Research Paper Multidimensional predictors of negative symptoms in antipsychotic-naive first-episode psychosis Alba Toll, MD; Laura Blanco-Hinojo, PhD; Daniel Bergé, PhD; Xavier Duran, PhD; Irene Canosa, MHN; Teresa Legido, PhD; Federico Marmol, PhD; Víctor Pérez-Solà, PhD; Emilio Fernández-Egea, PhD; Anna Mané, PhD Background: Despite a large body of schizophrenia research, we still have no reliable predictors to guide treatment from illness onset. The present study aimed to identify baseline clinical or neurobiological factors — including peripheral brain-derived neurotrophic factor (BDNF) levels and amygdala or hippocampal relative volumes — that could predict negative symptomatology and persistent negative symptoms in first-episode psychosis after 1 year of follow-up. Methods: We recruited 50 drug-naive patients with first-episode psycho- sis and 50 age- and sex-matched healthy controls to study brain volumes. We performed univariate and multiple and logistic regression analyses to determine the association between baseline clinical and neurobiological variables, score on the PANSS negative subscale and persistent negative symptoms after 1 year of follow-up. Results: Low baseline serum BDNF levels (p = 0.011), decreased left amygdala relative volume (p = 0.001) and more severe negative symptomatology (p = 0.021) predicted the severity of negative symp- toms at 1 year, as measured by the PANSS negative subscale. Low baseline serum BDNF levels (p = 0.012) and decreased left amyg- dala relative volume (p = 0.010) predicted persistent negative symptoms at 1 year. Limitations: We were unable to assess negative symptoms and their dimensions with next-generation scales, which were not available when the study was initiated. Conclusion: This study shows that a set of variables at baseline, including low BDNF levels, smaller left amygdala relative volume and score on the PANSS negative subscale are significant predictors of outcomes in first-episode psychosis. These findings might offer an initial step for tailoring treatments in first-episode psychosis. Introduction Therefore, it would be highly valuable to identify — at or near disease onset — those patients who are most likely to Schizophrenia is a chronic, lifelong, disabling disorder that develop severe negative symptoms; this would facilitate the develops via complex, heterogeneous, multifactorial (genetic, early implementation of personalized interventions.11 developmental and environmental) pathways.1 It is 1 of the There are different approaches to measuring negative top 15 leading causes of disability worldwide,2 impairing crit­ symptoms. The most common approach is to rate all nega­ ical areas of everyday functioning.3 Although early interven­ tive symptoms without considering the possible cause (e.g., tions may improve treatment outcomes,4 there is a notable positive symptoms, adverse effects, social deprivation or lack of reliable predictors to help guide an objective, data- others); PANSS is one of the most widely used scales for this driven personalized approach to treatment after illness onset.5 purpose. Another approach is to restrict negative symptoms Given the complexity of schizophrenia, it is essential to to those considered primary or deficit symptoms12,13 after identify unambiguous and distinctive predictors of out­ ruling out secondary causes. An alternative approach to comes,6 such as negative symptoms.7 Negative symptoms are studying negative symptoms is to use persistent negative among the most important determinants of functioning8,9 in symptoms; these may include secondary negative symp­ schizophrenia. They can be present at any stage of the disor­ toms, but only those that have failed to respond to the usual der, even in the early stages, with prevalence in up to 50% of treatments and are evident for more than 6 months. Persis­ patients.10 Longitudinal studies in early stages have found tent negative symptoms are also highly relevant because that when these symptoms are present at baseline, they tend they lead to functional impairment and represent an unmet to be persistent, although severity may vary over time. therapeutic need.14 Correspondence to: L. Blanco, MRI Department, Parc de Salut Mar, Passeig Marítim 25-29 08003, Barcelona, Spain; laura.blanco02@gmail.com Submitted Aug. 7, 2021; Revised Oct. 15, 2021; Accepted Nov. 1, 2021 Cite as: J Psychiatry Neurosci 2022 January 19;47(1). doi: 10.1503/jpn.210138 © 2022 CMA Impact Inc. or its licensors J Psychiatry Neurosci 2022;47(1) E21 Toll et al. Several clinical factors have been linked to negative symp­ The primary objective of the present study was to deter­ toms in general. For example, more severe negative symptoms mine whether baseline clinical and neurobiological factors — have been observed in males,15,16 in patients who developed including duration of untreated psychosis, clinical symp­ schizophrenia at a younger age17,18 and in those with a longer toms, cannabis use, peripheral BDNF levels and amygdala or duration of untreated psychosis.19,20 However, the associa­ hippocampal relative volumes — could predict negative tion between negative symptoms and these clinical factors symptomatology and persistent negative symptoms at 1 year has not been consistent in the studies to date.15,19,21 The influ­ of follow-up in patients with first-episode psychosis. A ence of cannabis use on negative symptomatology is poorly ­second aim was to measure and compare hippocampal and understood, but a recent meta-analysis showed that recently amygdalar volumes in patients with first-episode psychosis abstinent patients showed negative symptoms that were and healthy controls to determine significant differences. less severe than nonusers.22 In addition to these clinical We hypothesized that this combination of clinical and markers, a considerable amount of research has been con­ neuro­biological parameters would predict negative symp­ ducted to identify neurobiological factors that could predict toms and persistent negative symptoms at 1-year follow-up. outcomes and help to select optimal treatment strategies.5 We also hypothesized that patients with first-episode psy­ Morphological and functional brain alterations have been chosis would have smaller hippocampal and amygdalar rela­ widely reported in schizophrenia.23 Structural and func­ tive volumes than healthy controls. tional changes in the limbic system and medial temporal lobe are central to the symptomatology associated with psy­ Methods chosis, 24 and smaller relative hippocampal volumes are commonly observed in patients with first-episode psychosis Study population and schizophrenia.25 Furthermore, these hippocampal ab­ normalities are associated with the emergence of negative Fifty consecutive, drug-naive patients with first-episode psycho­ symptoms and persistent negative symptoms in patients sis who were treated between April 2013 and July 2017 at the with schizophrenia and first-episode psychosis.10,26 Smaller ­Estudi i Tractament de Primers Episodis Psicotics (ETEP) Program amygdala volumes have also been reported in patients with at Hospital del Mar (Barcelona, Spain) were included in this schizophrenia and first-episode psychosis compared to con­ study. The ETEP program is a specialized early intervention ser­ trols.27,28 In addition, volume alterations in the amygdala vice for people aged 18 to 35 who experience first-episode psy­ have been associated with both negative symptoms and chosis. It provides multimodal intervention, including extensive persistent negative symptoms in schizophrenia and in first- assessment and intensive medical and psychosocial treatment episode psychosis during follow-up.10,26 (for more details, see Bergé and colleagues33). Brain-derived neurotrophic factor (BDNF) has been im­ For the present study, inclusion criteria for patients were as plicated in the pathophysiology of schizophrenia, and it has follows: age 18 to 35 years; fulfillment of DSM-IV-TR criteria for been shown to play a crucial role in neuronal development, brief psychotic disorder, schizophreniform disorder, schizo­ transmission, regulation and synaptic plasticity.29 Data from phrenia with less than 1 year of symptoms or unspecified psy­ systematic reviews and meta-analyses show that patients chosis; no previous history of severe neurologic medical condi­ with schizophrenia 30 and first-episode psychosis 31 have tions or severe traumatic brain injury; presumed IQ above 80 lower peripheral BDNF levels than healthy controls. Cross- based on clinical records (evidence from previous IQ assess­ sectional studies have found an association between peri­ ments or suggested by the patient’s educational or employment pheral BDNF levels and the clinical features of schizophre­ level); and no substance abuse or dependence disorders, except nia, such as the severity of negative symptoms.32 These data for cannabis or nicotine use. All patients were antipsychotic- suggest that peripheral BDNF levels could be a neurobio­ and antidepressant-naive at inclusion. Treatment with benzo­ logical predictor of psychosis and psychosis phenotypes. diazepines was allowed. We excluded affective psychosis to However, more longitudinal studies are needed in first- homogenize the sample, because negative symptoms in affec­ episode psychosis to better determine the association between tive psychosis can differ in terms of whether they are primary peripheral BDNF levels and different outcome parameters, or secondary and persistent or transient34 (see Appendix 1, ­including negative symptoms. ­Figure S1 and Table S1, available at www.jpn.ca/lookup/ In summary, there is a clear need to identify predictors of doi/10.1503/jpn.210138/tab-related-content, for detailed infor­ outcomes in schizophrenia and guide individualized treat­ mation on patients excluded from the neuroimaging study). ment strategies from illness onset. Negative symptoms are We used 50 age- and sex-matched healthy volunteers as a distinctive features of schizophrenia and critical determi­ control group for anatomic measurements. All healthy controls nants of outcomes. Although various sociodemographic, clin­ underwent a complete medical interview. Those with relevant ical and neurobiological factors have been associated with medical or neurologic disorders, substance abuse disorders or negative symptoms, few studies have evaluated the predic­ psychiatric illnesses were excluded from the study. None of the tive value of these factors longitudinally starting from dis­ healthy controls was undergoing medical treatment at study ease onset, and even fewer have considered the sources of ­inclusion. Given the characteristics of the study design, we secondary negative symptoms or persistent negative symp­ used no statistical methods to predetermine sample sizes. toms. Finally, to our knowledge, no studies have assessed all The local ethics committee approved this study, and all 3 factors (clinical, neuroimaging and biochemical) together. participants provided written informed consent. E22 J Psychiatry Neurosci 2022;47(1) Predictors of negative symptoms in first-episode psychosis Clinical assessment and demographic data acquisition of high-resolution anatomic 3-dimensional im­ ages based on a T1-weighted fast spoiled gradient inversion All patients underwent a comprehensive assessment at recovery prepared sequence with the following param­ baseline and 1-year follow-up, performed by 2 experienced eters: repetition time 8.2 ms; echo time 3.8 ms; flip angle 8°; psychiatrists (A.M., D.B.). The assessment included field of view 24 cm; matrix 256 × 256 pixels; in-plane reso­ ­sociodemographic variables; the Structured Clinical Inter­ lution 0.94 × 0.94 mm2; and slice thickness 1 mm. view for DSM-IV-TR Axis I disorders for diagnosis; an as­ Anatomic images were visually inspected before analy­ sessment of substance use, including tobacco (cigarettes per sis by a trained operator to detect any motion effect. No day) and cannabis use (joints per week, after dichotomiza­ gross brain abnormalities were identified. The quality of tion between user/nonuser); the Positive and Negative the raw T 1-weighted MRI scans was quantitatively as­ Syndrome Scale (PANSS) for symptoms related to psycho­ sessed using the automated weighted image quality rating sis; 35 the Global Assessment of Functioning (GAF) 36 for (a measure of general image quality that combines the par­ functionality; and the Calgary Depression Scale for Schizo­ ameters of noise, inhomogeneities and spatial resolution) phrenia (CDSS) for depressive symptoms.37 We collected in the CAT12 toolbox (http://dbm.neuro.uni-jena.de/ patients’ antipsychotic dose at 1-year follow-up from their cat/); a score of more than 60% is considered sufficient medical records. We converted doses of antipsychotic quality for inclusion in subsequent analyses. All scans ­medications to chlorpromazine equivalents (mg/d) to fa­ passed this threshold. cili­tate comparisons.38 We performed volumetric segmentation using the fully To minimize the effect of secondary negative symptom automated and validated segmentation software package sources, we excluded all patients who scored 6 or higher on FreeSurfer v6.0 (http://surfer.nmr.mgh.harvard.edu) with the CDSS at baseline or at follow-up.39 We also included the default ”recon-all” stream, as described in previous anti­psychotic treatments at follow-up in the analyses. studies.41–44 Briefly, important preprocessing steps included We also assessed for the presence of persistent negative removal of non-brain tissue, intensity normalization, auto­ symptoms, defined as 1 or more negative symptoms of mod­ mated Talairach transformation and segmentation of the erate or higher severity (PANSS negative subscale > 3), not subcortical white matter and deep grey matter volumetric confounded by depression (CDSS < 6) or parkinsonism at structures.45 Segmentations of the hippocampus and amyg­ baseline and after 1 year of treatment.39 dala (left and right hemisphere) were visually inspected for accuracy by overlaying the segmentation label of each struc­ Collection of blood samples and determination of serum ture on the individual T1-weighted brain scan. Specifically, BDNF levels in patients at baseline we used the imaging quality control protocol for subcortical segmentations developed by the ENIGMA consortium We obtained baseline fasting blood samples upon patients’ (http://enigma.ini.usc.edu/protocols/imaging-protocols/) arrival at the health centre before administering any medi­ to generate a standard set of images from every partici­ cation (except for benzodiazepines). All blood samples pant’s brain scan, displaying boundaries of segmented were obtained in the morning (between 8 am and 12 pm) to structures on a series of slices in the axial, coronal and sagit­ avoid circadian fluctuations in BDNF levels, which have tal planes (Figure 1) for visual inspection of segmentation been reported to occur in men but not in women.40 Blood inaccuracies. For each participant, we extracted estimated samples were collected in glass K3–EDTA blood-drawing hippocampal and amygdalar volumes (left and right hemi­ tubes for whole blood. Serum was isolated by centrifuga­ sphere), as well as estimated total intracranial volume, from tion at 300 × g for 15 minutes, and then removed and the aseg.stats output files in FreeSurfer (https://surfer.nmr. stored at −80°C until analysis. mgh.harvard.edu/fswiki/asegstats2table). We used group- BDNF levels were measured using a sandwich enzyme- level means, standard deviations and histogram plots to linked immunoabsorbent assay kit according to the manu­ identify statistical outliers (i.e., if their volume was meas­ facturer’s instructions (ChemiKine; Chemicon). Absorben­ ured to be greater than 2.698 standard deviations [SDs] from cies were determined using the Wallac Victor 2 microplate the global mean) and to confirm the normal distribution of reader, with the wavelength set at 450 nm. BDNF concentra­ volumetric data. One control participant was identified as tions were detected according to the standard curve, which an outlier for the estimate of right amygdala volume. In this was constructed from duplicate samples containing appro­ case, we visually reinspected the images using Freeview priate concentrations. All samples were analyzed in dupli­ (v2.0) in FreeSurfer to confirm segmentation accuracy, and cate. The calculated overall intra- and interassay variation the data were retained in the analysis. coefficients were 3.7% and 8.5%, respectively. The detection We calculated the relative volume ratio for the ana­ limit of the BDNF assay was 15 pg/mL. tomic regions of interest as the volume in native space (mL) divided by the total intracranial volume (L). We per­ Image acquisition and processing at baseline formed MRI assessments when patients were able to stand still (i.e., they did not have high psychomotor agita­ Brain imaging was performed with a Philips Achieva 3.0 T tion or disorganized behaviour and could tolerate the MRI scanner (Philips Healthcare) equipped with an 8-channel MRI procedure). The scan was taken within 2 weeks of phased-array head coil. The imaging protocol involved the the initial assessment. J Psychiatry Neurosci 2022;47(1) E23 Toll et al. Figure 1: Illustrations of the amygdala (top row) and the hippocampus (bottom row) as segmented by FreeSurfer in a control participant. Segmentations are overlaid on the participant’s anatomic MRI. Statistical analysis BDNF levels. We also accounted for antipsychotic treat­ ment at follow-up as a possible confounding factor. We assessed data distribution normality using the Kolmogorov–​ Next, we developed a linear regression model using the Smirnov test. Only age, duration of untreated psychosis PANSS negative subscale at 1-year follow-up as the and tobacco use were non-Gaussian variables. We per­ depend­ent variable. Independent variables were all baseline formed univariate analyses to evaluate differences be­ variables (age, sex, cannabis use, tobacco use, duration of tween the patients who completed the entire follow-up untreated psychosis, PANSS negative subscale, PANSS pos­ and those who did not. We also performed univariate itive subscale, CDSS, GAF, BDNF, left and right amygdala analyses to assess between-group differences (patients relative volume, left and right hippocampus relative vol­ and healthy controls), and Bonferroni correction on brain- ume, and benzodiazepine treatment) and chlorpromazine volume comparisons. equivalents at 1 year follow-up. We also performed an ex­ To determine predictors of negative symptoms at 1-year ploratory analysis to assess predictors of the PANSS nega­ follow-up, we first conducted univariate analyses that in­ tive subscale items using the same methodology. cluded sociodemographic and baseline clinical and neuro­ To determine predictors of persistent negative symptoms biological variables that were potentially associated with at 1-year follow-up, we used the same methodology, except scores on the PANSS negative subscale at 1-year follow- that we used a logistic regression model with persistent up: sex, age, duration of untreated psychosis, baseline negative symptoms (yes/no) as the dependent variable. CDSS score, PANSS positive subscale score, benzodiaz­ All statistical analyses were performed in SPSS Statistics epine treatment, left and right amygdala relative volumes, for Windows, version 20 (IBM Corp.); p values ≤ 0.05 were left and right hippocampal relative volumes and serum considered statistically significant. E24 J Psychiatry Neurosci 2022;47(1) Predictors of negative symptoms in first-episode psychosis Results campus 4.17 ± 0.42 mL, right hippocampus 4.28 ± 0.45 mL, left amygdala 1.67 ± 0.20 mL, right amygdala 1.89 ± 0.25 mL) Characteristics of patients and healthy controls were comparable to those previously reported in similar clin­ ical populations using the same FreeSurfer method.41–44,46,47 A total of 50 patients were initially included in the study. Univariate analysis (Table 1) showed that patients had None of the patients scored 6 or higher on the CDSS at base­ significantly smaller relative volumes in the left and right line or at follow-up. More than half (56%) of the patients amygdala (t = −4.26, p = 0.001; t = −4.32, p = 0.001) com­ were male. The median age was 26 years (interquartile pared to healthy controls. Initially, left and right hippocam­ range 24–30.25 years). Most of the patients were being pal relative volumes were significantly lower in patients treated with benzodiazepines at inclusion (86%). compared to healthy controls (t = −3.63, p = 0.047; t = −2.13, Eleven patients were lost to follow-up. However, we found p = 0.048; Appendix 1, Figures S2 and S3), but after no statistically significant differences between patients who ­Bonferroni correction these findings were no longer signifi­ completed the study and those who did not (Appendix 1, cant (p = 0.047 and 0.048 > 0.05/4 = 0.0125). ­Table S2). At 1-year follow-up, mean (± SD) PANSS scores in patients with first-episode psychosis were as follows: positive Predictors of negative symptoms measured with PANSS at subscale 12.69 ± 7.03; negative subscale 17.18 ± 5.89; general 1-year follow-up pathology 31.67 ± 9.26; and total score 62.51 ± 19.39. Of the patient group, 41% had persistent negative symptoms, and In the univariate analysis (Table 2), PANSS negative sub­ the mean (± SD) GAF score was 61.18 ± 17.68. scale scores at 1 year were negatively correlated with base­ A total of 50 healthy controls were included in this study. line serum BDNF levels (r = −0.4; p = 0.012) and with left More than half (56%) were male, and the median age was amygdala relative volume (r = −0.51; p = 0.001) in patients 26 years (interquartile range 23–30 years). Other clinical with first-episode psychosis. Conversely, PANSS negative characteristics of the sample are shown in Table 1. subscale scores at 1 year were positively correlated with Hippocampal and amygdalar absolute volumes (group PANSS negative subscale scores at baseline (r = 0.41; p = 0.009). means ± SD; patients: left hippocampus 3.88 ± 0.38 mL, right In the multivariate analysis, the model that best predicted hippocampus 4.09 ± 0.45 mL, left amygdala 1.48 ± 0.24 mL, PANSS negative subscale scores at 1-year follow-up in­ right amygdala 1.68 ± 0.22 mL; healthy controls: left hippo­ cluded the following baseline variables: serum BDNF, Table 1: Sociodemographic, volumetric and clinical characteristics at baseline Patients Healthy volunteers Characteristic n = 50* n = 50* Statistical test p value Effect size† Age, yr 26 (24–30.25) 26 (23–30) t = 992 0.07 d = −0.36 Female 22 (44%) 22 (44%) χ2 = 0.16 0.42 V = 0.08 Relative volume, mL Left hippocampus 2.64 ± 0.27 2.77 ± 0.35 t = −3.63 0.047 d = −0.42 Right hippocampus 2.78 ± 0.31 2.84 ± 0.36 t = −2.13 0.048 d = −0.18 Left amygdala 1.01 ± 0.14 1.11 ± 0.14 t = −4.26 0.001 d = −0.71 Right amygdala 1.14 ± 0.15 1.24 ± 0.14 t = −4.32 0.001 d = −0.69 Duration of untreated psychosis, d 31 (8–115) — — — — Cannabis user 29 (58%) — — — — Tobacco use, cigarettes/d 4.5 (0–14) — — — — Diagnosis Schizophreniform disorder 18 (36%) — — — — Psychosis not otherwise specified 28 (56%) — — — — Brief psychotic disorder 4 (8%) — — — — Benzodiazepine treatment 43 (86%) — — — — BDNF, pg/mL 45.27 ± 27.14 — — — — PANSS positive subscale score 24.88 ± 6.74 — — — — PANSS negative subscale score 16.86 ± 6.65 — — — — PANSS general pathology score 43.68 ± 8.27 — — — — PANSS total score 85.24 ± 15.76 — — — — CDSS score 1.22 ± 2.02 — — — — GAF score 29.7 ± 8.89 — — — — BDNF = brain-derived neurotrophic factor, CDSS = Calgary Depression Scale for Schizophrenia; GAF = Global Assessment of Functioning; PANSS = Positive and Negative Syndrome Scale. *Values are given as mean ± standard deviation, n (%) or median (interquartile range). †Cohen d or Cramer V. J Psychiatry Neurosci 2022;47(1) E25 Toll et al. Table 2: Association of sociodemographic, volumetric and clinical characterics with PANSS negative score and persistent negative symptoms at 1-year follow-up: univariate analysis PANSS negative Persistent negative Characteristic score at 1 year p value Effect size* symptoms at 1 year p value Effect size* Age r = –0.07 0.69 — t = 0.76 0.45 d = 0.08 Sex t = 1.95 0.06 d = 0.64 χ2 = 0.24 0.81 V = 0.24 Male 18.58 ± 5.68 — — 68.8% — — Female 14.93 ± 5.71 — — 31.2% — — Relative volume Left hippocampus r = –0.25 0.12 — t = 1.01 0.32 d = 0.37 Right hippocampus r = –0.21 0.19 — t = 1.21 0.23 d = 0.44 Left amygdala r = –0.51 0.001 — t = 2.47 0.018 d = 0.91 Right amygdala r = –0.31 0.05 — t = 1.83 0.08 d = 0.57 Duration of untreated psychosis r = 0.29 0.07 — t = –1.53 0.14 d = –0.48 Cannabis use t = –1.2 0.32 d = −0.32 χ2 = –0.51 0.62 V = 0.21 Non-user 16.21 ± 5.5 — — 43.8% — — User 18.1 ± 6.25 — — 56.2% — — Tobacco use r = –0.27 0.10 — t = 1.49 0.14 d = 0.46 Chlorpromazine equivalent at 1 year r = 0.09 0.59 — t = 0.34 0.74 d = 0.11 Benzodiazepine treatment t = –1.61 0.12 d = −0.64 χ = 1.93 2 0.17 V = 0.41 BDNF, baseline r = –0.4 0.012 — t = 2.93 0.006 d = 1.01 PANSS negative subscale score, baseline r = 0.41 0.009 — — PANSS positive subscale score, baseline r = –0.27 0.09 — t = 1.23 0.23 d = 0.39 CDSS score, baseline r = –0.16 0.34 — t = 0.45 0.66 d = 0.14 GAF score, baseline r = –0.26 0.12 — t = 2.12 0.07 d = 0.72 BDNF = brain-derived neurotrophic factor; CDSS = Calgary Depression Scale for Schizophrenia; GAF = Global Assessment of Functioning; PANSS = Positive and Negative Syndrome Scale. *Cohen d or Cramer V. PANSS negative subscale scores and left amygdala relative Table 3: PANSS negative score at 1-year follow-up: linear regression volume, with an R2 of 0.478. This finding confirmed that model lower baseline serum BDNF level (95% confidence interval Predictor R2 B (95% CI) p value [CI] −0.12 to −0.02; p = 0.011), smaller left amygdala relative Model 1 0.253 < 0.001 volume (95% CI −33.14 to −9.03; p = 0.001), and higher base­ Relative left amygdala –24.14 (–37.98 to 0.001 line PANSS negative subscale score (95% CI 0.04 to 0.48; p = volume, mL –10.31) 0.021) were associated with higher PANSS negative sub­ Model 2 0.391 < 0.001 scale scores at 1-year follow-up in patients with first-episode Relative left amygdala –23.2 (–35.89 to 0.001 psychosis (Table 3 and Figure 2). When we conducted the volume, mL –10.51) different models adding 1 variable at a time, BDNF ex­ BDNF at baseline –0.08 (–0.13 to 0.007 plained 15.8% of the variance, baseline PANSS negative –0.02) subscale score explained 13.2% and left amygdala relative Model 3 0.478 < 0.001 volume explained 18.8%. We also performed a robust re­ Relative left amygdala –21.09 (–33.14 to 0.001 gression to confirm that these associations held (with volume, mL –9.03) PANSS negative subscale score at 1-year follow-up as a de­ BDNF at baseline –0.07 (–0.12 to 0.011 –0.02) pendent variable and BDNF level, left amygdala relative PANSS negative subscale 0.26 (0.04 to 0.48) 0.021 volume and baseline PANSS negative subscale score as in­ score at baseline dependent variables; Appendix 1, Table S4). BDNF = brain-derived neurotrophic factor; CI = confidence interval; PANSS = Positive We performed an exploratory univariate analysis to assess and Negative Syndrome Scale. predictors of PANSS negative subscale subitems at 1-year follow-up (Appendix 1, Table S3). In a multivariate analysis, baseline PANSS negative subscale score was a significant Predictors of persistent negative symptoms at 1-year follow-up predictor of those subitems. Baseline CDSS score was a sig­ nificant predictor of N2 (emotional withdrawal; 95% CI = In the univariate analysis (Table 2), persistent negative 0.02 to 0.26, p = 0.025) and sex was a significant predictor of symptoms at 1 year were significantly associated with N7 (stereotyped thinking; 95% CI = −1.87 to −0.56, p = 0.001). baseline serum BDNF levels (t = 2.93; p = 0.006) and left E26 J Psychiatry Neurosci 2022;47(1) Predictors of negative symptoms in first-episode psychosis Linear R2 = 0.476 Linear R2 = 0.220 30 15 10 PANSS-N at 1-year follow-up PANSS-N at 1-year follow-up 25 5 20 0 15 −5 10 −10 5 −15 -2 −1 0 1 2 3 −40.00 −20.00 0.00 20.00 40.00 60.00 80.00 Multiple regression (predicted z values) BDNF at baseline, pg/mL Linear R2 = 0.262 Linear R2 = 0.185 15 10 10 PANSS-N at 1-year follow-up PANSS-N at 1-year follow-up 5 5 0 0 −5 −5 −10 −15 −10 −400.0 −200.0 0.0 200.0 400.0 −10 0 10 20 Left amygdala relative volume at baseline, mL PANSS-N at baseline Figure 2: Correlations between PANSS-N scores at 1-year follow-up and BDNF levels, left amygdala relative volume and PANSS-N scores at baseline. BDNF = brain-derived neurotrophic factor; PANSS-N = Positive and Negative Syndrome Scale, negative subscale. amygdala relative volume (t = 2.47; p = 0.018) in patients Discussion with first-episode psychosis (Appendix 1, Figures S4 and S5). Based on the findings of the multivariate analysis, the In this study, we found that lower serum BDNF levels, ­final model included 2 baseline variables: serum BDNF smaller left amygdala relative volume and more negative levels and left amygdala relative volume (R 2 = 0.533). symptomatology at baseline predicted more negative symp­ These results further supported the association of lower toms (PANSS negative subscale) at 1-year follow-up. In ad­ baseline serum BDNF levels (OR 0.93, 95% CI 0.88 to 0.98; dition, lower baseline serum BDNF levels and a smaller left p = 0.012) and smaller left amygdala relative volume (OR amygdala relative volume were predictors of persistent 0.99, 95% CI 0.98 to 0.99; p = 0.010) with the presence of negative symptoms at 1 year. Finally, compared with persistent negative symptoms at 1 year in patients with healthy controls, patients with first-episode psychosis had first-episode psychosis (Table 4). We repeated this analysis smaller relative amygdalar volumes and a trend toward including chlorpromazine equivalents to control for anti­ smaller hippocampal volumes at baseline. psychotic adverse effects and found no significant differ­ In our cohort, lower baseline serum BDNF levels pre­ ences (Appendix 1, Table S4). When we conducted the dif­ dicted higher negative symptoms at 1 year, measured using ferent models adding 2 variables each time, BDNF the PANSS negative symptom subscale or persistent nega­ explained 27% of the variance and left amygdala relative tive symptoms. These findings were consistent with previous volume explained 22.7% of the variance. reports, which have also found that lower serum BDNF levels J Psychiatry Neurosci 2022;47(1) E27 Toll et al. Table 4: Persistent negative symptoms at 1-year follow-up: logistic regression model Predictor –2 log-likelihood Nagelkerke R2 χ2 OR (95% CI) p value Step 1 44.09 0.27 χ21 = 8.71 0.06 BDNF at baseline 0.96 (0.93 to 0.99) 0.018 Step 2 33.16 0.53 χ22 = 19.62 Left amygdala relative volume 0.99 (0.98 to 0.99) 0.010 BDNF at baseline 0.93 (0.88 to 0.98) 0.012 BDNF = brain-derived neurotrophic factor; CI = confidence interval; OR = odds ratio. are associated with negative symptoms in patients with first- Nevertheless, we found no association or mediation effect episode psychosis48 and schizophrenia.49,50 Our results were between negative symptoms and amygdala volume or BDNF also in line with those of a study that found more severe levels, as seen in the mediation analysis (Appendix 1, negative symptoms in patients with first-episode psychosis Figure S6). This would suggest that the mechanism by which who were carriers of the Met allele of the BDNF Val66Met negative symptoms develop in the case of the amygdala are polymorphism, a variant that has been associated with de­ independent of BDNF levels. As we commented above, creased BDNF activity.51 Nonetheless, a meta-analysis con­ BDNF may be associated with negative symptoms through a cluded that decreased peripheral BDNF levels did not ap­ functional process by increasing dopamine activity in the pear to be associated with the severity of positive or ­mesolimbic pathway.53 negative symptoms.52 However, it is important to note that This mechanism could also explain the lack of a significant the association between BDNF levels and negative symp­ association between hippocampal volumes and negative toms in that meta-analysis was assessed cross-sectionally symptoms that one would expect, given the known associa­ and did not take into account possible confounders, such as tion between hippocampus and BDNF, and given previous secondary sources of negative symptoms. findings. Furthermore, it could also be that in previous stud­ The findings of this study were in agreement with animal ies hippocampal volume was not directly associated with studies showing an influence of BDNF on dopamine release negative symptoms, but rather with other symptoms com­ in the mesolimbic dopamine pathway,53 which in turn is as­ monly associated with them, such as cognitive symptoms.60 sociated with improved motivational behaviours.54 However, In our exploratory analyses of the subitems of the PANSS the design of the present study did not allow us to test the negative subscale, only the total baseline PANSS negative sub­ mechanisms underlying the association between BDNF and scale score predicted PANSS negative subscale subitems at negative symptoms. follow-up (except for emotional withdrawal and stereotyped Left amygdalar relative volume at baseline predicted both thinking, which were predicted by baseline CDSS scores and negative symptoms (PANSS negative subscale) and persis­ sex, respectively). Patients with more symptoms of depression tent negative symptoms at 1-year follow-up. Previous studies at baseline had higher emotional withdrawal scores at 1-year have shown that structural abnormalities in the amygdalar– follow-up. Although it would be reasonable to assume that hippocampal complex are associated with negative symp­ higher emotional withdrawal scores were attributable to an af­ toms in patients with schizophrenia and first-episode psy­ fective disorder, it is important to note that no patients scored 6 chosis.10 One study found that patients with first-episode or higher on the CDSS at baseline or follow-up. Consequently, psychosis who presented with persistent negative symp­ this finding for emotional withdrawal could have been be­ toms had smaller left amygdalar and right hippocampal cause of difficulties in differentiating between depressive and volumes than patients without persistent negative symp­ negative symptoms in patients with psychosis, and particu­ toms.26 Furthermore, several studies have found an associa­ larly in those with first-episode psychosis. On the other hand, tion between functional activation patterns in the amygdala the fact that none of the other variables were significant predic­ and the severity of affective flattening in patients with tors of the PANSS negative subscale subitems could have been schizophrenia.55,56 secondary to the sample size. However, another explanation The amygdala plays a key role in emotion and motiva­ could be that the total PANSS negative subscale score meas­ tional behaviours.57 Alterations in its structure might contrib­ ures something different than the sum of the subitems. ute to the differential expression of negative symptoms. The The variables that predicted negative symptoms on the replicated findings regarding the association between nega­ PANSS negative subscale and persistent negative symptoms tive symptoms and left amygdala volume in first-episode at 1 year were highly consistent. This finding was not sur­ psychosis26–28 suggest that the lateralization of amygdala vol­ prising, given that baseline PANSS negative subscale scores ume differences in patients with first-episode psychosis may were among the main predictors of PANSS negative subscale be an early predictor of negative symptomatology. This find­ scores at 1 year. These findings were consistent with previous ing would be in line with those of some meta-analyses, which reports, which have found that negative symptoms tend to reported that amygdala activations are lateralized to the left, be persistent if they are already present at illness onset, even particularly for negative symptoms.58,59 though their severity may fluctuate over time.10 E28 J Psychiatry Neurosci 2022;47(1) Predictors of negative symptoms in first-episode psychosis With this multidimensional approach, we predicted to help guide clinicians in developing individualized treat­ 47.8% of the variance of negative symptomatology meas­ ment plans for patients with early-stage disease. However, ured with the PANSS negative subscale at 1-year follow-up future studies are needed to replicate these findings and in­ and 53.3% of the presence of persistent negative symptoms. clude the different dimensions of negative symptoms, as These findings may help guide clinicians from illness onset well as motivation and pleasure and emotional expressivity (e.g., applying more intensive treatments for negative factors with new generation-scales.67 symptoms in patients who are more vulnerable to severe negative symptoms or persistent negative symptoms in the Acknowledgements: The authors thank the patients and volunteers long term). Among such treatments,61 physical activity and who made this study possible. They also thank Bradley Londres for professional English-language editing. add-on antidepressants should be considered, because they have been shown to have beneficial effects on negative Affiliations: From the Institut de Neuropsiquiatria i Adiccions symptoms62 and to increase BDNF levels.63 However, future (INAD), Parc de Salut Mar, Barcelona, Spain (Toll, Bergé, Canosa, Legido, Pérez-Solà, Mané); the Fundació Hospital del Mar Medical studies should be carried out to test whether an increase in Research Institute (IMIM), Barcelona, Spain (Toll, Bergé, Duran, BDNF levels mediates the improvement of negative symp­ Canosa, Legido, Pérez-Solà, Mané); the Centro de Investigación toms with these treatments. Furthermore, the inclusion of Biomédica en Red, Área de Salud Mental (CIBERSAM), Spain other variables should be considered, such has whole-brain (Toll, Blanco-Hinojo, Bergé, Canosa, Pérez-Solà, Mané); the Depart­ gyrification,64 to predict negative symptoms at follow-up ment of Psychiatry and Forensic Medicine, Universitat Autónoma de Barcelona (UAB), Barcelona, Spain (Toll); the MRI Research even more accurately. Unit, Department of Radiology, Parc de Salut Mar, Barcelona, Finally, this study provides additional evidence to the rep­ Spain (Blanco-Hinojo); the Pharmacology Unit, Department of licated finding that patients with first-episode psychosis have Clinical Fundamentals, Faculty of Medicine, Barcelona University, smaller amygdala volumes than healthy controls.65 However, Barcelona, Spain (Marmol); and the Department of Psychiatry and Behavioral and Clinical Neuroscience Institute, University of Cam­ a high proportion of our patients were cannabis users, and bridge, United Kingdom (Fernández -Egea). that could have affected our results.66 Competing interests: None declared. Limitations Contributors: A. Toll and A. Mané designed the study. I. Canosa acquired the data, which L. Blanco-Hinojo, D. Bergé, X. Duran, T. This study had several limitations. The first is inherent to all Legido, F. Marmol, V. Pérez-Solà and E. Fernández-Egea ana­ lyzed. A. Toll wrote the article, which all other authors reviewed. longitudinal studies: not all of the participants completed the All authors approved the final version to be published and can entire follow-up. Nonetheless, we found no differences in certify that no other individuals not listed as authors have made baseline characteristics between completers and noncom­ substantial contributions to the paper. pleters, indicating that bias was unlikely. In addition, we did Content licence: This is an Open Access article distributed in ac­ not use a scale to assess extrapyramidal symptoms, although cordance with the terms of the Creative Commons Attribution we did take into account antipsychotic adverse effects by in­ (CC BY-NC-ND 4.0) licence, which permits use, distribution and cluding chlorpromazine equivalents in the analyses. Finally, reproduction in any medium, provided that the original publica­ we were unable to assess negative symptoms and their tion is properly cited, the use is noncommercial (i.e., research or educational use), and no modifications or adaptations are made. ­dimensions with next-generation scales, which were not See: https://creativecommons.org/licenses/by-nc-nd/4.0/ available when the study was initiated. Despite these limitations, this study had several important strengths, most notably the large sample of antipsychotic-naive References patients with first-episode psychosis, a patient profile that is particularly difficult to recruit. Another strength was the long 1. Mas S, Boloc D, Rodríguez N, et al. Examining gene-environment interactions using aggregate scores in a first-episode psychosis co­ follow-up period (1 year). Finally, the innovative, multi­ hort. Schizophr Bull 2020;46:1019-25. dimensional baseline assessment used in this study further 2. GBD 2016 Disease and Injury Incidence and Prevalence Collabora­ strengthened our findings. We included numerous param­ tors. Global, regional, and national incidence, prevalence, and eters (clinical, neuroimaging and laboratory assessment) to years lived with disability for 328 diseases and injuries for 195 build a robust model for predicting negative symptoms. countries, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet 2017;390:1211-59. Conclusion 3. Harvey PD, Strassnig MT, Silberstein J. Prediction of disability in schizophrenia: symptoms, cognition, and self-assessment. J Exp Psychopathol 2019;10:204380871986569. Our findings show that several baseline parameters — low 4. Fusar-Poli P, McGorry PD, Kane JM. Improving outcomes of first- serum BDNF levels, decreased left amygdalar relative vol­ episode psychosis: an overview. World Psychiatry 2017;16:251-65. ume and more negative symptomatology — were predic­ 5. Carvalho AF, Solmi M, Sanches M, et al. Evidence-based umbrella tive of negative symptoms at 1-year follow-up on the review of 162 peripheral biomarkers for major mental disorders. PANSS negative subscale (after taking into account second­ Transl Psychiatry 2020;10:152. ary sources). Two of those baseline variables — low serum 6. Fusar-Poli P, Hijazi Z, Stahl D, et al. The science of prognosis in BDNF and a smaller left amygdalar relative volume — were psychiatry: a review. JAMA Psychiatry 2018;75:1289-97. predictive of persistent negative symptoms at 1 year. These 7. Ross CA, Margolis RL, Reading SA, et al. Neurobiology of schizo­ findings are highly relevant, because they provide key data phrenia. Neuron 2006;52:139-53. J Psychiatry Neurosci 2022;47(1) E29 Toll et al. 8. Harvey PD, Strassnig M. Predicting the severity of everyday functional 29. Numakawa T, Odaka H, Adachi N. Derived neurotrophin factor in disability in people with schizophrenia: cognitive deficits, functional the neurogenesis and neuronal function, and its involvement in capacity, symptoms, and health status. World Psychiatry 2012;11:73-9. the pathophysiology of brain diseases. Int J Mol Sci 2018;19:3650. 9. Galderisi S, Rossi A, Rocca P, et al. The influence of illness‐related 30. Green MJ, Matheson SL, Shepherd A, et al. Brain-derived neuro­ variables, personal resources and context‐related factors on real‐ trophic factor levels in schizophrenia: a systematic review with life functioning of people with schizophrenia. World Psychiatry meta-analysis. Mol Psychiatry 2011;16:960-72. 2014;13:275-87. 31. Toll A, Mané A. Brain-derived neurotrophic factor levels in first epi­ 10. Hovington CL, Lepage M. Neurocognition and neuroimaging of sode of psychosis: a systematic review. World J Psychiatry 2015;5:154-9. persistent negative symptoms of schizophrenia. Expert Rev Neu- rother 2012;12:53-69. 32. Niitsu T, Ishima T, Yoshida T, et al. A positive correlation between serum levels of mature brain-derived neurotrophic factor and neg­ 11. Mezquida G, Cabrera B, Bioque M, et al. The course of negative ative symptoms in schizophrenia. Psychiatry Res 2014;215:268-73. symptoms in first-episode schizophrenia and its predictors: a pro­ spective two-year follow-up study. Schizophr Res 2017;189:84-90. 33. Bergé D, Mané A, Salgado P, et al. Predictors of relapse and function­ ing in first-episode psychosis: a two-year follow-up study. Psychiatr 12. Kirschner M, Aleman A, Kaiser S. Secondary negative symp­ Serv 2016;67:227-33. toms—a review of mechanisms, assessment and treatment. Schizophr Res 2017;186:29-38. 34. Strauss GP, Cohen AS. A transdiagnostic review of negative symp­ tom phenomenology and etiology. Schizophr Bull 2017;43:712-9. 13. Carpenter WT, Heinrichs DW, Wagman AM. Deficit and nondeficit forms of schizophrenia: the concept. Am J Psychiatry 1988;145:578-83. 35. Kay SR. Positive-negative symptom assessment in schizophrenia: psychometric issues and scale comparison. Psychiatr Q 1990;61:163-78. 14. Buchanan RW. Persistent negative symptoms in schizophrenia: an overview. Schizophr Bull 2007;33:1013-22. 36. Aas IH. Guidelines for rating Global Assessment of Functioning (GAF). Ann Gen Psychiatry 2011;10:2. 15. Ochoa S, Usall J, Cobo J, et al. Gender differences in schizophrenia and first-episode psychosis: a comprehensive literature review. 37. Addington D, Addington J, Schissel B. A depression rating scale Schizophr Res Treatment 2012;2012:916198. for schizophrenics. Schizophr Res 1990;3:247-51. 16. Morgan VA, Castle DJ, Jablensky AV. Do women express and ex­ 38. Gardner DM, Murphy AL, O’Donnell H, et al. International consen­ perience psychosis differently from men? Epidemiological evi­ sus study of antipsychotic dosing. Am J Psychiatry 2010;167:686-93. dence from the Australian National Study of Low Prevalence 39. Galderisi S, Mucci A, Bitter I, et al. Persistent negative symptoms in first (Psychotic) Disorders. Aust N Z J Psychiatry 2008;42:74-82. episode patients with schizophrenia: results from the European First 17. Kao YC, Liu YP. Effects of age of onset on clinical characteristics in Episode Schizophrenia Trial. Eur Neuropsychopharmacol 2013;23:196-204. schizophrenia spectrum disorders. BMC Psychiatry 2010;10:63. 40. Piccinni A, Marazziti D, Del Debbio A, et al. Diurnal variation of 18. Schultz SK, Ho BC, Andreasen NC. Clinical features characterizing plasma brain-derived neurotrophic factor (BDNF) in humans: an young-onset and intermediate-onset schizophrenia. J Neuropsychia- analysis of sex differences. Chronobiol Int 2008;25:819-26. try Clin Neurosci 2000;12:502-5. 41. Arnold SJM, Ivleva EI, Gopal TA, et al. Hippocampal volume is re­ 19. Boonstra N, Klaassen R, Sytema S, et al. Duration of untreated duced in schizophrenia and schizoaffective disorder but not in psy­ psychosis and negative symptoms—a systematic review and meta- chotic bipolar I disorder demonstrated by both manual tracing and analysis of individual patient data. Schizophr Res 2012;142:12-9. automated parcellation (FreeSurfer). Schizophr Bull 2015;41:233-49. 20. Jabar LSA, Sørensen HJ, Nordentoft M, et al. Associations between 42. Haukvik UK, Hartberg CB, Nerland S, et al. No progressive brain duration of untreated psychosis and domains of positive and neg­ changes during a 1-year follow-up of patients with first-episode ative symptoms persist after 10 years of follow-up: a secondary psychosis. Psychol Med 2016;46:589-98. analysis from the OPUS trial. Schizophr Res 2020;30:30567-3. 43. Sasabayashi D, Takayanagi Y, Takahashi T, et al. Subcortical brain 21. Immonen J, Jääskeläinen E, Korpela H, et al. Age at onset and the volume abnormalities in individuals with at-risk mental state. outcomes of schizophrenia: a systematic review and meta-analysis. Schizophr Bull 2020;46:834-45. Early Interv Psychiatry 2017;11:453-60. 44. van Erp TG, Hibar DP, Rasmussen JM, et al. Subcortical brain volume 22. Sabe M, Zhao N, Kaiser S. Cannabis, nicotine and the negative abnormalities in 2028 individuals with schizophrenia and 2540 healthy symptoms of schizophrenia: systematic review and meta-analysis controls via the ENIGMA consortium. Mol Psychiatry 2016;21:547-53. of observational studies. Neurosci Biobehav Rev 2020;116:415-25. 45. Fischl B, Salat DH, Busa E, et al. Whole brain segmentation: auto­ 23. Shenton ME, Whitford TJ, Kubicki M. Structural neuroimaging in mated labeling of neuroanatomical structures in the human brain. schizophrenia: from methods to insights to treatments. Dialogues Neuron 2002;33:341-55. Clin Neurosci 2010;12:317-32. 46. Pujol N, Mané A, Bergé D, et al. PEPs GROUP. Influence of BDNF 24. Rajarethinam R, DeQuardo JR, Miedler J, et al. Hippocampus and and MTHFR polymorphisms on hippocampal volume in first-episode amygdala in schizophrenia: assessment of the relationship of psychosis. Schizophr Res 2020;223:345-52. neuro­anatomy to psychopathology. Psychiatry Res 2001;108:79-87. 47. Szamosi A, Kelemen O, Kéri S. Hippocampal volume and the AKT 25. Koutsouleris N, Gaser C, Jäger M, et al. Structural correlates of signaling system in first-episode schizophrenia. J Psychiatr Res 2012;​ psychopathological symptom dimensions in schizophrenia: 46:279-84. a voxel-based morphometric study. Neuroimage 2008;39:1600-12. 48. Rizos EN, Michalopoulou PG, Siafakas N, et al. Association of 26. Makowski C, Bodnar M, Shenker JJ, et al. Linking persistent nega­ ­s erum brain-derived neurotrophic factor and duration of un­ tive symptoms to amygdala-hippocampus structure in first-episode treated psychosis in first-episode patients with schizophrenia. psychosis. Transl Psychiatry 2017;7:e1195. Neuropsychobiology 2010;62:87-90. 27. Honea R, Crow TJ, Passingham D, et al. Regional deficits in brain 49. Carlino D, Leone E, Di Cola F, et al. Low serum truncated-BDNF volume in schizophrenia: a meta-analysis of voxel-based mor­ isoform correlates with higher cognitive impairment in schizo­ phometry studies. Am J Psychiatry 2005;162:2233-45. phrenia. J Psychiatr Res 2011;45:273-9. 28. Bois C, Levita L, Ripp I, et al. Hippocampal, amygdala and nu­ 50. Chen DC, Wang J, Wang B, et al. Decreased levels of serum brain- cleus accumbens volume in first-episode schizophrenia patients derived neurotrophic factor in drug-naïve first-episode schizophre­ and individuals at high familial risk: a cross-sectional comparison. nia: relationship to clinical phenotypes. Psychopharmacology (Berl) Schizophr Res 2015;165:45-51. 2009;207:375-80. E30 J Psychiatry Neurosci 2022;47(1) Predictors of negative symptoms in first-episode psychosis 51. Mezquida G, Penadés R, Cabrera B, et al. Association of the brain- 59. Wager TD, Phan KL, Liberzon I, et al. Valence, gender, and lateral­ derived neurotrophic factor Val66Met polymorphism with nega­ ization of functional brain anatomy in emotion: a meta-analysis of tive symptoms severity, but not cognitive function, in first-episode findings from neuroimaging. Neuroimage 2003;19:513-31. schizophrenia spectrum disorders. Eur Psychiatry 2016;38:61-9. 60. Hovington CL, Bodnar M, Joober R, et al. Impairment in verbal 52. Fernandes BS, Steiner J, Berk M, et al. Peripheral brain-derived memory observed in first episode psychosis patients with persis­ neurotrophic factor in schizophrenia and the role of antipsychot­ tent negative symptoms. Schizophr Res 2013;147:223-9. ics: meta-analysis and implications. Mol Psychiatry 2015;20:1108-19. 61. Galderisi S, Mucci A, Dollfus S, et al. EPA guidance on assessment 53. Narita M, Aoki K, Takagi M, et al. Implication of brain-derived of negative symptoms in schizophrenia. Eur Psychiatry 2021;64:e23. neurotrophic factor in the release of dopamine and dopamine-related 62. Sabe M, Kaiser S, Sentissi O. Physical exercise for negative symp­ behaviors induced by methamphetamine. Neuroscience 2003;​ toms of schizophrenia: systematic review of randomized con­ 119:767-75. trolled trials and meta-analysis. Gen Hosp Psychiatry 2020;62:13-20. 54. Boekhoudt L, Wijbrans EC, Man JHK, et al. Enhancing excitability 63. Björkholm C, Monteggia LM. BDNF—a key transducer of antide­ of dopamine neurons promotes motivational behaviour through in­ pressant effects. Neuropharmacology 2016;102:72-9. creased action initiation. Eur Neuropsychopharmacol 2018;28:171-84. 64. Hauke DJ, Schmidt A, Studerus E, et al. PRONIA Group. Multi­ 55. Lepage M, Sergerie K, Benoit A, et al. Emotional face processing modal prognosis of negative symptom severity in individuals at and flat affect in schizophrenia: functional and structural neural. increased risk of developing psychosis. Transl Psychiatry 2021;​ correlates. Psychol Med 2011;41:1833-44. 11:312. 56. Gur RE, Loughead J, Kohler CG, et al. Limbic activation associated 65. Watson DR, Bai F, Barrett SL, et al. Structural changes in the hippo­ with misidentification of fearful faces and flat affect in schizophre­ campus and amygdala at first episode of psychosis. Brain Imaging nia. Arch Gen Psychiatry 2007;64:1356-66. Behav 2012;6:49-60. 57. Fernando AB, Murray JE, Milton AL. The amygdala: securing 66. Batalla A, Bhattacharyya S, Yücel M, et al. Structural and functional pleasure and avoiding pain. Front Behav Neurosci 2013;7:190. imaging studies in chronic cannabis users: a systematic review of adolescent and adult findings. PLoS One 2013;8:e55821. 58. Baas D, Aleman A, Kahn R. Lateralization of amygdala activation: a systematic review of functional neuroimaging studies. Brain Res 67. Strauss GP, Hong LE, Gold JM, et al. Factor structure of the Brief Rev 2004;45:96-10. Negative Symptom Scale. Schizophr Res 2012;142:96-8. J Psychiatry Neurosci 2022;47(1) E31