Poster D110, Monday, March 26, 8:00-10:00 am, Exhibit Hall C
Multivariate Prediction of General Intelligence from Patterns of Gray Matter Density
Kirsten Hilger1,2, Tim Hahn3, Christian Fiebach1,2, Ulrike Basten1; 1Goethe University Frankfurt, Frankfurt am Main, Germany, 2IDeA Center for Individual Development and Adaptive Education, Frankfurt am Main, Germany, 3Universitätsklinikum Münster, Münster, Germany
General intelligence has been associated with individual differences in morphological characteristics of the brain such as gray matter density (for meta-analyses, see Jung & Haier, 2007; Basten et al., 2015). However, the majority of previous investigations focused on correlative associations, which maximize explained variance within a given sample, without considering generalizability. To demonstrate the predictive value of individual differences in morphometric patterns of gray matter density for intelligence, we applied voxel-based morphometry (VBM) on structural magnetic resonance imaging (MRI) data from 308 adult participants (Nooner et al., 2012). In a regression model controlling for effects of age, sex, and handedness, intelligence (Wechsler Abbreviated Scale of Intelligence) was significantly associated with gray matter density in inferior frontal gyrus, middle temporal gyrus, lingual gyrus, precuneus, hippocampal region, and cerebellum (subsample of N = 200 used for model development). Using linear regression, we demonstrate that the multivariate pattern of gray matter density within these brain regions significantly predicts individual intelligence scores in the remaining, i.e., independent sample used for model testing (N = 108; correlation between predicted and actual intelligence scores: r = .36). Significant prediction was also achieved with a machine learning approach, i.e., support vector regression with nested cross-validation applied to the whole sample (correlation between predicted and actual intelligence scores: r = .28). In conclusion, our study demonstrates that the multivariate pattern of individual differences in gray matter density is predictive of individual intelligence scores, even in previously unseen individuals.
Topic Area: THINKING: Reasoning