Abstract
Purpose
About 30–50% of lung cancer patients suffer from insomnia with limited prevention options. We conducted a retrospective cohort study to investigate whether acupuncture for cancer-related symptoms such as pain could reduce the incidence of insomnia symptoms.
Methods
We included newly diagnosed lung cancer patients without prior insomnia symptoms from Guangdong Provincial Hospital of Chinese Medicine (2012–2021) and followed up until January 31, 2022. Cases were propensity score–matched by sex, age, ECOG PS score, pathological diagnosis, therapy, and comorbidities. The main outcome was the incidence of insomnia symptoms, identified by psychiatric diagnosis or insomnia symptoms in medical records.
Results
Among 1295 lung cancer survivors, 918 patients were matched (459 per group). Over an average follow-up of 218.6 days, insomnia symptoms were observed in 20.9% of the acupuncture group and 42.2% of the non-acupuncture group. Acupuncture was linked to a lower incidence of insomnia symptoms (adjusted HR 0.31, 95% CI 0.23–0.39). Higher frequency and more numerous acupuncture sessions further reduced the risk of insomnia symptoms (adjusted HR 0.15 for frequency of 1 day/session, adjusted HR 0.15 for sessions > 16). The acupuncture group had a significantly lower cumulative incidence of insomnia (log-rank test, P < 0.01).
Conclusion
Our results suggest that acupuncture may prevent the incidence of insomnia symptoms in patients with lung cancer. Daily sessions and a cumulative total of 16 or more treatments may enhance the benefit.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00520-025-10256-8.
Keywords: Insomnia symptoms, Lung cancer, Prevention, Acupuncture
Introduction
Insomnia symptoms, including difficulties initiating or maintaining sleep and dissatisfaction with sleep duration or quality [1], are common and disturbing in lung cancer survivors. About 30–50% of lung cancer survivors report having the condition [2–4], which is around twice to three times that of the general population [5–7]. The adverse consequences of insomnia symptoms significantly impair lung cancer survivors in multiple dimensions, including quality of life [8], physical and mental health [9], financial burden [10], and even survival duration [11]. Furthermore, insomnia symptoms can hinder the effectiveness of cancer treatment and negatively influence prognosis [12]. International guidelines emphasize the importance of long-term insomnia disorder prevention and management [13, 14]. The prevention of insomnia disorder in cancer patients is the desired objective [14]. However, research in prevention of insomnia symptoms or disorder remains limited. Preventive measures predominantly focus on sleep hygiene education [14]. A study has examined the feasibility of a brief, self-administered preventive intervention for insomnia symptoms, but it has been restricted by a small sample size and the absence of a control group [15].
Acupuncture, a safe and well-tolerated treatment [16], is commonly used for cancer-related conditions worldwide [16]. Recently, the uniquely preventive effect of acupuncture, which is achieved through stimulating acupoints to restore bodily balance and enhance overall health, has attracted attention, especially on the prevention of side effects of chemotherapy and radiotherapy [17, 18], postoperative cognitive dysfunction [19], and cardiovascular diseases [20]. However, research on its preventive effect on insomnia symptoms is limited. A study evaluated preoperative electroacupuncture in breast cancer patients, targeting postoperative anxiety and sleep quality. Sleep quality was not the primary outcome and was assessed only via quality-of-life scales, with a short follow-up period [21]. Thus, more evidence is needed to confirm acupuncture’s preventive effect on insomnia symptoms. We aimed to examine the association between acupuncture and incidence of insomnia symptoms in lung cancer survivors through a real-world study.
Methods
This study was conducted at Guangdong Provincial Hospital of Chinese Medicine, a major medical center comprising four hospital branches across different regions of Guangzhou, China, with shared access to the Electronic Medical Records Database (EMRD). The Research Ethics Committee of Guangdong Provincial Hospital of Chinese Medicine granted ethical approval for this study (approval no.YE2021-315-01). Since all research data were completely anonymised, the requirement for informed consent was waived. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines.
Study design and setting
To investigate whether acupuncture could reduce the risk of developing insomnia symptoms, we conducted a retrospective propensity score–matched cohort study. Data were retrospectively reviewed from the establishment of the EMRDin 2012 and followed up until January 31, 2022.
Data collection and management
Trained data collectors collected electronic medical record data using structured questionnaires via the Wenjuanxing (WJX) platform. They initially evaluated whether participants satisfied the criteria. Then, they systematically collected data on demographics, diagnosis, ECOG PS score, treatment, medical history, acupuncture specifics (frequency, sessions, type, points, purpose), and endpoint event details (insomnia symptoms, onset time, diagnosis and medication use). WJX offers diverse question types and logic settings that skip irrelevant questions based on prior answers, enhancing the survey experience. The questionnaire consists of 23 questions of various types, including single-choice, multiple-choice, date-selection, and fill-in-the-blank questions. All key information is set as required fields in the collection questionnaire, which can only be submitted once all required fields are completed. Cases with missing key data are excluded and documented. This approach largely ensures the integrity of the collected data.
Quality controllers verified data accuracy, completeness, and alignment with research objectives. A designated researcher overlooked data management, including database designing, maintaining, monitoring data collection, ensuring security and confidentiality, performing backups, assisting data collectors and quality controllers, and preparing data for analysis.
All researchers have backgrounds in clinical medicine and oncology and have received comprehensive training to conduct all research procedures following the standard operating procedure (SOP) (Supplementary 1). The study flowchart is shown in Fig. 1.
Fig. 1.
Flow diagram
Participants
Inclusion criteria:
Patients newly diagnosed with lung cancer (ICD-10 C34.900) at Guangdong Hospital of Chinese Medicine between 2012 and 2021, as recorded in the hospital’s Electronic Medical Records Database (EMRD). “Newly diagnosed” specifically refers to individuals for whom the first diagnosis of lung cancer was made at Guangdong Provincial Hospital of Chinese Medicine.
Patients who were aged 18 years or older at the time of lung cancer diagnosis.
Exclusion criteria:
Patients with a history of insomnia symptoms before or at the time of lung cancer diagnosis.
Patients who exhibited consciousness disturbances such as coma, somnolence, or delirium at the time of lung cancer diagnosis.
Patients with a follow-up period shorter than 14 days.
Patients whose data sources were missing key information like age, sex, and past medical history.
Patient diagnosis information was determined based on the patient’s diagnoses and ICD-10 codes in the EMRD.
Acupuncture exposures
Exposures were defined as receiving acupuncture after a lung cancer diagnosis. Information from medical records included treatment intent, acupoints, number of courses (acupuncture treatment of one hospitalization equals one course), cumulative days (total duration from first to last session per hospital stay), acupuncture frequency (days per session), and total sessions. Mean frequency was calculated as the sum of frequencies divided by the sum of courses.
Endpoint events
The primary endpoint was the occurrence of insomnia symptoms, defined by meeting either
a psychiatrist’s diagnosis of insomnia disorder (ICD-10 G47.0) or
documented sleep complaints—difficulty initiating or maintaining sleep, early-morning awakening, or dissatisfaction with sleep duration or quality.
Potential confounders
Based on clinicians’ recommendations and literature reviews [4, 22], a directed acyclic graph was used to determine the minimally sufficient adjustment set for identifying potential confounders [23]. The minimally sufficient adjustment set, determined using DAGitty v3.1, comprises covariates that minimize confounding bias when assessing the exposure-outcome relationship. The final minimally sufficient adjustment set comprised sex, age, ECOG PS score, pathological diagnosis, therapy, and comorbidities (Fig. 2).
Fig. 2.
A directed acyclic graph represents associations between covariates and primary exposure and outcome. Pink circles represent ancestors of the exposure and outcome (i.e., age), blue circles represent ancestors of the outcome (i.e., pain), green lines represent causal paths, and pink lines represent biasing paths. The minimally sufficient adjustment set represents covariates such that the adjustment for this set of variables will minimize confounding bias when estimating the association between the exposure and the outcome. The minimally sufficient adjustment set was determined using the DAGitty v3.1. The final minimally sufficient adjustment set comprised age, sex, comorbidities, therapy, pathological diagnosis, and ECOG PS score exposure. Abbreviations: ECOG PS score, Eastern Cooperative Oncology Group Performance Status score
Comorbidities included infection, diabetes, hypertension, and mental disorders. Therapy information included chemotherapy, radiation therapy, immunotherapy, target therapy, acupuncture, surgery, opioids, and Chinese herbal medicine. Pathological diagnosis included small cell lung cancer and non-small cell lung cancer. The ECOG PS score is one of the widely used methods to assess the functional status of a patient. It describes a patient’s level of functioning in terms of their ability to care for themself, daily activity, and physical ability (walking, working, etc.) [24]. Higher scores mean poorer performance status. Pain was excluded from primary analyses for the following reasons. First, acupuncture reduces pain, which in turn reduces insomnia, making pain a mediator in the acupuncture–insomnia relationship [25]. Adjusting for pain could paradoxically weaken acupuncture’s observed protective effect by masking its indirect therapeutic pathway [26]. Second, increased pain prompts more acupuncture use and also heightens insomnia risk, thus acting as a confounder [27]. Under severe pain, one might observe a spurious correlation of increased acupuncture and increased insomnia. Controlling for pain would be necessary to reveal the true relationship between acupuncture and insomnia [26]. However, if acupuncture still predicts reduced insomnia incidence without controlling for pain, this indicates that acupuncture’s true protective effect is robust enough to counteract the negative influence of pain.
Statistical analysis
Missing data management
Among 586 patients in the acupuncture group, 107 records (18.3%) lacked mean acupuncture frequency, session number, total treatment days, or acupuncture courses; the missing-data pattern was deemed missing at random. Within the acupuncture subset, five multiply-imputed datasets (m = 5, random seed = 123) were created for mean acupuncture frequency, total treatment days, and acupuncture courses using the R package mice with predictive mean matching (PMM). Session count was calculated post-imputation as total treatment days/mean frequency. After imputation, the distributions of the five completed datasets closely matched that of the original. The primary outcome (the association between acupuncture used and insomnia symptoms incidence) does not require these imputed variables.
Propensity scores match
To minimize the potential data differences between the two cohorts, 1:1 greedy nearest neighbors matching was performed, using matching without replacement and a caliper width of 0.02 times the pooled SD of the logit of the propensity scores for the cohort [28]. Propensity score matching (PSM) based on sociodemographic (age, sex) and clinical factors (ECOG PS score, pathological diagnosis, therapy, and comorbidities) minimized potential data differences between the two cohorts.
Standardized mean difference (SMD) was used to evaluate the covariate balance between the acupuncture and control groups, for which a difference of less than 0.10 was considered to indicate good balance [29].
Cox proportional hazards models
After 1:1 PSM, we performed univariable and multivariable Cox regression with robust standard errors for acupuncture use, sex, age, pathological diagnosis, ECOG PS score, chemotherapy, radiotherapy, targeted therapy, immunotherapy, surgery, opioid use, Chinese herbal medicine, infection, hypertension, diabetes, and mental disorder. We also stratified patients by median acupuncture frequency and by quartiles of total sessions and evaluated each stratification in the same univariable and multivariable Cox models. Hazard ratios (HRs), adjusted hazard ratios (aHRs), and 95% confidence intervals (CIs) were reported.
In cohort study design, matching the follow-up start (termed “index date” in our study) for exposed and control groups is key to avoiding differential immortal time periods [30]. When exact matching of exposure times is not feasible, randomly generating control group index dates is a viable solution [30]. This method ensures follow-up comparability and reduces immortal time bias [31].
In this study, the index date of the acupuncture group is the date of the first acupuncture session after the new lung cancer diagnosis. However, the non-acupuncture group has no direct counterpart to this index date. Using the lung cancer diagnosis date as the follow-up start would lead to an overestimation of event-free time in the non-acupuncture group, a dilution of the outcome incidence rate, and a potential exaggeration of the non-acupuncture group’s protective effect. To address this, for the non-acupuncture group, we used a date randomly selected within the same calendar year as the follow-up start date, bounded by the study enrollment date and the end of follow-up.This approach minimizes immortal time bias and makes the two groups’ follow-up start comparable.
Kaplan-Meier method and log-rank test
The Kaplan-Meier method was used to estimate the cumulative incidence of insomnia in acupuncture and control groups, and the log-rank test was used to test the difference.
Analysis of acupoints
The association between each acupoint and insomnia status was examined using a univariate chi-square test. The top 10 acupuncture points, which were most strongly associated with non-insomnia status and had the highest non-insomnia proportions, were identified. A random forest classifier with 100 decision trees was trained to analyze the association between acupoints and insomnia. After training, feature importance values were extracted and ranked. The top 15 acupoints by feature importance were selected.
To analyze the relationship between acupoint combinations and insomnia status, the frequency of common combinations was counted. The proportion and number of non-insomnia samples were then calculated for the top 10 acupoint combinations. We used the Apriori algorithm to mine association rules between acupoint combinations and non-insomnia status. Frequent item sets were found with a minimum support of 10%, and association rules were generated with a minimum confidence of 50%. Rules containing “no insomnia” were selected and ranked by confidence, with the top 10 rules identified.
Sensitivity analysis
Four sensitivity analyses were conducted. First, univariate and multivariate Cox proportional hazards regression analyses were fitted to the full cohort (n = 1295). Second, patients with only a single acupuncture session were excluded, and the multivariate model was re-estimated, as one single acupuncture session may provide only limited therapeutic benefit. Third, Chinese herbal medicine use was adjusted for in the primary model and then omitted in a subsequent multivariate Cox regression. Fourth, the effects of acupuncture frequency and session count on insomnia incidence were evaluated across an incomplete-case analysis. In addition to the full-cohort sensitivity analysis, all remaining sensitivity analyses are presented in Supplementary 2.
Furthermore, the E-value for the statistically significant primary outcome was calculated to assess the robustness of observed associations to potential unmeasured confounding [32].
All statistical analyses were performed using R Studio and Python, and differences with a two-sided P-value of less than 0.05 were considered statistically significant.
Results
Characteristics of the cohorts
Among 1295 newly diagnosed lung cancer patients between 2012 and 2021, 586 (45.3%) received acupuncture. After 1:1 propensity score matching (PSM), 459 acupuncture and 459 non-acupuncture patients were successfully paired. The characteristics of lung cancer survivors according to acupuncture exposure are shown in Table 1. After PSM, the two cohorts were well balanced across sex, ECOG PS score, pathological diagnosis, nearly all comorbidities, and therapeutic modalities, with standardized mean differences (SMD) < 0.1 for each variable. Minor residual imbalances were observed for chemotherapy (SMD = 0.100), infection (SMD = 0.110), and age group (SMD = 0.109) (Fig. 3).
Table 1.
Characteristics of patients before and after propensity score matching
| Propensity score–matched patientsa | Full-cohort patients | |||||
|---|---|---|---|---|---|---|
| Characteristics | Non-acupuncture (n = 459) |
Acupuncture (n = 459) |
SMD | Non-acupuncture (n = 709) |
Acupuncture (n = 586) |
SMD |
| Sex | ||||||
| Women | 143 (31.2%) | 146 (31.8%) | 0.014 | 222 (31.3%) | 169 (28.8%) | 0.054 |
| Men | 316 (68.8%) | 313 (68.2%) | 487 (68.7%) | 417 (71.2%) | ||
| Age group | ||||||
| 18–63 y | 244 (53.2%) | 219 (47.7%) | 0.109 | 373 (52.6%) | 268 (45.7%) | 0.138 |
| ≥ 64 y | 215 (46.8%) | 240 (52.3%) | 336 (47.4%) | 318 (54.3%) | ||
| Mean ± SD | 62 ± 11 | 64 ± 10 | 63 ± 11 | 64 ± 10 | ||
| ECOG PS score | ||||||
| 1–2 | 309 (67.3%) | 300 (65.4%) | 0.041 | 467 (65.9%) | 372 (63.5%) | 0.05 |
| 3–4 | 150 (32.7%) | 159 (34.6%) | 242 (34.1%) | 214 (36.5%) | ||
| Pathological diagnosis | ||||||
| NSCLC | 411 (89.5%) | 416 (90.6%) | 0.000 | 642 (90.6%) | 504 (86.0%) | 0.142 |
| SCLC | 48 (10.5%) | 43 (9.4%) | 67 (9.4%) | 82 (14.0%) | ||
| Therapy | ||||||
| Chemotherapy | 318 (69.3%) | 297 (64.7%) | 0.100 | 462 (65.2%) | 411 (70.1%) | 0.106 |
| Radiotherapy | 73 (15.9%) | 75 (16.3%) | 0.011 | 106 (15.0%) | 106 (18.1%) | 0.085 |
| Immunotherapy | 101 (22.0%) | 113 (24.6%) | 0.056 | 133 (18.8%) | 182 (31.1%) | 0.287 |
| Target therapy | 222 (48.4%) | 225 (49.0%) | 0.013 | 302 (42.6%) | 294 (50.2%) | 0.152 |
| Surgery | 26 (5.7%) | 30 (6.5%) | 0.033 | 59 (8.3%) | 45 (7.7%) | 0.024 |
| Opioids | 140 (30.5%) | 142 (30.9%) | 0.009 | 218 (30.7%) | 192 (32.8%) | 0.043 |
| Chinese herbal medicine | 424 (92.4%) | 417 (90.8%) | 0.058 | 635 (89.6%) | 543 (92.7%) | 0.109 |
| Comorbidity | ||||||
| Infection | 15 (3.3%) | 26 (5.7%) | 0.110 | 65 (9.2%) | 29 (4.9%) | 0.165 |
| Hypertension | 118 (25.7%) | 135 (29.4%) | 0.083 | 195 (27.5%) | 163 (27.8%) | 0.007 |
| Diabetes | 45 (9.8%) | 45 (9.8%) | 0.000 | 82 (11.6%) | 63 (10.8%) | 0.026 |
| Mental disorder | 2 (0.4%) | 2 (0.4%) | 0.000 | 2 (0.3%) | 2 (0.3%) | 0.011 |
| Follow-upb | ||||||
| Mean ± SD | 154 ± 274 | 283 ± 405 | - | 143 ± 248 | 282 ± 400 | - |
| Endpoint event | 194(42.2%) | 96(20.9%) | - | 284 (40.1%) | 125 (21.3%) | - |
SMD, standardized mean difference; SD, standard deviation; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; ECOG PS score, Eastern Cooperative Oncology Group Performance Status score
a1:1 propensity score matching (caliper = 0.02) without replacement was conducted to pair acupuncture and non-acupuncture patients, based on sex, age, ECOG PS score, pathological diagnosis, chemotherapy, radiotherapy, target therapy, immunotherapy, surgery, opioids, Chinese herbal medicine, infection, hypertension, diabetes, and mental disorder. Balance was evaluated by standardized mean differences (< 0.10 = adequate)
bThe follow-up duration is from the index date to the date of insomnia symptoms, or the event of critical illness, vital organ failure, coma, lethargy, or death, or January 31, 2022
Fig. 3.
Standardized mean differences in the unmatched and matched samples. Abbreviations: ECOG PS score, Eastern Cooperative Oncology Group Performance Status score. 1:1 propensity score matching (caliper = 0.02) without replacement was conducted to pair acupuncture and non-acupuncture patients, based on sex, age, ECOG PS score, pathological diagnosis, chemotherapy, radiotherapy, target therapy, immunotherapy, surgery, opioids, Chinese herbal medicine, infection, hypertension, diabetes, and mental disorder. Balance between the groups was assessed before and after matching by comparing SMDs for each variable for which a difference of less than 0.10 was considered to indicate adequate balance
In the full cohort, patients underwent acupuncture every 1.5 ± 1.0 days per session (median 1.0, IQR 1.0–2.0), totaling 18.2 ± 29.2 sessions over 4.4 ± 4.9 courses. In the PSM cohort, the frequency was 1.4 ± 0.2 days per session (median 1.0, IQR 1.0–2.0), with 17.1 ± 27.7 sessions over 4.2 ± 4.5 courses. Sessions were categorized by quartiles, and frequency was grouped into high (1 day per session), moderate (2–3 days per session), and low (> 3 days per session) based on clinical practice. Body acupuncture was the most common type. The most common purposes of acupuncture treatment were cancer pain and systemic cancer-related symptoms like fatigue, hot flashes, weight loss, and respiratory issues (Table 2).
Table 2.
Information of acupuncture treatment in acupuncture cohort
| Propensity-Score Matched Patients | Full-cohort Patients | |
|---|---|---|
| Acupuncture users (n=459) | Acupuncture users (n=586) | |
| Acupuncture coursea | ||
| Median (Q1, Q3) | 3 (1, 5) | 3 (1, 5) |
| Acupuncture session (times) | ||
| Median (Q1, Q3) | 7 (3, 15) | 7 (3, 18) |
| Median (min, max) | 7 (1, 233) | 7 (1,233) |
| Subgroup | ||
| Low (<3) | 117 (12.7%) | 147 (11.3%) |
| Medium-low (3-7) | 124 (13.5%) | 133 (10.2%) |
| Medium-high (7-15) | 103 (11.2%) | 159 (12.2%) |
| High (>16) | 115 (12.5%) | 147 (11.3%) |
| Acupuncture mean frequency (days per session)b | ||
| Median (Q1, Q3) | 1 (1, 2) | 1 (1, 2) |
| Median (min, max) | 1 (1,7) | 1 (1, 13) |
| Subgroup | ||
| Low (>3) | 9 (0.9%) | 17 (1.3%) |
| Medium (2-3) | 116 (12.6%) | 156 (12.0%) |
| High (1) | 331 (36.0%) | 134 (31.8%) |
| Type of acupuncture | ||
| Body acupuncture (BA) | 414 (90.2%) | 494 (84.3%) |
| Electroacupuncture (EA) | 4 (0.9%) | 5 (0.9%) |
| Scalp acupuncture (SA) | 16 (3.5%) | 20 (3.4%) |
| BA plus EA | 15 (3.3%) | 21 (3.6%) |
| BA plus SA | 17 (3.7%) | 19 (3.2%) |
| Main purpose of acupuncture | ||
| Cancer pain | 137 (29.8%) | 161 (27.5%) |
| Cancer-related symptoms (systemic) c | 136 (29.6%) | 159 (27.1%) |
| Neurological symptoms | 57 (12.4%) | 67 (11.4%) |
| Digestive symptoms | 47 (10.2%) | 58 (9.9%) |
| Respiratory symptoms | 100 (21.8%) | 120 (20.5%) |
| Circulatory symptoms | 7 (1.5%) | 9 (1.5%) |
| Urinary symptoms | 3 (0.7%) | 5 (0.9%) |
| Endocrine and nutritional metabolism symptoms | 2 (0.4%) | 3 (0.5%) |
| Psychiatric symptoms | 11 (2.4%) | 13 (2.2%) |
| Othersb | 61 (13.3%) | 78 (13.3%) |
a An acupuncture course is defined as all acupuncture treatments delivered during a single hospitalization.
b Mean acupuncture frequency= Σ acupuncture frequencys / Σ acupuncture courses.
c Cancer-related symptoms (systemic) include fatigue, hot flashes and weight loss.
d Others include those cases where symptoms were not clearly documented.
Relationship between acupuncture use and insomnia symptoms
Among the propensity score–matched survivors, 290 (31.5%) developed insomnia symptoms over an average follow-up of 218.6 days. The incidence of insomnia symptoms in the acupuncture group was 96 (20.9%) and 194 (42.2%) in the non-acupuncture group. Those who received acupuncture experienced a lower incidence of insomnia symptoms than their non-acupuncture counterparts (HR 0.32, 95% CI 0.25–0.41; adjusted HR 0.31, 95% CI 0.23–0.39). Results were consistent when the full-cohort analysis was performed.
In the acupuncture cohort, both increasing session counts and treatment frequency were associated with a lower risk of insomnia symptoms. Compared with no acupuncture, the adjusted hazard ratios (95% CI) for insomnia symptoms were 0.47 (0.3–0.72), 0.45 (0.30–0.68), 0.29 (0.18–0.47), and 0.15 (0.09–0.24) across ascending session categories (P < 0.001). Similarly, relative to no acupuncture, the adjusted hazard ratios of mean frequency were 0.15 (0.10–0.22) in the high-frequency group and 0.58 (0.43–0.79) in the moderate-frequency group. Both crude and adjusted hazard ratios for the low-frequency group exceeded 1.0, although the elevations lacked statistical significance. These relations were consistent across the full cohort (Table 3) and the incomplete cohort (Supplementary 2 Table 3a & 3b).
Table 3.
Hazard Ratios of acupuncture and acupuncture dose for Insomnia symptoms
| Variables | Propensity-score matched patients | Full-cohort patients | |||
|---|---|---|---|---|---|
| Crude HR | Adjusted HR a | Variables | Crude HR | Adjusted HR a | |
| (95% CI) | (95% CI) | (95% CI) | (95% CI) | ||
| Acupuncture b | 0.32 (0.25-0.41)*** | 0.31 (0.23-0.39)*** | Acupuncture | 0.33 (0.27-0.41)*** | 0.30 (0.24-0.37)*** |
| Acupuncture session (times) | |||||
| None | Ref. | Ref. | None | Ref. | Ref. |
| Low (<3) | 0.49 (0.32-0.73)*** | 0.47 (0.31-0.72)*** | Low (<3) | 0.55 (0.39-0.77)*** | 0.49 (0.35-0.69)*** |
| Medium-low (3-7) | 0.54 (0.37-0.79)** | 0.45 (0.30-0.68)*** | Medium-low (3-7) | 0.59 (0.42-0.83)** | 0.49 (0.34-0.71)*** |
|
Medium-high (7-15) |
0.31 (0.19-0.50)*** | 0.29 (0.18-0.47)*** |
Medium-high (7-18) |
0.38 (0.27-0.55)*** | 0.35 (0.24-0.50)*** |
| High (>16) | 0.15 (0.10-0.25)*** | 0.15 (0.09-0.24)*** | High (>18) | 0.10 (0.06-0.17)*** | 0.09 (0.06-0.15)*** |
| Mean acupuncture frequency c (days per session) | |||||
| None | Ref. | Ref. | None | Ref. | Ref. |
| Low (5-7) | 1.01 (0.45-2.27) | 1.04 (0.44-2.43) | Low (4-7) | 0.97 (0.53-1.78) | 0.80 (0.43-1.48) |
| Medium (2-3) | 0.64 (0.48-0.86)** | 0.58 (0.43-0.79)*** | Medium (2-3) | 0.61 (0.47-0.79)*** | 0.55 (0.42-0.72)*** |
| High (1) | 0.16 (0.11-0.23)*** | 0.15 (0.10-0.22)*** | High (1) | 0.16 (0.12-0.23)*** | 0.15 (0.11-0.21)*** |
HR, hazard ratios; CI, confidence interval. *P < 0.05, **P < 0.01, and ***P < 0.001
aShown are the hazard ratios from the multivariable Cox proportional hazards model, adjusted for sex, age, ECOG PS score, pathological diagnosis, chemotherapy, radiotherapy, target therapy, immunotherapy, surgery, opioids, Chinese herbal medicine, infection, hypertension, diabetes, and mental disorder
bAfter PSM, the univariable and multivariable Cox regression with robust standard errors were conducted
cMean acupuncture frequency = Σ acupuncture frequency/Σ acupuncture courses. An acupuncture course is defined as all acupuncture treatments delivered during a single hospitalization
The cumulative incidence of insomnia was lower in the acupuncture cohort than in the non-acupuncture cohort (log-rank test, P < 0.01). The median time to the incident of insomnia symptoms was 0.3 years in the non-acupuncture group and 2.3 years in the acupuncture group (Fig. 4).
Fig. 4.
The cumulative incidence of insomnia symptoms among propensity- score –matched samples
Risk factors of insomnia symptoms
According to the results of the multivariable analysis, having an ECOG PS score of 3–4 (adjusted HR 1.53, 95% CI 1.23–1.90 in full cohort, adjusted HR 1.58, 95% CI 1.19–2.09 in matched cohort), receiving opioids (adjusted HR, 1.44 95% CI 1.07–1.93), and having a mental disorder (adjusted HR 6.31, 95%CI 1.17–33.92) increased the risk of insomnia symptoms no matter in the unmatched or matched sample. Additionally, lung cancer patients who aged above 64 years old (adjusted HR 1.30, 95% CI 1.01–1.68) and diagnosed with SCLC (adjusted HR 1.70, 95% CI 1.17–2.45) were shown to have a higher risk of insomnia symptoms after matched (Table 4).
Table 4.
Hazard ratios for insomnia symptoms for variables included as covariates a
| Propensity-Score Matched Patientsb | Full-cohort patients | |||
|---|---|---|---|---|
| Variables | Crude HR (95% CI) |
Adjusted HR (95% CI) |
Crude HR(95% CI) |
Adjusted HR (95% CI) |
| Sex | ||||
| Women | Ref. | Ref. | Ref. | Ref. |
| Men | 1.00 (0.79-1.27) | 0.86 (0.65-1.12) | 0.89 (0.72-1.09) | 0.83 (0.66-1.03) |
| Age group | ||||
| 18-63 | Ref. | Ref | Ref. | Ref. |
| ≥64 | 1.10 (0.87-1.38) | 1.30 (1.01-1.68)* | 0.99 (0.82-1.21) | 1.13 (0.91-1.41) |
| ECOG PS score | ||||
| 1-2 | Ref. | Ref. | Ref. | Ref. |
| 3-4 | 1.36 (1.06-1.74)* | 1.58 (1.19-2.09)** | 1.32 (1.08-1.62)** | 1.53 (1.23-1.90)*** |
| Pathological diagnosis | ||||
| NSCLC | Ref. | Ref. | Ref. | Ref. |
| SCLC | 1.74 (1.26-2.41)*** | 1.70 (1.17-2.45)** | 1.27 (0.95-1.69)* | 1.35 (0.98-1.85) |
| Therapy | ||||
| Chemotherapy | 1.22 (0.95-1.57) | 1.25 (0.93-1.68) | 1.11 (0.89-1.38) | 1.17 (0.93-1.48) |
| Radiotherapy | 1.27 (0.97-1.66) | 1.17 (0.86-1.59) | 1.25 (0.99-1.57) | 1.23 (0.97-1.56) |
| Immunotherapy | 0.87 (0.65-1.16) | 1.00 (0.74-1.36) | 0.76 (0.61-0.97)* | 0.95 (0.74-1.20) |
| Target | 0.93 (0.74-1.17) | 1.03 (0.78-1.36) | 0.92 (0.76-1.12)* | 0.99 (0.80-1.22) |
| Surgery | 1.03 (0.68-1.55) | 1.12 (0.72-1.76) | 0.97 (0.69-1.36) | 1.01 (0.71-1.42) |
| Opioids | 1.40 (1.11-1.78)** | 1.39 (1.08-1.79)* | 1.33 (1.09-1.63)** | 1.35 (1.09-1.66)** |
| Chinese herbal medincine | 1.33 (0.84–2.10) | 1.31 (0.84-2.06) | 1.11 (0.79-1.55) | 1.25 (0.88-1.76) |
| Comorbidities | ||||
| Infection | 0.90 (0.46-1.76) | 0.90 (0.43-1.85) | 1.23 (0.82-1.84) | 1.03 (0.68-1.57) |
| Hypertension | 1.03 (0.80-1.33) | 1.03 (0.78-1.37) | 1.02 (0.82-1.26) | 0.97 (0.77-1.23) |
| Diabetes | 0.89 (0.59-1.35) | 0.89 (0.58-1.38) | 0.98 (0.71-1.33) | 0.99 (0.72-1.37) |
| Mental disorder | 3.54 (1.75-7.14) | 7.59 (4.40-13.12)*** | 3.45 (1.29-9.26)* | 7.28 (2.61-20.33)*** |
HR, hazard ratios; CI, confidence interval. *P < 0.05, **P < 0.01, and ***P < 0.001
aAdjust hazard ratios adjusted for acupuncture, sex, age, ECOG PS score, pathological diagnosis, chemotherapy, radiotherapy, target therapy, immunotherapy, surgery, opioids, Chinese herbal medicine, infection, hypertension, diabetes, and mental disorder
bAfter PSM, the univariable and multivariable Cox regression with robust standard errors were conducted
E-value analyses
An unmeasured confounder could fully account for the association of acupuncture with the occurrence of insomnia in survivors of lung cancer if it were associated with the exposure and outcome by an HR of 3.87 (lower confidence limit, 3.22).
Analysis of acupoints
We identified the top 10 acupoints with the highest non-insomnia rates and statistically significant differences using a chi-square test (Fig. 5A). Sanyinjiao (SP6) had the highest non-insomnia rate at 71.63% (P = 1.17e−03), and all listed acupoints had P-values < 0.05, indicating a significant association with non-insomnia.
Fig. 5.
Relationship between acupoints and non-insomnia outcomes. A Top 10 acupoints with the highest rates of non-insomnia and significant differences. B The top 15 acupoints by feature importance from the random forest analysis. C Heatmap of common acupoint combinations with non-insomnia outcomes. D Top 10 a priori association rules (non-insomnia related)
The top 15 acupoints by feature importance are displayed in Fig. 5B. Quchi (LI11) and Zusanli (ST36) had the highest importance scores at 0.0427 and 0.0422, respectively. Notably, Quchi (LI11), Sanyinjiao (SP6), and Fengchi (GB20) were identified in both analyses, suggesting a potential link to insomnia prevention.
We also analyzed common acupoint combinations. For the top 10 combinations, we calculated the proportion of individuals without insomnia (Fig. 5C). The most frequent combination was a combination of Sanyinjiao (SP6), Zhongwan (CV12), Guanyuan (CV4), etc., with 48 cases and 58.3% experiencing no insomnia. The combinations with the highest proportion of no insomnia were Fenglong (ST40), Lieque (LU7), Sishencong (EX-HN1), Zhi gou (TE6), Liangqiu (ST34), Baihui (GV20), Feishu (BL13), Zusanli (ST36), Yinlingquan (SP9) (13 cases, 100% with no insomnia), the same combination minus Liangqiu (ST34) and Yinlingquan (SP9) (6 cases, 100% with no insomnia), Zusanli (ST36) alone (17 cases, 100% with no insomnia), and Yongquan (KI1) alone (6 cases, 100% with no insomnia). The Apriori algorithm was used to analyze the correlation between acupoints and the no-insomnia state. The top 10 association rules involving “no insomnia” were identified. Results showed several acupoint combinations were closely linked to the no-insomnia state. The Zusanli (ST36), Lieque (LU7), and Baihui (GV20) combination had the highest confidence of 95.45%, indicating a high probability of no insomnia when this combination appeared. The Jiuwei (CV15) acupoint, alone or in combination, had a lift of 1.12 to 2.37, showing a significant role in promoting the no-insomnia state (Fig. 5D).
In summary, acupoints like Sanyinjiao (SP6), Quchi (LI11), Zusanli (ST36), and Jiuwei (CV15) show a close link to the non-insomnia state. Combinations such as “Zusanli (ST36), Lieque (LU7), Baihui (GV20)” and “Jiuwei (CV15), Lieque (LU7)” have high confidence and lift in association rules, promoting the non-insomnia state. These findings are valued for insomnia prevention.
Discussion
In this retrospective cohort study of 1295 lung cancer survivors, receiving acupuncture was associated with a 69% lower risk of insomnia symptoms during a mean of 218.6 days of follow-up. The results before and after propensity score matching, along with the K-M survival curve, demonstrated consistent trends that together reinforce the robustness of the research conclusions.
Our study further confirms the preventive effect of acupuncture on insomnia symptoms, which was consistent with a previous trial [21]. While they evaluated preoperative electroacupuncture in breast cancer patients to improve postoperative anxiety and sleep quality, we used real-world data to retrospectively explore the relationship between acupuncture and insomnia symptoms incidence in lung cancer patients. In our study, the main purposes of acupuncture were cancer pain (27.5%), cancer-related symptoms (systemic) (fatigue, hot flashes, and weight loss, 27.1%), and respiratory symptoms (20.5%). It is suggested that cancer pain, fatigue, nausea, anxiety, and depression are positively associated with insomnia in lung cancer survivors [3, 4]. Acupuncture has been proven as a feasible and effective approach to alleviating these relevant symptoms [33–35]. Thus, acupuncture may reduce the risk of insomnia symptoms by improving these symptoms. In addition, cancer and antineoplastic treatment–related inflammation, which is known to disturb sleep [36, 37], can be reduced by acupuncture treatment [38].
There is currently no consensus regarding the optimal dosage or frequency of acupuncture for the prevention of cancer-related insomnia. Most interventional studies on cancer-related insomnia also suffer from insufficient follow-up periods. A systematic review conducted in 2022, which included 22 trials, revealed that one-third of these studies evaluated efficacy after only 1 week [39]. Given that patients with lung cancer frequently experience repeated hospitalizations, we were able to obtain extended follow-up data, with an average of 282 days in the acupuncture group. An increased number of sessions was associated with improved prevention outcomes; specifically, more than 16 sessions resulted in an approximate 85% reduction in the incidence of insomnia symptoms. Daily sessions proved to be more effective. A dose-response meta-analysis of primary insomnia found that three or more sessions weekly for 3–4 weeks (at least 12 sessions) leads to the best results [40].The situation may be different in lung cancer patients. Lung cancer patients have a higher risk of insomnia than the general population [5–7]. About 70% of patients received chemotherapy in our study, and chemotherapy-induced neutropenia generally resolves within 14–24 days [41]. Sixteen daily sessions might offer continuous regenerative stimulation, enhancing the prophylactic effect. From the perspective of Traditional Chinese Medicine (TCM), lung cancer and its treatments are believed to deplete Qi and blood, thereby disturbing the heart-mind and leading to insomnia. Acupuncture is thought to bolster this Qi, thereby restoring energy and blood flow.
Our findings indicate that incorporating the acupoints Sanyinjiao (SP6), Quchi (LI11), Zusanli (ST36), or Jiuwei (CV15) into a treatment protocol is associated with a reduced incidence of insomnia. Moreover, the combinations “Zusanli (ST36)–Lieque (LU7)–Baihui (GV20)” and “Jiuwei (CV15)–Lieque (LU7)” appear to confer additional preventive benefits against insomnia. From the perspective of TCM, these acupoints can be grouped into three functional categories. The first category is the core acupoints for insomnia, like Sanyinjiao (SP6) and Baihui (GV20). Research shows effective combinations often feature Shenmen (HT7), Baihui (GV20), or Sanyinjiao(SP6) [42], supporting our findings. Sanyinjiao (SP6), at the intersection of the three foot-yin meridians, regulates the liver, spleen, and kidney, alleviating insomnia by nourishing yin and calming the heart-spirit. Baihui (GV20), at the vertex on the Governing Vessel, harmonizes the mind by lifting yang and settling it, helping insomnia by returning floating yang. However, Shenmen (HT7), known for treating insomnia, ranked low in importance in our study (12 of 15). This may be because we were observing acupuncture given for other cancer-related conditions, not for insomnia, leading to infrequent use of HT7 and an inability to fully assess its preventive potential. The second group of acupoints is characterized by their primary indication for respiratory symptoms. Our cohort consisted exclusively of lung cancer patients who sought acupuncture while still free of insomnia. Their acupuncture prescriptions commonly incorporated Quchi (LI11), Lieque (LU7), and Jiuwei (CV15), points frequently used to alleviate respiratory condition [43–45]. Quchi (LI11), a key point on the large intestine (LI) meridian, treats rebellious qi and heat, aiding lung cancer patients by clearing phlegm-heat and improving breathing and sleep quality. Lieque (LU7), on the lung meridian (LU), connects to the conception vessel (CV), balancing lung-qi and aiding sleep by directing defensive yang into the yin-qiao vessel. Jiuwei (CV15), located on the CV, alleviates chest tightness and calms the mind, crucial for lung cancer patients post-treatment to manage phlegm-dampness and stomach-qi issues affecting sleep. A review has shown that acupoints along the LI and LU meridians not only treat substance use disorders but also improve patients’ insomnia, supporting our observations [46]. The third cluster targets gastrointestinal harmony and spleen invigoration through the stomach and spleen meridians, using points like Zusanli (ST36), Tianshu (ST25), Shuidao (ST29), Fenglong (ST40), Sanyinjiao (SP6), and Daheng (SP15). Zusanli (ST36) is the lower he-sea point of the stomach meridian, often used for gastrointestinal issues like nausea and discomfort, particularly in cancer patients. These issues can negatively affect sleep quality, aligning with the Chinese medicine belief that “stomach disharmony leads to impaired sleep.” Points ST25, ST29, and ST40 on the stomach channel aid in unblocking the bowels, resolving phlegm, and lowering rebellious qi, thereby supporting ST36 in promoting downward qi movement to enhance sleep indirectly. SP6 and SP15 on the spleen channel strengthen the spleen, reduce dampness, and regulate qi, working with the stomach points to achieve “spleen–stomach harmony.” Combined use of spleen and stomach meridian points improved insomnia while treating IBS, Parkinson’s disease, and cystitis46. Regardless of the primary complaint about respiratory, gastrointestinal, or other cancer-related symptoms, targeting key points on the spleen, stomach, lung, conception, and governing vessels simultaneously lowers insomnia incidence by restoring visceral harmony and balancing qi-blood. Future RCTs could explore these acupoints for insomnia prevention.
Acupuncture may alleviate insomnia by calming the HPA axis and adjusting sleep-related neurotransmitters. It could reduce hypothalamic CRH and pituitary ACTH, lowering cortisol [47], while increasing serotonin, GABA, and melatonin to enhance sleep quality [48]. The effects are specific to acupoints, with Baihui (GV20) regulating the HPA response and enhancing hippocampal 5-HT/5-HT1AR [49], the function of GABAergic neurons [50], and cerebral oxygenation [51]. The combination of GV20, Shenmen (HT7), and Sanyinjiao (SP6) further promotes recovery by activating cAMP/CREB and BDNF/TrkB signaling pathways for neural repair [52]. Additionally, electrically stimulating ST36 (Zusanli) and LI11 (Quchi) engages the PI3K/Akt pathway to reduce neuronal apoptosis [53]. Collectively, these mechanisms may explain the preventative effect of acupuncture on insomnia observed in our study.
Our study has several strengths. We used a comprehensive dataset from a large public tertiary hospital with a broad radiation range; thus, the sample in this study is reasonably representative. We were able to access a large variety of data through the database. The data are unique since the electronic medical record database is private. Also, our study incorporated several advanced methodologies to strengthen causal inference in observational research. We employed DAGs from causal inference methods to assist in the selection of covariates and used the E-value to quantify the potential impact of unmeasured confounding on our study results. According to the E-value analysis, the hazard ratio (HR = 0.31) between acupuncture and insomnia could theoretically be attributed to an unmeasured confounder exhibiting HR = 3.87 for both exposure and outcome, which is larger than all the risk factors identified in this study (HR < 3.87 for all) except mental disorder (HR = 7.59). In addition, few studies focus on preventative effect; our study shows that acupuncture can reduce insomnia in lung cancer patients, suggesting broader preventive benefits.
Two main limitations warrant consideration. First, the data were collected from the inpatient department and do not include information on outpatient visits or referrals to other hospitals. Second, we did not register the study protocol in advance.
Notwithstanding these limitations, this study has several implications for future research and clinical care. Traditional Chinese medical literature, Yellow Emperor’s Inner Canon, states that the highest level of medical practice is to prevent disease rather than to treat it after it has occurred. It also elaborates on the holistic regulatory effects of acupuncture on the human body and mind from the perspective of TCM theory. Patients may experience improvements in emotional well-being and other aspects due to the overall regulatory benefits [54]. Acupuncture holds promise in many chronic, recurrent, and refractory symptomatic diseases that are difficult to cure [55–57]. Future research might focus more on the preventive aspects to alleviate the disease burden on both patients and society.
Conclusions
Our results suggest that acupuncture may prevent the incidence of insomnia symptoms in patients with lung cancer. Daily sessions and a cumulative total of 16 or more treatments may enhance the benefit.
Supplementary Information
Below is the link to the electronic supplementary material.
(PDF 778 KB)
(PDF 517 KB)
(DOCX 304 KB)
Acknowledgements
We thank Dr. Xiaobing Yang for his assistance in applying for the data and we also thank the research assistants: Jie Xu(data collection and input), Guolin Huang (data collection and input), Jiahao Mo (data collection and input).
Abbreviations
- ECOG PS score
Eastern Cooperative Oncology Group Performance Status score
- NSCLC
Non-small cell lung cancer
- SCLC
Small cell lung cancer
- EMRD
Electronic medical records database
- ICD-10
International Classification of Disease, Tenth Revision
- SMD
Standardized mean difference
- HR
Hazard ratios
- CI
Confidence interval
Author contributions
Ruifang Yu designed the study protocol, analyzed the data, and drafted and revised the manuscript. Xingfeng Guo contributed to the study's methodology, provided supervision, and reviewed the manuscript. Yue Chen and Canyang Zhang performed data collection, extraction and assisted in manuscript writing. Yanjuan Zhu and Xiaoshu Chai contributed to updated data extraction and analysis, reorganized methods and discussion and refined figures and tables. Xiaoshu Chai contributed to comprehensive language editing, results revisions and funding support. Yingqi Wang was responsible for project administration, data curation, and validation, provided supervision, and also contributed to drafting and revising the manuscript. All the authors read and approved the final version of the manuscript and accepted it for publication.
Funding
The authors declare that financial support was received for the publication of this article. This study was supported by Fuding for Young Top Talents of Guangzhou University of Chinese Medicines (A1-2601-25-415-111Z68), Funding for Young Top Talents of Guangdong Provincial Hospital of Traditional Chinese Medicine, International Collaborative Research Program of Guangdong Provincial Hospital of Traditional Chinese Medicine (YN2024HL05), the State Key Laboratory of Dampness Syndrome of Chinese Medicine, The Second Affiliated Hospital of Guangzhou University of Chinese Medicine (No. SZ2021ZZ0401), and the Science and Technology Planning Project of Guangdong Province (No. 2023B1212060063).
Data availability
The data that support the findings of this study are available from the study groups, but restrictions apply to the avail ability of these data, which were used under license for the current study; therefore, the data are not publicly available. However, data are available from the corresponding authors upon reasonable request and with permission from the study groups.
Declarations
Ethics approval
This study was approved by the Research Ethics Committee of Guangdong Provincial Hospital of Chinese Medicine (YE2021-315-01) in accordance with the Declaration of Helsinki and 2016 International ethical guidelines for health-related research involving human.
Consent to participate
Informed consent was waived for the total anonymity of all research data in this study.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Xiaoshu Chai, Email: chaixiaoshu@126.com.
Yingqi Wang, Email: wyq950101@126.com.
References
- 1.Schutte-Rodin S, Broch L, Buysse D, Dorsey C, Sateia M (2008) Clinical guideline for the evaluation and management of chronic insomnia in adults. J Clin Sleep Med: JCSM : Off Publ Am Acad Sleep Med 4(5):487–504 [Google Scholar]
- 2.Nishiura M, Tamura A, Nagai H, Matsushima E (2015) Assessment of sleep disturbance in lung cancer patients: relationship between sleep disturbance and pain, fatigue, quality of life, and psychological distress. Palliat Support Care 13(3):575–581. 10.1017/s1478951513001119 [DOI] [PubMed] [Google Scholar]
- 3.Papadopoulos D, Kiagia M, Charpidou A, Gkiozos I, Syrigos K (2019) Psychological correlates of sleep quality in lung cancer patients under chemotherapy: a single-center cross-sectional study. Psychooncology 28(9):1879–1886. 10.1002/pon.5167 [DOI] [PubMed] [Google Scholar]
- 4.Mercadante S, Valle A, Cartoni C, Pizzuto M (2021) Insomnia in patients with advanced lung cancer admitted to palliative care services. Int J Clin Pract 75(10):e14521. 10.1111/ijcp.14521
- 5.Savard J, Villa J, Ivers H, Simard S, Morin CM (2009) Prevalence, natural course, and risk factors of insomnia comorbid with cancer over a 2-month period. J Clin Oncol 27(31):5233–5239. 10.1200/jco.2008.21.6333 [DOI] [PubMed] [Google Scholar]
- 6.Og P, Ja R, Km M et al (2010) Prevalence, demographics, and psychological associations of sleep disruption in patients with cancer: University of Rochester Cancer Center-Community Clinical Oncology Program. J Clin Oncol 28(2):292–298. 10.1200/jco.2009.22.5011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Mm O (2002) Epidemiology of insomnia: what we know and what we still need to learn. Sleep Med Rev 6(2):97–111. 10.1053/smrv.2002.0186 [DOI] [PubMed] [Google Scholar]
- 8.Lis CG, Gupta D, Grutsch JF (2008) The relationship between insomnia and patient satisfaction with quality of life in cancer. Support Care Cancer 16(3):261–266. 10.1007/s00520-007-0314-z [DOI] [PubMed] [Google Scholar]
- 9.Irwin MR (2015) Why sleep is important for health: a psychoneuroimmunology perspective. Annu Rev Psychol 66:143–172. 10.1146/annurev-psych-010213-115205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ozminkowski RJ, Wang S, Walsh JK (2007) The direct and indirect costs of untreated insomnia in adults in the United States. Sleep 30(3):263–273. 10.1093/sleep/30.3.263 [DOI] [PubMed] [Google Scholar]
- 11.Sephton SE, Lush E, Dedert EA et al (2013) Diurnal cortisol rhythm as a predictor of lung cancer survival. Brain Behav Immun 30(Suppl):S163–S170. 10.1016/j.bbi.2012.07.019 [DOI] [PubMed] [Google Scholar]
- 12.Ruel S, Ivers H, Savard MH et al (2020) Insomnia, immunity, and infections in cancer patients: results from a longitudinal study. Health Psychol 39(5):358–369. 10.1037/hea0000811 [DOI] [PubMed] [Google Scholar]
- 13.Rosenberg RP, Benca R, Doghramji P, Roth TA (2023) Update on managing insomnia in primary care: insights from an expert consensus group. Prim Care Companion CNS Disord 25(1)10.4088/PCC.22nr03385
- 14.Howell D, Oliver TK, Keller-Olaman S et al (2013) A pan-Canadian practice guideline: prevention, screening, assessment, and treatment of sleep disturbances in adults with cancer. Support Care Cancer 21(10):2695–2706. 10.1007/s00520-013-1823-6 [DOI] [PubMed] [Google Scholar]
- 15.Marion LP, Ivers H, Savard J (2021) Feasibility of a preventive intervention for insomnia in women with breast cancer receiving chemotherapy. Behav Sleep Med 19(1):70–82. 10.1080/15402002.2019.1707203 [DOI] [PubMed] [Google Scholar]
- 16.Zhang XW, Hou WB, Pu FL et al (2022) Acupuncture for cancer-related conditions: an overview of systematic reviews. Phytomedicine 106:154430. 10.1016/j.phymed.2022.154430 [DOI] [PubMed] [Google Scholar]
- 17.Zhang SQ, Chen HB, Liu J, Dai WJ, Lu QQ, Li JC (2021) Research status and prospects of acupuncture for prevention and treatment of chemo- and radiotherapy-induced salivary gland dysfunction in head and neck cancer. Anatomical Record (Hoboken, NJ : 2007) 304(11):2381–2396. 10.1002/ar.24784 [Google Scholar]
- 18.Rithirangsriroj K, Manchana T, Akkayagorn L (2015) Efficacy of acupuncture in prevention of delayed chemotherapy induced nausea and vomiting in gynecologic cancer patients. Gynecol Oncol 136(1):82–86. 10.1016/j.ygyno.2014.10.025 [DOI] [PubMed] [Google Scholar]
- 19.Zhang Q, Li Y, Yin C et al (2022) Electro-acupuncture pretreatment ameliorates anesthesia and surgery-induced cognitive dysfunction via inhibiting mitochondrial injury and neuroapoptosis in aged rats. Neurochem Res 47(6):1751–1764. 10.1007/s11064-022-03567-3 [DOI] [PubMed] [Google Scholar]
- 20.Gao J, Zhao Y, Wang Y et al (2015) Anti-arrhythmic effect of acupuncture pretreatment in the rats subjected to simulative global ischemia and reperfusion--involvement of intracellular Ca2+ and connexin 43. BMC Complement Altern Med 15:5. 10.1186/s12906-015-0521-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Tong QY, Liu R, Gao Y, Zhang K, Ma W, Shen WD (2022) Effect of electroacupuncture based on ERAS for preoperative anxiety in breast cancer surgery: a single-center, randomized, controlled trial. Clin Breast Cancer 22(7):724–736. 10.1016/j.clbc.2022.04.010 [DOI] [PubMed] [Google Scholar]
- 22.Mercadante S, Girelli D, Casuccio A (2004) Sleep disorders in advanced cancer patients: prevalence and factors associated. Support Care Cancer 12(5):355–359. 10.1007/s00520-004-0623-4 [DOI] [PubMed] [Google Scholar]
- 23.Feeney T, Hartwig FP, Davies NM (2025) How to use directed acyclic graphs: guide for clinical researchers. BMJ 388:e078226. 10.1136/bmj-2023-078226 [DOI] [PubMed] [Google Scholar]
- 24.Oken MM, Creech RH, Tormey DC et al (1982) Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 5(6):649–655 [PubMed] [Google Scholar]
- 25.Yang M, Liou KT, Garland SN et al (2021) Acupuncture versus cognitive behavioral therapy for pain among cancer survivors with insomnia: an exploratory analysis of a randomized clinical trial. NPJ Breast Cancer. 10.1038/s41523-021-00355-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pearl J, Glymour M, Jewell NP (2016) Causal inference in statistics: a primer. Wiley, pp 58–61
- 27.Jakobsen G, Engstrøm M, Fayers P et al (2019) Sleep quality with WHO step III opioid use for cancer pain. BMJ Support Palliat Care 9(3):307–315. 10.1136/bmjspcare-2017-001399 [DOI] [PubMed] [Google Scholar]
- 28.Austin PC (2011) An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivar Behav Res 46(3):399–424. 10.1080/00273171.2011.568786 [Google Scholar]
- 29.Austin PC (2009) Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 28(25):3083–3107. 10.1002/sim.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Yadav K, Lewis RJ (2021) Immortal time bias in observational studies. JAMA 325(7):686–687. 10.1001/jama.2020.9151 [DOI] [PubMed] [Google Scholar]
- 31.Heide-Jørgensen U, Adelborg K, Kahlert J, Sørensen HT, Pedersen L (2018) Sampling strategies for selecting general population comparison cohorts. Clin Epidemiol 10:1325–1337. 10.2147/clep.S164456 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.VanderWeele TJ, Ding P (2017) Sensitivity analysis in observational research: introducing the E-value. Ann Intern Med 167(4):268–274. 10.7326/m16-2607 [DOI] [PubMed] [Google Scholar]
- 33.Molassiotis A, Bardy J, Finnegan-John J et al (2012) Acupuncture for cancer-related fatigue in patients with breast cancer: a pragmatic randomized controlled trial. J Clin Oncol 30(36):4470–4476. 10.1200/jco.2012.41.6222 [DOI] [PubMed] [Google Scholar]
- 34.Mao JJ, Farrar JT, Bruner D et al (2014) Electroacupuncture for fatigue, sleep, and psychological distress in breast cancer patients with aromatase inhibitor-related arthralgia: a randomized trial. Cancer 120(23):3744–3751. 10.1002/cncr.28917 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.He Y, Guo X, May BH et al (2020) Clinical evidence for association of acupuncture and acupressure with improved cancer pain: a systematic review and meta-analysis. JAMA Oncol 6(2):271–278. 10.1001/jamaoncol.2019.5233 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Vaughn CM, Vaughn BV (2025) Sleep and cancer. Cancers (Basel) 17(6). 10.3390/cancers17060911
- 37.Liu L, Mills PJ, Rissling M et al (2012) Fatigue and sleep quality are associated with changes in inflammatory markers in breast cancer patients undergoing chemotherapy. Brain Behav Immun 26(5):706–713. 10.1016/j.bbi.2012.02.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zhao H, Liu C, Ye H, Shi K, Yao Y, Wang M (2024) Advancements in the physiopathological study of acupuncture treatment for insomnia: a review. Medicine (Baltimore) 103(26):e38476. 10.1097/md.0000000000038476 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Zhang J, Zhang Z, Huang S et al (2022) Acupuncture for cancer-related insomnia: a systematic review and meta-analysis. Phytomedicine 102:154160. 10.1016/j.phymed.2022.154160 [DOI] [PubMed] [Google Scholar]
- 40.Zhang X, Wang Y, Liu C et al (2025) The dose-effect relationship between acupuncture and its effect on primary insomnia: a systematic review and meta-analysis. Front Psychiatry 16:1501321. 10.3389/fpsyt.2025.1501321 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Chinese Society of Clinical Oncology Guideline Working Committee (2021) CSCO guidelines for standardized management of tumor chemoradiotherapy-related neutropenia (2021) [in Chinese]. J Clin Oncol (China) 26(7):638–648. 10.3969/j.issn.1009-0460.2021.07.011. (in Chinese)
- 42.Shergis JL, Ni X, Jackson ML et al (2016) A systematic review of acupuncture for sleep quality in people with insomnia. Complement Ther Med 26:11–20. 10.1016/j.ctim.2016.02.007 [DOI] [PubMed] [Google Scholar]
- 43.Jiang TX, Wang BG, Yang QQ et al (2024) Observation of clinical efficacy of acupuncture combined with bloodletting therapy and Qingwen Xiere decoction in mild to moderate COVID-19 patients. Zhen Ci Yan Jiu 49(8):845–850. 10.13702/j.1000-0607.20230316 [DOI] [PubMed] [Google Scholar]
- 44.Wang P, Huang M, Xu H, Liu Y, Jia S (2024) Effect of transcutaneous electrical acupoint stimulation on postoperative cough in lung cancer patients undergoing video-assisted thoracoscopic surgery. Zhongguo Zhen Jiu 44(5):521–525. 10.13703/j.0255-2930.20231027-k0005 [DOI] [PubMed] [Google Scholar]
- 45.Teng D (2003) Clinical application of the jiuwei point. J Tradit Chin Med 23(1):45–46 [PubMed] [Google Scholar]
- 46.Huang W, Kutner N, Bliwise DL (2009) A systematic review of the effects of acupuncture in treating insomnia. Sleep Med Rev 13(1):73–104. 10.1016/j.smrv.2008.04.002 [DOI] [PubMed] [Google Scholar]
- 47.Han X, Gao Y, Yin X et al (2021) The mechanism of electroacupuncture for depression on basic research: a systematic review. Chin Med 16(1):10. 10.1186/s13020-020-00421-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Guo J, Guo J, Rao X et al (2024) Exploring the pathogenesis of insomnia and acupuncture intervention strategies based on the microbiota-gut-brain axis. Front Microbiol 15:1456848. 10.3389/fmicb.2024.1456848 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Le JJ, Yi T, Qi L, Li J, Shao L, Dong JC (2016) Electroacupuncture regulate hypothalamic-pituitary-adrenal axis and enhance hippocampal serotonin system in a rat model of depression. Neurosci Lett 615:66–71. 10.1016/j.neulet.2016.01.004 [DOI] [PubMed] [Google Scholar]
- 50.Zhang S, Li G, Xu X, Chang M, Zhang C, Sun F (2011) Acupuncture to point Baihui prevents ischemia-induced functional impairment of cortical GABAergic neurons. J Neurol Sci 307(1–2):139–143. 10.1016/j.jns.2011.04.021 [DOI] [PubMed] [Google Scholar]
- 51.Yujuan YE, Yuting W, Jing J, Xingke Y (2023) Efficacy of needling Baihui (GV20), Neiguan (PC6), Shenmen (HT7) and Taichong (LR3) on cerebral cortical blood oxygen level in rats with insomnia. J Tradit Chin Med 43(3):523–532. 10.19852/j.cnki.jtcm.20230404.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Lina Q, Yinan S, Lianhong T, Yanshu J, Yongsheng Y (2023) Efficacy of electroacupuncture stimulating Shenmen (HT7), Baihui (GV20), Sanyinjiao (SP6) on spatial learning and memory deficits in rats with insomnia induced by para-chlorophenylalanine: a single acupoint combined acupoints. J Tradit Chin Med 43(4):704–714. 10.19852/j.cnki.jtcm.20230308.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53.Xue X, You Y, Tao J et al (2014) Electro-acupuncture at points of Zusanli and Quchi exerts anti-apoptotic effect through the modulation of PI3K/Akt signaling pathway. Neurosci Lett 558:14–19. 10.1016/j.neulet.2013.10.029 [DOI] [PubMed] [Google Scholar]
- 54.Yang JW, Li QQ, Li F, Fu QN, Zeng XH, Liu CZ (2014) The holistic effects of acupuncture treatment. Evid Based Complement Alternat Med 2014:739708. 10.1155/2014/739708 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Gamus D, Shoenfeld Y (2025) Acupuncture therapy in autoimmune diseases: a narrative review. Autoimmun Rev. 10.1016/j.autrev.2024.103709 [DOI] [PubMed] [Google Scholar]
- 56.Li G, Liu J, Yang G et al (2025) Efficacy of acupuncture as adjunctive therapy for patients with acute exacerbation of chronic obstructive pulmonary disease: a systematic review and meta-analysis. Front Med (Lausanne) 12:1513888. 10.3389/fmed.2025.1513888 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.Minchom A, Punwani R, Filshie J et al (2016) A randomised study comparing the effectiveness of acupuncture or morphine versus the combination for the relief of dyspnoea in patients with advanced non-small cell lung cancer and mesothelioma. Eur J Cancer 61:102–110. 10.1016/j.ejca.2016.03.078 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
(PDF 778 KB)
(PDF 517 KB)
(DOCX 304 KB)
Data Availability Statement
The data that support the findings of this study are available from the study groups, but restrictions apply to the avail ability of these data, which were used under license for the current study; therefore, the data are not publicly available. However, data are available from the corresponding authors upon reasonable request and with permission from the study groups.