130 Am J Psychiatry 1 50:1 , January 1993
Intelligence and Brain Structure in Normal Individuals
Nancy C. Andreasen, M.D., Ph.D., Michael Flaum, M.D., Victor Swayze II, M.D.,
Daniel S. O’Leary, Ph.D., Randall Alliger, Ph.D., Gregg Cohen, M.S.,
James Ehrhardt, Ph.D., and William T.C. Yuh, M.D., M.S.E.E.
Objective: This study was designed to evaluate the relation between intelligence and a variety
ofmeasures ofbrain structure. Method: Magnetic resonance imaging scans were used to meas-
ure the volume of the intracranial cavity, cerebral hemispheres, lateral ventricles, temporal
lobes, hippocampus, caudate, and cerebellum, as well as the overall volume of gray matter,
white matter, and CSF, in 67 healthy, normal volunteers. Intelligence was measured with the
Wechsler Adult Intelligence Scale-Revised. Results: Full-scale IQ was found to be signifi-
cantly correlated with intracranial, cerebral, temporal lobe, hippocampal, and cerebellar vol-
ume but not with caudate and lateral ventricle volume. There were also significant correlations
offull-scale, verbal, and performance IQ with overall gray matter volume but not with white
matter or CSF volume. Gender differences were noted in the pattern and number of correla-
tions between the volume of the brain and its subregions and full-scale, verbal, and perform-
ance IQ. Conclusions: The results suggest that the size ofsome cerebral structures may account
f or a significant, but modest, proportion of the variance in human intelligence.
(Am J Psychiatry 1993; 150:130-1 34)
M odern neunoimaging techniques offer an unpar-
alleled opportunity to examine the relation be-
tween brain structure and brain function in normal mdi-
viduals. Because it is noninvasive and without significant
risk, magnetic resonance imaging (MR!) permits in vivo
quantitative measurements of brain size and the size of
subregions (e.g., the hippocampus) that are thought to
subserve known cognitive functions (e.g., memory).
Controversy has persisted for many years about whether
there are significant relationships between size and func-
tion in the human brain. Geschwind’s seminal work,
which indicates that the planum temponale is larger on
the left side of the brain, reflecting left hemispheric spe-
cialization for language, suggested that valid structure-
function relationships might exist (1 ). Subsequent reports
have suggested callosal enlargement in left-handed pen-
sons, reflecting a higher degree of intenhemisphenic trans-
fen as well as possible gender differences in brain structure
(2-4). While many studies have been done with post-mom-
tem tissue, more recently investigators have also used MRI
to conduct “in vivo autopsies” to study these relationships
in healthy, normal individuals, permitting the examina-
tion of structure-function relationships independent of the
Received March 24, 1 992; revision received July 14, 1 992; accepted
Aug. 5, 1992. From the Mental Health Clinical Research Center and
the Department of Radiology, University oflowa Hospitals and Clin-
ics, and the Department of Psychiatry, University of Iowa College of
Medicine. Address reprint requests to Dr. Andreasen, Mental Health
Clinical Research Center, University of Iowa Hospitals and Clinics,
200 Hawkins Dr., Iowa City, IA 52242.
Copyright © I 993 American Psychiatric Association.
possible epiphenomena produced by aging, illness, death,
on the fixation process (5-8).
The relation between intelligence and brain size has
been a particularly intriguing question. Several early
reports suggested possible relationships between cra-
nial size and intelligence or educational achievement,
but these reports have been questioned for a variety of
reasons (9-13). MRI has also been used to examine
relationships between cerebral size and educational
achievement, with the suggestion that a relationship
may occur (14, 15). A recent report indicated a high con-
relation (n=0.51 ) between !Q and cerebral size (16). This
particular report focused on college students, who
were designated as having a low IQ ( 1 03 on less, Wech-
sler full-scale scone) on a high IQ (130 or more, Wech-
sler full-scale score), thereby enhancing the probability of
finding positive correlations between brain size and !Q.
!n this article we report the finding of a similar nela-
tionship, also based on MRI, in a larger and more gen-
enalizable study group. In addition, we measured the
size of brain subregions such as the hippocampus and
assessed the total volume of gray matter in the brain,
which provides an indirect indicator of neunonal den-
sity and dendnitic expansion.
METHOD
The study group consisted of 67 healthy, normal vol-
unteens recruited through newspaper advertising. They
were screened with a structured interview to assess their