Moxibustion for depression

Abstract

Background:

Moxibustion, a traditional Chinese medicine therapy, is widely used for various diseases and has recently been applied in treating depression. This study aims to systematically assess the efficacy and safety of moxibustion, either alone or in combination with other therapies, for depression.

Methods:

We conducted a comprehensive search across 7 databases from their inception to December 20, 2023, focusing on randomized controlled trials that compared moxibustion with sham treatments, no treatment, or conventional therapies. The primary outcomes were the overall effective rate and Hamilton Depression Rating Scale scores, with adverse reaction incidence as the secondary outcome. Data were analyzed with RevMan 5.4 and Stata 18, with bias risk assessed using the Cochrane Risk of Bias tool. Evidence quality was graded with GRADEpro.

Results:

This study included 15 randomized controlled trials with 1510 patients. Meta-analysis results showed that moxibustion, alone or combined with other therapies, was more effective than other treatments for managing depression (relative risk = 1.07, 95% confidence interval [CI]: [1.03–1.10], P = .0003). Moxibustion, either alone or in combination with other treatments, significantly reduced Hamilton Depression Rating Scale scores compared to other treatments used alone (mean difference = −2.10, 95% CI: [−3.00 to −1.20]), with its efficacy comparable to that of antidepressant medications (mean difference = −0.60, 95% CI [−1.31 to 0.12], P = .10). Additionally, in terms of safety, the incidence of adverse reactions associated with moxibustion alone or in combination with other therapies for depression is relatively low (relative risk = 0.57, 95% CI: [0.39–0.84], P = .005). However, the overall quality of the evidence was assessed as low to very low.

Conclusion:

Moxibustion has been shown to alleviate depressive symptoms, particularly when combined with other therapies, offering both enhanced efficacy and safety. However, further high-quality studies are necessary to confirm its effectiveness.

1. Introduction

Depression is a common mental disorder, which can be long-lasting or recurrent, substantially impairing an individual’s daily functioning.^[1] Clinical depression is characterized by behavioral, cognitive, and emotional features and it is recognized as a major public health problem that has a substantial impact on individuals and on society.^[2] Consequently, study of the prevention and treatment of depression is a of paramount importance. In primary care, guidelines suggest the main treatment for depression is antidepressants,^[3] most patients with depression will be offered Selective serotonin reuptake inhibitor (SSRI) or Serotonin norepinephrine reuptake inhibitor antidepressants as a first line treatment.^[4] However, their efficacy remains limited, with recovery rates between 36% and 53% and recurrence rates ranging from 15% to 50%.^[5,6] Furthermore, these medications frequently cause side effects, including nausea, vomiting, constipation, lactation and sexual dysfunction.^[7] As a result, the relatively safe nondrug treatment measures, is increasingly favored by patients with depression.

The Yin and Yang is an important theory in traditional Chinese medicine (TCM), which posits that the state of health is characterized by a dynamic equilibrium between these 2 forces.^[8] Depression is understood as a result of this imbalance, specifically arising from a deficiency of Yang energy and an excess of Yin energy.^[9]

Moxibustion is a common treatment in TCM, utilized for millennia in China and other Asian countries. The mechanism of moxibustion involves burning mugwort on or above the skin at specific acupoints.^[10,11] This process generates a combination of effects – heat, radiation, moxa smoke, and moxa oil – which are absorbed through the skin and respiratory tract.^[12,13] Moxibustion is believed to alleviate depression by restoring the balance of yin and yang through these combined effects.^[14]

Therefore, an increasing number of patients with depression are seeking moxibustion and/or acupuncture as alternative or complementary therapies.^[15] However, due to many factors, such as the cultural theory of TCM, different terms, language barriers and so on, there are few clinical studies in this field, especially the studies published in English. According to the search results, this study is the first meta-analysis of clinical randomized controlled trials (RCTs) research on moxibustion for depression. Therefore, this study is expected to provide relatively comprehensive evidence for moxibustion treatment of depression and provide guidance for clinicians through the meta-analysis of moxibustion treatment of depression. Hoping to provide a safe and effective non-drug treatment for patients with depression.

2. Materials and methods

2.1. Study registration and study design

This protocol was registered on the PROSPERO International Prospective Register of Systematic Reviews (CRD42020190599). The present meta-analysis was reported according to the guidelines set forth in the Preferred Reporting Items for Systematic Reviews and Meta-analyses,^[16] along with the Cochrane Handbook for Systematic Reviews of Interventions.^[17]

2.2. Criteria for considering studies

2.2.1. Types of studies

We included RCTs that involved participants of all genders, ages, and ethnicities.

2.2.2. Inclusion criteria

We included participants with a primary diagnosis of clinical depression, excluding those where depression was a secondary condition or comorbidity. Depression was diagnosed via 1 or more of the following criteria: the Diagnostic and Statistical Manual (DSM) editions (DSM-III, DSM-IV, or DSM-V),^[18-20] the International Classification of Diseases-10,^[21] or the Criteria for Classification and Diagnosis of Mental Disorders (CCMD) editions (CCMD-2, or CCMD-3).^[22,23]

2.2.3. Types of interventions

Studies were eligible if participants in the experimental group received moxibustion either as a standalone treatment or in combination with other therapies. Moxibustion was considered an intervention in which burning moxa was applied to selected acupoints or designated body regions, including both direct and indirect approaches using moxa cones or sticks.^[24] Studies involving moxa-needle therapy were not included because this technique combines acupuncture with moxibustion, making it difficult to distinguish the therapeutic contribution of moxibustion itself from that of acupuncture.

Eligible comparators included sham or placebo moxibustion, antidepressant medications, acupuncture, electroacupuncture, and acupoint sticking therapy.

2.2.4. Types of outcome measures

2.2.4.1. Primary outcomes

All included studies must have pretreatment and posttreatment assessments using the Hamilton Depression Rating Scale (HAMD). The main observations were: total effective rates and scores of the HAMD before and after the treatment.

2.2.4.2. Secondary outcomes

The secondary outcome measures were treatment side effects, which mainly included adverse reactions of the nervous system: headache, dizziness, and sleep disorders. Adverse reactions of the digestive system included: nausea, vomiting, diarrhea, abdominal distension, and gastrointestinal discomfort. Adverse reactions of skin attachments included: rashes, itching, blisters, and burns.

2.3. Exclusion criteria

Non-published articles, animal experiments, reviews, and duplicate reports; articles with a total sample size of ≤30; and articles for which data could not be obtained from the full text or by contacting the author were excluded.

2.4. Search strategy

We conducted a comprehensive search across English databases, including PubMed, Web of Science, Cochrane Library, and Embase, as well as Chinese databases such as China National Knowledge Infrastructure, Wanfang Data Knowledge Service Platform (Wanfang Data), and Chinese Scientific Journal Database (VIP), from their inception to December 20, 2023. The search targeted RCTs utilizing various forms of moxibustion for depression, without restrictions on countries or article types. We used the keywords (“moxibustion”) and (“Depression*” or “Depressive Symptom*” or “Symptom*, Depressive” or “Emotional Depression*” or “Depression*, Emotional”) AND (“random*” or “randomized controlled” or “clinical trials, randomized” or “trials, randomized clinical” or “controlled clinical trials” or “RCT”) using the Boolean “AND/OR” operators. Both MESH and Non-MESH terms were included in this search. Papers that either had English full text, or Chinese full text available were eligible. Reference lists of full text papers were searched, and any relevant articles identified were screened. Detailed search strategies are provided in Appendix 1, Supplemental Digital Content 1.

2.5. Data collection and analysis

2.5.1. Selection of studies

After completion of the literature search, all retrieved records were imported into EndNote 21 and duplicate citations were removed. Two reviewers independently screened the titles and abstracts to identify potentially eligible studies. Articles that clearly failed to meet the predefined eligibility criteria were excluded at this stage. The full texts of potentially relevant studies were subsequently assessed against the inclusion and exclusion criteria. Any disagreement between the 2 reviewers was resolved through discussion and consensus. When eligibility could not be determined because of incomplete information, the corresponding authors were contacted for clarification when necessary.

2.5.2. Data extraction and management

Two authors extracted the data independently, and a third author resolved any disagreement. In cases of missing or unclear data, the study authors were contacted via email twice over a 4-week period to request additional information. If no response was received in that time, the data were marked as missing. Data were extracted on all of the following outcomes: main author, publication year, country, intervention type, treatment duration, treatment frequency, moxibustion methods, acupoints, outcome measures, and adverse reactions.

2.5.3. Assessment of risk of bias in included studies

The methodological quality of the included RCTs was evaluated using the Cochrane risk-of-bias tool. The Cochrane Handbook for Systematic Reviews includes 7 components for bias assessment: random sequence generation (selection bias); allocation concealment (selection bias); blinding of participants and personnel (performance bias); blinding of outcome assessment (detection bias); incomplete outcome data (attrition bias); selective reporting (reporting bias); other bias. Two reviewers performed the assessment independently and classified each domain as having a low, high, or unclear risk of bias. Differences in judgment were resolved through discussion, with consultation of a third reviewer when required.

The certainty of evidence for individual outcomes was further evaluated using the GRADEpro software, taking into consideration risk of bias, inconsistency, indirectness, imprecision, and publication bias. Evidence certainty was categorized as high, moderate, low, or very low.

2.5.4. Statistical analysis

Statistical analyses were conducted using RevMan 5.4 software (The Cochrane Collaboration) and Stata 18 (StataCorp LLC). Dichotomous outcomes were measured by relative risk (RR) with 95% confidence intervals (CI), while continuous outcomes were assessed using the mean difference (MD) between treatment groups, also with 95% CI. We analyzed outcomes based on randomized participants. In the case of multiple intervention groups within a study, we performed pairwise comparisons relevant to the study objective. Cochran’s Q testing and the I² statistic was used to evaluate the statistical heterogeneity of all studies. If heterogeneity is not significant (P > .1, I² < 50%), a fixed-effect model will be used for meta-analysis. If heterogeneity is present (P < .1, I² > 50%), the sources of heterogeneity will be explored, and subgroup analyses will be conducted based on the interventions. If heterogeneity is substantial, individual studies will be excluded to perform a sensitivity analysis and assess the stability of the results. For outcomes with 10 or more included studies, funnel plots and Egger’s tests were employed to assess publication bias.

2.5.5. Dealing with missing data

When outcome information was incomplete, the potential influence of missing data was considered during the risk-of-bias assessment and interpretation of the pooled findings. Missing outcome values were not imputed.

3. Results

3.1. Results of the search

Our searches yielded 2739 records. After removing duplicates, we screened the titles or abstracts of 1849 records. We read the full texts of 38 records and finally included 15 RCTs in the review.^[25-39] See Figure 1.

Figure 1.

The flowchart of search results for meta-analysis. CNKI = China National Knowledge Infrastructure, VIP = Chinese Science and Technology Journal Database, WOS = Web of Science.

3.2. Included studies

A total of 15 studies, including 1510 participants, were included in this review. The sample size of these studies ranged from 50 to 460 participants. All studies were reported in Chinese and were conducted in Chinese hospitals. Overall, 459 (30.4%) of participants in these studies were men. One study only reported the experimental participant gender.^[29] Two studies did not report the participant gender.^[35,37] Four studies compared moxibustion versus antidepressant.^{[27,29,30,33]} Five compared moxibustion plus SSRI versus SSRI,^{[26,28,34,35,38]} 1 compared moxibustion versus acupuncture,^[37] 1 compared moxibustion plus acupuncture versus acupuncture,^[32] 1 compared moxibustion plus ear auricular acupressure versus ear auricular acupressure,^[25] 1 compared moxibustion plus acupuncture plus SSRI versus acupuncture plus SSRI,^[36] 1 compared moxibustion versus electroacupuncture,^[31] and 1 compared moxibustion plus electroacupuncture versus electroacupuncture.^[39] The moxibustion methods mainly included: suspended moxibustion,^{[27,29,30,36,37]} direct moxibustion,^[32] indirect moxibustion,^{[26,28,31,34,38,39]} and apparatus-type moxibustion.^[25,33,35] See Table 1 for additional details.

Table 1.

Summary of included studies.

Study ID	Country	Sample Size	Diagnosis for inclusion*	Intervention†	Control†	Treatment duration (wk)	Frequency of treatment	Moxibustion methods‡	Acupointsδ	Outcome measures	Adverse reactions‖
Chen and Xie 2013[26]	China	50	ICD-10	a + b1	b1	8	45 min/treatment, 5 times/wk	A	1,2,3,4,5,6	HAMD	NR
Deng et al 2011[27]	China	158	CCMD-3	a	b2	4	30 min/treatment, 4 times/wk	B	3,7,8,9,10,11,12,13	HAMD	NR
Li et al 2023[28]	China	122	DSM-IV	a + b4	b4	8	90 min/treatment, 1 times/wk	A	1	HAMD	NR
Liu 2006[29]	China	460	CCMD-3	a	b3	4	30 min/treatment, 4 times/wk	B	7	HAMD	NR
Pan et al 2017[30]	China	81	CCMD-3	a	b4	8	30 min/treatment, 5 times/wk	B	7	HAMD	①②
Song 2006[31]	China	60	CCMD-3	a	e	6	15 min/treatment, 5 times/wk	A	7,8,10,11	HAMD	NR
Wang 2009[32]	China	58	CCMD-3	a + c	c	12	75 min/treatment, 2 times/wk	C	15,16	HAMD	③
Wang et al 2015[33]	China	60	CCMD-3	a	b5	6	30 min/treatment, 3 times/wk	D	2,3,14	HAMD	①②
Xie 2014[25]	China	71	CCMD-3, DSM-IV	a + d	d	6	30 min/treatment, 4 times/wk	D	7,11,17	HAMD	NR
Xie et al 2016[34]	China	80	ICD-10	a + b1	b1	12	30 min/treatment, 5 times/wk	A	1,2,3,4,5,6	HAMD	②
Xie and Liu 2021[35]	China	80	ICD-10	a + b8	b8	4	15 min/treatment, 5 times/wk	D	3,4,6,7,8,24	HAMD	②
Xu 2012[36]	China	60	CCMD-3	a + c + b6	c + b6	6	15 min/treatment, 5 times/wk	B	1,7	HAMD	NR
Xu et al 2015[37]	China	80	CCMD-3	a	c	4.3	15 min/treatment, 6 times/wk	B	1,7,18,19,20,21	HAMD	NR
Zhang et al 2019[38]	China	72	CCMD-3	a + b6	b6	6	50 min/treatment, 7 times/wk	A	22	HAMD	NR
Zhang 2021[39]	China	59	CCMD-3, DSM-IV	a + e	e	12.9	30 min/treatment, 2 times/wk	A	22	HAMD	NR

CCMD-3 = Criteria for Classification and Diagnosis of Mental Diseases, HAMD = Hamilton Depression Rating Scale, ICD-10 = International Classification of Diseases-10.

^*DSM-IV = Diagnostic and Statistical Manual.

^†a = moxibustion; b₁ = escitalopram oxalate, b₂ = venlafaxine, b₃ = amitriptyline, b₄ = fluoxetine, b₅ = mirtazapine, b₆ = selective serotonin reuptake inhibitions (SSRIs), b7 = sertraline, b8 = paxil; c = acupuncture; d = ear auricular acupressure; e = electroacupuncture. ‡ A = indirect moxibustion, B = suspended moxibustion, C = direct moxibustion, D = apparatus-type moxibustion.

^δ1 = DU14 dazhui, 2 = BL15 xinyu, 3 = BL18 ganyu, 4 = BL20 piyu, 5 = BL13 feiyu, 6 = BL23 shenyu, 7 = DU20 baihui, 8 = SP6 sanyinjiao, 9 = EX-HN1 sishencong, 10 = HT7 shenmen, 11 = ST36 zusanli, 12 = PC6 neiguan, 13 = LR3 taichong, 14 = RN17 danzhong, 15 = RN12 zhongwan, 16 = ST25 tianshu, 17 = LA14 qimen, 18 = EX-HN3 yintang, 19 = DU16 fengfu, 20 = DU24 shenting, 21 = KI1 yongquan, 22 = RN8 Shenque, 23 = LI4 Hegu, 24 = RN4 Guanyuan.

^‖①= Adverse reactions of the nervous system: headache, dizziness, sleep disorders; ②= Adverse reactions of the digestive system: Constipation, dry mouth, nausea, vomiting, diarrhea, abdominal distension, and gastrointestinal discomfort; ③ = Adverse reactions of skin attachments: rashes, itching, blisters, and burns; ④NR = Not Reported.

3.3. Excluded studies

The main reasons for excluding studies were duplicate publication,^[40-43] lack of a suitable control group,^[44-51 lack of suitable diagnostic criteria or relevant outcome,^[52-59] and study protocols (waiting further information).^[60-62]

3.4. Risk of bias in included studies

The risk of bias graph and summary are shown in Figures 2 and 3.

Figure 2.

Risk of bias summary.

Figure 3.

Risk of bias graph.

3.4.1. Allocation (selection bias)

Eight studies reported using a random number table or computer program,^{[25,28,32,34,36-39]} which were assessed as having a low risk of bias. Two research reports used a block randomization method,^[31,35] also evaluated as low risk. Three studies reported random allocation without providing sufficient methodological details and were therefore judged to have an unclear risk of bias.^[27,29,33] Two studies used the order of patient visits for grouping, considered to have a high risk of bias.^[26,30] Additionally, 3 studies reported adequate procedures for allocation concealment,^[25,28,32] while the remaining studies did not mention this aspect and were assessed as having an unclear risk of bias.

3.4.2. Blinding (performance bias and detection bias)

Two studies reported undertaking blinding of participants.^[25,35] Other studies did not mention blinding, so we assessed the performance bias in these studies as either unknown risk or high risk. Three studies applied the separation principle of investigator, operator, and statistician in treatment, data collection, and statistical analysis.^[25,28,35] Other studies did not report specific procedures in treatment, data collection, and statistical analysis of data.

3.4.3. Incomplete outcome data (attrition bias)

Of the included studies, all but 5 reported complete data.^{[25,28,32,33,39]} In 1 study, data were missing for 1/30 participants in both groups.^[32] Another study indicated missing data for 3/40 participants in the treatment group and 6/40 in the control group.^[25] Additionally, 1 study was missing data for 1/30 participants in the control groups.^[39] Furthermore, in another study, data were absent for 2/42 participants in the treatment group and 1/42 in the control group and placebo group.^[28] The number of missing participants in both groups was approximately the same, which is unlikely to have a significant impact on the results. However, in 1 study, data were missing for 1/10 participants in the treatment group and 1/3 in the control group,^[33] which we classified as high risk.

3.4.4. Selective reporting (reporting bias)

These studies provided comprehensive evidence by reporting results for all planned analyses, thereby avoiding selective reporting. Consequently, we consider the risk of selective reporting bias to be low.

3.4.5. Other potential sources of bias

We assessed the overall risk of bias as unclear in all studies.

3.5. Meta-analysis results

3.5.1. Effective rate

Fifteen trials reported effective rates, involving a total of 1510 patients. The results of the heterogeneity test revealed no significant heterogeneity in the data (P = .14, I² = 29%); therefore, the fixed-effects model was used to compute the combined effect size for meta-analysis. The results indicated that moxibustion or its combination with other treatments yields superior therapeutic outcomes compared to the use of other treatments alone (RR = 1.07, 95% CI: [1.03–1.10], P = .0003), as shown in Figure 4.

Figure 4.

Meta-analysis forest plot of the effective rate in depressed patients: moxibustion vs. control group. CI = confidence interval.

3.5.2. HAMD score

Fifteen trials reported HAMD scores before and after treatment. Moxibustion, whether used alone or in combination with other treatments, significantly reduced HAMD scores compared to other treatments used alone (MD = −2.10, 95% CI: [−3.00 to −1.20], P < .00001). However, the results of heterogeneity testing indicated high data heterogeneity (P < .00001, I² = 87%). Subgroup analyses will be conducted based on different intervention measures, as shown in Figure 5.

Figure 5.

Meta-analysis forest plot of HAMD scores in depressed patients: moxibustion vs. control group. CI = confidence interval, HAMD = Hamilton Depression Rating Scale, SD = standard deviation.

3.5.2.1. Moxibustion versus antidepressants

Four studies compared moxibustion with antidepressants, and heterogeneity testing revealed no significant heterogeneity among the studies (P = .55, I² = 0%), leading to the use of a fixed-effects model. The results indicated that there was no statistically significant difference between moxibustion and antidepressants in improving depressive symptoms (MD = −0.60, 95% CI: [−1.31 to 0.12], P = .10), as shown in Figure 6.

Figure 6.

Subgroup meta-analysis forest plot of HAMD scores: moxibustion group vs. control group in depression. CI = confidence interval, HAMD = Hamilton Depression Rating Scale, SD = standard deviation.

3.5.2.2. Moxibustion plus antidepressants versus antidepressants

Five studies involved treatment with moxibustion combined with antidepressants. Given the high heterogeneity observed in the data (P < .00001, I² = 94%), a random-effects model was applied for the meta-analysis. The results suggested that moxibustion combined with antidepressants may be more effective in treating depression than antidepressants alone (MD = −3.17, 95% CI: [−5.43 to −0.92], P = .006), as shown in Figure 7.

Figure 7.

Subgroup meta-analysis forest plot of HAMD scores: moxibustion group vs. control group in depression. CI = confidence interval, HAMD = Hamilton Depression Rating Scale, SD = standard deviation.

3.5.2.3. Moxibustion versus acupuncture

Only 1 study compared moxibustion to acupuncture.^[37] This study showed that the HAMD scores of the moxibustion group were significantly lower than those of the acupuncture group in patients with depression (MD = −6.07, 95% CI: [−9.24 to −2.90], P = .0002).

3.5.2.4. Moxibustion plus acupuncture versus acupuncture

Only 1 study compared moxibustion plus acupuncture to acupuncture.^[32] This study showed that the HAMD scores of the moxibustion plus acupuncture group were significantly lower than those of the acupuncture group in patients with depression (MD = −2.47, 95% CI: [−3.72 to −1.22], P = .0001).

3.5.2.5. Moxibustion plus ear auricular acupressure versus ear auricular acupressure

Only 1 study, comparing moxibustion plus ear auricular acupressure to ear auricular acupressure alone,^[25] found no significant difference in HAMD scores among patients with depression (MD = −0.62, 95% CI: [−1.45 to 0.21], P = .14).

3.5.2.6. Moxibustion plus acupuncture plus SSRI versus acupuncture plus SSRI

One study compared moxibustion plus acupuncture plus SSRI to acupuncture plus SSRI,^[36] showing significantly lower HAMD scores in the former group for depressed patients (MD = −2.68, 95% CI: [−4.52 to −0.84], P = .004).

3.5.2.7. Moxibustion versus electroacupuncture

Only 1 study compared moxibustion to electroacupuncture.^[31] There was no significant difference in HAMD scores in patients with depression between the moxibustion group and the electroacupuncture group (MD = 0.69, 95% CI: [−1.45 to 2.83], P = .53).

3.5.2.8. Moxibustion plus electroacupuncture versus electroacupuncture

Only 1 study compared moxibustion plus electroacupuncture to electroacupuncture.^[39] This study showed that the HAMD scores of the moxibustion plus electroacupuncture group were significantly lower than those of the electroacupuncture group in patients with depression (MD = −4.45, 95% CI: [−6.84 to −2.06], P = .0003).

3.5.2.9. Meta-regression analysis of HAMD score changes between the intervention and control groups

To explore potential moderating factors, we applied a random-effects meta-regression model to examine the impact of treatment duration and type of intervention on the effect size. The results are presented in Figure 8. However, neither variable reached statistical significance (P > .05).

Figure 8.

Meta-regression analysis of HAMD score changes between the intervention and control groups. CI = confidence interval, HAMD = Hamilton Depression Rating Scale.

3.5.3. Experimental group versus control group in adverse reaction rate

Five studies involving 359 patients reported adverse reaction rates. Moderate heterogeneity was detected (RR = 0.57, 95% CI: [0.39–0.84], P = .005), and a fixed-effect model was applied. The results indicated that moxibustion, either alone or combined with other treatments, had a better safety profile compared to other treatments alone, as shown in Figure 9.

Figure 9.

Meta-analysis forest plot of adverse event incidence: moxibustion group vs. control group. CI = confidence interval.

3.5.4. Sensitivity analysis

We conducted sensitivity analyses on the main outcomes by excluding individual studies. The results showed no significant changes in effective rate or HAMD scores, indicating stability in our findings. Moreover, the combined effect size was less influenced by individual studies, as shown in Figures 10 and 11.

Figure 10.

Results of a sensitivity analysis of the effective rate. CI = confidence interval.

Figure 11.

Results of a sensitivity analysis of the HAMD scores. HAMD = Hamilton Depression Rating Scale.

3.5.5. Publication bias

Publication bias for the effective rate was assessed using Egger’s test and a funnel plot. The funnel plot showed asymmetry, indicating potential bias, which was confirmed by Egger’s test (t = 3.33, P = .005). This bias may be due to all included studies being published in Chinese, with a preference for reporting statistically significant results, potentially leading to an incomplete reflection of the true effective rate, as shown in Figure 12.

Figure 12.

A funnel plot was used to assess publication bias in reporting the effective rate during meta-analysis. RR = relative risk, SE = standard error.

3.6. Quality of the evidence

The certainty of the evidence was evaluated according to the GRADE framework. Evidence was downgraded by 1 or 2 levels when concerns regarding inconsistency or risk of bias were identified. Overall, the certainty of evidence across the included studies ranged from very low to low, mainly because of inadequate reporting and limitations in methodological quality. Further details are presented in Table 2.

Table 2.

The result of the GRADE system.

Name of outcome	RCTs/article	Evidence quality evaluation				Number of events/sample size			Absolute effect	Evidence	Quality
		Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Experimental	Control
Total effective rate	15	Serious*	No serious	No serious	No serious	None	718/779	631/731	RR = 1.07, 95% CI(1.03–1.1)	⊕⊕⊄⊄	Low
Total HAMD scores	15	Serious*	Serious†	No serious	No serious	None	779	731	MD = 2.1, 95% CI(3.00–1.2)	⊕⊕⊄⊄	Low
Moxibustion VS antidepressants	4	Serious*	No serious	No serious	Serious‡	None	402	357	MD = 0.6, 95% CI(1.31–0.12)	⊕⊕⊄⊄	Low
Moxibustion plus antidepressants VS antidepressants	5	Serious*	Very serious†	No serious	Serious‡	None	181	182	MD = 3.17, 95% CI(5.43–0.92)	⊕⊄⊄⊄	Very low
Adverse reaction rate	5	Serious*	No serious	No serious	Serious‡	None	28/181	49/178	RR = 0.57, 95% CI(0.39–0.84)	⊕⊕⊄⊄	Low
Moxibustion VS acupuncture	1	Serious*	No serious	No serious	Serious‡	None	40	40	MD = 6.07, 95% CI(9.24–2.90)	⊕⊕⊄⊄	Low
Moxibustion plus acupuncture VS acupuncture	1	Serious*	No serious	No serious	Serious‡	None	29	29	MD = 2.47, 95% CI(3.72–1.22)	⊕⊕⊄⊄	Low
Moxibustion plus ear auricular acupressure VS ear auricular acupressure	1	No serious	No serious	No serious	Serious§	None	37	34	MD = -0.62, 95% CI(−1.45 to 0.21)	⊕⊕⊄⊄	Low
Moxibustion plus acupuncture plus SSRI VS acupuncture plus SSRI	1	Serious*	No serious	No serious	Serious‡	None	30	30	MD = -2.68, 95% CI(−4.52 to −0.84)	⊕⊕⊄⊄	Low
Moxibustion VS electroacupuncture	1	Serious*	No serious	No serious	Serious§	None	30	30	MD = 0.69, 95% CI(−1.45 to 2.83)	⊕⊄⊄⊄	Very low
Moxibustion VS electroacupuncture vs electroacupuncture	1	Serious*	No serious	No serious	Serious‡	None	30	29	MD = -4.45, 95% CI(−6.84 to −2.06)	⊕⊕⊄⊄	Low

CI = confidence interval, HAMD = Hamilton Depression Rating Scale, MD = mean difference, RCT = randomized controlled trial, RR = relative risk, SSRI = selective serotonin reuptake inhibitor.

^*The included studies exhibit considerable bias in terms of randomization, allocation concealment, and blinding.

^†High heterogeneity.

^‡Sample size < 400.

^§The sample size < 400, and the confidence interval is relatively wide.

4. Discussion

4.1. Summary of main results

There is no specific name for depression in TCM. According to the definition and description of clinical symptoms of depression in Western medicine, “yu zheng,” “zang zao,” “dai zheng,” “dian zheng,” “lily disease,” “confusion disease,” “easy forgetting,” “insomnia” in TCM are similar clinical manifestations of depression.^[63] Yin-Yang theory is one of the fundamental concepts in TCM, positing that the onset of depression is closely related to an imbalance between Yin and Yang within the human body.^[64] According to TCM, maintaining a harmonious balance between Yin and Yang is essential for health, while disruption of this equilibrium may result in various diseases.^[65] In patients with depression, symptoms such as persistent low mood and loss of interest are often associated with a deficiency of Yang qi.^[66] Moxibustion, a therapeutic technique in TCM, is believed to restore this balance by warming Yang, dispelling cold, and regulating the interplay between Yin and Yang.^[67] The therapeutic effects of moxibustion are thought to arise from the synergistic interaction between the heat generated by the burning moxibustion stick and the accompanying influence of the smoke on meridians and acupoints. Recent studies indicate that the synergistic mechanisms of moxibustion include thermal stimulation from burning mugwort, which enhances local circulation^[68]; absorption of combustion products such as volatile compounds and tar-like substances that may affect neurotransmitter levels like serotonin^[69]; and the unique aromatic scent of moxa smoke, which promotes emotional relaxation and psychological well-being.^[70] This process facilitates the circulation of qi and blood, ultimately leading to emotional regulation and improvements in mental well-being. As a modality of traditional acupuncture, moxibustion is an external therapy with several primary methods, including suspended, direct, indirect, and apparatus-type moxibustion.^[71] Compared to acupuncture, moxibustion is associated with fewer side effects, particularly with the advent of new moxibustion products that address the issue of burns, which were a concern in traditional methods.^[72] These modern products are compact yet highly effective, offering a gentle application without causing burns or damage to the meridians. Moreover, they are composed of pure, additive-free ingredients, enhancing their safety and health benefits.^[73] These advantages position moxibustion as a promising therapeutic option for the treatment of depression, with significant potential for clinical application.

The precise pathogenesis of depression remains unclear, though several hypotheses have been proposed to explain its etiology at different levels, including the monoamine neurotransmitter hypothesis, the neuroplasticity hypothesis, and the inflammatory hypothesis.^[74] The neuroplasticity hypothesis shows that depression is closely associated with structural alterations in specific brain regions, such as reduced hippocampal volume and glial cell density.^[75] This theoretical framework provides a possible explanation for the mechanism by which moxibustion exerts its antidepressant effects. For instance, a study by Jin et al demonstrated that moxibustion applied at Shenshu (BL 23), Sanyinjiao (SP 6), and Baihui (GV 20) effectively promoted the repair and regeneration of glial cells in the hippocampus of depression model rats, thereby modulating neuroplasticity.^[76] Additionally, the monoamine neurotransmitter hypothesis suggests that the development of depression is related to decreased levels or deficiencies in neurotransmitters, such as 5-hydroxytryptamine.^[77] This provides another potential mechanism through which moxibustion may exert its antidepressant effects. Ding et al reported that moxibustion at Baihui (GV 20), Dazhui (GV 14), and other acupoints significantly increased 5-hydroxytryptamine production in depressed rats.^[78] These findings suggest that moxibustion may exert its antidepressant effects by promoting the repair of glial cells and enhancing the production of neurotransmitters. Simultaneously, recent research indicates moxibustion has been shown to alleviate depression-like behaviors in chronic unpredictable mild stress mice primarily by reducing neuroinflammation, providing a theoretical basis for its application in the treatment of depression.^[79]

The studies included in this review indicate that moxibustion for the treatment of depression frequently targets acupoints such as Dazhui (GV 14), Baihui (GV 20), and Ganshu (BL 18). Baihui (GV 20) is recognized for its role in regulating psychological functions by modulating the central nervous system, improving cognition, and calming the mind.^[80] Dazhui (GV 14), as a primary acupoint for regulating yang energy throughout the body, facilitates the circulation of qi and blood, strengthens the body’s immunity, and stabilizes the defense system.^[81] Ganshu (BL 18), by regulating the flow of qi and blood in the liver meridian, helps alleviate liver qi stagnation and emotional distress, making it particularly effective in treating liver qi stagnation-type depression.^[82] The combination of these acupoints in moxibustion therapy exerts a synergistic effect, contributing to the comprehensive treatment of depression.

Among the 15 studies included in the meta-analysis, 4 demonstrated that the efficacy of moxibustion was comparable to that of conventional antidepressant medications, suggesting its potential as a non-pharmacological alternative for depression treatment. Five studies reported that moxibustion combined with SSRIs was more effective than SSRIs alone, further highlighting the advantages of moxibustion as an adjunctive therapy. Additionally, the effectiveness of moxibustion, either alone or in combination with acupuncture, was superior to acupuncture monotherapy. One study also indicated that moxibustion combined with electroacupuncture may provide greater therapeutic benefits compared to electroacupuncture alone. However, 1 study observed no significant differences between moxibustion combined with auricular point massage and auricular point massage alone, suggesting the need for further exploration into individual variability, depression subtypes, and treatment combinations. In terms of safety, the incidence of adverse reactions with moxibustion, either as a standalone treatment or combined with other therapies, was relatively low. The most commonly reported side effects were headache, dizziness, nausea, dry mouth, constipation, and insomnia, supporting the relative safety profile of moxibustion in treating depression.

4.2. Limitations of the study

It is important to acknowledge the limitations of this study. As all included trials were conducted in China and published in Chinese, the generalizability of the findings to other populations or regions is limited; thus, multi-center international studies are needed to validate the effectiveness of moxibustion in broader clinical settings. Additionally, the high heterogeneity observed across studies may be attributable to variations in moxibustion techniques, acupoint selection, disease duration, treatment periods, and concurrent medications. The small sample sizes and suboptimal methodological quality of the included studies may also compromise the reliability of the findings. Therefore, future research should adopt more rigorous study designs and incorporate a broader range of validated outcome measures beyond the HAMD to enhance the robustness and generalizability of the results.

5. Conclusion

In summary, moxibustion therapy has shown promising potential in alleviating depressive symptoms, particularly when used in combination with antidepressants, offering enhanced efficacy and a favorable safety profile. This review supports the potential role of moxibustion as a primary or adjunctive treatment for depression. Meanwhile, further high-quality, large-scale RCTs are needed to confirm these findings and enhance the credibility of this therapeutic approach.

Author contributions

Conceptualization: Yingcun Li, Minling Li, Xin Li.

Data curation: Yijin Xie, Xu Li, Aihua Yin.

Formal analysis: Yingcun Li, Minling Li, Fengxia Sun.

Funding acquisition: Xin Li.

Investigation: Yingcun Li, Minling Li.

Software: Fengxia Sun, Yijin Xie, Xu Li, Aihua Yin.

Supervision: Zengxun Liu, Xin Li.

Writing – original draft: Yingcun Li, Minling Li.

Writing – review & editing: Yingcun Li, Minling Li, Xin Li.