This study sought to assess whether level of education is related to regional brain metabolism during a sustained attention task. PET scans of individuals with more years of education demonstrated greater relative activity in the lingual gyrus, precuneus and left posterior cingulate. Because CPT error rate did not correlate with years of education or with relative regional activity, the present results are unlikely to be due to performance inconsistencies alone. Likewise, it is less probable that the associations identified are merely general intelligence effects because WAIS-R Vocabulary scores did not correlate with years of education or with relative regional activity.
Supratentorial cortical neurodevelopment appears to progress in a posterior to anterior direction, with the occipital cortex acquiring myelin prior to the frontal cortex.[14
] The medial posterior parietal and occipital cortexes, such as the precuneus, develop early and should be fully formed by the ages of our subjects.[15
] By identifying an association between education, an environmental variable, and regional metabolic activity in these areas, the present data are consistent with the notion of experience-driven regional neuroplasticity continuing to modulate brain function in the healthy adult brain. However, our study design cannot formally support a causal relationship between educational experience and the observed cerebral metabolic activity. Furthermore, the precise significance of activity on positron emission tomography in this context cannot be ascertained from the current data (e.g., whether this represents intrinsic synaptic rewiring in this area or simply increased recruitment of existing circuitry).
The observed areas of education-associated cerebral metabolism in our analysis all subserve advanced visual language processing. Both the precuneus and lingual gyrus respond to word as opposed to non-word visual stimuli, a response that is delayed but persistent when the presented stimuli are rotated.[16
] The left precuneus shows increased activation when reading high- rather than low-frequency Chinese characters, and the lingual cortex has the reverse preference.[17
] It is reasonable to posit that this augmented activity of the lingual gyrus, precuneus, and posterior cingulate in concert may be associated with the particular linguistic demands of advanced education. Additionally, the auditory nature of the continuous performance task may have contributed to highlight these regions.
Of great interest, the current results may help begin to provide a hypothetical rationale for education’s possible protective effect on cognition. The precuneus and neighboring posterior cingulate cortex occupy a vulnerable site in the medial parietal lobe on the border of the anterior and posterior cerebral arterial territory. They are tonically active during baseline wakefulness, and their metabolism correlates directly and dramatically with level of consciousness.[18
] They suffer significant decline in functioning with diminished consciousness and coma [19
] and early loss of activity in Alzheimer’s dementia.[20
] Both the precuneus and lingual gyrus (medial occipitotemporal gyrus) suffer age-related loss of volume bilaterally in non-demented adults with Down’s syndrome, a population prone to Alzheimer’s disease.[21
] Presumptively, more robust neuronal circuits in these structures may prove advantageous in the face of Alzheimer’s dementia or similar insult.
The lack of education-associated prefrontal changes was unexpected. Mathematical tasks recruit frontal and prefrontal regions in addition to the precuneus and left cingulate cortex.[9
] Furthermore, both the dorsolateral prefrontal cortex and precuneus are involved in pitch discrimination tasks.[22
] The lack of anterior findings could suggest that advanced educational experience may be associated more robustly with modifications in analysis of incoming visual and linguistic data than with executive processing. Alternatively, perhaps those with more mathematically oriented educational experience do have altered prefrontal activation, an effect that could have been masked by the small and heterogeneous sample population. These of course remain speculative, and additional studies with both functional imaging and neuropsychological testing data to address such possibilities are needed.
There are several limitations to this study. The small sample size and consequent narrow range of subjects’ years of education, for instance, posed a constraint on the analysis. However, the fact that data analysis nonetheless revealed statistically significant and bilateral findings across this narrow educational range we feel strengthens the results. Additionally, the WAIS-R Vocabulary Test data may not have been a sufficiently sensitive measurement of intelligence in such a small number of subjects. However, this test is frequently utilized as a test of general intelligence, and we included it to help rule out the possibility that years of education was not a mere surrogate variable for preexisting, innate intelligence that might predispose a subject to advance farther in schooling.[24
] Importantly, we cannot eliminate the possibility that our data are confounded by untested variables, such as nutritional status, socioeconomic status, or exercise, which might relate to both education and relative regional brain metabolism. Future work is needed to more thoroughly examine the effects of such confounding variables with a number of instruments and on a greater number of subjects.