Intrinsic functional connectivity of fronto-temporal networks in early onset psychosis at 2 years after the first episode

Introduction Intrinsic functional connectivity (iFC) of resting-state networks has been reported to be altered in schizophrenia, especially in fronto-temporal networks [1]. Our team have recently observed reduced connectivity in the right middle/inferior frontal gyrus (rIFG) within a network functionally related to language, in adolescents suffering a first episode of psychosis in comparison to healthy controls (HC) [2]. The aim of this study is to test whether these changes persist two year after clinical onset and to examine whether there are any structural correlates. Methods A case-control cross-sectional study was performed evaluating iFC and grey matter structure in adolescents two-years-after their first psychotic episode compared to HC, in sample with a partial overlap (33%) with regards to our previous report. Twenty-nine adolescents with Early Onset Psychosis (EOP) participated in this two-year follow-up assessment and were compared cross-sectionally to 36 HC. After excluding 3 cases due to excess movement, final sample (n=62) did not present significant differences in age (EOP=17.6 ± 0.3 years; HC= 18.3 ± 0.3 years) or sex (EOP=58% female; HC=50% female). Diagnoses were: schizophrenia (8), schizoaffective disorder (5), bipolar disorder (4), major depression (2) and psychosis not-otherwise specified (7). A DARTEL algorithm was applied to the segmented T1-structural volumes to generate a sample-specific template. Functional resting-state images were pre-processed using SPM12 and co-registered to this template. Three components obtained with independent component analysis (toolbox GIFT) corresponding to the salience, language and DMN were identified from visual inspection. Whole brain analysis t-tests were covaried by sex and age using an inclusive grey-matter mask. No differences were observed in total intracraneal, grey matter or white matter volumes amongst groups. Voxel Based Morphometry (VBM) analysis was covaried for age, sex and total intracranial volume. Secondarily, we performed a region of interest (ROI) analysis using the marsbar toolbox: data of iFC and grey matter volume was extracted, using a 10-mm-radius sphere centred at MNIx,y,z [42, 39, -3] [2] (rIFG) and performed a group comparison in Stata13. Results There were no differences between groups in iFC in either of the resting state networks. The ROI analysis corresponding to the rIFG in the language network revealed an effect of group, consisting of reduced iFC in EOP participants relative to HC, survived when controlled by age and sex (p = 0.026). There were no group-effects in grey matter volumes for this region of interest. It was found an inverse correlation between grey matter and iFC at rIFG, only significant in HC group (p = 0.02). Discussion Our results support persistence of reduced iFC in the language network in adolescents with EOP, 2 years after the first episode, in a mostly independent sample with regards our baseline study [2]. This provides further support to the notion that abnormal fronto-temporal connectivity may characterize EOP at different stages of the disease. Nevertheless, this finding needs to be interpreted taking into account that the present results correspond to ROI analyses and that the current study design does not allow to examine longitudinal change. Further research should focus on assessing longitudinal differences in resting-state networks in EOP.