Page 1 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Schizophrenia Bulletin. For Peer Review Only http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 2 of 31 1 2 3 4 Full Title: Correlations Between Brain Structure and Symptom Dimensions of Psychosis in 5 Schizophrenia, Schizoaffective and Psychotic Bipolar I Disorders 6 7 Running Title: Brain Structure and Symptom Dimensions in Psychosis 8 9 Jaya L. Padmanabhan 1, 2, Neeraj Tandon 1, 2, Chiara S. Haller 1, 2, 3, 4, Ian T. Mathew 1, 2, Shaun 10 11 M. Eack 5,6, Brett A. Clementz 7, Godfrey D. Pearlson 8, 9, John A. Sweeney 10, 11, Carol A. 12 Tamminga 11, Matcheri S. Keshavan 1, 2, 4 13 14 15 1 Department of Psychiatry, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, 16 17 18 Boston, MA 2 Division of Public Psychiatry, Massachusetts Mental Health Center, 75 Fenwood Road, Boston, 19 MA 20 3 Department of Psychology, Harvard University, Cambridge, MA 21 4 Harvard Medical School, Boston, MA 22 5 School of Social Work, Psychiatry, and Clinical and Translational Sciences Institute, University 23 24 25 of Pittsburgh, Pittsburgh, PA 6 Western Psychiatric Institute and Clinic, Pittsburgh, PA 26 7 Departments of Psychiatry and Neuroscience, Bio-Imaging Research Center, University of 27 Georgia, Athens, GA 28 8 Departments of Psychiatry and Neurobiology, Yale University, New Haven, CT 29 30 31 32 9 Olin Neuropsychiatry Research Center, Hartford Hospital/Institute of Living, Hartford, CT 10 Department of Psychiatry, University of Illinois at Chicago, Chicago, IL 11 Department of Psychiatry, University of Texas Southwestern Medical School, Dallas, TX 33 34 Corresponding Author: 35 36 Matcheri S. Keshavan, MD 37 38 Division of Public Psychiatry, Massachusetts Mental Health Center, 75 Fenwood Road, Boston 39 MA 02115 40 41 Email: mkeshava@bidmc.harvard.edu 42 Phone: 617-754-1256 43 44 Fax: 617-754-1250 45 46 47 Abstract Word Count: 233 words 48 49 Total Word Count: 3944 words 50 51 52 53 54 55 56 57 58 59 60 1 http://www.schizophreniabulletin.oupjournals.org Page 3 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 Abstract 5 6 Background: Structural alterations may correlate with symptom severity in psychotic disorders, 7 but the existing literature on this issue is heterogeneous. Additionally, it is not known how 8 cortical thickness and cortical surface area correlate with symptom dimensions of psychosis. 9 10 11 Methods: Subjects included 455 individuals with schizophrenia, schizoaffective, or bipolar I 12 disorders. Data were obtained as part of the Bipolar Schizophrenia Network for Intermediate 13 Phenotypes (BSNIP) study. Diagnosis was made through the Structured Clinical Interview for 14 15 DSM-IV. Positive and negative symptom subscales were assessed using the Positive and 16 Negative Syndrome Scale (PANSS). Structural brain measurements were extracted from T1- 17 18 weight structural MRIs using FreeSurfer v5.1 and were correlated with symptom subscales using 19 partial correlations. Exploratory factor analysis was also used to identify factors among those 20 regions correlating with symptom subscales. 21 22 Results: The positive symptom subscale correlated inversely with gray matter volume (GMV) 23 24 and cortical thickness in frontal and temporal regions, while the negative symptom subscale 25 correlated inversely with right frontal cortical surface area. Among regions correlating with the 26 27 positive subscale, factor analysis identified four factors, including a temporal cortical thickness 28 factor and frontal GMV factor. Among regions correlating with the negative subscale, factor 29 analysis identified a frontal GMV-cortical surface area factor. There were no significant 30 31 diagnosis by structure interactions with symptom severity. 32 33 Conclusion: Structural measures correlate with positive and negative symptom severity in 34 psychotic disorders. Cortical thickness demonstrated more associations with psychopathology 35 36 than cortical surface area. 37 38 Keywords: positive, negative, cortical thickness, surface area, psychopathology, gray matter 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 4 of 31 1 2 3 4 5 6 Introduction 7 8 Structural imaging studies have established the presence of subtle structural brain 9 10 11 alterations in psychotic disorders. For schizophrenia, some of the most consistent findings 12 13 include reductions in gray matter volume (GMV) of frontal, temporal, and limbic regions 1, while 14 15 bipolar disorder has also been associated with GMV reductions in prefrontal, temporal, and 16 17 18 limbic regions 2,3. 19 20 Studies have found correlations between structural alterations and symptom dimensions 21 22 of psychosis; however, findings have been heterogeneous. In schizophrenia, inverse correlations 23 24 25 between positive symptom severity and GMV of temporal lobe regions, most commonly the 26 27 superior temporal gyrus (STG), have been frequently reported 4-8, but a minority of studies have 28 29 30 observed no correlation or a direct correlation between positive symptom severity and GMV of 31 32 these regions 9-11. Results have been similarly mixed for the negative symptom dimension, with 33 34 several studies finding inverse correlations with GMV of frontal regions 12-14, and other studies 35 36 37 reporting no correlation or a positive correlation with frontal regions 15-17. 38 39 Multiple reasons may account for the heterogeneity in findings. Positive symptoms of 40 41 42 psychosis can wax and wane with time; thus results may be influenced by illness acuity of 43 44 subjects at the time of scan. Other reasons for heterogeneity of results may include variations in 45 46 technical aspects of imaging methodology, variable adjustment for confounding factors, and 47 48 49 differences in subject characteristics, such as duration of illness. 50 51 In addition to the heterogeneity of findings on clinical correlations with GMV, the 52 53 54 literature is limited regarding correlations with non-volumetric structural measures. Two 55 56 constituents of volume, cortical thickness (CT) and cortical surface area (CSA), have 57 58 59 60 3 http://www.schizophreniabulletin.oupjournals.org Page 5 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 demonstrated significant alterations in both schizophrenia 18-20 and bipolar disorder 21-23. They 5 6 may correlate differently with psychopathology due to their distinct neurobiological and genetic 7 8 origins. According to a prevailing theory of cortical development, CSA is determined by the total 9 10 11 number of cortical columns that form the cerebral cortex, while CT is determined by number of 12 13 neurons within each column 24-26. A recent longitudinal neuroimaging study of children found 14 15 evidence for independent developmental trajectories of CT and CSA 27. Additionally, 16 17 18 neuroimaging studies of twins have found that CSA and CT are both highly heritable but 19 20 probably genetically distinct 28. 21 22 23 Thus far, the literature has not established whether CT and CSA have distinct correlations 24 25 with positive and negative symptoms. Correlations have been reported between CT and 26 27 propensity for hallucinations 19 and positive symptoms 29 among individuals with schizophrenia. 28 29 30 However, one cross-diagnostic study of psychosis did not find an association between CT and 31 32 symptom dimensions of psychosis 22. Analysis of symptom correlations with CT and CSA may 33 34 reveal their differential contributions to psychopathology, which in turn could help identify more 35 36 37 specific neuropathological processes that drive the emergence of psychotic symptoms. In 38 39 addition, inclusion of multiple diagnostic categories may reveal whether symptom-structure 40 41 correlations are trans-diagnostic. 42 43 44 In this study, we examined correlations between symptom dimensions of psychosis and 45 46 regional GMV, CSA and CT in schizophrenia, schizoaffective disorder and bipolar I disorder 47 48 49 with psychosis. Correlations were examined using partial correlations between individual regions 50 51 and subscales, and factor analysis was used to summarize overall structure-symptom 52 53 relationships. Based on the existing literature 30, we hypothesized that temporal alterations would 54 55 56 correlate with the positive subscale, while frontal alterations would correlate with the negative 57 58 59 60 4 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 6 of 31 1 2 3 4 subscale. We also hypothesized that CT and CSA would show distinct correlations with symptom 5 6 dimensions. This is one of the largest sample sizes to date in which associations with dimensions 7 8 of psychosis have been examined. 9 10 11 12 13 Methods 14 15 16 Participants 17 18 Subjects included individuals with a DSM-IV diagnosis of schizophrenia, schizoaffective 19 20 21 disorder, or bipolar I disorder with psychotic features. Subjects were recruited as part of the 22 23 Bipolar-Schizophrenia Network for Intermediate Phenotypes (B-SNIP), using recruitment 24 25 methods that have been detailed elsewhere 31,32. Inclusion criteria included the following: (1) age 26 27 28 15 – 65; (2) English proficiency, as determined by ability to follow task instructions; (3) no 29 30 known history of neurologic disorders including head injury; (4) no history of substance abuse 31 32 within the last month or substance dependence within the last 6 months; and (5) negative urine 33 34 35 toxicology screen on day of testing. Patients were generally clinically stable and receiving 36 37 consistent psychopharmacological treatment for 4 weeks prior to testing. Study protocols were 38 39 approved by institutional review boards at each study site, and subjects signed informed consent 40 41 42 forms. 43 44 All subjects received the Structured Clinical Interview for DSM-IV (SCID-IV) 33. A 45 46 47 consensus process was used to establish diagnosis using results from the SCID-IV, chart review, 48 49 and review of psychiatric and medical histories. The Positive and Negative Symptom Scale 50 51 (PANSS) 34 was used to assess positive, negative, and general symptoms in patients. Inter-site 52 53 54 standardization of symptom ratings was carried out by periodic meetings for rater training, using 55 56 established ‘gold standard’ interviews. At the beginning of the study, there was a face-to-face 57 58 59 60 5 http://www.schizophreniabulletin.oupjournals.org Page 7 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 training session for all raters, with a requirement for reliability of >0.85. Rater training was 5 6 repeated annually to re-establish reliability 31. 7 8 A total of 455 patients had complete datasets available for structural MRI and symptom 9 10 11 scales for analysis of structure-symptom correlations 31,32. Of these 455 subjects, 181 individuals 12 13 were categorized as having schizophrenia, 117 individuals as having schizoaffective disorder, 14 15 and 157 individuals as having bipolar I disorder with psychotic features. 352 healthy controls 16 17 18 were also evaluated in the larger study; they received structural MRIs but did not receive PANSS 19 20 assessments. 21 22 23 24 25 MRI-structural imaging 26 27 28 High-resolution isotropic T1-weighted MPRAGE sequences were obtained. Sites used 29 30 comparable but slightly different MPRAGE acquisition parameters; full details for each site have 31 32 been described previously 31. The Alzheimer’s Disease Neuroimaging Initiative (ADNI) protocol 33 34 35 was used at all sites to standardize imaging analysis (http://www.loni.ucla.edu/ADNI). All 36 37 images were subjected to a rigorous data quality control process. First, images were opened, 38 39 converted to nifti format, and checked for scanner artifacts. If the images passed through this 40 41 42 check, they were run through auto-recon 1 in FreeSurfer v5.1 35. Images were then checked for 43 44 remaining non-brain tissues (dura or sinus). Trained raters, all reliable above 95%, edited images 45 46 47 to remove any remaining non-brain tissue. An independent rater then determined if images were 48 49 adequately cleaned for segmentation, and images were then processed through auto-recon 2 and 50 51 3. Freesurfer v5.1 software was used to extract regional GMV, CT and CSA measurements. 52 53 54 55 56 Statistical Analysis 57 58 59 60 6 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 8 of 31 1 2 3 4 All statistical analysis was done using the program R (Vienna, Austria; 2013, 5 6 http://www.R-project.org, version 2.15.3). Data were examined for bivariate normality using the 7 8 multivariate Shapiro-Wilk test (R package: mvnormtest), revealing that clinical symptom 9 10 11 measures were not normally distributed. Non-parametric tests were used for further analyses. 12 13 Diagnostic differences in demographic variables and symptom subscales were tested through the 14 15 Kruskal-Wallis and chi-squared tests. Partial correlations were performed to correlate individual 16 17 18 structural measures with symptom subscales. Additionally, a factor analysis approach was used 19 20 to identify structural factors among regions associated with symptom subscales. 21 22 23 24 25 Partial correlations between structural measures and symptom subscales 26 27 28 A series of variables were tested for potential inclusion as co-variates. These included 29 30 age, sex, race, study site, intracranial volume (ICV), socioeconomic status, patient educational 31 32 level, duration of illness, and antipsychotic medication status (a binary variable representing 33 34 35 whether or not the patient was currently on an antipsychotic). Duration of illness was computed 36 37 by subtracting age at illness onset from current age. Socioeconomic status was represented by 38 39 patient Hollingshead Index score, while patient educational level was represented by patient 40 41 42 years of education. Study site was treated as a categorical variable, with each site being ‘dummy 43 44 coded’ as a binary variable for regression analyses. Variables were retained as co-variates if they 45 46 correlated with either structural measures or symptom subscales, using the Kruskal-Wallis test 47 48 49 for categorical variables and the Spearman correlation for continuous variables. 50 51 Using partial Spearman correlations, symptom subscales were first correlated with GMV, 52 53 54 CSA, and mean CT of each lobe, and were Hochberg-adjusted for multiple comparisons (32 55 56 comparisons per subscale) 36. For each lobe that demonstrated statistically significant 57 58 59 60 7 http://www.schizophreniabulletin.oupjournals.org Page 9 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 correlations with a symptom subscale, sub-regions of that lobe were then correlated with that 5 6 symptom subscale in a step-down fashion (see Supplemental Table 1 for lists of the sub-regions 7 8 that constituted each lobe). These correlations were again Hochberg-adjusted for multiple 9 10 11 comparisons by the total number of sub-regions tested for correlations with that symptom 12 13 subscale. 14 15 GMV and CSA for each lobe were computed by adding the GMVs or CSAs of 16 17 18 component sub-regions. A mean CT for each lobe was determined by calculating a weighted 19 20 average of the cortical thicknesses of component sub-regions (i.e., CT of each sub-region 21 22 23 multiplied by CSA of that sub-region, divided by total CSA for that lobe). 24 25 Lobes and sub-regions that showed significant correlations in the whole-group analysis 26 27 were then tested for correlations with symptom subscales within each diagnostic group, and were 28 29 30 corrected for total number of correlations tested within that diagnostic group. Symptom 31 32 subscales and PANSS total scores were also correlated with total GMV. Finally, a supplemental 33 34 35 analysis was done using current antipsychotic dose in chlorpromazine equivalents, which was 36 37 only available for 295 out of 455 patients. Analyses were repeated in this subset of patients with 38 39 and without chlorpromazine dose as a co-variate to evaluate the impact of this co-variate. 40 41 42 43 44 Exploratory factor analysis 45 46 47 To construct an enriched sample of brain regions, all regional structural measures (80 48 49 GMV measures, 66 CSA measures, and 66 CT measures) were first screened for correlations 50 51 with the positive and negative symptom subscales. Regional measures that correlated with the 52 53 54 positive subscale at a significance level of p < 0.05, uncorrected, were entered into factor 55 56 analysis. These regional measures were regressed against age, ICV, sex, and race. The residuals 57 58 59 60 8 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 10 of 31 1 2 3 4 of these regressions were then used to create the correlation matrix for factor analysis, as has 5 6 been done with structural brain measures previously 37,38; that is, the correlation matrix for factor 7 8 analysis consisted of partial correlations among these regional measures, controlling for co- 9 10 11 variates of age, ICV, sex, and race. 12 13 Principal factor extraction (also called ‘principal axes factoring’) was used to extract 14 15 factors because of deviations from multivariate normality 39. A scree test was performed to obtain 16 17 18 an initial estimate of number of factors 40,41. Additionally, several factor analyses were performed 19 20 using different numbers of factors and were evaluated for overall factor structure. A final number 21 22 23 of factors was chosen if it produced factors with no or few item cross-loadings and at least three 24 25 variable loadings above 0.45 42. Direct oblique rotation was performed to assess if factors were 26 27 correlated with each other. Factor scores for subjects were derived using regression to maximize 28 29 30 determinacy of scores 43. 31 32 Each factor score was entered as a predictor into a regression model with a full set of co- 33 34 35 variates (see previous section) to verify associations with the positive symptom subscale. Next, 36 37 all factor scores were entered into a single model with co-variates to assess their relative 38 39 correlations with symptom subscale scores. To assess specificity of correlations between factor 40 41 42 scores and symptom subscales, factor scores were also tested for correlations with the negative 43 44 subscale. Finally, to assess diagnosis by structure interactions, models were re-run with the 45 46 inclusion of diagnosis and diagnosis by structure interaction terms. Assumptions of multiple 47 48 49 regression were tested through visualization of QQ plots and residual-versus-fitted plots. 50 51 An identical and separate factor analysis process was performed using those regions that 52 53 54 correlated with the negative symptom subscale. 55 56 57 58 59 60 9 http://www.schizophreniabulletin.oupjournals.org Page 11 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 Results 5 6 Demographics 7 8 9 Subject demographics and study site demographics are presented in Table 1 and 10 11 Supplemental Table 2, respectively. Mean values for symptom subscales varied across diagnostic 12 13 groups and study sites (p < 0.05 using Kruskal-Wallis rank sum test). Post-hoc comparisons of 14 15 16 symptom subscales in schizophrenia and bipolar I disorder indicated that all symptom subscales 17 18 were significantly higher in the schizophrenia group (p < 0.05 using Wilcoxon rank sum test). 19 20 [Table 1] 21 22 23 Partial correlations between structural measures and symptom subscales 24 25 All potential co-variates demonstrated correlations with at least one symptom subscale 26 27 28 and were retained as co-variates in further analyses. Additionally, age, sex, race, study site, 29 30 duration of illness, and intracranial volume correlated with structural measures (p-adjusted < 31 32 33 0.05). 34 35 Correlations with structural measures in combined group. PANSS positive symptom 36 37 subscale was correlated with frontal and temporal GMV reductions and temporal CT reductions, 38 39 40 while the PANSS negative subscale was correlated with reductions in right frontal CSA (p- 41 42 adjusted < 0.05, Table 2, Figure 1, Supplemental Figure 1). There were no correlations with the 43 44 PANSS general subscale. 45 46 47 [Table 2] 48 49 Correlations with structural measures within diagnostic groups. These structural 50 51 52 measures were then evaluated for symptom-subscale correlations within diagnostic groups. 53 54 Correlations remained significant within the schizophrenia group, but only the correlation with 55 56 57 58 59 60 10 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 12 of 31 1 2 3 4 the right insula CT would have survived correction for multiple comparisons. Correlations within 5 6 bipolar disorder and schizoaffective disorder were largely non-significant (Table 2). 7 8 Correlations with total GMV. In the combined group of all patients, total GMV 9 10 11 correlated inversely with the PANSS positive subscale (r = -0.177, p = 0.00015) and the PANSS 12 13 total score (r = -0.118, p = 0.011), but not with the PANSS negative subscale (r = -0.061, p = 14 15 0.19) or the PANSS general subscale (r = -0.0827, p = 0.08). 16 17 18 Among subjects with available data on antipsychotic dose in chlorpromazine equivalents, 19 20 effect sizes of correlations decreased when chlorpromazine equivalent medication dose was 21 22 23 included as a co-variate, but the overall pattern of correlations was very similar. 24 25 [Figure 1] 26 27 28 Exploratory Factor Analysis 29 30 Positive symptom subscale. Initial screening revealed that 58 out of 212 regions were 31 32 33 correlated with the positive subscale (p < 0.05, uncorrected). Scree test suggested between 4 and 34 35 6 factors, and further evaluation supported four reliable factors: a temporal CT factor, a frontal 36 37 GMV factor, a fronto-parietal CT factor, and a precuneus GMV-SA factor (Table 3, Supplemental 38 39 40 Table 3). There were no cross-loadings of variables. Direct oblique rotation was retained because 41 42 factors were correlated. 43 44 [Table 3] 45 46 47 All four factor scores were significant predictors of the positive subscale in separate 48 49 regression models with co-variates (temporal CT score: B = -1.04, p = 0.00035; frontal GMV 50 51 52 score: B = -1.16, p = 0.000047; fronto-parietal CT score: B = -0.66, p = 0.024; precuneus GMV- 53 54 SA score: B = -1.0, p = 0.00071). When all four factor scores were entered into the same 55 56 regression model with co-variates, the temporal CT score (B = -0.80, p = 0.028) and frontal 57 58 59 60 11 http://www.schizophreniabulletin.oupjournals.org Page 13 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 GMV score (B = -0.71, p = 0.032) remained significant predictors of the positive subscale 5 6 (adjusted R2 = 0.176). There were no significant diagnosis by factor interactions. None of the 7 8 factor scores were significant predictors of the negative subscale. 9 10 11 Negative symptom subscale. Fourteen out of 212 regions correlated with the negative 12 13 subscale upon initial screen (p < 0.05, uncorrected). Scree test suggested one factor, and 14 15 extraction of more than one factor led to degenerate factor structure and cross-loading of 16 17 18 variables. Further evaluation supported a frontal GMV-SA factor (Table 3, Supplemental Table 19 20 4). No rotation was performed because of the single factor structure. This factor score was a 21 22 23 significant predictor of the negative subscale (B = -0.99, p = 0.00032, adjusted R2 = 0.139). The 24 25 diagnosis by factor interaction was not significant. Lastly, this factor was a significant predictor 26 27 of the positive subscale (B = -0.87, p = 0.0019, adjusted R2 = 0.150). 28 29 30 31 32 Discussion 33 34 35 This study examined correlations between symptom dimensions and regional GMV, CT, 36 37 and CSA, in a group of individuals with schizophrenia, schizoaffective disorder, or bipolar I 38 39 disorder with psychotic features. Partial correlations were used to evaluate symptom correlations 40 41 42 with individual regions, and factor analysis was used to summarize and compare structure- 43 44 symptom relationships. The PANSS positive subscale correlated with reductions in both temporal 45 46 47 and frontal structural measures. Among regions correlating with the positive subscale, factor 48 49 analysis identified a temporal CT factor, a frontal GMV factor, a fronto-parietal CT factor, and a 50 51 precuneus GMV-SA factor, of which the temporal CT and frontal GMV factors independently 52 53 54 predicted positive symptom severity when all four factors were jointly entered in a regression 55 56 model. 57 58 59 60 12 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 14 of 31 1 2 3 4 Temporal regions have been previously implicated in the production of psychotic 5 6 symptoms. Functional MRI studies have noted activation of temporal regions during real-time 7 8 auditory hallucinations 44, while structural MRI studies have found associations between 9 10 11 temporal regions and both hallucinations 4,45 and thought disorder 46. While frontal associations 12 13 with the positive subscale have been observed less frequently in the literature, several other 14 15 studies have found correlations between the positive subscale and reductions in overall frontal 16 17 18 volume 47 or GMV in regions such as the inferior frontal gyrus 48,49. Abnormalities in frontal 19 20 regions could impact cognitive processes of working memory, attention, and language processing 21 22 50, contributing to positive symptoms such as disorganization and delusions. 23 24 25 Comparing diagnostic groups, structure-symptom correlations were larger in the 26 27 schizophrenia group than in the two other diagnostic categories. However, our factor analysis did 28 29 30 not find significant interactions between diagnosis and structural factors on symptom subscales, 31 32 indicating no major impact of diagnosis on structure-symptom correlations. Overall, findings in 33 34 the schizophrenia group may reflect their higher level of subtle brain pathology, which may in 35 36 37 turn be associated with their more chronic and persistent psychotic symptoms compared with 38 39 schizoaffective and bipolar I disorders. Differences in exposure to antipsychotic medication are 40 41 unlikely to completely account for the larger effect sizes of correlations in schizophrenia. While 42 43 44 individuals with schizophrenia typically have greater exposure to antipsychotics than individuals 45 46 with bipolar disorder, our secondary analyses found that antipsychotic medication dose did not 47 48 49 have a major impact on effect sizes of correlations. 50 51 The PANSS negative subscale correlated inversely with CSA of right frontal regions, and 52 53 factor analysis identified a frontal GMV-SA factor among those regions that correlated with the 54 55 56 negative subscale. These results corroborate prior studies reporting correlations between frontal 57 58 59 60 13 http://www.schizophreniabulletin.oupjournals.org Page 15 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 reductions and negative symptoms 13,51-53, and inverse correlations between regional cerebral 5 6 blood flow and negative symptoms on positron emission tomography (PET) scans 54. Overall, the 7 8 negative subscale demonstrated fewer correlations with structural measures than the positive 9 10 11 subscale. This observation may indicate the relatively greater contribution of social and non- 12 13 structural biological influences in the manifestation of negative symptoms. 14 15 As predicted, CT and CSA diverged in their associations with symptom subscales. CT 16 17 18 reductions exhibited more correlations with symptom subscales, particularly the positive 19 20 subscale, than CSA. Interestingly, one recent trans-diagnostic study of schizophrenia and bipolar 21 22 23 I found more widespread regional reductions in CT than in CSA 20. This study (and an earlier 24 25 study with the same subject sample) did not find associations of either type of structural measure 26 27 with symptom subscales 20,22. Our findings suggest that CT may be more closely associated with 28 29 30 symptoms of psychosis than CSA. These findings may reflect the distinct neurobiological 31 32 processes underlying these two aspects of structure. As mentioned earlier, recent research 33 34 indicates that CT and CSA may have distinct genetic influences 28, may follow independent 35 36 37 developmental trajectories in childhood 27, and may not be highly correlated with each other 20. 38 39 CT has been shown to fluctuate in response to environmental factors such as cannabis use 55,56 40 41 and childhood trauma 56, and may represent a “state” marker that tracks more closely with 42 43 44 fluctuating positive symptoms than CSA. Given the more widespread correlations of CT with the 45 46 symptom dimensions, particularly the positive subscale, future investigations may wish to focus 47 48 49 on pathophysiological processes of CT as driving the development of psychosis. 50 51 Notably, while correlations were significant, their magnitudes were small, indicating that 52 53 other factors likely make independent contributions to symptom severity. Symptom severity is 54 55 56 partly driven by social factors, such as education and socioeconomic status, whose 57 58 59 60 14 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 16 of 31 1 2 3 4 neurobiological effects may not be captured by structural measures. Additionally, small effect 5 6 sizes of structure-symptom correlations may complement the findings of neuropathology studies 7 8 of schizophrenia, which observe subtle reductions in cortical neuropil and volumes of neuronal 9 10 11 cell bodies, rather than loss of neurons 57,58. The effect sizes of our correlations may also reflect 12 13 the influence of processes that enlarge structural measures. For example, inflammation, which 14 15 may be important in early stages of schizophrenia 59, could lead to structural enlargement 16 17 18 through free water retention, thus reducing the strength of inverse correlations between structure 19 20 and symptoms. Overall, the small effect sizes of structure-symptom correlations corroborate the 21 22 concept of psychosis as a disorder of network connectivity 60, involving subtle neurochemical 23 24 25 and neurophysiological alterations in the interactions between brain regions. Our observations of 26 27 frontal and temporal contributions to psychosis is consistent with the possibility that fronto- 28 29 30 temporal connectivity may be of particular importance in the pathogenesis of positive symptoms 31 32 61. 33 34 This study had several strengths. This is one of the largest sample sizes thus far in which 35 36 37 this question has been examined. Inclusion of all three diagnostic categories permitted the 38 39 examination of associations in psychosis in a trans-diagnostic fashion, and inclusion of CT and 40 41 CSA permitted exploration of their relative associations with symptom subscales. Many potential 42 43 44 confounding factors were included in the analysis. Additionally, the subject sample consisted of 45 46 clinically stable, chronically ill individuals. Structural and physiological abnormalities may 47 48 49 change across the course of illness in psychosis 62, and structural alterations may be more 50 51 relevant in understanding persistent symptoms in a chronic population. Thus, our results may 52 53 reflect more stable correlations between brain structure and residual, treatment-resistant 54 55 56 psychopathology. 57 58 59 60 15 http://www.schizophreniabulletin.oupjournals.org Page 17 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 There were several limitations. Because data were cross-sectional, it was not possible to 5 6 draw conclusions about causal relationships. The region-of-interest analysis necessitated a heavy 7 8 correction for multiple comparisons, which may have obscured some findings. Additionally, data 9 10 11 on subjects’ lifetime history of antipsychotic use were not collected, although current 12 13 antipsychotic use was included. Longitudinal antipsychotic use may be associated with gray 14 15 matter changes 63 and may have contributed to structure-symptom correlations in this study. Last, 16 17 18 neuropsychological test results were not analyzed here, though they have been reported 19 20 elsewhere in this sample 64. 21 22 23 In conclusion, among a combined group of individuals with schizophrenia, 24 25 schizoaffective, and bipolar I disorders, the PANSS positive subscale was inversely correlated 26 27 with GMV and cortical thickness in frontal and temporal regions, while the PANSS negative 28 29 30 subscale was inversely correlated with frontal cortical surface area and GMV. Overall, cortical 31 32 thickness appeared more strongly associated with psychopathology, particularly the positive 33 34 subscale, than cortical surface area. However, the magnitudes of all correlations were low. These 35 36 37 results lend support to associations between structural brain alterations and severity of 38 39 psychopathology. 40 41 42 43 44 Funding 45 46 47 This work was supported by National Institute of Mental Health grants MH078113 (to 48 49 M.S.K), MH077945 (to G.D.P.), MH077852 (to Gunvant Thaker, MD), MH077851 (to C.A.T.), 50 51 and MH077862 (to J.A.S.). 52 53 54 55 56 Acknowledgements 57 58 59 60 16 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 18 of 31 1 2 3 4 The authors would like to thank Dr. Gunvant Thaker for his contributions to the Bipolar- 5 6 Schizophrenia Network for Intermediate Phenotypes consortium. 7 8 9 10 11 Conflicts of Interest 12 13 Dr. Keshavan has received research support from Sunovion and GlaxoSmithKline. 14 15 16 Dr. Padmanabhan has received grant support from the Janssen Academic Research Mentorship 17 18 program. 19 20 Dr. Pearlson has served on an advisory panel for Bristol-Myers Squibb. 21 22 23 Dr. Sweeney has been on advisory boards for Bristol-Myers Squibb, Eli Lilly, Pfizer, Roche, and 24 25 Takeda and has received grant support from Janssen. 26 27 Dr. Tamminga has the following disclosures to make: Intracellular Therapies (ITI, Inc.)— 28 29 30 Advisory Board, drug development; PureTech Ventures—Ad Hoc Consultant; Eli Lilly Pharma- 31 32 ceuticals—Ad Hoc Consultant; Sunovion—Ad Hoc Consultant; Astellas—Ad Hoc Consultant; 33 34 35 Cypress Bioscience—Ad Hoc Consultant; Merck—Ad Hoc Consultant; International Congress 36 37 on Schizophrenia Research—Organizer, unpaid volunteer; National Alliance on Mental Illness— 38 39 Council Member, unpaid volunteer; American Psychiatric Association—Deputy Editor. 40 41 42 The other authors report no disclosures. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 17 http://www.schizophreniabulletin.oupjournals.org Page 19 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 Table 1. Subject Demographics and Symptom Scale Characteristics 5 All Patients Schizophrenia Schizoaffective Bipolar I 6 7 n 455 181 117 157 8 Mean Age (sd) 36.2 (12.7) 35.8 (12.6) 36.0 (12.1) 36.7 (13.3) 9 10 11 Sex M : F a (% M : % F) 220 : 235 48% : 52% 118 : 63 65% : 35% 56 : 61 48% : 52% 46 : 111 29% : 71% 12 AA: 184 (40 %) AA: 94 (52 %) AA: 52 (44 %) AA: 38 (24 %) 13 Race (%)a,b CA: 242 (53 %) CA: 73 (40 %) CA: 58 (50 %) CA: 111 (71 %) 14 OT: 29 (6 %) OT: 14 (8 %) OT: 7 (6 %) OT: 8 (5 %) 15 16 Mean PANSS Positive (sd)c 16.2 (5.5) 17.2 (5.4) 18.6 (4.9) 13.3 (4.8) 17 Mean PANSS Negative (sd)c 14.9 (5.3) 18 19 Mean PANSS General (sd)c 32.2 (9.0) 16.8 (5.8) 32.8 (9.1) 15.4 (4.5) 35.0 (8.7) 12.3 (4.1) 29.5 (8.4) 20 Mean PANSS Total (sd)c 63.3 (16.9) 66.7 (16.9) 69.1 (15.4) 55.1 (14.7) 21 22 Mean Intracranial Volume (sd)c 1437 cc (186 cc) 1467 cc (197 cc) 1396 cc (177 cc) 1434 cc (174 cc) 23 24 Mean Duration of Illness (sd)c 18.6 yrs (12.3 yrs) 16.4 yrs (11.8 yrs) 19.9 yrs (11.7 yrs) 20.1 yrs (12.9 yrs) 25 Mean Years of Education (sd)c 13.4 yrs (2.4 yrs) 12.8 yrs (2.3 yrs) 13.2 yrs (2.2 yrs) 14.3 yrs (2.4 yrs) 26 27 Mean Hollingshead Score (sd)c,d 48.2 (15.7) 53.2 (14.6) 48.7 (14.6) 42.1 (15.7) 28 29 Antipsychotic Status (%)a,e Yes: 85%; No: 15% Yes: 92%; No:8% Yes: 86%; No: 14% Yes: 75%; No: 25% 30 asignificantly different across diagnostic groups by chi-squared test 31 bAA = African-American; CA = Caucasian-American; OT = Other 32 csignificantly different across diagnostic groups by Kruskal-Wallis test 33 dHollingshead occupation score multiplied by 7, added to Hollingshead education score multiplied by 4; higher score indicates lower social class 34 ecurrently taking antipsychotics (yes or no) 35 36 37 38 39 40 41 42 43 44 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 18 Schizophrenia Bulletin. For Peer Review Only Page 20 of 31 1 2 3 4 5 6 Table 2. Structure-Subscale Correlations Within Combined Group and Within Diagnostic Categories 7 Combined Group Schizophrenia Schizoaffective Bipolar I 8 9 Lobe Measurement Subscale r p-adjusteda r p-valueb r p-valueb r p-valueb 10 Left Frontal GMV -0.162 0.016 -0.176 0.018 -0.072 0.44 -0.067 0.41 11 12 Left pars orbitalis GMV PANSS Positive -0.182 0.004 -0.113 0.130 -0.139 0.13 -0.141 0.078 13 Left superior frontal -0.166 0.013 -0.170 0.022 -0.170 0.067 -0.015 0.85 14 15 16 17 Right Frontal GMV Right superior frontal GMV PANSS Positive -0.165 -0.180 0.013 0.005 -0.170 -0.160 0.022 0.032 -0.015 -0.155 0.88 0.095 -0.031 0.010 0.70 0.90 18 19 20 21 Right Temporal GMV Right superior temporal GMV PANSS Positive -0.158 -0.155 0.020 0.031 -0.141 -0.155 0.058 0.036 -0.054 -0.148 0.57 -0.020 0.11 -0.020 0.80 0.80 22 Right fusiform -0.178 0.005 -0.089 0.240 -0.145 0.12 -0.067 0.40 23 24 Right Temporal CT -0.173 0.007 -0.183 0.013 -0.079 0.40 -0.047 0.56 25 Right middle temporal Cortical PANSS -0.150 0.045 -0.199 0.0072 0.025 0.79 -0.091 0.26 26 27 Right superior temporal Thickness Positive -0.155 0.032 -0.138 0.064 -0.102 0.27 -0.104 0.19 28 Right insula 29 -0.175 0.006 -0.236 0.0014c -0.075 0.42 0.021 0.80 30 Right Frontal Area -0.151 0.041 -0.153 0.040 0.081 0.34 -0.023 0.77 31 32 33 Right pars orbitalis Right superior frontal Surface Area PANSS Negative -0.138 -0.138 0.026 0.026 -0.145 -0.101 0.051 0.17 -0.150 0.016 0.11 -0.093 0.25 0.86 -0.190 0.018 34 35 Right precentral -0.159 0.007 -0.150 0.043 -0.064 0.49 -0.162 0.042 36 a p-values Hochberg-adjusted as described in the text 37 b unadjusted p-value 38 c significant after Hochberg correction 39 40 41 42 43 44 19 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 Page 21 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 Table 3. Factor Analysis: Regions Loading Above 0.45 on Factors Derived Through Principal Factors Extraction 4 5 6 Positive Subscale: Four Factors Negative Subscale: One Factor 7 8 F1: Temporal CT F2: Frontal GMV F3: Fronto-parietal CT F4: Precuneus F: Frontal GMV-SA 9 10 L fusiform CT R lateral orbitofrontal GMV L superior parietal CT R precuneus GMV R superior frontal GMV 11 R middle temporal CT R superior frontal GMV R inferior parietal CT 12 13 R superior temporal CT L lateral orbitofrontal GMV L superior frontal CT L precuneus GMV R superior frontal CSA R precuneus CSA R precentral GMV 14 R fusiform CT L superior frontal GMV R lateral occipital CT R fusiform CSA R precentral CSA 15 16 L inferior temporal CT R superior frontal CSA R supramarginal CT R lateral orbitofrontal CSA 17 R inferior temporal CT R rostral middle frontal GMV R superior frontal CT R pars orbitalis CSA 18 R insula CT L superior frontal CSA L lateral occipital CT 19 20 L fusiform GMV L pars orbitalis GMV L supramarginal CT R pars orbitalis GMV R paracentral CSA 21 L insula CT R pars orbitalis GMV L superior parietal GMV 22 23 R temporal pole CT L rostral middle frontal GMV R lingual CT L superior parietal GMV L posterior cingulate GMV 24 R banks of superior 25 temporal sulcus CT 26 27 R fusiform GMV 28 R middle temporal GMV 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 20 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 Schizophrenia Bulletin. For Peer Review Only Page 22 of 31 1 2 3 Figure 1. Significant correlations between lobe measures and positive symptom severity for all patients and diagnostic groups; plotted values are 4 5 residuals after adjustment for co-variates. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 21 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 Page 23 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 References: 4 5 6 7 1. Shepherd AM, Laurens KR, Matheson SL, Carr VJ, Green MJ. Systematic meta-review and quality assessment of the structural brain alterations in schizophrenia. Neurosci Biobehav Rev. 8 Apr 2012;36(4):1342-1356. 9 2. Selvaraj S, Arnone D, Job D, et al. Grey matter differences in bipolar disorder: a meta-analysis of 10 voxel-based morphometry studies. Bipolar Disord. Mar 2012;14(2):135-145. 11 3. Ellison-Wright I, Glahn DC, Laird AR, Thelen SM, Bullmore E. The anatomy of first-episode and 12 chronic schizophrenia: an anatomical likelihood estimation meta-analysis. Am J Psychiatry. Aug 13 2008;165(8):1015-1023. 14 4. Barta PE, Pearlson GD, Powers RE, Richards SS, Tune LE. Auditory hallucinations and smaller 15 superior temporal gyral volume in schizophrenia. Am J Psychiatry. Nov 1990;147(11):1457-1462. 16 5. Flaum M, O'Leary DS, Swayze VW, 2nd, Miller DD, Arndt S, Andreasen NC. Symptom 17 18 19 20 dimensions and brain morphology in schizophrenia and related psychotic disorders. J Psychiatr Res. Jul-Aug 1995;29(4):261-276. 6. Lui S, Deng W, Huang X, et al. Association of cerebral deficits with clinical symptoms in 21 antipsychotic-naive first-episode schizophrenia: an optimized voxel-based morphometry and 22 resting state functional connectivity study. Am J Psychiatry. Feb 2009;166(2):196-205. 23 7. Nestor PG, Onitsuka T, Gurrera RJ, et al. Dissociable contributions of MRI volume reductions of 24 superior temporal and fusiform gyri to symptoms and neuropsychology in schizophrenia. 25 Schizophr Res. Mar 2007;91(1-3):103-106. 26 8. Onitsuka T, Shenton ME, Salisbury DF, et al. Middle and inferior temporal gyrus gray matter 27 volume abnormalities in chronic schizophrenia: an MRI study. Am J Psychiatry. Sep 28 2004;161(9):1603-1611. 29 9. Turetsky B, Cowell PE, Gur RC, Grossman RI, Shtasel DL, Gur RE. Frontal and temporal lobe 30 brain volumes in schizophrenia. Relationship to symptoms and clinical subtype. Arch Gen 31 32 33 Psychiatry. Dec 1995;52(12):1061-1070. 10. Fannon D, Chitnis X, Doku V, et al. Features of structural brain abnormality detected in first- 34 episode psychosis. Am J Psychiatry. Nov 2000;157(11):1829-1834. 35 11. Molina V, Reig S, Pascau J, et al. Anatomical and functional cerebral variables associated with 36 basal symptoms but not risperidone response in minimally treated schizophrenia. Psychiatry Res. 37 Nov 30 2003;124(3):163-175. 38 12. Benoit A, Bodnar M, Malla AK, Joober R, Lepage M. The structural neural substrates of 39 persistent negative symptoms in first-episode of non-affective psychosis: a voxel-based 40 morphometry study. Front Psychiatry. 2012;3:42. 41 13. Baare WF, Hulshoff Pol HE, Hijman R, Mali WP, Viergever MA, Kahn RS. Volumetric analysis 42 of frontal lobe regions in schizophrenia: relation to cognitive function and symptomatology. Biol 43 44 45 46 Psychiatry. Jun 15 1999;45(12):1597-1605. 14. Bora E, Fornito A, Radua J, et al. Neuroanatomical abnormalities in schizophrenia: a multimodal voxelwise meta-analysis and meta-regression analysis. Schizophr Res. Apr 2011;127(1-3):46-57. 47 15. Lacerda AL, Hardan AY, Yorbik O, Vemulapalli M, Prasad KM, Keshavan MS. Morphology of 48 the orbitofrontal cortex in first-episode schizophrenia: relationship with negative 49 symptomatology. Prog Neuropsychopharmacol Biol Psychiatry. Mar 30 2007;31(2):510-516. 50 16. Volpe U, Mucci A, Quarantelli M, Galderisi S, Maj M. Dorsolateral prefrontal cortex volume in 51 patients with deficit or nondeficit schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry. 52 Jun 1 2012;37(2):264-269. 53 17. Nesvag R, Saetre P, Lawyer G, Jonsson EG, Agartz I. The relationship between symptom severity 54 and regional cortical and grey matter volumes in schizophrenia. Prog Neuropsychopharmacol 55 Biol Psychiatry. Apr 30 2009;33(3):482-490. 56 57 58 18. Kuperberg GR, Broome MR, McGuire PK, et al. Regionally localized thinning of the cerebral cortex in schizophrenia. Arch Gen Psychiatry. Sep 2003;60(9):878-888. 59 60 22 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 24 of 31 1 2 3 19. Oertel-Knöchel V, Knöchel C, Rotarska-Jagiela A, et al. Association between psychotic 4 5 6 symptoms and cortical thickness reduction across the schizophrenia spectrum. Cereb Cortex. Jan 2013;23(1):61-70. 7 20. Rimol LM, Nesvåg R, Hagler DJ, et al. Cortical volume, surface area, and thickness in 8 schizophrenia and bipolar disorder. Biol Psychiatry. Mar 2012;71(6):552-560. 9 21. Lyoo IK, Sung YH, Dager SR, et al. Regional cerebral cortical thinning in bipolar disorder. 10 Bipolar Disord. Feb 2006;8(1):65-74. 11 22. Rimol LM, Hartberg CB, Nesvåg R, et al. Cortical thickness and subcortical volumes in 12 schizophrenia and bipolar disorder. Biol Psychiatry. Jul 2010;68(1):41-50. 13 23. Hartberg CB, Sundet K, Rimol LM, et al. Brain cortical thickness and surface area correlates of 14 neurocognitive performance in patients with schizophrenia, bipolar disorder, and healthy adults. J 15 Int Neuropsychol Soc. Nov 2011;17(6):1080-1093. 16 24. MOUNTCASTLE VB. Modality and topographic properties of single neurons of cat's somatic 17 18 19 sensory cortex. J Neurophysiol. Jul 1957;20(4):408-434. 25. Rakic P. Specification of cerebral cortical areas. Science. Jul 1988;241(4862):170-176. 20 26. Rakic P. A small step for the cell, a giant leap for mankind: a hypothesis of neocortical expansion 21 during evolution. Trends Neurosci. Sep 1995;18(9):383-388. 22 27. Wierenga LM, Langen M, Oranje B, Durston S. Unique developmental trajectories of cortical 23 thickness and surface area. Neuroimage. Feb 15 2014;87:120-126. 24 28. Panizzon MS, Fennema-Notestine C, Eyler LT, et al. Distinct genetic influences on cortical 25 surface area and cortical thickness. Cereb Cortex. Nov 2009;19(11):2728-2735. 26 29. Xiao Y, Lui S, Deng W, et al. Altered Cortical Thickness Related to Clinical Severity But Not the 27 Untreated Disease Duration in Schizophrenia. Schizophr Bull. Dec 18 2013. 28 30. Padmanabhan J, Hooker CI, Keshavan MS. Brain Evolution, Language and Psychopathology in 29 Schizophrenia. In: Brambilla P, Marini A, eds. Explorations in Mental Health. New York: 30 31 32 Routledge; 2013. 31. Tamminga CA, Ivleva EI, Keshavan MS, et al. Clinical Phenotypes of Psychosis in the Bipolar 33 and Schizophrenia Network on Intermediate Phenotypes (B-SNIP). Am J Psychiatry. Jul 2013. 34 32. Ivleva EI, Bidesi AS, Keshavan MS, et al. Gray Matter Volume as an Intermediate Phenotype for 35 Psychosis: Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP). Am J 36 Psychiatry. Nov 2013;170(11):1285-1296. 37 33. First M, Spitzer R, Gibbon M, et al. The Structured Clinical Interview for DSM-III-R Personality 38 Disorders (SCID-II). Part II: Multi-site Test-retest Reliability Study. Journal of Personality 39 Disorders. 1995;9(2):92 - 104. 40 34. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale (PANSS) for 41 schizophrenia. Schizophr Bull. 1987;13(2):261-276. 42 35. Fischl B. FreeSurfer. Neuroimage. Aug 2012;62(2):774-781. 43 44 45 36. Hochberg Y. A sharper Bonferroni procedure for multiple tests of significance. Biometrika. 1988;75:800-802. 46 37. Yeh PH, Zhu H, Nicoletti MA, Hatch JP, Brambilla P, Soares JC. Structural equation modeling 47 and principal component analysis of gray matter volumes in major depressive and bipolar 48 disorders: differences in latent volumetric structure. Psychiatry Res. Dec 30 2010;184(3):177- 49 185. 50 38. Colibazzi T, Zhu H, Bansal R, Schultz RT, Wang Z, Peterson BS. Latent volumetric structure of 51 the human brain: Exploratory factor analysis and structural equation modeling of gray matter 52 volumes in healthy children and adults. Hum Brain Mapp. Nov 2008;29(11):1302-1312. 53 39. Fabrigar L, Wegener D, MacCallum R, Strahan E. Evaluating the Use of Exploratory Factor 54 Analysis in Psychological Research. Psychological Methods. 1999;4(3):272-299. 55 56 57 58 40. Cattell R. The scree test for number of factors. Multivariate Behavioral Research. 1966;1(2):245276. 41. Stevens J. Applied multivariate statistics for the social sciences. New York, NY: Routledge, 59 60 23 http://www.schizophreniabulletin.oupjournals.org Page 25 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 Taylor & Francis Group, LLC; 2009. 4 5 6 42. Costello A, Osborne J. Best practices in exploratory factor analysis: four recommendations for getting the most out of your analysis. Practical Assessment, Research and Evaluation. 7 2005;10(7):1-9. 8 43. Thurstone L. The vectors of mind. Chicago: University of Chicago Press; 1935. 9 44. Jardri R, Pouchet A, Pins D, Thomas P. Cortical activations during auditory verbal hallucinations 10 in schizophrenia: a coordinate-based meta-analysis. Am J Psychiatry. Jan 2011;168(1):73-81. 11 45. Palaniyappan L, Balain V, Radua J, Liddle PF. Structural correlates of auditory hallucinations in 12 schizophrenia: A meta-analysis. Schizophr Res. May 2012;137(1-3):169-173. 13 46. Shenton ME, Kikinis R, Jolesz FA, et al. Abnormalities of the left temporal lobe and thought 14 disorder in schizophrenia. A quantitative magnetic resonance imaging study. N Engl J Med. Aug 15 27 1992;327(9):604-612. 16 47. Andreasen NC, Nopoulos P, Magnotta V, Pierson R, Ziebell S, Ho BC. Progressive brain change 17 18 19 in schizophrenia: a prospective longitudinal study of first-episode schizophrenia. Biol Psychiatry. Oct 2011;70(7):672-679. 20 48. Iwashiro N, Suga M, Takano Y, et al. Localized gray matter volume reductions in the pars 21 triangularis of the inferior frontal gyrus in individuals at clinical high-risk for psychosis and first 22 episode for schizophrenia. Schizophr Res. May 2012;137(1-3):124-131. 23 49. Suga M, Yamasue H, Abe O, et al. Reduced gray matter volume of Brodmann's Area 45 is 24 associated with severe psychotic symptoms in patients with schizophrenia. Eur Arch Psychiatry 25 Clin Neurosci. Sep 2010;260(6):465-473. 26 50. Antonova E, Sharma T, Morris R, Kumari V. The relationship between brain structure and 27 neurocognition in schizophrenia: a selective review. Schizophr Res. Oct 2004;70(2-3):117-145. 28 51. Wible CG, Anderson J, Shenton ME, et al. Prefrontal cortex, negative symptoms, and 29 schizophrenia: an MRI study. Psychiatry Res. Nov 30 2001;108(2):65-78. 30 31 32 52. Berge D, Carmona S, Rovira M, Bulbena A, Salgado P, Vilarroya O. Gray matter volume deficits and correlation with insight and negative symptoms in first-psychotic-episode subjects. Acta 33 Psychiatr Scand. Jun 2011;123(6):431-439. 34 53. Chua SE, Wright IC, Poline JB, et al. Grey matter correlates of syndromes in schizophrenia. A 35 semi-automated analysis of structural magnetic resonance images. Br J Psychiatry. May 36 1997;170:406-410. 37 54. Lahti AC, Weiler MA, Holcomb HH, Tamminga CA, Carpenter WT, McMahon R. Correlations 38 between rCBF and symptoms in two independent cohorts of drug-free patients with 39 schizophrenia. Neuropsychopharmacology. Jan 2006;31(1):221-230. 40 55. Rais M, van Haren NE, Cahn W, et al. Cannabis use and progressive cortical thickness loss in 41 areas rich in CB1 receptors during the first five years of schizophrenia. Eur 42 Neuropsychopharmacol. Dec 2010;20(12):855-865. 43 44 45 56. Habets P, Marcelis M, Gronenschild E, Drukker M, van Os J, (G.R.O.U.P) GRaOoP. Reduced cortical thickness as an outcome of differential sensitivity to environmental risks in 46 schizophrenia. Biol Psychiatry. Mar 2011;69(5):487-494. 47 57. Harrison PJ. The neuropathology of schizophrenia. A critical review of the data and their 48 interpretation. Brain. Apr 1999;122 ( Pt 4):593-624. 49 58. Glantz LA, Lewis DA. Decreased dendritic spine density on prefrontal cortical pyramidal neurons 50 in schizophrenia. Arch Gen Psychiatry. Jan 2000;57(1):65-73. 51 59. Pasternak O, Westin CF, Bouix S, et al. Excessive extracellular volume reveals a 52 neurodegenerative pattern in schizophrenia onset. J Neurosci. Nov 28 2012;32(48):17365-17372. 53 60. Lynall ME, Bassett DS, Kerwin R, et al. Functional connectivity and brain networks in 54 schizophrenia. J Neurosci. Jul 14 2010;30(28):9477-9487. 55 56 57 58 61. Lawrie SM, Buechel C, Whalley HC, Frith CD, Friston KJ, Johnstone EC. Reduced frontotemporal functional connectivity in schizophrenia associated with auditory hallucinations. Biol Psychiatry. Jun 15 2002;51(12):1008-1011. 59 60 24 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 26 of 31 1 2 3 62. Ren W, Lui S, Deng W, et al. Anatomical and functional brain abnormalities in drug-naive first- 4 5 6 episode schizophrenia. Am J Psychiatry. Nov 2013;170(11):1308-1316. 63. Ho BC, Andreasen NC, Ziebell S, Pierson R, Magnotta V. Long-term antipsychotic treatment and 7 brain volumes: a longitudinal study of first-episode schizophrenia. Arch Gen Psychiatry. Feb 8 2011;68(2):128-137. 9 64. Hill SK, Reilly JL, Keefe RS, et al. Neuropsychological Impairments in Schizophrenia and 10 Psychotic Bipolar Disorder: Findings from the Bipolar and Schizophrenia Network on 11 Intermediate Phenotypes (B-SNIP) Study. Am J Psychiatry. Jun 2013. 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 25 http://www.schizophreniabulletin.oupjournals.org Page 27 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Figure 1. Significant correlations between lobe measures and positive symptom severity for all patients and diagnostic groups; plotted values are residuals after adjustment for co-variates. 27 103x56mm (300 x 300 DPI) 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 28 of 31 1 2 3 4 5 6 7 Supplemental Table 1: Component sub-regions of lobes 8 Lobes (Right and Left) Component sub-regions of lobe (identical for right and left lobes) 9 10 Frontal GMV GMV of caudal middle frontal, rostral middle frontal, medial orbitofrontal, lateral orbitofrontal, pars opercularis, pars 11 orbitalis, pars triangularis, superior frontal, precentral, and paracentral regions 12 13 Temporal GMV GMV of banks superior temporal sulcus, inferior temporal, middle temporal, superior temporal, fusiform, entorhinal, 14 15 parahippocampal, transverse temporal, temporal pole, and insula regions 16 Limbic GMV 17 GMV of hippocampus, amygdala, caudal anterior cingulate, isthmus cingulate, rostral anterior cingulate, posterior 18 cingulate 19 20 Striatal GMV GMV of thalamus, caudate, putamen, pallidum, and accumbens 21 Parietal GMV 22 GMV of inferior parietal, superior parietal, supramarginal, precuneus, and postcentral regions 23 Occipital GMV GMV of cuneus, lingual, pericalcarine, and lateral occipital regions 24 25 Frontal Surface Area Identical to Frontal GMV, but composed of surface areas 26 Temporal Surface Area Identical to Temporal GMV, but composed of surface areas 27 28 Limbic Surface Area Surface areas of caudal anterior cingulate, isthmus cingulate, rostral anterior cingulate, and posterior cingulate regions 29 30 Parietal Surface Area Identical to Parietal GMV, but composed of surface areas 31 Occipital Surface Area Identical to Occipital GMV, but composed of surface areas 32 33 Frontal Cortical Thickness Identical to Frontal GMV, but composed of cortical thicknesses 34 35 Temporal Cortical Thickness Identical to Temporal GMV, but composed of cortical thicknesses 36 Limbic Cortical Thickness Cortical thicknesses of caudal anterior cingulate, isthmus cingulate, rostral anterior cingulate, posterior cingulate 37 38 Parietal Cortical Thickness Identical to Parietal GMV, but composed of cortical thicknesses 39 40 Occipital Cortical Thickness Identical to Occipital GMV, but composed of cortical thicknesses 41 Notes: 1. Component sub-regions were identical for lobes within both hemisphere. 2. Striatal regions are sub-cortical; thus, cortical thickness and 42 surface area measurements were unavailable for these regions. 43 44 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 Page 29 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 Supplemental Table 2. Study Site Differences 4 5 Site 1 (CT) 2 (GP) 3 (GT) 4 (JS) 5 (MB) 6 (MK) 6 7 Location Texas Connecticut Maryland Illinois Boston Detroit 8 N 85 46 114 120 24 66 9 10 11 Mean PANSS Positive (sd) 19.0 (4.3) 15.7 (4.6) 14.6 (5.9) 16.2 (5.4) 13.0 (5.5) 17.0 (5.4) 12 13 Mean PANSS Negative (sd) 15.1 (3.9) 14.6 (5.7) 14 15 16 Mean PANSS General (sd) 36.7 (7.7) 32.5 (8.5) 14.1 (5.1) 26.0 (6.7) 16.1 (6.1) 33.9 (8.5) 12.3 (6.3) 28.3 (10.2) 14.9 (4.8) 35.4 (8.7) 17 18 19 Mean PANSS Total (sd) 70.8 (14.4) 62.8 (15.8) 54.7 (14.2) 66.2 (16.5) 53.6 (19.7) 67.3 (17.2) 20 21 22 Within-site SP: 21 (25 %) SP: 16 (35 %) SP: 64 (56 %) SP: 37 (31 %) SP: 8 (33 %) SP: 35 (53 %) 23 Diagnostic SZA: 38 (45 %) SZA: 20 (43 %) SZA: 23 (20 %) SZA: 27 (23 %) SZA: 3 (13 %) SZA: 6 (9 %) 24 Breakdown BP: 26 (31 %) BP: 10 (22 %) BP: 27 (24 %) BP: 56 (47 %) BP: 13 (54 %) BP: 25 (38 %) 25 (N, %) 26 27 Notes: 1. SP = schizophrenia, SZA = schizoaffective disorder, BP = bipolar I with psychotic features. Percentages may not add up to 100 due to 28 rounding. 2. Subscale scores were significantly different across sites (p < 0.05 by Kruskal-Wallis test). 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 http://www.schizophreniabulletin.oupjournals.org 47 48 Schizophrenia Bulletin. For Peer Review Only Page 30 of 31 1 2 3 Supplemental Table 3. Factor Loadings for Principal Factors Extraction and Oblique Rotation of Four Factors: 4 5 Positive Subscale 6 F1: Temporal F2: Frontal F3: Fronto- F4: Precuneus 7 CT GMV parietal CT 8 SS Loadings 8.00 6.70 6.67 3.96 9 Proportion of Variance 0.14 0.12 0.12 0.07 10 L fusiform CT 0.82 11 R middle temporal CT 12 R superior temporal CT 13 14 R fusiform CT 15 L inferior temporal CT 16 R inferior temporal CT 0.81 0.79 0.73 0.71 0.69 17 R insula CT 0.58 18 L fusiform GMV 0.55 19 L insula CT 0.55 20 R temporal pole CT 21 R banks of superior temporal sulcus CT 22 23 R fusiform GMV 24 R middle temporal GMV 25 R lateral orbitofrontal GMV 0.55 0.50 0.47 0.46 0.75 26 R superior frontal GMV 0.70 27 L lateral orbitofrontal GMV 0.66 28 L superior frontal GMV 0.61 29 R superior frontal CSA 30 R rostral middle frontal GMV 31 32 L superior frontal CSA 33 L pars orbitalis GMV 34 R pars orbitalis GMV 0.57 0.57 0.56 0.52 0.50 35 L rostral middle frontal GMV 0.49 36 L superior parietal CT 0.83 37 R inferior parietal CT 0.81 38 L superior frontal CT 39 40 41 R lateral occipital CT R supramarginal CT 42 R superior frontal CT 43 L lateral occipital CT 0.81 0.70 0.67 0.59 0.52 44 L supramarginal CT 0.51 45 L superior parietal GMV 0.51 46 Right lingual CT 47 R precuneus GMV 48 49 50 L precuneus GMV R precuneus CSA 51 R fusiform CSA 0.46 0.67 0.61 0.60 0.50 52 There were no cross-loadings. The following regions did not load above 0.45 on any factor: L caudal middle 53 frontal GMV, L isthmus cingulate GMV, L superior temporal GMV, L hippocampus GMV, R hippocampus 54 GMV, R pars triangularis GMV, R supramarginal GMV, R banks of superior temporal sulcus GMV, R superior 55 temporal GMV, L lateral occipital GMV, L superior temporal CSA, L pars orbitalis CSA, L transverse temporal 56 CSA, L caudal middle frontal CSA, L pars triangularis CSA, L posterior cingulate CT, L parahippocampal CT, R 57 posterior cingulate CT, R pars triangularis CT, L medial orbitofrontal CT, L pars orbitalis CT. 58 59 60 http://www.schizophreniabulletin.oupjournals.org Page 31 of 31 Schizophrenia Bulletin. For Peer Review Only 1 2 3 4 5 6 7 Supplemental Table 4. Factor Loadings for Principal Factors Extraction of One Factor: Negative Subscale 8 Frontal GMV-SA factor 9 SS Loading 3.54 10 Proportion of Variance 0.25 11 R superior frontal GMV 0.72 12 R superior frontal CSA 13 R precentral GMV 14 15 R precentral CSA 16 R lateral orbitofrontal CSA 17 R pars orbitalis CSA 18 R pars orbitalis GMV 0.71 0.66 0.62 0.60 0.51 0.48 19 R paracentral CSA 0.46 20 L superior parietal GMV 0.45 21 L posterior cingulate GMV 0.44 22 Regions not loading above 0.45 on this factor: R fusiform CSA, 23 24 R fusiform GMV, R posterior cingulate CT, R fusiform CT 25 26 27 28 29 Supplemental Figure 1: Correlations between regional brain structure and positive and negative symptom 30 subscales. 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://www.schizophreniabulletin.oupjournals.org Schizophrenia Bulletin. For Peer Review Only Page 32 of 31 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Supplemental Figure 1. Correlations between regional brain structure and positive and negative symptom 40 subscales. 41 110x105mm (300 x 300 DPI) 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://www.schizophreniabulletin.oupjournals.org