6.1 Study Population
With the approval of the Ethics Committee of the Binzhou Medical University Hospital (under the Ethical Approval Number KYLL-186), 36 patients with pDOC were recruited to receive the XNKQ acupuncture intervention. All patients were enrolled from the Department of Rehabilitation Medicine at Binzhou Medical University Hospital between January 2024 and May 2025. This trial was registered in the Chinese Clinical Trial Registry (Identifier: ChiCTR2400090915) on 15 October 2024.
The inclusion criteria were as follows: (1) aged > 18 years(13); (2) had a DOC lasting more than 28 days(20); and (3) were diagnosed with VS/UWS or MCS according to the CRS-R scale. All patients met the diagnostic criteria for pDOC, according to the European Academy of Neurology guidelines for the diagnosis of coma and other disorders of consciousness in 2020(21).
The exclusion criteria were as follows: (1) unstable clinical conditions or vital signs; (2) scalp lesions or cranial defects and concurrent intracranial tumors, infections, or space-occupying lesions; (3) severe endocrine metabolic disorders or persistent status epilepticus(22); and (4) a history of neurological disorders or the use of anesthetic agents. Legally authorized representatives of all the subjects signed written informed consent forms. All enrolled patients received standard clinical care and conventional rehabilitation therapy as part of the baseline treatment protocol.
Thirty-six participants completed the study, including 16 patients with MCS and 20 patients with UWS. However, among the 36 participants, one MCS patient was excluded because of poor data quality. Prior to the experiment, the level of consciousness in patients with pDOC was evaluated using the CRS-R scale. Each patient underwent five assessments on nonconsecutive days within a 10-day period, and the highest score obtained was used to classify the patient as UWS or MCS(20). Table 1 summarizes the demographic and clinical assessment scores of all the participants. No significant differences in age distribution were observed between the two groups. However, significant disparities emerged in consciousness assessment scores, including Coma Recovery Scale-Revised (CRS-R) scores(23). The MCS group presented higher CRS-R scores (7.27 ± 2.31), whereas the UWS group presented lower CRS-R scores (3.50 ± 1.32).
Table 1
Basic characteristics of the patients.
Variable | Total (n = 35) | MCS (n = 15) | UWS (n = 20) |
|---|
Age, years, mean ± SD | 59.69 ± 13.79 | 59.20 ± 14.36 | 60.05 ± 14.36 |
Female, n (%) | 7(20%) | 4 (27%) | 3 (15%) |
CRS-R, mean ± SD | 5.11 ± 2.60 | 7.27 ± 2.31 | 3.50 ± 1.32 |
Time from onset to enrollment, mean ± SD(days) | 42.46 ± 1.42 | 42.53 ± 1.19 | 42.40 ± 1.60. |
Stroke, n (%) | 27(77%) | 13(87%) | 14(70%) |
Traumatic brain injury, n (%) | 1(3%) | 0(0%) | 1(5%) |
Hypoxic-ischemic encephalopathy, n (%) | 7(20%) | 2(13%) | 5(25%) |
CRS-R, Coma Recovery Scale-Revised; UWS, unresponsive wakefulness syndrome; MCS, minimally conscious state; SD, standard deviation.
6.2 Acupuncture and Point Selection
Among the 35 patients, 23 received acupuncture stimulation, while 12 received sham acupuncture stimulation. The acupuncture points applied in this study included the main acupoints Neiguan (PC6, bilateral), Renzhong (DU26, unilateral), Sanyinjiao (SP6, bilateral) and the secondary acupoints Jiquan (HT1, bilateral), Chize (Lu5, bilateral) and Weizhong (BL40, bilateral)(24). The anatomical location of these points was defined as follows: Renzhong (DU26) is situated at the junction between the upper one-third and middle one-third of the philtrum groove. Neiguan (PC6) is located 2 cun proximal to the transverse wrist crease, between the tendons of the flexor carpi radialis and palmaris longus muscles. Sanyinjiao (SP6) is positioned 3 cun above the medial malleolus, along the posterior border of the tibia. Jiquan (HT1) is situated at the apex of the axilla, where the axillary artery beats. Chize (Lu5) is located in the transverse stripes of the elbow, the radial depression of the bicipital brachii tendon. Weizhong (BL40) is located in the posterior region of the knee at the midpoint of the popliteal transverse stripes (Fig. 4). For the sham acupuncture control group, non-acupoints near the main acupoints were selected for minimal, superficial stimulation to control for nonspecific effects. These control points were chosen on the basis of a validated protocol from a previous study and through consultation with acupuncture experts. They are not situated on any recognized meridian or over major nerves but are all within a 2.5 cm radius of the corresponding true acupoint. The specific locations were defined as follows:
The sham point for PC6 was located lateral to PC6, in the intermeridian space between the Hand-Taiyin Lung Meridian and the Hand-Jueyin Pericardium Meridian. The sham point for DU26 was situated on the vertical line of the mouth, lateral (left) to DU26. The sham point for SP6 was located 6 cun proximal to the tip of the medial malleolus, in the intermeridian space between the Foot-Taiyin Spleen Meridian and the Foot-Jueyin Liver Meridian. This point is 3 cun proximal to the SP6 and 1.25 cun anterior to it on the medial side of the tibia. The sham point for HT1 was located 1 cun inferior and 1 cun anterior to the axillary apex, avoiding axillary hair, on the muscle belly of the bicipital brachii. The sham point for Lu5 was located at the midpoint of the line connecting Lu5 and LI11 (Quchi) when the elbow was flexed at 120 degrees on the radial side of the tendon of the radial wrist extensor longus muscle(11, 25).
The sham point for BL40 was located inferior and lateral to the midpoint of the popliteal transverse crease on the muscle belly of the peroneus longus muscle. The acupuncture procedures were conducted by a licensed acupuncturist with a decade of clinical experience from the Binzhou Medical University Hospital. All the subjects remained in a supine position on a scanning bed throughout the intervention. Sterile, single-use titanium needles (0.20×40 mm; Wujiang CloudDragon Medical Apparatus Co., Ltd.,Wujiang District, Suzhou, China) were utilized.
For the XNKQ acupuncture protocol, the needling techniques were executed as described below. At Neiguan (PC6), bilateral insertion was performed to a depth of 0.5–1.0 cun. A reducing stimulation method was applied for 1 minute, involving lifting-thrusting and rotating manipulations, with the left hand rotating the needle counterclockwise and the right hand rotating it clockwise. Subsequently, the needle at Renzhong (DU26) was obliquely inserted toward the nasal septum to a depth of 0.3–0.5 cun, followed by the sparrow-pecking technique until ocular moistening or lacrimation was observed. Finally, at Sanyinjiao (SP6), bilateral oblique insertion (0.5–1.0 cun depth) along the medial tibial border was performed by reinforcing manipulation via the lifting-thrusting method, characterized by heavy thrusting and light lifting, which was maintained for 1 minute. After the main points were needled, the patient’s secondary points, bilateral Jiquan (HT1), bilateral Chize (Lu5), and bilateral Weizhong (BL40), were needled using the lifting-thrusting method(24). The acupuncture intervention, encompassing both needle manipulation and retention time, lasted for 30 minutes in both the verum and sham acupuncture groups(26).
6.3 Data Acquisition
fNIRS data were acquired using a 48-channel NirSmart system (Danyang Huichuang, China). The fNIRS cap, designed according to the international 10–20 system, comprised 15 light sources and 16 detectors to obtain resting-state brain network data. The source‒detector distance was maintained within 3 cm (2.9–3.1 cm). Each channel primarily detected hemodynamic changes in the cortical region beneath the midpoint of the source‒detector pair, with channel localization mapped to Brodmann areas. The signals were recorded at wavelengths of 760 nm and 850 nm, with a sampling frequency of 11 Hz. The 48 channels were assigned to eight key regions of interest (ROIs) in the cerebral cortex: the left frontal lobe (LFL; channels 9, 10, 11, 27, 29, 30, 31, 32, 43, 44, 45, 46, and 47), right frontal lobe (RFL; channels 6, 7, 8, 21, 22, 23, 24, 26, 37, 39, 40, 41, and 42), left temporal lobe (LTL; channels 12, 13, 14, 15, 16, 33, and 35), right temporal lobe (RTL; channels 1, 2, 3, 4, 5, 17, and 18), left primary motor cortex (LPMC; channels 34 and 48), right primary motor cortex (RPMC; channels 19 and 38), left dorsolateral prefrontal cortex (LDLPFC; channels 10, 29, 30, 43, 44, and 45), and right dorsolateral prefrontal cortex (RDLPFC; channels 21, 23, 40, and 41) (27)(Fig. 5).
During fNIRS data acquisition, the participants initially rested in a comfortable supine position for 5 minutes to stabilize their physiological parameters before donning the fNIRS cap. The data acquisition protocol consisted of three phases: the preacupuncture resting-state phase, during which baseline brain activity was recorded for 8 minutes; the acupuncture intervention phase, during which real-time fNIRS signals were acquired throughout the acupuncture procedure (duration adjusted according to the actual manipulation time, typically 20 minutes); and the postacupuncture resting-state phase, during which postintervention resting-state signals were recorded for 8 minutes following needle removal. The final analyses were based on three 8 minute resting-state fNIRS epochs: pretreatment, middle segment of the during-treatment period, and posttreatment. Adopting uniform 8-minute windows aligns with established methodological standards in the fNIRS literature and ensures comparable signal quality across phases(13, 16, 28, 29).
6.4 Data Processing
The fNIRS data were processed and analyzed using NirSpark software. Raw light intensity signals were first converted into optical density (OD) curves, and oxygenated hemoglobin (HbO) and deoxygenated hemoglobin (HbR) concentrations were derived using the modified Beer‒Lambert law. During preprocessing, bandpass filtering (0.01–0.2 Hz) was applied to eliminate motion artifacts and baseline drift caused by physiological fluctuations (e.g., cardiac and respiratory cycles)(30). Subsequent analyses focused exclusively on HbO signals because their signal-to-noise ratio is superior to that of HbR.
FC analysis was performed by extracting HbO time series before, during, and after acupuncture. For each ROI, HbO time series were aggregated across corresponding channels. Pairwise Pearson correlation coefficients were computed between the HbO time series of all channel pairs, with these coefficients defined as the functional connectivity strength between associated brain regions(31). For further network-level analysis, time-averaged HbO signals from all 48 channels were subjected to correlation analysis. Pearson correlation coefficients underwent Fisher-z transformation to normalize their distributions, generating standardized metrics for quantifying dynamic changes in FC strength across the three experimental phases(32).
6.5 Statistical Analysis
Statistical analyses were performed using SPSS 27.0 software. Between-group comparisons were conducted using two-tailed Student's t tests, whereas multigroup comparisons were evaluated with one-way analysis of variance (ANOVA). A threshold of P < 0.05 was applied to define statistical significance.