A major aim of this presentation is to demonstrate that at least a part of the large interindividual variability in callosal size might be explained by brain size differences. Our analysis revealed that the CC/brain size relationship follows a geometrical rule implying that the cross-sectional CC area increases less than proportional to brain size (FBV or weight). It follows from this relationship that larger brains have relatively smaller cross-sectional CC areas. Nevertheless, this allometric relationship with FBV explained not more than 30% of the total variability in CC size observed in our sample, demonstrating that the thickness of the CC is still influenced mainly by other factors. The majority of callosal fibers are thought to originate from association cortices and subserve higher-order functions (Innocenti, 1986;
Pandya and Seltzer, 1986; LaMantia and Rakic, 1990; Aboitiz et al., 1992a). Thus, as was previously hypothesized by Peters (Peters, 1988), a possible lack of an allo-metric relationship between the size of the brain and the association cortices could account for at least some of the variation in CC size that remains unexplained by the CC/FBV relationship.
Additional variance of CC morphology may be added by environmental factors. In animal studies of postnatal development of the CC, the number of callosal axons in neonates exceeds that of young adults, suggesting that normal development involves the remodeling of axonal projections between the two hemispheres with a subsequent elimination of callosal axons (Innocenti and Frost, 1979; Clarke et al., 1989; LaMantia and Rakic, 1990; Innocenti, this volume; Rosen, this volume). However, this reduction in the number of callosal axons, reflecting the selective elimination of axon collaterals or callosal neurons during the early postnatal period, can be manipulated experimentally by altering sensory or motor experience during early development (Innocenti, Fiore, and Caminiti, 1977; Berrebi et al., 1988). Studies of humans have suggested a considerable degree of callosal plasticity during brain development until adulthood, possibly induced by environmental stimulation (Allen et al., 1991; Pujol et al., 1993; Schlaug et al., 1995).
The lack of a principal gender difference in the CC/FBV relationship implies that small brains exhibit larger CC ratios, irrespective of gender. Thus, FBV was the main factor explaining the gender difference in our sample of 120 young, healthy subjects. However, it remains to be determined in future experiments whether gender might exert additional impact on CC morphology.
On the basis of present anatomical knowledge, a functional interpretation of this inverse relationship between forebrain size and relative callosal size must remain speculative. Let us assume that the packing densities and branching patterns of callosal neurons and axons do not depend on brain size. In that case our data would indicate that the degree of interhemispheric connectedness decreases with increasing human brain size. This would concur with theoretical predictions made by Ringo and coworkers (Ringo, 1991; Ringo et al., 1994). They argued that as brain size is scaled up, there must be a decrease in interhemispheric connectivity, owing to the increasing time constraints of transcallosal conduction delay. Consequently, functionally related neuronal elements would cluster in one hemisphere, so increasing brain size would be the driving force in the phylogeny of hemispheric specialization. With regard to callosal connectivity, our morphometric data provide the first empirical support of this conjecture.
Our finding that CC size is related to FBV should motivate future studies to normalize CC measures to brain size measures, similar to what it was originally proposed by De Lacoste et al. (1990). Brain size turns out to be an important variable in aging, in psychopa-thology, and perhaps in cognitive capacity. Thus, it is necessary to rule out that subgroup differences in callosal morphology are not solely a function of brain size. Let us consider a study designed to compare CC size between mixed- and right-handers. The mean brain size of the mixed-handers might be slightly larger by sampling errors. Then it is most likely that the CC size measures of the mixed-handers are slightly larger than those of the right-handers. Therefore, it would be more appropriate to use brain size-normalized CC size measures or to compare handedness subgroups with similar brain size measures. In general, this problem can be extended to any study comparing CC size between subgroups, whether they are normal, dysphasic, autistic, schizophrenic, or whatever.
Previous in vivo MRI studies have tried to address this challenge by measuring total midsagittal cross-sectional area of both cerebral hemispheres to predict brain size (Clarke and Zaidel, 1994; Kertesz et al., 1986; Rauch and Jenkins, 1994). However, the postmortem study of De Lacoste and coworkers (De Lacoste-Utamsing and Holloway, 1982) demonstrated that such measures cannot be used to predict brain weight because the correlation between midsagittal surface area and brain weight is rather low (r = 0.40, r2 = 0.20). Therefore, it is necessary to measure brain volume more directly.
Our results supporting the conjecture that brain asymmetry might be linked to brain size may stimulate future research in a number of ways. An important question will be whether brains of different sizes differ in neuronal packing density and branching patterns of callosal axons. It will also be worth examining the relationship between brain size, established behavioral (language laterali-zation) asymmetries, and structural asymmetries (e.g., planum temporale and planum parietale asymmetry) (Steinmetzetal., 1991; Aboitiz et al., 1992c; Jancke et al., 1994). If there is indeed a relationship between brain size and asymmetry, larger brains should demonstrate stronger anatomical and functional asymmetries. This then may provide clues for the phylogeny of hemispheric dominance in the higher primates (Ringo, 1991; Ringo et al., 1994).
Was this article helpful?