Translational pathways of oxytocin therapy in schizophrenia: bridging negative symptom domains and neural mechanisms Authors: Lin Ji, Xiao Wang, Yue Li, Wentao Zhao, Xinzhe Du, Yao Gao, Junxia Li, Sha Liu Published: 2026 ABSTRACT Negative symptoms of schizophrenia, characterized by motivational deficits, diminished hedonic experience, and impaired affective expression, remain a major therapeutic challenge. Oxytocin, a neuropeptide with central nervous system activity, has emerged as a promising therapeutic target due to the close alignment of its functions with the pathophysiological mechanisms underlying negative symptoms. Based in the dimensional model of negative symptoms, this review systematically synthesizes advances in understanding the physiological functions of the oxytocin system and its therapeutic potential for negative symptoms in schizophrenia, with a focus on its multi-target neural mechanisms. Oxytocin exerts its effects through coordinated modulation of multiple neural circuits: enhancing mesolimbic dopamine pathway function to improve social reward processing and facilitating motivation via the prefrontal-striatal circuit, while stabilizing the amygdala-prefrontal emotional circuit and modulating the sensorimotor integration network and prefrontal motor pathway to ameliorate affective expression deficits. Current evidence highlights the therapeutic potential of oxytocin; however, clinical findings demonstrate marked heterogeneity, suggesting that its translation still requires overcoming multiple challenges. Future research should integrate genetic, neuroimaging, and behavioral paradigms to identify oxytocin-sensitive subtypes, optimize administration strategies, and coordinate behavioral interventions, thereby advancing the translation of oxytocin toward precision psychiatry in schizophrenia. Introduction Schizophrenia is a severe psychiatric disorder exhibiting considerable clinical heterogeneity. Affecting approximately 0.7% of the global population over a lifetime, it presents a considerable challenge to healthcare systems worldwide [1]. The etiology of schizophrenia arises from an intricate interplay between genetic predispositions and environmental influences. A core feature of schizophrenia is a disconnection from reality, manifesting primarily as positive symptoms, negative symptoms, and cognitive deficits. While current antipsychotic medications demonstrate considerable efficacy against positive symptoms, their effects on negative symptoms and cognitive impairment remain modest [2]. Nonetheless, negative symptoms represent a primary determinant of severe social dysfunction and unfavorable long-term prognosis, establishing them as a central therapeutic challenge and a critical priority in ongoing research. The negative symptoms of schizophrenia constitute a multidimensional construct. As outlined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) and related consensus statements from the National Institute of Mental Health (NIMH), these symptoms encompass five specific manifestations: blunted affect, alogia, avolition, anhedonia, and asociality. Brian Kirkpatrick and colleagues introduced a two-dimensional theoretical model that groups these manifestations into two primary domains: “Motivation and Pleasure Deficit (MAP)” and “Emotional Expressivity Reduction (EXP).” This conceptual framework has been incorporated into contemporary assessment instruments such as the Brief Negative Symptom Scale (BNSS) and the Clinical Assessment Interview for Negative Symptoms (CAINS). Subsequent research has demonstrated that a hierarchical model based on this two-dimensional framework provides a superior fit to the data [3], resulting in a closer alignment between theoretical constructs and empirical findings. This advancement has consequently enhanced the accuracy of both the assessment of negative symptoms and the investigation of their underlying mechanisms. Current therapeutics for negative symptoms remain limited. While neuromodulation techniques such as intermittent theta-burst stimulation (iTBS) [4] and high-frequency repetitive transcranial magnetic stimulation (rTMS) [5, 6] have shown promising therapeutic potential, their widespread application is constrained by limitations including high costs, extended treatment duration, and substantial interindividual variability in treatment response. These factors collectively impose significant economic and psychological burdens on patients and their families. Consequently, there is a significant need to develop novel pharmacological interventions that are more accessible and cost-effective, representing an urgent research priority aimed at ameliorating negative symptoms and enhancing treatment accessibility. Accumulating evidence increasingly implicates the oxytocin (OT) system in the pathophysiological mechanisms underlying schizophrenia [7, 8]. OT levels in both cerebrospinal fluid and plasma are consistently inversely correlated with negative symptom severity [9, 10]. Additional associations have been reported with asociality [11], cognitive empathy deficits [12], and facial emotion recognition impairments [13]. These observations have positioned OT as a potential intervention for negative symptoms in schizophrenia. Exogenous OT has been administered via two primary routes: peripheral injection (e.g., intravenous or intraperitoneal) and intranasal delivery. Intravenous injection rapidly increases peripheral plasma levels but yields negligible direct brain parenchyma distribution except at very high doses. Intranasal administration, however, bypasses the blood-brain barrier and delivers OT directly to specific brain regions—including the olfactory bulb, striatum, and brainstem—via olfactory and trigeminal pathways in animal models [14]. In clinical research, intranasal OT can reach the central nervous system directly via the nose-to-brain pathway and indirectly via entry into the peripheral circulation [15–18]. Preliminary studies have shown that intranasal OT can increase facial expressivity in individuals with schizophrenia [19] and enhance motivation to exert effort for social interaction by increasing social reward salience [20]. Systematic reviews and meta-analyses have failed to consistently confirm the efficacy of intranasal OT for negative symptoms in schizophrenia [21], and a dose-response meta-analysis similarly reported no significant overall effect [22]. A meta-analysis by Peled-Avron et al. found that oxytocin exerted a modest but stable improvement on overall psychopathology in schizophrenia, an effect likely mediated by its influence on negative symptoms, with no significant direct impact on positive symptoms or overall psychopathology beyond this pathway [23]. In contrast, more recent randomized controlled trial have reported that a single dose of intranasal OT did not significantly improve global neurocognitive function in schizophrenia [24], and another trial found that adding OT to social skills training did not significantly enhance social functioning or negative symptoms [25]. Notably, similarly heterogeneous findings have been observed in autistic populations [26–28]. This discrepancy likely stems from the complexity, pleiotropic nature, and individual heterogeneity of OT’s central mechanisms, which contribute to the heterogeneous and attenuated clinical outcomes observed in trials [29]. To clarify the mechanisms by which OT modulates negative symptoms in schizophrenia, this review adopts a dimensional model that distinguishes between two domains: motivational–pleasure deficits (anhedonia, amotivation, and asociality) and emotional expression deficits (blunted affect and alogia). We systematically synthesize findings from animal studies, clinical efficacy trials, and neurobiological investigations to examine the therapeutic mechanisms of oxytocin intervention. The overarching aim is to provide a theoretical framework for interpreting the heterogeneity of existing findings and identifying potential targets for precision-based intervention, thereby informing future research and clinical practice to advance the translational application of oxytocin in schizophrenia and other psychiatric disorders. Oxytocin functions OT, a nine-amino-acid neuropeptide, is evolutionarily conserved and mainly synthesized by neurons in the paraventricular nucleus (PVN), supraoptic nucleus (SON), and accessory nucleus (AN) of the hypothalamus (Fig. 1). While historically categorized as magnocellular and parvocellular subtypes, recent findings indicate a more extensive cellular and functional diversity than previously recognized [30]. OT demonstrates a distinct functional dualism. It functions as a hormone when released into the systemic circulation from the neurohypophysis, regulating reproductive processes like parturition and lactation in peripheral target organs. Simultaneously, through central projections, it exerts broad influences on limbic system regions (e.g., amygdala, nucleus accumbens) and brainstem areas, acting as a neuromodulator that regulates complex brain functions, including social behavior [31], motivation-reward processing [32], feeding behavior [33], as well as emotion and stress response [34]. The subsequent sections will examine the role of OT in the regulation of motivation-pleasure and emotional expression, elucidating its underlying neural mechanisms.Fig. 1The Synthesis and Distribution of Oxytocin (Schematic).Endogenous oxytocin is primarily synthesized in the supraoptic nucleus (SON) and paraventricular nucleus (PVN) of the hypothalamus. The synthesized oxytocin is stored in the posterior pituitary for peripheral release and is also released centrally via axonal projections to modulate multiple brain regions, including the prefrontal cortex (PFC), amygdala, anterior cingulate cortex (ACC), nucleus accumbens (NAc), striatum, hippocampus, and ventral tegmental area (VTA). The Synthesis and Distribution of Oxytocin (Schematic). Endogenous oxytocin is primarily synthesized in the supraoptic nucleus (SON) and paraventricular nucleus (PVN) of the hypothalamus. The synthesized oxytocin is stored in the posterior pituitary for peripheral release and is also released centrally via axonal projections to modulate multiple brain regions, including the prefrontal cortex (PFC), amygdala, anterior cingulate cortex (ACC), nucleus accumbens (NAc), striatum, hippocampus, and ventral tegmental area (VTA). Oxytocin and motivation-pleasure deficits Motivation-pleasure deficits primarily manifest as diminished internal drive for goal-directed behavior and reduced capacity to experience pleasure from rewarding stimuli. Research indicates that OT directly modulates the function of the mesolimbic dopamine reward system by acting on oxytocin receptors (OTR) in specific brain regions, thereby influencing social motivation and affective processes [35]. Multiple lines of experimental evidence support this view. Animal studies demonstrate that the endogenous OT system plays a crucial role in maintaining normal social motivation [36, 37]. Oxytocin receptor knockout (OTR-KO) mice exhibit significant social memory deficits and social avoidance behaviors; Pharmacological studies indicate that blocking endogenous OT system function with OTR antagonists significantly reduces social motivation in normal rats. In schizophrenia animal models (e.g., subchronic phencyclidine treatment), blocking this system further exacerbates inherent social deficits—both findings demonstrate the critical importance of OT system integrity for social motivation [38]. However, under environmentally enriched (EE) housing conditions, the opposite phenomenon occurs: the same OTR antagonist treatment actually increases social interaction time in normal mice. This suggests that EE may induce functional remodeling of the OT system, shifting its role from “promoting social behavior” to “inhibiting social behavior” [39]. Intervention studies with exogenous OT further confirm its regulatory role. Microinjection of OT into the ventral tegmental area (VTA) induces heightened social motivation in normal mice [40]. Furthermore, short-term, low-dose intraperitoneal administration of OT during a critical developmental window persistently rescues social behavior deficits in male Cntnap2 knockout mice [41]. Notably, OTR in the caudate lateral septum (cLS) represent a key neural substrate regulating “approach-avoidance” decision-making in social contexts. Microinjection studies in this region reveal that activating cLS OTR induces social avoidance and impairs long-term social recognition memory, whereas blocking these receptors promotes social approach behavior [42]. This indicates that the functional state of cLS OTR directly modulates the orientation of social motivation, and its over-activation can result in anhedonia and amotivation. Oxytocin and emotional expressivity deficits Expressivity deficits are a transdiagnostic clinical phenomenon characterized by significant impairments in the externalization and expression of emotional experience. It is commonly observed across various psychiatric disorders, including depression, autism spectrum disorder, post-traumatic stress disorder, and schizophrenia [43]. The OT system plays a critical regulatory role in this process. Animal studies provide mechanistic insights: acute maternal separation activates the OT system in young pups, driving them to actively seek maternal contact accompanied by increased ultrasonic vocalization (USV). However, when OT signaling is disrupted (e.g., in OTR-KO models or via OTR antagonist administration), pups fail to effectively locate and suckle, while dams exhibit prolonged time away from the nest and reduced active nursing behaviors (e.g., licking, arching-back suckling) within the nest [44]. Similarly, in adult male and female mice, inhibition of OT signaling also delays the latency to approach conspecifics and reduces licking behavior during interactions [45]. Further research demonstrates that exogenous OT intervention effectively ameliorates deficits in affective expression. A single intranasal dose of OT rapidly enhanced maternal communication behavior in primiparous California mouse dams. During pup separation, these mothers emitted significantly more guiding USV, facilitating long-distance communication without compromising direct pup care behaviors such as licking and suckling [46]. Moreover, maternal deprivation induces affective blunting and anxiety-depression-like behaviors, paralleling a reduction in OTR expression in the medial habenula (MHb). Conversely, EE intervention upregulates MHb OTR expression and concurrently rescues these behavioral deficits [47]. Efficacy of oxytocin on negative symptoms in schizophrenia Multiple studies have linked dysregulation of the endogenous OT system to the pathogenesis of negative symptoms in schizophrenia [48]. As an adjunctive intervention, intranasal administration of exogenous OT has been extensively investigated for its efficacy in alleviating negative symptoms. However, findings regarding its effects on overall negative symptom severity have been inconsistent [22]. When examined at the dimensional level, studies have reported relatively consistent therapeutic effects of OT on improving affective expression and social motivation [49–51]. Therefore, this section focuses on reviewing the therapeutic effects of OT on two dimensions of negative symptoms, namely motivational and anhedonic deficits and affective expression deficits, which are summarized in Table 1.Table 1Oxytocin Intervention for Negative Symptoms in Schizophrenia.ActhorParticipants N (M/F)1Dosing and DurationSymptom measuresMain Findings (with dimensional annotation)Effect size (Cohen’s d)2Evidence type31Zierhut et al. [50, 51]SSD 41 (31/10)OT (24 IU qd × 1wk) + PLC + MBGT (twice weekly)PANSS, the Self -Evaluation of Negative Symptoms (SNS); Plasma oxytocin levelsOT + MBGT shows superior improvement in negative symptoms, particularly avolition, social withdrawal and blunted affect. [MAP&EXP]Avolition: d = 0.70b Social withdrawal: d = 0.50bDiminished emotional range: d = 0.70bCore + Synergistic2Ota M et al. [49]SCZ 16 (7/9)OT (12 IU bid × 12wk)PANSS, Brief Assessment of Cognition in Schizophrenia (BACS); MRIOT improve negative symptoms and verbal fluency via right insula and left cingulate cortex (negative correlation with gray matter volume). [MAP&EXP]PANSS-N: d = 1.05bMAP: d = 0.66bEXP: d = 0.91bBACS: d = 0.53Core + Supportive3Jarskog et al. [61]SCZ 62 (47/15)aOT (24 IU bid × 12wk) + PLCPANSS, Social Skills Performance Assessment (SSPA)OT ineffective for social functioning but significantly reduced PANSS negative symptoms. [MAP]NRCore4Halverson et al. [62]SCZ 62 (47/15)aOT (24 IU bid × 12wk) + PLCPANSS, Specific Levels of Functioning Scale (SLOF); Interpersonal Reactivity Index (IRI)OT significantly improves cognitive empathy (IRI perspective-taking). [MAP]NRCore + Supportive5Fulford et al. [20]SCZ 42 (32/10), HC 43 (29/14)OT (40 IU × 2, ≥1wk interval) + PLCPANSS; The Awareness of Social Inference Test (TASIT), Social Vigor Task (SVT)Passive/apathetic social withdrawal associated with lower effort; OT shows a trend toward improving this effect. [MAP]NRCore + Supportive6Mouchlianitis et al. [54]SCZ 20 (20/0)OT (40 IU, single dose) + PLCPANSS; Trust Game; fMRIOT coordinates reward processing networks during social learning tasks, promoting cortical processing in investment trials (implicit/contextual rewards) and limbic processing in repayment trials (explicit rewards). [MAP]NRSupportive7Bradley et al. [58]SCZ 39 (39/0), HC 54 (54/0)OT (40 IU, single dose) + PLCPANSS; Auction GameOT improves social reward motivation. [MAP]NRSupportive8Caravaggio et al. [59]SCZ 43 (43/0), HC 16 (16/0)OT (50 IU, single dose, ≥20 d interval) + PLCSAPS, SANS, Social Functioning Scale (SFS); Bead Task; MRISCZ tendency to jump to conclusions (JTC) is positively correlated with ventral caudate volume; OT ineffective for JTC bias. [MAP]NRSupportive9Strauss et al. [64]SCZ 62 (38/24)OT (36 IU bid × 24wk) + PLC + group CBSST (twice weekly)Facial Emotion Recognition Test (FERT), Empathic Accuracy Task (EAT); Eye-trackingOT + CBSST ineffective for social function; enhanced gaze to happy faces (exploratory). [MAP]FERT-total: d = 0.37EAT-positive: d = 0.46Core + Supportive + Synergistic10Buchanan et al. [63]SCZ 62 (38/24)OT (36 IU bid × 24wk) + PLC + group CBSST (twice weekly)BPRS, SANS, The Birchwood Social Functioning Scale (BSFS)OXT+ CBSST ineffective for social functioning and negative symptoms, improvements driven by site effects.[MAP]NRRestrictive + Synergistic + Moderating11Abram SV et al. [111]SCZ 22 (22/0), HC 22(22/0);OT (40 IU, single dose) + PLCPANSS, SANS, fMRIOT enhances amygdala- left temporal connectivity in SCZ; baseline hypoconnectivity in this circuit specifically correlates with expressive negative symptoms. [EXP]Amygdala-left temporal connectivity: d = 1.00SANS: d = 1.32Core + Supportive12Woolley et al. [19]SCZ 33 (30/3) HC 35 (33/2)OT (40 IU × 2, ≥1wk interval) + PLCPANSS, Social Judgment Task (SJT), Facial Expression CodingOT effective for facial emotion expression. [EXP]NRCore + Supportive13Gibson et al. [67]SCZ 14 (11/3)OT (24 IU bid × 6wk) + PLCPANSS, Role-Playing AssessmentOxytocin effective for negative symptoms, nonverbal social skills (fear recognition), and specific social skills (self-reported perspective-taking ability). [EXP]PANSS-N: d = 1.77Specific: d = 0.26Nonverbal: d = 0.23Core + Supportive14Dwyer et al. [68]SCZ 40 (33/7)aOT (24 IU bid × 6wk) + PLCSocial role-playing tasks; computer-based behavioral coding and verbal content analysisinteractionsOT effective for total verbal output, characterized by increased use of positive emotion and social words. [EXP]Word count: d = 0.35Positive Emotion word: d = 0.22social word: d = 0.29Supportive15Cohen et al. [65]SCZ 40 (33/7)OT (24 IU bid × 6wk) + PLCSocial Competence Assessment Scale, computerized natural speech and facial expression measurementOT effective for negative facial expressions, ineffective for speech expression in SCZ. [EXP]Facial-negative: d = 0.51Vocal-pause times: d = 0.28Supportive16Cacciotti-Saija C et al. [125]SCZ 52 (36/16)OT (24 IU bid × 6wk) + PLC + group SCT (twice weekly)SAPS, SANS, Social Functioning Scale (SFS)OT may improve negative symptoms; greater OT use correlates with greater negative symptom reduction.SAPS: d = 0.27SANS: d = 0.31SFS: d = 0.04Core17Feifel D et al. [126]SCZ 19 (15/4)aOT (40 IU bid × 3wk) + PLCPANSS, CGIOT well- tolerated with antipsychotic properties; medium effect sizes; higher doses/longer treatment may yield greater benefit.PANSS: d = 0.43PANSS-N: d = 0.5CGI-I: d = 0.74Core18Korann V et al. [95]SCZ 31 (31/0) HC 21 (21/0)OT (24 IU, single dose) + PLCPANSS, SANS; fMRIOT- induced connectivity gains correlate with higher cognitive insight and lower negative symptoms. OT enhances effective connectivity from left caudate to left SMA, precentral gyrus, and IFG pars triangularis.NRCore + Supportive19Modabbernia A et al. [127]SCZ 40 (33/7)OT (20 IU bid wk1, then 40 IU bid wks2–8) + PLCPANSSOT add- on to risperidone effectively reduces positive symptoms and is well- tolerated.PANSS: d = 1.9PANSS-N: d = 1.4Core + Moderating20Lee MR et al. [53, 128]SCZ 28 (20/8)OT (20 IU bid × 3wk) + PLCPANSS, SANSOT ineffective for total, positive, and negative symptoms overall; improves negative symptoms in hospitalized subgroup [exploratory].SANS: d = 0.01;hospitalized:SANS: d = 0.85,SANS-EXP: d = 0.74Moderating21Buchanan RW et al. [129]SCZ 56 (45/11)OT (24 IU bid × 12wk) + PLCSANSOT ineffective for SANS total score; no benefit in subgroup analyses.SANS: d = 0.10Limiting22Dagani J et al. [130]SCZ 32 (26/6)OT (40 IU qd × 8mo) + PLCPANSSOT ineffective for clinical symptoms or psychosocial functioning.NRLimitingSCZ schizophrenia, HC healthy control, M male, F female, OT oxytocin, PLC placebo.1Participants: N(F/M) — sex data from original reports or calculated a;2Effect size (Cohen’s d): based on original data or converted (Hedges’ g) b; NR not reported/calculable;3Evidence type: Core (RCT, standardized scale, positive); Limiting (RCT, standardized scale, negative); Supportive (non- scale behavioral/neuro task, positive/trend); Synergistic (OT + psychosocial); Moderating (context- dependent). Table 1Oxytocin Intervention for Negative Symptoms in Schizophrenia.ActhorParticipants N (M/F)1Dosing and DurationSymptom measuresMain Findings (with dimensional annotation)Effect size (Cohen’s d)2Evidence type31Zierhut et al. [50, 51]SSD 41 (31/10)OT (24 IU qd × 1wk) + PLC + MBGT (twice weekly)PANSS, the Self -Evaluation of Negative Symptoms (SNS); Plasma oxytocin levelsOT + MBGT shows superior improvement in negative symptoms, particularly avolition, social withdrawal and blunted affect. [MAP&EXP]Avolition: d = 0.70b Social withdrawal: d = 0.50bDiminished emotional range: d = 0.70bCore + Synergistic2Ota M et al. [49]SCZ 16 (7/9)OT (12 IU bid × 12wk)PANSS, Brief Assessment of Cognition in Schizophrenia (BACS); MRIOT improve negative symptoms and verbal fluency via right insula and left cingulate cortex (negative correlation with gray matter volume). [MAP&EXP]PANSS-N: d = 1.05bMAP: d = 0.66bEXP: d = 0.91bBACS: d = 0.53Core + Supportive3Jarskog et al. [61]SCZ 62 (47/15)aOT (24 IU bid × 12wk) + PLCPANSS, Social Skills Performance Assessment (SSPA)OT ineffective for social functioning but significantly reduced PANSS negative symptoms. [MAP]NRCore4Halverson et al. [62]SCZ 62 (47/15)aOT (24 IU bid × 12wk) + PLCPANSS, Specific Levels of Functioning Scale (SLOF); Interpersonal Reactivity Index (IRI)OT significantly improves cognitive empathy (IRI perspective-taking). [MAP]NRCore + Supportive5Fulford et al. [20]SCZ 42 (32/10), HC 43 (29/14)OT (40 IU × 2, ≥1wk interval) + PLCPANSS; The Awareness of Social Inference Test (TASIT), Social Vigor Task (SVT)Passive/apathetic social withdrawal associated with lower effort; OT shows a trend toward improving this effect. [MAP]NRCore + Supportive6Mouchlianitis et al. [54]SCZ 20 (20/0)OT (40 IU, single dose) + PLCPANSS; Trust Game; fMRIOT coordinates reward processing networks during social learning tasks, promoting cortical processing in investment trials (implicit/contextual rewards) and limbic processing in repayment trials (explicit rewards). [MAP]NRSupportive7Bradley et al. [58]SCZ 39 (39/0), HC 54 (54/0)OT (40 IU, single dose) + PLCPANSS; Auction GameOT improves social reward motivation. [MAP]NRSupportive8Caravaggio et al. [59]SCZ 43 (43/0), HC 16 (16/0)OT (50 IU, single dose, ≥20 d interval) + PLCSAPS, SANS, Social Functioning Scale (SFS); Bead Task; MRISCZ tendency to jump to conclusions (JTC) is positively correlated with ventral caudate volume; OT ineffective for JTC bias. [MAP]NRSupportive9Strauss et al. [64]SCZ 62 (38/24)OT (36 IU bid × 24wk) + PLC + group CBSST (twice weekly)Facial Emotion Recognition Test (FERT), Empathic Accuracy Task (EAT); Eye-trackingOT + CBSST ineffective for social function; enhanced gaze to happy faces (exploratory). [MAP]FERT-total: d = 0.37EAT-positive: d = 0.46Core + Supportive + Synergistic10Buchanan et al. [63]SCZ 62 (38/24)OT (36 IU bid × 24wk) + PLC + group CBSST (twice weekly)BPRS, SANS, The Birchwood Social Functioning Scale (BSFS)OXT+ CBSST ineffective for social functioning and negative symptoms, improvements driven by site effects.[MAP]NRRestrictive + Synergistic + Moderating11Abram SV et al. [111]SCZ 22 (22/0), HC 22(22/0);OT (40 IU, single dose) + PLCPANSS, SANS, fMRIOT enhances amygdala- left temporal connectivity in SCZ; baseline hypoconnectivity in this circuit specifically correlates with expressive negative symptoms. [EXP]Amygdala-left temporal connectivity: d = 1.00SANS: d = 1.32Core + Supportive12Woolley et al. [19]SCZ 33 (30/3) HC 35 (33/2)OT (40 IU × 2, ≥1wk interval) + PLCPANSS, Social Judgment Task (SJT), Facial Expression CodingOT effective for facial emotion expression. [EXP]NRCore + Supportive13Gibson et al. [67]SCZ 14 (11/3)OT (24 IU bid × 6wk) + PLCPANSS, Role-Playing AssessmentOxytocin effective for negative symptoms, nonverbal social skills (fear recognition), and specific social skills (self-reported perspective-taking ability). [EXP]PANSS-N: d = 1.77Specific: d = 0.26Nonverbal: d = 0.23Core + Supportive14Dwyer et al. [68]SCZ 40 (33/7)aOT (24 IU bid × 6wk) + PLCSocial role-playing tasks; computer-based behavioral coding and verbal content analysisinteractionsOT effective for total verbal output, characterized by increased use of positive emotion and social words. [EXP]Word count: d = 0.35Positive Emotion word: d = 0.22social word: d = 0.29Supportive15Cohen et al. [65]SCZ 40 (33/7)OT (24 IU bid × 6wk) + PLCSocial Competence Assessment Scale, computerized natural speech and facial expression measurementOT effective for negative facial expressions, ineffective for speech expression in SCZ. [EXP]Facial-negative: d = 0.51Vocal-pause times: d = 0.28Supportive16Cacciotti-Saija C et al. [125]SCZ 52 (36/16)OT (24 IU bid × 6wk) + PLC + group SCT (twice weekly)SAPS, SANS, Social Functioning Scale (SFS)OT may improve negative symptoms; greater OT use correlates with greater negative symptom reduction.SAPS: d = 0.27SANS: d = 0.31SFS: d = 0.04Core17Feifel D et al. [126]SCZ 19 (15/4)aOT (40 IU bid × 3wk) + PLCPANSS, CGIOT well- tolerated with antipsychotic properties; medium effect sizes; higher doses/longer treatment may yield greater benefit.PANSS: d = 0.43PANSS-N: d = 0.5CGI-I: d = 0.74Core18Korann V et al. [95]SCZ 31 (31/0) HC 21 (21/0)OT (24 IU, single dose) + PLCPANSS, SANS; fMRIOT- induced connectivity gains correlate with higher cognitive insight and lower negative symptoms. OT enhances effective connectivity from left caudate to left SMA, precentral gyrus, and IFG pars triangularis.NRCore + Supportive19Modabbernia A et al. [127]SCZ 40 (33/7)OT (20 IU bid wk1, then 40 IU bid wks2–8) + PLCPANSSOT add- on to risperidone effectively reduces positive symptoms and is well- tolerated.PANSS: d = 1.9PANSS-N: d = 1.4Core + Moderating20Lee MR et al. [53, 128]SCZ 28 (20/8)OT (20 IU bid × 3wk) + PLCPANSS, SANSOT ineffective for total, positive, and negative symptoms overall; improves negative symptoms in hospitalized subgroup [exploratory].SANS: d = 0.01;hospitalized:SANS: d = 0.85,SANS-EXP: d = 0.74Moderating21Buchanan RW et al. [129]SCZ 56 (45/11)OT (24 IU bid × 12wk) + PLCSANSOT ineffective for SANS total score; no benefit in subgroup analyses.SANS: d = 0.10Limiting22Dagani J et al. [130]SCZ 32 (26/6)OT (40 IU qd × 8mo) + PLCPANSSOT ineffective for clinical symptoms or psychosocial functioning.NRLimitingSCZ schizophrenia, HC healthy control, M male, F female, OT oxytocin, PLC placebo.1Participants: N(F/M) — sex data from original reports or calculated a;2Effect size (Cohen’s d): based on original data or converted (Hedges’ g) b; NR not reported/calculable;3Evidence type: Core (RCT, standardized scale, positive); Limiting (RCT, standardized scale, negative); Supportive (non- scale behavioral/neuro task, positive/trend); Synergistic (OT + psychosocial); Moderating (context- dependent). Oxytocin Intervention for Negative Symptoms in Schizophrenia. Avolition: d = 0.70b Social withdrawal: d = 0.50b Diminished emotional range: d = 0.70b PANSS-N: d = 1.05b MAP: d = 0.66b EXP: d = 0.91b BACS: d = 0.53 FERT-total: d = 0.37 EAT-positive: d = 0.46 Amygdala-left temporal connectivity: d = 1.00 SANS: d = 1.32 PANSS-N: d = 1.77 Specific: d = 0.26 Nonverbal: d = 0.23 Word count: d = 0.35 Positive Emotion word: d = 0.22 social word: d = 0.29 Facial-negative: d = 0.51 Vocal-pause times: d = 0.28 SAPS: d = 0.27 SANS: d = 0.31 SFS: d = 0.04 PANSS: d = 0.43 PANSS-N: d = 0.5 CGI-I: d = 0.74 PANSS: d = 1.9 PANSS-N: d = 1.4 SANS: d = 0.01; hospitalized: SANS: d = 0.85, SANS-EXP: d = 0.74 SCZ schizophrenia, HC healthy control, M male, F female, OT oxytocin, PLC placebo. 1Participants: N(F/M) — sex data from original reports or calculated a; 2Effect size (Cohen’s d): based on original data or converted (Hedges’ g) b; NR not reported/calculable; 3Evidence type: Core (RCT, standardized scale, positive); Limiting (RCT, standardized scale, negative); Supportive (non- scale behavioral/neuro task, positive/trend); Synergistic (OT + psychosocial); Moderating (context- dependent). Oxytocin effects on motivational-pleasure deficits in schizophrenia Anhedonia Anhedonia, characterized by a persistent diminished capacity to experience pleasure and anticipate rewards, is closely associated with dysfunctions in reward processing system. Research indicates that while individuals with schizophrenia may partially retain the ability to experience consummatory pleasure, they consistently exhibit significant deficits in anticipatory pleasure. This deficit in anticipatory pleasure not only undermines motivation-driven behavior towards future rewards, but also represents a critical pathological link between “pleasure experience” and “behavioral motivation,” profoundly impacting social functioning and quality of life [52]. Recent evidence suggests a role for intranasal OT in addressing this deficit. One study found that a single 24 IU dose of intranasal OT administered prior to a mindfulness therapy produced acute effects that significantly reduced anhedonia ratings in patients [50]. In an independent three-week adjunctive treatment study using a higher dosage of 40 IU, OT demonstrated a favorable trend toward reducing anhedonia; however, the results did not achieve statistical significance [53]. Social learning studies based on trust games demonstrated that while a single OT administration did not induce significant behavioral changes, it normalized neural activity patterns during critical stages of social reward learning. Specifically, it enhanced activation in the ventral caudate and suppressed amygdala response, providing neurophysiological evidence for its potential to alleviate social anhedonia [54].In summary, current evidence supports the potential of intranasal OT to positively influence anhedonia in schizophrenia, though the stability and generalizability of its efficacy require further validation. Avolition Avolition in schizophrenia arises from abnormalities in reward prediction error signaling, which result from a diminished capacity for anticipatory pleasure [55]. This directly leads to a significant reduction in goal-directed behavior, manifesting as initiation difficulties, effort avoidance, and impaired persistence. In contrast to marked difficulties in complex, executive-demanding tasks, performance in routine daily activities is often comparatively preserved. Research indicates that in effort-competitive tasks, while individuals with schizophrenia pursue rewards similarly to healthy controls, they exhibit markedly lower initial motivation and inadequate strategy adjustment capabilities [56, 57]. Intranasal OT intervention has been shown to effectively enhance participation motivation by amplifying patients’ subjective valuation of social victory, specifically ameliorating their motivational deficits [58]. In the Beads Task, a paradigm requiring sustained judgment and decision-making, patients demonstrate a pronounced jumping-to-conclusions bias and significantly reduced motivation for information gathering compared to healthy individuals. While intranasal OT intervention did not produce an overall effect, exploratory analyses suggested that it may specifically improve cautious decision-making motivation in patients with lower social functioning by modulating ventral caudate nucleus function and enhancing social cue processing [59]. In summary, current evidence indicates that the efficacy of intranasal OT for improving motivational and decision-making deficits is circumscribed and significantly moderated by individual factors such as baseline social functioning and brain structure. Asociality The core characteristics of social withdrawal in schizophrenia encompass diminished social motivation and a significant reduction in proactive social behaviors, which are closely linked to social anhedonia and avolition. Research has revealed distinct psychological mechanisms underlying active social avoidance versus passive social reduction: proactive avoidance is primarily associated with anticipatory anhedonia, whereas passive social reduction correlates more strongly with general motivational deficits and restricted emotional expression. Notably, these behavioral phenotypes exhibit significant gender differences, with females more frequently exhibiting proactive avoidance and males tending toward passive withdrawal [60, 61]. Regarding interventions for social deficits, clinical evidence indicates that long-term, high-frequency OT monotherapy has limited efficacy in directly improving social deficits in schizophrenia [62]. However, when administered as an adjunctive enhancer alongside structured psychosocial interventions, OT demonstrates significant synergistic effects. For instance, its combination with Cognitive Behavioral Social Skills Training (CBSST) significantly enhances patients’ overall social functioning with sustained treatment effects [63, 64]. Similarly, the combined administration of OT and mindfulness-based therapy has been demonstrated to enhance patients’ perspective-taking capacity, consequently increasing their social sensitivity and motivation for interpersonal interactions [50]. This context-dependent efficacy is further supported by behavioral experiments. Studies using the Social Vigor Task (SVT) demonstrated that patients exhibit normal immediate vigor responses to social encouragement, with their deficit lying in insufficient spontaneous, desire-driven motivation rather than impaired motivational responsiveness. A single dose of OT did not enhance this immediate responsiveness to social stimuli [20], corroborating that social deficits originate from impaired intrinsic social motivation rather than compromised capacity to respond to external social cues. In conclusion, OT may not directly ameliorate symptoms but rather functions as a “social learning catalyst,” optimizing the neurobiological state for social information processing and thereby enhancing an individual’s social learning capacity within specific contexts. This mechanism explains its characteristically context-dependent therapeutic efficacy. Oxytocin effects on emotional expression deficits in schizophrenia Blunted affect Blunted affect in schizophrenia manifests primarily as a significant reduction in the intensity and range of non-verbal emotional expression, including facial expression, vocal prosody, and body language. Its core pathological mechanism is not a complete absence of internal emotional experience, but rather involves an impaired translation of internal emotional states into external expressive signals within the neural pathways responsible for this conversion [43]. Research indicates variability in emotional consistency across different expressive channels: reductions in facial expressivity and vocal prosody are typically more pronounced than in body language, and the former two are more strongly associated with impaired social functioning. Interventions targeting these deficits show channel-specific effects. Intranasal OT has demonstrated inconsistent impacts across different expressive modalities. One study found that OT produced modest improvement in patients’ negative facial expressions, but had limited efficacy in enhancing natural vocal expression [65]. Conversely, another study using emotional videos to elicit responses found that a single dose of intranasal OT significantly enhanced patients’ facial expressiveness in response to positive emotional stimuli [19]. Furthermore, the effects of OT appear to be functionally specific. While it effectively improved patients’ spontaneous expression generation, it did not concurrently alter their subjective judgments of the trustworthiness of static neutral faces [66]. This dissociation between “expression and cognition” suggests that OT may preferentially act on the behavioral generation system that drives external emotional expression, while having a more limited impact on the cognitive appraisal system underlying nuanced social evaluations. Alogia The core feature of alogia in schizophrenia is a marked reduction in thought and speech output, manifesting as brief replies, empty speech, delayed responses, and impaired verbal fluency. Its pathological basis extends far beyond simple reticence, representing a direct linguistic manifestation of an underlying poverty of thought and reflecting a fundamental “productive deficit.” Research indicates that lower endogenous OT levels are associated with more severe alogia. In patients with milder symptoms, OT administration has been shown to improve social skills, primarily reflected in more proactive social interaction, more effective communication, and richer emotional expression during exchanges [67]. Corroborating this, research has found that OT intervention increased both the quantity and emotional quality of speech during a free-talk task. Specifically, it boosted total verbal output and increased the frequency of positive emotion words, social vocabulary, and affiliative language in speech content [68]. However, OT’s beneficial effects on speech function are not universal. One study reported that it failed to improve patients’ basic verbal learning and expressive fluency [65]. More notably, a role-play task study found that patients in the OT intervention group demonstrated poorer performance than the placebo group in terms of speech clarity, fluency, and content organization [61]. In summary, research on the impact of OT on verbal expression in schizophrenia remains relatively limited, and the results exhibit significant heterogeneity, precluding definitive conclusions. Future targeted studies are needed to clarify its effects across different speech dimensions and determine its clinical applicability. Neurobiological mechanisms of oxytocin in improving negative symptoms in schizophrenia The neural mechanisms through which OT influences the negative symptoms in schizophrenia may be related to its modulation of distinct neural circuits [69, 70]. The core hypothesis posits that OT exerts its effects through the parallel regulation of interconnected neural pathways. First, it enhances the function of the mesolimbic reward system, increasing the anticipation of and sensitivity to rewards—particularly social rewards—thereby improving motivational drive [71, 72]. Second, it modulates amygdala activity not through simple suppression, but by reducing its baseline neural activity levels. This may attenuate its hyper-responsiveness to negative social cues while optimizing its processing of social-emotional signals, thereby providing accurate information for higher-order cortical regions [73, 74]. Subsequently, by specifically enhancing the integrative function of the prefrontal cortex, OT strengthens the integration of motivational and emotional signals, translating these integrated signals into coordinated goal-directed actions and affective expressions [75]. This synergistic, circuit-level modulation facilitates the more efficient translation of latent motivational drives into goal-directed behavior, thereby systematically addressing the core symptoms ranging from amotivation and anhedonia to impairments in emotional expression. Mechanisms underlying oxytocin effects on motivational-pleasure deficits Motivation-pleasure deficits in schizophrenia are thought to be closely related to synergistic dysfunction of the mesolimbic dopamine system and the prefrontal-striatal circuit [76]. Emerging evidence indicates that OT may exert potential ameliorative effects through dual pathways: on one hand, it may enhance nucleus accumbens (NAc) dopamine signaling, thereby influencing reward experience; on the other hand, it has been shown to strengthen prefrontal-striatal functional connectivity, which may in turn facilitate motivational formation and behavioral selection (see schematic in Fig. 2) [77, 78]. Based on this framework, it can be hypothesized that oxytocinergic modulation may be associated with positive behavioral changes in patients, including enhanced pleasure during social interactions, increased active social engagement, and a greater willingness to sustain effort toward long-term goals [79].Fig. 2An integrative model of oxytocinergic modulation in Schizophrenia’s Motivation-Pleasure Deficit (Schematic).This model illustrates four key pathways: (1) Mesolimbic reward pathway: ventral tegmental area (VTA) → nucleus accumbens (NAc), processing reward information; (2) Prefrontal motivation integration pathway: VTA → NAc → orbitofrontal cortex (OFC)/ventromedial prefrontal cortex (vmPFC) → dorsal anterior cingulate cortex (dACC), integrating motivational signals; (3) Cortical–basal ganglia motor pathway: dACC/NAc → globus pallidus → primary motor cortex, translating motivation into action; (4) Oxytocin release pathway: hypothalamic paraventricular nucleus (PVN) and supraoptic nucleus (SON) → posterior pituitary, mediating peripheral and central release. An integrative model of oxytocinergic modulation in Schizophrenia’s Motivation-Pleasure Deficit (Schematic). This model illustrates four key pathways: (1) Mesolimbic reward pathway: ventral tegmental area (VTA) → nucleus accumbens (NAc), processing reward information; (2) Prefrontal motivation integration pathway: VTA → NAc → orbitofrontal cortex (OFC)/ventromedial prefrontal cortex (vmPFC) → dorsal anterior cingulate cortex (dACC), integrating motivational signals; (3) Cortical–basal ganglia motor pathway: dACC/NAc → globus pallidus → primary motor cortex, translating motivation into action; (4) Oxytocin release pathway: hypothalamic paraventricular nucleus (PVN) and supraoptic nucleus (SON) → posterior pituitary, mediating peripheral and central release. Oxytocin modulation of the mesolimbic dopamine reward pathway Dysfunction within the mesolimbic dopamine reward system constitutes a core pathological basis for amotivation and anhedonia in schizophrenia [80]. Within this system, OT receptors are densely distributed on dopamine neurons in the VTA and medium spiny neurons in the NAc, forming the anatomical basis for its regulation of reward pathways [81, 82]. Animal studies demonstrate that endogenous OT in the VTA can rapidly trigger dopamine release in the NAc during social interactions, specifically enhancing the reward value of social behavior [83, 84]. Neuroimaging studies further support this mechanism: meta-analyses and empirical data indicate that patients with schizophrenia exhibit significantly reduced grey matter volume in the NAc and consistently show attenuated NAc activation during reward anticipation tasks. Concurrently, there is an abnormal increase in local low-frequency oscillation power (particularly in the delta/theta bands), a feature directly linked to their diminished experience of pleasure and reduced proactive behavior in social activities [85, 86]. Notably, functional magnetic resonance imaging (fMRI) studies in schizophrenia have found that intranasal OT effectively mitigates these abnormalities. Patients exhibit significantly enhanced NAc activation when receiving social rewards (e.g., smiling faces), and resting-state connectivity between the NAc and prefrontal cortex markedly improves [54, 87]. These neural improvements directly reflected in behavioral changes, manifesting as increased engagement and motivation in social interactions. Particularly in social learning tasks requiring behavioral adjustments based on others’ feedback, patients demonstrate enhanced choice accuracy and learning efficiency [88]. Collectively, these findings indicate that OT acts on receptors in the VTA to enhance tonic neural encoding of social rewards in the NAc, thereby influencing social motivation and behavioral responses in individuals with schizophrenia. Oxytocin modulation of the prefrontal-striatal circuit Motivation-pleasure deficits in schizophrenia are also closely associated with dysfunction within the prefrontal-striatal circuitry. This circuit is modulated by prefrontal parvalbumin-expressing interneuron gating, and its dysregulation impairs the computation of social reward value and motivational formation [75]. Previous research suggests that abnormal functional integration within the striato-cortical network may be an important neural basis for motivation-pleasure deficit [89]. Neuroimaging studies further reveal abnormal circuit coordination in patients, manifested as weakened resting-state functional connectivity between the dorsolateral/orbitofrontal cortex and ventral striatum, alongside insufficient task-related neural mobilization for high-effort rewards. These dual abnormalities jointly correlate with motivational deficits [57, 90]. According to Brakowski et al., self-report assessments and clinical interviews in individuals with schizophrenia may capture distinct symptom dimensions associated with different neural mechanisms [89]. OT, in turn, modulates social behavior through distinct neural pathways: during cooperative learning tasks, it synergistically enhances the medial prefrontal cortex’s encoding of prosocial value prediction errors and prefrontal-striatal circuit coupling with the dopamine system [91]. In contrast, during competitive interactions, it acts independently on the orbitofrontal cortex to amplify in-group value computation, thereby promoting group-biased self-serving behavior [92]. A meta-analysis provides evidence that intranasal OT synergistically enhances ventral striatum responsiveness to social rewards while optimizing value assessment functions in the ventromedial/orbitofrontal cortex [93]. Resting-state fMRI studies reveal that OT specifically strengthens functional connectivity within prefrontal-ventral striatal circuits. It enhances the motivational evaluation pathway from the left ventral caudate nucleus to the right dorsal anterior cingulate cortex while amplifying the action preparation pathway from the left globus pallidus to the right primary motor cortex, thereby synergistically advancing reward evaluation and behavioral initiation processes [94]. Dynamic causal modeling of resting-state fMRI data reveals that OT can enhance effective connectivity between the ventral striatum and the prefrontal cortex in individuals with schizophrenia, and this change is associated with an increase in the frequency of behavior in which patients initiate social interactions, suggesting that enhanced reward circuit function may be linked to improvements in motivation [95]. Mechanisms underlying oxytocin effects on emotional expression deficits Emotional expressivity deficits in schizophrenia are thought to be associated with multistage neural pathway dysfunction, ranging from internal emotional processing to external behavioral output [96, 97]. The underlying mechanisms may involve impaired emotional integration within amygdala-prefrontal circuits and disrupted expression execution in sensorimotor integration networks. Recent evidence suggests that OT may improve potential ameliorative effects through multilevel modulation: at the internal processing level, it may enhance prefrontal regulation of the amygdala and improve the precision of emotional information processing [98]; at the external expression level, it may facilitate sensorimotor integration network activity and its coordination with motor pathways [99, 100]. Based on this framework, it can be hypothesized that such integrated modulation might ultimately be associated with improved emotional recognition accuracy, enhanced facial expression generation, and better coordination of behavioral expression during social interactions in patients [101]. Oxytocin modulation of the amygdala-prefrontal emotional circuit Emotional expression deficits in schizophrenia are believed to be associated with hierarchical dysfunction within social information processing pathways [102]. During the initial stage of multimodal social information integration, fMRI studies reveal impaired detection of speech-gesture mismatches in patients, with insufficient activation in the right superior temporal sulcus/middle temporal gyrus during conflicting information processing, leading to biases in social context interpretation. This perceptual ambiguity propagates to the core emotional processing circuitry, specifically the amygdala-prefrontal cortex network. Consequently, it interferes with the essential dynamic balance between the amygdala’s swift emotional detection and the prefrontal cortex’s advanced regulatory functions. This disruption leads to disorganized emotional experiences, which fundamentally underlie deficits in emotional expression [103, 104]. Research indicates that OT coordinates this circuit function through multi-level mechanisms. Epigenetic studies confirm that the OT system regulates the developmental trajectory of amygdala-prefrontal connectivity during childhood via DNA methylation mechanisms [105]. Neuroimaging studies in adult samples demonstrate that OT both enhances top-down inhibitory control from prefrontal regions (e.g., the cingulate cortex) over the amygdala [106] and simultaneously increases the processing precision of emotional information across the distributed amygdala-prefrontal network [107]. Concurrently, OT acts as a neural coordinator, strengthening functional connectivity within this circuit during both resting and task states, as well as its consistency with behavioral performance [108]. Notably, orally administered OT produces a sex-dimorphic modulation of amygdala reactivity: it reduces activation intensity in males but enhances reactivity in females [109]. In contrast, low-dose intranasal OT delivered via an efficient nasal device can cross the blood-brain barrier directly, producing a dose-dependent inhibitory effect on the amygdala [110]. A combination of fMRI and psychophysiological interaction analysis revealed that OT specifically enhances functional connectivity between the amygdala and ventral prefrontal cortex in schizophrenia, promoting emotional integration and generation of motivation [111]. This mechanism may be associated with the clinically observed improvements in social-emotional functioning. Collectively, these findings demonstrate that OT stabilizes dysfunctional amygdala-prefrontal circuits through multi-target mechanisms, providing a critical neurobiological explanation for how it ameliorates disorganized emotional experiences and facilitates affective expression. Oxytocin modulation of the sensorimotor integration network and prefrontal motor pathway The affective expression network, comprising the prefrontal motor cortex and the posterior sensorimotor cortex, collectively forms the neural basis underlying the output of emotional expression [112]. Within this network, the sensorimotor integration network, encompassing regions such as the ventral premotor cortex and the inferior parietal lobule, is traditionally considered a critical hub for emotional resonance and action intention understanding [113]. In parallel, the prefrontal cortex, particularly the ventrolateral prefrontal cortex as a core node, organizes and executively controls complex facial expressions and verbal expression through its connections with the motor cortex and supplementary motor area [114]. Impairment in this circuit can lead to expressive deficits even when emotional experience is intact, highlighting a disconnection between emotional experience and its expression. From the perspective of sensorimotor integration, multiple neuroimaging studies have found that individuals with schizophrenia exhibit significantly reduced activation or decreased functional connectivity in core sensorimotor integration regions (e.g., posterior inferior frontal gyrus and anterior inferior parietal lobule) during action observation, imitation, and language tasks [115]. At the neuro-electrophysiological level, studies consistently show attenuated desynchronization of electroencephalographic mu frequency band (8–13 Hz) oscillations (i.e., mu suppression) during action observation in patients [116], collectively suggesting a disruption in the fundamental process of transforming perceptual input into motor simulation. Notably, research employing facial electromyography and behavioral analyses has demonstrated that OT specifically enhances mimicry activity of the zygomaticus major and corrugator supercilii muscles in healthy individuals under conditions of mutual gaze, facilitating the generation of context-appropriate facial expressions [117]. Neuroimaging studies have further revealed that OT selectively modulates neural activity in the inferior frontal gyrus and the inferior parietal lobule [118].This effect suggests that the underlying neural mechanism may involve OT’s optimized modulation of this sensorimotor integration network, potentially by enhancing the efficiency of sensorimotor integration or by improving the prediction and integration of sensorimotor information during social interaction, ultimately affecting precision and coordination of social interaction. Conclusions and agenda for future research Grounded in the two-dimensional model of negative symptoms, the present review systematically synthesizes evidence from animal studies, clinical interventions, and mechanistic investigations concerning oxytocin-based interventions in schizophrenia. Integration of these three bodies of literature reveals the following patterns. Animal studies have elucidated, at multiple levels, the critical role of OT in regulating social reward and affective expression, thereby providing a refined neurobiological foundation for understanding the dimension-specific mechanisms underlying negative symptoms. By contrast, clinical intervention studies have yielded marked heterogeneity at the macro-level of rating scales. Some trials have reported improvements in total negative symptom scores, whereas others have produced null findings. However, mechanistic studies employing functional neuroimaging and fine-grained behavioral coding have captured more consistent effects within specific symptom dimensions. This convergence and divergence, spanning animal-to-human research and macro-to-micro level assessments, serves as the basis for in-depth discussion. The stark contrast between the high consistency of animal findings and the mixed results of clinical studies highlights a fundamental gap between idealized conditions and real-world complexity. This discrepancy stems from the fact that clinical research must contend with confounding factors such as genetic heterogeneity, environmental variability, placebo effects, and subjective assessment biases. Notably, when studies move beyond traditional clinical scales and adopt dimension-specific targeted research paradigms, including behavioral tasks and neuroimaging markers, the evidence begins to align more closely with the efficacy patterns observed in animal studies. This contrast suggests that composite scales such as the PANSS and SANS, when employed as clinical endpoints, may lack sensitivity to dimension-specific neurobehavioral changes. Moreover, OT system dysfunction may represent an intermediate phenotype present only in a subset of patients, rather than a universal feature across all individuals presenting with negative symptoms. It is important to note that interpretation of existing clinical evidence must also consider the critical limitation of sex imbalance. Recent research has indicated that female patients with schizophrenia and at-risk individuals exhibit hypomethylation of the OXTR gene, an epigenetic signature significantly associated with negative symptoms, particularly anhedonia and social withdrawal [119]. This epigenetic marker not only reveals a distinct biological basis for OT system abnormalities in females but also suggests that exogenous OT interventions targeting this subgroup warrant further investigation regarding their potential efficacy. Concurrently, the coexistence of synergistic and moderating evidence serves as a cautionary reminder. The effects of OT do not operate independently of the environment, and its clinical translation must integrate strategies that account for social context and behavioral intervention. The sources of heterogeneity outlined above precisely illuminate the translational challenges that must be overcome in moving from laboratory findings to clinical application. The first concerns the challenge of administration strategy. The majority of current studies employ intranasal OT with highly convergent dosing and frequency regimens. However, this empirically driven approach circumvents a fundamental question. Does intranasal OT effectively cross the blood-brain barrier? Positron emission tomography imaging by Winterdahl and colleagues has demonstrated limited and highly variable central nervous system distribution following intranasal administration, with the optimal therapeutic window remaining unclear [17]. Encouragingly, researchers are actively exploring novel delivery systems, including selective OT receptor agonists [120], nanoparticle carriers [121], and sustained-release formulations [122]. Pilot studies of sublingual OT have also yielded promising results [123], suggesting that these novel approaches may enhance bioavailability and central targeting. The second key difficulty involves the strategic integration of behavioral co-intervention. OT may not directly generate social behavior, but rather induce a state of neuroplasticity conducive to social learning. In the absence of appropriate social interaction or behavioral training during this window, the pharmacological effects may dissipate. This mechanistic interpretation suggests that future trials should conceptualize OT as an adjuvant enhancer of behavioral interventions, rather than a stand-alone treatment. The third critical issue pertains to the precision of outcome measures. Clinical scales such as the PANSS rely on clinical interviews and are susceptible to assessor bias, patient insight levels, and cultural factors. In contrast, objective indicators, including behavioral paradigms, eye-tracking, and functional neuroimaging, offer dual utility. In early-phase clinical trials, task-induced functional brain changes can serve as pharmacodynamic markers, providing real-time verification of whether the drug successfully engages target neural circuits. Concurrently, baseline neural or behavioral characteristics hold promise as predictive biomarkers, informing subsequent patient stratification and selection. In summary, OT research is currently at a critical stage, transitioning from conceptualizing the molecule as a broad-spectrum hormone to recognizing it as a precision modulator. The primary direction forward entails abandoning the non-selective approach and instead recruiting patients with oxytocin-sensitive subtypes. Drawing inspiration from the multi-step precision treatment framework proposed by Xiang and colleagues in autism research [124], this approach would involve first identifying individuals with putative OT system deficits based on genetic and epigenetic markers. It would then require confirming responders through endogenous OT level assessment and single-dose behavioral challenge paradigms. Finally, long-term intervention would be administered exclusively to the sensitive subgroup. A core priority is the optimization of administration protocols and standardization of co-intervention models, including systematic comparison of pharmacokinetic profiles across different routes of administration and delineation of the optimal timing and content of behavioral interventions. At the theoretical level, critical reflection on lessons learned from the autism field is warranted. Site effects in large-scale multi-center trials, substantial patient heterogeneity, and neglect of critical developmental windows are pitfalls that schizophrenia research must actively avoid. Only through such rigorous and targeted approaches can OT be transformed from a potentially effective candidate into a precision therapeutic tool that provides clear answers to the questions of for whom and how it works.