Schoenemann, P. Thomas. 1997. An MRI Study of the Relationship Between Human Neuroanatomy and Behavioral Ability. Dissertation, University of California, Berkeley.


Abstract

Hominid brain volume increased more than 3-fold in ~2.5 million years.  From an evolutionary perspective, natural selection for one or more behavioral abilities is the most likely explanation.  A large number of studies have demonstrated that brain volume is related to performance on standardized intelligence tests.  The present study expands on this work in three important ways.  First, a sibling-pair design is used, allowing for the calculation of independent within-family (WF) and between-family (BF) correlations.  Second, a broader set of cognitive dimensions (specifically guided by an understanding of human evolution) is investigated.  Third, neuroanatomical components are quantified from higher resolution 3D MR images (whole brain volume resolution of 1.3 mm3).  The subject group consisted of 36 pairs of sisters.
 
The results for BF correlations between neuroanatomical and cognitive variables are consistent with previous MRI-based studies, but WF correlations are almost entirely negative or close to zero.  The BF correlation of brain volume to the first principal component of 11 cogntive tests (1st PC) is r=0.46 (p<0.01), but the corresponding WF correlation is only r=-0.05 (NS).  Social interest variables show a similar pattern, while throwing ability and maturation rate are unrelated to neuroanatomic variation both BF and WF.

This pattern is not easily explained by methodological or statistical artifacts.  This study also shows that family differences in socio-economic status (SES) do not explain this pattern, because SES is not correlated with brain volume.  It is therefore highly unlikely that the genetic correlation between neuroanatomy and behavior is zero.  Possible explanations are: 1) Behavioral variability is genetically correlated only weakly with brain volume, precluding reliable detection in relatively small sample sizes; 2) The types of tests used might not assess behavior important for hominid brain evolution; 3) Within-family environmental effects (e.g., sibling competition) may mask any underlying genetic correlation.  This is consistent with the finding that individuals more than four years apart from their sister show a correlation between brain volume and 1st PC of r=0.65 (p<0.023, N=12).  Those closer in age only show a correlation of r=0.15 (NS, N=60).  Furthermore, several cognitive tests are negatively correlated with WF age differences.

copyright 1999 P. Thomas Schoenemann