22q11.2 Microdeletion syndrome (22qDS) is caused by a recurrent genetic mutation associated with a high degree of social impairment, and also represents one of the greatest known genetic risk factors for schizophrenia. This syndrome therefore represents an excellent model for investigating how a known genetic "lesion" may lead to abnormal development of social behavior and to the expression of a diagnosable psychotic disorder. However, little is known about vulnerabilities that contribute to the development of psychosis in this population, nor have the neurobiological substrates of social cognitive impairment been explored in 22qDS. The present investigation sought to examine social cognitive risk factors, at the level of both behavior and neuroanatomy, which may contribute to psychotic symptomatology in adolescents and young adults with 22qDS.
We conducted three separate studies to investigate these questions. In the first study (22qDS= 31, controls=31), using behavioral measures, we sought to determine whether social cognition better predicts positive symptoms than does non-social cognition in 22qDS. The primary aims of study 2 (22qDS=31, controls= 34) were: 1) to investigate neuroanatomic alterations in socially relevant brain regions (i.e., amygdala, fusiform gyrus, superior temporal gyrus, insula, anterior cingulate, and frontal regions), using structural magnetic resonance imaging (sMRI); and 2) to determine whether such alterations were associated with psychotic symptoms and social cognition in 22qDS patients. Finally, in study 3 (22qDS=26, controls=23), we used diffusion tensor imaging (DTI) to: 1) examine alterations in white matter tracts connecting these `social brain' regions in 22qDS patients relative to typically developing youth, and 2) to determine whether white matter microstructural abnormalities were associated with psychotic symptoms and social cognition in 22qDS patients.
Several novel findings emerged from these studies. First, in study 1, we found that Theory of Mind (ToM) performance was the best predictor of positive symptoms in 22qDS, accounting for 39% of the variance in symptom severity. In study 2, in comparison to typically developing controls, 22qDS participants showed disruptions in multiple brain regions associated with social cognition. In particular, those with 22qDS had increased cortical volumes in bilateral orbitofrontal cortices and insula, which appeared to be driven by increased cortical thickness in these regions, and decreased cortical volume in bilateral fusiform gyrus and anterior cingulate, which appeared to be driven by decreased surface area in these regions. We also found that increased cortical thickness in the right medial orbitofrontal cortex was significantly associated with increased positive symptom severity in 22qDS, while increased right amygdala volumes were associated with better social cognition performance in 22qDS. Finally, in study 3, in comparison to typically developing controls, 22qDS participants showed reduced white matter integrity in the left inferior frontal fasciculus and right uncinate fasciculus, fiber tracts that connect occipital to the temporal lobes and medial temporal with orbitofrontal regions, respectively. 22qDS participants also had significantly decreased axial diffusivity, a putative index of axonal damage, in multiple tracts, including the bilateral inferior and superior longitudinal fasciculus (which connects the parietal to the frontal lobes), and the uncinate fasciculus. Greater severity of positive symptoms was associated with decreased axial diffusivity in the left inferior frontal fasciculus and right superior longitudinal fasciculus; in contrast, increased axial diffusivity in the left inferior longitudinal fasciculus was associated with better social cognition in 22qDS.
Considering that both social impairment and neuroanatomic abnormalities predate the onset of psychosis in 22qDS, these findings provide novel information about the relationship between social cognition and psychosis risk in 22qDS. Importantly, study 2 is the first to investigate multiple measures of structural neuroanatomy (i.e., volume, cortical thickness, surface area) in 22qDS and provides important information about functionally distinct subcomponents that may contribute to alterations in cortical volume in social relevant neuroanatomic regions. Also, when testing the joint contribution of behavioral and neuroanatomic measures to prediction of positive symptoms in 22qDS, we found that right medial orbitofrontal cortical thickness and ToM task performance accounted for 43% of the variance in positive symptoms in 22qDS, significantly improving the prediction of positive symptoms in comparison to the ToM behavioral measure alone. Study 3 represents one of the first investigations of multiple DTI indices (i.e., fractional anisotropy, axial and radial diffusivity) in 22qDS. Our pattern of results suggests that white matter microstructural disruption in 22qDS may be driven by axonal damage, rather than demyelination. Finally, given that ToM was a robust predictor of positive symptoms in our sample and exploratory analyses found relationships between positive symptoms and neuroanatomic regions associated with social cognition in 22qDS, these findings suggest that social cognition may be a valuable intermediate trait for predicting the development of psychosis.