A factor analysis of the rat's corpus callosum

Abstract

Previous work from our laboratory (Berrebi et al., Brain Research, 438 (1988) 216–224) demonstrated region-specific sexual dimorphisms in the size of the rat's corpus callosum, which are modifiable by extra stimulation in early life. These differences are assumed to reflect regional corticocortical fibers of passage which are altered differentially by gender and our experimental manipulations. In this paper, we report our findings when the original data are reanalyzed using a newly developed computer program. This program not only reproduced, with very high accuracy, the original means, but also permitted us to examine computer generated callosal width scores via a factor analysis procedure. Such a procedure yields useful information concerning the clustering of callosal fibers and thus contributes significantly to our hypothesis that discrete cortical regions are selectively sensitive to experimental variables. Factor analyses of the callosal variables and brain weight of 155 rats found 7 width factors, and an eighth factor which contained the variables of brain weight, callosal length, and callosal perimeter. Callosal area did not load significantly on any of these factors. The percentile locations of the width factors, starting the anterior (genu) end were: widths 1–5, 6–17, 24–38, 46–57, 62–72, 79–95 and 96–99. Use of these factor scores in analyses of variance revealed that the male callosum is wider than the female's with the differences most pronounced in the genu and the most posterior portion of the splenium. Both age and early handling experience influenced the callosal width factors. The factor analysis as well as other statistical analyses confirmed our prior conclusion that callosal length and perimeter represent one cluster of callosal parameters which are independent of a second cluster centering around the width measures and area.