Background

Metabolic dysfunction-associated fatty liver disease (MAFLD) is the term suggested in 2020 to replace non-alcoholic fatty liver disease (NAFLD).1 In 2023, the consensus used metabolic dysfunction-associated steatohepatitis (MASH) replaced non-alcoholic steatohepatitis (NASH).2 It refers to fatty liver disease related to systemic metabolic dysregulation. MAFLD has emerged as the most common chronic liver disease worldwide, with a reported global prevalence as high as 32.4%.3 The increased prevalence and mortality rates of MAFLD across the globe are alarming, and there is a significant economic burden on healthcare systems.4

MAFLD includes hepatocellular injury, inflammation and fibrosis, which may then progress to cirrhosis and its complications, including hepatocellular carcinoma (HCC). Current evidence on the natural history of MAFLD before developing hepatic decompensation indicates that the disease is reversible. The disease’s histological features (steatosis, inflammation and fibrosis) can improve under effective treatment.5 Resmetirom, as the first drug, has been approved by the US Food and Drug Administration (FDA) for NASH.6 Despite the promising hepatoprotective effect of resmetirom on histological liver endpoints, the adverse events (AEs), ethnic differences, effectiveness and cost-effectiveness, production scalability, social acceptance and accessibility need to be noticed.7 To date, critical therapy is still losing weight according to several guidelines. The lack of effective interventions for MASH makes lifestyle modification even more crucial, such as diet management and physical exercise. However, these methods are complicated for many people to implement.8 As a result, a convenient and effective adjunctive therapy with minimal side effects is needed.

Acupuncture has been used to treat obesity in China for many years. Previous studies indicated acupuncture could lose weight rather than a placebo effect.9 The meta-analysis, including eight randomised controlled trials (RCTs) with 939 patients, showed that acupuncture or acupuncture plus conventional medicine was superior to standard of care alone in improving overall clinical efficacy for MAFLD/NAFLD. Still, it has insufficient methodological quality and sample size.10 Popular acupuncture techniques include electroacupuncture (EA) (mild electric stimulation of acupuncture needles) and manual acupuncture (MA) (acupuncture stimulation by hand). A meta-analysis revealed that EA effectively decreased alanine transferase (ALT), aspartate transferase (AST), total cholesterol, and triglyceride (TG), which were significant. However, MA did not affect ALT, AST and TG, and these results may be explained by the lower stimulation intensity of MA.11 Then, there were no studies comparing EA versus no intervention or placebo/sham acupuncture (SA); the specific effect of EA for MAFLD/NAFLD was unclear. In the past, we designed a pilot study to reveal the positive impact of EA for NAFLD.12 Due to the small sample size and broken blind, the results were not full of reliability. Therefore, a multicentre, randomised, sham-acupuncture controlled, patients-blinded study will be designed to provide strong evidence of EA for MAFLD.

MethodsStudy design

This is a multicentre, randomised, patient-assessor blinded, sham-acupuncture controlled trial. This study will be conducted at Beijing Friendship Hospital, Capital Medical University, Dongzhimen Hospital, Beijing University of Chinese Medicine and Tongzhou Chinese Medicine Hospital. The Research Ethics Committee of Beijing Friendship Hospital has approved the study protocol (version 1.0, May 2022) with an approval number of 2022-P2-111-01. The Research Ethics Committee of Dongzhimen Hospital has approved the study protocol (version 1.0, May 2022) with an approval number of 2022DZMEC-246. Dongzhimen Hospital and Tongzhou Chinese Medicine Hospital have identified ethics with each other. It was registered at the Chinese Clinical Trial Registry (http://www.chictr.org.cn) on 29 May 2022 with the registration number ChiCTR2200060353. The trial started on 1 June 2022 and will be completed on 31 December 2025.

The study will recruit patients with MAFLD who meet the inclusion criteria and randomly assign them to each of two groups (EA and SA group) from Beijing Friendship Hospital, Dongzhimen Hospital and Tongzhou Chinese Medicine Hospital, receiving EA and SA. The included participants will modify their lifestyle into the 4-week run-in period after uniforming health education about eating and activating.8 After the run-in period, the participants will receive the second screen. Eligible participants will enter a 12-week treatment period in this trial. The details of the general condition, anthropometry parameters, life behaviour and medication for metabolic dysfunction will be recorded on the baseline. All treatments will be given three times a week (every other day) for 12 weeks. Participants will be assessed at the following time points: the baseline (the second screen), the middle of the treatment (after the first treatment, 4 weeks and 8 weeks after treatment), the end of the treatment (12 weeks after treatment) and follow-up (4 weeks after treatment finished). All participants will complete the assessments using MRI, blood tests, anthropometry and questionnaires about eating and physical behaviour. A flow diagram of the trial is shown in figure 1. The detailed trial process is seen in table 1.

Figure 1

Flow chart. BFH, Beijing Friendship Hospital, Capital Medical University; DZM, Dongzhimen Hospital, Beijing University of Chinese Medicine; EA, electroacupuncture; ITT, intent-to-treat; PP, per protocol; SA, sham-acupuncture; TZH, Tongzhou Traditional Chinese Medicine Hospital.

Table 1

Schedule of enrolment, intervention and assessments (SPIRIT figure)

Participant recruitment

Eligible patients will be recruited from outpatient clinics through advice from doctors and advertisements on posters, newspapers and WeChat. In the first screen criteria, eligible patients will enter the run-in period. Participants in this study include men and women between the ages of 18 and 65. MAFLD diagnosis was based on the Asia-Pacific region MAFLD clinical diagnosis and treatment guidelines13; ALT≥1.5× ULN appeared twice within 3 months; body mass index (BMI) ≥25 kg/m2; sign the informed consent form. After 4 weeks, the eligible patients will receive the second screen. Participants received the abdomen MRI scan and the liver fat fraction (MRI-PDFF) ≥8% after the run-in period.

Critical exclusion criteria included hepatitis B, C and other liver diseases, advanced liver cirrhosis and decompensation, and evidence of HCC. Patients receiving acupuncture within 1 month before enrolment were also excluded.

A complete list of the inclusion and exclusion criteria in the first and second screens was provided in the supplement table 1 of online supplemental additional file 2.

Randomisation and allocation concealment

Eligible participants will be randomised in a 1:1 ratio to receive EA or SA. Block randomisation will be used in this trial. The block size is 6. A biostatistician who did not participate in this trial will create a randomisation sequence. An independent research assistant will use a centralised block randomisation procedure by Interactive Web Response System in the WeChat programme (from the Ashermed contract research organisation in Shanghai) to implement the allocation schedule. The random number will be assigned after the participants have met all inclusion criteria and completed baseline assessments. To ensure the allocation concealment, acupuncturists will enter the WeChat programme to register information about patients before the first treatment, and the randomised number will be received immediately. Before randomisation, trial information will be provided to each eligible patient. Researchers (doctors) will get written informed consent from eligible patients willing to participate.

Blinding

Except for acupuncturists, patients, research assistants, outcome assessors and statisticians will be blinded to the group assignment. To minimise known sources of bias, the number/or location of the acupuncture points and the site of the electric needle stimulator will be similar in the EA and SA groups, thereby optimising the blinding of subjects. To test the success of patient blinding, all participants will be asked by their acupuncturists whether they received EA or SA after the first and end of the treatment. Similarly, to avoid the subjective bias of researchers and subjects, the outcome assessors, the research assistant for collecting data and statisticians charged with the analysis of the whole experiment will be blinded. The two groups will be identified as groups 1 and 2 during the statistical analysis. The participant’s allocated intervention will not be revealed until the statistical analysis reports are finalised.

Interventions

Each participant will be scheduled for 36 treatment sessions, 30 min for each session, three times a week over 12 weeks. Besides, all enrolled participants will receive primary health education about diet and lifestyle in the run-in period and get through the second screening in the allocation period. Acupuncture will be discontinued if the patients experience any serious adverse events after acupuncture.

Electroacupuncture

Patients allocated to the EA group with disposable and sterile acupuncture needles (length: 40 mm, diameter: 0.30 mm; HuanQiu, Suzhou, China) will receive treatment with acupuncture three times a week for 12 weeks for a total of 36 treatment sessions. Acupuncture experts predefined acupoints for MAFLD according to the Traditional Chinese Medicine theory and clinical experience. In this group, acupoints consist of Zhongwan (CV12), Guanyuan (CV4), bilateral Tianshu (ST25), Daheng (SP15), Zhangmen (LR13), Hegu (LI4), Zusanli (ST36), Sanyinjiao (SP6) and Taichong (LR3). Locations of acupoints are all according to the WHO Standard Acupuncture Locations14 shown in table 2 and figure 2A. Except bilateral Zhangmen (LR3) will be horizontally punctured 10–20 mm, these acupoints will be perpendicularly punctured 20–30 mm. After insertion, twirling, lifting and uniform reinforcing-reducing manipulation will be performed on all needles to elicit ‘de-qi.’ The compositional sensation of ‘de-qi’ includes numbness, soreness, distention, aching and heaviness, which are believed to be an essential component of acupuncture efficacy. Then, the needles will be removed after a 30-min session with clean cotton balls to prevent bleeding.

Figure 2

The points used in electroacupuncture (EA) and sham acupuncture (SA) groups. (A)The blue points are the acupoints used in the EA group, and (B) the orange points are the non-acupoints used in the SA group.

Table 2

Location of points in EA group and SA group

EA will be applied to bilateral Tianshu (ST25) and Daheng (SP15) on the abdomen at a dilatational wave of 2 Hz and an intensity of 0.1–1 mA depending on the participant’s comfort level through an electric needle stimulator (SDZ-II 6-Channel Programmable Electro-acupuncture) for 30 min. The intensity is adjusted to a level at which patients feel comfortable.

Sham acupuncture

The participants in the SA group will also receive the same treatment duration and frequency of sessions as the EA group. Similarly, SA will be performed at predefined bilateral sites (non-acupoints 1–8) that do not correspond to EA group traditional points or meridians, the locations shown in table 2 and figure 2B. These points will be superficially punctured 2–3 mm. SA will be applied to bilateral NA3 and NA4 on the abdomen at a dilatational wave of 2 Hz through an electric needle stimulator (SDZ-II 6-Channel Programmable Electro-acupuncture) for 30 min. Still, the electric current will be cut-off.

The difference between EA and SA will be revealed in figure 3.

Figure 3

The manipulation used in electroacupuncture (EA) and sham acupuncture (SA) groups. (A) The photo of EA treatment: the needle was moderately inserted into the acupoints and manipulated by twirling and lifting to elicit ‘de-qi’, the two lines from the electric needle stimulator and the current was turned on. (B) The photo of SA treatment: the needle superficial inserted into the non-acupoint and no twirling and lifting, the two lines from the electric needle stimulator and the current were turned off.

OutcomesPrimary outcomes

The primary outcome is the changes in relative liver fat content measured by MRI-PDFF between baseline and 12 weeks after intervention. The detailed calculation is (MRI-PDFF at week 12 - MRI-PDFF at baseline)/MRI-PDFF at baseline ×100%.

MRI-PDFF is a quantitative imaging biomarker that enables accurate, repeatable and reproducible quantitative assessment of liver fat over the entire liver.15 Some trials supported the use of MRI-PDFF in non-invasive monitoring of treatment response in early-phase NASH clinical trials.16

In this trial, MRI-PDFF, as an essential outcome, needs a strict quality control process. MRI was performed in the morning during the second screening, and all patients were fasting for more than 12 hours. They will undergo non-contrast scans using a 3.0-T MRI scanner (MR750, GE Healthcare, Waukesha, Wisconsin, USA) with an eight-channel phased-array body coil centred over the liver. The whole liver was covered during the axial IDEAL IQ examination.17 The details of the measurement of PDFF are shown in supplement figure 1 of online supplemental additional file 2.

Secondary outcomes

The secondary outcomes included the patients with a ≥30% relative decline in MRI-PDFF and the change of liver stiffness by MRE, the change of liver and metabolic biomarkers and anthropometry parameters at week 12, and the relative change of MRI-PDFF, the change of MRE and anthropometry parameters at week 16.

MRE measurement uses a two-dimensional spin-echo–echo planar imaging MRE sequence, and the MRI scanner is the same as the MRI-PDFF. Our group’s recent publication showed MRE sequence parameters.18 The MRE measurement details are shown in supplement figure 2 of online supplemental additional file 2.

Liver and metabolic biomarkers including liver enzyme levels (ALT, AST, γ-glutamyl transferase (GGT) and ALP); glucose metabolism (fasting blood glucose (FBG), fasting insulin, glycosylated haemoglobin % and homeostatic model assessment of insulin resistance (HOMA-IR) index score); lipid levels (CHOL, TG, high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C)).

After a minimum of 12-hour fasting, elbow venous blood will be taken in the early morning to determine the ALT, AST, GGT, ALP, cholesterol, TG, HDL-C, LDL-C and FBG. The serum will be measured according to standard procedures by an autoanalyser (BECKMAN AU5800, US) at the central laboratory of Beijing Friendship Hospital to analyse the samples. Fasting insulin will be measured by electrochemiluminescence (Roche, Germany), and HbA1c% will be assessed by high-performance liquid chromatography (VARIANT, BioRad Lab., Hercules, California, USA). HOMA-IR will be used as the index of insulin resistance based on the following formula: HOMA-IR = (fasting serum insulin [mIU/mL] × fasting serum glucose [mmol/L]/22.5).19

Anthropometry parameters include the BMI and waist–hip ratio (WHR). Body weight, waist and hip circumference will be measured while the participants wear light clothing and no shoes after more than 12 hours of fasting. Height without shoes will be measured using a stadiometer. BMI, kg/m2, will be calculated as weight divided by the square of height. WHR will be calculated as waist circumference divided by hip circumference.

In this trial, MRI-PDFF and MRE will also be measured at baseline, week 12 and week 16; the liver and metabolic biomarkers and anthropometry parameters were assessed at baseline and week 12.

Exploratory outcomes

Exploratory outcomes including the score of several questionaries about quality of life (Chronic Liver Disease Questionnaire, CLDQ-NAFLD/NASH),20 the behaviour of eating (the Three-Factor Eating Questionnaire-21, TFEQ-21),21 physical activity (International Physical Activity Questionnaire-7, IPAQ-7)22 and attitude of behaviour change (the University of Rhode Island Change Assessment scale, URICA).23 These questionnaires will be assessed at baseline, week 4, week 8, week 12 and week 16.

Blinding assessment

To test the patient-blinding effects, all patients will be asked to guess whether they have received EA or SA at the end of the first treatment session and the 36th treatment session.

Credibility and expectancy

The credibility and expectancy of patients will be assessed using the Credibility/Expectancy Questionnaire at the end of the first treatment session and the 36th treatment session. The questionnaire is a quick and easy-to-administer scale for measuring treatment expectancy and rationale credibility for clinical outcome studies.

Adverse events

In this trial, participants or acupuncturists monitored and recorded any treatment-related and unrelated AEs. Common side effects of EA or SA include (but are not limited to) subcutaneous haematoma, local bleeding, skin bruising, needle site itching, continuous needle site pain, muscle spasms and dizziness. In a serious AE, the necessary measures will be taken immediately for the subject’s safety. The time of occurrence, severity, duration, outcome and treatment-related status will be reported to the Medical Ethics Committee within 24 hours.

Data management

The EDC system provided by Aisha Medical Company is used to collect data, and there are corresponding data management measures in the clinical research and data processing stages. Data will be collected clearly and entirely in an electronic case report form (eCRF). All original data sources will be preserved, including original records on paper, informed consent forms, inspection results and acupuncture records sheets. All researchers will receive training regarding data management. The research assistant is responsible for verifying the accuracy and integrity of data to prevent any detection of errors, omissions or items requiring changes or clarification and oversees the completed eCRF in a timely, truthful, detailed and earnest manner. No information about the identity of the participants will be disclosed in source documents. The data administrators will implement data lockup on completion of the study. In all scenarios, the physician must maintain source documents for each participant in the study, consisting of case and visit notes (hospital or clinic medical records) containing demographic and medical information, as well as the results of any other tests or assessments. All data related to the trial will be saved for at least 3 years after publication. Readers can access the original data by contacting the corresponding author.

An independent Data and Safety Monitoring Committee (DSMC) is established to review and interpret data generated from the study, and a meeting will be held every 3 months to report the research progress. Each protocol amendment will conform to good clinical practice principles and maintain the ethical standards for RCTs. The DSMC will review the progress and decide whether a premature closure is needed.

Quality control

To ensure the safety and proper operation of acupuncture, all acupuncturists will participate in the standard operation process (SOP) of skin disinfection, location of acupoints and non-acupoints, fundamental puncture and needle manipulations before the trials. Licensed acupuncturists will perform treatments with at least 3 years of experience in acupuncture.

To ensure the results are objective, veritable and accurate, all the assessors about outcome measurements will be blinded to the allocation of participants. The MRI-PDFF and MRE assessment must be scanned in the central lab at Beijing Friendship Hospital. Two radiologists will read the images independently, followed by SOP, and the average of two measurements will be adopted for analysis. The liver and metabolic biomarkers will be tested in the central lab at Beijing Friendship Hospital. SOP will strictly follow the assessor in subcentres of anthropometry parameters at each visit.

The statistics will be blinded to the group assignment to ensure a reliable conclusion. The statistical analysis identified these two groups as groups 1 and 2.

The study protocol has been reviewed and revised by experts in hepatology, acupuncture, statistics and methodology. All study researchers will attend a training series to promote the implementation. Data will be monitored on three levels by the head of subcentres, researchers from the Beijing Friendship Hospital, Capital Medical University and the DSMC preset for this study. The Bioethics Committee of Beijing Friendship Hospital, Capital Medical University, will audit the study independently regularly.

Statistical methodsSample size calculation

The primary outcome is the changes in relative liver fat content measured by MRI-PDFF between baseline and 12 weeks after intervention. Considering it is a continuous variable, we use the two-sample t-tests assuming equal variance for calculation according to the formula:

Based on the previous pilot study12 and the clinical experts’ opinions, the mean changes in relative liver fat content estimated in the EA and SA groups were 28%±20% and 16%±20%, respectively. For an equivalence trial, two-tailed α of 0.05, β=0.1 and assuming 20% dropout. We concluded that we needed to recruit 144 patients (72 per group) for this trial to ensure statistically significant results. The calculation was performed using the PASS software (V.11.0).

Statistical analysis

All analyses will be performed using the intent-to-treat (ITT) and the per-protocol (PP) populations. ITT requires all participants to receive a baseline assessment of the primary outcome and at least one acupuncture session or SA administration. Assuming that the occurrence of missing data was random, we used multiple imputations by chained equations to impute one complete dataset with five interactions. The imputation models included all variables under analyses.24 The PP population is usually defined as patients who complete at least 80% of the treatment protocol without significant protocol violations.

All efficacy and safety analyses will be conducted according to the ITT principle. Descriptive analyses were performed on all baseline variables. Continuous variables whose distribution met normality assumptions were given as mean±SD or else given as medians and quartiles (quartile 1 (Q1), quartile 3 (Q3)); categorical variables were presented as the absolute value and relative frequency. Baseline differences between the groups will be assessed using the Student’s t-test for normally distributed continuous variables and the non-parametric Mann-Whitney U test for non-normally distributed variables. The χ2 or Fisher’s exact test will be used for categorical variables. Changes in IPAQ-7, TFEQ-21, URICA and CLDQ-NAFLD at each evaluation time point from baseline will be analysed using the analysis of covariance model, followed by the Bonferroni t-test to detect the differences between the two groups at each time point. The incidence of AEs will be examined using the χ2 test.

The analyses were implemented with SAS V.9.4 (SAS Institute). P<0.05 was considered statistically significant.

Discussion

MAFLD is common, affects a quarter of the population and causes a considerable burden for patients and society. According to traditional Chinese medicine’s theory and clinical experiences, EA may be an effective therapy for MAFLD.11 12 The mechanism of EA anti-inflammation,25 anti-insulin resistance26 and anti-hyperlipidaemia27 were revealed in primary studies. However, high-quality clinical evidence was still lacking.

In previous studies about MAFLD, no well-designed studies compared the efficacy of acupuncture versus placebo/SA on variables measured by internationally accepted modalities and chose the internationally recognised measurement as the outcome.28 Acupuncture is increasingly accepted by patients in many parts of the world, as accumulating high-level evidence on the effect and safety of acupuncture is recognised.29–31 However, the sham device had more significant effects than the placebo pill on self-reported pain and severity of symptoms for the entire course of treatment.32 Therefore, a distinction must be made between overall treatment responses and placebo responses. We still found that the median reduction of relative liver content by MRI-PDFF was −15.8% by the SA group in the pilot study,12 which is higher than the placebo response (−2.4%) in trials for new therapeutic agents.33 These facts indicated the necessity to have SA as a control to avoid overestimation of the therapeutic efficacy of EA. The SA in this trial is superficially non-acupoint, which allows the participants to feel the skin penetration and recognise they received the acupuncture treatment. In our experience, the type of SA is easy to handle and blind.29 34

Moreover, previous studies didn’t adopt the recognised measurement by industry experts as the outcome. In December 2018, the FDA issued guidance for clinical trial design for patients with non-cirrhotic NASH, permitting the use of MRI-PDFF as the method for enrolment and evaluating therapeutic effect for early-phase clinical trials in NASH instead of liver biopsy.16 Therefore, our study would choose the relative change of liver fat content by MRI-PDFF after treatment as the primary endpoint to evaluate the therapeutic efficacy of EA for MAFLD.

Above all, we designed this multicentre, randomised, sham-acupuncture controlled, patients-blinded, assessor-blinded study to investigate the efficacy and safety of EA on reduction of relative liver fat content measured by MRI-PDFF in patients with MAFLD. This trial still has limitations: first, the acupuncturists will not be masked due to the responsibility of providing the intervention. However, the patients and outcome assessors will be blinded to reduce the bias for the results. Second, this study has multicentres, effectively avoiding the researcher effect. However, they are all located in Beijing, and regional impact will take a lot of work to avoid. Finally, the long duration of the treatment may compromise the participants’ compliance.

At the end of this trial, we hope the results will provide more reliable and robust evidence of EA as an effective and safe therapy for MAFLD.

Trial status

This trial is currently recruiting patients.

Ethics and dissemination

This study was approved by the institutional review board and ethics committee of Beijing Friendship Hospital, Capital Medical University (no. 2022-P2-111-01) and Dongzhimen Hospital, Beijing University of Chinese Medicine (no. 2022DZMEC-246) conformed to the Declaration of Helsinki. It was registered at the Chinese Clinical Trial Registry (http://www.chictr.org.cn) on 29 May 2022 with the registration number ChiCTR2200060353. Before participation, all patients provided written informed consent (online supplemental additional file 3). All personal information from potential and enrolled participants out of the scope of this trial will not be collected, shared or maintained to protect confidentiality before, during and after the trial. Trial results will be published in peer-reviewed journals and disseminated to the media and the public.