De qi is a type of meridian-qi induction perceived by both acupuncturists and/or patients in the process of needling. The Lingshu Jing emphasized the fundamental role of de qi in acupuncture treatment [1]. Many studies have also found that de qi improves the curative effect more than when de qi is absent [2-5].

To improve the curative effect, studies have sought to identify the factors that facilitate de qi. Most current research has focused on external intervention factors, such as needling location (acupoints and non-acupoints), depth of needling, the use (or not) of manipulation, and the different methods of needling manipulation [5-7]. Moreover, almost all studies concluded that deep needling and/or manipulation could initiate more significant de qi reactions than shallow needling and/or no manipulation [2,6,8]. Conversely, researchers also noticed that body-intrinsic factors affected de qi, mainly involving the constitution of Traditional Chinese Medicine (TCM) [9,10], personality characteristics [11-13], and psychological factors, including the patient’s perception of the credibility of acupuncture [14,15], expectations [16], anxiety levels, and depressive symptoms [12,17,18]. Certain personality traits such as social boldness, warmth, and liveliness were related to de qi [11-13]. Researchers also had varying opinions regarding the influence of a patient’s perception of the credibility of acupuncture on de qi [14,15].

Similarly, three studies have different views on the influence of anxiety and depression on de qi: one study found that the de qi score decreased when patients were anxious or depressed [12]; another found that the degree of anxiety was positively correlated with the de qi index [18]; while a third study reported that anxiety did not affect either the de qi rate and its intensity [17]. These results suggest that psychological factors affecting de qi require further study.

Primary dysmenorrhea (PD) is a common gynecological complaint among young women, which manifests as spasmodic pain in the lower abdomen [19]. PD can adversely impact work and study [20-22] and can even negatively affect emotional moods [23]. Non-steroidal anti-inflammatory drugs (NSAIDs) are the preferred treatment [24], although long-term use can cause headaches, vertigo, gastrointestinal discomfort, and other adverse symptoms [25]. Consequently, holistic therapies such as acupuncture have emerged as viable alternatives to medications. The effectiveness and safety of needling for treating PD has been proven [26,27], with Sanyinjiao (SP6), an acupoint on the Spleen Meridian of Foot-Taiyin, one of the most commonly used acupoints for its treatment [28].

Our primary analysis in this trial showed that de qi is beneficial for curative effects [29]. Our research team also carried out studies on how the characteristics of de qi can influence the analgesic effect in PD patients. The results suggested that quicker onset of de qi achieves a better analgesic outcome [30], and a de qi response of moderate intensity is more likely to prolong the analgesic effects [31]. The purpose of the secondary analysis carried out in this study was to see how acupuncture-related anxiety and patients’ cognition of the relationship between needle sensation and therapeutic impact influenced the de qi sensation at SP6 in PD patients.

1. Study design

This study was a randomized, assessor-patient-blinded, parallel-controlled trial. Participants were randomly assigned to the de qi or non-de qi group at a 1:3 ratio. When designing this investigation and presenting our findings, we adhered to the Consolidated Standards of Reporting Trials and the Revised Standards for Reporting Interventions in Clinical Trials of Acupuncture [32,33]. The Beijing University of Chinese Medicine (BUCM) Ethics Committee approved the trial (no. 2014BZYLL0302), and it was registered with the Chinese Clinical Trial Registry (no. ChiCTR-IPR-14005361) on October 21, 2014.

2. Patients

Participants were primarily recruited through posters displayed at four universities. The inclusion and exclusion criteria, incentives (free gynecological ultrasound and acupuncture treatment), and contact details were all clearly visible on the posters.

The participants had to meet the following inclusion criteria: (1) aged between 18 and 30 years old; (2) nulliparous female; (3) diagnosed with PD as per the Primary Dysmenorrhea Consensus Guidelines [34], with a cold and dampness stagnation pattern according to the revised Chinese national guideline for PD [35]; (4) duration of PD between six months and 15 years; (5) and a pain level of 40 or more on the visual analog scale for pain (VAS-P, 0-100 mm) for at least three consecutive menstrual cycles. The exclusion criteria were: (1) dysmenorrhea; (2) abnormal menstrual cycles (beyond the normal range of 21-35 days); (3) pregnancy, trying to conceive, or lactation; (4) prior theoretical knowledge of acupuncture; (5) without using any analgesics within 24 hours before needling; (6) having a serious or life-threatening condition (e.g., cerebro-cardiovascular disease, liver failure, coagulopathy, etc.), mental illness, active gastric ulcers, and asthma; (7) and poor compliance (e.g., fear or resistance to needling).

The main symptoms of cold and dampness stagnation are (1) coldness and pain in the lower abdomen before or during the menstrual cycle that was relieved by warmth; (2) retarded menstruation; (3) scant menstrual flow; (4) and dark or black menses with clots. Secondary symptoms are (1) intolerance to cold, (2) cold extremities, (3) heavy leukorrhea, (4) white and greasy tongue coating, and (5) either a wiry or a deep and tense pulse. We gave 2 points for each primary symptom, and 1 point for each secondary symptom. The total score was calculated by adding the primary and secondary symptom scores, with an overall score of 6 or greater indicating cold and dampness stagnation.

3. Study procedures

Interested participants were initially evaluated over the phone. Those who had a clear desire to participate then took part in a complete interview. Researchers collected complete medical and gynecological histories from eligible applicants and informed them of the procedures and potential risks. Participants who passed the preliminary evaluation, met all the inclusion criteria, and signed an informed consent form then underwent further examinations including ultrasonography.

4. Sample size

The sample size was calculated using a frequently used sample size estimation method [36] with the following formula:

with significance level α = 0.05, power 1−β = 0.80, and ρ = 0.70. The minimum clinically effective change in abdominal pain is 10 mm on the VAS-P scale for women with dysmenorrhea [37]. Our previous research showed that the standard deviation of VAS-P for abdominal pain before treatment was 15.934 mm [5]. Therefore, δ = 10 and σ = 15.934, while the initial sample size for each group was set at 14.

In our previous study, the de qi or non-de qi group had equal numbers. Participants in the de qi group all experienced de qi, whereas approximately one-third in the non-de qi group did not [5]. This leads to a large discrepancy in the number of cases between patients who experienced genuine de qi in the de qi group and patients who did not experience genuine de qi in the non-de qi group. Therefore, in order to make the number of cases in these two subgroups as close as possible, the non-de qi group needed to be three times the size of the de qi group. Thus, the de qi to non-de qi group ratio was about 1:3. After accounting for a 20% dropout rate, the final sample sizes for the de qi and non-de qi groups were 17 and 51, respectively.

5. Randomization and blinding

Before treatments, the participants were randomly assigned to the de qi or non-de qi groups in a 1:3 ratio using a centralized telephone randomization technique. An administrator who was not involved with the data collection created the random allocation sequence using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA). The allocation codes and procedural information were kept in sequentially numbered opaque envelopes. The acupuncturist contacted the administrator to ask for the patient’s assignment 5-10 minutes before the treatment. Only the acupuncturist was aware of each subject’s group assignment and treatment plan, and neither the participants nor the recorder, who was responsible for recording and assessing all outcomes, knew the assignment and needling procedure. Additionally, a screen shielded both the patients and the recorders from the acupuncturist during treatment.

6. Intervention

The needling processes were performed in the laboratory of the School of Acupuncture-Moxibustion and Tuina at BUCM. To assure trial quality, the trial was frequently monitored by an experienced acupuncturist. All treatments performed by the same licensed acupuncture practitioner with five years of work experience in acupuncture practice.

Patients with abdominal pain caused by menstrual dysmenorrhea ≥ 40 mm on the VAS-P were treated with acupuncture. All patients were in a supine position and had bilateral SP6 needling for 30 minutes. The location of SP6 is the inner side of the leg, 3 B-cun (proportional bone cun) above the highest point of the medial malleolus, and distal to the medial edge of the tibia [38]. Sterile disposable stainless-steel needles with tubes were utilized during treatment (Zhongyan Taihe, Wuxi Jiajian Medical Instrument Co. Ltd., Jiangsu, China).

In the de qi group, 0.3 × 40 mm needles were inserted into SP6 to a depth of 1-1.2 cun with manipulation consisting of “lifting-thrusting” and twirling (180° in both directions, once per second) for 30 s. The manipulation was repeated twice every 10 minutes for 30 seconds each, to achieve the de qi sensation.

In the non-de qi group, 0.18 × 13 mm needles were inserted into SP6 at a depth of 0.1-0.2 cun with no manipulation in order to avoid inducing de qi. Many studies concluded that deep needling and/or manipulation could initiate more significant de qi reactions than shallow needling and/or no manipulation [2,6,8]. Therefore, we adopted the shallow needling using fine needles without manipulation to avoid de qi.

7. Measurements

After the treatment, the patient’s anxiety levels during the procedure were evaluated using a five-point scale (from 0: thoroughly relaxed to 4: feeling anxious beyond endurance). Next, the patients were asked “Do you think the needle sensation is a manifestation of the curative effect?” answered on a four-point scale (0: no; 1: maybe; 2: should be; 3: must be), to ascertain their views about the relationship between needle sensation and therapeutic effect.

The experience of the de qi sensation was measured using the Acupuncture De qi Clinical Assessment Scale (ADCAS) [39], which was developed using a literature review about the connotation of de qi, and the experiences and comments of 43 Chinese acupuncturists [40]. The ADCAS demonstrated satisfactory reliability and validity in a clinical trial (manuscript in preparation). The intensity of the needle sensation was assessed with a five-point scale (0: not at all; 1: mild; 2: moderate; 3: strong; 4: unbearably sharp pain). The 12 components of needle sensation (dull pain, warmth, coldness, soreness, numbness, fullness, heaviness, electric sensation, formication, throbbing, spreading, and radiating) were assessed with three-point scales. Patients were asked to substitute their own expressions if none of the 12 sensations listed above accurately conveyed their feelings. During our research, a recorder with a stopwatch (using 1/100 s as a unit of time) sat near the patient’s upper torso at a distance from which the vocal responses of the patient could be heard. After needling at SP6 by the acupuncturist, the patient told the recorder “Yes” when they experienced the de qi sensation and “No” when the sensation disappeared. The recorder used the stopwatch to record the occurrence and disappearance time of patients’ sensations, which were used to describe the rapidity and duration of de qi.

There are usually differences between the acupuncturist’s intentions during needling and the response that actually occurs, meaning that patients may not experience any of the sensations that the acupuncturist is trying to induce, including de qi. Therefore, assessing the patient’s response in clinical settings becomes important. The patients’ overall needling sensation may indicate the adequacy of the acupuncture treatment in the view of some specialists [41]. Accordingly, we judged whether or not a subject experienced de qi based on the overall intensity of the needle sensation.

The overall intensity of needle sensation is defined as the sum of the intensity of the needling sensation on bilateral SP6, with an intensity score ranging from 0 to 8. Clinical acupuncturists agree that sharp pain (a score of 4 or more on the ADCAS) should be avoided [42]. Five conditions were rated as non-de qi: 0 + 0 (no sensation on either side), 0 + 1 (faint sensation unilaterally), 0 + 4 (no sensation on one side and sharp pain on the other), 4 + 4 (sharp pain bilaterally), and 1 + 4 (weak sensation on one side and sharp pain on the other). Otherwise, these conditions (0 + 2, 0 + 3, 1 + 1, 1 + 2, 1 + 3, 2 + 2, 2 + 3, 2 + 4, 3 + 3, 3 + 4) were deemed to reflect de qi [43]. In this way, based on the patients’ actual de qi experience, we divided them into subgroups. Any adverse events were also recorded.

8. Data analysis

Clinical outcomes were analyzed using SPSS 22.0 statistics software (SPSS Inc., Chicago, IL, USA). Means and standard errors (SEs) were used to define baseline data. T-tests or analysis of variance (ANOVA) or non-parametric rank-sum tests were used to compare the means of independent samples. A chi-squared test was used to compare rates. Pearson’s (Measurement data of normal distribution) or Spearman’s (Measurement data of non-normal distribution or ranked data) correlation coefficients were used for correlation analyses. p-values of < 0.05 denoted statistical significance.

1. Participants

The research enrolled 179 candidates from September 2014 to May 2015, with 18 dropping out and 68 failing to meet the inclusion criteria. Eleven of the remaining 93 candidates also had to withdraw owing to school or work commitments. Eighty-two participants signed informed consent and underwent B-ultrasound and gynecological examinations. However, six participants of them did not meet the PD diagnostic criteria, and another eight were ruled out because of menstrual cycle disorders and unknown acupuncture history. Four of the remaining 68 subjects were removed because their VAS-P was less than 40mm before treatment (Fig. 1). Therefore, a total of 15 patients comprised the de qi group, and 53 comprised the non-de qi group.

Figure 1. Experimental design. PCOS: Polycystic ovary syndrome; PD: Primary dysmenorrhea; VAS-P: VAS for pain. Subgroup A: patients in the de qi group who experienced de qi; Subgroup B: patients in the de qi group who did not experience de qi; Subgroup C: patients in the non-de qi group who experienced de qi; Subgroup D: patients in the non-de qi group who did not experience de qi.

Patients whose date in the ADCAS revealed a contradiction between de qi experience and self-reported needling feelings (e.g., the overall needle sensation was scored as 0, while numbness was scored as 1) were eliminated, which led to one patient in the de qi group and seven in the non-de qi group subsequently being eliminated (Fig. 1). The remaining 14 patients in the de qi group all experienced de qi (100%, subgroup A; 0%, subgroup B). However, not all subjects in the non-de qi group were recognized as having an absence of de qi sensation: 25 individuals (59.5%, subgroup C) experienced a de qi sensation, while 17 patients (40.5%, subgroup D) did not.

2. Baseline characteristics

Participants’ ages, age at symptoms onset, period, menstrual cycle, duration, and VAS-P ratings before treatment did not differ significantly between the de qi and non-de qi group (Table 1). There were also no significant differences in the baseline characteristics of subgroups A, C, and D (Table 2).

Values are represent mean ± SE. *Test statistics by independent samples t-test (normal distribution data) or non-parametric statistics (non-normally distribution data). VAS-P, VAS score for pain..

&md=tbl&idx=1' data-target="#file-modal"">Table 1

Baseline characteristics of patients in the de qi and non-de qi groups (M ± SE).

Demographic information	De qi group (n = 15)	Non-de qi group (n = 49)	Statistic	p
Age (year)	21.40 ± 0.65	21.55 ± 0.43	370.00	0.974
Onset age (year)	15.13 ± 0.49	16.22 ± 0.35	–1.82	0.079
Period (day)	5.80 ± 0.30	5.39 ± 0.14	441.00	0.219
Menstrual cycle (day)	29.87 ± 0.85	30.22 ± 0.52	337.00	0.619
Disease duration (month)	79.53 ± 9.36	64.08 ± 6.34	0.39	0.537
VAS-p value (mm)	52.80 ± 2.95	54.90 ± 1.43	309.50	0.362

Values are represent mean ± SE. *Test statistics by independent samples t-test (normal distribution data) or non-parametric statistics (non-normally distribution data). VAS-P, VAS score for pain..

VAS-P: VAS score for pain; Subgroup A: patients in the de qi group who experienced de qi; Subgroup C: patients in the non-de qi group who experienced de qi; Subgroup D: patients in the non-de qi group who did not experience de qi (the same as in the following tables). Statistic, the F-value of the ANOVA or the chi-square value of the non-parametric test. *Test statistics by the ANOVA (normal distribution data) or non-parametric test..

&md=tbl&idx=2' data-target="#file-modal"">Table 2

Baseline characteristics of patients in subgroups (M ± SE).

Demographic information	Subgroup A (n = 14)	Subgroup C (n = 25)	Subgroup D (n = 17)	Statistic	p
Age (year)	21.21 ± 0.66	21.32 ± 0.63	21.53 ± 0.65	0.28	0.87
Onset age (year)	15.00 ± 0.50	15.92 ± 0.50	16.24 ± 0.55	1.21	0.31
Period (day)	5.71 ± 0.30	5.40 ± 0.23	5.41 ± 0.19	0.74	0.69
Menstrual cycle (day)	29.93 ± 0.91	30.40 ± 0.60	30.59 ± 1.17	0.52	0.77
Disease duration (month)	79.07 ± 10.04	62.88 ± 10.69	66.18 ± 9.52	3.27	0.19
VAS-p value (mm)	53.50 ± 3.08	54.36 ± 1.94	57.18 ± 2.73	0.54	0.59

VAS-P: VAS score for pain; Subgroup A: patients in the de qi group who experienced de qi; Subgroup C: patients in the non-de qi group who experienced de qi; Subgroup D: patients in the non-de qi group who did not experience de qi (the same as in the following tables). Statistic, the F-value of the ANOVA or the chi-square value of the non-parametric test. *Test statistics by the ANOVA (normal distribution data) or non-parametric test..

3. De qi sensation

The intensity, rapidity, and duration of the de qi sensation are shown in Table 3, while the intensity of the 12 components of de qi is shown in Table 4.

Table 3

The intensity, quickness, and duration of de qi (M ± SE).

Subgroup	N	Intensity of de qi	Quickness of de qi(s)	Duration of de qi(s)
Subgroup A	14	5.07 ± 0.16	4.66 ± 0.89	19.53 ± 2.44
Subgroup C	25	3.08 ± 0.16	81.92 ± 41.68	12.90 ± 1.93

Table 4

The intensity of 12 components of de qi (M ± SE).

Subgroups	N	Dull pain	Warmth	Coldness	Soreness	Numbness	Fullness/distention	Heaviness	Electric sensation	Formication	Throbbing	Spreading	Radiating
Subgroup A	14	1.64 ± 0.57	0	0.21 ± 0.15	2.93 ± 0.54	2.71 ± 0.51	4.07 ± 0.34	1.57 ± 0.5	1.71 ± 0.44	0.21 ± 0.21	1.07 ± 0.38	3.29 ± 0.38	0.79 ± 0.38
Subgroup C	25	0.72 ± 0.21	0.32 ± 0.22	0.48 ± 0.26	0.84 ± 0.24	2.4 ± 0.37	1.52 ± 0.32	1.12 ± 0.31	0.68 ± 0.23	0.44 ± 0.2	0.4 ± 0.15	1.68 ± 0.29	0.2 ± 0.16

4. Anxiety caused by needling

The anxiety levels caused by needling are shown in Table 5. More than 50% of the patients in subgroup A felt mild anxiety, while more than half in subgroup C felt thoroughly relaxed. None of the participants experienced “anxiety beyond endurance.”

Table 5

Anxiety caused by needling.

Subgroup	N	Thoroughly relaxed	Mild anxiety	Anxiety	Marked anxiety	Feeling anxiety beyond endurance
Subgroup A	14	2 (14.29%)	8 (57.14%)	2 (14.29%)	2 (14.29%)	0
Subgroup C	25	16 (64.00%)	8 (32.00%)	1 (4.00%)	0	0
Subgroup D	17	11 (64.70%)	6 (35.30%)	0	0	0

In subgroup A, no correlation was found between the degree of anxiety and de qi (p > 0.05). In subgroup C, the anxiety level was significantly positively correlated with throbbing (p = 0.041, r = 0.411) (Table 6). No significant difference in anxiety levels was found between subgroups C and D (0.05, 95% CI –0.30 to 0.40; p = 0.890).

*p < 0.05 by the Pearson or Spearman correlation analysis..

&md=tbl&idx=6' data-target="#file-modal"">Table 6

The results of the correlation analysis of anxiety and de qi.

	Intensity	Quickness	Duration	Dull pain	Warmth	Coldness	Soreness	Numbness	Fullness/distention	Heaviness	Electric sensation	Formication	Throbbing	Spreading	Radiating
Subgroup A
r	–0.212	–0.120	0.02	0.17	-	0.188	–0.5	–0.07	–0.048	–0.074	–0.18	–0.077	0.187	–0.352	0.057
p	0.466	0.682	0.947	0.561	-	0.519	0.069	0.813	0.87	0.801	0.539	0.795	0.522	0.217	0.846
Subgroup C
r	–0.055	–0.019	0.087	0.176	0.073	0.09	0.139	–0.09	0.263	0.126	–0.04	0.098	0.411*	–0.131	–0.219
p	0.792	0.929	0.68	0.399	0.729	0.668	0.507	0.67	0.204	0.549	0.85	0.64	0.041	0.533	0.293

*p < 0.05 by the Pearson or Spearman correlation analysis..

5. Expectations on the relationship between needle sensation and therapeutic effect

The cognition of needle sensation and the therapeutic effects are shown in Table 7. More than half of the patients in subgroups A and C believed that the needling sensation was relevant to the therapeutic effect, while relatively few patients thought that the needle sensation was unrelated to the curative effect.

Table 7

The cognition of the relationship between needle sensation and therapeutic effect.

Subgroup	N	Not relevant	Maybe relevant	Should be relevant	Must be relevant
Subgroup A	14	1 (7.14%)	1 (7.14%)	8 (57.14%)	4 (28.57%)
Subgroup C	25	1 (4.00%)	8 (32.00%)	13 (52.00%)	3 (12.00%)
Subgroup D	17	7 (41.18%)	5 (29.41%)	3 (17.65%)	2 (11.76%)

In subgroup A, patients’ expectations were negatively correlated with the rapidity of the de qi sensation (p = 0.049, r = –0.535), and positively correlated with an electric sensation (p = 0.005, r = 0.705). In subgroup C, the patients’ expectations were positively correlated with the intensity of de qi (p = 0.001, r = 0.612), coldness (p = 0.019, r = 0.464), and numbness (p = 0.01, r = 0.507) (Table 8). There was a significant difference between subgroups C and D in terms of participants’ views on the relationship between needle sensation and curative effect (0.72, 95% CI 0.16‒1.28; p = 0.015). Additionally, patients who experienced the de qi sensation had higher expectations for the association between needle sensation and curative effect than those who did not.

*p < 0.05, **p < 0.01 by the Pearson or Spearman correlation analysis..

&md=tbl&idx=8' data-target="#file-modal"">Table 8

The results of the correlation analysis of cognition and de qi.

	Intensity	Quickness	Duration	Dull pain	Warmth	Coldness	Soreness	Numbness	Fullness/distention	Heaviness	Electric sensation	Formication	Throbbing	Spreading	Radiating
Subgroup A
r	–0.232	–0.535*	0.079	0.379	-	0.251	–0.025	0.077	–0.097	0.508	0.705**	0.386	0.296	0.171	0.369
p	0.425	0.049	0.788	0.181	-	0.387	0.932	0.794	0.74	0.064	0.005	0.173	0.304	0.558	0.194
Subgroup C
r	0.612**	–0.024	0.054	0.135	0.315	0.464*	0.174	0.507**	–0.012	0.014	0.202	–0.109	–0.031	0.153	0.322
p	0.001	0.911	0.797	0.519	0.125	0.019	0.406	0.01	0.955	0.946	0.333	0.602	0.885	0.467	0.117

*p < 0.05, **p < 0.01 by the Pearson or Spearman correlation analysis..

6. Safety

No adverse effects were reported.

We conducted a secondary analysis of data obtained from an RCT about acupuncture de qi, where we had previously proven that patients who experienced de qi also experienced a bigger analgesic effect than those who did not [5,29]. The purpose of this paper, however, was to analyze the psychological factors influencing de qi. Our overall findings suggested that the feelings of anxiety and the patient’s view regarding the association between needle sensation and curative effect can influence the de qi of needling at SP6 in PD patients with cold and dampness stagnation.

Acupuncture-related anxiety is a psychological phenomenon often experienced by patients who undergo acupuncture. A previous study on PD patients found that a personality trait of anxiousness shows some correlation with de qi [13]. In the present study, 46.2% (18/39) of the patients who experienced de qi (subgroups A + C), were thoroughly relaxed throughout the process, 41% (16/39) felt mild anxiety, 7.7% (3/39) felt moderate anxiety, and 5.1% (2/39) experienced marked anxiety. It was also found that under conditions of light stimulation with a fine needle and shallow needling without manipulation, the throbbing sensation increased with the patient’s anxiety levels.

In China, patients undergoing acupuncture often believe that de qi significantly influences the curative effect. Some patients hope to experience de qi during treatment since they consider that the curative increases with the de qi sensation [44,45]. Influenced by the background of TCM culture, Chinese patients prefer experiencing acupuncture de qi more than American patients [42]. Gu and Ma believed that acupuncture de qi has a two-pronged effect of psychological suggestion and physiological regulation [46]. The patient’s expectation of the effect of acupuncture treatment often affects their sensory experience of the whole acupuncture process [47]. Also, patients’ understanding of the relationship between needle sensation and therapeutic effect, in part, reflects their expectations. The results of the present study showed that the more the patients’ believed that sensation was related to effect, the quicker and stronger the de qi sensation, and the stronger the intensity of coldness, numbness, and electric sensations. This suggests that when patients believe that the needle sensation affects acupuncture treatment, the intensity of de qi is stronger and occurs more quickly.

The expectations surrounding acupuncture treatment are of great significance for clinical practice [14,48]. By analyzing several RCTs of acupuncture for chronic pain, Linde et al. [49] found that patients’ expectation of pain relief was the only factor that correctly predicted the therapeutic effect. Another study found that positive psychological expectations can significantly enhance the analgesic effect of acupuncture, which is mainly manifested as a decrease in the subjective pain score and a change in the objective functional magnetic resonance signal under harmful stimulation conditions [50]. By contrast, a negative perception can change the relationship between test expectations and self-reported pain. The low-frequency components of heart rate variability after acupuncture were increased in patients with a negative perception of acupuncture, but not in patients with a positive perception [51]. The positive effect of expectancy psychology on de qi and the curative effect of acupuncture may be related to the regulation of the reward system [48].

The present study provides a preliminary discussion of the influence of psychological factors on de qi. However, there are several limitations. First, the evaluation of de qi was subjective. Although we explored acupoint surface temperature and evoked somatosensory potential to evaluate de qi in our previous research, further validation is necessary. Second, the sample size of this trial was relatively small.

This secondary analysis explored the influence of psychological factors (the participants’ acupuncture-related anxiety and expectations on the relationship between needle sensation and curative effect) on de qi. The results showed that participants with higher expectations for the association between needle sensation and curative effect were more likely to experience de qi. Clinical trials using both objective and subjective methods to assess de qi with a larger sample size should be urgently conducted to investigate the influence of body factors on de qi.

This research was supported by the National Basic Research Program of China (973 Program) — the Effect of De Qi on Acupoint Specific Effect Based on Meridians and its Characteristics and Molecular Response Mechanisms (No. 2012CB518506), the Scientific Research Development Fund Program of Beijing University of Chinese Medicine (No. 2018-ZXFZJJ-010).

Ni-juan Hu and Chun-hua Li contributed equally to this work. Jiang Zhu, Liang-xiao Ma and Pei Wang conceived and designed the experiments. Pei Wang and Gui-wen Wu performed the experiments. Ni-juan Hu and Chun-hua Li analyzed the data. Ni-juan Hu wrote the paper. Chun-hua Li revised the paper. All authors approved the final version on the manuscript accepted for publication.

The authors declare no conflict of interest.