Efficacy and safety of acupuncture for functional constipation: a randomised, sham-controlled pilot trial

This was a randomised, sham-controlled, participant/assessor-blinded pilot trial conducted in a university hospital setting in South Korea from November 2013 to October 2015. Thirty participants with FC were registered and assigned to either the real acupuncture (RA) or sham acupuncture (SA) group; 15 participants were in each group. Twelve RA or SA sessions were performed over 4 weeks, and follow-up assessments were performed every 2 weeks for 4 weeks after intervention completion in each group. This study followed CONSORT guidelines (www.​CONSORT-statement.​org), and the CONSORT flow chart of the study is shown in Fig. 1.

Eligible participants were aged between 19 and 65 years old, of either sex, and met the Rome III criteria for FC.

The exclusion criteria included having other causes of constipation, a history of bowel surgery or cholecystectomy, other severe medical diseases (e.g., insulin-dependent diabetic mellitus, heart, lung, liver or kidney disease, or malignancy), hypersensitivity reaction, being pregnant, expecting to be pregnant, actively nursing, receiving acupuncture, moxibustion, cupping or herbal medicine treatments for constipation within the past 4 weeks, or taking any medications or probiotics for constipation within the past 2 weeks.

Eligible participants who were willing to comply with the study protocol and provided written consent were asked to maintain a defecation diary for a one-week screening period. This diary was used by Korean medical doctors (KMDs) at visit 1 to confirm that the patients satisfied the Rome III criteria.

This study was approved by the institutional review board (IRB) of the Daejeon Oriental Hospital, Daejeon University (djomc-112) and performed in compliance with the Declaration of Helsinki. All subjects were provided with verbal and written information on possible benefits, harms and the possible treatments for constipation other than the treatments offered in our clinical trial before signing the consent form. We obtained written informed consent from the subjects who decided to participate in the trial by their voluntary will. The safety assessment was performed at every visit by KMDs.

As this was a pilot clinical trial to explore the efficacy and safety of acupuncture for FC and assess trial feasibility, the sample size was not calculated to provide sufficient statistical power to support the hypothesis but was instead based on the number of participants available for enrolment and the minimum number of participants required for the purposes of a pilot study [24]. Accordingly, 12 participants were required for each group; considering a 20% withdrawal rate, we planned to enrol 30 participants.

The 30 enrolled participants were randomly allotted into either the RA or SA group at a 1:1 ratio. The randomisation list was generated with a block-randomisation method using SAS® Version 9.1.3 (SAS Institute, Inc., Cary, NC, USA) by an independent statistician. The block size was randomly determined.

After the randomisation list was created, it was coded, sealed in an envelope, and stored in a locked cabinet. The group allocations were sealed in opaque double envelopes and numbered in order according to the randomisation list. Upon participant enrolment, an envelope was opened to determine their group allocation. The KMDs in charge of the intervention verified the group allocations of the participants.

Assessors and participants were blinded. As practitioners cannot be blinded due to the nature of the intervention, they merely conducted acupuncture without unnecessary conversation or sharing information with other researchers. Assessors simply asked essential questions, which were specified in the case report form. Participants were scheduled to visit at different times to minimise both the exchange of information and the risk of bias.

Each group received 12 intervention sessions (3 times over 4 weeks). We used same acupuncture needles (0.25 mm in diameter and 40 mm in length, Donbgang Acupuncture Inc., Bundang, Sungnam, Republic of Korea), and conducted intervention and evaluation with the same number of times in same schedule.

The RA group received RA on eight fixed acupuncture points (bilateral ST25, ST27, BL52 and BL25) and four individualised acupuncture points. Sa-am acupuncture was used as the individualised acupuncture. The Sa-am acupuncture is a type of acupuncture based on traditional Korean medical (TKM) principles. The traditional theory about yin and yang, five element, and meridian are used in the application of Sa-am acupuncture. That is, five shu points of each meridian are used according to the creation and control cycles of five-element theory. Acupuncture points are selected based on the relationship of Governor, Mother and Son. For example, in case of deficiency of organ, governor point of its own meridian and of its governor meridian are sedated, while mother point of its own meridian and of its mother meridian are tonified. The set of these four acupuncture points is called ‘Jeong-Gyeok’. In case of excess of organ, the governor points should be tonified while son point of its own meridian and of its son meridian should be sedated. The set of these four acupuncture points is called ‘Seung-Gyeok’. In the present study, we used large intestine (LI) Jeong-Gyeok for LI deficiency, stomach (ST) Jeong-Gyeok for stomach deficiency, and liver (LR) Seung-Gyeok for liver excess [25‐27]. The detailed location and procedures of RA can be found in Additional file 1 and Fig. 2. KMDs inserted the acupuncture needles using guide tubes, manipulated them (twitch, forward-and-backward) for de-qi, and maintained them for 30 min. Subsequently, the needles were again manipulated for de-qi and then removed. “de-qi” indicates that “sensory of soreness, numbness or tingling, fullness or pressure and heaviness” has occurred by proper acupuncture insertion [23, 28].

The SA group received minimal acupuncture consisted of shallow acupuncture insertion (1–2 mm in depth) without manual stimulation or de-qi sensation on 12 non-acupuncture points for 30 min [29, 30]. The detailed location and procedures of SA can be found in Additional file 2 and Fig. 3. KMDs with over 4 years of clinical experience conducted the acupuncture procedures.

Each participant’s usual intake of dietary fibre drinks, coffee, sweet potatoes, or fermented milk was maintained. In the first screening stage, patients taking regular laxatives or had taken a laxative in the 2 weeks before enrollment were excluded according to our eligibility criteria, and subjects were not asked to discontinue their prescribed medication or change their standard medical care. All patients who participated in the present study were allowed to use laxative (rescue medicine) prescribed by the protocol when needed throughout the entire study period. We provided participants with Magnesium Hydroxide 500 mg as a rescue medication and had them take it if needed. The Magnesium Hydroxide has been recommended as effective treatment for constipation and widely used in clinical practice with agreement of more than 90% of Korean clinicians [31]. Each participant was asked to complete a daily diary recording the shape and frequency of defecation, related symptoms, and use of rescue medication. The number of days and amount of rescue medication taken by each participant were monitored and used for analysis of results. The intervention details have been described according to the Standards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) criteria (Additional file 3).

The weekly DFs, weekly spontaneous complete bowel movement (SCBMs) and mean Bristol stool scale (BSS) scores were recorded every week during the treatment period and the two follow-up visits. The constipation assessment scale (CAS) scores were evaluated at weeks 1, 3, 5, 7 and 9. Rescue medication use and AEs were investigated at each visit. The primary outcome was changes in DF at week 5; all others were secondary outcomes.

Statistical analyses were performed using the full analysis set (FAS) population, which included a population as similar to intent-to-treat (ITT) as possible. All analyses were performed using SAS® Version 9.4 with a significance level of 5% and two-sided tests.

For the purpose of this pilot study, the analyses performed were mainly descriptive to present actual changes and treatment effects using the mean difference (MD) and Hedge’s g. The effect size was interpreted as follows: 0.2, small; 0.5, medium; 0.8, large; 1.2, very large; and 2.0, huge [32, 33].

Inter-group differences were examined using a Mixed-effect Model for Repeated Measures (MMRM). Paired-sample t-tests or Wilcoxon signed-rank tests were conducted to analyse the changes in each group after treatment. All AEs reported during the study period were charted, and the difference in the incidence of AEs between the two groups was assessed using the Chi-square test or Fisher’s exact test. The ‘last observation carried forward’ approach was used to replace the missing values, except in the MMRM. The MMRM method does not require an alternative step because it uses the maximum likelihood, which accounts for missing values.

A subgroup analysis was performed for participants with severe symptoms. DF and SCBM were relatively even in participants overall, while the BSS and CAS scores showed heterogeneity. Thus, severity was classified based on these two factors. Subgroup A included participants whose proportion of hard stool (BSS score 1–2) was > 25% of total defecation during one week, and subgroup B included participants whose CAS score was ≥7 at visit 1.

Thirty-five participants were assessed for eligibility, but 5 (14.3%) were not enrolled (three withdrew consent; two did not meet the eligibility criteria).

Twenty-eight (93.3%) participants completed the trial. One participant in the RA group was excluded due to a violation of the inclusion criteria identified at visit 1, before any treatment was performed. One participant in the SA group withdrew consent.

The treatment-compliance rate, i.e., the percentage of treatment sessions participants actually received of all 12 planned sessions, was 99.4 ± 2.4% and 94.4 ± 19.3% in the RA and SA groups, with an overall rate of 96.8 ± 14.0%.

Both groups included more females than males, with 85.7 and 91.7% females in the RA and SA groups, respectively. The mean age (49.6 and 50.0 years old), height, body weight and body mass index were similar in both groups. Factors that may influence defecation, including dietary, exercise, smoking and drinking habits, were not significantly different between groups (Table 1).

After the four-week treatment period (week 5), DF decreased from 5.86 ± 5.61 to 5.43 ± 3.39 with RA, while it increased from 3.79 ± 1.62 to 5.00 ± 1.77 with SA. The MD was − 1.70 (− 3.94 to 0.55), Hedge’s g was 0.56, and the changes were not significantly different between the groups (p = 0.2933, Table 2).

The blinding index scores [34] were 0.357 (95% confidence interval [CI] -0.068 to 0.782) for RA and − 0.500 (95% CI -0.883 to − 0.117) for SA, indicating that the blinding was adequate (p = 0.9999) (Table 4).