Rapid CommunicationStructural brain variation and general intelligence
Introduction
Correlations between regional brain function and performance on mental tests associated with a general intelligence factor (g) as defined originally by Spearman (1904) have been demonstrated many times in normal subjects Duncan et al., 2000, Gray et al., 2003, Haier et al., 1988, Haier and Benbow, 1995, Parks et al., 1988, Prabhakaran et al., 1997. Most of these studies show that good test performance recruits areas distributed throughout the brain, although a case has been made that activation within areas of the frontal lobes is the primary source of differences in g-loaded test performance (Duncan et al., 2000). There is evidence that deactivation within some brain areas, including frontal lobes, is associated with better mental task performance Haier et al., 1988, Haier et al., 1992a, Parks et al., 1988, especially in subjects with higher intelligence test scores (Haier et al., 1992b). Even when a passive task with no inherent problem solving is used, subjects with higher intelligence scores show more activation in posterior information processing areas than subjects with lower scores Boivin et al., 1992, Haier et al., 2003b.
Functional brain imaging studies always must be interpreted to take account of the specific task demands of the mental task used during the imaging protocol. This makes inconsistencies among study results difficult to reconcile given the wide variety of tasks used. To the extent that individual differences in general intelligence have a structural component, examining structural correlates of intelligence would eliminate any task-related influences from consideration. For this reason, structural imaging of regional gray and white matter volumes would provide unique information about the distribution of brain areas related to general intelligence.
For example, total brain volume assessed by MRI in many studies has been shown to correlate about r = 0.40, with intelligence scores and total gray and white matter volumes also show small correlations with IQ (Gignac et al., 2003), but attempts to relate volume of specific brain areas to test scores have been mostly unsuccessful Flashman et al., 1997, MacLullich et al., 2002. Until now, such attempts have used various region-of-interest (ROI) methods that are difficult to reliably apply to many brain gyri when outlined by hand and often do not conform well to the extensive individual differences among subjects in brain size and morphology when applied stereotactically. A recent methodological advance is optimized voxel-based morphometry (VBM), which uses algorithms to segment gray matter (GM) and white matter (WM) from structural MRIs Ashburner and Friston, 2000, Good et al., 2001. VBM has been validated extensively Ashburner and Friston, 2001, Good et al., 2002 and it has been used, for example, to characterize gray and white matter volume changes in aging (Good et al., 2001), dementia (Good et al., 2002), and Down syndrome (White et al., 2003).
Section snippets
Subjects
We tested two samples and used a statistical conjunction approach (Price and Friston, 1997) to show where correlations between IQ and gray or white matter were common to both samples. The first sample was 23 normal volunteers (14 women and 9 men; mean age = 27, SD = 5.9, range = 18–37) recruited from the University of New Mexico (UNM). Sample 1 MRIs were obtained with a 1.5-T scanner, head coil, and software (Signa 5.4; General Electric Medical Systems, Waukesha, WI). A T1 sagittal localizer
Results
The conjunction results (N = 47; Fig. 1) showed robust positive correlations (P < 0.05, corrected for multiple comparisons) between FSIQ and gray matter volumes in Brodmann areas (BA) 10, 46, and 9 in frontal lobes; BA 21, 37, 22, and 42 in temporal lobes; BA 43 and 3 in parietal lobes; and BA 19 in the occipital lobe. The size and locations of these areas are shown in Table 1. Similar but less robust correlations with white matter areas (P ≤ 0.001, uncorrected) were also found (Table 1).
Discussion
These findings support the view that individual differences in gray and white matter volumes, in a relatively small number of areas distributed throughout the brain, account for considerable variance in individual differences in general intelligence. The locations of our strongest conjunction GM findings (P < 0.05, corrected for multiple comparisons) in the frontal lobes (BA 10, 46, and 9) are consistent with earlier functional imaging findings and reinforce the importance of frontal areas for
Acknowledgements
The UCI portion of this work was funded by a grant from NICHD to Dr. Haier (HD037427).
References (33)
- et al.
Voxel-based morphometry—The methods
NeuroImage
(2000) - et al.
Why voxel-based morphometry should be used
NeuroImage
(2001) - et al.
Verbal fluency and positron emission tomographic mapping of regional cerebral glucose-metabolism
Cortex
(1992) - et al.
Intelligence and regional brain volumes in normal controls
Intelligence
(1997) - et al.
Factors influencing the relationship between brain size and intelligence
- et al.
A voxel-based morphometric study of ageing in 465 normal adult human brains
NeuroImage
(2001) - et al.
Automatic differentiation of anatomical patterns in the human brain: validation with studies of degenerative dementias
NeuroImage
(2002) - et al.
Cortical glucose metabolic rate correlates of abstract reasoning and attention studied with positron emission tomography
Intelligence
(1988) - et al.
Regional glucose metabolic changes after learning a complex visuospatial/motor task: a positron emission tomographic study
Brain Res.
(1992) - et al.
Intelligence and changes in regional cerebral glucose metabolic rate following learning
Intelligence
(1992)