Effectiveness and safety of moxibustion for primary insomnia: a systematic review and meta-analysis

Background Primary insomnia is a widespread and refractory disease. Moxibustion therapy for insomnia shows some advantages compared with conventional therapies. This systematic review and meta-analysis of randomized controlled trials (RCTs) was conducted to evaluate the effectiveness and safety of moxibustion therapy for insomnia. Methods We conducted a comprehensive literature review of the CENTRAL, PubMed, EMBASE, Web of science, CNKI, VIP, and Wanfang Data databases from their inception to July 2015 for RCTs that compared moxibustion with western medications, oral Chinese medicine, or other methods of traditional Chinese medicine (TCM) in patients with primary insomnia. The primary outcome measure was effective rate and secondary outcome measure was adverse events. Data collection and analysis included risk of bias evaluation, meta-analysis, sensitivity analysis, publication bias and adverse events analysis according to corresponding criteria. Results The study included 22 RCTs (1,971 patients). The quality of the studies was low. The overall meta-analysis demonstrated that moxibustion was more effective for insomnia than western medications, oral Chinese medicine and other TCM therapies (RR = 1.17, 95 % CI 1.12 to 1.23, P < 0.00001). Subgroup analyses demonstrated that moxibustion was more effective for insomnia than western medications (RR = 1.16, 95 % CI 1.09 to 1.24, P < 0.00001), oral Chinese medicine (RR = 1.11, 95 % CI 1.04 to 1.18, P = 0.002), and other TCM therapies (RR = 1.22, 95 % CI 1.15 to 1.30, P < 0.00001). There were no serious adverse effects associated with moxibustion therapy for insomnia, and the rate of adverse events was low. Conclusion It is difficult to get the conclusion regarding the effectiveness and safety of moxibustion for primary insomnia due to insufficient evidence, such as the high risk of bias in the included studies, small sample sizes, and few reports on adverse effects. Moxibustion should be considered as a novel therapeutic option for insomnia, and more rigorous clinical trials of moxibustion therapy for insomnia are needed to assess its effects.


Description of the condition
Insomnia is a sleep disorder characterized by the inability to fall asleep, sleep loss and poor-quality sleep. Insomnia is caused by multiple physiological, psychological, and environmental factors [1][2][3][4].
Insomnia is related to the function of the cerebral cortex and results from mental or nervous tension. Insomnia is thought to be associated with a group of centrally located neurons coupled with dynamic transformation of neurotransmitters, including norepinephrine (NE) and 5-hydroxytryptamine (5-HT) [5].
However, insomnia remains under-diagnosed and under-treated. An estimated 47-67 % of individuals with insomnia do not seek medical attention. Among those that do attempt to resolve their sleep problems, only 50-90 % receive treatment [17,38].

Description of the intervention
Western conventional medicine recommends pharmacological treatment (such as hypnotic sedative agents) and cognitive behavioral therapy (CBT) for insomnia [39].
Pharmacological agents are effective for insomnia but are only recommended for short-term relief. The longterm use of these medications is associated with adverse effects such as disturbed sleep architecture, rebound insomnia, withdrawal effects [40], damage to cerebral nerves, memory and psychomotor impairment, hypofunction, dependency, and addiction [41]. For example, benzodiazepines may cause headaches, nightmares, daytime fatigue, nausea, confusion, and falls [42]. Zdrugs can result in bizarre behaviors, dizziness, falls, and gastrointestinal upset [43].
Evidence from clinical studies supports the use of CBT for insomnia [44,45]. However, CBT is not effective in all patients [46], and access to treatment is limited [45,47] because qualified CBT therapists are rare [48] and expensive [49].
Consequently, insomnia sufferers require alternative treatments [50,51]. Moxibustion is a component used in traditional Chinese medicine (TCM). Some Chinese studies by randomized controlled trials (RCTs) or clinical observations suggest that the moxibustion has the potential to be an effective and safe therapy for insomnia, such as improving sleep quality, adjusting the brain's sleep function, improving symptoms of dreaminess, dizziness, headache, heavy head and poor memory, and promoting the periodicity from light to deep sleep [52].

How the intervention might work
Modern medicine believes moxibustion modulates neurotransmitters to resist insomnia, thereby improving sleep quantity [53]. Experiments in rats indicate that moxibustion protects against chronic stress by acting on the hippocampal neurons to increase the amount of brain-derived neurotrophic factor as well as 5-HT and its metabolites. Holistic healthcare uses moxibustion to generate far-infrared and near infrared energy to regulate dysfunctional organs and build wellbeing [54]. Suspended moxibustion at Baihui can treat nervous system diseases by improving the blood supply to brain tissue, increasing the elasticity of blood vessels, and enhancing the excitability of related sites on the cerebral cortex [54]. In TCM, moxibustion is thought to regulate qi and the blood, tonifying healthy qi to eliminate pathogenesis by means of warming. Moxibustion applied at Bǎihuì can balance yin and yang, tonify both the heart and the spleen, dredge blood vessels, and tranquilize [54] mind [5].

Why it is important to do this study
The effectiveness of moxibustion therapy for insomnia remains controversial; therefore, its application is limited. There are currently no published systematic reviews or meta-analyses investigating the effectiveness and safety of moxibustion therapy for insomnia.

Objective
This systematic review and meta-analysis of randomized controlled trials (RCTs) was conducted as a rigorous evaluation of the effectiveness and safety of moxibustion therapy for insomnia.

Methods
This systematic review and meta-analysis is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [55].

. #1 AND #2 AND #3
A monthly e-mail alert was set up at the National Center for Biotechnology Information (NCBI) from the U.S. National Library of Medicine (NLM) to obtain updates of new publications.

Inclusion criteria
1) RCTs of patients that were dissatisfied with their quality of sleep; 2) in which the intervention group included patients undergoing therapy with different methods of moxibustion as monotherapy or combination therapy (including grain-moxibustion, thunder-fire moxibustion, heat-sensitive moxibustion), and the control group included patients undergoing therapy with western medications, or oral Chinese medicine, or other TCM therapies (including acupoint massage, point-application, head-needle acupuncture, auricularplaster therapy, and acupuncture); 3) The primary outcome measure was the clinical effective rate. It was a dichotomous outcome and the overall effectiveness of moxibustion therapy as a subjective assessment, which was defined as the proportion of participants who got improved in sleep quality and was based on response evaluation criteria used in the treatment of insomnia with TCM. What's more, it was reported by trial participants themselves.  [39,56]. Then the patients of "cure, markedly effective, effective" were taken as people who got improved in sleep quality and the patients of "ineffective" were taken as people who got unimproved in sleep quality. The total number of "cure, markedly effective, effective" were used to calculate effective rate. Other assessment criteria of clinical therapeutic effect with comparable definitions were also considered [39]. 4) The secondary outcome measure was adverse events associated with the use of moxibustion therapy for insomnia. It was reported in the articles or measured by validated scales, e.g., Health Survey Questionnaire, Treatment Emergent Symptom Scale (TESS) et al.

5) Only English and Chinese as language selection
Exclusion criteria 1) studies that were not RCTs; 2) patients diagnosed with primary insomnia resulting from another physiological or psychological disease; 3) trials in special patient populations such as menopausal women; 4) trials in which moxibustion as combination therapy was not the only intervention to differ between the treatment and control group; 5) studies reporting fraudulent data or with insufficient data; 6) duplicate studies.

Study selection
Two review authors (SYJ, YJM) independently examined titles and abstracts to select eligible RCTs. When datasets overlapped or were duplicated, only the most recent information was included. Then the full text of potentially relevant studies was retrieved. Two author reviewers (SYJ and YJM) independently examined the full text records to determine which studies met the inclusion criteria. Disagreements about the study selection were resolved by discussion with a third review author (YZM) and consensus.

Data extraction and management
Two review authors (SYJ and YJM) independently extracted the data from eligible RCTs including details on the study population, interventions, and outcome measures. Disagreements about data extraction were resolved through discussion with a third review author (YZM) and consensus.

Assessment of quality of evidence in included studies
The methodological quality of RCTs was assessed independently using the Cochrane Handbook for Systematic Reviews of Interventions [57] from 7 parts, including random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessments, incomplete outcome data, selective reporting, and other bias.
Two review authors (SYJ, YJM) independently evaluated the methodological quality of the included articles. Disagreements about the assessment of quality of evidence in included studies were resolved through discussion with a third review author (YZM) and consensus.
RCTs with fraudulent data of low quality were not included in the meta-analysis. A random-effects model was used to pool the studies with significant heterogeneity, as determined by the inconsistency index (I 2 ≥ 30 %). A fixed effect model was used to pool the studies in the absence of substantial heterogeneity (I 2 < 30 %).

Data analysis
Sensitivity analyses were conducted to explore the impact of confounding factors.
Publication bias was comprehensively assessed using funnel plot by RevMan v5.3. and Begg's rank correlation test of asymmetry by stata.13.0. Publication bias was thought to be insignificant at P > 0.05 [58].

Trial identification
The searches identified 590 articles. Titles and abstracts were screened, and 77 RCTs were considered potentially eligible for inclusion. After analyzing the full-text articles and a risk of bias assessment, 55 RCTs were excluded. Twenty-two RCTs were found eligible based on our inclusion criteria (Fig. 1).
The baseline was comparable because there were no significant differences in gender, age, or disease duration between the intervention and control groups (P > 0.05).

Risk of bias in included studies
The overall risk of bias in the 22 included trials was high (Fig. 2).
In 19 trials [5, 54, 59-64, 67-72, 74-78], the risk of bias due to incomplete outcome data was assessed as low because there were no missing data and all expected outcomes were reported. Three articles [65,66,73] were assessed as having unclear risk because they did not report sufficient detail to let us make sure the baseline was balanced after dropouts and did not use an intention-to-treat (ITT) analysis, though the dropouts were less than 20 % in each article.

Effective rate of moxibustion for insomnia
The effective rate of moxibustion for insomnia was described in all 22 included trials.
To account for clinical heterogeneity and subgroup differences probably arising from the use of different criteria to evaluate the effectiveness of moxibustion therapy for insomnia, a sensitivity analysis of trials using only GCTNPCM criteria was conducted. The effectiveness of moxibustion classified according to GCTNPCM criteria was described in 12 trials (moxibustion vs. western medications in 3 trials [60][61][62], moxibustion vs. oral Chinese medicine in 3 trials [63,66,67] and moxibustion vs. other TCM therapies in 6 trials [5,70,72,74,75,77]).
In sensitivity analysis (Fig. 4), the overall meta-analysis demonstrated that moxibustion was significantly more effective for insomnia than western medications, oral Chinese medicine and other TCM therapies (RR = 1.21, 95 % CI 1.14 to 1.28, P < 0.00001). There was no evidence of significant heterogeneity between the trials (χ 2 = 3.15, P = 0.99, I 2 = 0 %) and tests for subgroup differences showed there were no potential differences between the groups (χ 2 = 1.14, P = 0.57, I 2 = 0 %).

Adverse events
Adverse events associated with moxibustion therapy for insomnia were described in 3 studies. Wu [59] reported 3 cases of headache, 5 cases of fatigue, 1 case of constipation, and 2 cases of diarrhea in the moxibustion group, and 53 cases of bitter taste in the mouth, 6 cases of thirst, 8 cases of myasthenia, 16 cases of drowsiness, 15 cases of vomiting, and 91 cases of abstinence symptoms in the estazolam group. Wang [76] reported 1 case (5.5 %) of aching in the moxibustion plus pointapplication group and 1 case of aching, 1 case of nausea, and 1 case of diarrhea (total 3 cases, 16.5 %) in the point-application group (p < 0.05). Xu [78] reported 5 cases (8.33 %) of scalp pain in the moxibustion plus head-needle acupuncture group and 4 cases (7.47 %) of scalp pain in the head -needle acupuncture group.

Summary of results
The results demonstrated that moxibustion was more effective for insomnia than western medications, oral Chinese medicine, and other TCM therapies both in the overall meta-analysis and subgroup meta-analysis. However, the efficacy of moxibustion therapy for insomnia cannot be confirmed due to a high risk of bias in the included studies and the small sample sizes. Our findings indicate no serious adverse effects associated with moxibustion therapy for insomnia and a low rate of adverse events; however, only a small number of studies reported on the safety of moxibustion.

Quality of the evidence
The results of our study should be interpreted with caution due to the high risk of bias in the included trials. In particular, risk of bias due to the random sequence generation, allocation concealment, blinding of participants, personnel, and outcome assessments and incomplete outcome data are considerations. Firstly random sequence generation and allocation concealment are important to prevent selection bias. Trials with inadequate allocation concealment report, on average, that the intervention is 18 % more "beneficial" than in trials with adequate concealment (95 % CI 5 % to 29 %) [79]. Secondly, blinding of participants, personnel, and outcome assessments are important to prevent performance bias and detection bias. Thirdly, as insomniacs are liable to be anxious and drop out if they feel the treatment invalid, according to our clinical experience, especially in a long treatment duration, detailed information about how to prevent dropouts or whether to use ITT to deal with data of dropouts plays irreplaceable role when warranting the validity and the reliability of data. But most of the included study did not report these 4 parts clearly or rightly.

Limitations
First, diagnostic criteria for insomnia varied between included trials. As the objective of this study was to investigate the effectiveness of moxibustion for insomnia, subgroup analyses according to diagnostic criteria were not performed. This is in accordance with other metaanalyses of TCM for insomnia [39,80]. Second, overall risk of bias in the 22 included trials was high. Third, the included trials relied on different criteria to classify the effectiveness of moxibustion which could lead to heterogeneity between the trials or subgroup differences, and the number of studies according to different effectiveness criteria is not enough if we made subgroup. However, a sensitivity analysis focusing on GCTNPCM criterion indicated the results of the meta-analysis were robust. Fourth, there was an obvious publication bias. The reason may be some Asian countries, including China, publish an unusually high proportion of positive results [81]. Fifth only 3 studies pay attention to adverse effect.

Implications
Our data suggest that additional high-quality trials are warranted to determine the benefits of moxibustion for insomnia. When designing and reporting future RCTs on moxibustion for insomnia, we recommend that the CONSORT 2010 statement [82,83], which consists of a 25-item checklist to determine trial quality and rigor, should be used as a guideline. All clinical trials should be registered prior to the enrollment of the first patient, according to the International Committee of Medical Journal Editors statement [84]. Randomization methods should be clearly described and fully reported. Although blinding may be difficult, the blinding of patients and outcome assessors should be attempted. Well-defined and widely recognized diagnostic or classified criteria, such as ICD-10 or the CCMD-3, should be employed to make a precise clinical diagnosis of insomnia and increase comparability between trials. Additionally, because insomnia is a highly heterogeneous disease and presents various etiologies and severities, moxibustion is likely to differentially affect various patient subgroups. Therefore, future clinical trials should focus on particular subgroups or include a very large sample size to delineate the effects of moxibustion on different patient types. Interventions should include appropriately long treatment periods, treatment frequencies, and follow-up periods. The severity of insomnia may vary despite the presence of absence treatment; thus, a longer follow-up period with serial measurements of outcomes is important to determine the long-term effects of moxibustion. Internationally recognized and validated outcome measurements should be consolidated and used consistently. Appropriate statistical analyses should be carried out for the baseline data and ITT analysis is recommended in case of dropouts or withdrawal. Trials should have a sufficiently large sample size, ideally based on formal power calculations.

Conclusion
It is difficult to get the conclusion regarding the effectiveness and safety of moxibustion for primary insomnia due to insufficient evidence, such as the high risk of bias in the included studies, small sample sizes, and few reports on adverse effects. Moxibustion should be considered as a novel therapeutic option for insomnia, and more rigorous clinical trials of moxibustion therapy for insomnia are needed to assess its effects.