The study subjects consisted of patients who underwent elective total knee arthroplasty (TKA) at Gansu Provincial Hospital of TCM from November 2023 to January 2024. The Pittsburgh Sleep Quality Index Questionnaire (PSQI) was used to assess the overall sleep status of patients over the past month. All patients were divided into normal sleep group (PSQI ≤ 5) and poor sleep group (PSQI score > 5). Postoperative pain was assessed using a numeric rating scale (NRS), and hydromorphone (HM) consumption, number of rescue analgesia, and quality of recovery QoR-15 scores were recorded 24 hours after surgery. Spearman correlation analysis was used to detect the correlation between each index and preoperative PSQI score. A logistic regression model was used to analyze the relationship between PSQI score, serum BDNF and CB1R concentrations, and the occurrence of moderate to severe pain 24 hours after surgery. Generalized linear regression models combined with restricted cubic spline models were used to analyze the relationship and dose-response relationship between PSQI, BDNF, and CB1R and QoR-15 score and HM consumption at 24 hours after surgery after adjusting for confounders. Serum BDNF and CB1R mediation effects were analyzed by Process plug-in.

This study was approved by the Medical Ethics Committee of Gansu Provincial Hospital of TCM (2023-044-01) and was performed in compliance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants at the time of surgery. Patients who were scheduled to undergo primary unilateral TKA at Gansu Provincial Hospital of TCM from November 2023 to January 2024 were selected as the study subjects. Inclusion criteria: (1) aged ≥ 18 years; (2) patients scheduled for elective unilateral TKA under spinal anesthesia; (3) American Society of Anesthesiologists (ASA) classification I-III ; (4) voluntary participation and signed informed consent. Patients were excluded based on the following criteria: (1) unwillingness to participate in this study; (2) concomitant serious underlying diseases that could affect postoperative rehabilitation; (3) revision TKA ; (4) severe neuropsychiatric disorders, especially depression, which could affect sleep quality. Exclusion criteria: (1) Patients who failed spinal puncture; (2) Patients who voluntarily withdrew from the study halfway; (3) Patients with poor compliance who took other analgesic drugs by themselves to interfere with the study results; (4) Patients who were lost to follow-up and could not collect outcome measures.

Data collection Perioperative visits and basic information collection were performed by a uniformly trained anesthesiologist before surgery in all study subjects. The demographic characteristics of the participants, such as age, sex, BMI, comorbidities, and sleep quality scores were recorded.

The following parameters were documented after surgery: postoperative pain score, consumption of hydromorphone (HM) 24 hours after surgery, and the number of rescue analgesia and quality of recovery 24 hours after surgery.

Before anesthesia, 3 ml of venous blood was collected from the non-infusion side of the patient (on an empty stomach). The whole blood was collected into a blood collection tube containing EDTA sodium salt. After standing at room temperature for 30 minutes, the blood was centrifuged at 3000 rpm for 10 minutes. The supernatant was then placed in a cryopreservation box and stored at -80℃ until use.

Laboratory Testing Serum BDNF and CB1R concentrations were measured by enzyme-linked immunosorbent assay (ELISA).

Associated definitions Sleep disorders refer to various functional disturbances occurring within the sleep-wake cycle, including difficulties falling asleep, abnormally short or long sleep durations (hypersomnia), sleep-related breathing disorders, and parasomnias[27]. The Pittsburgh Sleep Quality Index (PSQI) is utilized to assess patients' sleep quality. The PSQI primarily evaluates a patient's sleep quality over the past month[28, 29]. This scale consists of 19 self-rated items and 5 other-rated items. Excluding the 19th self-rated item and the 5 other-rated items, the remaining 18 self-rated items are categorized into seven dimensions: sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbance, use of hypnotic medication, and daytime dysfunction. A 4-point Likert scale is applied, with each dimension scoring from 0 to 3. The cumulative score across all dimensions constitutes the total PSQI score, ranging from 0 to 21. Higher scores indicate poorer sleep quality. The PSQI demonstrates a sensitivity of 89.6% and a specificity of 86.5% in identifying patients with sleep disorders. A total score exceeding 5 suggests the presence of a sleep disorder[28]. Based on PSQI scores, patients are categorized into two groups: the normal sleep group (PSQI ≤ 5) and the poor sleep group (PSQI > 5).

The numerical rating scale (NRS) was used to evaluate the level of postoperative pain in patients. NRS scores were recorded at 2h, 6h, 12h, 24h, and 48h postoperatively, and the incidence of moderate to severe pain at corresponding time points was calculated. On this scale, 0 represents no pain, while 10 represents the worst imaginable pain. Based on the NRS scores, patients were classified into three pain categories: no pain (0), mild pain (1-3/10), and moderate to severe pain (≥4/10)[30].

Quality of recovery (QoR) is a crucial concept in perioperative patient management. Assessing QoR is significant for patient prognosis and economics, offering potential for clinical and research applications and enhancing perioperative patient satisfaction. The QoR-15 scale is employed to evaluate patients' recovery quality at 24 hours postoperatively. The QoR-15 score is a recommended outcome measure in studies focusing on postoperative patient comfort[26]. This scale comprises 15 questions assessing five dimensions of patients' postoperative health status[31], including physical comfort, physical independence, psychological support, emotions, and pain. Each item is scored from 0 to 10, with a total possible score of 150, where higher scores indicate better recovery.

Anesthesia Protocol All enrolled patients underwent a unified anesthesia protocol. After entering the operating room, patients received routine intravenous infusion oxygen inhalation and were connected to a multifunction electrocardiogram monitor. In the lateral position, a subarachnoid block was performed at the L3-4 intervertebral space using a spinal anesthesia needle. Upon successful puncture and withdrawal of clear cerebrospinal fluid (CSF), 2ml of 0.75% bupivacaine mixed with CSF to a total volume of 3ml was slowly injected, with 2ml administered. After completion, the spinal anesthesia needle was removed, and the level of anesthesia was controlled below T8. If the subarachnoid block failed, the anesthesia method was changed to general anesthesia, and the case was excluded from the study. Before the incision, blood was driven away from the surgical site using an Esmarch bandage, and an electric tourniquet was applied with a pressure of 300 mmHg. At the end of the surgery, the tourniquet was released slowly while the patient's vital signs and discomfort were observed. Patients were then transferred to the postanesthesia care unit (PACU) for observation for 30 minutes.

Postoperative analgesia protocol: All patients followed a standardized protocol for postoperative pain management. Oral etoricoxib 120 mg (Qilu Pharmaceutical Co., Ltd., specification: 60 mg/tablet, 2 times/day) was given for preemptive analgesia on the day before surgery. This analgesic regimen was repeated when patients resumed oral intake postoperatively. A "cocktail" therapy (compounding betamethasone injection 1 mL, epinephrine 0.5 mL, ropivacaine hydrochloride 150 mg, diluted with normal saline to 100 mL) was administered before suturing. All patients received intravenous patient-controlled analgesia (PCIA) postoperatively, consisting of hydromorphone (HM) 6 mg diluted with 0.9% NaCl solution to 100 mL, with an infusion rate of 2.0 mL/h, a PCIA bolus of 0.5 mL, and a lockout time of 15 minutes. The criterion for initiating PCIA therapy was an NRS score of ≥3. When the NRS score was ≥4[32], patients were instructed to press the PCIA pump for medication. If the effective PCIA bolus presses were ≥2 times within 30 minutes or the NRS score remained ≥4, rescue analgesia with 40 mg of intramuscular non-steroidal anti-inflammatory drug parecoxib sodium was administered.

Statistical Analyses A descriptive analysis of patients' basic characteristics was conducted based on PSQI scores. Categorical data were expressed as frequencies or percentages, and comparisons between groups were made using the Chi-square test. For continuous variables, the Kolmogorov-Smirnov test was used to assess normal distribution, and Levene's test was applied to check for homogeneity of variance. Normally distributed continuous variables were presented as mean ± standard deviation (± s) and compared using the t-test. Non-normally distributed continuous variables were reported as the median and interquartile range (IQR) and compared using the Mann-Whitney U test. Spearman's correlation analysis was employed to examine the association between various indicators and preoperative PSQI scores.

A multivariate logistic regression model was employed to examine the association between the Pittsburgh Sleep Quality Index (PSQI) score, serum brain-derived neurotrophic factor (BDNF) levels, cannabinoid receptor type 1 (CB1R) concentrations, and the incidence of moderate to severe pain within 24 hours following surgery, with the latter serving as the dependent variable. Model 1 did not adjust for confounding factors; Model 2 adjusted for age and gender; Model 3 added body mass index (BMI), operation time (<80 min, ≥80 min), education level (below high school, high school, and above), diabetes (yes, no), hypertension (yes, no), and dyslipidemia (yes, no) to Model 2.

The 25th, 50th, and 75th percentiles were selected as knots. A generalized linear regression model combined with a restricted cubic spline model was employed to analyze the relationship and dose-response relationship between PSQI, BDNF, CB1R, and postoperative 24-hour QoR-15 scores and HM consumption after adjusting for confounding factors. Additionally, the same variables involved in logistic regression were corrected. A Poverall <0.05 indicated the presence of a dose-response relationship between variables, while a Pnonlinear <0.05 suggested a nonlinear dose-response relationship between two variables.

Inclusion Criteria:

• aged ≥ 18 years; patients scheduled for elective unilateral TKA under spinal anesthesia; American Society of Anesthesiologists (ASA) classification I-III ; voluntary participation and signed informed consent

Exclusion Criteria:

• unwillingness to participate in this study; concomitant serious underlying diseases that could affect postoperative rehabilitation; revision TKA ; severe neuropsychiatric disorders, especially depression, which could affect sleep quality.

Patients who were scheduled to undergo primary unilateral TKA at Gansu Provincial Hospital of TCM from November 2023 to January 2024 were selected as the study subjects.