A novel sex-specific role for 27-hydroxycholesterol in early-onset schizophrenia

Abstract

Background

Although Schizophrenia is a serious mental illness which more common in adults, it can also manifest in childhood with disease onset before age 18 as “early-onset schizophrenia” (EOS). While our previous research identified circulating oxysterols—24(S)-hydroxycholesterol (24 S-OHC) and 27-hydroxycholesterol (27-OHC)—as potential biomarkers in adult schizophrenia, their role in EOS remains unknown. This study aimed to characterize the developmental trajectories of these oxysterols during early life and evaluate their clinical relevance in EOS, with particular emphasis on sex-specific effects.

Methods

We recruited 142 healthy individuals (aged 1 to 41 years) to define age-related oxysterol patterns and 71 EOS patients (aged 4 to 18 years) to identify disease-associated changes. Plasma concentrations of 24 S-OHC and 27-OHC were measured using liquid chromatography-tandem mass spectrometry.

Results

Healthy participants exhibited significant age-dependent variations, with younger individuals showing higher 24 S-OHC and lower 27-OHC levels. While EOS patients showed no overall differences in oxysterol levels compared to age-matched healthy controls, a striking sex-based divergence emerged: male patients exhibited significantly elevated levels of 27-OHC compared to females. Additionally, the 24 S-OHC/27-OHC ratio showed a positive correlation with negative symptom severity, which did not survive correction for multiple comparisons but may still indicate a potential relationship. Notably, sex-stratified analysis revealed opposing correlations between 27-OHC levels and PANSS scores, most prominently for positive symptoms.

Conclusions

This study represents the first evidence of sexually dimorphic oxysterol regulation in EOS and highlights 27-OHC as a sex-sensitive factor linked to clinical symptom expression, even in the absence of group-level oxysterol alterations.

Introduction

Schizophrenia is a severe neuropsychiatric disorder that typically manifests in early adulthood and is characterized by persistent and relapsing symptoms [1, 2]. While the majority of cases present in adulthood, approximately 4% of patients experience onset during childhood or adolescence, a condition designated as early-onset schizophrenia (EOS) [3]. Defined by symptom emergence before age 18, EOS is increasingly recognized as a distinct clinical entity associated with more pronounced negative symptoms and poorer functional outcomes compared to adult-onset forms [4, 5]. Despite its clinical significance, the neurobiological mechanisms underlying EOS remain inadequately understood, highlighting the need for further research into its pathogenic pathways.

Accumulating evidence supports a neurodevelopmental model of schizophrenia, in which early disruptions in brain maturation predispose individuals to later dysfunction [6]. A 30-year longitudinal study proposed that the pathogenesis of schizophrenia unfolds across two critical developmental phases spanning childhood through early adolescence [7]. Given its notably earlier onset, EOS likely involves more profound neurodevelopmental alterations [8, 9].

Cholesterol homeostasis plays an essential role in central nervous system (CNS) development and function, contributing to myelination, membrane stability, dendritic arborization, and axonal integrity [10, 11]. Perturbations in cholesterol metabolism may disrupt synaptic function and neuronal development, potentially contributing to the pathophysiology of schizophrenia [12, 13]. For example, reduced cholesterol levels have been correlated with enhanced thalamohippocampal connectivity in first-episode schizophrenia, suggesting lipid-mediated hippocampal dysregulation [14]. Furthermore, age- and disease-related declines in synaptic and astrocytic programs involving cholesterol synthesis provide additional support for this connection [15].

As the most cholesterol-rich organ, the brain maintains autonomous cholesterol metabolism due to the blood-brain barrier (BBB), which limits direct exchange with peripheral circulation [16]. Cerebral cholesterol elimination occurs primarily through its conversion to 24(S)-hydroxycholesterol (24 S-OHC), which readily crosses the BBB [17]. In contrast, peripheral tissues produce 27-hydroxycholesterol (27-OHC), the most abundant circulating oxysterol. Both metabolites regulate neurodevelopmental processes, including myelination and neuronal differentiation [18, 19]. Experimental evidence indicates that 24 S-OHC deficiency leads to significant cognitive impairment in murine models [20], while 27-OHC disrupts hippocampal synaptic function [21]. The sex difference in 27-OHC is primarily driven by the differential regulation of its synthesizing (CYP27A1) and metabolizing (CYP7B1) enzymes by sex hormones, particularly estrogen, which alters both the systemic flux of 27-OHC and its interaction with estrogen receptors (ERs) to modulate inflammatory responses [22–24].

Our previous research demonstrated elevated 24 S-OHC and reduced 27-OHC levels in adult schizophrenia patients, with similar patterns observed in clinical high-risk individuals, suggesting the presence of early neurodevelopmental alterations [25]. Corroborating these findings, Guidara et al. reported correlations between 24 S-OHC levels, the Positive and Negative Syndrome Scale (PANSS) scores, and cognitive performance in drug-naïve patients [26], while Moreno et al. identified elevated oxysterol levels in ultra-high-risk populations [27]. Notably, no previous study has examined oxysterol profiles in pediatric schizophrenia, creating a critical knowledge gap regarding their role in EOS.

Given the essential functions of cholesterol metabolites in neurodevelopment, this study aims to: (1) characterize the developmental trajectories of 24 S-OHC (primarily brain-derived) and 27-OHC (mainly peripheral) in both healthy individuals and EOS patients, and (2) investigate their relationships with demographic and clinical parameters, with particular emphasis on sex-specific effects. These investigations may provide novel insights into EOS pathophysiology and identify potential biomarkers for early intervention.

Materials and methods

Subjects

A total of 213 participants, comprising schizophrenia patients and healthy controls (HCs), were recruited from Beijing Anding Hospital, China. The cohort included 142 healthy individuals (aged 1–41 years) for establishing demographic-oxysterol relationships, and 71 patients diagnosed with EOS, along with matched HCs. All EOS patients were of Han Chinese ethnicity, aged 4–18 years. Diagnoses were confirmed by two senior psychiatrists according to Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) criteria using a consensus approach. At the time of blood sampling, all EOS patients were receiving antipsychotic treatment. Detailed information on medication type and dose ranges is provided in Table 1.

Table 1.

Clinical characteristic of the patients with EOS

	Healthy controls	EOS patients	p
n	49	71
Age	13.39 ± 4.37	14.33 ± 2.61	0.177
Gender (Male/Female)	26/23	36/35	0.857
Onset Age		12.68 ± 2.43
Duration of Disease (month)		18.41 ± 17.84
Family History (Yes/No)	0/49	14/57	0.001
Medication (Frequency/Dose Range)
Olanzapine		34.9%/5–20 mg
Aripiprazole		15.9%/7.5–17.5 mg
Quetiapine		11.1%/200–600 mg
Paliperidone		3.2%/6 mg
Risperidone		20.6%/0.25-7 mg
Clozapine		9.5%/150–350 mg
Amisulpride		4.8%/600–1200 mg
Clinical Assessment
PANSS-P		24.39 ± 8.71
PANSS-N		23.17 ± 7.85
PANSS-G		36.98 ± 8.28
PANSS-T		84.55 ± 16.58
PSPT		47.79 ± 17.83
CGI		4.65 ± 1.00

EOS = early-onset schizophrenia; PANSS = Positive and Negative Syndrome Scale; P = positive symptom; N = negative symptom; G =general psychopathology; T = total scores; PSPT = Personal and Social Performance Scale Test; CGI = Clinical Global Impressions Severity Scale

Exclusion criteria for EOS patients were: (1) electroconvulsive therapy within the preceding 60 days; (2) current substance abuse; (3) significant neurological disorders; (4) use of lipid-lowering agents; and (5) major cardiovascular, renal, or hepatic diseases. HCs were recruited via advertisement and screened to exclude those meeting DSM-5 diagnostic criteria. Additional exclusion criteria for HCs included: (1) personal history of psychiatric illness or psychotropic drug use; (2) first-degree family history of mental disorders; (3) substance abuse; (4) significant neurological illness; (5) lipid-lowering medication use; and (6) serious systemic and metabolic diseases.

The study was approved by the Ethical Committee of Beijing Anding Hospital, Capital Medical University (Approval No. 2016103FS-2; Date: 20 October 2016). Written informed consent was obtained from all participants (Fig. 1).

Fig. 1.

An overview of participant selection. EOS, early-onset schizophrenia; DSM-5, Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; PANSS, the Positive and Negative Syndrome Scale; CGI, the Clinical Global Impressions Severity scale; PSP, Personal and Social Performance Scale

Clinical assessments

Psychiatric symptom severity in EOS patients was evaluated using the PANSS and the Clinical Global Impressions Severity scale (CGI). Social functioning was assessed with the Personal and Social Performance Scale (PSP), a single-item instrument with a maximum score of 100 [28].

Sample collection

Peripheral blood samples were collected from all participants after an overnight fast into tubes containing EDTA and then centrifuged at 3000 rpm for 10 min at room temperature. The plasma samples were then collected and stored at -80 °C until oxysterol detection.

Oxysterols analysis

Levels of plasma cholesterol in EOS patients were measured by an enzymatic method (Modular; Roche Diagnostics) as previously reported [29]. The plasma levels of 24S-OHC and 27-OHC were detected using a previously described method [25]. Briefly, 50 µL of plasma was pipetted into an Eppendorf tube, followed by the addition of 10 ng of an internal standard (D5/D7 deuterium-labeled cholesterol). The extraction conditions were as follows: 200 µL of acidic buffer solution (pH = 3) containing 50 mM ammonium acetate and 1% formic acid, and 1 mL of methyl tert-butyl ether. The mixture was frozen at -80°C, after which the supernatant was transferred to a new Eppendorf tube and dried at 30°C. Then, the chloroform solution (50 µL) containing 12.6 g/L N, N’-diisopropylcarbodiimide, 12.2 g/L 4-dimethylaminopyridine, and 12.4 g/L nicotinamide was added for the derivatization reaction at 35 °C for 2 h. After drying at 30 °C, 100 µL of methanol was added to dissolve the powder for further analysis. Oxysterol quantification was performed using liquid chromatography-tandem mass spectrometry with an Agilent G1312B Pump and a C18 column.

Statistical analysis

Data were analyzed using SPSS 20.0 and figures were generated using GraphPad Prism 10.1.2. Group differences in demographic and oxysterol measures between EOS patients and HCs were assessed with χ² tests or independent t-tests, as appropriate. Sex differences in oxysterol levels were evaluated via two-way ANOVA with age as a covariate, followed by post hoc Tukey tests for mean comparisons. Correlations between oxysterol levels and clinical variables in EOS patients were examined using Pearson or Spearman tests based on data distribution. Age-related trends in oxysterol levels across healthy subjects were modeled using second-order polynomial regression. A p-value < 0.05 was considered statistically significant.

Results

Age-Dependent Oxysterol Dynamics in Health and EOS

Analysis of healthy individuals (n = 142, aged 1–41 years) revealed pronounced age-related oxysterol dynamics. Plasma 24 S-OHC levels peaked within five years post-birth, declining precipitously thereafter, with the most substantial decrease occurring before age 15 (Fig. 2A). This decreasing trajectory was consistent between sexes (Fig. 2B). Conversely, 27-OHC levels demonstrated a gradual age-dependent increase, with significant elevation before age 10 (Fig. 2D). Although both sexes followed this pattern, males exhibited a steeper increase rate (Fig. 2E). The 24 S-OHC/27-OHC ratio mirrored 24 S-OHC’s developmental pattern, decreasing dramatically before age 15 (Fig. 2G), with comparable sex trends (Fig. 2H).

Fig. 2.

Age-related changes of plasma oxysterol levels in healthy individuals and EOS patients. Absolute values of oxysterols at different ages in healthy individuals were shown in A, D, and G, respectively. Regression curves between 24 S-OHC, 27-OHC, ratio value (24 S-OHC/27-OHC), and age in healthy individuals, differentiated by sex, were shown in B, E, and H, respectively. Regression curves between 24 S-OHC, 27-OHC, ratio value (24 S-OHC/27-OHC), and age in both HCs and EOS patients were shown in C, F, and I, respectively. Plasma 24 S-OHC levels and ratio value decreased, while 27-OHC levels increased with age. It was worth noting that the 27-OHC levels in males increased slightly faster than those in females in healthy individuals. **, ***, and **** were significantly different from the corresponding values of the previous age at p < 0.01, 0.001, and 0.0001, respectively. EOS, early-onset schizophrenia; 24 S-OHC, 24(S)-hydroxycholesterol; 27-OHC, 27-hydroxycholesterol

The EOS cohort comprised 71 patients (36 males, 35 females) aged 4–18 years (mean ± SD: 14.3 ± 2.6), among whom 64 had complete clinical assessments. When compared with matched healthy controls (HCs, n = 49 selected from the original cohort), no significant differences were observed in age or sex distribution (p = 0.177 and p = 0.857, respectively; see Table 1). EOS patients displayed age-dependent trends similar to those of HCs: significant negative correlations with age were identified for both 24 S-OHC levels (EOS: r = − 0.340, p = 0.004; HCs: r = − 0.512, p < 0.001) and the 24 S-OHC/27-OHC ratio (EOS: r = − 0.506, p < 0.001; HCs: r = − 0.529, p < 0.001; Fig. 2C and I). In contrast, no significant correlation with age was found for 27-OHC in either group (EOS: r = 0.222, p = 0.067; HCs: r = 0.278, p = 0.056; Fig. 2F).

Sex-specific 27-OHC dysregulation in EOS

Comparative analysis revealed no significant differences in plasma oxysterol concentrations between EOS patients and HCs (24 S-OHC: 38.64 ± 13.87 vs. 41.85 ± 17.42 ng/mL; 27-OHC: 32.63 ± 11.98 vs. 34.33 ± 10.26 ng/mL; ratio: 1.27 ± 0.58 vs. 1.33 ± 0.66; all p > 0.05) (Fig. 3A, C, E).

Fig. 3.

Plasma oxysterol profile in patients with EOS. The plasma 24 S-OHC levels, 27-OHC levels, and ratio value (24 S-OHC/27-OHC) in HCs and EOS groups were shown in A, C, and E, respectively. No significant difference in oxysterol levels was found between the two groups. The sex difference of plasma 24 S-OHC levels, 27-OHC levels, and ratio value (24 S-OHC/27-OHC) in HCs and EOS groups was shown in B, D, and F, respectively. No significant sex difference was found in 24 S-OHC and ratio value between the two groups. Notably, male patients showed higher 27-OHC levels than female patients, while no significant sex difference was found in HCs. EOS, early-onset schizophrenia; 24 S-OHC, 24(S)-hydroxycholesterol; 27-OHC, 27-hydroxycholesterol

Sex-stratified analysis uncovered a particularly noteworthy finding: male patients showed significantly elevated 27-OHC levels compared to females (F = 10.748, p = 0.001, Fig. 3D), despite comparable 24 S-OHC levels between sexes (Fig. 3B). Correspondingly, the 24 S-OHC/27-OHC ratio was significantly reduced in male patients compared with females (F = 9.539, p = 0.003, Fig. 3F). This sex effect was independent of total cholesterol, High-Density Lipoprotein (HDL) and Low-Density Lipoprotein (LDL) levels, which did not differ between male and female patients (Table 2, all p > 0.05).

Sex-Modulated Clinical Correlations of Oxysterols

Initial analysis identified a negative association between disease duration and the 24 S-OHC/27-OHC ratio (r = -0.268, p = 0.037), which became non-significant after age adjustment (r = -0.029, p = 0.825). Most notably, the 24 S-OHC/27-OHC ratio positively correlated with PANSS negative symptom scores (r = 0.266, p = 0.040), suggesting a specific relationship with negative symptom severity (Table 3). It should be noted that these correlations have not been significantly adjusted for multiple comparisons (p > 0.006).

Table 3.

Associations between oxysterols and clinical assessments in patients with EOS

	24 S-OHC	27-OHC	24 S-OHC/27-OHC
Onset age*	0.092 (-0.216)	0.491 (0.089)	0.092 (-0.217)
Duration (months)*	0.181 (-0.172)	0.378 (0.114)	0.037 (-0.268)
PANSS-P	0.410 (0.107)	0.853 (-0.024)	0.086 (0.224)
PANSS-N	0.191 (0.170)	0.653 (-0.059)	0.040 (0.266)
PANSS-G	0.636 (-0.062)	0.542 (0.080)	0.287 (-0.140)
PANSS-T	0.425 (0.104)	0.999 (0.000)	0.199 (0.168)
PSPT	0.593 (-0.070)	0.480 (0.093)	0.317 (-0.132)
CGI*	0.284 (0.139)	0.253 (-0.149)	0.188 (0.172)

Values represent p-value (correlation coefficient). Statistically significant results (p < 0.05) are shown in bold

EOS = early-onset schizophrenia; PANSS = Positive and Negative Syndrome Scale; P = positive symptom; N = negative symptom; G = general psychopathology; T = total scores; PSPT = Personal and Social Performance Scale Test; CGI = Clinical Global Impressions Severity Scale

* represent Spearman correlation analysis, others are Pearson correlation analysis

While female patients exhibited higher PANSS positive symptom scores than males (p = 0.045, Table 2) with medium effect size (0.66), no other clinical differences emerged. Sex-specific correlation analyses revealed divergent oxysterol-clinical relationships: although the overall cohort showed the aforementioned correlation with negative symptoms, subanalysis by sex revealed no significant associations (Fig. 4). Strikingly, male and female patients demonstrated opposite directional relationships between 27-OHC levels and PANSS scores, particularly for positive symptoms (Fig. 4B).

Table 2.

Sex-difference in clinical assessments and plasma levels of oxysterol in patients with EOS

	Male Patients	Female Patients	t	p	Effect Size
n	37	35
Age	14.11 ± 3.07	14.58 ± 2.00	-0.773	0.442	0.34
Onset Age	12.41 ± 2.92	13.00 ± 1.70	-0.997	0.323	0.50
Duration	18.53 ± 16.58	18.26 ± 19.46	0.058	0.954	0.02
Clinical Assessment
PANSS-P	22.27 ± 7.15	26.73 ± 9.75	-2.053	0.045	0.66
PANSS-N	23.00 ± 8.88	23.36 ± 6.68	-0.184	0.855	0.08
PANSS-G	36.12 ± 9.06	37.93 ± 7.37	-0.865	0.390	0.35
PANSS-T	81.39 ± 14.90	88.03 ± 17.86	-1.607	0.113	0.53
PSPT	46.91 ± 17.83	48.79 ± 18.11	-0.412	0.682	0.15
CGI	4.61 ± 1.06	4.70 ± 0.95	-0.369	0.714	0.14
Cholesterol detection
24 S-OHC (ng/ml)	36.64 ± 11.87	40.35 ± 15.63	-1.110	0.271	0.34
27-OHC (ng/ml)	37.26 ± 13.52	29.03 ± 7.00	3.158	0.003	1.69
24 S-OHC/27-OHC	1.09 ± 0.52	1.43 ± 0.58	-2.560	0.013	0.84
Total cholesterol (mmol/L)	3.92 ± 0.68	4.01 ± 0.61	-0.537	0.594	0.21
HDL (mmol/L)	1.32 ± 0.29	1.31 ± 0.26	0.131	0.896	0.06
LDL (mmol/L)	2.21 ± 0.58	2.26 ± 0.54	-0.382	0.704	0.13

Values represent mean ± S.D. Statistically significant results (p < 0.05) are shown in bold

EOS = early-onset schizophrenia; PANSS = Positive and Negative Syndrome Scale; P = positive symptom; N = negative symptom; G = general psychopathology; T = total scores; PSPT = Personal and Social Performance Scale Test; CGI = Clinical Global Impressions Severity Scale; HDL = high density lipoprotein; LDL = low density lipoprotein

Fig. 4.

Relationship between plasma levels of oxysterol and PANSS scores in patients with EOS. The first line showed the associations between 24S-OHC level, 27-OHC level, ratio value (24S-OHC/27-OHC), and PANSS positive symptom scores (A, B, and C, respectively). Correspondingly, the second, third, and last lines showed the correlation between oxysterol levels and PANSS negative symptoms (D, E, and F, respectively), PANSS general psychiatric symptoms (G, H, and I, respectively), and PANSS total scores (J, K, and L, respectively). Notably, male and female patients showed opposite directional relationships between 27-OHC levels and PANSS scores, particularly for positive symptoms. PANSS, the Positive and Negative Syndrome Scale; 24S-OHC, 24(S)-hydroxycholesterol; 27-OHC, 27-hydroxycholesterol

Discussion

This comprehensive investigation yields several novel insights into cholesterol metabolism in EOS. First, we establish that EOS patients show parallel developmental oxysterol trajectories to healthy individuals, arguing against general oxysterol dysregulation in EOS pathogenesis. Second, and most importantly, we identify a previously unrecognized sex-specific alteration in 27-OHC metabolism unique to EOS patients. Third, we demonstrate clinically relevant correlations between oxysterol ratios and negative symptoms, with intriguing sex-dependent relationships between 27-OHC and symptom profiles. A graphical summary is provided to integrate the observed sex-specific alterations in 27-OHC levels and their opposing associations with symptom severity in EOS (Fig. 5).

Fig. 5.

Graphical summary illustrating sex-specific differences in circulating 27-hydroxycholesterol levels and their associations with clinical symptoms in early-onset schizophrenia (EOS). Male patients exhibited higher 27-OHC levels and a negative correlation between 27-OHC and PANSS positive symptom scores, whereas female patients showed lower 27-OHC levels and a positive correlation with PANSS positive symptoms scores

The age-dependent patterns of 24 S-OHC we observed align with established neurodevelopmental processes [30, 31]. The 2.3-fold higher 24 S-OHC levels in adolescents versus adults reflect active myelination and neurodevelopment, stabilizing after cerebral maturation [30]. The parallel trajectories in EOS and HCs suggest intact overall neurodevelopmental cholesterol metabolism in EOS. However, circulating oxysterols represent global measures that may mask regional abnormalities in structures implicated in EOS, particularly prefrontal and hippocampal regions [32, 33].

The most striking finding emerges from sex-specific analyses: the significant elevation of 27-OHC in male EOS patients represents the first evidence of sexually dimorphic oxysterol regulation in early schizophrenia. This effect was independent of total cholesterol levels, suggesting altered enzyme activity or clearance mechanisms rather than substrate availability. Notably, this sex difference was absent in healthy controls, indicating a disease-specific alteration rather than general sexual dimorphism.

Several mechanisms may explain this sex-specific effect, including variations in the rate of brain maturation and brain structural differences between males and females [34–36]. In addition, the protective effect of estrogen in females may have also contributed to the sex difference in schizophrenia [37]. Estrogen has been shown to induce the expression and activity of CYP7B1, a critical enzyme responsible for the metabolic inactivation of 27-OHC, thereby facilitating its conversion to downstream bile acid intermediates and promoting clearance [38, 39]. Through this estrogen-CYP7B1 axis, females may maintain lower circulating and tissue levels of 27-OHC, conferring a relative protective effect. In contrast, reduced estrogen signaling in males may result in diminished CYP7B1-mediated metabolism, leading to accumulation of 27-OHC and enhanced biological impact. Supporting this, recent evidence indicates more pronounced dyslipidemia in male schizophrenia patients [40], with additional modulation by inflammatory states, oxidative stress, and lipoprotein profiles [41, 42]. The opposing correlations between 27-OHC and positive symptoms in male versus female patients further suggest sex-dependent pathophysiological roles for this metabolite.

The positive correlation between the 24 S-OHC/27-OHC ratio and negative symptoms suggests that balanced central-peripheral cholesterol metabolism may influence negative symptom pathogenesis. This aligns with adult literature linking cholesterol disturbances to negative symptoms [43–45], possibly through neurotransmitter systems modulated by oxysterols-including dopamine (via tyrosine hydroxylase induction [46]), glutamate (through NMDA receptor modulation [47–49]), and serotonin (via lipid raft disruption [50, 51]).

Beyond mechanistic insights, these findings also suggest potential clinical relevance of oxysterol measures in EOS. Although overall oxysterol levels did not distinguish patients from HCs, the observed sex-specific alterations in 27-OHC and its symptom-related associations indicate that oxysterol profiling may serve as a sex-informed biomarker to aid in patient stratification and symptom characterization. Moreover, given the regulatory roles of CYP27A1 and CYP7B1 in 27-OHC metabolism and the interaction of 27-OHC with estrogen receptor signaling, these pathways may represent promising targets for future therapeutic modulation. Such approaches could enable more personalized interventions that account for biological sex and lipid-related vulnerability in EOS. It is noteworthy that all EOS patients were receiving antipsychotic medication at the time of sampling. It is well established that antipsychotic medications can influence systemic lipid metabolism and disrupt metabolic homeostasis, which may indirectly alter circulating levels of oxysterols, including 27-OHC [52, 53]. However, to date, no studies have specifically investigated the direct impact of antipsychotic drugs on 27-OHC concentrations. Therefore, when interpreting the present findings, the potential contribution of medication-related metabolic effects cannot be excluded.

This study has several limitations. Its cross-sectional design precludes assessment of longitudinal dynamics. The sample size was relatively small, and cholesterol levels in healthy controls were not measured, limiting a comprehensive evaluation of cholesterol homeostasis. Potential confounding factors such as body composition, metabolic parameters, and inflammatory markers were not accounted for. Additionally, all EOS patients were receiving antipsychotic treatment at the time of sampling, which may have indirectly influenced circulating oxysterol levels. Future studies should include larger, drug-free cohorts and longitudinal measures to better clarify disease-related, sex-specific, and treatment-related influences on oxysterol regulation.

Conclusion

This study provides initial evidence for sex-specific oxysterol alterations in EOS, revealing elevated 27-OHC levels specifically in male patients and sex-dependent relationships between oxysterols and clinical symptoms across sexes. While some clinical correlations did not survive correction for multiple comparisons and should therefore be interpreted with caution, the observed sex-specific patterns suggest that sexually dimorphic cholesterol metabolism may be relevant to differential disease expression in EOS. While overall oxysterol levels may not distinguish EOS patients from healthy individuals, the identified sex-specific patterns offer new avenues for understanding disease mechanisms and developing personalized interventions.

Author contributions

Sun ZL, Liu MX, and Zheng Y obtained funding for this study. Yan JJ and Wu YZ enrolled the participants and collected the plasma samples. Sun ZL and Liu MX detected the oxysterol levels, analyzed the data, and wrote the manuscript. Li YH and Wang HX analyzed the data. Li R and Zheng Y designed the study and edited the manuscript. Tang YL contributed to data interpretation, and he critically revised the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (no.82171525, 81671248), R&D Program of Beijing Municipal Education Commission (KM202310025007), Beijing Hospitals Authority Innovation Studio of Young Staff Funding Support (202336), and Beijing Municipal Science & Technology Commission (no. Z161100000216151).

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Ethical Committee of Beijing Anding Hospital, Capital Medical University, on October 20, 2016 (No. 2016103FS-2) in accordance with the Declaration of Helsinki. Clinical trial number: not applicable. Written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.