Effect of Propofol on Postoperative Sleep Quality in Patients Undergoing Elective Cesarean Section with Spinal Anesthesia:A Retrospective Cohort Study Using Propensity Score Matching at a Single Center

Introduction

Women undergoing cesarean section (CS) exhibit a significantly higher incidence of postpartum sleep disturbances disruption compared with vaginal delivery.^1,2 Postpartum sleep disturbances are characterized by reduced sleep efficiency and increased wakefulness.³ Sleep quality after CS is associated with various factors, including pregnancy-associated comorbidities, antenatal sleep quality, anesthesia or surgical duration, postpartum pain, post-cesarean recovery, neonatal care demands, lactation and feeding patterns, environmental stressors, and other maternal discomfort-inducing factors.^4–7 A prospective study tracking postoperative sleep trajectories demonstrated persistent sleep disturbances in most CS patients through 6 months postpartum.⁸ Within the 6 months postpartum, the prevalence of insomnia was 20.4%, and the prevalence of poor sleep quality was 67.8%, it can longitudinally predict more severe postpartum anxiety symptoms and depression.⁹ Importantly, the frequent co-occurrence of depressive symptoms, fatigue, and compromised sleep quality in postpartum women appears to collectively impair infant developmental outcomes.¹⁰ Moreover, sleep disturbances may serve as clinical indicators of postpartum psychosis risk.¹¹ However, the physiological complexity of the maternal-neonatal dyad and pharmacological restrictions during lactation have resulted in a paucity of effective interventions, particularly with pharmacologic interventions.

During our investigation into maternal sleep quality, we identified that a subset of women undergoing CS received intraoperative propofol administration by attending physicians based on real-time clinical assessment. The causes encompass multiple dimensions, including maternal voluntary requests, severe anxiety states, and profound fatigue. Propofol has been found to significantly alter sleep architecture.¹² In recent years, propofol has been established to improve postoperative sleep quality in non-obstetric surgical populations. For instance, studies have demonstrated that, compared to sevoflurane, propofol-based anesthesia reduces postoperative sleep disturbances and improves sleep efficiency in laparoscopic surgery patients,¹³ and has been shown to prolong total sleep time on the first postoperative night in elderly patients undergoing major abdominal surgery.¹⁴ Furthermore, propofol was associated with a greater improvement in sleep quality on the first post-procedure day compared to remimazolam for patients undergoing ambulatory painless gastroscopy.¹⁵

Therefore, we hypothesize that intraoperative administration of propofol following extraction of the infant during CS may confer favorable effects on maternal postpartum outcomes. Even so, conducting a randomized controlled study on the use of propofol during CS for pregnant women poses a challenge to medical ethics. This means that some pregnant women were randomly assigned to receive propofol, rather than based on medical necessity. Accordingly, conducting a preliminary analysis of the hypothesis utilizing real-world clinical data represents a judicious strategic approach.

This study aimed to evaluate whether pharmacological intervention of propofol following extraction of the infant influences sleep quality on postpartum day 30 among CS patients undergoing spinal anesthesia, with the goal of informing future study. The data were retrieved from the electronic medical record database of a tertiary Level III Grade A teaching hospital in China.

Materials and Methods

Ethics and Study Design

This study was approved by the Medical Ethics Committee of Tongde Hospital of Zhejiang Province (No: ZTD-2025S-068). As a retrospective study, the institutional ethics committee granted a waiver of written informed consent in accordance with institutional ethical policies. With approval from the ethics committee, verbal consent was obtained via standardized telephone interviews prior to administering follow-up scale assessments. The entire consent process, including the verbal explanation of the study and the participant’s agreement, was audio-recorded. These recordings are stored securely with the research data under a unique participant ID and are available for audit by the ethics committee upon request. Patients who declined participation were excluded from the study. The trial was registered prior to patient enrollment at the Chinese Clinical Trial Registry (https://www.chictr.org.cn/showproj.html?proj=262780, Registration number: ChiCTR2500097811, Registration date: 02/26/2025). The study adhered to the Declaration of Helsinki.

Inclusion and Exclusion Criteria

The inclusion criteria were patients aged over 18 yr with American Society of Anesthesiologists (ASA) physical status II–III undergoing elective CS with spinal anesthesia.

The exclusion criteria were patients who received other intravenous anesthetics except propofol following extraction of the infant, failure of spinal anesthesia, chronic life-threatening comorbidities (eg, chronic obstructive pulmonary disease, pulmonary hypertension, or a history of congestive heart failure), severe mental disorder, long-term use of hypnotics and sedatives, inability to communication. Patients who refused to participate were excluded.

This retrospective cohort study included the patients who underwent elective CS under spinal anesthesia in 2024. In clinical practice, a subset of patients undergoing cesarean section under spinal anesthesia may receive propofol after extraction of the infant due to specific indications including maternal request, acute anxiety, or preoperative sleep deprivation. The decision to administer propofol was made by attending anesthesiologists based on their clinical assessment. Based on actual anesthesia records, we divided the patients who were received propofol following extraction of the infant into the propofol group, and those who were not received intravenous anesthetics following extraction of the infant into the control group.

Data were retrospectively collected by research assistants using a standardized data collection form from two sources: electronic medical records system and telephone follow-ups of patients discharged after CS. The collected variables included age, BMI, gestational age, ASA classification, parity, smoking status, diabetes, hypertension, hypothyroxinemia during pregnancy, education level, employment status, prenatal exercise, infant sex, duration of anesthesia, postoperative complications, duration of hospital stays. The 30-day postoperative assessment was conducted via a home telephone interview to evaluate sleep quality and depressive symptoms using PSQI and the EPDS. We tried to be as objective as possible in the assessment, without insinuating the patient, but rather a review of the entire postoperative recovery. The primary outcome was postoperative sleep quality, and the secondary outcomes were depressive symptoms.

Measures

Quality of Postoperative Sleep

The postoperative sleep quality was assessed using Pittsburgh Sleep Quality Index (PSQI), which was initially described by Buysse in 1989.¹⁶ The PSQI has seven components including sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction. Score in each group ranged from 0 to 3 and total score ranging from 0 to 21 was obtained by summing the seven component scores. PSQI total score of more than 5 was generally considered to poor sleep quality, higher PSQI score indicated lower sleep quality. PSQI has good construct validity and reliability (Cronbach’s alpha 0.73) and has been used to assess sleep quality among childbearing women.¹⁷

Edinburgh Postnatal Depression Scale

The postoperative depressive symptoms were assessed using Edinburgh Postnatal Depression Scale (EPDS).¹⁸ Its 10 items were scored ranged from 0 to 3 with total scores ranging from 0 to 30. The higher the score indicated the more severe the symptoms. EPDS has good construct validity and reliability (Cronbach’s alpha 0.86) among childbearing women.¹⁸

Statistical Analysis

Statistical analyses were performed using IBM SPSS Statistics for mac (version 27). Continuous data were tested for normality using the Shapiro–Wilk test. Normally distributed data are presented as the means and standard deviations (mean ±SD), while skewed data are reported as the medians (M) and interquartile ranges (IQR). Count data are presented as numbers and percentages (%).

Given the differences in the baseline characteristics between eligible participants in the two groups (Table 1), propensity-score matching was performed using the PSMATCH procedure in SPSS Statistics. The propensity score was estimated using a logistic regression model incorporating the following baseline covariates: age, BMI, gestational age, ASA classification, parity, smoking status, diabetes, hypertension, hypothyroxinemia during pregnancy, education level, employment status, prenatal exercise, infant sex, duration of anesthesia, postoperative complications, duration of hospital stays, global PSQI score before surgery and EPDS score before surgery. Participants in the intervention group were matched 1:1 to those in the control group using the nearest neighbor method within a specified caliper width. The caliper was set to 0.02 on the propensity score scale. The matching order was unspecified (as per the software default). Standardized mean differences (SMD) were estimated for all the baseline covariates before and after matching to assess pre-match imbalance and post-match balance. SMD of less than 10.0% for a given covariate indicate a relatively small imbalance. For non-essential matching variables, an SMD of less than 20.0% was considered acceptable. The magnitude of SMD was measured using Hedges’g as the effect size.

Table 1 Demographic Data

In the original cohort, independent Student’s t-tests and Mann–Whitney U-tests were employed to compare between-group differences for normally distributed continuous variables and non-normally distributed continuous variables. In the matched cohort, paired Student’s t test and Wilcoxon signed-rank test were employed to compare between-group differences for normally distributed continuous variables and non-normally distributed continuous variables, respectively.

Results

Participant Characteristics

Through the electronic medical record system (Docare Clinical Anesthesia Information System V5.0, Madison Medical Technology), a total of 245 patients meeting the inclusion criteria were retrieved from 1 January 2024 to 31 December 2024. Six patients in the propofol group and 22 patients in the control group were excluded. Two patients in the propofol group and four patients in the control group were lost to follow-up (Figure 1).

Figure 1 Flowchart of patient selection, inclusion, and matching for the study cohort. Abbreviation:CS, cesarean section.

Before propensity-score matching, there were differences between the two groups in several of the baseline variables (Table 1). With the use of propensity-score matching, 29 patients who underwent pharmacological intervention of propofol following extraction of the infant were matched with 29 patients in the control group. After matching, the standardized differences for the key variables were less than 10.0%, indicating the differences between the two groups were very small (Table 1). The matching rate was 82.9% in the propofol group. After propensity score matching, the distributions of propensity scores showed similarity and overlap between the two groups (Figure 2).

Figure 2 Propensity score distributions in the control and propofol groups, before and after matching. Abbreviation:PS, propensity score.

Primary Outcome

The results of between-groups comparisons remained consistent before and after propensity score matching (Table 1). After matching, the PSQI scores on the 30th day after surgery were 8.3 ± 3.4 in the propofol group versus 10.9 ± 3.1 in the control group, with a difference of −2.6 (95% CI: −3.1 to −2.1), and the difference was statistically significant (t = −11.154, P < 0.001) (Table 2). The estimated Hedges’ g effect size was −2.043 (95% CI: −2.679 to −1.395). The Pearson correlation was observed between groups (r = 0.926, P < 0.001) (Figure 3A). Among the seven components of the PSQI scale, significant intergroup differences were observed in sleep quality, sleep latency, sleep duration and sleep disturbances (P < 0.05), while no significant differences were found in the remaining components (Table 2 and Figure 4).

Table 2 Comparison of PSQI and EPDS Score After Surgery Between Two Groups

Figure 3 Comparison of postoperative PSQI and EPDS between the two groups. (A) Global PSQI scores. (B) EPDS scores. Abbreviation: PSQI, Pittsburgh Sleep Quality Index; EPDS, Edinburgh Postnatal Depression Scale.

Figure 4  Comparison of Pittsburgh Sleep Quality Index (PSQI) component scores between the propofol and control groups after matching. The seven components are: (A) Sleep quality, (B) Sleep latency, (C) Sleep duration, (D) Sleep efficiency, (E) Sleep disturbances, (F) Use of sleeping medication, and (G) Daytime dysfunction. P-values were derived from the Wilcoxon signed-rank test, indicating statistical significance for most components between the two groups.

Secondary Outcome

After matching, the postoperative EPDS scores on the 30th day after surgery were 7.9 ± 3.0 in the propofol group compared with 10.2 ± 3.1 in the control group, with a difference of −2.3 (95% CI:-3.0 – −1.7), and the difference was statistically significant (t = −7.261, P < 0.001) (Table 2). The estimated Hedges’g effect size was −1.330 (95% CI: −1.823 to −0.825). The Pearson correlation was observed between groups (r = 0.842, P < 0.001) (Figure 3B).

Discussion

Our trial assessed the efficacy of propofol as a pharmacological intervention on sleep quality in patients undergoing elective CS with spinal anesthesia. Our study found that intravenous propofol administration following extraction of the infant significantly reduced occurrence of postpartum sleep disorders.

Due to the influence of physiological, psychological and social factors, postpartum sleep disturbances are common. The changes of neuroendocrine during postpartum period affect sleep regulation.¹⁹ The interaction between progesterone and GABA_A receptor are highly selective. The decline in progesterone levels within postpartum alter GABA_A receptor structure and function, significantly impairing sleep maintenance.²⁰ Moreover, a deficiency in GABA_A receptor regulation during postpartum may make individuals prone to postpartum-related mood disorders.²¹ Propofol’s GABA_A receptor-mediated modulation has been empirically demonstrated to enhance sleep.²² Current evidence demonstrates that propofol administration significantly improves postoperative sleep quality in non-obstetric surgical patients.^13,14 Based on the above research, we reasonably suppose that intravenous administered propofol following extraction of the infant might have improved postpartum sleep quality by GABA receptor-mediated modulation.

Our study was the first to investigate the impact of intravenous drug intervention on sleep quality after elective CS, utilizing subjective PSQI. Our results showed that intravenous propofol administration following extraction of the infant lowered the PSQI score on postoperative days 30. As for each component of PSQI scores, the postoperative PSQI components of sleep quality, sleep latency, sleep duration and sleep disturbances scored significantly lower in the propofol group.

Postpartum depression (PPD), akin to postpartum sleep disturbance, exhibits high prevalence rates and exerts profound adverse effects on both maternal psychological well-being and physiological recovery. A multicenter cross-sectional study of 1065 postpartum mothers in Northeast China revealed a 23.57% prevalence of depressive symptoms at 6 weeks postpartum.²³ Our results showed that intravenous propofol administration following extraction of the infant lowered the EPDS score on postoperative days 30. This phenomenon may involve dual mechanistic pathways: (1) propofol exerts direct anxiolytic and antidepressant effects²⁴ and (2) propofol indirectly alleviates PPD by improving the postpartum sleep quality. Sleep disturbances constitute a significant modifiable risk factor for PPD, improving sleep quality can help alleviate PPD.^25–28

Although studies in non-obstetric surgeries have demonstrated that intraoperative propofol administration improves postoperative sleep quality and reduces depression, its use during CS remains restricted due to safety concerns. Pregnant women frequently present with full stomach and difficult airways, increasing risks of regurgitation/aspiration and airway obstruction with propofol use. Additionally, propofol crosses the placental barrier, raising concerns about fetal drug exposure. Importantly, administering propofol after extraction of the infant eliminates fetal drug exposure. In clinical practice, anesthesiologists report that propofol use is primarily driven by maternal request, acute anxiety, or preoperative sleep deprivation. This explains why the propofol group in our original cohort predominantly comprised mothers with more severe baseline sleep disturbances and mood disorders. Propofol rapidly ameliorates acute emotional distress and preoperative sleep deprivation.

While maternal sleep disturbances have been extensively studied, with numerous interventions targeting pregnancy and postpartum periods, the compounded effects of surgical stress and anesthesia emerged during CS. This critical perioperative window—where sleep and mood disturbances peak—has paradoxically received minimal therapeutic attention. Our study bridges this gap by demonstrating that intraoperative propofol administration reduces both postoperative sleep disorders and PPD, thereby enabling continuum-of-care from pregnancy through postpartum.

This study is a retrospective cohort analysis with certain limitations. One of the limitations of our study is that despite rigorous propensity-score matching, residual confounding remains possible due to unmeasured variables such as detailed socioeconomic status, specific psychosocial stressors and specific breastfeeding practices. Second, the relatively small proportion of propofol use in our cohort resulted in a limited sample size after matching. This not only increases the risk of a type II error but also limits the generalizability of our findings. We were unable to maintain standardized mean differences (SMDs) below 10% for all covariates, achieving this balance only for key variables, which may somewhat compromise the overall comparability between groups. Third, the assessment of the primary outcome via telephone interview at 30 days postoperatively is susceptible to recall bias. Recalling sleep quality and emotional states over such an extended period is challenging and can be influenced by a patient’s most recent experiences or their current mood state. Subjective recall has the potential to underestimate or overestimate the true postoperative situation, impacting the accuracy of our effect estimates. Finally, patients assigned to the propofol regimen, characterized by greater severity of sleep disorders and depression, demonstrated a propensity for meticulous retrospection and focused attention on risk factors. Nevertheless, our findings suggest a potential protective effect of propofol use during CS on postpartum sleep quality and depression, warranting further validation through randomized controlled trials to establish its true effect.

Conclusions

In summary, our study suggested that use of propofol significantly improved postoperative sleep quality and reduced the postpartum depression in in patients undergoing elective CS with spinal anesthesia. Future prospective studies are needed to confirm the causal efficacy of propofol in improving postoperative sleep quality after cesarean section and to elucidate its underlying mechanisms.