Randomised clinical trial: an assessment of acupuncture on specific meridian or specific acupoint vs. sham acupuncture for treating functional dyspepsia


Correspondence to:

Prof. F. R. Liang, No.37, Shier Qiao Road, Chengdu, 86-610075, Sichuan, China.

E-mail: lfr@cdutcm.edu.cn



Functional dyspepsia (FD) is a common disease without an established optimal treatment.


To determine (i) the effect of acupuncture in relieving FD symptoms and improving life quality; (ii) the effect difference between acupoint and non-acupoint; and (iii) the effect difference among different acupoints.


A total of 712 eligible patients were included and randomly assigned to six groups (Group A: specific acupoints of the stomach meridian; Group B: non-specific acupoints of the stomach meridian; Group C: specific acupoints of alarm and transport points; Group D: specific acupoints of the gallbladder meridian; Group E: sham acupuncture of non-acupoints; and Group F: itopride). A treatment period of 4 weeks (continuous five sessions per week), and a follow-up period of 12 weeks were arranged. The outcomes were the (i) patients’ response, (ii) symptoms improvement measured using the Symptom Index of Dyspepsia and (iii) quality-of-life improvement based on Nepean Dyspepsia Index.


All groups had an improvement in dyspepsia symptoms and the QoL at the end of treatment, and the improvement was sustained for 4 weeks and 12 weeks. The overall response rate was significantly higher in acupuncture group A (70.69%), and lower in sham acupuncture group (34.75%), compared with itopride and other acupuncture groups. Similarly, the difference in symptoms and QoL improvement was significant between group A and the other acupuncture groups.


Acupuncture is effective in the treatment of functional dyspepsia, and is superior to non-acupoint puncture. The benefit of acupuncture relies on acupoint specificity.


Functional dyspepsia (FD) is a common gastrointestinal disorder that is defined as epigastric pain or burning, postprandial fullness or early satiation in the absence of underlying organic disease, according to the Rome III criteria.[1] Globally, the prevalence of FD has been reported to vary between 11% and 29.2%[2]; and it reaches 45% in the outpatient department.[3] Although FD is not a life-threatening disease, patients with FD have a poor quality-of-life (QoL), which imposes an economic burden on society.

The classic pharmacologic treatments for FD include antacids, prokinetics, antidepressants and anxiolytics, but the benefit of these agents still remains unclear,[4-6] and patients do not gain sufficient benefit in improved QoL from the above treatments.[7] As the pharmacologic treatments for FD remain unsatisfactory, alternative therapies, including acupuncture, are attractive to both patients and practitioners.

Acupuncture for treating FD has been practiced in China for thousands of years, and is increasingly accepted worldwide.[6] However, previous evidence is minimal. As a meta-analysis on acupuncture for FD[8] indicated, because of the poor quality of the trials, the efficacy of acupuncture for FD has not been verified. Based on traditional meridian and acupoint theories, the efficacy of acupuncture is mostly determined by acupoint. There is a specific effect of acupoint compared with non-acupoint, and the specific effect may differ from acupoints on different meridians or of different types. Due to the heterogeneity of results from previous randomised controlled trials,[9, 10] this theory is not completely supported.

The aim of our study was as follows: (i) to determine the efficacy of acupuncture for patients with FD compared with itopride and its safety; (ii) to investigate the specific effect of acupoint compared with that of the non-acupoint; and (iii) to explore whether there is acupoint specificity by comparing different acupoints or not.

Patients and Methods


Outpatients with FD between 18 and 65 years old were considered eligible for inclusion based on the Rome III criteria,[1] which suggested that the predominant FD symptoms should include epigastric pain or burning, postprandial fullness or early satiation.

Patients were recruited by gastroenterologists. Before inclusion, patients underwent a physical examination, routine blood tests, measurement of fasting blood glucose, ECG, liver and kidney function tests, abdominal ultrasonography and upper gastrointestinal endoscopy to exclude structural or biochemical abnormalities. Helicobacter pylori non-infected patients were included. The H. pylori infected patients who still suffered from standing dyspepsia symptoms were also included only if eradication treatment was completed 4 weeks before enrollment. Exclusion criteria were the presence of a co-existing serious illness, pregnancy and gastroesophageal reflux disease (GERD). Patients taking medications that are known to alter gastric function or influence the outcome assessment or using acupuncture during the past 1 month were also excluded.

Study design

This study was a randomised, controlled trial conducted at eight hospital outpatient departments in China between April 2008 and October 2009. The protocol was designed and passed the ethics approval in 2007, before the performance of this trial.[11]

The study was conducted under the rules of the Declaration of Helsinki, and approval was received from the Ethics Committee of Chengdu University of TCM. All patients provided written informed consent before participation. The trial received financial support from the National Basic Research Program of China (973 Program; No. 2006CB504501), and was given a unique registration number (NCT00599677) at http://www.clinicaltrial.gov/.

The study consisted of a run-in period prior to randomisation, a treatment period of 4 weeks (continuous five sessions per week with an interval of 2 days between 2 weeks), and a follow-up period of 12 weeks. Approximately 7 days after diagnosis of FD, which was at the end of the run-in period, dyspeptic symptoms and disease-specific QoL were evaluated to determine the patient's baseline status. Patients were randomly allocated to one of four acupuncture groups (groups A, B, C and D), a sham acupuncture group (group E) or an itopride group (group F) in a 1:1:1:1:1:1 ratio by central randomisation.

Randomisation and Blinding

In this trial, a central randomisation was performed by the Chengdu Good Clinical Practice (GCP) Center, which is affiliated with the National Clinical Trial Center of Chinese Medicine. The allocation sequence was generated by a permuted-block randomisation procedure to ensure the blindness of group assignments. The full randomisation process was pre-programmed and carried out by the central computer system.

All of the investigators’ mobile phone numbers were pre-approved and registered by GCP. Once an eligible patient was included, the investigator would send a mobile phone text, including the name, birthday and gender to the GCP Center computer system. A unique random number and group assignment were automatically sent back. This procedure assured complete randomisation and adequate allocation concealment. In addition, investigators received a confirmation email at the same time, containing a random number, the group assignment code and the patient's basic information. This feedback email was then printed and attached to the corresponding case report form (CRF).

Patients in groups A, B, C, D and E were blinded to which point-puncturing they would receive. The outcome assessors and statistical analysts were unaware of the intervention assignments throughout the trial.


Patients were asked not to take antacids (such as PPIs and H2 blockers), prokinetics, antidepressants, anxiolytics or TCM herbs focusing on relieving FD symptoms during treatment. The name and dosage of medication was documented, if any FD-related medications were taken without improvement.


Acupuncture was performed by certified acupuncturists who had a TCM licence and had worked for at least 2 years in clinics. Before the trial, all the acupuncturists were required to take special training to acquire a full understanding of the performance of treatment, and only those who passed the examination were qualified to take part in the trial. The training included the method of central randomisation, the location of acupoints (Table 1), the correct manipulation of needling and the operation of an electro-acupuncture machine.

Table 1. Locations of verum acupoints and non-acupoints
Acupoints in group AAcupoints in group BAcupoints in group CAcupoints in group DNon-acupoints in group E
Chongyang (ST42)Tiaokou (ST38)Weishu (BL21)Qiuxu (GB40)At the medial arm on the anterior border of the insertion of the deltoid muscle at the junction of the deltoid and biceps muscles
Fenglong (ST40)Dubi (ST35)Zhongwan (CV12)Guangming (GB37)Halfway between the tip of the elbow and the axillae
Zusanli (ST36)Yinshi (ST33) Yanglingquan (GB34)Ulnar side, halfway between the epicodylus medialis of the humerus and the ulnar side of the wrist
Liangqiu (ST34)Futu (ST32) Waiqiu (GB36)The edge of the tibia 1–2 cm lateral to the Zusanli (ST36) horizontally

Acupuncture groups

In group A, specific acupoints on the stomach meridian (ST42, ST40, ST36 and ST34) were selected according to the meridian and acupoint theories, and a review of the ancient and modern literature. In group B, non-specific acupoints on the stomach meridian (ST38, ST35, ST33 and ST32) were chosen. These acupoints were located near acupoints in group A with the purpose of investigating the difference in efficacy between specific and non-specific acupoints on the same meridian. In group C, specific acupoints (BL21 and CV12) were selected. This acupuncture prescription was common for FD treatment. In group D, acupoints (GB40, GB37, GB36 and GB34) on the gallbladder meridian were used to determine the difference in meridian-related efficacy (Figure 1).

Figure 1.

Locations of verum acupoints and non-acupoints.

Except for CV12, the acupoints were punctured unilaterally by sterile disposable stainless needles (40 mm or 25 mm in length and 0.25 mm in diameter; Hwatuo, Suzhou, China, PRC). The needles were alternately placed on the left and right to relieve pain and reduce tolerance. The depth of puncture was described in the protocol, according to the location of the acupoints. Twirling and rotating, lifting and thrusting manipulation was applied to promote qi arrival. After the arrival of qi, auxiliary needles (13 mm in length and 0.18 mm in diameter) were punctured 2 mm lateral (proximal limbs) to each acupoint at a depth of 2 mm without manual stimulation. Then, a Hans transcutaneous electric acupoint stimulation (TEAS) machine (LH 200A; Nanjing, China) was used. Electrodes were applied on each acupuncture needle and the auxiliary needle for 30 min with a stimulation frequency of 2/100 Hz and an intensity between 0.5 mA and 1.5 mA. The acupuncture treatment consisted of 20 sessions over a period of 4 weeks (one session per day, five continual sessions per week, 2 days interval between 2 weeks).

Sham acupuncture group

Non-acupoints with a shallow puncture was performed as sham acupuncture group. The location of non-acupoints has been described in our previous study[12] (Figure 1). Non-acupoints were punctured perpendicularly, 0.5–1 cm unilaterally. After insertion, twirling and rotating, lifting and thrusting, manipulation was performed six times to make patients experience an identical procedure to true acupuncture without considering qi arrival. Then, the use of auxiliary needles and the Hans TEAS machine was the same as in the true acupuncture groups.

Itopride group

As disturbances in gastrointestinal motility and sensory function were regarded to be closely related with the symptoms in FD patients, itopride was considered as an effective prokinetics with effects on gastrointestinal motility promotion and gastric accommodation.[13] Patients in group F only received itopride treatment. They were instructed to take 50 mg of itopride three times daily (provided by Qinghua Yuanxing Pharmaceutical Co. Ltd, Shenzhen, China), 30 min before meals. The medication was given for 4 weeks with 20 sessions (one session per day, five continual sessions per week, 2 days interval between 2 weeks).

Outcome measures

As the Nepean Dyspepsia Index (NDI) and Symptom Index of Dyspepsia scale forms were self-report questionnaires, patients were asked to complete them by themselves. Designated assessors were trained to provide questionnaires guidance to every patient, explain or answer questions when necessary and check whether the questionnaires were completed. When a face-to-face interview was administered by an assessor for poor vision or uneducated patients, assessors should only read each item of the scales without asking leading questions, for the purpose of avoiding results difference between patient self-report and external assessment. The assessment was made by telephone during the follow-up period.

The primary outcome was the patients’ response based on the improvement of Symptom Index of Dyspepsia. The Symptom Index of Dyspepsia was a four-rate scale with a ranging score of 0–4 of each dyspepsia symptoms (postprandial distension; early satiety; epigastric pain; and epigastric burning). A score of 0 indicated not present, 1 indicated occasional and mild symptoms without interfering with social activities, 2 indicated prolonged and obvious symptoms interfering with work or rest, and 3 indicated severe symptoms with an impact on work or rest. The improvement of at least two scores or no occurrence of any symptom included in the Symptom Index of Dyspepsia scale was regarded as the positive response.

The secondary outcome were the change in 25-item NDI quality-of-life and Symptom Index of Dyspepsia score at the end of treatment relative to baseline. The 25-item NDI[14] measured the specific life quality of FD patients in four domains, namely, interference (13 items), know/control (seven items), eat/drink (three items) and sleep/disturb (two items). The scores were measured with a 5-point likert scale that ranged from ‘not at all’ to ‘extremely’. We had translated the 25-item NDI into Chinese; its reliability and validity was assessed before this trial.[15] To compare the treatment effect differences on disease-specific quality-of-life, we calculated a total quality-of-life score in NDI according to a formula of Talley et al., with higher scores indicating higher quality-of-life.[16]

Statistical analysis

Sample sizes

According to the literature,[17] an improvement of 18.0 on the NDI index score after treatment with 50 mg of itopride was reported. As there was no reference to the change in the NDI score by acupuncture for FD in the extant literature, we anticipated an improvement of no less than 15.0 on the NDI index score after acupuncture treatment in this trial. Standard deviation was defined as 22.0, as described in a previous study.[17] As a result, the sample size in this trial was estimated according to the following formula, using 90% power and a 5% level of significance. Each group needed at least 120 patients to allow for a 15% withdrawal rate.

Data analysis

All of the data were filled in electro-CRF twice by different persons, and the difference between the first and second electro-CRF fill-in was checked automatically by computer to maintain accuracy.

The data were analysed using spss 16.0 (SPSS Inc, Chicago, IL, USA). All of the data analysis was based on an intention-to-treat population, replacing missing data by the last-observation-carried-forward method. Chi-squared tests of independence were used to determine the existence of significant differences between groups in the proportion of responders. For continuous measures (i.e. for the change in mean dyspepsia scores and NDI scores from baseline to each visit time), we used parametric statistics (t-tests and anova). A two-sided test was applied for all available data, and a P-value < 0.05 was considered statistically significant.

Quality control

A strict quality control system was maintained throughout our clinical study to ensure a reliable result. First, all acupuncturists were trained to use the centre randomisation method, to fill in the CRF and electro-CRF, to locate the points and manipulate the needles and to use the Hans machine to maintain homogeneity of the trial process. In addition, two levels of a quality inspection system, including quality examination and quality monitoring, were established to ensure management in compliance with the Standard Operation Procedure (SOP). Qualified quality inspectors∕monitors, who were blinded to group assignment, provided quality control reports in written form once∕every 3 months in each hospital respectively.[18]


Patient characteristics

Between April 2008 and October 2009, 720 patients with FD were randomly assigned. Four non-FD patients were erroneously randomised, and four patients were randomly assigned repeatedly. Six patients violated the treatment protocol (two in group A, two in group C and two in group E). The intention-to-treat population comprised all remaining 706 patients (Figure 2). Table 2 displays the patients’ baseline status in each group. The baseline was comparable with respect to age, gender, number of patients with postprandial distress syndrome (PDS)/epigastric pain syndrome (EPS), duration of disease, total Symptom Index of Dyspepsia score and NDI quality-of-life score.

Figure 2.

Trial flow chart.

Table 2. Patient characteristics at baseline
 Group AGroup BGroup CGroup DGroup EGroup F
  1. PDS, postprandial distress syndrome; EPS, epigastric pain syndrome.

No. patients116120114119118119
Age, year (mean ± s.d.)38.1 ± 13.535.2 ± 14.037.4 ± 13.138.2 ± 13.236.8 ± 13.136.2 ± 13.9
Women, no. (%)83 (71.6)79 (65.8)82 (71.9)88 (73.9)84 (71.2)82 (68.9)
Diagnosis, no. (%)
PDS79 (68.1)82 (68.3)74 (64.9)85 (71.4)71 (60.2)74 (62.2)
EPS37 (31.9)38 (31.7)40 (35.1)34 (28.6)47 (39.8)45 (37.8)
Duration of disease, mh (mean ± s.d.)74.2 ± 73.962.6 ± 54.271.1 ± 63.468.2 ± 56.864.4 ± 75.665.9 ± 66.1
Concomitant irritable bowel syndrome no. (%)25 (21.6)28 (23.3)25 (21.9)24 (20.2)28 (23.7)23 (19.3)
Summary Symptom Index of Dyspepsia (mean ± s.d.)4.26 ± 1.544.48 ± 1.774.38 ± 1.624.69 ± 1.544.27 ± 1.744.24 ± 1.78
Nepean Dyspepsia Index (mean ± s.d.)73.7 ± 11.275.4 ± 10.674.4 ± 11.173.1 ± 8.975.7 ± 10.776.7 ± 9.8

Response at the end of 4-week treatment

The response rate showed significant difference between six groups (Table S1). The overall response rate in acupuncture groups (A, B, C) and itopride group were significantly higher than in sham acupuncture group (70.69% in group A, 50% in group B, 51.75% in group C, 55.46% in group F vs. 34.75% in group E, P < 0.05). In addition, the data in Table S1 demonstrated that the overall response rate achieved in group A was significantly higher when compared with treatment in group B, C and group F (P < 0.05) (Supporting Information). The response rate was comparable between women (52.0%) and men (47.1%) (P = 0.236).With respect to the subcategory of FD, the response rate was differrent between groups in early satiation of PDS (P < 0.001), and group A was more efficacious than other five groups (45.6% vs. 29.3% in group B, 27.0% in group C, 16.5% in group D, 14.1% in group E and 23.0% in group F, P < 0.05) (Figure 3).

Figure 3.

Response rates in subgroup analysis of subcategory (week 4). Note: In early satiation of PDS, 45.6% patients in group A had a response (36 of 79), as compared with 29.3% in group B (24 of 82), 27.0% in group C (20 of 74), 16.5% in group D (14 of 85), 14.1% in group E (10 of 71) and 23.0% in group F (17 of 74) (P < 0.05).

Quality-of-life assessment during treatment (4 weeks)

We used the change in NDI score from baseline to indicate the improvement difference between groups, which was considered more clinically important. With regard to a change of at least 10 points on the NDI total scale corresponding to a clinically meaningful change in patient status,[19] although patients in all groups achieved QoL improvement (P < 0.001), only the change in groups A and C was significant from a clinical point of view (Table S2). With the exception of itopride, the effect of four acupuncture treatments (groups A, B, C and D) revealed significant superiority to sham acupuncture (all P < 0.05 vs. sham acupuncture). The difference between group A and groups B and D was significant (14.8 vs. 9.47 and 9.58, P < 0.001), but no significant difference was detected between acupuncture group C, and groups B and D (Supporting Information).

Symptom assessment during treatment (4 weeks)

The Symptom Index of Dyspepsia total score improved from baseline in all six groups (P < 0.001) (Table S2). The score change demonstrated that acupuncture groups (A, B, C and D) and itopride group were superior to sham acupuncture group (all P < 0.05 vs. sham acupuncture). Among acupuncture groups, only the change in group A reached statistical difference compared with itopride group (2.43 vs. 1.76, P = 0.005). No significant difference was detected between groups B, C and D, and group A had a significantly better effect than groups D (2.43 vs. 1.43, P < 0.001) and B (2.43 vs. 1.88, P = 0.020) (Supporting Information).


The continual benefits in symptoms and QoL persisted after treatments. Similar statistically significant differences were seen in the score change in Symptom Index of Dyspepsia and NDI scores at 4 weeks and 12 weeks after treatment (Figure 4 and 5).

Figure 4.

Mean changes from baseline in the Summary Score of Symptom Index of Dyspepsia at 4- and 12-week follow-ups. Note: Bars represent the standard error. The symptoms were continually improved 4 weeks and 12 weeks after treatment. Group E differed statistically with the other five groups 4 weeks after treatment, but the difference between group E and group F was not statistically significant 12 weeks after treatment. Group A showed superiority to other groups in the 4- and 12-week follow-ups.

Figure 5.

Mean changes from baseline in the Summary Score of NDI at 4- and 12-week follow-ups. Note: Bars represent the standard error. The quality-of-life was continually improved 4 weeks and 12 weeks after treatment. Group E differed statistically with groups A, B, C and D. Group A showed superiority to other acupuncture groups and group F.

Safety assessments

Ten adverse events (10/593) were reported in the acupuncture or sham acupuncture group, as follows: pain or haematoma at the puncture point (8/10); stomach distention after treatment (1/10); and mild fainting during needling (1/10). No severe adverse events were reported. One patient (1/119) had a mild headache in the itopride group, and the symptoms disappeared once stopping medication.


Our trial is one of the larger randomised controlled trials involving acupuncture for the treatment of FD, and the first study to explore acupoint specificity in different aspects. The results of our study indicate differences in improvement of FD symptoms and QoL between acupoints and itopride, acupoints and non-acupoints and different acupoints.

As a prokinetic drug, itopride was thought to be superior to placebo in the treatment of FD.[17, 20] In this trial, itopride was set as the positive control group to determine the effect of acupuncture. After 4-week treatment, 55.46 per cent of the Chinese FD patients receiving itopride were symptom-free or marked improved, which is consistent with the results from the RCT of Holtmann.[17] Compared with itopride, the response rate was higher in acupuncture group A, and lower in sham acupuncture, and similar differences were seen in the improvement of symptoms. The results indicate that although both acupuncture and itopride can improve dyspeptic symptoms, acupuncture on specific acupoints of the stomach meridian showed a better effect, especially in improving QoL. Due to a lack of observation of gastrointestinal motor or gastrointestinal hormone levels, we cannot determine the reason for the superior efficacy. However, based on previous studies, despite a stimulatory effect on gastrointestinal motility, acupuncture (especially ST36) probably exerts an anti-nociceptive effect on visceral hypersensitivity,[21] which may explain the difference in effect.

Although most of clinical studies have reported an effect of acupuncture, many of them[22-25] could not distinguish the effect of it from sham acupuncture, like the IBS trial from Lembo AJ and FD trial from Yang-Chun Park. The heterogeneity may be attributed to the cultural backgrounds of enrolled patients, sample and the trial design, including the location of non-acupoint, treatment sessions and frequency, selection of acupoint etc. Before this trial, we arranged a randomised controlled trial[12] by comparing two different methods of non-acupoint location, and finally defined the non-acupoint group. In our trial, in sham acupuncture group, only 34.76 percent of the patients were symptom-free or marked improved, which was significantly lower than that in acupuncture groups A, B, C and itopride. Besides, the improvement of QoL in NDI score in sham acupuncture group was non-clinically meaningful. The results indicate that although acupoint and non-acupoint puncturing brought relief of symptoms and improvement in QOL, acupoint was more effective.

Efficacy differences between acupoints were also detected. Compared with acupoint on the gallbladder meridian (group D) and non-specific acupoints of the stomach meridian (group B), acupuncture on specific acupoints of the stomach meridian (group A) showed a better effect in improving QoL and relieving symptoms, which is in accordance to meridian and acupoint theories. The results suggest the existence of relative specificity of acupoints, and are consistent with recent neuroimaging studies.[26-28] However, contrary to our expectations, group C (the combination of specific acupoints on the abdomen and back) did not show a better effect than groups D and B, suggesting the effect of acupoints proximal to the stomach differing from those on the limbs along the meridian. More studies are needed to verify the results and explore the reason for the differences.

The strengths of our study include a rigorous randomisation system to confirm adequate allocation concealment and comparability between groups, and a strict quality control system to assure the standardisation of the procedure and reliable results. The Hans TEAS was applied to each point and the nearby auxiliary needling point to unify stimulation parameters, as well as to localise the current effect.

There were limitations in our study. First, due to the purpose of our trial and in consideration of patient compliance, a waitlist control group was not designed, which may not exclude self-healing from the therapeutic effect. Second, the anxiety status of the patients was not measured in our study, which may have an impact on the assessment results.[29] Further trials should consider the influence of anxiety and perform a subgroup analysis.

In summary, acupuncture is effective in the treatment of FD and showed superiority over non-acupoint puncture. In addition, this trial provided evidence for the existence of specificity between acupoints on different meridians and of different kinds.


Declaration of personal and funding interests: None.