Brain Age Disparities in Psychosis Across DSM Diagnoses and B-SNIP Biotypes.

Brain Age Disparities in Psychosis Across DSM Diagnoses and B-SNIP Biotypes - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Schizophr Bull . 2025 May 31;52(1):sbaf022. doi: 10.1093/schbul/sbaf022 Brain Age Disparities in Psychosis Across DSM Diagnoses and B-SNIP Biotypes Liisi Promet 1, 2 , Shashwath A Meda 3 , Ney Alliey-Rodriguez 4, 5 , Brett A Clementz 6 , Elliott S Gershon 7 , Scot K Hill 8 , Elena I Ivleva 9 , Sarah K Keedy 10 , Matcheri S Keshavan 11 , Jennifer E McDowell 12 , David A Parker 13, 14 , Carol A Tamminga 15 , Godfrey D Pearlson 16, 17, ✉ Author information Article notes Copyright and License information 1 Department of Psychiatry, School of Medicine, Yale University, New Haven, CT 06511, United States 2 Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States 3 Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States 4 Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States 5 Institute of Neuroscience, University of Texas Rio Grande Valley, Harlingen, TX 78550, United States 6 Department of Neuroscience, Bio-Imaging Research Center, University of Georgia, Athens, GA 30602, United States 7 Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States 8 Department of Psychology, Rosalind Franklin University of Medicine and Science, Chicago, IL 60064, United States 9 Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX 75390, United States 10 Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States 11 Department of Psychiatry, Beth Israel Deaconess Medical Centre, Harvard Medical School, Boston, MA 02215, United States 12 Department of Psychology, University of Georgia, Athens, GA 30602, United States 13 Department of Neuroscience, Bio-Imaging Research Center, University of Georgia, Athens, GA 30602, United States 14 Department of Psychology, Emory University, Atlanta, GA 30322, United States 15 Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX 75390, United States 16 Department of Psychiatry, School of Medicine, Yale University, New Haven, CT 06511, United States 17 Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States ✉ To whom correspondence should be addressed: Godfrey D. Pearlson, Olin Neuropsychiatry Research Center, Institute of Living, 200 Retreat Avenue, Whitehall Building, Hartford, CT 06106, United States ( Godfrey.Pearlson@yale.edu ) Collection date 2026 Jan. © The Author(s) 2025. Published by Oxford University Press on behalf of the Maryland Psychiatric Research Center. All rights reserved. For commercial re-use, please contact reprints@oup.com for reprints and translation rights for reprints. All other permissions can be obtained through our RightsLink service via the Permissions link on the article page on our site—for further information please contact journals.permissions@oup.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model ( https://academic.oup.com/pages/standard-publication-reuse-rights ) PMC Copyright notice PMCID: PMC12809854  PMID: 40448350 Abstract Background and Hypothesis The brain age gap (BAG) quantifies the difference between predicted brain age and chronological age. Prior research implicates higher BAG in psychotic disorders, suggesting accelerated brain aging. We hypothesized distinct brain aging profiles among biological subtypes of psychosis and intermediate BAG in their relatives. Study Design Brain age gap values were quantified in 348 healthy controls (HCs), 950 psychosis probands classified by both DSM diagnoses of psychotic bipolar disorder, type I (BP, n = 247), schizoaffective disorder (SAD, n = 313), and schizophrenia (SZ, n = 390), and Bipolar-Schizophrenia Network for Intermediate Phenotypes (B-SNIP) Biotypes (301 Biotype 1, 304 Biotype 2, and 345 Biotype 3), and 491 of their non-psychotic first-degree relatives. We calculated brain age values from structural T1-weighted images using the pre-trained, open-source brain age package, brainageR. In probands, we assessed associations between BAG and clinical characteristics, comorbid disorders, medications, and polygenic risk scores for SZ (PRS-SZ). Study Results All DSM diagnosis and Biotype groups had higher BAG than HC. While no significant differences were observed between BP, SAD, or SZ, Biotypes 1 and 2 had significantly higher BAG compared to Biotype 3. Relatives exhibited intermediate BAG values between HC and probands, with the highest BAG in relatives of those with SAD. Brain age gap was not linked to comorbid disorders or PRS-SZ, but was associated with symptom severity, cognition, functioning, and psychotropic medication use. Conclusions Bipolar-Schizophrenia Network for Intermediate Phenotypes Biotypes better captured age-related brain structural differences in psychosis than DSM diagnoses. Associations between BAG and medication underscore the potential influence of pharmacotherapy on brain aging in psychosis. Keywords: schizophrenia, bipolar disorder, schizoaffective disorder, structural magnetic resonance imaging, accelerated aging Introduction In the brain age paradigm, an individual’s chronological age is compared to their “brain age,” that is, age predicted from magnetic resonance (MR) images using machine learning models. The difference between these measures, termed the brain age gap (BAG), may offer insights into brain health, 1 with positive BAG values indicating accelerated aging and negative ones implying delayed maturational processes. A higher BAG has been linked to lifestyle factors such as smoking and alcohol consumption, 2 , 3 medical conditions (eg, stroke, diabetes, and hypertension), 2 , 3 a higher mortality risk, 4 cognitive impairment, 5 and several neuropsychiatric disorders. 6–11 In line with the hypothesis of accelerated aging in psychosis, 12 , 13 individuals with first-episode schizophrenia (SZ) or more advanced stages of the disease consistently exhibit higher BAG compared to healthy controls (HCs), indicating accelerated brain aging. 14–18 In bipolar disorder (BP), studies have demonstrated both elevated BAG, 11 , 19–21 BAG comparable to HC, 22 , 23 and, in a more cognitively compromised group, negative BAG, hypothetically linking cognitive dysfunction to delayed brain maturation. 24 While investigating brain aging in these disorders may suggest differential underlying neurobiological mechanisms contributing to cognitive decline and brain structural changes, findings regarding differences in BAG between BP and SZ have so far remained inconsistent. 19 , 21–23 Despite differences in clinical presentation and course, BP and SZ exhibit overlap in symptoms, 25 genetic liability, 26 , 27 cognitive impairments, 28 neurobiological alterations, 29–31 and treatment response. 32 More than half of individuals with BP experience psychotic symptoms 33 and evidence suggests psychotic BP in particular may more closely resemble SZ. 34–37 While a proportion of cases display clinically prototypic BP or SZ, many patients fall on the SZ-BP continuum, 38 partially captured by the diagnostic construct of schizoaffective disorder (SAD). 39 In addition, as a group, unaffected first-degree relatives (REL) of individuals with SZ or BP display similar, though less pronounced, neurobiological alterations, 40–42 suggesting shared endophenotypes in these disorders. 43 , 44 An alternative classification of psychotic disorders has been proposed by the Bipolar-Schizophrenia Network for Intermediate Phenotypes (B-SNIP) consortium, which revealed replicable neurobiological subtypes of psychosis that transcend conventional diagnostic boundaries of BP, SAD, and SZ. 45 , 46 These subtypes, termed “Biotypes,” display distinct cognitive, electrophysiological, and oculomotor profiles, also reflected in their relatives. 45 , 47 Accumulating evidence suggests that grouping individuals by Biotype rather than DSM diagnosis better captures between-group differences across several modalities, including brain structure. 48–54 Importantly, the distinct neurobiological profiles of the Biotypes allow proposing biologically driven hypotheses for targeted treatments. 55 While previous studies have investigated BAG in conventional, symptom-based diagnostic groups of psychosis, research on brain age in biological subtypes of psychosis is lacking. However, a recent study has found promising evidence of genetic correlations between BAG and neuroanatomical subtypes of SZ. 56 As the B-SNIP Biotypes are based on neurobiology and differ both in cognition and brain structures, we hypothesized that grouping the same psychotic individuals by Biotype compared to DSM diagnosis would yield more distinct brain aging profiles. With non-psychotic relatives often exhibiting neurobiological alterations intermediate between HC and probands, 57–59 we also expected to see higher BAG in relatives, with greater differences from HC when grouped by Biotype vs DSM diagnosis. Next, we examined the associations between BAG and probands’ clinical characteristics. Considering the higher rates of age-related somatic disorders such as obesity, 60 diabetes, 61 hyperlipidemia, 62 and hypertension, 63 and increased mortality in both BP 64 and SZ, 65 we expected to see associations between BAG and comorbid medical disorders. We also explored the links between BAG and medication use, as reported in previous studies. 10 , 18 , 66–69 Finally, given ongoing research into the genetics of brain age, with evidence of several shared single-nucleotide polymorphisms (SNPs) between BAG and both BP and SZ, 11 but mixed findings on the links between genetic risk for SZ and BAG, 56 , 70–72 we examined the associations between BAG and polygenic risk scores for schizophrenia (PRS-SZ). Methods Participants Supplementary Figure S1 visualizes the workflow for the current study. The B-SNIP study details have been previously published. 25 , 45–47 The sample included individuals from B-SNIP1 (study sites in Baltimore, Boston, Chicago, Dallas, Detroit, and Hartford) and B-SNIP2 (Boston, Chicago, Dallas, Hartford, and Athens). Study protocols were approved by the institutional review boards at each site and written informed consent was obtained from study participants prior to initiating study procedures. Study participants aged 15-65 years included 350 HCs, 979 probands grouped by either DSM diagnosis (252 psychotic BP I, 323 SAD, and 404 SZ) or Biotype (311 Biotype 1 [BT1], 319 Biotype 2 [BT2], and 349 Biotype 3 [BT3]), and 499 first-degree relatives (REL) (174 relatives of BP probands [REL-BP], 121 relatives of SAD probands [REL-SAD], 204 relatives of SZ probands [REL-SZ], or 143 relatives of BT1 probands [REL-BT1], 162 relatives of BT2 probands [REL-BT2], and 194 relatives of BT3 probands [REL-BT3]). Psychiatric diagnoses were established using the Structured Clinical Interview for DSM-IV-TR (SCID-IV). 73 Exclusion criteria for all participants included serious neurological or medical illness that can affect brain structure and function (eg, history of stroke, seizure disorder), mental retardation (IQ < 65 on the Wide Range Achievement Test), 74 current/recent substance abuse (within 3 months) or dependence (within 6 months), and extensive past history of drug dependence. Exclusion criteria for HC were a lifetime history of psychosis or recurrent mood syndromes and a family history of psychotic or bipolar disorder in first- or second-degree relatives. Based on SCID-IV, only relatives without a lifetime history of Axis I psychotic disorders were included. Relatives were additionally assessed with the Structured Interview for DSM-IV Personality Disorders (SIDP-IV). 75 Based on this assessment, among n = 499 relatives, 35 met criteria for a DSM-IV Axis II Cluster A psychosis spectrum disorder (paranoid, schizotypal, or schizoid personality disorder); analyses were repeated excluding these relatives. Biotype clustering of participants is described in detail by Clementz et al. 45 ; the current study used the Biotype assignment developed in Parker et al. 47 Briefly, the Biotypes were derived using a biomarker panel representing measures from the domains of cognition, inhibitory control, oculomotor activity, and electroencephalography (EEG). The chosen biomarkers included: (1) the Brief Assessment of Cognition in Schizophrenia (BACS), 76 (2) stop signal task, (3) pro- and anti-saccades, EEG-based event-related potentials measured with (4) auditory paired stimuli, (5) auditory oddball paradigms, and (6) intrinsic EEG activity. These measurements underwent principal component analysis; the resulting 11 biofactors were then used for k-means clustering with a 3-cluster solution as indicated by the gap statistic and other metrics. Of the 3 resulting psychosis clusters, BT1 was characterized by poor cognition and low EEG-based neural activity, BT2 by poor cognition, deficient inhibitory control, and increased EEG activity, and BT3 was characterized by moderately attenuated sensitivity to stimulus salience but was otherwise most similar to HC. Clinical Assessment Participants were assessed on socioeconomic status with the Hollingshead Two-Factor Socioeconomic Rating Scale, 77 past 30-day smoking status with the Fagerström Test for Cigarette Dependence 78 (B-SNIP2 only), lifetime cannabis abuse or dependence (from SCID-IV; combined into 1 categorical variable), concomitant medication use (average daily chlorpromazine dose equivalent, 79 total number of medications and psychotropic medications, and categorical variables indicating medication classes), and comorbid medical disorders (total number of all comorbid disorders and categorical variables indicating lifetime medical disorders). Current symptom severity in probands was assessed with the Positive and Negative Syndrome Scale (PANSS), 80 Montgomery–Åsberg Depression Rating Scale (MADRS), 81 and Young Mania Rating Scale (YMRS). 82 General and social functioning in probands were evaluated using the Global Assessment of Functioning (GAF; from SCID-IV) and the Birchwood Social Functioning Scale (SFS), 83 respectively, while cognition was assessed by calculating BACS composite z-scores (of all 6 BACS subtests). The Childhood Trauma Questionnaire (CTQ; collected in B-SNIP2 only) 84 data were also obtained from probands. Illness duration in probands was computed by subtracting age at initial SZ or BD symptoms onset from age at the MRI scan. Neuroimaging Whole brain structural T1-weighted images were collected for all study participants on 3-Tesla scanners using standardized sequences from the Alzheimer’s Disease Neuroimaging Initiative protocol ( https://adni.loni.usc.edu/data-samples/adni-data/neuroimaging/mri/mri-scanner-protocols/ ). Details on the MRI scanners and acquisition parameters are presented in Supplementary Table S1 . For all images, the image quality rating (IQR) was obtained using CAT12 85 ( https://neuro-jena.github.io/cat12-help/ , CAT12.6-rc1 [r1434], SPM12 v7487, MATLAB R2018b [9.5.0.944444]). Subjects with IQR below 70 ( n = 7) were excluded from further analyses. Brain Age Computation Brain age was computed using brainageR, a pre-trained, open-source software package for predicting age from T1-weighted structural MR images (v2.1, https://github.com/james-cole/brainageR ), 86 following the procedures outlined in the software’s online documentation. BrainageR utilizes Gaussian Processes regression and was trained on 3377 healthy individuals aged 18-90 years (mean 40.1 ± 21.8 years) derived from 7 datasets. The model’s reported performance is r = 0.973 and a mean absolute error (MAE) of 3.933 years, and r = 0.947 and MAE = 4.9 years in an independent test set. With brainageR, raw T1 images are segmented and normalized using SPM12 ( https://www.fil.ion.ucl.ac.uk/spm/software/spm12/ ; version 7487 in current analyses). Vectorized and masked white matter, grey matter, and cerebrospinal fluid tissue probability maps are then used for voxel-based brain age prediction. FSL ( https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/FSL ) is used to create images for quality control. FSL outputs were visually inspected and subjects with unsatisfactory image preprocessing (eg, suboptimal skull stripping) were excluded ( n = 32). While scanner effects are a significant source of variation in multi-site neuroimaging studies, several considerations guided our decision not to include harmonization steps in our protocol. The published brainageR model was developed using data from 7 independent datasets without harmonization, making it robust for multi-site datasets like ours. We also aimed for consistency with other BAG studies in psychosis, as many prominent studies addressed site differences by including site as a covariate rather than harmonizing data. 11 , 17 , 21 , 87 , 88 Furthermore, recent findings showed that brain age models trained on non-harmonized data performed best when applied to similarly non-harmonized test datasets, 89 providing empirical support for our approach. Genetics Full details on calculating PRS-SZ are given in the Supplementary Material . Briefly, whole blood DNA samples were genotyped using the Illumina Infinium PsychArray BeadChip platform (“PsychChip”) for B-SNIP1 and the Illumina Infinium Global Screening Array-24 (“GSA”) for B-SNIP2. Genotypes underwent standard QC and imputation to the 1000Genomes phase 3 v5 (hg19) mixed population. Post-QC and imputation, a final set of 10,321,126 SNPs was analyzed. PRS-SZ were calculated for 10 P -value thresholds, with the Psychiatric Genomics Consortium Schizophrenia 2 90 sample as training sample and the B-SNIP sample as target sample. Analyses were limited to probands of Caucasian ancestry, of whom PRS-SZ were available for 470 individuals. The first 5 principal components were used as covariates. Statistical Analysis All statistical analyses were performed in RStudio version 2023.09.0 + 463. Group differences in sociodemographic and clinical variables were assessed using ANOVA or X 2 tests. Brain age prediction accuracy was assessed by Pearson’s correlation, MAE, R 2 , and root mean squared error (RMSE). 91 Brain age models often exhibit an age bias, that is, a tendency to overestimate age in younger subjects and underestimate age in older subjects. 92 , 93 To mitigate the age bias without eliminating disease-related variability, we fit a linear model of predicted age against chronological age in HCs and used the slope and intercept from this model to adjust for the age bias in the whole sample, as recommended in the literature. 94 Brain age gap (BAG) was then computed by subtracting chronological age from corrected brain age, with positive values indicating accelerated brain aging, and negative values delayed aging. Group differences in BAG were analyzed using linear mixed models (package lme4, 95 version 1.1.31) with fixed effects for sex, IQR, and group, and a random effect for each independent site/scanner. In comparisons involving relatives, we included a nested random effect to account for familial relatedness, with the relationship type within each family (eg, proband, mother, and son) nested within a unique code assigned to each family. The emmeans 96 package, version 1.9.0, was used to estimate marginal mean BAG values and perform pairwise comparisons between groups. Degrees of freedom were determined using the Kenward–Rogers method and the Bonferroni–Holm method was applied to correct for multiple comparisons. Effect sizes for post-hoc tests were calculated using Cohen’s d . To investigate the associations between BAG and clinical variables, comorbid medical disorders, and medication use in probands, separate linear models were fit with BAG as the outcome variable, covariates of sex, site, and IQR, and the variable in question as the predictor. Continuous variables were z-score normalized, so β-values reflect a change in BAG corresponding to a one standard deviation increase in the predictor variable. For categorical variables, β-values denote the mean difference in BAG between groups. Bonferroni–Holm correction for multiple comparisons was applied separately for clinical variables (17 tests), comorbid medical disorders (20 tests), and concomitant psychotropic and non-psychotropic medication use (27 tests). PRS-SZ values in probands were z-score normalized; their relationship with BAG was assessed using a linear model with the same covariates and, in addition, 5 genetic principal components. The reported incremental R 2 denotes the increase in variance explained by adding PRS-SZ to the linear model. Significance levels were set at α < 0.05 throughout. Results Sociodemographic and Clinical Characteristics After excluding images with poor quality or unsatisfactory preprocessing with brainageR ( n = 39), the remaining sample ( n = 1789) consisted of 348 HC, 950 probands (247 BP, 313 SAD, 390 SZ or 301 BT1, 304 BT2, 345 BT3), and 491 relatives (171 REL-BP, 120 REL-SAD, 200 REL-SZ or 139 REL-BT1, 160 REL-BT2, 192 REL-BT3). Subject distribution across studies and sites is shown in Supplementary Table S2 . Sociodemographic and clinical characteristics of HC and both the diagnostic groups and Biotypes are presented in Table 1 ; demographics in HC, relatives, and probands are presented in Supplementary Table S3 . Concomitant psychotropic medication use in probands is outlined in Supplementary Table S4 . Table 1. Sociodemographic and clinical characteristics of the sample Variable %NA HC ( n = 348) BP ( n = 247) SAD ( n = 313) SZ ( n = 390) X 2 /F, P BT1 ( n = 301) BT2 ( n = 304) BT3 ( n = 345) X 2 /F, P Age (mean ± SD) 0% 37.9 ± 11.8 35.1 ± 12.5 38.3 ± 11.8 36.1 ± 12 F = 4.750, P = .003 37.2 ± 12.3 37.7 ± 11.7 35 ± 12.3 F = 4.316, P = .005 Sex (F, M, %F) 0% 199 F, 149 M (57.2 %F) 161 F, 86 M (65.2 %F) 172 F, 141 M (55 %F) 134 F, 256 M (34.4 %F) X 2 = 70.349, P < .001 131 F, 170 M (43.5 %F) 168 F, 136 M (55.3 %F) 168 F, 177 M (48.7 %F) X 2 = 14.96, P = .002 Hispanic ethnicity ( n , %) 0.2% 37 (10.6%) 30 (12.1%) 45 (14.4%) 38 (9.8%) X 2 = 4.094, P = .252 39 (13%) 35 (11.6%) 39 (11.3%) X 2 = 0.922, P = .82 Caucasian race ( n , %) 0% 220 (63.2%) 180 (72.9%) 164 (52.4%) 170 (43.6%) X 2 = 61.945, P < .001 130 (43.2%) 153 (50.3%) 231 (67%) X 2 = 48.159, P < .001 African American race ( n , %) 0% 107 (30.7%) 55 (22.3%) 141 (45%) 201 (51.5%) X 2 = 69.713, P < .001 163 (54.2%) 138 (45.4%) 96 (27.8%) X 2 = 62.43, P < .001 Education, years (mean ± SD) 1.2% 15.2 ± 2.5 14.3 ± 2.3 13.3 ± 2.2 12.8 ± 2.2 F = 70.183, P < .001 13 ± 2.2 13.1 ± 2.3 14 ± 2.3 F = 60.961, P < .001 Hollingshead score (mean ± SD) 3.9% 35.3 ± 14.2 42.4 ± 14.9 46.9 ± 14 51.6 ± 13.6 F = 84.279, P < .001 50.6 ± 14.2 48.9 ± 13.9 44 ± 14.7 F = 73.787, P < .001 Symptom onset (mean ± SD) 2.6% – 16.5 ± 7.2 16.9 ± 8.1 19 ± 7.1 F = 10.894, P < .001 17.8 ± 7.5 18.1 ± 7.6 17.1 ± 7.6 F = 1.536, P = .216 Illness duration (mean ± SD) 2.6% – 18.7 ± 12.2 21.4 ± 12.3 17.1 ± 11.7 F = 10.975, P < .001 19.3 ± 13 19.5 ± 11 18 ± 12.4 F = 1.558, P = .211 PANSS positive (mean ± SD) 3.4% – 12.6 ± 4.5 18 ± 6.2 17 ± 5.9 F = 67.65, P < .001 16.2 ± 6.1 17.6 ± 6.1 14.8 ± 5.8 F = 17.211, P < .001 PANSS negative (mean ± SD) 3.4% – 11.9 ± 4.2 15.8 ± 6.1 16.9 ± 6.3 F = 57.353, P < .001 15.4 ± 6.2 16.2 ± 6.2 14.2 ± 5.8 F = 8.967, P < .001 PANSS general (mean ± SD) 3.5% – 28.2 ± 8.1 33.6 ± 9.4 31.9 ± 9.1 F = 24.414, P < .001 31.1 ± 9.4 32.6 ± 8.9 30.8 ± 9.1 F = 3.293, P = .038 PANSS total (mean ± SD) 3.7% – 52.7 ± 14.4 67.3 ± 19 65.7 ± 18.2 F = 54.494, P < .001 62.7 ± 18.8 66.4 ± 18.2 59.9 ± 18.2 F = 9.907, P < .001 YMRS total (mean ± SD) 4% – 6.1 ± 6.8 9.5 ± 7.7 7.1 ± 6.5 F = 17.76, P < .001 7.4 ± 6.9 8 ± 7.2 7.5 ± 7.2 F = 0.527, P = .591 MADRS total (mean ± SD) 3.7% – 10.3 ± 8.9 14.1 ± 10.6 8.8 ± 8.5 F = 27.259, P < .001 10.7 ± 10 10.7 ± 9.2 11.5 ± 9.7 F = 0.738, P = .478 GAF (mean ± SD) 1.7% – 60.6 ± 13.1 51.6 ± 12.3 50 ± 12.1 F = 59.042, P < .001 51.4 ± 12.8 51.1 ± 12.3 56.8 ± 13.5 F = 20.211, P < .001 SFS total (mean ± SD) 13.2% – 137 ± 22.2 119.9 ± 24.6 119.8 ± 22.7 F = 42.491, P < .001 120.8 ± 24.9 120.9 ± 22.9 130 ± 24.1 F = 13.982, P < .001 BACS total (mean ± SD) 1.4% – −0.9 ± 1.3 −1.5 ± 1.3 −1.9 ± 1.3 F = 42.225, P < .001 −1.8 ± 1.2 −2.1 ± 1.3 −0.6 ± 1.1 F = 143.656, P < .001 Cannabis abuse/dependence ( n , %) 0.1% – 42 (17%) 92 (29.5%) 101 (25.9%) X 2 = 11.988, P = .002 76 (25.3%) 74 (24.3%) 85 (24.6%) X 2 = 0.084, P = .959 Smoking (last 30 days; n , %) 55.7% – 19 (25%) 87 (47.8%) 75 (46%) X 2 = 12.363, P = .002 63 (48.8%) 69 (51.1%) 49 (31.2%) X 2 = 14.321, P < .001 Open in a new tab Group differences between HC and diagnostic groups, HC and Biotypes, diagnostic groups, or Biotypes were assessed via ANOVA (F-statistic) or X 2 test. NA% indicates percentage of missing data. Bold denotes significant group differences. Abbreviations: BACS—Brief Assessment of Cognition in Schizophrenia, BP—bipolar disorder, BT1—Biotype 1, BT2—Biotype 2, BT3—Biotype 3, GAF—Global Assessment of Functioning, HC—healthy controls, MADRS—Montgomery–Åsberg Depression Rating Scale, PANSS—Positive and Negative Syndrome Scale, SAD—schizoaffective disorder, SFS—Birchwood Social Functioning Scale, SZ—schizophrenia, YMRS—Young Mania Rating Scale. Accuracy of Brain Age Prediction in HC A Pearson’s correlation of r = −0.28 ( P < .001) between BAG and chronological age in the whole sample indicated a significant age bias and supported implementing an age correction procedure. Age correction based on HC removed the age bias from all DSM diagnostic groups and Biotypes ( Supplementary Figure S2 ). After age correction, brain age prediction with brainageR in HC reached an accuracy of Pearson’s r = 0.87 ( P < .001), MAE = 5.154, R 2 = 0.755, and RMSE = 6.703, comparable to accuracies reported in other psychosis studies. 21 , 66 , 88 , 97 The same metrics in all groups are presented in Supplementary Table S5 . BAG in the DSM Diagnostic Groups and Biotypes A mixed model on BAG across DSM diagnoses showed a significant effect of diagnosis ( F (3, 1291.14) = 25.800, P < .001), with estimated marginal mean BAG and SE of 0.014 ± 0.649 in HC, 3.109 ± 0.703 in BP, 4.460 ± 0.666 in SAD, and 4.048 ± 0.641 in SZ ( Figure 1A ). Post-hoc analyses indicated higher BAG in all diagnostic groups compared to HC (HC vs BP: t (1289) = −5.193, P < .001, d = −0.437; HC vs SAD: t (1292) = −7.807, P < .001, d = 0.628; HC vs SZ: t (1292) = −7.354, P < .001, d = 0.569). No significant differences were observed between BP and SAD ( t (1292) = −2.168, P = .091, d = 0.191), BP and SZ ( t (1290) = −1.556, P = .240, d = 0.132), and SAD and SZ ( t (1292) = 0.741, P = .459, d = 0.058). Figure 1. Open in a new tab Estimated marginal mean BAG ± SE in (A) HC and DSM diagnostic groups and (B) HC and the Biotypes. *** P < .001, Bonferroni–Holm corrected. BAG—brain age gap, BP—bipolar disorder, BT1—Biotype 1, BT2—Biotype 2, BT3—Biotype 3, HC—healthy control, SAD—schizoaffective disorder, SZ—schizophrenia A separate mixed model on BAG across Biotypes showed a significant effect of Biotype ( F (3, 1289.44) = 31.018, P < .001), with estimated marginal mean BAG and SE of 0.020 ± 0.641 in HC, 4.624 ± 0.662 in BT1, 4.755 ± 0.653 in BT2, and 2.561 ± 0.646 in BT3 ( Figure 1B ). Post-hoc analyses indicated significantly higher BAG in all Biotypes compared to HC (HC vs BT1: t (1292) = −8.061, P < .001, d = 0.654; HC vs BT2: t (1288) = −8.418, P < .001, d = 0.672; and HC vs BT3: t (1292) = −4.650, P < .001, d = 0.361). In addition, there were significant differences in BAG between BT1 and BT3 ( t (1287) = 3.679, P < .001, d = 0.293) and BT2 and BT3 ( t (1290) = 3.901, P < .001, d = 0.312), but not between BT1 and BT2 ( t (1289) = −0.226, P = .821, d = 0.019). To examine whether BAG varied across age groups, we divided the sample into quartiles and compared BAG in the youngest and oldest quartiles. Group differences were analyzed using a linear mixed model with fixed effects for sex, IQR, diagnosis/Biotype, age group, an interaction between diagnosis/Biotype and age group, and a random effect for site/scanner. The mixed model showed a significant effect of diagnosis/Biotype, but no effect of age group or their interaction. Post-hoc analyses indicated no significant differences between younger and older individuals within any diagnostic group or Biotype. Pairwise comparisons between the DSM diagnostic groups in the younger and older quartiles were in line with the main findings, while in the Biotypes, significant differences between BT2 and BT3 only emerged in the younger quartile ( Supplementary Results , Supplementary Figure S3 , Supplementary Table S6 ). BAG in HC, Relatives, and Probands A mixed model comparing BAG values among HC, the pooled group of relatives, and all probands revealed a significant effect of group ( F (2, 1352.00) = 48.039, P < .001), with estimated marginal mean BAG and SE of 0.060 ± 0.629 in HC, 1.437 ± 0.624 in relatives, and 3.968 ± 0.557 in probands ( Figure 2A ). Post-hoc analyses revealed significant differences in BAG between all groups (HC vs relatives: t (1752) = −2.743, P = .006, d = 0.256; HC vs probands: t (1732) = −8.917, P < .001, d = 0.726; and relatives vs probands: t (926) = −6.398, P < .001, d = 0.470). Brain age gap (BAG) differences between these groups were also examined after excluding the 35 relatives with psychosis spectrum personality disorders; the results remained unchanged ( Supplementary Results and Supplementary Figure S4A ). Figure 2. Open in a new tab Estimated marginal mean BAG ± SE in (A) HC, relatives, and probands, (B) HC, relatives grouped by the diagnosis of their affected relative, and proband diagnostic groups, and (C) HC, relatives grouped by the Biotype of their affected relative, and proband Biotype groups. * P < .05, ** P < .01, *** P < .001, Bonferroni–Holm corrected. BAG—brain age gap, BP—bipolar disorder, BT1—Biotype 1, BT2—Biotype 2, BT3—Biotype 3, HC—healthy control, PROB—probands, REL—first-degree non-psychotic relatives, SAD—schizoaffective disorder, SZ—schizophrenia The model investigating BAG differences between HC, relatives grouped by the diagnosis of their affected family member, and proband diagnostic groups showed a significant effect of group ( F (6, 1494.41) = 17.219, P < .001) ( Figure 2B ). The estimated marginal mean BAG with SE of the groups is shown in Table 2 . We then compared HC to all relative groups, but each relative group only to their corresponding proband group (i.e., REL-BP vs BP, but not REL-BP vs SAD). Post-hoc analyses showed significant differences in BAG between HC and REL-SAD ( t (1681.01) = −2.566, P = .042, d = 0.364), REL-BP and PROB-BP ( t (939.68) = −2.507, P = .042, d = 0.307), REL-SAD and PROB-SAD ( t (1129.17) = −3.419, P = .003, d = 0.474), and REL-SZ and PROB-SZ ( t (1068.41) = −5.382, P < .001, d = 0.600), but not between HC and REL-BP ( t (1673.23) = −2.221, P = .053, d = 0.279) or HC and REL-SZ ( t (1735.39) = −1.299, P = .194, d = 0.154). The results remained similar when relatives with psychosis spectrum personality disorders were excluded from analyses ( Supplementary Results , Supplementary Figure S4B , and Supplementary Table S7 ). Table 2. Estimated marginal mean BAG in HC, relatives grouped by the diagnosis or Biotype of their affected family member, and proband diagnostic groups or Biotypes Group Estimated marginal mean BAG ± SE REL grouped by diagnosis HC 0.031 ± 0.631 REL-BP 1.521 ± 0.772 REL-SAD 1.972 ± 0.845 REL-SZ 0.852 ± 0.732 BP 3.162 ± 0.681 SAD 4.500 ± 0.646 SZ 4.055 ± 0.621 REL grouped by Biotype HC 0.034 ± 0.620 REL-BT1 1.357 ± 0.798 REL-BT2 1.630 ± 0.763 REL-BT3 1.226 ± 0.738 BT1 4.668 ± 0.640 BT2 4.793 ± 0.632 BT3 2.555 ± 0.625 Open in a new tab Abbreviations: BAG—brain age gap, BP—bipolar disorder, BT1—Biotype 1, BT2—Biotype 2, BT3—Biotype 3, HC—healthy control, PROB—probands, REL—first-degree non-psychotic relatives, SAD—schizoaffective disorder, SZ—schizophrenia. The mixed model on differences in BAG between HC, relatives grouped by the Biotype of their affected family member, and proband Biotypes showed a significant effect of group ( F (6, 1495.51) = 19.881, P < .001) ( Figure 2C ). The estimated marginal mean BAG and SE of the groups are shown in Table 2 . Similarly, we compared HC to all relative groups, but each relative group only to their corresponding proband group. Post-hoc analyses revealed significant differences in BAG between REL-BT1 and PROB-BT1 ( t (1127.54) = −4.803, P < .001, d = 0.615), REL-BT2 and PROB-BT2 ( t (1026.00) = −4.846, P < .001, d = 0.588), but not between HC and REL-BT1 ( t (1694.43) = −1.858, P = .126, d = 0.246), HC and REL-BT2 ( t (1697.42) = −2.358, P = .074, d = 0.297), HC and REL-BT3 ( t (1711.74) = −1.861, P = .126, d = 0.222), or REL-BT3 and PROB-BT3 ( t (981.47) = −2.192, P = .086, d = 0.247). These results also pertained when relatives with psychosis spectrum personality disorders were excluded from analyses ( Supplementary Results , Supplementary Figure S4C , and Supplementary Table S7 ). Associations Between BAG and Clinical Variables, Comorbid Medical Disorders, and Concomitant Medications in Probands In probands, we assessed the association of BAG with 17 clinical variables ( Figure 3A ). After correcting for multiple comparisons, BAG was positively associated with PANSS positive symptoms score (β = 1.128, SE = 0.251, t (904) = 4.497, P < .001, adjusted P < .001), PANSS total symptoms score (β = 1.012, SE = 0.268, t (901) = 3.781, P < .001, adjusted P = .002), PANSS negative symptoms score (β = .789, SE = 0.271, t (904) = 2.907, P = .004, adjusted P = .045), and negatively associated with SFS total score (β = −0.915, SE = 0.264, t (811) = −3.467, P < .001, adjusted P = .007), BACS composite score (β = −1.144, SE = 0.24, t (923) = −4.761, P < .001, adjusted P < .001), and GAF (β = −1.203, SE = 0.252, t (920) = −4.773, P < .001, adjusted P < .001). The association with the total number of hospitalizations did not survive Bonferroni–Holm correction (β = 0.509, SE = 0.259, t (784) = 1.969, P = .049, adjusted P = .493) and the association with PANSS general symptoms score only reached trend level significance after correction (β = 0.731, SE = 0.269, t (903) = 2.714, P = .007, adjusted P = .075). Associations with all clinical variables are presented in Supplementary Table S8 . Figure 3. Open in a new tab Associations of BAG With (A) clinical variables, (B) comorbid disorders, and (C) concomitant medications in probands. The x-axis shows the β-value for each variable in a separate linear model. Continuous variables are in italics. The color code stands for uncorrected P -values. Variables that were significantly associated with BAG after Bonferroni–Holm correction are marked by an asterisk. BACS—Brief Assessment of Cognition in Schizophrenia, BAG—brain age gap, CPZ—chlorpromazine, CTQ—Childhood Trauma Questionnaire, GAF—Global Assessment of Functioning, HEENT—head, ears, eyes, nose, and throat (disorder), MADRS—Montgomery–Åsberg Depression Rating Scale, PANSS—Positive and Negative Syndrome Scale, SFS—Birchwood Social Functioning Scale, SSRIs—selective serotonin reuptake inhibitors, YMRS—Young Mania Rating Scale. We also analyzed the association of BAG with 20 medical comorbid disorders in probands ( Figure 3B ). No comorbid disorder exhibited a significant association with BAG, apart from positive associations with endocrine disorders (β = 2.091, SE = 0.76, t (926) = 2.753, P = .006, adjusted P = .117) and the total number of comorbid disorders (β = 0.715, SE = 0.259, t (928) = 2.762, P = .006, adjusted P = .117), neither of which survived correction for multiple comparisons. Associations with all comorbid disorders are presented in Supplementary Table S9 . Finally, we examined the associations between BAG and 27 variables capturing concomitant medication use in probands ( Figure 3C ). After correcting for multiple comparisons, BAG was positively associated with the use of anticonvulsants (β = 2.557, SE = 0.502, t (930) = 5.091, P < .001, adjusted P < .001), anxiolytics/sedatives/hypnotics (β = 2.126, SE = 0.561, t (930) = 3.787, P < .001, adjusted P = .004), mood stabilizers (any) (β = 1.904, SE = 0.482, t (930) = 3.954, P < .001, adjusted P = .002), number of psychotropic medications used (β = 1.364, SE = 0.238, t (930) = 5.726, P < .001, adjusted P < .001), and number of all medications used (β = 0.882, SE = 0.243, t (930) = 3.633, P < .001, adjusted P = .007). Significant associations with BAG that did not survive Bonferroni–Holm correction included the use of antipsychotics (any) (β = 1.604, SE = 0.625, t (930) = 2.567, P = .01, adjusted P = .229), anticholinergics/anti-parkinsonian drugs (β = 1.491, SE = 0.696, t (930) = 2.143, P = .032, adjusted P = .680), cardiovascular/lipid-lowering medications (β = 1.159, SE = 0.572, t (930) = 2.027, P = .043, adjusted P = .815), selective serotonin reuptake inhibitors (SSRIs) (β = 1.117, SE = 0.527, t (930) = 2.120, P = .034, adjusted P = .685), and second-generation antipsychotics (β = 1.102, SE = 0.546, t (930) = 2.017, P = .044, adjusted P = .815). Full details of the associations with all concomitant medications are presented in Supplementary Table S10 . Association Between BAG and PRS-SZ in Probands Finally, we examined the association between BAG and genetic risk for SZ in 470 probands of Caucasian ancestry. PRS-SZ were not significantly associated with BAG across any of the examined P -value thresholds ( Table 3 ). These findings remained consistent in Supplementary analyses that only included probands with a diagnosis of schizophrenia ( n = 157, Supplementary Table S11 ). Table 3. Association of BAG with PRS-SZ in probands ( n = 470) Variable β SE df t P Incr. R 2 PRS-SZ ( P < 5 × 10 −8 ) −0.024 0.344 451 −0.07 0.944 0.03 PRS-SZ ( P < 1 × 10 −6 ) −0.401 0.349 451 −1.152 0.25 0.032 PRS-SZ ( P < 1 × 10 −4 ) 0.124 0.346 451 0.359 0.72 0.03 PRS-SZ ( P < .001) −0.16 0.358 451 −0.447 0.655 0.03 PRS-SZ ( P < .01) −0.015 0.391 451 −0.038 0.97 0.03 PRS-SZ ( P < .05) −0.245 0.422 451 −0.581 0.561 0.03 PRS-SZ ( P < .1) −0.278 0.419 451 −0.663 0.508 0.03 PRS-SZ ( P < .2) −0.221 0.432 451 −0.511 0.61 0.03 PRS-SZ ( P < .5) −0.043 0.452 451 −0.095 0.924 0.03 PRS-SZ ( P < 1) −0.075 0.453 451 −0.165 0.869 0.03 Open in a new tab Abbreviations: BAG—brain age gap, PRS-SZ—polygenic risk scores for schizophrenia. Discussion This study examined brain aging in the conventional DSM-based diagnostic groups (BP, SAD, and SZ) and the transdiagnostic psychosis Biotypes developed within the B-SNIP consortium. Using brainageR, an open-source, pre-trained package for predicting age from structural T1-weighted MR images, we showed that BP, SAD, and SZ, and the three Biotypes (BT1, BT2, and BT3) all exhibit higher brain age gap compared to HC. No significant differences were observed between the DSM diagnostic groups, whereas BT1 and BT2 had significantly higher BAG compared to BT3. In addition, BAG in the pooled group of non-psychotic first-degree relatives of the psychosis probands was intermediate in magnitude between HC and probands. In probands, BAG was positively associated with symptom severity and negatively associated with cognition and functioning. While we found no associations between brain aging and comorbid medical disorders in probands, BAG was positively associated with psychotropic medication use (ie, mood stabilizers, particularly anticonvulsants, and anxiolytics). Finally, genetic risk for schizophrenia as assessed by PRS-SZ was not associated with accelerated brain aging in probands. Our finding of higher BAG in all DSM diagnostic groups compared to HC could be attributed to our relatively large sample size, which allowed the detection of effects that some previous studies may have been underpowered to identify, such as significant differences between HC and BP. 22 , 23 In addition, the restriction to psychotic BP probands likely reduced heterogeneity, allowing for clearer differentiation from HC, particularly considering the evidence of more pronounced brain structural changes in psychotic compared to non-psychotic BP. 98–100 In our large sample, BAG did not differ between individuals with BP, SAD, and SZ (albeit a trend level difference was observed between BP and SAD), consistent with some previous studies, 19 , 21 but contrasting the findings of others. 22 , 23 Interestingly, the largest BAG compared to HC was observed in SAD (although with negligible differences between SAD and SZ), which contrasts with a previous study, 19 warranting further research. This study was the first to investigate brain age in the B-SNIP Biotypes. The distinct brain aging profiles in the Biotypes align with previous findings showing brain structural and neurocognitive differences between them. 45 , 46 Interestingly, despite BT1 having more extensive gray matter volume reductions than BT2, 45 brainageR found comparable BAG in these Biotypes, suggesting contributions from other brain structural factors as well. Previous studies have shown differences in white matter microstructure, 101 volume and shape of the amygdala and hippocampus, 49 and volumetric abnormalities in the hippocampal subfields 48 among the Biotypes. The differences in BAG between the Biotypes also mirror their cognitive profiles, with BT1 and BT2 exhibiting impaired cognition and BT3 displaying preserved cognitive abilities. 45–47 Notably, the fact that BAG was comparable between the DSM diagnostic groups, whereas the Biotypes exhibited divergent brain aging patterns, suggests that Biotypes better capture age-related brain structural differences within psychotic disorders compared to conventional diagnoses. Supplementary analyses suggested potential age-related differences in BAG within the Biotypes; however, longitudinal data will be needed to determine whether these represent meaningful distinctions in brain aging progression. While several open-source brain age prediction tools are available, we chose brainageR for this study due to its demonstrated reliability and accuracy in independent samples. 102 The observed MAE of ~5.2 years of brainageR in HC of our sample is well within the range reported in other psychosis studies, with MAEs spanning from 3.66 103 to 8.45 years. 66 Enhanced model precision might in fact interfere with the goals of the study, as the accuracy of a model is not necessarily related to its ability to detect discrepancies in brain aging across different clinically relevant phenotypes. 104 While tightly fitting brain age models likely only focus on brain features related to age and ignore other pathologies, loosely fitting models may only capture crude features such as ventricular size and miss more nuanced pathological patterns, for which moderately fitting brain age models are best suited. 105 As an important limitation, brainageR assesses aging at the level of the whole brain, whereas several prior publications have differentiated between aging in different tissue types or brain regions, or used methods that increase the interpretability of machine learning models. 11 , 88 , 97 , 106 The current methods also did not allow us to differentiate between aging-related and disease-specific brain structural patterns. 67 , 68 While some previous studies have demonstrated progressive changes in brain age in psychosis, 10 , 67 , 68 our Supplementary Analyses did not reveal differences in BAG between younger and older individuals within any group. However, our data were limited to cross-sectional snapshots of individuals in a midcourse stage of illness, which inherently restricted our ability to fully examine age-related effects. Ideally, brain age in the B-SNIP Biotypes should be explored using models more informative about the aging processes across different brain regions and tissue types over the course of the illness. With a large sample of non-psychotic relatives of probands with BP, SAD, and SZ, we showed intermediate BAG in the pooled group of relatives compared to HC and probands. Contrary to our hypothesis, grouping the relatives according to the Biotype of their affected family member did not allow better differentiation from HCs compared to grouping them by DSM diagnosis. In fact, only relatives of those with SAD displayed significantly higher BAG compared to HC, mirroring the finding of the highest BAG in SAD probands. However, comparisons between HC and the relative groups often bordered on statistical significance and displayed comparably small effect sizes, potentially explaining the earlier negative findings in smaller samples. 17 , 19 Higher BAG in the pooled group of relatives may indicate a familial predisposition to accelerated brain aging, though the effect was small. In probands, however, BAG was not associated with genetic risk for SZ assessed by PRS, consistent with some previous publications. 11 , 56 , 70 , 71 While these results may appear contradictory, findings of increased BAG in non-psychotic first-degree relatives need not necessarily imply shared genetics. Given that BAG is also associated with several environmental and behavioral factors, 2 , 3 , 107–109 this may instead reflect shared familial environments. Moreover, accelerated brain aging is not exclusive to SZ but is observed across various psychiatric conditions with shared heritability. 11 , 14 , 110 Higher BAG in the pooled group of relatives of those with BP, SAD, and SZ may instead reflect this general vulnerability to psychiatric disorders, beyond what is captured by PRS-SZ. The lack of a significant relationship between BAG and PRS-SZ in probands may also partly reflect limited statistical power in our study. However, 2 large-scale studies also failed to demonstrate genetic correlations between BAG and SZ or BP. 11 , 71 Nonetheless, several shared loci were found between BAG and the disorders 11 and Mendelian randomization indicated a possible causal influence of BP on BAG. 71 Another study did observe a small correlation between BAG and PRS-SZ ( r = 0.1), however, this was not significant after controlling for disease status or correcting for multiple comparisons. 72 Similarly, no difference in BAG was found between youth SZ groups with low or high PRS-SZ. 70 Interestingly, a recent large-scale study found genetic correlations between BAG and neuroanatomical subtypes of SZ, 56 which—like our study—underscores the relevance of biological subtypes for studying brain age in schizophrenia. While our results do contribute to the growing literature on the genetic associations between brain age and schizophrenia, much larger sample sizes are needed for a more comprehensive genetic characterization of brain aging and its relationship with psychosis and its biological subtypes. In probands, we replicated some earlier findings of positive associations between BAG and severity of positive, negative, and total (and general, at trend level) symptoms captured by the PANSS scale 11 , 68 , 111 and negative associations with measures of cognition 111 , 112 and social and global functioning. 11 , 67 , 87 , 111 However, these associations vary substantially across studies, possibly reflecting different brain age models sensitive to different brain structural features and variability in proband characteristics recruited into different studies. Taking advantage of the extensive phenotypic data in the B-SNIP dataset, we found no significant associations between BAG and comorbid medical disorders in the probands, apart from endocrine disorders and total number of medical comorbidities, neither of which survived correction for multiple comparisons. However, BAG was significantly positively associated with the total number of concomitant medications, particularly psychotropics, with even larger effects for specific medication groups. While lithium has previously shown a neuroprotective effect on brain age in BP, 69 we found a positive association between BAG and mood stabilizers. This was, however, driven by anticonvulsant use, echoing previous findings in temporal lobe epilepsy patients 113 and studies linking these drugs to reductions in brain volume and cortical thickness 100 , 114 (but see ref. 115 ). The association between BAG and anxiolytic, sedative, or hypnotic drugs may further suggest a common GABAergic mechanism, with some animal studies showing decreases in neurotrophic factors, 116 regulators of synaptic plasticity, 117 and spine density 118 following benzodiazepine administration. However, evidence of benzodiazepine-related brain structural changes in humans remains inconsistent. 119–123 Some associations not surviving correction for multiple comparisons are still in line with previous literature, such as studies linking anticholinergic burden with brain atrophy in various populations 124–126 and antipsychotic use with increased BAG. 10 , 18 , 66–68 Another study, however, found a potential neuroprotective effect of antidepressant use in major depression, 127 in contrast to our findings. Several studies have highlighted the role of psychotropic medications in brain structural changes, particularly antipsychotics, 128–130 though the effects reported vary across drug types and brain regions. 131–133 Crucially, these medications are effective in reducing symptoms for many individuals 134–136 and have been linked to reduced mortality rates. 137–139 While our findings suggest a potential association between psychotropic medication use and brain aging, several limitations must be acknowledged. Importantly, our cross-sectional, observational design cannot establish causality between medication use and BAG. Symptoms, comorbid disorders, and medication use are highly interrelated, and our findings may instead reflect the fact that those with greater illness severity tend to receive more medications. Longitudinal studies with detailed information on medication history and dosage are therefore needed to confirm these results. In addition, with many B-SNIP probands prescribed multiple drugs simultaneously, individual effects of any one drug could not be fully isolated, for which further research comparing monotherapy versus unmedicated groups is needed. Furthermore, the influence of medications on brain aging versus disease-related brain structural changes remains to be elucidated, as evidence suggests differential effects of drugs on these non-overlapping brain structural patterns. 67 , 68 While BAG captures a dynamic effect evolving over time, the medication data in the B-SNIP sample are limited to current medication use. Consequently, this cross-sectional analysis does not allow for assessing the complex dynamic effects of aging and medications on the brain, exemplified by findings of reduced BAG following early antipsychotic medication in first-episode SZ. 18 Finally, the significant associations between BAG and medication use put into question the group differences reported in the study. The DSM diagnostic groups significantly differed in psychotropic medication use, whereas the Biotypes showed less pronounced differences. To better assess differences in brain aging between groups unconfounded by medication use, comparisons should ideally involve medication-naïve or minimally medicated individuals. In conclusion, by demonstrating distinct brain aging profiles in the Biotypes, we further validated the Biotype construct and provided support for complementing studies on the neurobiology of psychosis with biologically based subtypes alongside traditional symptom-based diagnostic categories. Our findings indicate that brain aging is more strongly linked to psychotropic medication use than to poor health or genetic risk for SZ. However, these associations remain preliminary, and the causality of this link has yet to be determined. Given the potential implications for optimal treatment selection, further research is warranted to disentangle the effects of medication vs disease on brain aging, using suitable study designs and a biologically nuanced approach to the BAG metric. Supplementary material Supplementary material is available at https://academic.oup.com/schizophreniabulletin . sbaf022_suppl_Supplementary_Figures_S1-S4_Tables_S1-S11 sbaf022_suppl_supplementary_figures_s1-s4_tables_s1-s11.docx (23.7MB, docx) Contributor Information Liisi Promet, Department of Psychiatry, School of Medicine, Yale University, New Haven, CT 06511, United States; Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States. Shashwath A Meda, Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States. Ney Alliey-Rodriguez, Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States; Institute of Neuroscience, University of Texas Rio Grande Valley, Harlingen, TX 78550, United States. Brett A Clementz, Department of Neuroscience, Bio-Imaging Research Center, University of Georgia, Athens, GA 30602, United States. Elliott S Gershon, Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States. Scot K Hill, Department of Psychology, Rosalind Franklin University of Medicine and Science, Chicago, IL 60064, United States. Elena I Ivleva, Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX 75390, United States. Sarah K Keedy, Department of Psychiatry and Behavioral Neuroscience, University of Chicago, Chicago, IL 60637, United States. Matcheri S Keshavan, Department of Psychiatry, Beth Israel Deaconess Medical Centre, Harvard Medical School, Boston, MA 02215, United States. Jennifer E McDowell, Department of Psychology, University of Georgia, Athens, GA 30602, United States. David A Parker, Department of Neuroscience, Bio-Imaging Research Center, University of Georgia, Athens, GA 30602, United States; Department of Psychology, Emory University, Atlanta, GA 30322, United States. Carol A Tamminga, Department of Psychiatry, University of Texas Southwestern Medical Center, Dallas, TX 75390, United States. Godfrey D Pearlson, Department of Psychiatry, School of Medicine, Yale University, New Haven, CT 06511, United States; Olin Neuropsychiatry Research Center, Institute of Living, Hartford, CT 06106, United States. Funding This study was funded by the National Institutes of Health (NIH) grants R01077945 and R01096957 and the National Institute of Mental Health (NIMH) grants MH096957 to G.D.P., MH096942 to M.S.K., MH096913 to C.A.T., and MH096900 to B.A.C. D.P. was supported by NIH/NIMH R01MH117315 and the NIH/National Center for Advancing Translational Sciences (NCATS) UL1TR002378 and TL1TR002382. L.P. was supported by the Carl Duisberg Fellowship for Medical Sciences by the Bayer Foundation. Conflicts of Interest C.A.T., E.S.G., B.A.C., E.I.I., M.S.K., S.K.K., J.E.M., and G.D.P serve on the Board of Directors of B-SNIP Diagnostics. The remaining authors declare no conflicts of interest in relation to the subject of this study. References 1. Cole  JH, Franke  K.  Predicting age using neuroimaging: innovative brain ageing biomarkers. Trends Neurosci.  2017;40:681–690. https://doi.org/ 10.1016/j.tins.2017.10.001 [ DOI ] [ PubMed ] [ Google Scholar ] 2. Cole  JH.  Multimodality neuroimaging brain-age in UK biobank: relationship to biomedical, lifestyle, and cognitive factors. Neurobiol Aging.  2020;92:34–42. https://doi.org/ 10.1016/j.neurobiolaging.2020.03.014 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 3. Leonardsen  EH, Peng  H, Kaufmann  T, et al.  Deep neural networks learn general and clinically relevant representations of the ageing brain. Neuroimage.  2022;256:119210. https://doi.org/ 10.1016/j.neuroimage.2022.119210 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 4. Cole  JH, Ritchie  SJ, Bastin  ME, et al.  Brain age predicts mortality. Mol Psychiatry.  2018;23:1385–1392. https://doi.org/ 10.1038/mp.2017.62 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Liem  F, Varoquaux  G, Kynast  J, et al.  Predicting brain-age from multimodal imaging data captures cognitive impairment. Neuroimage.  2017;148:179–188. https://doi.org/ 10.1016/j.neuroimage.2016.11.005 [ DOI ] [ PubMed ] [ Google Scholar ] 6. Franke  K, Gaser  C.  Longitudinal changes in individual BrainAGE in healthy aging, mild cognitive impairment, and Alzheimer’s disease. GeroPsych. 2012;25:235–245. https://doi.org/ 10.1024/1662-9647/a000074 [ DOI ] [ Google Scholar ] 7. Eickhoff  CR, Hoffstaedter  F, Caspers  J, et al.  Advanced brain ageing in Parkinson’s disease is related to disease duration and individual impairment. Brain Commun. 2021;3:fcab191. https://doi.org/ 10.1093/braincomms/fcab191 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Cole  JH, Raffel  J, Friede  T, et al. ; MAGNIMS study group. Longitudinal assessment of multiple sclerosis with the brain‐age paradigm. Ann Neurol.  2020;88:93–105. https://doi.org/ 10.1002/ana.25746 [ DOI ] [ PubMed ] [ Google Scholar ] 9. Han  LKM, Dinga  R, Hahn  T, et al.  Brain aging in major depressive disorder: results from the ENIGMA major depressive disorder working group. Mol Psychiatry.  2021;26:5124–5139. https://doi.org/ 10.1038/s41380-020-0754-0 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 10. Koutsouleris  N, Davatzikos  C, Borgwardt  S, et al.  Accelerated brain aging in schizophrenia and beyond: a neuroanatomical marker of psychiatric disorders. Schizophr Bull.  2014;40:1140–1153. https://doi.org/ 10.1093/schbul/sbt142 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Kaufmann  T, van der Meer  D, Doan  NT, et al. ; Karolinska Schizophrenia Project (KaSP). Common brain disorders are associated with heritable patterns of apparent aging of the brain. Nat Neurosci.  2019;22:1617–1623. https://doi.org/ 10.1038/s41593-019-0471-7 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 12. Kirkpatrick  B, Kennedy  BK.  Accelerated aging in schizophrenia and related disorders: future research. Schizophr Res.  2018;196:4–8. https://doi.org/ 10.1016/j.schres.2017.06.034 [ DOI ] [ PubMed ] [ Google Scholar ] 13. Kirkpatrick  B, Messias  E, Harvey  PD, Fernandez-Egea  E, Bowie  CR.  Is schizophrenia a syndrome of accelerated aging? Schizophr Bull.  2008;34:1024–1032. https://doi.org/ 10.1093/schbul/sbm140 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Baecker  L, Garcia-Dias  R, Vieira  S, Scarpazza  C, Mechelli  A.  Machine learning for brain age prediction: introduction to methods and clinical applications. eBioMedicine. 2021;72:103600. https://doi.org/ 10.1016/j.ebiom.2021.103600 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 15. Ballester  PL, Romano  MT, De Azevedo Cardoso  T, et al.  Brain age in mood and psychotic disorders: a systematic review and meta‐analysis. Acta Psychiatr Scand.  2022;145:42–55. https://doi.org/ 10.1111/acps.13371 [ DOI ] [ PubMed ] [ Google Scholar ] 16. Blake  KV, Ntwatwa  Z, Kaufmann  T, Stein  DJ, Ipser  JC, Groenewold  NA.  Advanced brain ageing in adult psychopathology: a systematic review and meta-analysis of structural MRI studies. J Psychiatr Res.  2023;157:180–191. https://doi.org/ 10.1016/j.jpsychires.2022.11.011 [ DOI ] [ PubMed ] [ Google Scholar ] 17. Hajek  T, Franke  K, Kolenic  M, et al.  Brain age in early stages of bipolar disorders or schizophrenia. Schizophr Bull.  2019;45:190–198. https://doi.org/ 10.1093/schbul/sbx172 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 18. Xi  YB, Wu  XS, Cui  LB, et al.  Neuroimaging-based brain-age prediction of first-episode schizophrenia and the alteration of brain age after early medication. Br J Psychiatry.  2022;220:339–346. https://doi.org/ 10.1192/bjp.2021.169 [ DOI ] [ PubMed ] [ Google Scholar ] 19. Demro  C, Shen  C, Hendrickson  TJ, Arend  JL, Disner  SG, Sponheim  SR.  Advanced brain-age in psychotic psychopathology: evidence for transdiagnostic neurodevelopmental origins. Front Aging Neurosci.  2022;14:872867. https://doi.org/ 10.3389/fnagi.2022.872867 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Rokicki  J, Wolfers  T, Nordhøy  W, et al.  Multimodal imaging improves brain age prediction and reveals distinct abnormalities in patients with psychiatric and neurological disorders. Hum Brain Mapp.  2021;42:1714–1726. https://doi.org/ 10.1002/hbm.25323 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 21. Tønnesen  S, Kaufmann  T, De Lange  AMG, et al.  Brain age prediction reveals aberrant brain white matter in schizophrenia and bipolar disorder: a multisample diffusion tensor imaging study. Biol Psychiatry Cogn Neurosci Neuroimaging. 2020;5:1095–1103. https://doi.org/ 10.1016/j.bpsc.2020.06.014 [ DOI ] [ PubMed ] [ Google Scholar ] 22. Nenadić  I, Dietzek  M, Langbein  K, Sauer  H, Gaser  C.  BrainAGE score indicates accelerated brain aging in schizophrenia, but not bipolar disorder. Psychiatry Res Neuroimaging.  2017;266:86–89. https://doi.org/ 10.1016/j.pscychresns.2017.05.006 [ DOI ] [ PubMed ] [ Google Scholar ] 23. Shahab  S, Mulsant  BH, Levesque  ML, et al.  Brain structure, cognition, and brain age in schizophrenia, bipolar disorder, and healthy controls. Neuropsychopharmacology.  2019;44:898–906. https://doi.org/ 10.1038/s41386-018-0298-z [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 24. Chakrabarty  T, Frangou  S, Torres  IJ, Ge  R, Yatham  LN.  Brain age and cognitive functioning in first-episode bipolar disorder. Psychol Med.  2023;53:5127–5135. https://doi.org/ 10.1017/S0033291722002136 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 25. Tamminga  CA, Ivleva  EI, Keshavan  MS, et al.  Clinical phenotypes of psychosis in the bipolar-schizophrenia network on intermediate phenotypes (B-SNIP). Am J Psychiatry.  2013;170:1263–1274. https://doi.org/ 10.1176/appi.ajp.2013.12101339 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 26. Forstner  AJ, Hecker  J, Hofmann  A, et al.  Identification of shared risk loci and pathways for bipolar disorder and schizophrenia. PLoS One.  2017;12:e0171595. https://doi.org/ 10.1371/journal.pone.0171595 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 27. Lichtenstein  P, Yip  BH, Björk  C, et al.  Common genetic determinants of schizophrenia and bipolar disorder in Swedish families: a population-based study. Lancet. 2009;373:234–239. https://doi.org/ 10.1016/S0140-6736(09)60072-6 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 28. Schretlen  DJ, Cascella  NG, Meyer  SM, et al.  Neuropsychological functioning in bipolar disorder and schizophrenia. Biol Psychiatry.  2007;62:179–186. https://doi.org/ 10.1016/j.biopsych.2006.09.025 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 29. De Peri  L, Crescini  A, Deste  G, Fusar-Poli  P, Sacchetti  E, Vita  A.  Brain structural abnormalities at the onset of schizophrenia and bipolar disorder: a meta-analysis of controlled magnetic resonance imaging studies. Curr Pharm Des.  2012;18:486–494. https://doi.org/ 10.2174/138161212799316253 [ DOI ] [ PubMed ] [ Google Scholar ] 30. Ethridge  LE, Hamm  JP, Pearlson  GD, et al.  Event-related potential and time-frequency endophenotypes for schizophrenia and psychotic bipolar disorder. Biol Psychiatry.  2015;77:127–136. https://doi.org/ 10.1016/j.biopsych.2014.03.032 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Meda  SA, Ruaño  G, Windemuth  A, et al.  Multivariate analysis reveals genetic associations of the resting default mode network in psychotic bipolar disorder and schizophrenia. Proc Natl Acad Sci USA.  2014;111:E2066–E2075. https://doi.org/ 10.1073/pnas.1313093111 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 32. Citrome  L, Goldberg  JF, Stahl  SM.  Toward convergence in the medication treatment of bipolar disorder and schizophrenia. Harv Rev Psychiatry.  2005;13:28–42. https://doi.org/ 10.1080/10673220590923164 [ DOI ] [ PubMed ] [ Google Scholar ] 33. Chakrabarti  S, Singh  N.  Psychotic symptoms in bipolar disorder and their impact on the illness: a systematic review. World J Psychiatry. 2022;12:1204–1232. https://doi.org/ 10.5498/wjp.v12.i9.1204 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Anticevic  A, Yang  G, Savic  A, et al.  Mediodorsal and visual thalamic connectivity differ in schizophrenia and bipolar disorder with and without psychosis history. Schizophr Bull.  2014;40:1227–1243. https://doi.org/ 10.1093/schbul/sbu100 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Bora  E.  Neurocognitive features in clinical subgroups of bipolar disorder: a meta-analysis. J Affect Disord.  2018;229:125–134. https://doi.org/ 10.1016/j.jad.2017.12.057 [ DOI ] [ PubMed ] [ Google Scholar ] 36. Pearlson  GD, Wong  DF, Tune  LE, et al.  In vivo D2 dopamine receptor density in psychotic and nonpsychotic patients with bipolar disorder. Arch Gen Psychiatry.  1995;52:471–477. https://doi.org/ 10.1001/archpsyc.1995.03950180057008 [ DOI ] [ PubMed ] [ Google Scholar ] 37. Bellivier  F, Geoffroy  PA, Scott  J, Schurhoff  F, Leboyer  M, Etain  B.  Biomarkers of bipolar disorder: specific or shared with schizophrenia? Front Biosci (Elite Ed).  2013;5:845–863. https://doi.org/ 10.2741/E665 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Keshavan  MS, Morris  DW, Sweeney  JA, et al.  A dimensional approach to the psychosis spectrum between bipolar disorder and schizophrenia: the Schizo-Bipolar Scale. Schizophr Res.  2011;133:250–254. https://doi.org/ 10.1016/j.schres.2011.09.005 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Kasanin  J.  The acute schizoaffective psychoses. AJP.  1933;90:97–126. https://doi.org/ 10.1176/ajp.90.1.97 [ DOI ] [ PubMed ] [ Google Scholar ] 40. Ivleva  EI, Moates  AF, Hamm  JP, et al.  Smooth pursuit eye movement, prepulse inhibition, and auditory paired stimuli processing endophenotypes across the schizophrenia-bipolar disorder psychosis dimension. Schizophr Bull.  2014;40:642–652. https://doi.org/ 10.1093/schbul/sbt047 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 41. Khadka  S, Meda  SA, Stevens  MC, et al.  Is aberrant functional connectivity a psychosis endophenotype? A resting state functional magnetic resonance imaging study. Biol Psychiatry.  2013;74:458–466. https://doi.org/ 10.1016/j.biopsych.2013.04.024 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Narayanan  B, O’Neil  K, Berwise  C, et al.  Resting state electroencephalogram oscillatory abnormalities in schizophrenia and psychotic bipolar patients and their relatives from the bipolar and schizophrenia network on intermediate phenotypes study. Biol Psychiatry.  2014;76:456–465. https://doi.org/ 10.1016/j.biopsych.2013.12.008 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Gottesman  II, Gould  TD.  The endophenotype concept in psychiatry: etymology and strategic intentions. Am J Psychiatry.  2003;160:636–645. https://doi.org/ 10.1176/appi.ajp.160.4.636 [ DOI ] [ PubMed ] [ Google Scholar ] 44. Pearlson  GD.  Etiologic, phenomenologic, and endophenotypic overlap of schizophrenia and bipolar disorder. Annu Rev Clin Psychol.  2015;11:251–281. https://doi.org/ 10.1146/annurev-clinpsy-032814-112915 [ DOI ] [ PubMed ] [ Google Scholar ] 45. Clementz  BA, Sweeney  JA, Hamm  JP, et al.  Identification of distinct psychosis biotypes using brain-based biomarkers. Am J Psychiatry.  2016;173:373–384. https://doi.org/ 10.1176/appi.ajp.2015.14091200 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Clementz  BA, Parker  DA, Trotti  RL, et al.  Psychosis biotypes: replication and validation from the B-SNIP consortium. Schizophr Bull.  2022;48:56–68. https://doi.org/ 10.1093/schbul/sbab090 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 47. Parker  D, Trotti  R, McDowell  J, et al.  Differentiating biomarker features and familial characteristics of B-SNIP psychosis biotypes. Preprint. Res Sq . 2024;rs.3.rs-3702638. Published 2024 Jan 5. https://doi.org/ 10.21203/rs.3.rs-3702638/v1 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 48. Del Re  EC, Zeng  V, Alliey-Rodriguez  N, et al.  Anterior-posterior axis of hippocampal subfields across psychoses: a B-SNIP study. Biomark Neuropsychiatry. 2021;5:100037. https://doi.org/ 10.1016/j.bionps.2021.100037 [ DOI ] [ Google Scholar ] 49. Guimond  S, Gu  F, Shannon  H, et al.  A diagnosis and biotype comparison across the psychosis spectrum: investigating volume and shape Amygdala-Hippocampal differences from the B-SNIP study. Schizophr Bull.  2021;47:1706–1717. https://doi.org/ 10.1093/schbul/sbab071 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 50. Ivleva  EI, Clementz  BA, Dutcher  AM, et al.  Brain structure biomarkers in the psychosis biotypes: findings from the bipolar-schizophrenia network for intermediate phenotypes. Biol Psychiatry.  2017;82:26–39. https://doi.org/ 10.1016/j.biopsych.2016.08.030 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 51. Ji  L, Meda  SA, Tamminga  CA, et al.  Characterizing functional regional homogeneity (ReHo) as a B-SNIP psychosis biomarker using traditional and machine learning approaches. Schizophr Res.  2020;215:430–438. https://doi.org/ 10.1016/j.schres.2019.07.015 [ DOI ] [ PubMed ] [ Google Scholar ] 52. Koen  JD, Lewis  L, Rugg  MD, et al.  Supervised machine learning classification of psychosis biotypes based on brain structure: findings from the bipolar-schizophrenia network for intermediate phenotypes (B-SNIP). Sci Rep.  2023;13:12980. https://doi.org/ 10.1038/s41598-023-38101-0 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 53. Thomas  O, Parker  D, Trotti  R, et al.  Intrinsic neural activity differences in psychosis biotypes: findings from the bipolar-schizophrenia network on intermediate phenotypes (B-SNIP) consortium. Biomark Neuropsychiatry. 2019;1:100002. https://doi.org/ 10.1016/j.bionps.2019.100002 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 54. Meda  SA, Clementz  BA, Sweeney  JA, et al.  Examining functional resting-state connectivity in psychosis and its subgroups in the bipolar-schizophrenia network on intermediate phenotypes cohort. Biol Psychiatry Cogn Neurosci Neuroimaging. 2016;1:488–497. https://doi.org/ 10.1016/j.bpsc.2016.07.001 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 55. Clementz  BA, Trotti  RL, Pearlson  GD, et al.  Testing psychosis phenotypes from bipolar-schizophrenia network for intermediate phenotypes for clinical application: biotype characteristics and targets. Biol Psychiatry Cogn Neurosci Neuroimaging. 2020;5:808–818. https://doi.org/ 10.1016/j.bpsc.2020.03.011 [ DOI ] [ PubMed ] [ Google Scholar ] 56. Wen  J, Zhao  B, Yang  Z, et al.  The genetic architecture of multimodal human brain age. Nat Commun.  2024;15:2604. https://doi.org/ 10.1038/s41467-024-46796-6 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 57. Boos  HBM, Aleman  A, Cahn  W, Pol  HH, Kahn  RS.  Brain volumes in relatives of patients with schizophrenia: a meta-analysis. Arch Gen Psychiatry.  2007;64:297. https://doi.org/ 10.1001/archpsyc.64.3.297 [ DOI ] [ PubMed ] [ Google Scholar ] 58. Moberg  PJ, Kamath  V, Marchetto  DM, et al.  Meta-analysis of olfactory function in schizophrenia, first-degree family members, and youths at-risk for psychosis. Schizophr Bull.  2014;40:50–59. https://doi.org/ 10.1093/schbul/sbt049 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 59. Snitz  BE, MacDonald  AW  III, Carter  CS.  Cognitive deficits in unaffected first-degree relatives of schizophrenia patients: a meta-analytic review of putative endophenotypes. Schizophr Bull.  2006;32:179–194. https://doi.org/ 10.1093/schbul/sbi048 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 60. Chouinard  VA, Pingali  SM, Chouinard  G, et al.  Factors associated with overweight and obesity in schizophrenia, schizoaffective and bipolar disorders. Psychiatry Res.  2016;237:304–310. https://doi.org/ 10.1016/j.psychres.2016.01.024 [ DOI ] [ PubMed ] [ Google Scholar ] 61. Vancampfort  D, Correll  CU, Galling  B, et al.  Diabetes mellitus in people with schizophrenia, bipolar disorder and major depressive disorder: a systematic review and large scale meta-analysis. World Psychiatry. 2016;15:166–174. https://doi.org/ 10.1002/wps.20309 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 62. Sugai  T, Suzuki  Y, Yamazaki  M, et al.  High prevalence of obesity, hypertension, hyperlipidemia, and diabetes mellitus in Japanese outpatients with schizophrenia: a nationwide survey. PLoS One.  2016;11:e0166429. https://doi.org/ 10.1371/journal.pone.0166429 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 63. Ayerbe  L, Forgnone  I, Addo  J, Siguero  A, Gelati  S, Ayis  S.  Hypertension risk and clinical care in patients with bipolar disorder or schizophrenia; a systematic review and meta-analysis. J Affect Disord.  2018;225:665–670. https://doi.org/ 10.1016/j.jad.2017.09.002 [ DOI ] [ PubMed ] [ Google Scholar ] 64. Crump  C, Sundquist  K, Winkleby  MA, Sundquist  J.  Comorbidities and mortality in bipolar disorder: a Swedish National Cohort Study. JAMA Psychiatry. 2013;70:931–939. https://doi.org/ 10.1001/jamapsychiatry.2013.1394 [ DOI ] [ PubMed ] [ Google Scholar ] 65. Crump  C, Winkleby  MA, Sundquist  K, Sundquist  J.  Comorbidities and mortality in persons with schizophrenia: a Swedish National Cohort Study. Am J Psychiatry.  2013;170:324–333. https://doi.org/ 10.1176/appi.ajp.2012.12050599 [ DOI ] [ PubMed ] [ Google Scholar ] 66. Constantinides  C, Han  LKM, Alloza  C, et al. ; ENIGMA Schizophrenia Consortium. Brain ageing in schizophrenia: evidence from 26 international cohorts via the ENIGMA Schizophrenia Consortium. Mol Psychiatry.  2023;28:1201–1209. https://doi.org/ 10.1038/s41380-022-01897-w [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 67. Schnack  HG, van Haren  NEM, Nieuwenhuis  M, Hulshoff Pol  HE, Cahn  W, Kahn  RS.  Accelerated brain aging in schizophrenia: a longitudinal pattern recognition study. Am J Psychiatry.  2016;173:607–616. https://doi.org/ 10.1176/appi.ajp.2015.15070922 [ DOI ] [ PubMed ] [ Google Scholar ] 68. Lieslehto  J, Jääskeläinen  E, Kiviniemi  V, et al.  The progression of disorder-specific brain pattern expression in schizophrenia over 9 years. npj Schizophr.  2021;7:32. https://doi.org/ 10.1038/s41537-021-00157-0 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 69. Van Gestel  H, Franke  K, Petite  J, et al.  Brain age in bipolar disorders: effects of lithium treatment. Aust N Z J Psychiatry.  2019;53:1179–1188. https://doi.org/ 10.1177/0004867419857814 [ DOI ] [ PubMed ] [ Google Scholar ] 70. Constantinides  C, Baltramonaityte  V, Caramaschi  D, et al.  Assessing the association between global structural brain age and polygenic risk for schizophrenia in early adulthood: a recall-by-genotype study. Cortex.  2024;172:1–13. https://doi.org/ 10.1016/j.cortex.2023.11.015 [ DOI ] [ PubMed ] [ Google Scholar ] 71. Leonardsen  EH, Vidal-Piñeiro  D, Roe  JM, et al.  Genetic architecture of brain age and its causal relations with brain and mental disorders. Mol Psychiatry.  2023;28:3111–3120. https://doi.org/ 10.1038/s41380-023-02087-y [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 72. Teeuw  J, Ori  APS, Brouwer  RM, et al.  Accelerated aging in the brain, epigenetic aging in blood, and polygenic risk for schizophrenia. Schizophr Res.  2021;231:189–197. https://doi.org/ 10.1016/j.schres.2021.04.005 [ DOI ] [ PubMed ] [ Google Scholar ] 73. First  MB, Spitzer  RL, Gibbon  M, Williams  JBW.  Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition (SCID-I/P). Biometrics Research, New York State Psychiatric Institute; 2002. [ Google Scholar ] 74. Wilkinson  GS, Robertson  GJ.  Wide Range Achievement Test. 4th Edition. Psychological Assessment Resources; 2006. [ Google Scholar ] 75. Zanarini  MC, Frankenburg  FR, Sickel  AE, Yong  L.  The Diagnostic Interview for DSM-IV Personality Disorders (DIPD-IV).  McLean Hospital; 1996. [ Google Scholar ] 76. Keefe  RSE, Goldberg  TE, Harvey  PD, Gold  JM, Poe  MP, Coughenour  L.  The brief assessment of cognition in schizophrenia: reliability, sensitivity, and comparison with a standard neurocognitive battery. Schizophr Res.  2004;68:283–297. https://doi.org/ 10.1016/j.schres.2003.09.011 [ DOI ] [ PubMed ] [ Google Scholar ] 77. Hollingshead  AB.  Two Factor Index of Social Position. Privately printed; 1957. [ Google Scholar ] 78. Heatherton  TF, Kozlowski  LT, Frecker  RC, Fagerström  KO.  The Fagerström test for nicotine dependence: a revision of the Fagerström tolerance questionnaire. Br J Addict.  1991;86:1119–1127. https://doi.org/ 10.1111/j.1360-0443.1991.tb01879.x [ DOI ] [ PubMed ] [ Google Scholar ] 79. Andreasen  NC, Pressler  M, Nopoulos  P, Miller  D, Ho  BC.  Antipsychotic dose equivalents and dose-years: a standardized method for comparing exposure to different drugs. Biol Psychiatry.  2010;67:255–262. https://doi.org/ 10.1016/j.biopsych.2009.08.040 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 80. Lançon  C, Auquier  P, Nayt  G, Reine  G.  Stability of the five-factor structure of the Positive and Negative Syndrome Scale (PANSS). Schizophr Res.  2000;42:231–239. https://doi.org/ 10.1016/S0920-9964(99)00129-2 [ DOI ] [ PubMed ] [ Google Scholar ] 81. Montgomery  SA, Åsberg  M.  A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382–389. https://doi.org/ 10.1192/bjp.134.4.382 [ DOI ] [ PubMed ] [ Google Scholar ] 82. Young  RC, Biggs  JT, Ziegler  VE, Meyer  DA.  A rating scale for mania: reliability, validity and sensitivity. Br J Psychiatry.  1978;133:429–435. https://doi.org/ 10.1192/bjp.133.5.429 [ DOI ] [ PubMed ] [ Google Scholar ] 83. Birchwood  M, Smith  J, Cochrane  R, Wetton  S, Copestake  S.  The social functioning scale the development and validation of a new scale of social adjustment for use in family intervention programmes with schizophrenic patients. Br J Psychiatry.  1990;157:853–859. https://doi.org/ 10.1192/bjp.157.6.853 [ DOI ] [ PubMed ] [ Google Scholar ] 84. Bernstein  DP, Fink  L, Handelsman  L, et al.  Initial reliability and validity of a new retrospective measure of child abuse and neglect. Am J Psychiatry.  1994;151:1132–1136. https://doi.org/ 10.1176/ajp.151.8.1132 [ DOI ] [ PubMed ] [ Google Scholar ] 85. Gaser  C, Dahnke  R, Thompson  PM, Kurth  F, Luders  E; The Alzheimer’s Disease Neuroimaging Initiative. CAT—a computational anatomy toolbox for the analysis of structural MRI data. GigaScience  2024;13:giae049. https://doi.org/ 10.1093/gigascience/giae049 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 86. Biondo  F, Jewell  A, Pritchard  M, et al.  Brain-age is associated with progression to dementia in memory clinic patients. Neuroimage Clin. 2022;36:103175. https://doi.org/ 10.1016/j.nicl.2022.103175 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 87. McWhinney  S, Kolenic  M, Franke  K, et al.  Obesity as a risk factor for accelerated brain ageing in first-episode psychosis—a longitudinal study. Schizophr Bull.  2021;47:1772–1781. https://doi.org/ 10.1093/schbul/sbab064 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 88. Haas  SS, Ge  R, Sanford  N, et al.  Accelerated global and local brain aging differentiate cognitively impaired from cognitively spared patients with schizophrenia. Front Psychiatry.  2022;13:913470. https://doi.org/ 10.3389/fpsyt.2022.913470 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 89. Yu  Y, Cui  HQ, Haas  SS, et al. ; ENIGMA‐Lifespan Working Group. Brain-age prediction: systematic evaluation of site effects, and sample age range and size. Hum Brain Mapp.  2024;45:e26768. https://doi.org/ 10.1002/hbm.26768 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 90. Ripke  S, Neale  BM, Corvin  A, et al.  Biological insights from 108 schizophrenia-associated genetic loci. Nature.  2014;511:421–427. https://doi.org/ 10.1038/nature13595 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 91. de Lange  AMG, Anatürk  M, Rokicki  J, et al.  Mind the gap: performance metric evaluation in brain-age prediction. Hum Brain Mapp.  2022;43:3113–3129. https://doi.org/ 10.1002/hbm.25837 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 92. Le  TT, Kuplicki  RT, McKinney  BA, Yeh  H-W, Thompson  WK, Paulus  MP; Tulsa 1000 Investigators. A nonlinear simulation framework supports adjusting for age when analyzing BrainAGE. Front Aging Neurosci.  2018;10:317. https://doi.org/ 10.3389/fnagi.2018.00317 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 93. Liang  H, Zhang  F, Niu  X.  Investigating systematic bias in brain age estimation with application to post-traumatic stress disorders. Hum Brain Mapp.  2019;40:3143–3152. https://doi.org/ 10.1002/hbm.24588 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 94. de Lange  AMG, Cole  JH.  Commentary: correction procedures in brain-age prediction. Neuroimage Clin. 2020;26:102229. https://doi.org/ 10.1016/j.nicl.2020.102229 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 95. Bates  D, Mächler  M, Bolker  B, Walker  S.  Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67:1–48. https://doi.org/ 10.18637/jss.v067.i01 [ DOI ] [ Google Scholar ] 96. Lenth  RV. emmeans: Estimated Marginal Means, aka Least-Squares Means. October 2017:1.10.4. https://doi.org/ 10.32614/CRAN.package.emmeans [ DOI ] [ Google Scholar ] 97. Ballester  PL, Suh  JS, Ho  NCW, et al.  Gray matter volume drives the brain age gap in schizophrenia: a SHAP study. Schizophrenia  2023;9:3. https://doi.org/ 10.1038/s41537-022-00330-z [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 98. Altamura  AC, Maggioni  E, Dhanoa  T, et al.  The impact of psychosis on brain anatomy in bipolar disorder: a structural MRI study. J Affect Disord.  2018;233:100–109. https://doi.org/ 10.1016/j.jad.2017.11.092 [ DOI ] [ PubMed ] [ Google Scholar ] 99. Maggioni  E, Altamura  AC, Brambilla  P.  Exploring the neuroanatomical bases of psychotic features in bipolar disorder. Epidemiol Psychiatr Sci. 2017;26:358–363. https://doi.org/ 10.1017/S2045796017000087 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 100. Hibar  DP, Westlye  LT, Doan  NT, et al. ; ENIGMA Bipolar Disorder Working Group. Cortical abnormalities in bipolar disorder: an MRI analysis of 6503 individuals from the ENIGMA Bipolar Disorder Working Group. Mol Psychiatry.  2018;23:932–942. https://doi.org/ 10.1038/mp.2017.73 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 101. Kelly  S, Guimond  S, Pasternak  O, et al.  White matter microstructure across brain-based biotypes for psychosis—findings from the bipolar-schizophrenia network for intermediate phenotypes. Psychiatry Res Neuroimaging.  2021;308:111234. https://doi.org/ 10.1016/j.pscychresns.2020.111234 [ DOI ] [ PubMed ] [ Google Scholar ] 102. Dörfel  RP, Arenas-Gomez  JM, Fisher  PM, et al.  Prediction of brain age using structural magnetic resonance imaging: a comparison of accuracy and test-retest reliability of publicly available software packages. Hum Brain Mapp.  2023;44:6139–6148. https://doi.org/ 10.1101/2023.01.26.525514 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 103. Kuo  CY, Lee  PL, Hung  SC, et al.  Large-scale structural covariance networks predict age in middle-to-late adulthood: a novel brain aging biomarker. Cereb Cortex. 2020;30:5844–5862. https://doi.org/ 10.1093/cercor/bhaa161 [ DOI ] [ PubMed ] [ Google Scholar ] 104. Jirsaraie  RJ, Gorelik  AJ, Gatavins  MM, et al.  A systematic review of multimodal brain age studies: uncovering a divergence between model accuracy and utility. Patterns (N Y). 2023;4:100712. https://doi.org/ 10.1016/j.patter.2023.100712 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 105. Bashyam  VM, Erus  G, Doshi  J, et al.  MRI signatures of brain age and disease over the lifespan based on a deep brain network and 14 468 individuals worldwide. Brain.  2020;143:2312–2324. https://doi.org/ 10.1093/brain/awaa160 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 106. Zhu  JD, Wu  YF, Tsai  SJ, Lin  CP, Yang  AC.  Investigating brain aging trajectory deviations in different brain regions of individuals with schizophrenia using multimodal magnetic resonance imaging and brain-age prediction: a multicenter study. Transl Psychiatry.  2023;13:82. https://doi.org/ 10.1038/s41398-023-02379-5 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 107. Bittner  N, Jockwitz  C, Franke  K, et al.  When your brain looks older than expected: combined lifestyle risk and BrainAGE. Brain Struct Funct.  2021;226:621–645. https://doi.org/ 10.1007/s00429-020-02184-6 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 108. Rogenmoser  L, Kernbach  J, Schlaug  G, Gaser  C.  Keeping brains young with making music. Brain Struct Funct.  2018;223:297–305. https://doi.org/ 10.1007/s00429-017-1491-2 [ DOI ] [ PubMed ] [ Google Scholar ] 109. Luders  E, Cherbuin  N, Gaser  C.  Estimating brain age using high-resolution pattern recognition: younger brains in long-term meditation practitioners. Neuroimage.  2016;134:508–513. https://doi.org/ 10.1016/j.neuroimage.2016.04.007 [ DOI ] [ PubMed ] [ Google Scholar ] 110. Anttila  V, Bulik-Sullivan  B, Finucane  HK, et al. ; Brainstorm Consortium. Analysis of shared heritability in common disorders of the brain. Science.  2018;360:eaap8757. https://doi.org/ 10.1126/science.aap8757 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 111. Kim  WS, Heo  DW, Shen  J, et al.  Stage-specific brain aging in first-episode schizophrenia and treatment-resistant schizophrenia. Int J Neuropsychopharmacol.  2023;26:207–216. https://doi.org/ 10.1093/ijnp/pyac080 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 112. Wang  J, Kochunov  P, Sampath  H, et al.  White matter brain aging in relationship to schizophrenia and its cognitive deficit. Schizophr Res.  2021;230:9–16. https://doi.org/ 10.1016/j.schres.2021.02.003 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 113. Hwang  G, Hermann  B, Nair  VA, et al.  Brain aging in temporal lobe epilepsy: chronological, structural, and functional. Neuroimage Clin. 2020;25:102183. https://doi.org/ 10.1016/j.nicl.2020.102183 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 114. Pardoe  HR, Berg  AT, Jackson  GD.  Sodium valproate use is associated with reduced parietal lobe thickness and brain volume. Neurology.  2013;80:1895–1900. https://doi.org/ 10.1212/WNL.0b013e318292a2e5 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 115. Atmaca  M, Ozdemir  H, Cetinkaya  S, et al.  Cingulate gyrus volumetry in drug free bipolar patients and patients treated with valproate or valproate and quetiapine. J Psychiatr Res.  2007;41:821–827. https://doi.org/ 10.1016/j.jpsychires.2006.07.006 [ DOI ] [ PubMed ] [ Google Scholar ] 116. Licata  SC, Shinday  NM, Huizenga  MN, et al.  Alterations in brain-derived neurotrophic factor in the mouse hippocampus following acute but not repeated benzodiazepine treatment. PLoS One.  2013;8:e84806. https://doi.org/ 10.1371/journal.pone.0084806 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 117. Huopaniemi  L, Keist  R, Randolph  A, Certa  U, Rudolph  U.  Diazepam-induced adaptive plasticity revealed by α1 GABAA receptor-specific expression profiling. J Neurochem.  2004;88:1059–1067. https://doi.org/ 10.1046/j.1471-4159.2003.02216.x [ DOI ] [ PubMed ] [ Google Scholar ] 118. Curto  Y, Garcia-Mompo  C, Bueno-Fernandez  C, Nacher  J.  Chronic benzodiazepine treatment decreases spine density in cortical pyramidal neurons. Neurosci Lett.  2016;613:41–46. https://doi.org/ 10.1016/j.neulet.2015.12.048 [ DOI ] [ PubMed ] [ Google Scholar ] 119. Hofe  I, Stricker  BH, Vernooij  MW, Ikram  MK, Ikram  MA, Wolters  FJ.  Benzodiazepine use in relation to long-term dementia risk and imaging markers of neurodegeneration: a population-based study. BMC Med.  2024;22:266. https://doi.org/ 10.1186/s12916-024-03437-5 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 120. Richardson  K, Wharton  SB, Grossi  CM, et al.  Neuropathological correlates of cumulative benzodiazepine and anticholinergic drug use. J Alzheimers Dis. 2020;74:999–1009. https://doi.org/ 10.3233/JAD-191199 [ DOI ] [ PubMed ] [ Google Scholar ] 121. Huhtaniska  S, Korkala  I, Heikka  T, et al.  Antipsychotic and benzodiazepine use and brain morphology in schizophrenia and affective psychoses—systematic reviews and birth cohort study. Psychiatry Res Neuroimaging.  2018;281:43–52. https://doi.org/ 10.1016/j.pscychresns.2018.08.015 [ DOI ] [ PubMed ] [ Google Scholar ] 122. Huhtaniska  S, Jääskeläinen  E, Heikka  T, et al.  Long-term antipsychotic and benzodiazepine use and brain volume changes in schizophrenia: the Northern Finland Birth Cohort 1966 study. Psychiatry Res Neuroimaging.  2017;266:73–82. https://doi.org/ 10.1016/j.pscychresns.2017.05.009 [ DOI ] [ PubMed ] [ Google Scholar ] 123. Gallet  Q, Bouteloup  V, Locatelli  M, et al. ; MEMENTO study group. Benzodiazepine use and neuroimaging markers of Alzheimer’s disease in nondemented older individuals: an MRI and 18F Florbetapir PET study in the MEMENTO cohort. Neuropsychopharmacology.  2022;47:1114–1120. https://doi.org/ 10.1038/s41386-021-01246-5 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 124. Cooley  SA, Paul  RH, Strain  JF, Boerwinkle  A, Kilgore  C, Ances  BM.  Effects of anticholinergic medication use on brain integrity in persons living with HIV and persons without HIV. AIDS.  2021;35:381–391. https://doi.org/ 10.1097/QAD.0000000000002768 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 125. Kilimann  I, Wucherer  D, Ittermann  T, et al.  Inverse association between the anticholinergic burden and hippocampus volume in a population-based cohort across the entire adult age range. GeroScience. 2022;44:1715–1726. https://doi.org/ 10.1007/s11357-021-00497-w [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 126. Risacher  SL, McDonald  BC, Tallman  EF, et al. ; Alzheimer’s Disease Neuroimaging Initiative. Association between anticholinergic medication use and cognition, brain metabolism, and brain atrophy in cognitively normal older adults. JAMA Neurology. 2016;73:721–732. https://doi.org/ 10.1001/jamaneurol.2016.0580 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 127. Han  LKM, Schnack  HG, Brouwer  RM, et al.  Contributing factors to advanced brain aging in depression and anxiety disorders. Transl Psychiatry.  2021;11:1–11. https://doi.org/ 10.1038/s41398-021-01524-2 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 128. Haijma  SV, Van Haren  N, Cahn  W, Koolschijn  PCMP, Hulshoff Pol  HE, Kahn  RS.  Brain volumes in schizophrenia: a meta-analysis in over 18 000 subjects. Schizophr Bull.  2013;39:1129–1138. https://doi.org/ 10.1093/schbul/sbs118 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 129. Fusar-Poli  P, Smieskova  R, Kempton  MJ, Ho  BC, Andreasen  NC, Borgwardt  S.  Progressive brain changes in schizophrenia related to antipsychotic treatment? A meta-analysis of longitudinal MRI studies. Neurosci Biobehav Rev.  2013;37:1680–1691. https://doi.org/ 10.1016/j.neubiorev.2013.06.001 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 130. Ho  BC, Andreasen  NC, Ziebell  S, Pierson  R, Magnotta  V.  Long-term antipsychotic treatment and brain volumes: a longitudinal study of first-episode schizophrenia. Arch Gen Psychiatry.  2011;68:128–137. https://doi.org/ 10.1001/archgenpsychiatry.2010.199 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 131. Navari  S, Dazzan  P.  Do antipsychotic drugs affect brain structure? A systematic and critical review of MRI findings. Psychol Med.  2009;39:1763–1777. https://doi.org/ 10.1017/S0033291709005315 [ DOI ] [ PubMed ] [ Google Scholar ] 132. Lyoo  IK, Dager  SR, Kim  JE, et al.  Lithium-induced gray matter volume increase as a neural correlate of treatment response in bipolar disorder: a longitudinal brain imaging study. Neuropsychopharmacol.  2010;35:1743–1750. https://doi.org/ 10.1038/npp.2010.41 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 133. Chopra  S, Fornito  A, Francey  SM, et al.  Differentiating the effect of antipsychotic medication and illness on brain volume reductions in first-episode psychosis: a longitudinal, randomised, triple-blind, placebo-controlled MRI study. Neuropsychopharmacol.  2021;46:1494–1501. https://doi.org/ 10.1038/s41386-021-00980-0 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 134. Samara  MT, Nikolakopoulou  A, Salanti  G, Leucht  S.  How many patients with schizophrenia do not respond to antipsychotic drugs in the short term? An analysis based on individual patient data from randomized controlled trials. Schizophr Bull.  2019;45:639–646. https://doi.org/ 10.1093/schbul/sby095 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 135. McCutcheon  RA, Pillinger  T, Mizuno  Y, et al.  The efficacy and heterogeneity of antipsychotic response in schizophrenia: a meta-analysis. Mol Psychiatry.  2021;26:1310–1320. https://doi.org/ 10.1038/s41380-019-0502-5 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 136. Lieshout  RJV, MacQueen  GM.  Efficacy and acceptability of mood stabilisers in the treatment of acute bipolar depression: systematic review. Br J Psychiatry.  2010;196:266–273. https://doi.org/ 10.1192/bjp.bp.108.057612 [ DOI ] [ PubMed ] [ Google Scholar ] 137. Tiihonen  J, Lönnqvist  J, Wahlbeck  K, et al.  11-year follow-up of mortality in patients with schizophrenia: a population-based cohort study (FIN11 study). Lancet. 2009;374:620–627. https://doi.org/ 10.1016/S0140-6736(09)60742-X [ DOI ] [ PubMed ] [ Google Scholar ] 138. Correll  CU, Solmi  M, Croatto  G, et al.  Mortality in people with schizophrenia: a systematic review and meta‐analysis of relative risk and aggravating or attenuating factors. World Psychiatry. 2022;21:248–271. https://doi.org/ 10.1002/wps.20994 [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 139. Chen  PH, Tsai  SY, Chen  PY, et al.  Mood stabilizers and risk of all-cause, natural, and suicide mortality in bipolar disorder: a nationwide cohort study. Acta Psychiatr Scand.  2023;147:234–247. https://doi.org/ 10.1111/acps.13519 [ DOI ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials sbaf022_suppl_Supplementary_Figures_S1-S4_Tables_S1-S11 sbaf022_suppl_supplementary_figures_s1-s4_tables_s1-s11.docx (23.7MB, docx) Articles from Schizophrenia Bulletin are provided here courtesy of Oxford University Press ACTIONS View on publisher site PDF (1.5 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Looking for the Self in Psychiatry: Perils and Promises of Phenomenology-Neuroscience Partnership in Schizophrenia Research#167283
• Psychiatric disorders at the crossroads of neurovascular, immune, and metabolic dysregulation: Examining comorbidities from a biopsychosocial and dimensional perspective across the lifespan.#375595
• Mechanisms of Response to Therapeutic Intervention in Clinical High Risk (CHR) for Psychosis#1153201
• Variability and Magnitude of Choline Levels Across the Psychosis Spectrum: A Meta-analysis.#174440
• Simultaneous FMRI and NIRS to Estimate Brain Cerebral Metabolism#888587