Brain Abnormalities And Iq In Offspring, Siblings, Co-Twins, And Parents Of Patients With Schizophrenia
Sonja De Zwarte,Rachel Brouwer,Manon Hillegers,Wiepke Cahn,Hilleke Hulshoff Pol,Rene Kahn,Neeltje Van Haren
DOI: https://doi.org/10.1093/schbul/sbx021.127
2017-01-01
Schizophrenia Bulletin
Abstract:Background: Schizophrenia (SZ) is associated with brain abnormalities and cognitive impairment, and to some extent first-degree relatives of patients also show these deficits. However, the extent of the deficits differs between studies and possibly between types of relatives. We propose that this discrepancy might be explained by generational differences in familial risk. Here, we compare siblings, twins, parents, and offspring of patients with SZ to healthy controls (HC) and patients on total brain (TB), gray (GM) and white matter (WM), hippocampal (HIP) volumes, total surface area (SA), and mean cortical thickness (CT) and investigate (1) whether brain abnormalities vary between different types of relatives and (2) whether these brain abnormalities are associated with IQ. Methods: A total of 1011 participants were included from 5 SZ family studies: 342 relatives who are unaffected for psychosis (i.e., 41 offspring, 211 siblings, 44 parents, 26 dizygotic (DZ) co-twins, 20 monozygotic (MZ) co-twins), 438 HC, and 231 patients. FreeSurfer was used to estimate the brain measures from MRI. Age and gender effects were regressed out in each cohort and standardized residuals were calculated. Linear mixed model analyses were performed comparing patients, HC, and relatives, taking family relatedness into account. Effect sizes were calculated for the individual relative groups when compared to HC. To investigate the role of IQ, all analyses were performed both with and without correcting for IQ. Results: As a group, relatives had significantly lower WM, SA, TB and GM than HC (all Ps < .02), while CT and HIP reached trend level significance. Patients were significantly decreased as compared to relatives in TB, HIP, GM, and CT but not in WM and SA. When correcting for IQ, no significant differences were found between relatives and HC, and patients had smaller HIP and CT than relatives and HC (Ps < .01). Comparing the different relative types, offspring showed the greatest effect sizes compared to HC in all measures, except CT where MZ co-twins showed the largest effect. After IQ correction, most effects disappeared; only decreased CT in MZ remained significant. Conclusion: These results suggest a familial (possibly genetic) component to GM, TB, WM and SA decreases in SZ. In contrast, a thinner cortex and lower hippocampal volume were only observed in patients, implicating an effect of illness rather than a familiar marker. Preliminary results indicate that SZ offspring show more abundant brain abnormalities as compared with parents, siblings and twins, in line with a high relative risk for developing SZ in offspring. Importantly, lower IQ explained most of the brain abnormalities found in relatives, suggesting that genetic vulnerability for SZ is also related to IQ. To increase sample sizes, we are currently including subjects in collaboration with the ENIGMA consortium for a large-scale meta-analysis.