Clinical observation on 90 cases of primary dysmenorrhea treated by buccal acupuncture therapy: a randomized controlled study

Abstract

OBJECTIVE:

To explore the short-term and long-term efficacy of buccal acupuncture therapy in the treatment of patients with primary dysmenorrhea (PD).

METHODS:

A total of 90 patients with PD who were admitted to the Second Affiliated Hospital of Zhejiang University of Traditional Chinese Medicine from January 2020 to December 2021 were included in this study and randomly divided into three groups, including the observation group, the control group, and the placebo group, with 30 patients in each group. Patients in the observation group were treated according to the syndrome types of Traditional Chinese Medicine, those in the control group were treated with Saridon, and those in the placebo group received a placebo. The visual analog scale (VAS) scores of these patients were evaluated at different time points (before treatment, 5 min after needle retention, 30 min after needle retention, 1 h after treatment, 2 h after treatment, 1 d after treatment, and 2 d after treatment). The Cox Menstrual Symptom Scale (CMSS) scores were assessed before the treatment and after three courses of treatment. Besides, these two scales were also evaluated during follow-up. Moreover, the pain score and adverse reactions of these patients were assessed during treatment.

RESULTS:

After treatment, the VAS and CMSS scores in the observation group decreased significantly (P < 0.05) 5 and 30 min after needle retention and after treatment.

CONCLUSION:

Buccal acupuncture therapy is verified to have short-term efficacy in the treatment of PD. In addition, this therapy has an immediate analgesic effect and long-term efficacy in reducing the severity and frequency of PD.

1. INTRODUCTION

Primary dysmenorrhea (PD) refers to painful menstruation without organic lesions in the genital, typically involving lower abdominal cramps, bloating, low back pain, or other discomforts before and after menstruation or during menstruation. This disease can be accompanied by other symptoms, such as headache, dizziness, nausea, vomiting, abdominal distension, and aching pain in the low back. Severe symptoms will affect the work and life of patients.¹ Based on epidemiological surveys, the incidence rate of PD in China is over 50%,² and globally the rate ranges between 50% and 90%, with over 50% of cases presenting moderate to severe pain.³ Due to its long duration and high relapse rate, PD has caused serious adverse impacts on their work, study, and life. Under the influence of increased stress and emotional factors, the incidence of PD is increasing with each passing year.⁴ Traditional Chinese Medicine (TCM) categorizes dysmenorrhea into two types: deficiency and excess. The excess type is mainly manifesting as cold coagulation and blood stasis pattern,⁵ and Qi stagnation and blood stasis pattern.³ This disease is mainly associated with significantly increased prostaglandin F2 α (PGF2 α) and prostaglandin E2 in both the endometrium and menstrual blood of PD patients. The excessively high level of PGF2 α will cause excessive contraction of uterine smooth muscle and vasospasm, resulting in uterine ischemia, hypoxia, and dysmenorrhea. Increased prostaglandins entering the blood circulation can also cause cardiovascular and digestive symptoms.⁶ In addition, higher levels of vasopressin, endogenous oxytocin, and β-endorphin is also related to PD. In the clinical treatment based on Western Medicine, prostaglandin enzyme inhibitors, and hormone drugs are common medications, which may induce many adverse reactions.⁷ The immediate efficacy of such common painkillers as ibuprofen, ketoprofen, and diclofenac is favorable, but the long-term efficacy is poor. Most patients need to receive analgesics for a long time to maintain treatment. In contrast, the application of acupuncture as non-pharmaceutical pain management has been recognized for a long time, and its efficacy can be maintained for more than 12 months.⁸ This TCM-based therapy has been widely used in the management of pain diseases, and its clinical value is more significant. In the treatment of PD, the overall effectiveness of acupuncture therapy can reach over 95%.^9,10 It has been found that acupuncture therapy for PD has a higher overall effectiveness rate than analgesics. The effectiveness is particularly significant immediately after treatment, and the recurrence rate is low.¹¹ Buccal acupuncture therapy can be applied to the treatment of various diseases, and it can be performed by acupuncture at specific points on the cheek. This therapy features less pain, rapid remission, and higher safety.¹² Buccal acupuncture, as a novel microneedle therapy, has demonstrated positive effects in various pain disorders such as cervical spondylosis¹³ and lumbar and back soft tissue injuries.^14,15 Its rapid analgesic effect has also made it a viable adjunct for perioperative analgesia.^15,16 Hence, buccal needle therapy was employed in this experiment for the treatment of PD. In this study, visual analog scale (VAS) and COX menstrual symptom scale (CMSS) were used to evaluate the clinical efficacy of buccal acupuncture therapy in the treatment of PD.

2. MATERIALS AND METHODS

2.1. Clinnical data

2.1.1. General information

Patient recruitment was conducted among the outpatients of gynecology and obstetrics of the Second Affiliated Hospital of Zhejiang University of Traditional Chinese Medicine and college students of Zhejiang University of Traditional Chinese Medicine. A total of 90 PD patients were included in this study according to the inclusion criteria. The random number table was generated by SPSS 27.0 (IBM Corp. Armonk, NY, USA), and the random numbers were placed in numbered, opaque, and sealed envelopes according to the generation order. These envelopes were distributed according to the recruitment order of patients entering the group. These patients were divided into the observation group, the control group, and the placebo group at a ratio of 1∶1∶1, with 30 patients in each group. This study was approved by the Medical Ethics Committee of the Second Affiliated Hospital of Zhejiang University of Traditional Chinese Medicine [Approval No. 2019-KL-060-01] and registered in Chinese Clinical Trial Registry (Registration No. MR-33-20-002005).

2.1.2. Diagnostic criteria

The diagnostic criteria of Western Medicine refer to the diagnostic criteria of PD in Clinical Diagnosis and Treatment Guide-Gynecology and Obstetrics:¹⁷ (a) young women who have the onset within 1 to 2 years after menarche or have no children; (b) pain with the onset during menstruation or several hours before menstruation and the duration no more than 72 h; (c) spastic pain or pain similar to labor pain; (d) no pelvic organic disease confirmed by pelvic examination and B-ultrasound.

The TCM classification standard refers to the TCM classification standard of dysmenorrhea in Traditional Chinese Gynecology (2007):

Cold coagulation and blood stasis pattern: cold pain with refusal to press in the lower abdomen before or during menstruation, and relieved by heat; or delayed menstruation with low amount, of dark red color with blood clots; Bluish-white complexion, cold limbs and fear of cold, dull tongue with white fur, and sunken and tight pulse. (b) Qi stagnation and blood stasis pattern: abdominal distension and pain with refusal to press in the lower abdomen before or during menstruation; the menstrual amount is low. The menstrual blood is able to flow but not smooth, with dark purple color and blood clots; The pain will be relieved temporarily when the clot is gone; breast pain, with chest distress; purple dark tongue or with ecchymosis; wiry pulse.

2.1.3. Inclusion criteria

Patients with symptoms consistent with the diagnostic criteria of western medicine for PD and the two syndrome types of TCM dysmenorrhea (the cold coagulation and blood stasis pattern and the Qi stagnation and blood stasis pattern); (b) Patients aged 18-35 years; (c) Patients with the onset on the day before menstruation or the day of menstruation for 4 consecutive months in the past 6 months; (d) Patients without receiving other relevant treatment before enrollment; (e) Patients with no administration of any painkillers during treatment; (f) Patients with VAS scores ranging from 4 to 10 points; (g) Patients providing the informed consent form.

2.1.4. Exclusion criteria

Patients with secondary dysmenorrhea; (b) Patients with an irregular menstrual cycle or the cycle being less than 21 d or more than 35 d; (c) Patients not eligible for acupuncture due to irregular ulceration, injury, and swelling on the face; (d) Patients with a previous history of cosmetic surgery, face thinning needles, and anti-wrinkle needles; (e) Patients with neurosis and neuropsychiatric diseases.

2.1.5. Test dropout/subject withdrawal criteria

Patients with poor compliance and voluntary withdrawal during treatment; (b) Patients with adverse reactions during treatment or those who were not suitable for subsequent treatment; (c) Patients with a loss to follow-up.

2.1.6. Elimination criteria

Patients who did not meet the inclusion criteria and were mistakenly included; (b) Patients with failure to receive treatment as prescribed in this study or with incomplete data, which may affect efficacy evaluation and safety evaluation; (c) Patients with a substitute individual during treatment.

2.2. TREARMENT

2.2.1. Observation group

Selection of acupoints: (a) Cold coagulation and blood stasis pattern : buccal acupuncture points on the cheek: bilateral waist point CA-7 (at the midpoint of the line between the back point and the sacral point) (Note: back point CA-6: below the temporomandibular joint at the lower edge of the zygomatic arch root), sacral point CA-8 (1.66 cm above the front of the mandibular angle), abdominal palpation tenderness point corresponding to the holographic cheek projection point, bilateral lower energizer point (CA-4 inside the front edge of the internal mandibular angle). (b) Qi stagnation and blood stasis pattern: buccal acupuncture points of cheeks: bilateral waist points, sacral points, abdominal palpation tenderness points corresponding to holographic cheek projection points, bilateral upper energizer points (CA-2, the intersection point of the posterior coronal process of the mandible and the lower edge of the zygomatic arch), middle energizer point (CA-3, the midpoint of the line between the upper energizer point and the lower energizer point), and CA-4.¹⁸ Acupuncture method: × 30 mm acupuncture needle is used for acupuncture at CA-7 and CA-8 0.18 mm for the corresponding holographic cheek projection point of abdominal palpation tenderness point. The direction of acupuncture at CA-7 is inclined to the corner of the mouth, and the needle is 20-25 mm; Direct puncture at Triple Energizer point, about 20 mm; The abdominal palpation tenderness point corresponds to the holographic cheek projection point, and the needle is inserted about 20 mm; 0.16 mm for sacral point × 20 mm acupuncture needle is used for acupuncture without reinforcing and reducing manipulation. The needle was inserted, followed by proper adjustment according to the patient’s complaint.¹⁹ The needle was retained for 30 min, without lifting-thrusting or twirling reinforcing-reducing method. Appliance selection: acupuncture needle 0.16 mm × 20 mm and 0.18 mm × 30 mm. Course of treatment: the acupuncture was performed once a week,^8,14 specifically on the first day of the menstrual period (The visual analog scale method was used to evaluate the immediate analgesic effects of buccal acupuncture therapy on PD), for 3 consecutive months.

2.2.2. Control group

Taking oral San Litan (compound acetaminophen tablets) Bayer Healthcare Co., Ltd. (Leverkusen, Germany, National Medical Approval Number: H20056948), starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months.

2.2.3. Placebo group

The placebo used was a lactose pill (which looks similar to the analgesic pill and contains 500 mg of lactose per pill). These were developed and provided by Jiangsu Yasheng Pharmaceutical Development Co., Ltd. (Suzhou, China) They were taken orally from the first day of menstruation after the participant joined the study. They were taken for three consecutive days, three times a day, for a total of three months.

All patients in the observation group were treated with acupuncture throughout the treatment course by the same doctor to ensure the consistency of acupoint selection and treatment. Those in the control group and the placebo group were treated by another doctor. The CMSS score of patients was evaluated before treatment. The VAS score of patients was also evaluated after injection, 5 min, and 30 min after needle retention. The evaluation was performed by the acupuncturist and prescriber. Postgraduate students and treatment doctors jointly completed the follow-up and scoring tables.

2.3. Observation indexes and methods

2.3.1. Pain score

In terms of VAS²⁰ in this study, a 10 cm long moving scale was used, with 10 scales on one side. “0” indicates no pain, and “10” indicates the most severe pain that cannot be tolerable. During the test, patients marked a line at a certain point on the side without scales according to the pain severity. The pain severity was recorded on the side with scales. The higher the score, the more severe the PD. Acupuncture was conducted after the onset of pain in each menstrual cycle. The VAS score was evaluated before treatment and 5 min, 0.5 h, 2 h, 1 d, and 2 d after treatment. The treatment was performed for a total of 3 menstrual periods. After 3 treatment cycles, follow-up was performed once a month until 6 months after treatment, and the VAS score was recorded monthly.

2.3.2. CMSS

CMSS was composed of 18 items, including abdominal pain, headache, lumbosacral pain, stomachache, leg pain, nausea, vomiting, anorexia, fatigue, dizziness, diarrhea, facial color change, blush, insomnia, general pain, depression, irritability, and neuroticism. It was employed to evaluate the severity and duration of dysmenorrhea.²¹ The severity evaluation criteria included: 0 points for none, 1 point for mild (symptoms can be felt), 2 points for moderate (symptoms but do not affect daily life), 3 points for more significant (symptoms affect daily life), and 4 points for severe (symptoms are serious and affect daily life). The pain duration evaluation criteria included: 0 points for no pain, 1 point for the duration < 3 h, 2 points for the duration of 3-7 h, and 3 points for the duration of 7-24 h. Before treatment, patients were asked to review the average severity and duration of PD in the past three menstrual cycles to complete relevant evaluations. The CMSS score was recorded at the end of treatment, and 6 months after treatment. The short-term and long-term efficacy can be evaluated based on the scores of these patients.

2.3.3. Statistical analysis

All statistical analyses were performed using SPSS 27.0 (IBM Corp. Armonk, NY, USA). Measurement information calculated as mean ± standard deviation ($\bar{x}±s$). Significance tests were two-tailed, and statistical significance was set at a P value less than 0.05.

2.3.3.1. Sample size

This study employs a repeated measures design, taking into account the effect size reported in previous research: (effect size, ES) $f=0.25$ , (power) $\mathrm{ }1-\mathrm{\beta }=0.8$ , (significance level) $\mathrm{\alpha }=0.05$ .²² Assuming a follow-up loss rate of 10% among the study subjects, a minimum sample size of 25 participants per group is required to meet the experimental.

2.3.3.2. Analytical method

The data collected in this experiment consists of repeated measurements. There are only two independent variables, one between-group factor — group, and one within-group factor — time. Therefore, a two-factor repeated-measures analysis of variance (ANOVA) was utilized for the analysis. When there was missing data, the data for the entire experimental process of the corresponding participant was removed. Also, when a significant interaction between time and group was observed, further separate effect analysis and post hoc test were performed, with Bonferroni correction applied.

3. RESULTS

3.1. Comparison of general data between three groups

In this study, a total of 90 patients were included. In the observation group, 3 patients dropped out (due to personal reasons) and 2 patients were eliminated (due to other treatments). In the control group, 7 patients dropped out (due to personal reasons). In the placebo group, 6 patients dropped out (due to personal reasons) and 2 patients were eliminated (due to other treatments). Finally, 25 patients in the observation group, 23 patients in the control group, and 22 patients in the placebo group were included in the statistical analysis (Figure 1). There was no statistically significant difference in the general data of age and course of disease between these three groups (P > 0.05), as shown in Table 1.

Figure 1. Flowchart of participants in the randomized controlled trial.

HCG: human chorionic gonadotropin.

Table 1.

Baseline characteristics of patients in the three study groups, collected on the first day of the intervention cycle ($\bar{x}±s$)

Variable	Observation group (n = 30)	Control group (n = 30)	Placebo group (n = 30)	P value
Age (years)	22.3±2.5	21.7±2.8	22.2±2.1	0.648
Age at menarche (years)	12.5±1.2	12.9±1.5	12.6±1.3	0.447
Menstrual cycle (days)	28.8±1.8	29.0±1.9	28.6±1.6	0.644
Menstruation (days)	6.5±0.9	6.6±0.9	6.4±0.9	0.553
Duration of disease (years)	5.1±1.6	4.2±1.6	4.9±1.4	0.049

Notes: observation group: treated according to the syndrome types of TCM (selection of acupoints: CA-2, CA-3, CA-4, CA-7, CA-8; frequency of acupuncture: once a week, for 3 consecutive months); control group: treated with Saridon (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months); placebo group: received a placebo (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months). Menstruation: days in one menstrual cycle; Menstrual cycle: length between menstrual cycles. Baseline data was analyzed using one-way analysis of variance when comparing between the three groups, and differences were considered statistically significant at P < 0.05.

3.2. Comparison of VAS scores between three groups before and after treatment

3.2.1. Comparison of short-term VAS scores before and after first treatment

The pre- and post-treatment VAS scores of the three patient groups were compared pairwise. The observation group and control group, the observation group and placebo group, and the control group and placebo group’s VAS scores all used two-factor repeated-measures ANOVA. Through testing, there were interactions between time and group for the observation group and control group (F_time*group = 55.23, P < 0.01), the observation group and placebo group (F_time*group = 93.61, P < 0.01), and the control group and placebo group (F_time*group = 109.775, P < 0.01). Therefore, separate effect analysis and post hoc test were further performed for all three groups of data.

The separate effect analysis and post hoc test of the observation group and control group showed: before treatment, there was no significant difference in VAS scores between the observation group and the control group (P = 0.31); the differences in VAS scores between the observation group and the control group at 5 min, 30 min, and 2 h after treatment were －3.53, －0.77, and 0.47, respectively, all of which were statistically significant (P < 0.05). The largest difference was observed 5 min after treatment, indicating that the analgesic effect of the observation group started quickly. The difference in VAS scores between the observation group and the control group one day after treatment was 0.17, which was not statistically significant (P = 0.12). Two days after treatment, both groups had scores of 0, so no further statistical comparison was made. One and two days after treatment, the pain scores of the observation group were close to 0, similar to the control group, indicating that the analgesic effect of the observation group was comparable to that of the control group, and there was no rebound.

The separate effect analysis and post hoc test of the observation group and placebo group showed that before treatment, there was a statistically significant difference in VAS scores between the observation group and the placebo group (F = 5.67, P = 0.31). However, after considering the baseline description in the previous text, the two groups can be compared; at 5 min, 30 min, 2 h, and 1 d after treatment, the differences in VAS scores between the observation group and placebo group were－3.13, －4.66, －4.95, and－1.97, respectively, all of which were statistically significant (P < 0.01). This indicates that the short-term analgesic effect of the observation group was significant and could last for over 24 h.

The separate effect analysis and post hoc test of the control group and placebo group showed that before treatment, there was no significant difference in VAS scores between the control group and the placebo group (P = 0.19); 5 min after treatment, the difference in VAS scores between the control group and placebo group was 0.40, which was not statistically significant (P = 0.26); 30 min, 2 h, and 1 d after treatment, the differences in VAS scores between the control group and placebo group were －3.89, －5.42, and －2.14, respectively, all of which were statistically significant (P < 0.01). The results revealed that the control group exhibited a noticeable short-term analgesic effect, with the onset time occurring 30 min after treatment (Table 2).

Table 2.

VAS scores at short-term effect analysis following the start of the first intervention, in the three groups ($\bar{x}±s$)

Time	Observation (n = 30) ($\bar{x}±s$)	Control (n = 30) ($\bar{x}±s$)	Placebo (n = 30) ($\bar{x}±s$)	Observation vs control (time*group)		Observation vs control			Observation vs placebo (time*group)		Observation vs placebo			Control vs placebo (time*group)		Control vs placebo
				FTime* Group	P value	Adjusted mean difference (95% CI)	P value	F value	FTime* Group	P value	Adjusted mean difference (95% CI)	P value	F value	FTime* Group	P value	Adjusted mean difference (95% CI)	P value	F value
0 min (before first treat-ment)	6.80±1.19	6.47± 1.31	6.00± 1.39	55.23	< 0.001	0.33 (－0.31, 0.98)	0.305	1.02	93.61	< 0.001	0.80 (0.13, 1.47)	0.021	5.67	109.775	< 0.001	0.47 (－0.24, 1.17)	0.189	1.77
After 5 min	2.87±1.59	6.40± 1.33	6.00± 1.39	-	-	－3.53 (－4.29, －2.78)	< 0.001	87.12	-	-	－3.13 (－3.91, －2.35)	< 0.001	64.74	-	-	0.40 (－0.31, 1.11)	0.263	1.28
After 30 min	1.07±1.02	1.83± 1.14	5.72± 1.44	-	-	－0.77 (－1.33, －0.21)	0.008	7.52	-	-	－4.66 (－5.30, －4.01)	< 0.001	207.99	-	-	－3.89 (－4.57, －3.21)	< 0.001	132.60
After 2 h	0.67±0.84	0.20± 0.66	5.62± 1.86	-	-	0.47 (0.07, 0.86)	0.021	5.66	-	-	－4.95 (－5.70, －4.21)	< 0.001	175.57	-	-	－5.42 (－6.14, －4.70)	< 0.001	225.29
After 1 d	0.20±0.55	0.03± 0.18	2.17± 1.98	-	-	0.17 (－0.05, 0.38)	0.121	2.47	-	-	－1.97 (－2.73, －1.22)	< 0.001	27.49	-	-	－2.14 (－2.87, －1.41)	< 0.001	34.62

Notes: observation group: treated according to the syndrome types of TCM, (selection of acupoints: CA-2, CA-3, CA-4, CA-7, CA-8; frequency of acupuncture: once a week, for 3 consecutive months); control group: treated with Saridon (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months); placebo group: received a placebo (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months). TCM: Traditional Chinese Medicine; ANOVA: a two-factor repeated-measures analysis of variance. VAS: visual analog scale. A two-way repeated-measures ANOVA was utilized for the VAS scores at short-term effect, when a significant interaction (P < 0.05) between time and group was observed, further separate effect analysis and post hoc test were performed with Bonferroni correction applied. An adjusted mean difference (95% CI) was presented, and P values indicate significance after the post hoc test.

3.2.2. Comparison of long-term VAS scores post-treatment and during follow-up

During the double-factor repeated measurement analysis of dysmenorrhea VAS scores after treatment and during the follow-up period, all data points from individuals outside of the cohort were excluded: 5 from the experimental group, 7 from the control group, and 8 from the placebo group.

Analysis of VAS scores post-treatment between the observation group and the control group revealed a significant interaction between time and group (F_time*group = 41.89, P < 0.01). Therefore, further individual effect and post-hoc analysis was performed.

Individual effect and post-hoc analysis of the observation group and the control group showed that pre-treatment, the difference in VAS scores between the observation and control groups was 0.43, which was not statistically significant (P = 0.23); After 1, 2, and 3 treatment courses, and at 3 and 6 month follow-up, the differences in VAS scores between the observation and control groups were －1.21, －2.91, －4.29, －3.77, and －3.92 respectively, all of which were statistically significant (P < 0.01); Moreover, the score differences between the two groups gradually increased, indicating that the treatment of PD in the observation group has a cumulative effect, able to alleviate the severity of pain attacks.

Analysis of VAS scores post-treatment between the observation group and the placebo group revealed a significant interaction between time and group (F_time*group = 38.67, P < 0.01). Therefore, further individual effect and post-hoc analysis was performed.

Individual effect and post-hoc analysis of the observation group and the placebo group showed that pre-treatment, there was a statistically significant difference in VAS scores between the observation group and the placebo group (P = 0.02), but after a comprehensive consideration of the baseline description, the two groups could be compared. The differences in VAS scores between the observation group and the placebo group after 1, 2, and 3 treatment courses, and at 3 and 6 month follow-up were －0.91, －2.47, －3.77, －3.00, and －3.13 respectively, all of which were statistically significant (P < 0.01).

Analysis of VAS scores post-treatment between the control group and the placebo group revealed no significant interaction between time and group (F_time*group = 0.51, P = 0.63). Therefore, a main effect analysis was performed.

Based on the results of the intra-subject effect test, VAS scores did not exhibit a downward trend at the end of treatment 1, treatment 2, treatment 3, and during the third and sixth months of follow-up after the start of the experiment (F_time = 1.81, P = 0.16). Additionally, according to the results of the intersubject effect test, there was no significant difference in VAS scores between the control group and the placebo group (F_group = 2.89, P = 0.10). These results suggest that Sanlitong only has a short-term analgesic effect, and does not have a cum-ulative effect (Table 3).

Table 3.

VAS scores at long-term effect analysis following the start of the first intervention, in the three groups ($\bar{x}±s$)

Time	Observation (n = 25) ($\bar{x}±s$)	Control (n = 23) ($\bar{x}±s$)	Placebo (n = 22) ($\bar{x}±s$)	Observation vs control (time*group)		Observation vs control			Observation vs placebo (time*group)		Observation vs placebo			Control vs placebo (time*group)
				FTime*Group	P value	Adjusted Mean Difference (95% CI)	P value	F value	FTime*Group	P value	Adjusted Mean Difference (95% CI)	P value	F value	FTime*Group	P value
Before first treatment	6.9±1.2	6.4±1.3	5.9±1.4	41.89	<0.01	0.43 (－0.28, 1.12)	0.23	1.47	38.67	<0.01	0.91 (0.17, 1.64)	0.02	6.15	0.51	0.63
After completion of first treatment	5.2±1.4	6.4±1.1	6.1±1.1	-	-	－1.21 (－1.93, －0.49)	<0.01	11.43	-	-	－0.91 (－1.64, －0.18)	0.02	6.32	-	-
After completion of second treatment	3.4±1.2	6.3±1.5	5.9±1.3	-	-	－2.91 (－3.66, －2.16)	<0.01	61.31	-	-	－2.47 (－3.18, －1.76)	<0.01	49.27	-	-
After completion of third treatment	2.1±1.6	6.4±1.2	5.9±1.4	-	-	－4.29 (－5.12, －3.46)	<0.01	108.47	-	-	－3.77 (－4.64, －2.89)	<0.01	75.40	-	-
3 months after treatment	2.4±1.7	6.1±1.8	5.4±1.6	-	-	－3.77 (－4.77, －2.78)	<0.01	58.33	-	-	－3.00 (－3.95, －2.07)	<0.01	41.27	-	-
6months after treatment	2.6±1.8	6.6± 1.2	5.8±1.2	-	-	－3.92 (－4.82, －3.02)	<0.01	76.72	-	-	－3.13 (－4.04, －2.22)	<0.01	47.37	-	-

Notes: observation group: treated according to the syndrome types of TCM, (selection of acupoints: CA-2, CA-3, CA-4, CA-7, CA-8; frequency of acupuncture: once a week, for 3 consecutive months); control group: treated with Saridon (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months). TCM: Traditional Chinese Medicine; ANOVA: a two-factor repeated-measures analysis of variance. VAS: visual analog scale. Values are mean ± standard deviation ($\bar{x}±s$). VAS scores at long-term effect between the three groups were analyzed by two-way repeated-measures ANOVA, with further separate effects analysis and post hoc tests with Bonferroni correction when a significant interaction between time and group was observed (P <0.05). An adjusted mean difference (95% CI) was presented, and P values indicate significance after the post hoc test.

3.3. Comparison of CMSS scores between three groups before and after treatment

The CMSS scores pre- and post-treatment for the observation group and the control group were analyzed using a two-factor repeated measures ANOVA. The test showed a significant interaction between time and group (F_time*group = 106.48, P < 0.01). Therefore, further individual effect and post-hoc analysis was performed.

Individual effect and post-hoc analysis showed that pre-treatment, the difference in CMSS scores between the observation group and the control group was not statistically significant (F = 3.33, P = 0.06). After treatment, there was no statistically significant difference in CMSS scores between the observation group and the control group (F = 0.07, P = 0.80), suggesting that the analgesic effect of buccal acupuncture in treating PD is comparable to Sanlitong. At the 6-month follow-up, the difference in CMSS scores between the observation group and the control group was －13.92, which was statistically significant (F = 71.79, P < 0.01), suggesting that buccal acupuncture, in addition to improving pain, can significantly alleviate other accompanying symptoms of PD, such as gastrointestinal discomfort, back pain, anxiety, and irritability, etc. (Table 4).

Table 4.

COX following the start of the first intervention, observation group vs control group ($\bar{x}±s$)

Time	Mean±SD	Mean±SD	Observation group vs Control group (Time*Group)		Observation group vs Control group
			FTime*Group	P value	Adjusted mean difference (95% CI)	P value	F value
Before first treatment	35.6±7.8	31.7±6.8	106.48	<0.01	3.90 (0.63, 6.18)	0.06	3.33
After completion of third treatment	17.1±4.6	16.7±5.7	-	-	0.38 (－2.62, 3.39)	0.80	0.07
6 months after treatment	16.6±4.5	30.5±6.7	-	-	－13.92 (－17.23, －10.61)	<0.01	71.79

Notes: observation group (n = 25): treated according to the syndrome types of TCM, (selection of acupoints: CA-2, CA-3, CA-4, CA-7, CA-8; frequency of acupuncture: once a week, for 3 consecutive months); control group (n = 23): treated with Saridon (starting from the first day of menstruation after enrollment, take 1 tablet per dose, 3 times per day, for 3 consecutive days, for a duration of 3 months).SD: standard deviation; TCM: Traditional Chinese Medicine; COX: Cox Menstrual Symptom Scale; ANOVA: analysis of variance. A two-way repeated-measures ANOVA was used for COX scores during follow-up in both groups, and when a significant interaction between time and group was observed (P < 0.05), further separate effects analyses and post hoc tests with Bonferroni correction were performed. An adjusted mean difference (95% CI) was presented, and P values indicate significance after the post hoc test.

3.4. Safety evaluation

There was no adverse event in the three groups during treatment, and there were no needle sickness, needle stagnation, local hematoma, allergy and other conditions in the acupuncture group.

4. DISCUSSION

The findings of this study demonstrated that buccal acupuncture therapy could exert immediate analgesic effects, and the pain in PD patients could be significantly mitigated after buccal acupuncture treatment. The improvement effect in the observation group was similar to that in the control group. There was no significant difference in the alleviation of other clinical symptoms between the observation group and the control group. However, the alleviation in these two groups was significantly better than that in the placebo group. Of note, the long-term efficacy in the observation group was significantly better than that of the control group. This finding was consistent with the pre-experimental results and the results of acupuncture on PD in other studies.^10,23 The mechanism underlying its long-term efficacy may be acupuncture is a complex intervention that combines neurophysiologic stimulation. It can generate anti-inflammatory effects by inhibiting cyclooxy genase synthesis in peripheral and central nociceptive sites.²⁴ In addition, the benefits of effective clinician-patient interaction are far-reaching and well documented.¹⁶ The exact underlying mechanisms of acupuncture for PD are worth of further investigation.

PD belongs to the category of “abdominal pain during menstruation” in TCM. This disease is mostly related to the invasion of external pathogens, dysfunction of the viscera, and emotional disorders, and it may involve the liver, spleen, and kidneys. The pathogenesis of this disease is mainly based on the empirical theory that “stagnation of Qi and blood may bring about pain”. Cold, dampness, and blood stasis in the uterus may induce the stagnation of Qi and blood circulation, which further results in abnormal pain during menstruation. Acupuncture can be employed to eliminate meridian obstruction, promote Qi and blood circulation, and relieve pain.²⁵ As a dysmenorrhea therapy approved by the World Health Organization,²⁶ acupuncture and moxibustion have certain clinical advantages in the treatment of PD, such as fewer adverse reactions. Hence, these therapies are worthy of clinical promotion.²⁷ As mentioned in Lingshu ·Jiu Zhen Shi Er Yuan,²⁸ “the most important thing in acupuncture is to get the acupuncture feeling, when it appears, the curative effect will appear in the wake of it, and the curative effect is so reliable as the blue sky appearing when the clouds are blown away by the wind, which can be considered the principles of acupuncture.” Acupuncture can generate effects immediately in the treatment of PD. Different acupuncture methods, such as acupuncture, warming needle moxibustion, ear acupuncture therapy and intradermal needle therapy, have been reported in many studies.^10,25,29,30

Buccal acupuncture is a new micro-needle acupuncture technique developed by Professor WANG Yongzhou based on modern anatomy, biological holography,³¹ and mind-body theory. According to its mechanism of action, the whole human body should be reduced in proportion, and there are holo-graphic embryos cor-responding to the human body structure on both cheeks. Corresponding viscera and limb dis-orders can be treated by acupuncture at specific points on the cheek.¹⁸ Meanwhile, the theories of Qi Street, Meridians and Triple Energizer in TCM are integrated into this therapy, so as to eliminate diseases by regulating the Qi mechanism of Triple Energizer and viscera through acupuncture.

To the best of our knowledge, this is the first scientific attempt to use the bioholography theory based on the location of the uterus in the pelvic cavity (in the center of the pelvic cavity, and front of S1-S5). The corresponding point was found in the cheek, namely CA-8, According to modern research findings,³² the uterus is a converging site of afferent projections in the spinal ganglia of L3-L6, so the CA-7 is selected as the first choice for the treatment of PD with buccal acupuncture. It has been confirmed in some studies that buccal acupuncture can increase the content of β-endorphin in the cerebrospinal fluid of a rheumatoid arthritis rabbit model, thus exerting analgesic effects.³³ According to another study, buccal acupuncture can adjust the level of 5-hydroxytryptamine in the human body and effectively improve the visceral pain of patients with gynecological disorders after laparoscopic surgery.³⁴ In the experiment, the above two points were selected for the treatment of PD patients. Pain mitigation can be observed through the VAS score 5 min after needle retention. This indicated that buccal acupuncture therapy can exert immediate analgesic effects. After 30 min of needle retention, 2 h, 1 d, and 2 d after treatment，namely at the end of the first treatment, a comparison was drawn between these three groups. It was found that the analgesic effect in the observation group and the control group was better than that in the placebo group. Besides, the analgesic effect in the observation group was comparable to that in the control group, indicating that buccal acupuncture therapy could realize favorable analgesic effects in PD patients.

Moreover, according to the Triple Energizer Theory in TCM, dysmenorrhea belongs to the category of “abdominal pain induced by the stagnation of Qi and blood”. For PD patients with the syndrome of “cold coagulation and blood stasis”, as the pain site is located in the category of lower energizer, it is mostly considered that blood stasis is caused by the cold coagulation and lower energizer. Therefore, the bilateral CA-4 sites are supplemented to promote blood circulation in the lower energizer. After the blood circulation flows freely, the pain can be resolved spontaneously. For PD patients with “Qi stagnation and blood stasis” syndrome, they are prone to present with obstructed functional activity of Qi. According to the Triple Energizer Theory, the Triple Energizer have a powerful gasification function and can serve as a channel. Hence, the smooth circulation at the Triple Energizer contributes to the flowing of blood and Qi throughout the body. Otherwise, the local or systemic stagnation of Qi and blood may be induced, which would result in various diseases. Therefore, the Triple Energizer acupoints are regarded as the core acupoints for regulating the viscera and Qi in the whole body.³⁵ As a result, it is proposed to use the Triple Energizer acupoints on both sides to regulate the Qi in the whole body. When Qi is regulated, relevant diseases may be resolved. In the experiment, patients in the observation group obtained favorable immediate analgesic effects after acupuncture and moxibustion. Besides, buccal acupuncture therapy was also verified to be effective in the treatment of PD. Furthermore, this therapy can also be employed to mitigate pain in PD patients and alleviate the complications of this disease.

The strength of the present trial lays in the strict inclusion criteria. But our study had limitations. First, 22% of participants were lost to follow-up over the course of the study. We tried to mitigate these by including the missing data in the data analysis and adjusting the multiple comparisons by using the Two-way repeated-measures ANOVA. Second，In this study, we did not study the mechanism of buccal acupuncture treatment for PD, and we can conduct mechanism related research in subsequent study. Third, we did not systematically assess risk for adverse events, although participants did not report harms; nonetheless, given the high dropout rates, we cannot rule out the possibility of adverse events.

In conclusion, buccal acupuncture is verified to be effective and safe in the treatment of PD. Therefore, it is expected to apply this therapy to the clinical treatment of PD.

5. ACKNOWLEDGMENTS

All authors of this manuscript would like to express our sincere gratitude to the Second Affiliated Hospital of Zhejiang Chinese Medical University.