Abstract
BACKGROUND: 22q11.2 deletion syndrome (22q11DS) is a neurogenetic condition associated with deficits in neuropsychological functioning and psychiatric disorders. This deletion confers a high risk for the development of psychosis, as approximately 30-45 % of individuals develop psychosis in adulthood. Previous reports of resting-state functional magnetic resonance imaging (rs-fMRI) functional connectivity patterns in 22q11DS have demonstrated that atypical connectivity is associated with both the emergence and severity of psychotic symptoms. However, due to sample overlap and large age ranges of samples spanning multiple critical periods of brain maturation, more independent studies with samples within the window of time when psychotic symptoms have been shown to emerge (ages 17-26) are needed. Resting-state networks (RSNs) in 22q11DS during this stage of brain development may thus provide insight into the dynamic changes in functional integration that influence the incidence of prodromal symptoms and neurocognitive deficits characteristic of this syndrome. METHODS: Independent component analysis (ICA) was performed to identify RSNs in a combined sample of 55 individuals with 22q11DS (27 males; age range 17-26) and 29 controls (17 males; age range 17-23, consisting of 8 siblings without the deletion and 21 typically developed individuals) from two research sites. We conducted a full factorial analysis to determine group differences between 22q11DS and controls. A Poisson regression analysis was conducted in the 22q11DS group to determine relationships of rs-fMRI network connectivity with psychiatric symptoms based on factors of the 18-item Brief Psychiatric Rating Scale. Nonparametric Spearman correlations were performed to test associations between within-network functional connectivity (FC) and performance on measures of verbal memory (California Verbal Learning Test) and executive function (Behavior Rating Inventory of Executive Function Adult version) in 22q11DS. RESULTS: Between-group network connectivity analyses revealed significant differences in 9 RSNs. Decreased network FC in 22q11DS was observed in the following networks: high-level visual processing network (HLVPN), low-level visual processing network (LLVPN), visual/precuneus network, left frontal-parietal network (LFPN), right frontal-parietal network (RFPN), and self-referential network (SRN). In contrast, greater network FC in 22q11DS was observed in subclusters of the LLVPN, visual/precuneus network, limbic network (LN), default mode network (DMN), and visuospatial processing network (VSPN). Increased functional connectivity of the right cuneus (visual/precuneus network) and right superior parietal lobule (DMN) in 22q11DS was positively associated with both thought disturbance and disorganization factors of the Brief Psychiatric Rating Scale (BPRS). Decreased functional connectivity in the left posterior cingulate (LLVPN) was associated with higher thought disturbance scores in 22q11DS. No associations with our neurocognitive measures passed correction for multiple comparisons (Bonferroni-corrected p ≤ 0.0014). CONCLUSIONS: Our findings suggest that atypical network connectivity within RSNs may be indicative of increased risk for developing psychosis and supports the utility of RSNs as biomarkers of prodromal symptoms in 22q11DS.