation was also conducted to assess the quality of evidence.

RESULTS:

Seventy-two trials with 5720 participants were included. All included trials were assessed as having high risk of bias. The majority of the included trials reported the benefit of cupping plus other therapy or cupping alone on improving cure rate (average risk ratio more than 1.15) and reducing visual analogue scale [average mean difference (MD) reduction 0.16 to 7.0 cm], improving quality of life, quality of sleep or other symptoms related to pain condition. And there was low/very low quality evidence that the incidence of adverse events in the cupping groups were lower than that in the control groups. Although the heterogeneity between studies and the methodological quality of the study itself lead to the low evidence strength of the current conclusions, the results of this study are a valuable supplement to the founding of previous review.

CONCLUSION:

Cupping therapy alone or combined with other therapy was considered benefit in relieving pain, improving the quality of life, and increasing the cure rate of patients with pain conditions, though supported by the low quality of evidence. According to the limited evidence, cupping therapy seems to have less harm than drugs when treating pain conditions.

1. INTRODUCTION

Pain is almost universal, which has not only brought huge pain and suffering to patients, but also caused hundreds of millions of economic losses to society.^1,⇓,⇓-4 International Association for the Study of Pain defined the pain as a distressing experience associated with actual or potential tissue damage with sensory, emotional, cognitive, and social components.⁵ Acute pain usually lasts no more than 7 d, but it can usually extend to 30 d and may recur periodically or even turn into chronic pain.⁶ Chronic pain generally lasts for more than 3 months. As a serious public health problem, it is associated with severe emotional distress and/or functional disability and cannot be well explained by other causes.⁷ Chronic pain seriously affects physical and mental function, reduces productivity and quality of life, affecting about 50 million people and causing a loss of 635 billion dollars in the United States.^8,9

Pain condition mainly involves back pain, neck pain, musculoskeletal pain (non-back pain, neck pain), peripheral neuralgia, postoperative pain, toothache, kidney stones, sickle cell crisis, and so on.^10,11 According to the World Health Organization (WHO) report, drugs for pain mainly included opioids and non-opioids. Opioids are generally considered the most effective analgesics and are commonly used for pain management in most patients, including the elderly and the disabled.^12,13 Although it was only for the treatment of cancer pain, the “analgesic ladder”¹⁴ is recommended by WHO for acute and chronic pain as the basic treatment.¹⁰ The adverse reactions of opioids include tolerance, addiction, constipation, and respiratory depression.¹⁵ The most commonly used non-opioid drugs are non-steroidal anti-inflammatory drugs (NSAIDs), which could be oral administration or topical used. Oral NSAIDs often caused gastrointestinal discomfort, while topical NSAIDs may cause skin reaction.¹⁵ Due to the adverse reactions of drugs, many patients seek non-drugs therapy for pain relieving. Cupping as a non-drugs therapy was also increasingly known for its efficacy in treating pain-related diseases.^16,17

We published a systematic review involved 16 trials with 921 participants to assess cupping therapy for pain management in 2014,¹⁸ however, no affirmative conclusion about cupping therapy could be drawn due to the limited number of included trials. Considering there are more trials published in the past 9 years, we conduct this systematic review to provide an up-to-date evidence-based evaluation of cupping therapy in treating acute and chronic pain.

2 METHODS

2.1. Study registration

The protocol of this systematic review was registered at PROSPERO (CRD42021261308), accessible at http://www.crd.york.ac.uk/PROSPERO/display_record.asp?ID=CRD42021261308. This review was conducted and reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)¹⁹ statement Guidelines.

2.2. Inclusion criteria

(a) Participants of any gender over 18 years with pain conditions were included in the review, regardless of the cause and type of diseases.

(b) The intervention was any form of cupping therapy (retained cupping, dry cupping, wet cupping, flash cupping, moving cupping, medicinal cupping, needling cupping, et al).

(c) The control could be no treatment/wait-list, sham cupping, usual care (e.g. exercise, heat therapy) or drugs in treating of pain condition. Comparisons also included a combination of cupping therapy and other treatment versus other treatment alone. Trials that used combined therapy employing cupping therapy with other Traditional Chinese Medicine (TCM) therapy (such as acupuncture or herbal medicine) compared with other non-TCM interventions were not included.

(d) Primary outcomes included pain severity scores (e.g. visual analogue scale, VAS), number of patients who have at least 50% of maximum possible pain relief over baseline, and number of patients who have at least 30% of maximum possible pain relief over baseline. Secondary outcomes included number of tender points (for some specific diseases, e.g., fibromyalgia), patients reported episodes of pain, patients or physicians’ global evaluation, psycho-social function outcomes (e.g. the Hamilton Depression Scale, HAMD), Quality of Life [QoL, e.g. Short Form-36 (SF-36)] and adverse events. Adverse events could be assessed by reporting early study discontinuations, the worsening of pain in participants, and other adverse events during the treatment and follow-up period. The included trial should report at least one of the above outcomes.

(e) Randomized controlled trials were considered for inclusion, only the first phase data of randomized cross-over trials could be used if those met the inclusion criteria. No limitation for the type of language and publication. Plagiarized or repeatedly published studies were excluded. Studies without available analyzable data were also excluded.

2.3. Searching strategy

Since the literature searching of previous review was up to 2014, updated comprehensive search was conducted from January 2014 to February 2023 in 3 Chinese databases and 4 English databases, including the Cochrane Central Register of Controlled Trials (CENTRAL), PubMed, Science Direction, Biomed Central, China Network Knowledge Infrastructure (CNKI), Chinese Scientific Journals Database (VIP) and Wanfang Database. In Chinese and English databases, searches were conducted using titles, abstracts, or keywords. The search strategy of PubMed is as follows: (cupping [Title/Abstract] OR cupping therapy [Title/ Abstract]) AND (pain [Title/Abstract] OR analgesic* [Title/ Abstract]) AND (randomized controlled trial [Title/ Abstract] OR controlled clinical trial [Title/ Abstract] OR randomized [Title/Abstract] OR placebo [Title/Abstract] OR randomly [Title/Abstract]). The search strategy of other database is shown in Supplementary Table 1.

2.4. Study selection and data extraction

Two authors (WANG Yiying and DOND Shuai) selected the eligible studies and extracted the data independently. Disagreements were settled by a discussion with a third author (CAO Huijuan). Extracted information included study design (random allocation generation and concealment, blinding method, inclusion/exclusion criteria, sample size, etc.), characteristics of participants (gender, age, type of disease/condition, etc.), details of intervention and control (type of cupping therapy, selection of acupoints, frequency and duration of treatment, type of control, details of co-interventions, etc.), outcomes data. Data of the included trials in the previous review were also integrated into this study.

2.5. Quality assessment

Two review authors (WANG Yiying and DOND Shuai) independently assessed all the included study (including 14 studies involved in the previous review) using the Cochrane risk of bias assessment tool 2.0.²⁰ This tool addresses specific five domains, namely risk of bias arising from the randomization process, risk of bias due to deviations from the intended interventions (effect of assignment to intervention), risk of bias due to missing outcome data, risk of bias in the measurement of the outcome and risk of bias in the selection of the reported result. For the above domains, the quality of each included trial was categorized as low/some concerns/high risk of bias. If all the evaluation domains of the trial are low risk of bias, the overall quality of the trial is judged as low risk of bias. When one or more items are evaluated as high risk of bias, or the evaluation results of multiple domains are likely to have some concerns and have a great impact on the credibility of the research results, the overall evaluation is high risk of bias. Otherwise, the overall quality of the trial is judged as some concerns.

2.6. Data analysis

We used Revman 5.4.1 (Cochrane Collaboration, London, UK) software to conduct Meta-analysis. One author (WANG Yiying) was responsible for entering data into the software and the second author (DONG Shuai) checked the data. Dichotomous data was present as a risk ratio (RR) with its 95% confidence interval (CI), and continuous data as a mean difference (MD) with its 95% CI. The random-effects model was used for Meta-analysis considering potential sources of clinical heterogeneity. Meta-analysis was conducted when they were sufficient homogeneity among included studies (including characteristics of the participants, details of intervention and control, type of outcome measurements). The heterogeneity was judged according to I ² value. Pooling analysis was not used if there was significant statistical heterogeneity (I ²≥75%) among trials.

Subgroup analysis was conducted to detect the source of heterogeneity among trials. Subgroups were categorized by participants’ baseline characteristics [including age, sex, type of disease (acute pain, chronic pain) and severity of the disease (medium and high)]; or by type of cupping therapy and treatment duration (short term: less than 4 weeks, medium term: 4-12 weeks, and long term: more than 12 weeks).

Sensitivity analysis was used for the robustness of the results by excluding ambiguous studies. The funnel plot was used to explore the possible publication bias when there were 10 or more studies in an analysis.

In addition, the quality of the evidence for the main outcome were assessed using Grading of Recommendations, Assessment, Development and Evaluation (GRADE)²¹ by the GRADEpro Guideline Development Tool (GDT) online. Whether the level of evidence was downgraded was mainly according to the limitations in study design or execution, the inconsistency or imprecision of results, the indirectness of evidence and the potential publication bias.

3. RESULTS

3.1. Description of the included trials

Totally 955 records were retrieved after primary literature searching from 7 databases, 170 duplicate citations were excluded, 28 citations were screened out after browsing titles and abstracts, and 16 studies were not included after reading the full text. Finally, only 72 studies (74 reports)^22-93 met our inclusion criteria. Among them, 14 were from the previous systematic review and the remaining 58 were newly included. The study flow chart was shown in Figure 1. Two studies^94,95 in the original review were not included in this review since their control group contained multiple complex interventions. One of them compared cupping therapy to drugs and hot compress, and another one compared cupping treatment to acupuncture plus wet cupping. It’s hard to explain the effect of cupping therapy based on such complext control treatments, thus we excluded these two studies in this review.

Figure 1. PRISMA 2020 flow diagram for updated systematic reviews.

PRISMA: preferred reporting items for systematic review and Meta-analyses; CNKI: Chinese national knowledge infrasturcture; VIP: China science and technology journal database; CENTRAL: Cochrane central registre of controlled trials; VAS: visual analog scale.

3.2. Study characteristics

Seventy-two trials involving 5720 participants were included in this review. The sample size in each study varied from 32 to 300 with an average of 79 patients per group. The detailed basic characteristics of the included trials are presented in Table 1. Sixty-two^{25,⇓-27,33,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-77,79,⇓,⇓,⇓,⇓-84,86,⇓,⇓,⇓,⇓,⇓,⇓- 93} trials were conducted in China and published in Chinese and the remaining 10 trials^{22,⇓-24,28,⇓,⇓,⇓-32,78,85} were published in English, in which 5 studies were conducted in Germany,^{22,24,29,30,31} two were conducted in Korea,^23,31 the remaining 3 studies were conducted in Iran,²⁸ India⁷⁸ and Brazil,⁸⁵ respectively.

Table 1.

Characteristics of 72 included trials

Study ID	Disease type		Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measurement		Intervention group operation					Control group operation		Treatment duration
			T	C	T	C	T	C			Type of cupping	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
Cupping therapy vs waiting list/no treatment or usual care
Teut M et al 201222		Osteoarthritis of theknee	5/16	8/ 11	68.10±7.20	69.30± 6.80	2.90± 0.46	2.81± 0.55		①②③⑥	a	The knee joint	10	Twice weekly	NA	Wait-list	NA	2
Kim JI et al 201123		Non-specific low back pain	5/16	3/8	44.20±9.40	48.00± 5.40	5.81± 1.12	5.27± 8.00		①②⑥	b	Bilateral BL23, BL24, BL25	5	Three times weekly	SAC	Wait-list	NA	14
Lauche R et al 201124		Non-specific neck pain	7/15	4/ 20	48.60±11.2	53.00± 11.40	4.55± 2.09	4.23± 1.08		①③⑥	a	Tender points	10-20	Every 3-4 d	NA	Wait-list	NA	14
Wu ZS 200725		Acute ankle sprain	10/ 21	11/19	NR	NR	8.61± 1.06	8.74± 0.92		①④⑥	b	Tender points	10	Once daily	NA	Wait-list	NA	5
Zhang QE, Lai Z 201726		Low back pain	50/ 52	52/54	NR	NR	4.03± 0.89	3.62± 0.91		①③④	e	From DU14 to DU2 on the back midline	10	Once every 2 d	NA	Usual care 1k	30 min, once every 2 d.	14
Dong G et al 201427		Low back pain	28/ 22	27/23	54.50 (39.00-70.00)	54.00 (38.00-70.00)	7.45± 1.48	7.38± 1.35		①④	f	BL23, DU3, BL40 and Ashi (bilateral)	5-15	Once every 2 d	NA	Usual care 2	20-30 min, once daily.	10
Farhadi K et al 200928		Non-specific low back pain	30/ 18	37/13	44.90± 14.80	41.80± 13.90	2.70± 0.80	2.70± 0.90		①⑥	b	Tender points	3-5	Once every 2 d	NA	Usual care 3	See notes for details.	6
Lauche R et al 201329		Chronic neck pain	6/24	10/21	54.50± 12.30	53.70± 13.40	5.58± 1.97	5.63± 1.86		①②③⑥	d	NA	10-15	Twice weekly	NA	Usual care 4	20 min, twice weekly.	84
Cramer H et al 201130		Chronic neck pain	4/20	6/ 18	44.50± 10.80	47.90± 13.50	4.12± 1.45	4.20± 1.57		①②③⑥	ad	Ashi	10-15	Once every 3-4 d.	NA	Usual care 5	See notes for details.	14
Kim TH et al 201231		Neck pain	7/13	11/9	25.50 (22.50-40.50)	28.00 (25.00-31.50)	5.93± 1.63	6.49± 1.49		①②③⑥	b	Ashi	5-10	Three times weekly.	NA	Usual care 1l	10 min, 3 times weekly.	49
Study ID	Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome mea-surement	Intervention group operation					Control group operation		Treatment duration
		T	C	T	C	T	C		Type of cupping	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
MichalsenA et al 200932	Carpal tunnel syndrome	2/ 24	4/ 22	57.20±7.70	59.30±8.30	6.15± 2.49	5.86± 2.51	①②③⑥	b	Tender points	5-10	Treatment once	NA	Usual care 1m	Once 15 min	7
Cupping therapy vs drugs
Xue M et al 202133	Dysmenorrhea	0/ 29	0/ 29	26.48±4.82	27.71±4.01	NR	NR	②④	b	SP8	10	Once a month	NA	Ibuprofen	0.3-0.6 g, twice daily	90
Wang L et al 202034	Postherpetic neuralgia	16/19	17/ 18	58.60	57.80	9.15± 0.31	9.07± 0.27	①④⑤⑥	b	Ashi	15-20	Once every 2 d	NA	Gabapentin capsule	0.3 g, 3 times daily (once or twice on the first two days)	16
Li XJ et al 201935	Postherpetic neuralgia	16/16	14/ 18	46.90± 13.90	46.70± 14.30	6.90± 1.30	6.97± 1.40	①④⑥	b	Zhuang medicine acupoints	10-15	Once every 3 d	NA	Gabapentin capsule	300 mg，3 times daily	60
Liu MH, Wang GY 201736	Postherpetic neuralgia	18/17	19/ 14	52.00	49.00	7.91± 1.35	7.88± 1.47	①④⑤⑥	b	NA	5-10	Once daily	NA	Flupentixol; Melitracen	10.5 mg, twice daily; 0.5 mg, 3 times daily.	28
Luo CY et al 201737	Postherpetic neuralgia (AIDS)	8/ 14	10/ 12	48.00± 18.00	50.00± 18.00	8.30± 1.50	8.30± 1.40	①④	b	Ashi	20-30	Once every 2 d	NA	Meco-balamin; Carba-mazepine	10 mg, 3 times a day; 0.5 mg, 3 times a day; 0.1 g, 3 times a day.	20
Zhang XQ, Hu CL 201538	Postherpetic neuralgia	14/24	16/ 22	60.20±9.70	61.40± 10.20	NR	NR	②④	b	The affected skin	NA	Once every 5 d	NA	Amitriptyline	25 mg, once daily at bedtime.	30
Study ID		Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measurement	Intervention group operation					Control group operation		Treatment duration
			T	C	T	C	T	C		Type of cupping	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
Feng LF 201539	Postherpetic neuralgia		22/16	20/18	56.32±9.81	58.24±9.29	7.21± 1.16	7.18± 1.21	①④	b	Ashi	10	Once every 3 d	NA	Mecobalamin; Indomethacin	5 mg, twice daily; 5 mg, twice daily.	30
Xing YK et al 201740	Herpes zoster		14/14	15/13	40.02±19.62	42.50±18.32	7.80± 1.75	7.78± 1.70	①②④⑥	b	NA	NA	Once daily	NA	sodium phosphonate (intravenous drip); Valaciclovir	3 g, once daily; 0.15 g, twice daily	14
Li ZQ 201841	Acute herpes zoster		26/33	28/31	44.83±15.36	44.44±15.33	NR	NR	④	b	Tender points	1-2	Once every 3 d	NA	Acyclovir; Mecobalamin;	0.2 g, 3 times daily; 0.5 mg; 3 times daily.	10
Chen W et al 201442	Acute herpes zoster		12/18	10/20	Range 40.00-76.00	Range 46.00-80.00	7.63± 1.19	7.40± 1.11	①④⑥	b	Tender points	10-15	Once every 2 d	NA	Valaciclovir hydrochloride (Fasting before meals).	0.3 g, twice daily.	10
Liu Y, et al 201643	Migraine		10/20	12/18	41.47±13.51	42.47 ±13.60	5.73± 2.01	6.07± 1.61	①②④	b	EX-HN5 (bilateral) and its nearby	NA	NA	NA	Diclofenac sodium	75 mg, once daily.	90
Jiang HW et al 201544	Migraine		7/23	9/21	20.00-65.00	18.00-57.00	NR	NR	④⑥	b	BL17	20	Twice weekly	NA	Ceftriaxone hydrochloride	10 mg every night.	28
Song N et al 201345	Migraine		16/29	18/27	35.40±3.10	36.10 ±2.30	6.76± 1.48	6.44± 1.78	①②④	b	G20, GV14	15	Twice weekly	NA	Flunarizine	10 mg, once daily at bedtime.	56
Zhou YC 201646	Neck myofascial pain syndrome		7/23	9/21	38.33±8.63	36.90 ±9.03	6.03± 1.32	5.93± 1.36	①②	b	Ashi	10	Once every 3 d	NA	Lidocaine	2 mL at each point, once every 3 d.	15
Wu K 201347	Osteoarthritis		8/22	7/23	56.70±6.60	57.40±5.80	6.97± 0.85	7.00± 0.87	①⑥	b	Ex-LE4, Ex-LE5, ST34, SP10, SP9, Ashi	3-4	Once every 2 d	NA	Diclofenac	50 mg, twice daily.	14
Study ID	Disease type		Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measure-ment	Intervention group operation					Control group operation		Treat-ment duration
			T	C	T	C	T	C		Type of cupping	Points of selection	Time of cupping retained (min)	Com-bination frequency	Combin-ation on therapy	Control group	Usage
Cupping therapy plus drugs vs drugs
Hou YW et al 202248		Postherpetic neuralgia	13/17	14/16	49.79±3.46	55.34±1.78	8 (8,9)	8 (7,9)	①②④	j	Ashi	10	Once every 2 d	SAC	Dihydrocodeine Tartrate Tablets	60 mg, 3 times daily	20
Luo TW, Wu XY 202249		Haemorr-hoids	15/12	15/18	38±16	39±14	6.18± 1.85	6.27± 1.89	①④	b	BL57	5	Once every 4 d	SAC	Diosmin tablets; Yinglong Ma musk hemorrhoid Cream; Potassium permanganate solution	1 g, twice daily; (applied externally), 20 min, twice daily; (sitting bath) twice daily	14
Xia ZY et al 202250		Pain after fracture	25/19	23/21	67.72±0.68	67.32±1.11	NR	NR	①②	b	GB34, GB39, BL57, Ashi, BL-25, BL54, GB30, etc	8	Once daily	SAC	Self controlledanalgesics (intravenous injection)	5 mL (first), 15 min, 2 mL/h (sustained dose), a single dose of 0.5 mL	7
Zhang CF 202251		Herpes zoster	22/18	24/16	45.26±3.74	45.48±3.62	6.94± 0.54	7.05± 0.57	①②④	b	Ashi	10	Once daily	SAC	Valacyclovir hydrochloride tablets, Mecobalamin tablets	0.3 g, twice daily 0.5 mg, 3 times daily	10
Li H 202252		Acute herpes zoster	14/18	15/17	Range27-78	Range20-76	7.21± 1.21	7.15± 1.14	①④	b	Ashi	3-5	Once daily	SAC	Acyclovir sodium needle, Me-cobalamin	0.5 g, 3 times daily; 0.5 mg, 3 times daily	7
Zhang LQ et al 202153		Postherpetic neuralgia	16/15	17/13	61.69±8.43	61.42±7.96	7.09± 1.33	7.05± 1.30	①②④⑥	b	Ashi	15-20	Once every 2 d	SAC	Pregabalin	150 mg, twice daily.	30
Yang C et al 201954		Postherpetic neuralgia	33/32	34/31	39.78±5.24	40.15±4.92	39.78±5.24	40.15± 4.92	①③④	b	The herpes area	NA	Once every 2 d	SAC	Rat nerve growth factor (Injected intra-muscularly)	30 μg, once daily.	14
Study ID		Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measure-ment	Intervention group operation					Control group operation		Treatment duration
			T	C	T	C	T	C		Type of cup-ing	Points of selection	Time of cupping retained (min)	Combina-tion frequency	Combina-tion on therapy	Control group	Usage
Su ZY et al 202055	Acute gouty arthritis		18/ 12	11/ 19	38.10±5.30	39.20±5.60	8.43± 1.17	8.46± 1.12	①②④⑥	b	NA	3-5	Once every 2 days	SAC	Fibulinastat	40 mg, once daily.	5
Wang D, Huang CZ 202056	Acute gouty arthritis		31/ 19	30/ 20	65.13±14.24	15.55±4.49	8.35± 1.50	8.23± 1.45	①④⑥	b	ST44, BL66, GB43	10-15	Once every 2 d	SAC	Diclofenac sodium	75 mg, once daily.	16
Jiang GL et al 201957	Acute herpes zoster pain		18/ 12	16/ 14	47.30±13.20	44.20±15.30	7.26± 1.34	7.18± 1.47	①⑥	b	The herpes area	10-15	NA	SAC	Valaciclovir hydrochlorideMecobalamin	0.3 g, twice daily; 0.5 mg, 3 times daily; 10 mg, 3 times daily.	15
Liu Z 201958	Herpes zoster acute blister		22/ 20	20/ 21	44.80±13.70	44.30±15.80	6.60± 1.31	6.46± 1.55	①④⑥	a	The herpes area	10	Once every 4 d	SAC	Valaciclovir hydrochloride	300 mg, twice a day; 10 mg, 3 times a day.	12
Tong LS 201659	Herpes zoster		24/ 18	22/ 18	45.00±8.40	44.50±5.30	NR	NR	④⑥	b	Ashi	15	Once 2 d	SAC	Acyclovir ointment; Acyclovir tablets	3 times daily; 0.4 g, 5 times daily.	10
Feng L, Huang SY 201860	Cervicalspondylotic radiculo-pathy		10/ 10	8/ 12	53.00±14.00	46.00±14.00	6.40± 1.54	6.30± 1.95	①②④	c	DU14, GB21, SI11, SI15, SI14, SI13, Ashi	5-10	Once daily	SAC	Loxoprofen sodium	60 mg, 3 times a day.	30
Zhang HY 201461	Lumbodor-sal myofasciitis		22/ 38	24/ 36	39.20±3.40	40.10±3.60	6.81± 2.26	6.77± 2.34	①④	de	Bladder meridians	10	Once 2 d	SAC	Votalin emulsion	3-4 times daily.	10
Li BJ 202262	Cervical spondyosis		23/17	22/ 18	47±6	46±7	7.48± 0.51	7.42± 0.55	①④⑥	b	GV14, SI11	8	Once every 5 d	SAC	Electro-acupuncture	20 min, once daily	20
Cupping therapy plus acupuncture vs acupuncture
Xu J, yANG LL 202263	Lumbago		18/17	19/ 16	56.47±3.18	56.42±3.16	8.16± 1.05	8.09± 1.02	①②④	b	BL40, BL17, Ashi, BL23	5	Once every 2 d	SAC	Warm needling (acupuncture)	30 min, once daily	30
Study ID		Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measure-ment	Intervention group operation					Control group operation		Treatment duration
			T	C	T	C	T	C		Type of cup-ing	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
Wu XY et al 202264	Scapulo-humeral periarthritis		14/19	11/20	52.36±4.99	51.35±5.10	5.70± 2.08	5.71± 2.20	①④	d	Ashi	NA	Once every week	SAC	Acupuncture	30 min, Once every 2 d	14
Huang HP et al 202265	Scapulo-humeral periarthritis		20/16	19/17	50.78±2.98	51.21±3.23	7.43± 1.04	7.45± 1.11	①②④	b	TE14, LI15, SI11, Ashi	3-7	Once daily	SAC	Warm needling (acupuncture)	30 min, Once every other day	10
Dong JP et al 202166	Dysmenorr-hea		0/26	0/26	18.00-35.00	20.00-33.00	9.56± 1.71	9.75± 1.81	④⑥	ad	Bilateral of spine	NA	Once every 3 d	SAC	Acupuncture	40 min, once daily	90
Zhang QH 202067	Postherpetic neuralgia		19/16	20/15	56.90±3.30	57.30±3.10	8.20± 0.50	8.10± 0.70	①④⑤⑥	b	Ashi	10	Once every 2 d	SAC	Fire-acupuncture	NA, Once every 2 d	20
Zhao XX 201868	Postherpetic neuralgia		NR	NR	NR	NR	NR	NR	①④⑤⑥	b	Ashi	10	Once every 2 d	SAC	Fire-acupuncture	NA, Once every other day	10
Zhang HX et al 200969	Herpes zoster		10/15	12/13	18.00-66.00	19.00-67.00	NR	NR	①④	b	Ashi	5-10	Once daily	SAC	Electro-acupuncture	30 min, NA.	10
Xie FL et al 201970	Lumbar disc herniation		86/64	80/70	48.70±4.60	46.20 ±4.10	5.24± 0.54	5.35± 0.61	①③④⑥	b	Ashi	10	Once in 2-3 d	SAC	Acupuncture	30 min, 3 times weekly	14-21
Wang ZJ et al 201871	Lumbar disc herniation		43/37	46/34	48.15±4.43	47.88 ±4.43	4.61± 0.79	4.53± 0.76	①⑥	c	Wrist and ankle	10	Once daily	SAC	Acupuncture (Wrist ankle acupuncture therapy)	30 min, once daily.	7
Cupping therapy plus acupuncture vs acupuncture
Lu P et al 201672	Lumbar disc herniation		17/14	15/16	47.48±7.05	45.70±8.37	6.452±0.624	6.355±0.733	①⑥	eg	EX-B2, BL17 BL23, BL25 and EX-B6 (bilateral)	5	Once daily	SAC	Acupuncture	30 min, once daily	28
Study ID		Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measure-ment	Intervention group operation					Control group operation		Treatment duration
			T	C	T	C	T	C		Type of cupping	Points of selection	Time of cupping retained (min)	Combination frequency	Combina-tion on therapy	Control group	Usage
Chen WJ et al 201973	Chronic low back pain		17/20	16/21	58.00±6.20	58.00±7.80	8.11±1.22	8.05±0.97	①⑥	i	Ashi	NA	Once daily	SAC	Acupuncture	30 min, once daily	10
Liu HY 202074	Low back pain		20/20	21/19	42.48±2.59	41.22±1.96	NR	NR	①④	a	Dai meridian	≤15	Every 2 d	SAC	Acupuncture	15 min, once daily	NR
Ma LY, Bi Y 202075	Knee osteoarthritis		16/14	15/15	53.36±5.15	55.73±5.66	6.48±1.52	6.60±1.40	①④⑥	c	Ashi, EX-LE2, EX-LE4, EX-LE5, SP10, ST35, ST36, GB34, GB33, SP9Y	5-10	Once daily	SAC	Acupuncture	30 min, once daily	10
Zhu WJ et al 201776	Dorsal myofasciitis		15/20	16/19	36.20±6.70	35.80±7.30	6.31±1.62	6.42±2.05	①④⑥	h	Ashi	10	Once every 2 d	SAC	Electro-acupuncture	30 min, once daily	16
Chen JJ 200977	Scapulohumeral periarthritis		16/14	15/13	52.00±1.56	53.00±1.25	4.63±1.42	4.63±1.42	①④⑥	b	Ashi	10	Once every 2 d	SAC	Electro-acupuncture	30 min, once daily	60
Cupping therapy plus other therapy vs other therapy
Malik S et al 202278	Chronic plantar fasciitis		0/15	0/15	22.2±1.85	22.8±2.18	7.67±0.90	7.07±0.96	①	a	Affected side	10	Thrice a week	SAC	Excercise	25-30 min, 3 times/ week	28
Han LL et al 202279	Breast carbuncle		0/32	0/32	36.75±4.87	36.91±4.49	7.00 (5.25,8.00)	6.50 (6.00,8.00)	①②④	b	Ashi	NA	Once every 5 days	SAC	Traditional Chinese medicine (TCM) formula	twice daily	14
Jiang Q, Chen Y 202280	Acute lumbago		27/20	25/22	42.09±6.35	42.18±6.44	6.10±0.72	6.03±0.68	①③④⑥	j	Du Meridian and Bladder Meridians	5-10	Once every 3 days	SAC	Celecoxib capsules; Yulong powder (External Applied)	100 mg, twice daily; once daily	21
Sun ZZ et al 202281	Cervical spondyosis		20/25	21/23	45.91±5.47	46.35±5.41	5.82±0.81	5.74±0.77	①⑥	d	Shoulder dorsal to lumbar	45	Once every 3 days	SAC	Massage	30 min	21
Study ID	Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Out-come measure-ment	Intervention group operation					Control group operation		Treatment duration
		T	C	T	C	T	C		Type of cup-ping	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
Liang NJ 202282	Cervical spondyosis	17/27	19/25	52.84±3.91	53.02±3.12	6.74±0.55	6.88±0.51	①④⑥	c	Cervical Jiaji	5-8	Once daily	SAC	Massage	20 min	12
Dong PP, Li FF 202283	Acute lumbar posterior joint disorder	46/49	43/52	42.6±5.9	42.3±5.7	5.94±0.55	5.87±0.58	①④⑥	f	Ashi	10-15	Once daily	SAC	Massage	Once daily	14
Wan FD et al 202284	Cervical spondyosis	17/13	16/14	46.87±5.34	46.43±5.74	5.67±1.56	5.61±1.63	①④⑥	c	Acupoints were selected based on meridian flow	8-10	Once daily	SAC	Massage	Once daily	14
Moura CC et al 202185	Chronic back pain	15/73	15/79	48.20±11.62	47.87±13.18	NR	NR	①⑥	a	Selection according to the region of pain	10	Once a week	SAC	Ear acupuncture	Once a week	42
Xiao BE et al 202186	Chronic non-specific low back pain	36/32	38/30	57.65±5.37	57.27±5.62	7.18±1.05	7.10±1.08	①⑥	j	Du Meridian and Bladder Meridians	5-10	Once every 3 d	SAC	Exercise	Core muscle group training (NA)	28
Chen JX et al 202187	Chronic non-specific lower back pain	27/21	26/22	63.50±3.40	64.20±3.40	6.33±1.72	6.29±1.66	①③⑥	b	Ashi	10	Once every 2 d	SAC	Exercise	McKenzie therapy once a day	10
Feng WH, Shao P 201788	Low back and leg pain	NR	NR	NR	NR	8.73±0.51	8.80±0.47	①⑥	c	DU3, EX-B2, BL25 point, and Ashi	8	Twice weekly	SAC	Exercise	Physical therapy (NA)	14
Chen JX et al 202089	Low back pain	22/26	25/23	46.10±3.20	42.30±3.50	7.73±1.21	7.78±1.23	①⑥	b	Ashi	10	Once daily	SAC	Exercise	McKenzie therapy, once a day	10
He YM et al 202090	Breast pain	0/40	0/40	35.60±3.40	35.80±3.20	6.17±1.08	5.96±1.15P	③④⑥	bj	LA14	5	Twice weekly	SAC	Xiaopi oral liquid	15-30 mL, 3 times daily	60
Wang J et al 201991	Shoulder pain	14/9	13/10	56.71±5.49	56.31±5.63	NR	NR	①	a	GB21, Ashi	8-10	Treatment only once	SAC	Massage	NA	2
Study ID	Disease type	Sample size (M/F)		Age (years, MD ± SD) or M (Q1,Q3)		Pain intensity at baseline		Outcome measure-ment	Intervention group operation					Control group operation		Treatment duration
		T	C	T	C	T	C		Type of cupping	Points of selection	Time of cupping retained (min)	Combination frequency	Combination on therapy	Control group	Usage
Chen X et al 201792	Scapulohumeral periarthritis	17/23	18/22	52.72±10.35	51.34±11.37	5.82± 0.71	6.13± 0.74	①④⑥	a	Ashi, SI11, SI9, and LI15 of the affected side	10	Once a week	SAC	Gua-sha therapy	10 min, once weekly	28
Ouyang Q et al 200193	Hemiplegic shoulder pain	18/8	22/8	58.20 (27.00-75.00)	56.80 (29.00-71.00)	6.37± 3.22	6.25± 3.01	①⑥	b	Ashi	10	Once every 2 d	SAC	Exercise	30 min once daily	30

Notes: outcome measurements: ① pain intensity ② adverse events ③ quality of life ④ effectiveness rate ⑤ sleep quality ⑥improvement of other clinical symptoms for specific disease (including osteoarthritis activity; waist function; lumbar spine mobility; neck function; cervical spine functional status; total clinical symptom score; cervical spine range of motion; wrist functional activity; expression levels of serum SP and NK-1; self- rating depression degree; quantitative score of symptoms and signs; knee function; interleukin-1β and interleukin-8 levels; degree of skin lesion recover; recurrence rate; traditional Chinese medicine syndrome score; inflammatory factor detection; back function activity; pain frequency; range of motion shoulder joint; degree of fatigue; blood flow velocity; static and dynamic muscle endurance; waist and leg function; spine function; muscle strength. a: dry cupping, b: wet cupping, c: medicinal cupping, d: moving cupping, e: flash cupping, f: needling cupping, g: retained cupping, h: bioceramic cupping, i: easy cupping, j: balance cupping. usual care 1^k: heat therapy: use a heated sterile towel to compress the painful part; usual care 1^l: heat therapy: hot water bottle applied to neck and upper trapezius for 10 min; usual care 1^m: heat therapy: heating pad applied for 15 min to shoulder areas bilaterally with participant in supine position; usual care 2: infrared thermal radiation therapy; usual care 3: 1) early return to usual activities encouraged, excluding heavy manual labor; 2) activity change to minimize symptoms; 3) acetaminophen or NSAIDs; 4) short-duration muscle relaxants or opioids; 5) bed rest, not more than 2 d; 6) spinal manipulation; Usual care 4: Progressive muscle relaxation: Participants asked to practice relaxation; Usual care 5: Participants continued self-directed standard medical care (physical therapy, exercise, analgesics) with general practitioner or orthopaedist. SAC: Same as control; NA: Not Applicable. Manual-acupuncture used in needling cupping: plum blossom needles or three-edged needles. Selection according to region of pain: The suctioncups were then applied bilaterally to the dorsal region on the acupoints GB21 (Gall Blader 21), GB30 (GallBlader 30), LR8 (Liver 8). Depending on the patient's paillocation, suction cups were also applied to the acupoints a) BL10 (Bladder 10) and BL11 (Bladder 11) in the cervical region, b) BL12 (Bladder 12), BL17 (Bladder 17), and BL21 (Bladder 21) in the thoracic region or c) BL23 (Bladder 23), BL24 (Bladder24), and BL25 (Bladder 25) in the lumbar region. NSAIDS: no nsteroidal antiinflammatory drugs.

Constitution of the type of pain condition was shown in Supplementary Table 2. The top 4 number of the condition involved in this review were herpes zoster (20 trials),^{34,⇓,⇓,⇓,⇓,⇓,⇓,⇓-42,48,51,⇓,⇓-54,57,⇓-59,67,⇓-69} low back pain/non-specific low back pain (14 trials, low back pain,^{26,27,63,70,⇓,⇓,⇓-74,80,91} non-specific low back pain,^23,28,86,87 neck pain (9 trials),^{24,29,⇓-31,60,62,81,82,84} and osteoarthritis (9 trials).^{22,47,55,56,64,65,75,77,92}

In summary, there are five types of comparisons among the 72 included trials, details of the treatment information were showed in Table 1. Eleven trials compared cupping therapy with wait-list (4 trials^{22,⇓,⇓-25})/no treatment or usual care (7 trials^{26,⇓,⇓,⇓,⇓,⇓-32}), usual care was defined as hyperthermia, infrared thermal radiation or relaxation exercise. Fifteen trials^{33,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-47} compared cupping therapy with drugs. Forty-six trials combined cupping therapy with other treatments. The other treatments involved acupuncture (16 trials),^{62,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-77} drugs (14 trials),^{48,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-61} exercise (6 trials),^{78,86,⇓,⇓-89,93} massage (5 trial),^{81,⇓,⇓-84,91} Gua-sha (1 trial),⁸⁹ herbal patent (Xiaopi oral liquid) (1 trial) ⁷¹ and self-prescribed herbal decoction (1 trial).⁷⁹

The drugs used in the control group mainly include analgesic, neurotrophic drugs or antidepressants (such as ibuprofen, aspirin, etc., Table 1). The acupuncture used in the control group mainly included manual-acupuncture (8 trials),^{64,⇓ 66,70,⇓,⇓,⇓,⇓-75} electro-acupuncture (4 trials),^62,69,76,77 warm needling (acupuncture with moxibustion, 2 trials)^63,65 and fire acupuncture (acupuncture treatment with a hot needle, 2 trials).^67,68

Sixty trials^{22,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-32,34,⇓,⇓-37,39,40,42,43,45,⇓-47,49,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-58,60,⇓,⇓,⇓,⇓-65,67,68,70,⇓,⇓,⇓,⇓,⇓,⇓,⇓-78,80,⇓-82,84,86,⇓,⇓ -89,91,⇓-93} reported the pain severity scores, including VAS or Numerical Rating Scale (NRS), and PPI (4 trials).^23,27,28,46 Forty-five trials^{25,⇓-27,33,⇓-35,37,⇓,⇓,⇓,⇓,⇓,⇓,⇓-45,47,⇓-49,51,52,54,55,58,59,61,64,⇓,⇓,⇓,⇓,⇓,⇓,⇓-72,75,⇓-77,79,82,84,86,87,89,90,92} reported the cure rate (Cure was defined as complete disappearance of symptoms). No trials reported tender point counts, patients-reported episodes of pain, or numbers of patients who had at least 30% or 50% of maximum pain relief over baseline. QoL measured by SF-36 was reported in 6 trials,^{22,24,29,30,54,70} and adverse events were described in 12 trials.^{33,38,43,45,46,48,50,51,55,60,63,65}

3.3. Methodological quality assessment

Majority of the included trials provided limited information about their study design and methodology. According to our pre-defined quality assessment criteria, all studies were with high risk of bias (Figure 2, supplementary Figure 1).

Figure 2. Risk of bias of included 72 Randomized Controlled Trials.

For randomization process, twenty-three trials^{27,35,38,42,44,⇓,⇓-47,54,55,58,62,66,68,69,72,77,78,85,87,89,91,92} were judged as “some concern” due to limit information, of the remaining 49 studies reported using random number table to generate the random allocation, only 4 studies^29,30,72,92 reported using sealed envelopes to implement allocation con-cealment, which were judged as “low risk of selective bias”.

For the intended interventions, none of the included trial reported blinding information, however, without placebo control they were impossible to blind the clinicians and participants, all of them were judged as “some concern” for this domain.

Only 3 study^78,85,86 reported methods of sample size calculation. For missing outcome data, twelve studies reported number of dropped out and loss of follow-up. Six of them^{23,24,28,⇓,⇓-31} reported the reason for drop out which was the time conflict or fear of adverse reactions of subjects, and they used last observation carried forward method to conduct intention-to-treat analysis. Three studies^46,64,84 reported the reasons for the drop out and only analyzed the data from per protocol set. Five studies^{34,41,58,86,89} neither mentioned the reason for data missing nor deal with the missing data with appropriate method, however, the number of drop out was less than 10% of the sample size and balanced between the groups. The remaining 50 studies did not report whether there was missing data during the treatment. Therefore, 12 trials were judged as “low risk of attrition bias” and 60 trials were judged as “some concern” due to insufficient information.

For measurement of outcome, all of the trials did not report whether they blinded the outcome assessors, nor the information of potential conflict of interest. Thus, they were all judged as “some concerns”.

For selection of the reported results, 7 trials registered their protocol and were implemented in accordance with their protocol. Though we did not find the protocol of the remaining trials, they all reported key outcomes of pain condition (pain severity or improvement of the symptoms). Therefore, all of the trials were judged as “low risk of bias” for this domain.

3.4 Effects of interventions VAS

Fifty-nine trials reported VAS scores before and after treatment, due to the obvious clinical and statistical heterogeneity, we did not conduct pooling analyses for most of them. The results of the individual trial were listed in supplementary Table 3. Due to the insufficient number of included trials in most of the Meta-analysis, the subgroup analysis was only conducted when cupping therapy was compared to no treatment/ waiting list. For the results of subgroup analysis that have been conducted, we only reported those that can explain the sources of heterogeneity.

Ten trials^{22,⇓,⇓,⇓,⇓-27,29,⇓,⇓-32} compared cupping therapy with waiting list/ no treatment or usual care. Overall, cupping therapy was better than no treatment or usual care in relieving pain intensity (VAS MD －1.75 cm, 95% CI －3.33 cm to －0.18 cm, I ² = 99%). Subgroup analysis was conducted to detect the source of heterogeneity. One trial²⁷ used needling cupping therapy as an intervention, which showed needling cupping was better than usual care in reducing VAS score (MD －0.17 cm, 95% CI －0.27 cm to －0.07 cm). Since the type of cupping treatments is different from other trials, thus we excluded it from the Meta-analysis. Its results showed that the remaining 9 trials were divided into two subgroups according to the type of disease. Among them, 1 trial²³ found a huge reduction of VAS scores in cupping group (MD －7.07 cm, 95% CI －7.45 cm to －6.69 cm, P < 0.00001, 61 participants) for acute pain, and the Meta-analysis results of the other 8 trials also showed that cupping therapy was better than control in pain relieving (VAS MD－1.26 cm, 95% CI －1.88 cm to －0.65 cm, I ² = 65%, P < 0.0001, 527 participants) for patients with chronic pain conditions (supplementary Figure 2).

Eleven trials compared cupping therapy to drugs. Among them, 7 trials^{34,35,37,39,40,45,47}showed that cupping therapy was better than drugs in reducing VAS score (average reduction 0.46 cm to 2.68 cm) for patients with herpes zoster or osteoarthritis, 3 trials^36,42,46 showed that there was no significant difference between cupping therapy and drugs in reducing VAS score for patients with myofascitis and herpes zoster associated disease, the remaining one study⁴³ showed that drugs was superior to cupping therapy on this outcome (MD 0.60 cm, 95% CI 0.08 cm to 1.12 cm) for migraine patients. This study⁴³ evaluated the immediate analgesic effect, and the control drug was diclofenac sodium (analgesic), while other included studies evaluated the degree of pain relief at the end of the treatment duration.

Twelve trials^{49,⇓,⇓,⇓,⇓,⇓,⇓,⇓,⇓-58,60,61} compared cupping therapy plus drugs to drugs alone. All of them suggested that combination therapy was better in reducing VAS score (average MD reduction from 0.99 cm to 2.99 cm). Subgroup analysis was attempt but failed to interpret the source of heterogeneity, results of individual studies were shown in supplemantary Table 3 and Figure 3.

Figure 3. Funnel plot of 6 types comparisons.

A: cupping vs no treatment about VAS; B: cupping vs drugs about cure rate; C: cupping vs drugs about VAS; D: cupping plus acupuncture vs acupuncture about cure rate; E: cupping plus acupuncture vs acupuncture about VAS; F: cupping plus other therapy vs other therapy about VAS; VAS: visual analog scale; RR: risk ratio; MD: mean difference.

Fourteen trials^{62,⇓,⇓-65,67-68,70,⇓,⇓,⇓,⇓,⇓,⇓-77} compared cupping therapy plus acupuncture to acupuncture alone. All trials suggested that combination therapy was superior to acupuncture alone in pain relief (average MD reduction 0.48 cm to 2.97 cm, supplemantary Table 3).

Thirteen trials^{78,80,⇓-82,84,⇓,⇓,⇓,⇓-89,91,⇓-93}compared cupping therapy plus other therapies (exercise, Gua-sha, massage, or herbal medicine) to other therapies alone. Eleven of them found better effect of combination therapy in pain relief (average MD reduction 0.25 cm to 2.73 cm), however, 2 trial^91,93 found no difference between groups for patients with myofascitis and herpes zoster associated disease.

3.5 Painful pressure index (PPI)

Three trials compared cupping therapy to waiting-list. One trial²⁷ showed no difference between groups in reducing PPI score (MD －0.02, 95% CI －0.13 to 0.09), another 2 trials^23,28 showed cupping therapy was better than waiting-list in terms of PPI score reduction (supplementary Table 6).

One trial⁴⁶ compared cupping therapy to drugs, which showed no difference between groups in reducing PPI score (MD 0.03, 95% CI －0.18 to 0.24, P = 0.78).

3.6. Japanese orthopaedic association (JOA)

Four trials^59-62 compared combination of cupping therapy plus acupuncture to acupuncture alone and reported JOA scores to assess the functional activity (involved pain intensity) for patients with low back pain. The results of the individual trials suggested that combination therapy had better effect on improving the JOA scores (average MD promotion 3.10 scores to 8.47 scores) (supplementary Table 6).

3.7. Cure rate

In the 5 types of comparison, Forty-four included trials reported the cure rate after treatment (supplementary Table 4). Five Meta-analyses showed significant better effect in cupping therapy group on increasing the number of cured participates than control group (average RR 1.38-2.20, supplementary Table 3 and supplementary Figure 3).

3.8. Quality of life

Four studies^22,24,29,30 reported quality of life through SF-36 Physical Component Scale and SF-36 Mental Component Scale, the results showed that compared to waiting list/no treatment or usual care, cupping therapy may increase the scores of SF-36 Physical Component score (MD 4.78 scores, 95% CI 2.60 scores to 6.96 scores, I ² = 0%, P < 0.0001, 4 trials, 187 participants).

However, no difference was found between groups in SF-36 Mental Component score (MD 2.57 scores, 95% CI －1.70 scores to 6.84 scores, I ² = 48%, P = 0.24).

One study⁵⁴ reported the quality of life through the SF-36 health survey scale.⁹⁶ The result showed that cupping therapy plus drugs had higher general emotion index, emotion score, life satisfaction and health index than drugs alone (average MD 2.52 scores to 12.53 scores, 130 participants).

One study⁷⁰ assessed the improvement of quality of life from three dimensions, including daily self-care ability, social ability and daily life ability. The results showed that patients in the combination of cupping therapy and acupuncture group increased more scores than those in acupuncture alone group (average MD 9.91 scores to10.94 scores, 300 participants).

3.9. Sleep quality

Two trials^34,36 reported sleep quality scores (0-10 scores, higher means better), the pooling results showed cupping therapy had a better effect on improving the sleeping quality than drugs (MD 2.69 scores, 95% CI 2.29 scores to 3.09 scores, P < 0.00001, I ² = 0%, 134 participants). Other 2 studies^67,68 showed cupping therapy plus acupuncture may increase the Pittsburgh sleep quality index (PSQI) scores more compared to acupuncture alone (MD 23.08 scores, 95% CI 19.31 scores to 26.84 scores, I ² = 0%, P < 0.00001, 119 participants).

3.10. Adverse events

Only 12 studies^{33,38,43,45,46,48,50,51,55,60,63,65}mentioned adverse events, in which only 2 trials reported the adverse events in cupping therapy group. One of them⁴³ reported two cases of dizziness after wet cupping therapy, and anotherone⁵⁵ reported 2 cases of skin damage (details unreported) in combination group. The remaining adverse events occurred in the control (drugs) group, including skin damage and gastrointestinal discomfort caused by oral administration of febristat tablets, and intestinal adverse reactions (such as abdominal distension and stomachache) caused by oral administration of losoprofen sodium tablets).

The pooling results showed cupping therapy may decrease the incidence of adverse events compared to drugs (RR 0.11, 95% CI 0.03 to 0.39, I ² = 24%, P = 0.0006, 344 participants), however, no difference was found between cupping therapy plus drugs and drugs on incidence rate of adverse events (RR 0.45, 95% CI 0.15 to 1.37, I ² = 21%, P = 0.16, 328 participants).

3.11. Other outcomes

One trial³¹ reported the Measure Yourself Medical Outcome Profile (MYMOP2), EuroQolfive-dimension scale (EQ-5D), Short Form Stress Response Inventory (SRI-SF) and Fatigue Severity Scale (FSS), which found only for decreasing MYMOP2 score cupping therapy had a better effect than wait-list/no treatment or usual care (MD －1.00, 95% CI －1.66 to －0.34, 40 participants). No difference was found on changing of EQ-5D (MD 0.09, 95% CI －0.03 to 0.21, P = 0.15), SRI-SF (MD －4.50, 95% CI －11.04 to 2.04, P = 0.18) or FSS (MD －0.69, 95% CI －1.45 to 0.07, P = 0.08) between groups (supplementary Table 6).

3.12. Publication bias

Six funnel plot analyses were conducted (Figure 3), all of them showed asymmetry and suggested potential publication bias.

3.13. Overall quality of evidence by GRADE

Fourteen Meta-analyses were conducted in this review, concerning VAS score, cure rate, Quality of life, sleep quality, and incidence of adverse events. We listed all of them in one summary of the finding table (supplementary Table 5). All of the evidence was evaluated as low-quality/very low-quality due to the potential selection bias, detection bias of the included studies, the inconsistency of the results among trials, and potential publication bias.

4. DISCUSSION

4.1. Summary of main results

Totally 72 trials involving 5720 participants were included in this review, and all of the trials may have high risk of bias. Thus, only low/very low quality evidence was found in this review showed that cupping therapy alone or combined with other treatments may have better results than no treatment, usual care, drugs, acupuncture or other treatments on decreasing VAS scores, increasing cure rate, improving quality of life, quality of sleep or other symptoms related to pain condition. Compared with the control group, cupping therapy alone or in combination with other therapies can reduce the VAS scores by 0.16-2.97 cm and improve the cure rate by at least 20%, while the incidence of adverse events is only one tenth of that of the control group. So we can explain the mechanism of cupping analgesia in some degree. Although the heterogeneity between studies and the methodological quality of the study itself lead to the low evidence strength of the current conclusions, the results of this study are a valuable supplement to the founding of previous review.

4.2. Differences and similarities with previous studies

Comparing to the previous review, the number of studies included in this review increased from 14 to 72, and the number of participants increased from 921 to 5720. The above changes make us more confident that cupping therapy has a better effect on pain relief, although the inconsistency of clinical characteristics among included trials makes Meta-analysis difficult to conduct, we can try more subgroup analyses to explore the source of heterogeneity and gain some valuable information.

Besides, three studies^97,⇓-99 published in last five years also evaluated the efficacy and safety of cupping for pain relief. Their main findings suggest that cupping therapy may have advantages in relieving chronic pain (e.g. low back pain, neck pain, cancer pain, trigeminal neuralgia, brachialgia) compared to usual care, heat pad or analgesics, however, the conclusions are not certain due to poor quality of the evidence.

None of them systematically searched and included all the relevant Chinese literatures, the types of involved diseases and type of comparisons were limited. This review comprehensively includes the trials published in Chinese and English in topic of cupping for pain management, involving 15 diseases for both acute and chronic pain condition. Therapeutic effect of cupping therapy as single treatment or adjunctive treatment are all evaluated in this review, and summarizes the characteristic of the included clinical research in more detailed. Therefore, this review provides a most comprehensive and up to date evidence for the research question, thus, is able to provide more information for clinical practice.

4.3. Strengths and limitations of this study

As mentioned above, as an updated systematic review, this study included the clinical research evidence of cupping therapy for pain published in recent 8 years and adopted the latest methodological quality evaluation tools and evidence grading methods. The implementation of the study was carried out in strict accordance with the implementation standards of systematic review, in order to provide the current best evidence for clinical practice.

However, due to the high risk of bias in the original trial, the level of evidence in this study is low, which affects the reliability of the results. At the same time, the non-English clinical study results published by some other countries that commonly use cupping therapy for pain management (such as South Korea, Japan, etc.) are not included in this study due to language restrictions. Most of the included studies come from China, and the results may be affected by the contextual effects.¹⁰⁰

4.4. Implications for the clinical practice

Although the evidence quality of this study was low, we can provide some suggestions for clinical pain management practice. The types of cupping therapy in this review involved dry cupping, wet cupping, medicinal cupping, moving cupping, flashing cupping, needling cupping, retained cupping, bioceramic cupping, easy cupping and balance cupping. As the most commonly used type of cupping therapy, wet cupping therapy (56.25%) was generally used in treatment of herpes zoster (29.82%), chronic neck pain (10.91%) and non-specific low back pain (7.27%). For herpes zoster, the needles for pricking were mostly plum blossom needles or three-edged needles, cups were mostly retained in the herpes area for 8-15 min and the endurance of treatment time was mostly 7-15 d. Interestingly, in the treatment of herpes zoster by wet cupping therapy in this study, we found that the greater the bleeding amount (10-20 mL), the greatly the better the effect (the reduction of pain severity scores).

Subgroup analysis showed compared to chronic pain, cupping therapy may have even better effect of pain relieving for acute pain, and cupping therapy may have advantages in the treatment of herpes zoster and osteoarthritis compared to drugs.

The incidence of adverse events of cupping therapy seems lower than control therapy, mainly manifested as dizziness after wet cupping therapy, which may cause by bloodletting. However, some patients may not like the cupping marks, it is necessary to explain the details to patients in details before treatment.

4.5. Implications for future research

From the perspective of future clinical practice, different treatment standards can be formulated for different diseases treated by cupping therapy, including the selection of cupping equipment, cupping therapy retention time, point selection standards for different diseases, etc., so that cupping therapy can serve the clinic more finely. In addition, through clinical observation, it is found that cupping therapy is also used to treat weight loss, constipation and other diseases. In the future, more in-depth research on these diseases can be carried out to provide more evidence for further clinical application of cupping therapy.

Although cupping therapy is easy to operate and easy to be accepted by patients, the duration of cupping treatment also needs to be further combined with the actual situation (for example, it may be affected by seasons). Clinical research needs to consider the compliance of patients to a large extent.

From the perspective of methodology for future clinical trials, adequate sample sizes should be recruited. Researchers should use appropriate randomization, blinding, and statistical methods. Cupping therapy intervention methods should be described according to the STRICTOC,¹⁰¹ the components of complex interventions should be clearly reported. The control interventions should be designed as either no treatment, standard treatment, or usual care as applied, and the use of drugs should be based on current evidence for their practice for acute or chronic pain. Trials should be registered prospectively.

In conclusion, this review involved 72 trials with 5720 participates. We would like to show that cupping therapy or combined cupping therapy can be considered to treat patients with acute and chronic pain in relieving pain, improving the quality of life and sleep, and improving the cure rate, even only supported by low quality of evidence. There was low/very low quality evidence that the incidence of adverse events in the cupping groups were lower than that in the control groups. We also look forward to the future high quality trials with large sample size and rigorous methodological quality to provide even more confirmatory evidence for cupping therapy.

5. SUPPORTING INFORMATION

Supporting data to this article can be found online at http://journaltcm.com.

Funding Statement

Supported by National Natural Science Foundation of China Project: Research on the Correlation between Traditional Chinese Medicine Constitution, Cupping Marks Color, and Clinical Efficacy Based on Regression Models (No. 81804000)