Asymmetry in paracingulate sulcus (PCS) length and superior temporal sulcus (STS) depth was examined in schizophrenia spectrum patients stratified by auditory verbal hallucination (AVH) frequency. An ordinal reduction in sulcal asymmetry was observed with increasing AVH frequency, reaching statistical significance only for STS depth. The decline in effect size from frequent to less frequent AVH may explain prior inconsistent findings from binary AVH classifications that disregard clinical heterogeneity.
The motivation-and-pleasure (MAP) factor affects social functioning more than the expressivity (EXP) factor in schizophrenia patients, but less is known regarding the differential associations of negative symptom factors on social functioning in treatment-resistant schizophrenia (TRS) patients. Four-hundred-and-two schizophrenia (230 TRS and 172 non-TRS) patients completed the Clinical-Assessment-Interview-for-Negative-Symptoms (CAINS), the Positive-and-Negative-Syndrome-Scale (PANSS), and the Social-and-Occupational-Functioning Assessment-Scale (SOFAS). A regularized-partial-correlation network and a flow-diagram were constructed to examine the role of negative symptoms, and their direct/indirect associations with social functioning in TRS patients. The network structure and global strength were compared between TRS and non-TRS patients. We further compared the regularized-partial-correlation network of CAINS items in TRS patients with that in non-TRS patients. The results showed that the regularized-partial-correlation network of clinical symptoms, illness-duration and social functioning in TRS patients was highly stable. The MAP factor, EXP factor, PANSS-disorganized factor, and SOFAS were central in the network. Moreover, the flow-diagram indicated that the MAP factor had the strongest and most direct association with social functioning in TRS patients, followed by the PANSS-disorganized factor. However, the EXP factor largely contributed to the variance of social functioning via the MAP factor. The symptom-domain network in TRS patients showed stronger connectivity than the network of non-TRS patients. Comparing the two CAINS-item networks, TRS patients’ negative symptoms showed more cross-item connections than non-TRS patients. Our findings concur with previous network analysis results in patients with schizophrenia, and extended the role of negative symptoms to the TRS population. Interventions targeting negative symptoms in TRS patients are warranted.
Reliable neuroimaging biomarkers for the diagnosis of schizophrenia and the prediction of treatment response remain limited, particularly at the individual level. Radiomics converts neuroimaging data into high-dimensional quantitative data that may capture subtle disease-related patterns not detectable by conventional analyses. This study aimed to meta-analyze the diagnostic and treatment-response prediction performance of radiomics-based neuroimaging in schizophrenia. A systematic search of PubMed, EMBASE, IEEE Xplore, and Scopus was conducted until 09/17/2026, following PRISMA-DTA guidelines and a pre-registered protocol (PROSPERO/ CRD420261286782). Eligible studies applied radiomics combined with machine learning/quantitative neuroimaging data (MRI/PET) for schizophrenia spectrum diagnosis/treatment response classification. A random-effects meta-analysis was performed to pool the area under the receiver operating characteristic (ROC) curve (AUC). Hierarchical summary ROC (HSROC) models were used to estimate pooled sensitivity and specificity. Risk of bias and reporting quality were assessed using QUADAS-2 and METRICS. Fourteen studies were included. For diagnostic discrimination, radiomics-based models showed good performance (AUC = 0.81, 95% CI [0.73; 0.88], I2 = 84.2%,k = 8). The HSROC model yielded a pooled mean sensitivity = 0.736, 95% CI [0.666; 0.795], and specificity = 0.735, 95% CI[0.668; 0.793]. Radiomics also showed good performance in differentiating treatment response, including longitudinal prediction (AUC = 0.80) and cross-sectional discrimination (AUC = 0.88). Although substantial methodological heterogeneity among studies emerged, overall study quality was moderate/good. Radiomics-based neuroimaging demonstrated good diagnostic and treatment-response discrimination in schizophrenia, supporting their potential complementary function for precision psychiatry. However, limited external validation may currently constrain clinical translation, since less rigorous validation procedures were associated with more optimistic performances, as displayed by our analyses. Future studies should prioritize standardized radiomics workflows, prospective designs, integration with clinical markers, and rigorous validation.
Amino acid dysregulation has been implicated in schizophrenia, but baseline plasma amino acid profiles in first-episode schizophrenia and their clinical relevance remain unclear. We hypothesized that antipsychotic-naïve patients would show distinct amino acid alterations associated with subsequent clinical improvement. Plasma levels of 31 amino acids were measured by ultra-performance liquid chromatography–tandem mass spectrometry in 80 antipsychotic-naïve patients with first-episode schizophrenia and 40 healthy controls. Between-group differences were assessed using Mann–Whitney U tests with Benjamini–Hochberg false discovery rate correction. Amino acid-derived indices, receiver operating characteristic analysis, and Spearman correlation analyses were used to characterize metabolic patterns, discriminatory performance, and associations with PANSS improvement. Discriminatory performance was further evaluated using repeated stratified 10-fold cross-validation with feature selection performed independently within each training set, and multiple testing in the exploratory clinical correlation analyses was controlled using the Benjamini–Hochberg procedure. Eleven amino acids differed significantly between patients and controls after false discovery rate correction. Patients showed lower kynurenine, arginine, homoarginine, aminoadipic acid, and methionine, and higher ornithine, alanine, glutamate, serine, glycine, and GABA. Arginine/ornithine, kynurenine/tryptophan, citrulline/arginine, and tyrosine/phenylalanine ratios also differed between groups. A multivariable amino acid model showed an apparent AUC of 0.949, while repeated stratified 10-fold cross-validation yielded a mean out-of-fold AUC of 0.869 (SD = 0.011). In exploratory analyses among patients with paired PANSS data (n = 26), the baseline glutamate/GABA ratio was negatively associated with improvement in PANSS positive symptoms and remained significant after false discovery rate correction (ρ = −0.618, q = 0.015). However, this association was attenuated and no longer significant after additional adjustment for baseline PANSS positive score. Antipsychotic-naïve first-episode schizophrenia was associated with plasma amino acid abnormalities involving arginine-related metabolism, kynurenine pathway activity, aromatic amino acid conversion, and glutamate/GABA-related balance. These findings support the relevance of peripheral amino acid metabolism in first-episode schizophrenia but require validation in independent cohorts.
Observational evidence is increasingly central to prognostic precision in early intervention in mental health, particularly in Clinical High-Risk for Psychosis (CHR-P) populations where trajectories are heterogeneous and clinical decisions unfold sequentially. In this context, baseline pharmacological exposure should not be treated merely as a nuisance variable or dismissed because it may reflect confounding by indication, confounding by severity, or protopathic bias. These mechanisms limit causal interpretation, but they may also explain why treatment exposure carries prognostic information. Antipsychotic, antidepressant, benzodiazepine, or combined exposure at CHR-P inception may encode prior clinical deterioration, symptom burden, comorbidity, treatment-seeking pathways, clinician concern, partial treatment response, or pharmacologically attenuated expression of emerging psychosis. Current prognostic models often prioritize demographic, symptomatic, cognitive, and functional predictors while giving insufficient attention to baseline pharmacotherapy, despite its availability, clinical interpretability, and temporal relevance. Incorporating medication exposure, dose, duration, indication, adherence, severity, and longitudinal treatment changes may improve risk stratification, reduce hidden heterogeneity, and support more precise evaluation of preventive interventions. The key implication is methodological and clinical: treatment exposure should be interpreted with causal humility but prognostic pragmatism. In CHR-P research, medication history is not simply a source of bias; it is part of the longitudinal clinical record and may help identify prognostically non-equivalent subgroups.
