Peripartum Esketamine: A Narrative Review of Analgesia, Postpartum Depressive Symptoms, Sleep-Related Outcomes, and Safety

Abstract

Pain, depressive symptoms, and sleep disruption frequently overlap during the peripartum period, although their relationships are associative rather than established as causal. Esketamine, an N-methyl-D-aspartate receptor antagonist, has attracted interest as a peripartum intervention with potential effects on pain, mood, and sleep. This narrative review evaluates the clinical and mechanistic evidence for peripartum esketamine in analgesia, postpartum depressive symptoms, sleep-related outcomes, and maternal–infant safety, with particular attention to the directness and interpretability of the evidence. Most direct clinical evidence derives from cesarean-delivery studies, many of which are single-center trials. Evidence is most consistent for short-term analgesic benefit, whereas opioid-sparing effects are less consistent across individual studies and evidence syntheses. Several randomized trials suggest that low-dose peri-delivery esketamine may reduce early EPDS-defined screen-positive postpartum depressive symptoms, while evidence based on structured diagnostic outcomes remains limited. Evidence for sleep-related benefit remains preliminary and relies predominantly on subjective measures, with no direct postpartum polysomnographic evidence. Studied low-dose regimens have generally shown acceptable short-term maternal and neonatal tolerability in selected populations, but pregnancy-specific pharmacokinetics, lactation, and long-term offspring neurodevelopment remain insufficiently characterized. Overall, esketamine may represent an individualized adjunct in peripartum care, but current evidence does not support its routine prophylactic use. Future multicenter studies should incorporate diagnostic mood outcomes, objective sleep assessment, standardized treatment protocols, and long-term maternal and infant follow-up.

Introduction

Pain, depressive symptoms, and sleep disruption frequently overlap during the peripartum period. Severe acute postpartum pain predicts later depressive symptoms, while antenatal and postpartum sleep disturbance is associated with postnatal depression. These observations support coordinated consideration of pain, mood, and sleep, but they do not establish a shared causal pathway or a bidirectional causal cycle.1–3 In a large screening program of approximately 10,000 postpartum women, 14.0% screened positive for depression, and some screen-positive women reported thoughts of self-harm.4 Existing preventive and therapeutic approaches differ in accessibility, timing of effect, and tolerability, while pharmacotherapy during lactation may require individualized benefit–risk assessment according to maternal and infant factors.5,6 These considerations have prompted interest in peripartum strategies that might influence more than one clinically relevant outcome.

Esketamine is the S-enantiomer of ketamine and an antagonist at the N-methyl-D-aspartate receptor (NMDAR). Recent randomized trials suggest that low-dose peri-delivery esketamine may improve short-term analgesic outcomes and may reduce early EPDS-defined screen-positive postpartum depressive symptoms in some populations.7,8 In women with antenatal depressive symptoms, a single 0.2 mg/kg infusion after childbirth reduced the incidence of a major depressive episode diagnosed by structured interview at 42 days postpartum.9 Glutamatergic signaling participates in nociception, emotional processing, and sleep–wake regulation, providing a biological rationale for considering these domains together; however, mechanistic overlap does not establish a common clinical treatment effect.

The available direct clinical evidence is geographically and methodologically concentrated, with many studies conducted in Chinese obstetric populations and at single centers, which may limit generalizability and the independence of apparently concordant findings. Rather than performing another pooled efficacy analysis, this narrative review uses a cross-domain framework to compare the directness, consistency, and clinical interpretability of evidence for analgesia, postpartum depressive symptoms, sleep-related outcomes, and maternal–infant safety. The peri-delivery intravenous regimens discussed in this review should be regarded as investigational rather than established routine obstetric indications. In the United States, approved esketamine is available as an intranasal formulation for specific depressive disorders and is subject to supervised administration and post-dose monitoring requirements.10 Because direct clinical studies have focused predominantly on cesarean delivery, labor analgesia, and early postpartum care, the primary clinical scope of this review is peripartum rather than antenatal use. Evidence concerning antenatal exposure, placental transfer, lactation, and developmental safety is included primarily to contextualize maternal–infant safety rather than to infer antenatal efficacy.

Literature Search and Evidence Selection

A focused literature search was conducted in PubMed and the Web of Science Core Collection for articles published from database inception to July 21, 2026. Search terms related to esketamine or S-ketamine were combined with terms covering pregnancy, childbirth, cesarean delivery, the postpartum period, analgesia, depressive symptoms, sleep, lactation, and maternal or neonatal safety. The reference lists of relevant clinical studies and reviews were also examined to identify additional publications. The database-specific search terms are provided in Supplementary Table S1.

This narrative review included randomized controlled trials, observational studies, pharmacokinetic and safety studies, systematic reviews, and meta-analyses involving pregnant, laboring, or postpartum women. Selected preclinical and non-obstetric studies were also considered when direct peripartum evidence was limited and when they provided relevant pharmacological, mechanistic, sleep-related, or safety information. The literature was organized thematically according to analgesia, postpartum depressive symptoms, sleep-related outcomes, and safety.

Pharmacological Basis

Stereoselectivity and Pharmacological Potency

Ketamine is a chiral drug used clinically as a racemic mixture of S(+)- and R(-)-enantiomers; esketamine is the S(+)-enantiomer. Volunteer studies found greater anesthetic potency and faster recovery with S-ketamine than with R-ketamine, and electrophysiological experiments showed approximately twofold greater inhibition of NMDA-receptor currents.11,12 Early ex vivo cardiac-perfusion work also demonstrated stereoselective effects.13 These differences allow a lower dose to achieve a comparable anesthetic effect, but they do not by themselves imply better tolerability, which remains dependent on dose, route, and clinical setting. Intranasal esketamine was approved in 2019 for treatment-resistant depression,14,15 and its perioperative and obstetric use has since become an active area of study.16,17

NMDAR Signaling as a Mechanistic Link Across Domains

NMDAR blockade provides a biological basis for studying esketamine across pain, mood, and sleep, but it does not establish a common clinical effect.15,18 Foundational experiments showed that NMDAR activation contributes to the induction and maintenance of central sensitization.19 In cellular and animal models, NMDAR hypofunction reduces interneuron activity and increases pyramidal-cell output, AMPAR throughput is required for ketamine-like antidepressant effects, and mTOR-dependent synapse formation contributes to rapid behavioral responses.20–22 Perinatal animal models further implicate hippocampal synaptic plasticity, BDNF/AKT/mTOR signaling, and TASK-1 channels.23–25 Sleep-related evidence is more indirect: a clinical ketamine study associated greater early-night slow-wave activity with BDNF changes, and a polysomnographic trial after gynecological surgery found a higher proportion of slow-wave sleep with S-ketamine.26,27 These non-obstetric findings generate testable mechanisms but do not demonstrate an effect on postpartum sleep architecture. Figure 1 summarizes these proposed pathways and the differing levels of supporting evidence.

Figure 1.

Proposed pathways linking peripartum esketamine with analgesic, mood-related, and sleep-related outcomes. (A) illustrates the proposed analgesic pathway, including nociceptive processing in the spinal dorsal horn, NMDAR antagonism, and attenuation of central sensitization. Direct obstetric evidence is most consistent for short-term analgesic outcomes, whereas opioid-sparing effects are less consistent. (B) illustrates the proposed mood-related pathway, including prefrontal–hippocampal circuits, glutamatergic signaling, AMPAR activation, and BDNF–mTOR-related synaptic plasticity. Clinical evidence for early postpartum depressive symptoms remains heterogeneous and derives predominantly from screening-based outcomes. (C) illustrates the proposed sleep-related pathway, including sleep–wake networks and possible modulation of slow-wave and REM-related activity. Postpartum sleep-related evidence remains predominantly subjective, while objective sleep-architecture evidence is mainly indirect and non-obstetric. Solid lines indicate relationships supported predominantly by direct obstetric clinical evidence, dashed lines indicate mixed direct and indirect evidence, and dotted lines indicate predominantly indirect evidence. These proposed pathways are conceptual and do not establish causality.

Abbreviations: AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; BDNF, brain-derived neurotrophic factor; mTOR, mammalian target of rapamycin; NMDAR, N-methyl-D-aspartate receptor; REM, rapid eye movement.

Perinatal Pharmacokinetic Considerations

Pregnancy-specific pharmacokinetic data for esketamine are sparse. Intravenous ketamine crosses the placenta rapidly.28 A contemporary obstetric study found no evidence of worse short-term neonatal outcomes with the regimen evaluated,29 but this observation neither proves that fetal exposure is brief nor addresses long-term safety. In chronically instrumented pregnant sheep, clinically relevant S-ketamine doses did not significantly reduce uterine blood flow.30 Small lactation studies found low, rapidly declining concentrations of ketamine and its metabolites in human milk, with estimated relative infant doses below 1%.31,32 These data suggest limited exposure through breast milk, but they are insufficient to define a universal interruption interval for breastfeeding. Timing decisions should account for the clinical setting, maternal recovery, infant factors, and local guidance.

Esketamine in Peripartum Analgesia

Analgesia for Cesarean Delivery

Intravenous esketamine, usually at approximately 0.15–0.5 mg/kg, has been studied within multimodal analgesic regimens for cesarean delivery.33–36 A recent meta-analysis of randomized cesarean-delivery trials found reductions in several early postoperative pain outcomes, but pooled sufentanil consumption was not significantly reduced.37 These findings suggest that the evidence for short-term analgesic benefit is more consistent than that for opioid sparing. Individual trials have also reported reductions in visceral traction pain and improvements in other perioperative pain-related outcomes, although the studied doses, cointerventions, and outcome measures varied substantially.33–36,38 In a retrospective cohort of women converted from labor analgesia to emergency cesarean delivery, intravenous esketamine was not associated with worse short-term neonatal outcomes.39

Post-Cesarean Analgesia

After cesarean delivery, esketamine has been administered as a single intravenous dose, a continuous infusion, or an adjunct to patient-controlled intravenous analgesia (PCIA). A meta-analysis of esketamine–sufentanil PCIA reported lower postoperative pain scores, although the included studies involved mixed surgical populations and therefore provide only partly direct evidence for post-cesarean analgesia.40 Individual obstetric trials have reported improved early pain control with esketamine–sufentanil, esketamine–tramadol, or dexmedetomidine-containing PCIA regimens.41–43 Other randomized studies reported fewer PCIA activations or lower opioid exposure while also evaluating mood-related outcomes.44,45 However, these opioid-related findings were not uniform across the broader evidence base and are difficult to separate from differences in background analgesia and combination regimens. Neuraxial approaches have also been explored: intrathecal S(+)-ketamine with bupivacaine prolonged analgesia,46 a low-dose S-ketamine infusion was evaluated as preventive analgesia during spinal anesthesia,36 and a pilot trial examined epidural esketamine plus morphine.47 Overall, the evidence more consistently supports short-term analgesic benefit than a reproducible opioid-sparing effect.

Labor Analgesia During Vaginal Delivery

Prospective labor-analgesia studies have evaluated epidural esketamine as an adjunct to ropivacaine. One dose-response study characterized the dose–response relationship,48 and two sequential dose-finding studies estimated ED90 values for epidural esketamine when combined with low-concentration ropivacaine during early labor.49,50 These dose estimates are study-specific and should not be interpreted as standardized clinical dosing recommendations. In a randomized noninferiority trial, epidural esketamine was noninferior to fentanyl with respect to ropivacaine consumption.51 Conversely, a later three-arm study in a low-risk spontaneous-labor cohort found no significant between-group differences in EPDS or VAS outcomes when 0.5 mg/kg esketamine was combined with labor analgesia.52 Overall, these studies support the feasibility of epidural esketamine as a neuraxial adjunct but do not establish a preferred regimen, superiority over established comparators, or benefits for neonatal or longer-term maternal outcomes.

Mechanistic Considerations

NMDAR antagonism provides a biologically plausible mechanism for the analgesic effects of ketamine-class drugs, in part through attenuation of central sensitization rather than direct agonism at the mu-opioid receptor.19,53 However, mechanistic plausibility should not be interpreted as evidence of a consistent opioid-sparing effect in obstetric populations. NMDAR-dependent plasticity also contributes to the affective dimension of pain, but current peripartum studies cannot determine whether observed mood-related effects are independent of analgesic improvement or are partly mediated by better pain control. A non-obstetric spinal-surgery trial reported lower pain and opioid consumption one year after intraoperative S-ketamine.54 This finding provides indirect support for investigating persistent post-cesarean pain but should not be extrapolated as evidence that peripartum esketamine prevents chronic pain or reduces long-term opioid use.

Prevention of Postpartum Depressive Symptoms

Clinical Need for PPD Prevention

Postpartum depression can adversely affect maternal functioning and mother–infant interaction, and severe episodes may be accompanied by self-harm or suicidal thoughts.4,55 Pharmacotherapies targeting GABA–glutamate neurotransmission have expanded the therapeutic landscape for perinatal depression,56,57 but evidence for pharmacological prevention initiated around childbirth remains comparatively limited. This has contributed to interest in rapidly acting interventions that can be administered during the childbirth admission.

Evidence from Randomized Trials

Early single-center randomized trials tested intraoperative boluses, continuous infusions, and PCIA-based esketamine regimens for preventing postpartum depressive symptoms.44,45,58–62 Findings were mixed. Several studies reported lower EPDS scores or fewer EPDS-defined screen-positive cases at selected early assessments, whereas Liu et al found no significant reduction in screen-positive depressive symptoms through 6 months.60 These studies therefore provide evidence primarily for changes in screening outcomes rather than for prevention of clinically diagnosed postpartum depression.

An important distinction is provided by the multicenter trial of Wang et al,9 which enrolled women with antenatal depressive symptoms and used a structured diagnostic interview rather than an EPDS cutoff as its primary outcome. A single 0.2 mg/kg infusion after childbirth was associated with a lower incidence of major depressive episodes at 42 days postpartum. This trial therefore provides more clinically direct evidence for a diagnostic mood outcome, although its findings apply to a selected higher-risk population and follow-up was limited to 42 days. Other recent trials suggest that treatment effects may vary according to baseline depressive risk, regimen, and follow-up, but the available data are insufficient to establish baseline risk as a definitive treatment-effect modifier.

The findings of Ren et al8 and Yang et al63 also require cautious interpretation because both trials originated from the same institution and involved overlapping investigator groups; they should therefore not be treated as fully independent replication. Ren et al reported a lower 6-week EPDS-defined screen-positive rate after a 0.25 mg/kg infusion,8 whereas in the later three-arm trial only the bolus arm showed a statistically significant reduction in the screen-positive endpoint compared with control, while the equivalent-dose infusion arm did not.63 Similarly, a three-arm study in low-risk women undergoing spontaneous vaginal delivery found no significant between-group differences in either EPDS or VAS outcomes.52 Table 1 summarizes representative positive and negative randomized trials and distinguishes screening-based from diagnostic outcomes.

Table 1.

Representative Randomized Trials of Esketamine for Postpartum Depressive Outcomes

Study	Design and Population	Sample Size	Esketamine Regimen	Outcome Definition and Follow-Up	Main Findings	Main Limitations	Reference
Han et al, 2022	Single-center RCT; women undergoing cesarean delivery with postoperative PCIA	375 randomized; 275 analyzed	S-ketamine 0.5 mg/kg added to sufentanil-based PCIA for 48 h, corresponding to approximately 0.01 mg/kg/h	EPDS-defined screen-positive postpartum depressive symptoms on postpartum days 3, 14, and 28	Screen-positive depressive symptoms were less frequent on day 3 (8.2% vs 17.6%) and day 14 (9.8% vs 24.2%), but not on day 28 (17.2% vs 19.0%); early postoperative pain was also reduced.	Single center; substantial post-randomization exclusion; screening rather than diagnostic outcome; effect was not sustained to day 28.	[44]
Liu et al, 2023	Randomized trial; elective cesarean delivery	150 randomized; 123 analyzed	Intravenous esketamine 0.25 mg/kg during surgery followed by esketamine 1.25 mg/kg in PCIA for 48 h	EPDS-defined screen-positive postpartum depressive symptoms at 3 days, 42 days, 3 months, and 6 months postpartum	No significant reduction in screen-positive depressive symptoms or postpartum anxiety was observed at the assessed time points; cumulative sufentanil consumption was reduced.	Small analyzed sample; loss to follow-up; screening rather than psychiatric diagnosis.	[60]
Ling et al, 2023	Single-center, randomized, double-blind trial; women receiving epidural labor analgesia	120 enrolled	Single intravenous dose of 0.2 mg/kg after delivery	EPDS-defined screen-positive postpartum depressive symptoms at 1 and 6 weeks postpartum	Screen-positive depressive symptoms were less frequent at 1 week (3.4% vs 15.3%) and 6 weeks (5.2% vs 18.6%); exploratory stress- and inflammation-related biomarkers also changed.	Single center; screening outcome; biomarker analyses were exploratory; limited generalizability.	[62]
Chen et al, 2024	Single-center, double-blind, placebo-controlled RCT; elective cesarean delivery	298	Intravenous 0.25 mg/kg immediately after fetal delivery plus 50 mg in PCIA for 48 h	EPDS-defined screen-positive postpartum depressive symptoms on day 7, with additional assessments on days 14, 28, and 42	Screen-positive symptoms were less frequent on day 7 (23.0% vs 35.3%; OR 0.55, 95% CI 0.33–0.91), but not on days 14, 28, or 42.	Predominantly low baseline EPDS scores; composite intraoperative-plus-PCIA regimen; benefit was transient.	[7]
Wang et al, 2024	Five-center, double-blind RCT; women with antenatal EPDS ≥10 undergoing childbirth	364 randomized; 361 included in primary analysis	Single intravenous infusion of 0.2 mg/kg over 40 min after childbirth	Major depressive episode diagnosed by structured interview at 42 days postpartum	Major depressive episodes occurred in 6.7% versus 25.4% of participants (RR 0.26, 95% CI 0.14–0.48); EPDS scores were also lower at 7 and 42 days.	Follow-up limited to 42 days; most participants had mild antenatal symptoms; women with prepregnancy mood disorders were excluded.	[9]
Ren et al, 2025	Single-center randomized pragmatic trial; cesarean delivery	308	Intravenous infusion of 0.25 mg/kg over 20 min during surgery	EPDS-defined screen-positive postpartum depressive symptoms at 6 weeks postpartum	Screen-positive symptoms were less frequent at 6 weeks (10.4% vs 19.5%; RR 0.53, 95% CI 0.30–0.93), but not at 1 week; analgesic-pump activations were reduced.	Screening rather than diagnostic outcome; single center; transient neuropsychiatric events were more frequent immediately after treatment; same institution and overlapping investigator group as Yang et al,63 limiting independence of replication.	[8]
Chu et al, 2026	Three-center, double-blind RCT; primiparae undergoing elective cesarean delivery with antenatal EPDS <10	322 in modified intention-to-treat analysis	Intravenous 0.25 mg/kg followed by 80 mg in PCIA for 24 h	EPDS-defined cumulative screen-positive postpartum depressive symptoms within 3 months postpartum	Cumulative incidence was lower (11.59% vs 20.89%; adjusted RR 0.57, 95% CI 0.35–0.94), driven mainly by the day-7 difference; no significant differences were observed at 1, 2, or 3 months individually.	Composite regimen and cumulative endpoint; benefit was not sustained at individual later assessments; limited to low-risk primiparae.	[64]
Gao et al, 2026	Single-center, three-arm, open-label randomized study; low-risk women undergoing spontaneous vaginal delivery	299 analyzed	Esketamine 0.5 mg/kg plus labor analgesia versus labor analgesia alone or no labor analgesia	EPDS score at 42 days; longitudinal EPDS and VAS assessments	No significant between-group differences in EPDS or VAS outcomes; screen-positive depressive symptoms occurred in 4.0%, 2.0%, and 6.1% of the three groups, respectively.	Open-label and per-protocol design; retrospective registration; very low baseline event rate and probable floor effect.	[52]
Yang et al, 2026	Randomized, double-blind, three-arm superiority trial; cesarean delivery without antenatal depression	503 randomized; 459 analyzed	Intravenous bolus 0.25 mg/kg versus infusion of the same total dose versus saline	EPDS-defined screen-positive postpartum depressive symptoms at 6 weeks postpartum	The bolus arm had fewer screen-positive cases than control (9.15% vs 19.33%; RR 0.47, 95% CI 0.25–0.91). The equivalent-dose infusion arm did not show a significant reduction versus control, and bolus did not differ significantly from infusion (9.15% vs 11.54%). No group differences were observed at 1 or 4 weeks.	Screening endpoint; loss to follow-up; bolus-versus-infusion comparison was not significant; same institution and overlapping investigator group as Ren et al,8 so the two trials should not be regarded as fully independent replication.	[63]

Abbreviations: CI, confidence interval; EPDS, Edinburgh Postnatal Depression Scale; OR, odds ratio; PCIA, patient-controlled intravenous analgesia; RCT, randomized controlled trial; RR, relative risk; VAS, visual analogue scale.

Evidence from Meta-Analyses

Most meta-analyses published from 2024 to 2026 reported lower rates of EPDS-defined screen-positive postpartum depressive symptoms and lower mean EPDS scores after peri-delivery ketamine or esketamine.65–78 However, this apparent consistency should not be interpreted as repeated independent confirmation, because many of these meta-analyses pooled substantially overlapping sets of predominantly single-center trials. They therefore inherit many of the same limitations, including heterogeneity in baseline depressive risk, esketamine dose and route, cointerventions, screening thresholds, and follow-up intervals. Trial sequential analysis has suggested an early reduction in screening-defined depressive outcomes, particularly during the first postpartum week, but it does not establish a durable preventive effect or convert screening outcomes into evidence of clinically diagnosed postpartum depression.77 Network meta-analysis provides exploratory comparisons among regimens, although direct head-to-head evidence remains limited.76 Accordingly, repeated meta-analytic confirmation should not be regarded as multiple independent bodies of evidence. The pooled findings support further targeted trials but do not, by themselves, justify routine prophylactic use. Possible publication bias and inconsistent longer-term findings add further uncertainty.72,76,77 The optimal regimen, target population, and duration of any preventive effect therefore remain unresolved.

Potential Mechanisms

Rapid antidepressant effects of ketamine-class interventions have been demonstrated in non-peripartum populations. In treatment-resistant depression, ketamine produced antidepressant effects within approximately 110 minutes, with benefits extending into the following week.79 These findings provide biological plausibility for a rapid mood-related effect but should not be extrapolated to the timing, magnitude, or durability of postpartum preventive effects. Candidate mechanisms include BDNF-related synaptic plasticity, enhanced AMPAR throughput, and mTOR-dependent synapse formation.20–24 Peripartum clinical studies have also reported changes in inflammatory markers and serum neurotransmitters,80,81 while preclinical studies suggest effects on prenatal-stress-related phenotypes and offspring behavior.24,82 Anxiolytic effects may further contribute to observed emotional outcomes.83,84 However, current clinical studies do not establish whether improvements in mood are direct pharmacological effects or are partly mediated by better analgesia, reduced anxiety, inflammatory changes, or other cointerventions.

Sleep-Related Evidence

Clinical Importance of Postpartum Sleep Disturbance

Objective longitudinal studies demonstrate substantial fragmentation of maternal sleep during the early postpartum months.3 Sleep disturbance during pregnancy is associated with more severe postnatal depressive symptoms,2 while depressive symptoms may also worsen perceived sleep quality. These observations support a close clinical relationship between sleep and mood but do not establish the direction or magnitude of causality.

Potential Effects on Sleep Architecture

Evidence that ketamine-class drugs modify sleep physiology is derived largely from non-obstetric settings. In major depressive disorder, ketamine was associated with increased early-night slow-wave activity alongside BDNF-related changes.26 In a randomized polysomnographic study after gynecological abdominal surgery, S-ketamine increased the proportion of slow-wave sleep, although subjective sleep quality did not improve in parallel.27 These findings suggest a potential effect on sleep architecture but cannot establish that peripartum esketamine directly improves postpartum sleep. Moreover, any apparent sleep-related benefit may be influenced by concurrent changes in pain, mood, sedation, or other perioperative factors. To date, no postpartum polysomnographic study has demonstrated an esketamine-related change in REM sleep, slow-wave sleep, or other objective sleep-stage parameters.

Direct and Indirect Clinical Evidence

Direct obstetric evidence for sleep-related effects remains limited and is based predominantly on subjective sleep measures, often assessed as secondary outcomes. A double-blind trial of esketamine plus dexmedetomidine in PCIA reported improved postoperative sleep-quality scores alongside analgesic benefit,43 while another randomized cesarean-delivery study reported concurrent changes in pain, anxiety, depressive symptoms, sleep, and inflammatory markers.80 Because these outcomes changed together, and because some regimens included additional active cointerventions, these studies cannot establish a direct sleep-specific effect of esketamine. Observational evidence is similarly difficult to interpret: in high-risk parturients, a multimodal analgesic protocol containing esketamine was associated with better subjective sleep,85 but the independent contribution of esketamine cannot be isolated. Indirect evidence from gynecological abdominal surgery, gynecological laparoscopy, and induced abortion suggests that ketamine-class interventions may improve selected postoperative sleep outcomes,27,86,87 but these non-postpartum populations cannot establish efficacy for postpartum sleep disturbance. Overall, the available evidence supports at most a preliminary signal for subjective sleep improvement. Prespecified postpartum sleep outcomes, objective assessment such as polysomnography, and longer follow-up remain lacking. Table 2 summarizes the direct and indirect evidence and their principal limitations.

Table 2.

Clinical Evidence for Analgesic and Sleep-Related Outcomes of Peripartum Esketamine

Clinical Setting or Outcome	Evidence Base	Studied Regimen	Outcome Role	Main Findings	Directness	Key Limitations	Reference
Intraoperative analgesia during cesarean delivery	Randomized and dose-finding studies	Intravenous esketamine approximately 0.15–0.5 mg/kg, alone or with dexmedetomidine; single-bolus or preventive-infusion strategies	Primary or prespecified analgesic/dose-finding outcomes	Several studies reported less visceral traction pain, lower early pain scores, or reduced analgesic requirements.	Direct obstetric evidence	Regimens and cointerventions differed; most studies were single-center and focused on short-term outcomes.	[33–36,38]
Post-cesarean pain and opioid consumption: pooled evidence	Systematic review and meta-analysis of randomized cesarean-delivery trials	Heterogeneous intravenous or epidural esketamine regimens	Pooled analgesic and opioid-consumption outcomes	Rest pain at 24 h was lower with esketamine (SMD −0.42, 95% CI −0.69 to −0.16; P<0.00001; I2=84%), as was movement-evoked pain at 24 h (SMD −0.68, 95% CI −1.29 to −0.07; P<0.00001; I2=92%). In contrast, pooled sufentanil consumption was not significantly reduced (SMD −0.03, 95% CI −0.21 to 0.14; P=0.71; I2=0%).	Direct obstetric evidence	High heterogeneity for pain outcomes; all included trials were conducted in China; regimens and follow-up varied.	[37]
PCIA: mixed surgical meta-analysis	Meta-analysis of adult surgical RCTs	Esketamine plus sufentanil versus sufentanil alone in postoperative PCIA	Pooled postoperative pain and safety outcomes	Pooled VAS scores were lower at several postoperative time points; however, the meta-analysis did not provide a pooled opioid-consumption outcome.	Partly direct/mixed surgical evidence	Included mixed surgical populations rather than an exclusively obstetric population; pooled findings therefore cannot be attributed directly to post-cesarean patients.	[40]
Post-cesarean PCIA	Obstetric randomized trials	Esketamine combined with sufentanil, tramadol, or dexmedetomidine-containing postoperative analgesic regimens	Analgesic, opioid-related, and/or mood outcomes; role varied across trials	Individual trials reported improved early pain control, fewer PCIA activations, or lower opioid exposure in selected regimens; opioid-related findings were not uniform across the broader evidence base.	Direct obstetric evidence	Considerable heterogeneity in background analgesics, esketamine dose, and PCIA settings; combination regimens make the independent contribution of esketamine difficult to isolate.	[41–45]
Intrathecal or epidural analgesia for cesarean delivery	Small RCT and pilot trial	Intrathecal S-ketamine with bupivacaine or epidural esketamine with morphine	Primary or prespecified analgesic outcomes	Prolonged analgesia or improved early pain control was reported.	Direct obstetric evidence	Small samples; older intrathecal evidence; insufficient neurological and long-term maternal or neonatal follow-up.	[46,47]
Epidural labor analgesia: dose exploration	Prospective dose-response and sequential ED90 studies	Epidural esketamine combined with low-concentration ropivacaine	Primary dose-finding outcome	Dose–response relationships and ED90 estimates were reported.	Direct obstetric evidence	Estimates are study-specific and should not be interpreted as standardized clinical dosing recommendations; comparative and neonatal follow-up were limited.	[48–50]
Epidural labor analgesia: active comparison	Randomized noninferiority study	Epidural esketamine or dexmedetomidine compared with fentanyl as ropivacaine adjuncts	Primary local-anesthetic-consumption/noninferiority outcome	Esketamine was noninferior to fentanyl with respect to ropivacaine consumption.	Direct obstetric evidence	The study did not establish superiority and was not designed to determine longer-term maternal, mood, or neonatal benefit.	[51]
Vaginal delivery: pain and depressive symptoms	Three-arm randomized study	Esketamine 0.5 mg/kg combined with labor analgesia	Prespecified pain and EPDS outcomes	No significant between-group improvement in EPDS or VAS outcomes was observed.	Direct obstetric evidence	Open-label design, low-risk cohort, low event rate, per-protocol analysis, and retrospective registration.	[52]
Subjective sleep after cesarean delivery	Randomized obstetric studies using sleep-quality scales	Esketamine plus dexmedetomidine in PCIA or perioperative esketamine within multidomain protocols	Predominantly secondary sleep outcomes	Some improvement in subjective postoperative sleep scores was reported alongside changes in pain, mood, anxiety, or inflammatory markers.	Direct obstetric evidence	Sleep was generally not the primary endpoint; subjective scales, concurrent changes in other outcomes, and cointerventions prevent attribution to a direct sleep-specific effect.	[43,80]
Sleep and depressive symptoms in high-risk parturients	Observational cohort	Multimodal analgesic protocol containing esketamine	Observational sleep and mood outcomes	Better subjective sleep and fewer depressive symptoms were associated with the multimodal protocol.	Direct but nonrandomized obstetric evidence	Confounding by indication and other components of the protocol; the independent effect of esketamine and any causal or mediation relationship cannot be established.	[85]
Objective sleep architecture	Clinical ketamine study and polysomnographic gynecological-surgery RCT	Ketamine or S-ketamine administered outside the postpartum setting	Sleep-related physiological outcomes	Increased early-night slow-wave activity or slow-wave-sleep proportion was reported; subjective improvement did not consistently parallel objective changes.	Indirect non-obstetric evidence	Different populations and clinical contexts; findings cannot establish an effect on postpartum sleep architecture.	[26,27]
Postoperative sleep disturbance outside obstetrics	Randomized gynecological-laparoscopy and induced-abortion studies	Single or intraoperative low-dose esketamine	Prespecified postoperative sleep outcomes	Reduced postoperative sleep disturbance was reported in non-postpartum gynecological populations.	Indirect perioperative evidence	Not postpartum populations; observed effects may reflect analgesia, sedation, or surgical recovery rather than a direct sleep-specific effect.	[86,87]

Abbreviations: CI, confidence interval; ED90, dose effective in 90% of participants; EPDS, Edinburgh Postnatal Depression Scale; PCIA, patient-controlled intravenous analgesia; RCT, randomized controlled trial; SMD, standardized mean difference; VAS, visual analogue scale.

Integrated Interpretation of Pain, Mood, and Sleep

Clinical Relationships Among Pain, Mood, and Sleep

Pain, depressive symptoms, and sleep disruption are associated but remain clinically distinct outcomes. Severe postpartum pain predicts later depressive symptoms, sleep problems during pregnancy are associated with postnatal depressive symptoms, and sleep is markedly fragmented early after childbirth.1–3 These observations justify coordinated assessment but do not establish a bidirectional causal chain or imply that improvement in one domain necessarily causes improvement in another. Established interventions also differ in their targets and time courses. Figure 2 summarizes these clinical relationships and the temporal distribution of the current evidence.

Figure 2.

Timeline of current clinical evidence and major evidence gaps for peripartum esketamine. (A) summarizes antenatal assessment and candidate clinical factors under investigation, which should not be interpreted as validated criteria for selecting patients. (B) illustrates trial-derived esketamine regimens used around labor or cesarean delivery, including intravenous bolus or infusion, neuraxial adjuncts, and postoperative patient-controlled intravenous analgesia. These regimens do not constitute standardized dosing recommendations. (C) summarizes clinical outcomes studied during the early postpartum period (0–72 h), including pain, mood, subjective sleep, and short-term maternal and neonatal safety. (D) summarizes follow-up during postpartum weeks 1–6, including depressive symptoms, limited diagnostic mood outcomes, subjective sleep, maternal recovery, and neonatal observation. (E) highlights major evidence gaps beyond 6 weeks, including the durability of mood-related benefits, objective postpartum sleep architecture, long-term maternal outcomes, and offspring neurodevelopmental safety. Direct clinical evidence is concentrated around labor or cesarean delivery and the early postpartum period, with the most consistent findings relating to short-term analgesic outcomes. Mood-related studies have predominantly used EPDS-defined screening outcomes, whereas structured diagnostic outcomes remain limited. Sleep-related evidence is mainly subjective, and objective postpartum sleep architecture has not been directly evaluated. Italicized text denotes interpretive evidence-status statements, qualifications, or cautions rather than additional measured clinical outcomes. Horizontal dotted or dashed separator lines distinguish these interpretive notes from the surrounding information; they do not represent dashed evidence boxes. The timeline reflects the approximate timing and distribution of the available evidence.

Abbreviations: EPDS, Edinburgh Postnatal Depression Scale; PCIA, patient-controlled intravenous analgesia.

Mechanistic Overlap and Cross-Domain Interpretation

NMDAR-dependent signaling contributes to central sensitization, cortical plasticity, and sleep-related slow-wave activity, providing a biological rationale for considering pain, mood, and sleep within a common framework.19,26 Clinically, however, the strength and directness of evidence differ substantially across these domains. Short-term analgesic benefit is the most consistently supported outcome, whereas opioid-sparing effects are less reproducible.37 Mood-related evidence includes reductions in EPDS-defined screening outcomes in several randomized trials and overlapping meta-analyses, together with more limited evidence based on structured diagnostic outcomes.9,65–78 Sleep-related findings remain preliminary and rely largely on subjective, often secondary outcomes.43,80,85 Because changes in pain, mood, and sleep frequently occur concurrently and may also be influenced by cointerventions, current studies cannot determine whether improvements in mood or sleep represent direct effects of esketamine or are partly mediated by better analgesia, reduced anxiety, sedation, or other perioperative factors. Formal mediation analyses and studies designed to establish temporal sequence are largely lacking. The value of an integrated framework is therefore primarily methodological: it encourages prespecified multidomain outcomes and mediation analyses without implying that esketamine independently treats all three problems. Table 3 compares the consistency, directness, and clinical interpretability of evidence across these domains.

Table 3.

Cross-Domain Summary of Evidence for Peripartum Analgesic, Mood-Related, and Sleep-Related Outcomes

Domain and Clinical Question	Principal Evidence Base	Consistency	Directness	Current Interpretation	References
Does esketamine improve short-term post-cesarean analgesia?	RCTs and evidence syntheses	Generally favorable	Predominantly direct	Short-term analgesic benefit is reasonably supported, although effect magnitude varies with regimen, background analgesia, and study design.	[33–47]
Does esketamine reduce post-cesarean opioid exposure?	Individual RCTs and evidence syntheses	Mixed	Predominantly direct	Some individual studies reported lower opioid use or fewer PCIA activations, but opioid-sparing effects are not consistently reproduced across pooled evidence.	[37,41–45]
Is epidural esketamine an effective adjunct for labor analgesia?	Dose-finding and noninferiority studies	Favorable but limited	Direct	Supports neuraxial adjunct activity and local-anesthetic sparing, but does not establish a standardized dose, preferred regimen, or comparator superiority.	[48–51]
Does esketamine reduce EPDS-defined screen-positive postpartum depressive symptoms?	RCTs and substantially overlapping meta-analyses	Generally favorable but heterogeneous	Direct	Several studies suggest an early reduction in screening-defined depressive outcomes, but not all trials are positive, durability is inconsistent, and repeated meta-analyses do not represent independent confirmation.	[7,8,44,45,52,58–78]
Does esketamine prevent clinically diagnosed postpartum depressive episodes?	Limited multicenter randomized evidence	Promising but incompletely replicated	Direct but sparse	One multicenter trial using a structured diagnostic interview showed benefit at 42 days in women with antenatal depressive symptoms; external replication is required before broader inference.	[9]
Does esketamine improve subjective postpartum sleep?	Few obstetric studies using mainly questionnaire-based secondary outcomes	Preliminary	Direct but limited	A possible subjective sleep benefit has been reported, but it cannot be reliably separated from changes in pain, mood, sedation, or cointerventions.	[43,80,85]
Does esketamine improve objective postpartum sleep architecture?	No direct postpartum polysomnographic study	Not established	Indirect	Current evidence does not establish effects on REM sleep, slow-wave sleep, or other objective postpartum sleep-stage parameters.	[26,27,86,87]
What is known about short-term maternal and neonatal tolerability?	RCTs, cohorts, and evidence syntheses	Generally reassuring for studied regimens	Direct for short-term outcomes	Studied low-dose regimens generally show acceptable short-term tolerability, but this should not be interpreted as established long-term maternal or offspring safety.	[7–9,29,39,63,64,75,76]

Abbreviations: PPD, postpartum depression; RCT, randomized controlled trial.

Safety and Clinical Considerations

Overview of Maternal and Infant Safety

Across randomized trials and evidence syntheses, studied low-dose esketamine regimens were most commonly associated with transient dizziness, perceptual or dissociative symptoms, and short-lived hemodynamic changes; reported rates varied according to dose, route, and ascertainment.75,76 Selected obstetric trials and cohorts have not identified clear worsening of short-term neonatal outcomes, including Apgar scores and other early clinical measures.29,39,40 These findings are reassuring regarding short-term tolerability in selected populations but should not be interpreted as establishing maternal or neonatal safety, because most studies were not powered for uncommon adverse events and did not assess long-term neurodevelopment. Ketamine crosses the placenta rapidly,28 and maternal recovery therefore cannot be used as a surrogate for limited fetal exposure. Preclinical studies of NMDA-receptor antagonism have raised developmental neurotoxicity concerns, including apoptotic neurodegeneration during susceptible developmental windows;88 prolonged ketamine exposure has also produced neuronal cell death in fetal or neonatal nonhuman-primate models.89 However, the doses and exposure durations used in these experiments differ substantially from the single or short-course low-dose regimens evaluated in most obstetric trials, and their relevance to human peripartum exposure remains uncertain. Clinically relevant S-ketamine doses did not significantly reduce uterine blood flow in pregnant sheep,30 but this remains indirect evidence. Small human lactation studies found low and rapidly declining breast-milk concentrations,31,32 with estimated relative infant doses below 1%. These data are insufficient to define a universal fixed interruption interval for breastfeeding; decisions should instead consider maternal recovery, infant factors, the clinical context, and applicable local guidance.

Contraindications and Precautions

Because esketamine can transiently increase blood pressure and produce perceptual or dissociative symptoms, careful assessment is warranted in patients with cardiovascular or cerebrovascular conditions in which transient blood-pressure elevation could pose substantial risk. A history of psychosis or increased risk of substance misuse also merits individualized assessment.10 Severe pre-eclampsia has been underrepresented or excluded in many trials; this lack of evidence is neither proof of safety nor sufficient grounds for a universal absolute contraindication. Animal studies have reported behavioral changes in offspring after gestational S-ketamine exposure,90 while broader preclinical studies raise developmental-neurotoxicity concerns with prolonged or developmentally timed NMDA-receptor antagonist exposure.88,89 Human evidence for early-pregnancy exposure remains limited to case-level observation.91 These signals warrant explicit communication of uncertainty and long-term offspring follow-up in future studies.

Clinical Interpretation and Research Priorities

The heterogeneity of studied regimens precludes recommendation of a standardized peripartum dose. Several positive trials used initial intravenous doses in the range of 0.2–0.25 mg/kg, including 0.25 mg/kg administered intraoperatively, either alone or followed by a postoperative PCIA component,7,64 and a single 0.2 mg/kg postpartum infusion in women with antenatal depressive symptoms.9 However, current evidence does not establish the 0.2–0.25 mg/kg range as an optimal or standardized regimen, because route, timing, cointerventions, baseline risk, and outcome definitions varied substantially; other studies used different infusion or pump-based strategies.63 Epidural dose-response and ED90 studies in labor provide study-specific research estimates rather than clinical dosing standards.48–50 Findings in women with gestational diabetes,92 preoperative anxiety,83 or category-1 cesarean delivery,93 as well as findings from esketamine–dexmedetomidine combination regimens,94 should not be generalized to all obstetric populations. Routine peripartum use would require adequately powered multicenter trials with standardized regimen reporting, consistent adverse-event assessment, longer maternal and infant follow-up, and outcome measures appropriate to each clinical domain. Exploratory machine-learning models have attempted to predict response to subanesthetic ketamine or esketamine,95 but none is ready for clinical use without external validation. Pregnancy-specific pharmacokinetic and pharmacogenetic evidence also remains limited. A human study found no significant effect of CYP2B6 genotype on ketamine metabolism or clearance,96 and genotype-guided dosing therefore remains investigational. Table 4 summarizes maternal and neonatal safety considerations, major evidence gaps, and related research priorities.

Table 4.

Safety Considerations, Evidence Gaps, and Research Priorities for Peripartum Esketamine

Safety Issue or Population	Current Evidence	Evidence Source	Principal Uncertainty	Clinical Implication	Research Priority	Reference
Transient neuropsychiatric effects	Dizziness, perceptual changes, dissociation, hallucinations, or dream-like experiences have been reported more frequently after esketamine; most reported events were brief and self-limited.	Obstetric RCTs and evidence syntheses	Definitions and ascertainment varied; patient acceptability and functional impact were inconsistently assessed.	Short-term neuropsychiatric effects require observation during and immediately after administration and should be incorporated into benefit–risk assessment.	Standardized adverse-event assessment and direct comparison of bolus, infusion, and PCIA regimens.	[7–9,63,64,75–78]
Maternal hemodynamics and hypertensive disorders	Transient increases in blood pressure or heart rate may occur, although major instability was uncommon in selected trial populations.	RCTs and evidence syntheses	Women with severe pre-eclampsia, uncontrolled hypertension, or major cardiovascular disease were commonly excluded or underrepresented.	Short-term tolerability findings in selected populations cannot be generalized to severe hypertensive disorders of pregnancy.	Dedicated prospective studies or registries in hypertensive and other high-risk obstetric populations.	[7–9,63,64,75,76]
Placental transfer and fetal exposure	Ketamine crosses the placenta rapidly after intravenous administration; animal data suggest that clinically relevant S-ketamine doses may not substantially reduce uterine blood flow.	Human placental-transfer study and animal experiment	Esketamine-specific maternal–fetal pharmacokinetic data are sparse, and animal uterine-perfusion findings may not translate directly to human pregnancy.	Maternal recovery should not be used as a surrogate for negligible fetal exposure; animal findings provide mechanistic context rather than proof of fetal safety.	Paired maternal, cord-blood, and neonatal pharmacokinetic studies.	[28,30]
Short-term neonatal outcomes	Studied regimens have not consistently worsened Apgar scores, neonatal neurobehavioral measures, or NICU admission.	Obstetric RCTs and cohorts	Most studies were not powered for uncommon neonatal events and assessed only early outcomes.	Findings are reassuring for short-term tolerability but should not be interpreted as established developmental safety.	Standardized neonatal endpoints and larger multicenter safety cohorts.	[29,39,40]
Lactation exposure	Ketamine and its metabolites were detected at low and rapidly declining concentrations in breast milk; estimated relative infant doses were below 1%.	Small human pharmacokinetic studies	Few participants; limited infant plasma and clinical outcome data; findings may not apply to all esketamine regimens.	Current evidence is insufficient to define a universal fixed breastfeeding-interruption interval.	Larger lactation pharmacokinetic studies incorporating infant exposure and clinical outcomes.	[31,32]
Developmental and early-pregnancy safety	Human early-pregnancy evidence is limited to case-level observation, while animal and broader preclinical studies raise potential neurobehavioral and developmental-neurotoxicity concerns.	Case report and preclinical studies	Species, dose, exposure duration, and developmental stage differ substantially from typical low-dose peripartum regimens; long-term human offspring data are lacking.	Preclinical findings should be regarded as signals for surveillance rather than estimates of human risk; neither safety nor harm can be inferred from isolated human cases.	Pregnancy exposure registries and long-term offspring neurodevelopmental follow-up.	[88–91]
Special obstetric populations	Exploratory studies have included women with gestational diabetes, preoperative anxiety, and category-1 cesarean delivery.	Small RCTs, cohorts, or case series	Limited sample size, selection bias, and lack of replication.	Findings should not be generalized to all high-risk pregnancies.	Prespecified studies in clinically defined high-risk subgroups.	[83,92,93]
Combination regimens	Esketamine has frequently been combined with opioids, dexmedetomidine, or other analgesic agents.	RCTs	Independent efficacy and adverse effects attributable to esketamine are difficult to isolate.	Outcomes from combination regimens should not be attributed exclusively to esketamine.	Factorial or head-to-head comparative-effectiveness trials.	[33,41–43,94]

Abbreviations: NICU, neonatal intensive care unit; PCIA, patient-controlled intravenous analgesia; RCT, randomized controlled trial.

Limitations of This Review

This review has several limitations. As a narrative review, it did not follow a formal systematic-review protocol or include structured study-level risk-of-bias or certainty assessments, which may influence the interpretation of the evidence. Although PubMed and the Web of Science Core Collection were searched and reference lists were screened, relevant studies indexed elsewhere or reported in the grey literature may have been missed. The direct peripartum evidence is also geographically and methodologically concentrated: most clinical studies were conducted in Chinese populations, and many were single-center trials, which may limit generalizability and increase the influence of center- or investigator-specific effects. In addition, several published meta-analyses include substantially overlapping sets of primary trials, and some individual trials originate from overlapping institutions and investigator groups; these publications should therefore not be interpreted as fully independent bodies of evidence. The absence of commercial funding does not eliminate the possibility of publication, investigator, or academic bias, particularly in an evidence base dominated by relatively small positive trials. Evidence concerning sleep architecture, pharmacokinetics, lactation, and long-term offspring safety also relies partly on non-obstetric, racemic-ketamine, or preclinical data and should therefore be interpreted cautiously. Finally, substantial heterogeneity in patient populations, esketamine dose, route, timing, cointerventions, outcome definitions, and follow-up limits direct comparison across studies and precludes recommendation of a standardized peripartum regimen or firm criteria for candidate selection. The conclusions should therefore be viewed as a clinically oriented synthesis rather than definitive practice guidance.

Conclusions

Current evidence supports a cautiously promising but still investigational role for peripartum esketamine. Evidence is most consistent for short-term analgesic benefit, whereas opioid-sparing effects remain inconsistent. Several randomized trials suggest that peri-delivery esketamine may reduce early EPDS-defined screen-positive postpartum depressive symptoms, while evidence based on structured diagnostic outcomes remains limited and requires further replication. Sleep-related evidence is preliminary and relies predominantly on subjective measures, and long-term maternal, lactational, and offspring neurodevelopmental safety remains insufficiently characterized. Current evidence does not support routine prophylactic use or a standardized peripartum regimen. Future research should prioritize adequately powered multicenter trials, diagnostic mood outcomes, objective postpartum sleep assessment, standardized treatment protocols, and long-term maternal and infant follow-up. Until such evidence is available, esketamine should be considered an individualized adjunct within multimodal peripartum care rather than an established prophylactic intervention.

Funding Statement

No funding was received for this study.

Abbreviations

AKT, protein kinase B; AMPAR, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor; BDNF, brain-derived neurotrophic factor; CI, confidence interval; CYP2B6, cytochrome P450 2B6; ED90, effective dose in 90% of participants; EPDS, Edinburgh Postnatal Depression Scale; GABA, gamma-aminobutyric acid; mTOR, mammalian target of rapamycin; NICU, neonatal intensive care unit; NMDAR, N-methyl-D-aspartate receptor; OR, odds ratio; PCIA, patient-controlled intravenous analgesia; PPD, postpartum depression; RCT, randomized controlled trial; REM, rapid eye movement; RR, relative risk; VAS, visual analogue scale.

Data Sharing Statement

No new datasets were generated or analyzed for this narrative review. The article draws exclusively on the published literature cited in the manuscript.

Ethics Approval and Consent to Participate

Ethics approval and informed consent were not required because this narrative review used only previously published literature and involved no new human participants, animals, or identifiable personal data.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no competing interests.