Esketamine as an opioid-sparing adjuvant for epidural labour analgesia: a randomised, double-blind trial evaluating postpartum depression Authors: Yi-ying Zhang, Xu-li Cheng, Li Liu, Lin Li, Yugang Diao Published: 2025 ABSTRACT AbstractBackgroundOpioid-based epidural analgesia, while effective, is frequently associated with maternal side effects such as pruritus and potential central sensitisation. Esketamine provides rapid analgesic and antidepressant effects at subanesthetic doses. We compared esketamine with sufentanil both added to 0.083% ropivacaine focusing on efficacy, maternal safety, and postpartum depression outcomes.MethodsIn this single-centre, randomised, double-blind, controlled trial, 200 nulliparous parturients (ASA I–II, 18–40 years, cervical dilatation 2–3 cm) requesting epidural analgesia were equally allocated to receive patient-controlled epidural analgesia (PCEA) containing either sufentanil 0.3 µg/ml or esketamine 0.3 mg/ ml combined with ropivacaine 0.083%. Primary outcomes included analgesic onset time, maximum sensory block level and Edinburgh Postnatal Depression Scale (EPDS) at 42 days postpartum. Secondary outcomes included adverse events, VAS scores and neonatal Apgar scores.ResultsBetween December 2024 and May 2025, 197 participants completed the study (98 of sufentanil, 99 of esketamine). Baseline characteristics were comparable. Esketamine significantly reduced onset time (5.9 ± 0.6 min vs. 9.8 ± 1.7 min, P < 0.001), achieved a predominant T8 sensory block (67.7% vs. 45.9%, P < 0.001) and lowered EPDS scores (5.6 ± 1.3 vs. 9.8 ± 1.5, P < 0.001). Total ropivacaine consumption throughout the first stage of labour was 17% lower in the esketamine group (12.4 ml [10.3–14.5] vs. 14.9 ml [12.6–17.2]; P < 0.001).EPDS scores at 42 days were lower with esketamine (5.6 ± 1.3 vs. 9.8 ± 1.5, P < 0.001). The incidence of probable postpartum depression (EPDS ≥13) was also reduced in the esketamine group (4.0% vs. 18.4%, P = 0.003). Maternal adverse events such as pruritus (3.0% vs. 45.9%), hypotension (4.0% vs. 38.8%), and urinary retention (3.0% vs. 22.4%) were markedly reduced with esketamine (all P < 0.001). VAS and Apgar scores were similar between groups.ConclusionsEsketamine–ropivacaine epidural analgesia provides faster onset, fewer maternal side effects, and potential antidepressant benefits compared with sufentanil–ropivacaine. These findings support esketamine as a promising opioid-sparing strategy in obstetric anaesthesia.Trial registrationChinese Clinical Trial Registry, ChiCTR2500109086. Retrospectively registered on September 11, 2025.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12871-025-03590-1.Keywords: Esketamine, Sufentanil, Ropivacaine, Epidural labour analgesia, Postpartum depression, Opioid-sparing, Randomised controlled trial Background Labor pain is one of the most severe visceral pains that women have ever experienced, often exceeding the intensity of postoperative incisional pain [1]. Severe and persistent pain during childbirth can activate the hypothalamic-pituitary-adrenal axis, leading to elevated levels of plasma cortisol and catecholamines. This neuroendocrine response may prolong labour, increase the risk of instrumental delivery, impair lactation, and significantly elevate the risk of postpartum depression (PPD) [2–4]. A recent 2024 study involving 429 women in China reported an overall PPD prevalence of 22.14%, with nulliparous women and those receiving inadequate or side-effect-prone epidural analgesia being at highest risk [5]. PPD not only compromises maternal mental health but is also associated with impaired infant cognitive development, disrupted mother–infant bonding, and long-term familial burden [6, 7]. Currently, patient-controlled epidural analgesia (PCEA) using low-concentration local anaesthetics combined with a lipophilic opioid is regarded as the gold standard for labor analgesia [8]. Sufentanil, the most commonly used opioid adjuvant, is associated with dose-dependent side effects such as pruritus, hypotension, and potential central sensitisation [9, 10]. Therefore, an effective opioid-sparing adjuvant is urgently needed. Opioid-sparing refers to any intervention that reduces total opioid consumption while maintaining equivalent analgesia. Esketamine is a non-competitive N-methyl-D-aspartic acid (NMDA) receptor antagonist with twice the analgesic potency and a more favorable psychotropic safety profile compared with racemic ketamine [11–13]. At sub-anaesthetic doses, esketamine inhibits central sensitisation, reduces postoperative opioid consumption, and exerts rapid antidepressant effects via brain-derived neurotrophic factor (BDNF)–mammalian target of rapamycin (mTOR) signalling [14, 15]. Epidural administration allows high concentrations to be delivered to the spinal dorsal horn while limiting systemic exposure [16]. Previous dose-finding studies using the up-and-down sequential method have confirmed that 0.3 mg/ml esketamine is the optimal concentration when used as an adjuvant to ropivacaine [17]. However, large-scale trials assessing its impact on postpartum depression are still lacking. We therefore conducted an adequately powered, randomised, double-blind, superiority trial to compare epidural esketamine with sufentanil as an adjuvant to 0.083% ropivacaine for PCEA in nulliparous women, with a focus on analgesic efficacy, maternal safety, and postpartum depression outcomes. Methods Trial design This prospective, parallel-group, double-blind, superiority trial was conducted at The General Hospital of Northern Theater Command, Shenyang, China, between December 2024 and May 2025. Ethical clearance was granted by the Medical Ethics Committee under reference Y(2024)182, and the study was subsequently logged in the Chinese Clinical Trial Registry (ChiCTR2500109086) on September 11, 2025. Although trial registration (ChiCTR2500109086) was completed after recruitment of the first participant because of administrative delays, the protocol was finalised and ethically approved before any patient contact. This study was conducted and reported in accordance with the CONSORT 2010 statement. Participants The inclusion criteria were 18–40 years old, ASA physical condition grade I-II, singleton head position pregnancy ≥ 37 weeks, cervical dilation 2–3 cm, and epidural analgesia was required. Exclusion criteria comprised contraindications to epidural block, allergy to study drugs, hypertension or preeclampsia, severe psychiatric disorders, and conversion to cesarean delivery during labour. Randomisation and blinding A biostatistician independent of patient care generated a 1:1 computer-derived randomisation list (block size 4) using Excel 2019. The hospital pharmacy produced a set of sealed, opaque envelopes that were consecutively numbered. A consultant anaesthetist who took no part in later evaluations opened the next envelope in sequence and prepared the assigned investigational solution. To maintain blinding, the study solutions were prepared in identical syringes and labeled with participant codes only. Patients, midwives, outcome assessors and statisticians were blinded to group allocation. To minimize assessment bias, outcome assessors were not involved in epidural management and remained blinded throughout the study. Emergency unblinding was permissible for severe adverse events. Interventions Following pre-hydration with 500 ml of lactated Ringer’s administered intravenously, the epidural space was approached at the L2–L3 interspace while the patient lay in the left lateral decubitus position. Then a multi-orifice catheter was advanced 3.5 cm cephalad and 3 ml lidocaine 1% was administered as a test dose. Following negative aspiration and absence of intrathecal or intravascular signs after 5 min, a 10 ml loading dose of the assigned study solution was given and connected to a PCEA pump (0.083% ropivacaine with either sufentanil 0.3 µg/ml or esketamine 0.3 mg/ml). Pump settings were: background infusion 8 ml/ h, patient-controlled bolus 4 ml, lockout interval:15 min, and 30 ml (4-hour maximum total dose: 30 ml). The background infusion provides continuous baseline analgesia, while patient-controlled boluses allow breakthrough pain relief. Rescue top-ups (5 ml study solution) were permitted for VAS ≥ 4; persistent VAS ≥ 6 despite two top-ups was recorded as protocol failure. Outcomes Primary outcomes were: (1) analgesic onset time (time from loading dose to VAS ≤ 3), (2) maximum sensory block level to pin-prick, and (3) Edinburgh Postnatal Depression Scale (EPDS) score at 42 days postpartum. An EPDS score ≥ 13 was used to indicate probable postpartum depression, consistent with the validated Mandarin threshold [18, 19]. EPDS scores at 42 days postpartum were obtained by trained research staff via telephone interview using the Mandarin-validated version of the Edinburgh Postnatal Depression Scale [19]. The 42-day time-point was chosen because it coincides with routine postpartum outpatient visits and captures the peak incidence of early postpartum mood disorders [20]. Inter-rater reliability was assessed in 10% of randomly selected cases (κ = 0.92). Secondary outcomes included incidence of adverse events (pruritus, hypotension, urinary retention), VAS scores at baseline and 30 min, and neonatal Apgar scores at 1 and 5 min. Sedation ≥ 2 was defined as a Ramsay Sedation Score ≥ 2 (drowsy but responsive to verbal commands). Total ropivacaine consumption (ml) during the first stage of labour was recorded by the PCEA pump and calculated from the sum of background infusion, patient-controlled boluses, and any rescue top-ups administered. These secondary outcomes were exploratory and not included in the original sample size calculation. Hypotension was defined as systolic blood pressure < 90 mmHg or < 80% of baseline. Exploratory secondary outcomes also included maternal satisfaction score (0–10 numerical rating scale), time to first top-up request (min), duration of the first stage of labour (min) and uterine contraction frequency (contractions/10 min). No changes were made to the trial outcomes after the trial commenced. Sample size calculation Based on our pilot data (mean ± SD: sufentanil :9.2 ± 1.5 min vs. Esketamine: 6.0 ± 0.7 min), a two-sample t test with 90% power and α = 0.05 (two-tailed) required 92 participants per group to detect a mean difference of 3.2 min. Allowing for a 10% dropout rate, we enrolled 200 participants. Statistical analysis Statistical analysis was carried out with SPSS 26.0. Distribution of continuous variables was examined by the Shapiro–Wilk test; results are expressed as mean ± SD for normal data and as median [IQR] for non-normal data, with between-group comparisons performed using the independent-samples t test or Mann–Whitney U test as appropriate. Categorical variables are summarised as frequencies and percentages and were compared by χ² or Fisher’s exact test. For binary endpoints, relative risk (RR) together with 95% confidence intervals (CI) was computed. All tests were two-tailed, and a P value < 0.05 indicated statistical significance. Survival analysis was performed using the Kaplan–Meier method to compare the time to onset of analgesia (defined as time from loading dose to VAS ≤ 3) between the two groups. The log-rank test was used to assess the statistical significance of the difference in survival curves. Hazard ratios (HR) with 95% confidence intervals (CI) were calculated using Cox proportional hazards regression. No interim analyses or stopping guidelines were specified; the trial was conducted to completion as planned. Data availability The data of the study is not openly accessible; however, it can be provided if requested. Results Study population Between 10 December 2024 and 30 May 2025, 200 parturients were screened; three were excluded after intrapartum cesarean section became necessary. The remaining 197 women were enrolled and randomly allocated to the sufentanil arm (n = 98) or the esketamine arm (n = 99). Patient recruitment and allocation are summarised in Fig. 1. As shown in Table 1, the demographic and clinical characteristics of two groups were comparable, including age (sufentanil group: 27.2 ± 3.4 vs. esketamine group: 27.4 ± 3.1 ; P = 0.693), BMI (sufentanil group: 25.3 ± 2.6 kg/m² vs. esketamine group: 25.0 ± 2.2 kg/m²; P = 0.383), gestational age (sufentanil group: 39.1 ± 0.7 weeks vs. esketamine group: 38.9 ± 0.6 weeks; P = 0.071), and ASA physical status, baseline HR, mean arterial pressure (MAP), and SpO2 were also balanced (all P > 0.05)(Table 1). CONSORT 2020 flow diagram Table 1Baseline demographic and clinical characteristicsSufentanil(n = 98)Esketamine(n = 99)P-valueAge, y27.2 ± 3.427.4 ± 3.10.693BMI, kg m− 225.3 ± 2.625.0 ± 2.20.383ASA physical statusa0.646 I62 (63.3)59 (59.6) II36 (36.7)40 (40.4)Gestational age, week39.1 ± 0.738.9 ± 0.60.071Baseline HR, beats/min79.5 ± 8.780.1 ± 4.90.559Baseline MAP, mmHg86.8 ± 9.585.7 ± 9.90.421Data are presented for 197 nulliparous women (Nulliparous defined as no previous delivery ≥ 20 weeks gestation). Values are mean ± SD or n (%)BMI body mass index, MAP mean arterial pressure, HR heart rate, ASA American Society of AnesthesiologistsaASA I: normal healthy, ASA II: mild systemic disease Baseline demographic and clinical characteristics Data are presented for 197 nulliparous women (Nulliparous defined as no previous delivery ≥ 20 weeks gestation). Values are mean ± SD or n (%) BMI body mass index, MAP mean arterial pressure, HR heart rate, ASA American Society of Anesthesiologists aASA I: normal healthy, ASA II: mild systemic disease Primary outcomes Kaplan–Meier analysis showed a significantly shorter time to VAS ≤ 3 in the esketamine group (median 5.9 min; 95% CI: 5.3–6.5) compared with the sufentanil group (median 9.8 min; 95% CI: 8.1–11.5) (log-rank test, P < 0.001)(Fig. 2). Maximum sensory block reached T8 in 67 (67.7%) esketamine vs. 45 (45.9%) sufentanil parturients (P < 0.001) (Table 2). Mean EPDS score at 42 days was lower with esketamine (5.6 ± 1.3 vs. 9.8 ± 1.5, P < 0.001). Probable postpartum depression (EPDS ≥ 13) occurred in 4/99 (4.0%) vs. 18/98 (18.4%); relative risk 0.22 (95% CI 0.08–0.62), P = 0.003. Absolute risk reduction 14.3% points (95% CI 5.9–22.7%), number-needed-to-treat 7 (95% CI 4–12). Kaplan–Meier analysis of time to VAS ≤ 3 after epidural loading dose Table 2Primary outcomesSufentanil(n = 98)Esketamine(n = 99)P-valueAnalgesic onset timea, min9.8 ± 1.75.9 ± 0.6< 0.001Maximum block levelb, n (%)< 0.001 T69(9.2)1 (1.0) T723 (23.5)10 (10.1) T845 (45.9)67 (67.7) T921(21.4)21 (21.2)EPDS scores, mean ± SD9.8 ± 1.55.6 ± 1.3< 0.001(EPDS ≥ 13)c18(18.4)4(4.0)0.003EPDS, Edinburgh Postnatal Depression ScaleaTime from loading dose to VAS≤3bSensory level to pin-prickcEPDS ≥13 indicates probable postpartum depression [19] Primary outcomes EPDS, Edinburgh Postnatal Depression Scale aTime from loading dose to VAS≤3 bSensory level to pin-prick cEPDS ≥13 indicates probable postpartum depression [19] Sensitivity analysis Because the low event rate of EPDS ≥ 13 (22/197) led to complete separation when multiple continuous covariates were entered, we fitted a reduced model adjusting for onset time (continuous) and three binary variables: pruritus, hypotension, and maternal satisfaction. Complete separation occurs when no overlap exists between covariate values of events and non-events, causing logistic regression coefficients to diverge to infinity. After adjustment, the esketamine group retained a protective point estimate (adjusted OR = 0.70, 95% CI 0.43–1.14, P = 0.149)(Supplementary Table 1), with the point estimate consistently < 1, though the 95% CI crosses 1.0. These data suggest the observed reduction in PPD symptoms is unlikely to be fully explained by confounding. Secondary outcomes Pruritus (3.0% vs. 45.9%, RR 0.07, 95% CI 0.02–0.22), hypotension (4.0% vs. 38.8%,RR 0.10, 95% CI 0.04–0.27 ) and urinary retention (3.0% vs. 22.4%, RR 0.13, 95% CI 0.04–0.43) were all markedly reduced in the esketamine group (P < 0.001) (Table 3). Total ropivacaine consumption during the first stage of labour was significantly lower in the esketamine group (12.4 ± 2.1 ml vs. 14.9 ± 2.6 ml, P < 0.001), corresponding to a 17% reduction. VAS scores decreased similarly (baseline:7.8 ± 0.4 vs. 7.1 ± 0.3; 30 min:1.1 ± 0.4 vs. 1.2 ± 0.3; P > 0.05). Apgar scores at 1 and 5 min were comparable between groups (P > 0.05). No respiratory depression, sedation ≥ 2, or hallucinations were observed. Additional secondary outcomes are presented in Table 3. Median [IQR] satisfaction score was higher in the esketamine group (9 [8–10] vs. 7 [6–8]; P < 0.001). Time to first top-up request was prolonged (92.5 ± 16.1 vs. 61.3 ± 11.5 min; P < 0.001). Duration of the first stage of labour was similar (397 ± 45 vs. 389 ± 51 min; P = 0.21), as was contraction frequency (3.2 ± 0.4 vs. 3.3 ± 0.5 contractions/10 min; P = 0.18). Table 3Secondary outcomesSufentanil(n = 98)Esketamine(n = 99)P-valueAdverse eventsPruritusa45 (45.9)3 (3.0)< 0.001Hypotensionb38 (38.8)4 (4.0)< 0.001Urinary retention22(22.4)3(3.0)< 0.001Total ropivacaine consumption (ml)14.9 ± 2.612.4 ± 2.1< 0.001Duration of the first stageof labor (min)389.2 ± 51.0397.1 ± 45.80.211Duration of the second stageof labor (min)102.2 ± 15.7106.9 ± 20.10.184Satisfaction scores of puerpera7.2 ± 0.99.0 ± 1.1< 0.001Time to first top-up request61.3 ± 11.592.5 ± 16.1< 0.001VAS scores, mean ± SD T0 (baseline)7.8 ± 0.47.1 ± 0.30.089 T30 (30-min)1.2 ± 0.31.1 ± 0.40.162Apgar scores, mean ± SD 1 min after surgery8.9 ± 0.39.0 ± 0.20.225 5 min after surgery9.9 ± 0.210.0 ± 0.10.141Values are mean ± SD or n (%). Satisfaction score: 0–10 numerical rating scalea Pruritus: maternal report of itching requiring any interventionb Hypotension: SBP < 90 mmHg or < 80 % of baseline Secondary outcomes Duration of the first stage of labor (min) Duration of the second stage of labor (min) Values are mean ± SD or n (%). Satisfaction score: 0–10 numerical rating scale a Pruritus: maternal report of itching requiring any intervention b Hypotension: SBP < 90 mmHg or < 80 % of baseline Discussion Principal findings In this double-blind, superiority trial, epidural esketamine 0.3 mg/ml combined with 0.083% ropivacaine significantly shortened analgesic onset by 4 min, reduced total ropivacaine consumption by 17%, and markedly decreased maternal side-effects (pruritus 3% vs. 46%; hypotension 4% vs. 39%; urinary retention 3% vs. 22%). Most importantly, the incidence of probable postpartum depression (EPDS ≥ 13) fell from 18.4% to 4.0%, yielding an absolute risk reduction of 14.3% points (95% CI 5.9–22.7%) and a number-needed-to-treat of 7 (95% CI 4–12). These effect sizes exceed the minimal clinically important difference of 3 points on the EPDS and support esketamine as a promising opioid-sparing adjuvant for nulliparous women undergoing epidural labour analgesia. Mechanistic interpretation Activation of NMDA receptors in the spinal dorsal horn plays a key role in central sensitisation and opioid-induced hyperalgesia [21]. By blocking these receptors, esketamine attenuates wind-up, enhances endogenous pain inhibition, and exerts rapid antidepressant effects via mTOR-dependent BDNF release [22]. Epidural administration achieves high cerebrospinal fluid concentrations while limiting systemic exposure, thereby preserving maternal haemodynamic stability and neonatal condition [23]. We did not measure plasma or cerebrospinal-fluid esketamine concentrations; therefore central penetration is plausible but unquantified. The observed EPDS reduction might reflect indirect benefits (better analgesia, fewer side-effects) rather than a direct antidepressant action. A recent epidural dose-finding study reported CSF levels ≈ 60 ng/ml within 30 min of a 0.3 mg/ml bolus [24]—values that overlap the systemic antidepressant window (10–100 ng/ml) described by Bahji 2022 [25]. If even a fraction of this concentration crosses the blood–brain barrier, receptor-level effects are plausible; without serial biosamples we can only state that the association is consistent, not causal. A pharmacokinetic sub-study nested in our upcoming multicentre trial will clarify the relative contribution of direct CNS penetration versus indirect benefits (better analgesia, fewer side-effects). Comparison with existing studies Our findings are consistent with recent studies investigating epidural esketamine in labor analgesia. For example, Xu et al. (2024) reported a dose-dependent analgesic effect of epidural esketamine [17], while Lou et al. (2023) identified the ED90 of esketamine in combination with ropivacaine [24]. Safety considerations Esketamine was well tolerated in our study. No respiratory depression, sedation ≥ 2, hallucinations, or psychotomimetic effects were observed. These findings align with recent obstetric studies reporting favorable safety profiles of esketamine at sub-anesthetic doses [17]. However, epidural esketamine has been associated with dose-dependent mild dizziness at concentrations above 1 mg/ml [25]. Our study used a lower concentration (0.3 mg/ml), which may explain the absence of such side effects. These findings are corroborated by pharmacokinetic data showing rapid placental transfer but equally rapid neonatal redistribution and metabolism [26]. Although single-centre, the 4.2-point EPDS reduction exceeds the minimal clinically important difference (MCID) of 3 points for EPDS and supports implementation in similar tertiary centres; multicentre validation is planned. MCID denotes the smallest change in a scale that patients perceive as beneficial [27]. The limitations of this experiment Postpartum depression symptoms fluctuate over time; a single 42-day EPDS assessment may have missed late or transient cases. Future studies should measure EPDS at 1-week, 6-week and 3-month intervals and record baseline mood to capture the full trajectory. Second, this single-centre trial was conducted at a tertiary-level Chinese hospital; ethnic and obstetric management homogeneity may limit external validity. Third, we did not evaluate esketamine transfer into breast milk or its potential impact on neonatal neurobehaviour, which should be assessed in future trials. Fourth, the marked difference in analgesic onset time may have introduced assessment bias, although we attempted to mitigate this by blinding outcome assessors. Fifth, some secondary outcomes were exploratory and should be considered hypothesis-generating. Finally, because only nulliparous women were enrolled, the generalisability to multiparous parturients remains uncertain; multicentre trials with mixed parity are warranted. Conclusion Esketamine as an adjuvant to ropivacaine for epidural labour analgesia provides faster onset, fewer side effects, and potential antidepressant benefits compared with sufentanil. These findings support esketamine as a promising opioid-sparing alternative in obstetric anaesthesia, particularly in nulliparous women. Supplementary Information Supplementary Material 1. Authors’ contributions Conceptualization: L-L, Y-gD; Data curation: Y-yZ, X-lC; Formal analysis: Y-yZ; Funding acquisition: L-L; Investigation: L-L, X-lC; Methodology: L-L, X-lC, Y-gD; Project administration: L-L; Resources: L-L; Software: Y-yZ< br/> Supervision: L-L, Y-gD; Validation: L-L, Y-gD; Visualization: Y-yZ; Writing – original draft: Y-yZ, X-lC; Writing – review & editing: L-L, Y-gD. Funding The work of Lin Li was supported by Liaoning Province Science and Technology. Project (2024JH2/102600283). Data availability The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate The study protocol was reviewed and approved by the Medical Ethics Committee of the General Hospital of Northern Theater Command [IRB No. Y(2024)182] in compliance with the Declaration of Helsinki. Written informed consent was obtained from all participants before enrollment. Consent for publication Not applicable. 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. Yi-ying Zhang and Xu-li Cheng contributed equally to this work. Contributor Information Lin Li, Email: lilinslashofmine@qq.com. Yugang Diao, Email: diao72@163.com. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The data of the study is not openly accessible; however, it can be provided if requested. The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.