Non-traditional Acupuncture Therapies for Smoking Cessation: A Systematic Review of Randomized Controlled Trials


 BackgroundNon-traditional acupuncture (NTA) therapies widely used for smoking cessation include acupressure, transcutaneous electrical acupoint stimulation (TEAS), laser acupuncture, intradermal needle, and acupoint catgut embedding (ACE). Our aim was to evaluate their therapeutic effects and safety for smoking cessation. MethodsRandomized controlled trials (RCTs) comparing NTA therapies with sham NTA or conventional therapy for smoking cessation were included. Ten databases were searched from their inception to February 2021. Two review authors independently screened studies, extracted data, and assessed the risk of bias. Meta-analysis was conducted with RevMan 5.4 software. Grading of Recommendations Assessment, Development and Evaluation (GRADE) was applied to assess the quality of evidence. The primary outcome was abstinence rate at short-term (1-3months), mid-term (3-6months), and long-term (≥6months). ResultsTwenty-five RCTs involving 2600 smokers were identified. Acupressure was found more effective than sham acupressure or conventional therapy in improving short-term (RR 1.41, 95% CI [1.04 to1.91]; low certainty; 8 trials, n=637) and mid-term abstinence rate (RR 1.63, 95% CI [1.27 to 2.09]; low certainty; 8 trials, n=749). Intradermal needle was not superior to sham control or counseling for short-term (RR 1.62, 95% CI [0.85 to 3.08]; low certainty; 5 trials, n=346) and mid-term abstinence rate (RR 1.49, 95% CI [0.68 to 3.27]; low certainty; 3 trials, n=187).TEAS failed to show a better effect than sham TEAS or counseling for short-term abstinence rate (RR 1.27, 95% CI [0.96 to 1.67];moderate certainty;5 trials, n=485), TEAS appeared more beneficial for mid-term abstinence (RR 1.58, 95% CI [1.10 to 2.27]; moderate certainty; 3 trials, n=325). Laser acupuncture was superior to sham control for long-term abstinence rate (RR 2.25, 95% CI [1.23 to 4.11]; moderate certainty; 2 trials, n=160). ACE was comparable to Bupropion for mid-term abstinence rate (RR 0.99, 95% CI [0.70 to 1.40]; low certainty; 2 trials, n=177). No serious adverse events were reported in the included trials. Conclusions﻿Low to moderate certainty evidence suggests that acupressure, TEAS, laser acupuncture and ACE maybe effective in achieving short-term, middle-term or long-term smoking cessation. Further large, long-term follow-up RCTs are warranted to verify their benefits and safety.Systematic review registration: INPLASY 202120054.


Background
Cigarette smoking is one of major public health issues worldwide, which is closely related to the onset and progression of many diseases. Reports suggest that about 480,000 people die prematurely every year, and one-fth of the deaths are related to smoking [1] . The Global Burden of Diseases, Injuries, and Risk Factors Study reported that in 2019, out of 87 risk factors, smoking was the second leading risk factor for attributable deaths globally [2] . There were as many as 40 million adult smokers in the United States in 2017, and smokers accounted for 14% of the 15.8% adult population [3] . However, in 2019 more than 50.6 million US adults smoked (accounting for 20.8% of the adult population) [4] . Responses from the National Health Interview Survey in 2015 indicated that 68% smokers (a total of 33,672 adult smokers) wanted to quit smoking [5] . Currently, smoking cessation therapies recommended in guidelines and US Food and Drug Administration approved are pharmacological therapy, behavioral therapy, and complementary and alternative therapies [6] . However, studies have found that pharmacological therapy included nicotine replacement therapy (NRT), Varenicline and Bupropion have been limited for smoking cessation due to their high cost, side effects, tendency to relapse [7][8][9] as well as low popularity and being di cult access in China [10] . Additionally, withdrawal symptoms are hard to control with psychological and behavioral intervention alone. Acupuncture as a complementary and alternative therapy has been used for smoking cessation for nearly 50 years [11] . Clinical and experimental studies have found that acupuncture was effective for withdrawal symptoms through promoting the release of endogenous opioids [12][13] or suppressing the craving for cigarettes [14] .
As acupuncture techniques have developed, various non-traditional acupuncture (NTA) techniques have been used to stop smoking. NTA therapies differ from traditional body acupuncture. Traditional body acupuncture generally using liform needles inserted through the skin at acupoints. NTA therapies generally refer to acupressure, transcutaneous electrical acupoint stimulation (TEAS), laser acupuncture, intradermal needle, acupoint catgut embedding (ACE) or acupoint injection. These are also known as acupuncture or acupuncture related therapies. NTA therapies have been widely used for nicotine dependence in several trials worldwide due to convenience and acceptability. The majority of systematic reviews [15][16][17] aimed to evaluate the effectiveness of traditional body acupuncture for smoking cessation, one Cochrane systematic review [18] focused on the effectiveness of both traditional body acupuncture and some NTA therapies on smoking cessation. So far, we have not found any systematic reviews that comprehensively and individually evaluate common NTA therapies for smoking cessation. Therefore, this systematic review aims to collect all randomized controlled trials (RCTs) on NTA therapies for smoking cessation and to evaluate their therapeutic effect and safety.

Methods
This systematic review was reported following PRISMA 2020 19

Eligibility criteria
Parallel group RCTs regardless of blinding were included. The study population comprised of smokers of any age who had no serious diseases or were pregnant and wanted to quit smoking. The study interventions were NTA therapies used alone or in combination with conventional therapy, which included medication or behavioral counseling. NTA therapies refer to acupressure, transcutaneous electrical acupoint stimulation (TEAS), laser acupuncture, intradermal needle, acupoint catgut embedding (ACE) or acupoint injection. Interventions with traditional inserted liform needles on the body for smoking cessation were ineligible. The controls were medications (NRT, Bupropion or Varenicline), behavioral counseling, sham acupuncture or non-speci c acupoint stimulation, or no treatment. The primary outcome was smoking cessation, de ned as the abstinence rate. The secondary outcomes were nicotine withdrawal symptoms measured by Minnesota Nicotine Withdrawal Scale Score, daily cigarette consumption, the Fagerström test for nicotine dependence (FTND), the level of exhaled carbon monoxide (CO), relapse rate, craving for cigarettes and adverse events. Trials that failed to report at least one required outcome were excluded.

Study selection and data extraction
After removing duplicates, two authors (ZYY and HDL) independently screened studies by title and abstract, uncertainty was determined for eligibility through checking full texts. Reasons for excluding trials were recorded at the full-text screening stage, and any discrepancies were discussed by two review authors or arbitrated by third party (JPL). In the data extraction process, data were extracted by two authors (ZYY and HDL) independently using a pre-de ned electronic data extraction form which included basic information of study design (study ID, setting, sample size, centers, and funding); participants characteristics; details of NTA therapies and controls; outcomes in different measuring time; follow ups; dropouts and adverse events.

Quality assessment
The methodological quality of each included trial was evaluated independently by two review authors (ZYY and HDL). Cochrane Risk of Bias tool (ROB) [21] was employed to assess the risk of bias of each trial based on its seven domains (the adequacy of sequence generation, allocation concealment, blinding of participants, blinding of outcome assessors, incomplete outcome data, selective reporting, and other bias). Grading of Recommendations Assessment, Development and Evaluation (GRADE) [22] was applied to evaluate the certainty of the body of evidence based on risk of bias, directness, precision, consistencies, and publication bias.

Data synthesis
For dichotomous data, data were presented as risk ratio (RR) with 95% con dence interval (CI). For continuous data, mean difference (MD) with 95% CI was estimated. Meta-analysis was conducted by Cochrane Review Manger 5.4 software when the trials have similarities in study design and clinical characteristics. Otherwise, data were synthesized qualitatively. The I 2 statistic was utilized to test the statistical heterogeneity. The heterogeneity [21] was considered as substantial when I 2 statistic value was greater than 50% [21,23] . The xed effects model was used when I 2 ≤ 30%, otherwise the random effects model was applied in meta-analysis. To explain heterogeneity, subgroup analysis was prede ned by the comparisons (active control and inactive control).
Active controls referred to medication and/or behavioral counseling; inactive controls referred to sham acupuncture, non-speci c acupoint stimulation, or no treatment. Sensitivity analysis was employed to explore the in uence of the randomization concealment (clear or not) and blinding (blinded or not). Funnel plots were generated to detect possible publication bias if 10 or more trials were included in a meta-analysis.

Screening
Initially 660 records were retrieved and 191 duplicates were removed. A total of 423 records were excluded in title and abstract screening process, remaining 46 records. In the full-text screening process, 21 trials were excluded because: not real RCTs, had only English abstracts, interventions were combined with traditional body acupuncture, ineligible controls, and ineligible or incomplete outcomes. Finally, 25 trials were identi ed and were all quantitatively synthesized. The ow chart of study selection was shown in (Fig. 1).
The protocols were not reported in the rest of the trials and were considered as unclear risk of reporting bias. In terms of the "other bias" domain, the funding was reported and the baseline data were comparable in 11 trials [24-25, 27, 29, 31-32, 34-35, 37, 40, 48] . There was potential con icts of interest in one trial [43] and this was considered as high risk of bias. The remaining 13 trials were assessed as 'unclear' since either the funding or the baseline data was not reported. The risk of bias summary was shown in (Fig. 2).
Only one trial [43] reported the short-term relapse rate and suggested that TEAS was not superior to sham TEAS in reducing relapse rate ( [47][48] comparing the effect of ACE with Bupropion or Varenicline on smoking cessation were identi ed. The pooled data suggested that ACE was comparable to medication in improving abstinence rate at mid-term follow-up (RR 0.99, 95% CI [0.70 to 1.40]; low certainty; 2 trials, n = 177) (Additional le1: Table 1, Table 5).

Adverse events
Four trials [24][25][26]29] reported the transient and minor auricular adverse events both in the acupressure group (itching in 12 cases, mild tenderness in 13 cases, feeling hot in 4 cases) and in the control group (uncomfortable feeling in 5 cases, dizziness in 1 case), the pooled data suggested that there was no signi cant difference between acupressure and controls (RR 2.51, 95% CI [0.24 to 26.59]; I 2 = 70%; 4 trials, n = 240) (Additional le1: Table 1). One trial [43] reported adverse events both in TEAS group (pain in 2 cases, soreness in 3 cases) and sham TEAS (soreness in 2 cases, headache in 2 cases). One trial [47] reported adverse events both in ACE group (pain and soreness in 2 cases, minor swelling in 1 case) and bupropion group (nausea in 5 cases, insomnia in 2 cases). No serious adverse events were reported in the included trials.

Publication bias
Funnel plots were not performed to detect publication bias since there was no meta-analysis combining more than ten trials at one time.

Additional analysis
Prede ned subgroup analysis via the controls (active control and inactive control) was conducted for 3 outcomes under acupressure, and one outcome under intradermal needle. We were unable to conduct other meaningful subgroup analysis due to limited trials. Sensitivity analysis for allocation concealment for short-term abstinence rates suggested that clear allocation concealment may in uence the results (RR 1.51, 95% CI [0.95 to 2.41]; I 2 = 44%; 5 trials [24-25, 28, 30, 34] , n = 520).
3.4.9 Certainty of evidence GRADE approaches were employed to assess the certainty of evidence from primary outcomes. The quality of the evidence was downgraded to low or very low quality due to lack of blinding, imprecision, inconsistency or indirectness. The detailed evidence summary of outcomes is presented in ( 4. Discussion

Main ndings
Twenty-ve RCTs involving 2600 smokers were identi ed. The overall risk of bias was not serious. We downgraded the certainty of evidence to moderate or low due to small number of events or absence of blinding method. NTA therapies had different effects on smoking cessation at different time points. Compared with sham acupressure or conventional therapy, acupressure improved the short-term and mid-term abstinence rate by 11.9% and 11.0% respectively. Additionally, acupressure was also favorable to decrease the short-term withdrawal symptom score (-2.68) and reduce the level of exhaled CO (-3.84 ppm). Intradermal needles failed to demonstrate a better effect than sham control or counseling in achieving both short-term and mid-term smoking cessation. TEAS was found superior to sham TEAS or counseling in improving mid-term abstinence rate by 11.2%. However, TEAS failed to show a better effect in reducing daily cigarettes consumption, the level of exhaled CO, FTND, and relapse rate. Two double blinded trials suggested that compared with sham laser acupuncture, laser acupuncture successfully improved abstinence rate by 20.6% at long-term follow-up. Two trials involving 180 smokers compared the effect of ACE with Bupropion or Varenicline, with 41.6% and 42.0% smokers achieving smoking cessation in the ACE group and control group respectively at 8-12 weeks follow-up suggesting that ACE maybe comparable with Bupropion or Varenicline in assisting smoking cessation. We have not found evidence of serious adverse events associated with the use of NTA therapies.

Relation to previous research
A Cochrane systematic review [18] published in 2014 evaluated the effect of acupuncture and related interventions on smoking cessation. The interventions included some NTA therapies (e.g. acupressure, TEAS, laser acupuncture) used alone or in combination with traditional body acupuncture. In this systematic review, we identi ed NTA therapies (acupressure, TEAS, intradermal needles, laser acupuncture, ACE) and excluded any traditional body acupuncture interventions. We also focused on withdrawal symptoms, nicotine dependence, daily cigarettes consumption, the level of exhaled CO, relapse rate, and craving for cigarettes. We consistently found that acupressure was superior to sham acupressure in achieving short-term smoking cessation. Additionally, we found that acupressure was bene cial in improving mid-term abstinence rate, relieving withdrawal symptoms, and decreasing the level of exhaled CO. TEAS, laser acupuncture, and ACE were also potentially effective for mid-term or long-term smoking cessation.

Strengths and limitations
Several systematic reviews have been conducted to evaluate the effect of traditional body acupuncture on smoking cessation [15,18,49] . However, very few systematic reviews have focused on the potential bene ts of NTA therapies on smoking cessation. NTA therapies are widely used for smoking cessation due to convenience and good compliance [50] . Therefore, this systematic review was a comprehensive and individual evaluation of NTA therapies for smoking cessation. Withdrawal symptoms, nicotine dependence, and other outcomes were also assessed. The protocol of this systematic review was registered on INPLASY. Prede ned subgroup analysis via controls (active control and inactive control) was conducted. GRADE approaches were employed to assess the certainty of evidence. There were several limitations for this systematic review. Although, blinding of the participants was applied in more than 50% identi ed trials, it was di cult to con rm whether the participants were unaware of the interventions they received, since "sham acupuncture" can be easily identi ed. Additionally, the certainty of the evidence was downgraded to moderate or low certainty due to small number of participants and events.

Implications for practice
Quitting smoking is a process that requires long-term adherence to treatment if there are to be any bene cial effects to health. An in-depth interview study explored factors in uencing participant compliance in acupuncture trials suggested that patient's fear of traditional body acupuncture pain may reduce treatment compliance [51] . NTA therapies, such as acupressure was widely used for smoking cessation as it is painless, convenient and may be a selfadministered procedure (every 2 or 3 days treatment cycle). It has been reported that withdrawal symptoms were the most serious after quitting for about one month but gradually die down over time, and relapse usually occurs at this stage [52] . Hence, it was important to control withdrawal symptoms to prevent relapse. Acupressure was found effective in relieving short-term withdrawal symptoms. We also found that TEAS, laser acupuncture and ACE were potentially effective in aiding cessation at mid-term or long-term follow up. Therefore, NTA therapies can play a complementary role in enhancing smoking cessation and relieving withdrawal symptoms.

Implications for research
The smoking cessation was measured by biochemically veri cation of abstinence rates in 6 trials [24-25, 27, 29, 44, 47] the remaining trials were self-reports.
Therefore, cotinine or exhaled CO veri cations of smoking cessation are warranted. Non-speci c acupoints stimulation was used as sham or placebo acupuncture in many trials. However, it was uncertain whether this kind of sham control was really inactive for cessation, and participants may identify sham acupuncture easily. Hence, programmatic RCT may be more appropriate to evaluate the effect of NTA therapies for smoking cessation. Additionally, long-term follow up data was rarely reported in the included trials, and needs to be improved in future studies.

Conclusion
Low to moderate certainty evidence suggests that acupressure, TEAS, laser acupuncture and ACE maybe effective in achieving short-term, middle-term or longterm smoking cessation. And acupressure was also bene cial to relieve withdrawal symptoms and decrease the level of exhaled CO. However, intradermal needle was found ineffective in aiding cessation at any time point. No serious adverse events were reported in the included trials. However, large sample size, fully reported and long-term follow up RCTs are warranted to con rm these effects. Risk of bias summary Figure 3 Acupressure for short-term smoking cessation. ACP: acupressure