The study, led by Xiao Wang and Shuyi Wang of Shanxi Medical University together with colleagues across seven Chinese hospitals, enrolled 118 patients with early-onset schizophrenia and 76 healthy controls. Every patient met diagnostic criteria confirmed through the Structured Clinical Interview for DSM-5, and all were eighteen or younger at the time of recruitment. Blood was drawn in the early morning, between eight and nine o’clock, to control for the circadian rhythms that can distort hormone measurements. Plasma was separated by centrifugation at 4 degrees Celsius and frozen at minus 80 degrees until analysis with a commercially available, extraction-free enzyme-linked immunosorbent assay kit, a choice the researchers made deliberately to sidestep longstanding technical controversies over how plasma oxytocin should be prepared and measured.

The central result was unambiguous. Patients with early-onset schizophrenia had plasma oxytocin levels far below those of healthy controls, a difference that held up at the stringent threshold of P less than 0.001. Crucially, the team ran sex-stratified sensitivity analyses because the patient group contained a higher proportion of females. The oxytocin deficit persisted in both males and females when analyzed separately, and a two-by-two analysis of variance found a strong effect of diagnosis but no effect of sex and no interaction between the two. In other words, the hormonal signature appears to be a property of the illness itself, not an artifact of who happened to enroll in the study.

But the researchers went further than a simple group comparison. They fed every variable that differed between patients and controls—demographics, personality dimensions, childhood trauma scores, and oxytocin itself—into a binary logistic regression using a backward likelihood ratio selection procedure. Four factors survived: sex, extraversion, plasma oxytocin, and emotional abuse. Together they explained 55.4 percent of the variance separating patients from healthy adolescents. Each additional unit of oxytocin corresponded to a 3.3 percent decrease in the odds of belonging to the patient group, while higher scores of emotional abuse raised those odds. Males were only about 24 percent as likely as females to fall into the patient category, a pattern the authors connect to prior reports of X-chromosome influences and sex-specific transmission in psychosis.

The diagnostic potential of the hormone was quantified with receiver operating characteristic analysis, the standard statistical tool for judging how well a biomarker separates two groups. Plasma oxytocin alone achieved an area under the curve of 0.776, a value generally interpreted as moderate discriminative ability, with an optimal cutoff of 183.428 picograms per milliliter that yielded 83.9 percent sensitivity and 67.1 percent specificity. The full four-variable regression model performed better still, reaching an area under the curve of 0.892. For context, an area of 0.5 reflects coin-flip performance while 1.0 is perfect classification. A blood test approaching the 0.9 mark, even in a cross-sectional sample, is the kind of number that makes biomarker researchers sit up and pay attention.

Perhaps the most intriguing finding emerged when the team split patients by symptom severity using a cutoff of 65 on the Positive and Negative Syndrome Scale, the gold-standard instrument for rating psychotic symptoms. Among patients with milder symptoms, oxytocin levels rose in tandem with symptom severity, a positive correlation of 0.482 that reached statistical significance. Among severely affected patients, the correlation vanished entirely. The authors offer a provocative interpretation: in the early stages of illness, the body may ramp up oxytocin secretion as a compensatory attempt to buffer emerging psychopathology, producing a temporary positive association between the hormone and symptom load. As the disease progresses, broader neurobiological damage—possibly including altered oxytocin receptor sensitivity, chronic microglial hyperactivity, and gene-environment interactions involving oxytocin receptor variants such as rs53576—may swamp that signal and mask the relationship.

The neurobiology behind these observations is far from settled, but the study situates oxytocin within the established framework of schizophrenia pathophysiology. The disorder is classically associated with dopaminergic hyperfunction in limbic circuits, dopaminergic hypofunction in the prefrontal cortex, and impaired signaling through N-methyl-D-aspartate-type glutamate receptors. Animal work suggests endogenous oxytocin may protect against exactly these disruptions: it appears to counteract glutamatergic hypofunction and central hyperdopaminergic states, and oxytocin administration in rodents can restore prepulse inhibition deficits, a well-established endophenotypic marker of schizophrenia, when those deficits are induced by dopamine agonists or NMDA receptor antagonists. Whether reduced peripheral oxytocin in patients is a cause of the illness, a consequence of it, or merely a correlated byproduct remains an open question the authors are careful to flag.

The study also wove together two psychosocial threads that have long haunted schizophrenia research: personality and childhood adversity. Patients scored lower on extraversion and higher on neuroticism than controls, consistent with meta-analytic findings across the field, and reported significantly more of every category of childhood maltreatment measured by the Childhood Trauma Questionnaire—emotional abuse, physical abuse, sexual abuse, and both forms of neglect. Prior work has shown that intranasal oxytocin can increase self-reported extraversion and that lower endogenous oxytocin tracks with higher trauma scores in adult patients, with the hormone mediating the link between early adversity and later psychopathology. The new data extend that biopsychosocial picture into adolescence, a developmental window where adult environmental confounds have had less time to accumulate and where the intrinsic biology of the illness may be most visible.

The therapeutic implications are tantalizing but tempered. Intranasal oxytocin has produced genuinely mixed results in schizophrenia trials; a recent meta-analysis found roughly two-thirds of intervention studies reported no significant improvement in social cognition or overall symptoms. The prevailing hypothesis, which this study indirectly supports, is that baseline endogenous oxytocin levels may determine who responds—meaning a simple blood measurement could eventually help clinicians select patients likely to benefit from oxytocin-based therapy rather than applying it uniformly. The finding that oxytocin’s relationship with symptoms is strongest in milder, earlier cases further suggests that any such intervention may need to begin early in the disease course, before irreversible neurobiological changes take hold. Notably, the patients in this cohort had a mean illness duration of only 11.5 months, placing them squarely within the window that overlaps with first-episode psychosis, where hormonal abnormalities have been independently linked to cognitive impairment and drug response.

The authors are candid about the limitations. All participants came from a single geographic region, the groups were not sex-balanced, and medication histories—though documented in type and dose equivalent—did not record when treatment began or how long it lasted, leaving open the possibility that antipsychotic exposure influenced hormone levels. The mild-symptom subgroup that produced the striking correlation numbered only 28 patients, small enough that the finding demands replication. And because the design was cross-sectional, no claim of causality can be made. Still, the study was prospectively registered with the Chinese Clinical Trial Registry in 2019, and its convergence of a robust group difference, a quantified diagnostic threshold, and a stage-specific correlation with symptom severity gives the field something concrete to test. If longitudinal studies confirm that plasma oxytocin falls before or during the earliest phase of psychotic illness, a routine blood draw could become part of the toolkit for catching schizophrenia when intervention still has the power to change its trajectory.

Subject of Research: Plasma oxytocin levels as a biomarker for early-onset schizophrenia severity and early intervention

Article Title: The role of plasma oxytocin in early-onset schizophrenia: implications for disease severity and early intervention

Article References: Wang, X., Wang, S., Guo, J., Zhang, J., Chen, B., Li, J., Zhao, Z., Li, C., Cai, H., & Liu, S. (2026). The role of plasma oxytocin in early-onset schizophrenia: implications for disease severity and early intervention. BMC Psychiatry, 26(1), Article 770. https://doi.org/10.1186/s12888-026-08364-8

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08364-8

Keywords: early-onset schizophrenia, oxytocin, biomarker, plasma oxytocin, PANSS, childhood trauma, extraversion, neuropeptide, early intervention, psychiatry, neuroendocrinology, logistic regression