TBV and full-scale IQ within the combined sample were significantly correlated (r
= .284; p
≤ 0.025). Similarly, TBV was significantly associated with callosal thickness at numerous locations (). More specifically, we observed significant positive correlations in the callosal posterior body and also in most extreme anterior and posterior callosal sections, located within the anterior third and splenium. Significant negative correlations were completely absent. Altogether, these findings are in close agreement with results from previous studies where relationships between brain size and intelligence (McDaniel, 2005
) as well as between brain size and callosal size (Rauch and Jinkins, 1994
; Jancke et al., 1997
) have been reported. Thus to ensure that callosal thickness/IQ associations are not driven by overall brain size effects, all subsequent analyses were performed while removing variance explained by TBV.
Figure 1 Correlations between TBV and callosal thickness. Illustrated are correlation coefficients (top map) and associated significance values (bottom map). The upper color bar encodes the r-values that depict the magnitude and direction of correlations between (more ...)
Full-scale IQ correlations within the combined sample
We detected significant positive correlations between callosal thickness and full-scale IQ across the posterior surface of the CC, corresponding to posterior body, isthmus, and anterior sections of the splenium (regions defined according to the traditional Witelson CC classification scheme). In addition, another smaller region at the border between anterior third and anterior body exhibited positive correlations (). Permutation testing confirmed positive correlations for the combined sample (). Negative correlations, as depicted in the r-map (), did not reach the threshold of significance. To further illustrate relationships between callosal thickness and full-scale IQ, we sampled the callosal point that showed the highest p-value (point #65). The correlations between full-scale IQ and callosal thickness measures obtained from this same surface point in each subject and after partialling out TBV are shown in (r = .476, p ≤ 0.00009).
Figure 2 Correlations between callosal thickness and intelligence. Partial correlation coefficients and associated significance values for full-scale IQ measurements are shown in the first and second columns of callosal maps, while significance values for performance (more ...)
Correlations between callosal thickness and intelligence measures (corrected p-values computed by permutation testing for p<0.05.). The asterisk * denotes significant findings.
Figure 3 Correlations between callosal thickness and intelligence. Partial regression plots showing relationships between full-scale IQ and callosal thickness (callosal residuals) at the most significant callosal surface point (point # 65), within the combined (more ...)
Full-scale IQ correlations within males
The spatial location and extent of significant positive correlations between callosal thickness and full-scale IQ in males resembled the outcomes of the combined-sample analysis in posterior callosal regions (posterior body, isthmus, and splenium). In addition, the thickness of a small region located close to the anterior end of the CC was significantly positively correlated with intelligence (). Permutation testing confirmed positive correlations within males (). Partial regression plots showing positive relationships between full-scale IQ and callosal thickness in males at the most significant callosal surface point are shown in (r = .457, p ≤0.015). Negative correlations, as visualized in the r-map (), were only present in two regions within the anterior body and splenium but again did not reach the threshold of significance.
Full-scale IQ correlations within females
Significant relationships in females were less pronounced and more spatially restricted than in males. Callosal thickness was positively correlated with full-scale IQ in regions of the posterior body and anterior splenium (but omitting the isthmus; ). Partial regression plots showing positive relationships between full-scale IQ and callosal thickness in females at the most significant callosal surface point are shown in (r = .490, p ≤0.003). Positive correlations, however, were not confirmed by permutation testing (). Negative correlations, as shown in the r-map (), were present in extended regions towards the anterior and posterior end (anterior third and splenium) but did not reach the threshold of significance.
Full-scale IQ gender interaction
The different directions observed for the correlation between callosal thickness and full-scale IQ in the anterior third in males (positive correlation) and females (negative correlation) appear to provide a basis for the spatially discrete significant gender interaction detected in anterior callosal regions (). These interaction effects, however, were not confirmed by permutation testing ().
Performance IQ and verbal IQ correlations
As further demonstrated, significance profiles for performance and verbal IQ correlations, within the combined sample (), but also in males () and females (), resemble the results observed for full-scale IQ correlations. That is, significant negative correlations were completely absent, while significant positive correlations were detected predominantly in posterior callosal regions (posterior body, isthmus, and anterior splenium). These findings were confirmed by permutation testing within the combined sample and in males, but not in females (for details see ). Significant gender interactions with respect to performance IQ were completely absent (), while significant gender interaction effects on verbal IQ correlations were revealed in anterior regions (), but not confirmed by permutation testing ().