How is brain morphology related to ‘intelligence’?


It is usually believed that, in general, someone with a bigger brain would be more intelligent than someone with a smaller one. Assessing this belief from a scientific point of view is difficult as it relies on simultaneous measures of brain morphology and some form of intelligence in humans. Here, we review the literature on the subject and study in detail a particular data set. We show that there is a small, but significant correlation between brain morphology and intelligence.

1. Introduction

One of the greatest mysteries of human culture is the origin and location of intelligence within the brain. It has long been speculated that the frontal and temporal lobe regions have the closest association to intelligence[19], however there is no definite association despite interesting studies linking lesions in the frontal lobe and a loss of fluid intelligence[10]. The central question to be addressed here is whether there is a connection between intelligence and brain morphology. To do so, we will study correlations between a measure of intelligence and global brain features.

The human brain has an average weight of 1407 grams in men[23] and about 100 grams lighter in women[2]. It includes the following structures and regions:[5] The Cerebellum (meaning ‘little brain’) is responsible for the voluntary movements including posture, balance, coordination, speech, and smooth muscular activity.[17] The Cerebral Cortex is the outer layer of the brain’s most dominant part, the Cerebrum, and is where most of the higher cognitive functions are based. It is the bulging wrinkled surface we see when looking at the brain from any angle. With a thickness of a few millimetres, the cortex is composed of grey matter. Grey Matter refers to the unmyelinated neurons and other cells included in the central nervous system. Its name is derived from its colour (after staining) in contrast to the white matter. Grey matter is distinguished from white matter as it contains many cell bodies but relatively few myelinated axons (axons are the long tails of neuronal cells that connect different cells together)[5]. White Matter refers to the regions made up of mostly myelinated axons also known as tracts and is named after its high myelin whitish content (Fields 2008).

Scientists and philosophers have long struggled to come up with an appropriate definition and measure of intelligence[6]. For instance, a 2007 review[20] collected more than 70 different, and often contradictory, definitions and emphasised that intelligence is often falsely defined as a single identity. Indeed, there are many behaviours that can be linked to intelligence. For example, someone who appears ‘clever’ in academic challenges may not have the same mental capacity when it comes to more social or practical challenges. In both cases, one is seen to be using intelligence to solve problems. Therefore, one cannot use a single outstanding characteristic to define intelligence and we must consider multiple attributes[9].

In the 19th century, scientists believed that all mammalian brains (including the human brain) had the same overall composition.[17] They assumed that the number of neurons was proportional to the size of the brain. Assuming that cognitive abilities are proportional to the number of neurons, the largest brain should be the most cognitively able and hence the most intelligent. However, many other species, such as elephants and whales have larger brains, and, arguably, they do not show the same intelligence as humans. Therefore, from a comparative biology point of view, it is not clear at all that the naive idea of intelligence could be simply related to volume.

A typical measure of intelligence is through academic abilities. This is not because academic prowess is the most important intelligence trait or even the most important one, but because academic prowess, such as mathematical ability, are relatively easy to measure and compare due to their quantitative rather than qualitative nature. When it comes to measuring creativity or wisdom, we struggle due to a lack of a proper assessment scale. This is why psychologists have been mostly using Intelligence quotient (IQ) tests for the last hundred years. The goal of IQ tests is to offer a standardised measure of general intelligence. Many different tests have been used and created in the attempt to find the most accurate and consistent way of measuring someone’s intelligence.[9] Some of these tests include the Stanford-Binet Intelligence Scale, the Universal Nonverbal Intelligence, the Differential Ability Scales and the Wechsler Adult Intelligence Scale.[9] Each of these tests have their own way of measuring intelligence, whether it be through various knowledge questions or through problem solving quizzes.

Here, we will use a particular measure of intelligence and correlate it to simple, gross, features of the brain obtained from MRI.

2. Methods

2.1 Measure of intelligence

We will use the Wechsler Adult Intelligence Scale (WAIS). In particular, the WAIS-R, a revised version of the WAIS has emerged as a standard reference test in the field of psychology and is by far the most commonly used test in many of the ‘intelligence to brain’ studies as it balances many different attributes. Released in 1981 the WAIS-R originally consisted of five performance subtests and six verbal subtests. The idea of so many subtests was to provide a full coverage of the intelligence spectrum. Later versions of the WAIS-R include 13 different verbal and performance subtests.[9] The combination of these subtests allows for accurate and consistent Verbal and Performance IQs. In addition, the test provides a method of calculating a full-scale IQ.

2.2 Measure of morphology

In order to extract suitable structural data of the brain, we will use data from Medical resonance imaging (MRI), a standard imaging technique that creates a full three-dimensional image of tissue and water content in the brain. MRI scanners use strong magnetic fields, magnetic field gradients and radio waves to generate these images. The invention and development of the MRI in the 1980’s have revolutionised neurosciences by allowing researchers and clinicians to have a detailed and specific rendition of the brain and the skull in three dimensions in vivo.[18] Further processing of the data can be used to extract from each scan relevant morphological information such as: brain volume, surface area and cortical thickness.[38] Starting in the early 1990’s, MRI became the central tool for studies of brain morphology and function. Due to the rapid improvement in technology and the ability to collect and share large amounts of data, organisations have been created for the collection and sharing of patient data such as the UK Biobank.

2.3 Methods

First, we will consider previous studies that have used both MRI data and Intelligence score in order to compare them and try to establish the known consensus. Second, we will use another set of data

From the published paper ‘Association of Neurocognitive and Physical Function with Gait Speed in Midlife’[30] which was mainly focussed on the connection between gait speed and cognitive functions. The study gives data for N=905 people from Dunedin, New Zealand, at the same age (45 years). The data was extracted from the graphs using ‘WebPlotDigitiser’[33] and analysed in MS Excel to obtain the line of best fit and the correlation coefficients between any two data entries.

2.4 Best fit and correlation coefficients

Consider a data set with two variables (e.g. brain volume and IQ for a population). We are interested to see if these two variables are related to each other: does the increase of one systematically imply the increase, or decrease, of the other one? To determine whether there is such a relationship, we plot each data point on a graph, and find the best line going through this cloud of points; this is called the ‘line of best fit’ and provides a simple way to capture and visualise a trend.

The simplest and most reliable statistical measure is correlation. It tests the existence of a linear relationship between two variables. The correlation coefficient (r) represents the likeliness of a relationship holding between two variables in a specific situation. By construction, the range of values that a correlation coefficient can take is from -1 to 1.[9]

A correlation above 0 represents a positive relationship. A perfect correlation (with a coefficient of 1 or -1) only occurs when every single case is linearly related. Sizes of correlation can be misunderstood easily. Typically, a high absolute value of the correlation is around 0.7-0.9, however in many real-life situations, 0.5 or above it is considered to be a strong or large correlation, whereas between 0.2 and 0.5 is moderate or modest and anything between 0 and 0.2 is small or weak. The square of r gives the percentage of variability that can be explained by the trend. Finally, statistical significance tells you how unlikely it is for an event to occur given the null hypothesis. This requires the ‘t-test’ to be measured which provides a p value. Since significance can vary depending on variables such as sample size, it is not easy to correctly interpret an r value as a high correlation may not be statistically significant (for instance 2 data points always give |r|=1 but are never statistically significant).

3 Results

3.1 Literature survey

Before analysing a specific data set, we considered previous studies that explicitly linked intelligence with brain volume and reported their correlation. The studies were found systematically by using the key words ‘intelligence, brain volume, correlation’ on google scholar. In Table 1, MRI studies with their correlation coefficients are reported together with the test used, as well as any other particular feature of the study. Despite early works,[35] systematic studies of brain volume and intelligence began with Willerman et al. in 1991, the first study that used MRI.

Table 1. Intelligence-Volume correlations. This summary of the relevant recent studies gives the sample size N, the number of k different data sets (in the case of a meta-analysis), the type of test used and the correlation coefficient from a linear fit between intelligence and brain volume (see Methods). For the exact citation of each study see the reference section where each study can be found from the names of the authors and the year of publication.









Willerman et al.[37]



Right-handed Anglo psychology students


Wechsler (1981)


Andreasen et al.[1]



Adult healthy


Wechsler (1981)


Raz et al.[31]



Age (17-78)


Cattell’s Culture Fair Intelligence Test


Harvey et al.[15]



Ages (19-49)


National Adult Reading Test


Wickett et al.[36]




Ages (20-40)


Multidimensional Aptitude Battery test


Egan et al.[11]



Soldiers early 20s (38 male)


Wechsler (1981)


Rushton and Ankney[34]



MRI and X-ray


Various tests


Paradiso et al.[26]



Adult healthy


Wechsler (1981)


Gignac et al.[13]














Pietschnig et al.[28]



Largest meta-analysis


Wechsler (1981)


Nave et al.[24]



TBV (total brain volume) to fluid intelligence

UK Biobank


Fluid Intelligence

This table shows that the first few studies seem to corroborate the commonly held belief that intelligence is closely related to brain size. Indeed, the correlation coefficient of 0.51 suggests a relatively strong correlation between the two variables (see also Gignac and Bates[14]). However, as demonstrated in Figure 1, further studies have resulted in a greatly lowered coefficient.

Figure 1. Brain Volume-IQ correlation over the years shows a systematic decline.

A possible reason for this decrease is publication bias (early studies were only published if they showed agreement with previous studies, hence many data sets might have been discarded) or, perhaps, more simply, larger sample size due to availability of larger data sets as shown in Figure 4. Other factors that could have an effect on correlation are population selection, location and gender. However this information is not typically available in all studies.

Figure 2. Over the years, improvements in MRI technology has allowed for a large increase in sample size.

In order to test for the consistency of the results and study general trends, we compute directly the correlation coefficient based on a single large uniform data set.

3.2 Analysis of the Dunedin data

As an example of the data used, Figure 3 gives the data of brain volume (in cubic mm) versus Intelligence Score on the WAIS test for the Dunedin data set. The line of best fit for the squared correlation are also given.

Figure 3. Brain volume (in liters) vs IQ. Data given by blue points and line of best fit (red) indicates a correlation r=0.1889

The same analysis process was repeated for other measures of IQ against either brain morphology or ‘Mean Cortical Thickness’ and ‘Surface Area’. The results are shown in Figures 4 and 5 and in each case, the r value was computed and collated in Table 2.

Figure 4. Cortical thickness vs IQ for the Dunedin data (blue points). The line of best fit (red) indicates a modest correlation.

Figure 5. Surface area (10^3 mm^2) vs IQ for the Dunedin data (blue points) and the line of best fit (red) with a slightly better correlation of r=0.1995 than the previous two figures.

Table 2. Correlation values extracted from the Dunedin studies.


Computed Correlation r

Brain Volume VS IQ


Mean Cortical Thickness VS IQ


Surface Area VS IQ


4. Discussion

4.1 Results

Our analysis of the data indicates that both brain volume and surface area have a higher correlation to the measure of intelligence used than cortical thickness. The similarity between the correlations of brain volume and surface area is expected. Indeed, brain volume itself is very closely related to surface area and has a correlation of 0.94. From these results, we conclude that average cortical thickness is poorly correlated to intelligence. This result suggests that average cortical thickness may have less of an impact on an individual’s cognitive abilities in comparison to brain volume and surface area. This may be due to the fact there is very little variation in thickness of the cortex and mean cortical thickness is a stable property that would not change in regard to intelligence. However, brain volume and surface area are familiar global markers that appear to be better correlated with intelligence.

4.2 Possible mechanisms

The causality between intelligence and brain morphology is not currently established. While it seems reasonable that a greater brain volume would lead to more processing power and hence better cognitive ability, it could also be the case that higher intelligence leads to more brain activity that in turn leads to the development and remodelling of the brain which may increase local brain volume in specific regions. This relationship has been documented in brain studies of mathematicians, musicians and taxi drivers.[3,29]

Assuming that bigger brains leads to higher intelligence,[9] here is a summary, extracted from the literature, of five main proposed mechanisms:

Willerman et al. claim that ‘larger size might reflect more cortical columns’, in turn this would imply a better ability to process information connected to IQ tests.[37] Since this was the first MRI study, it is a good reference point to show how brains were originally believed to be linked to intelligence. The main idea around Willerman’s discussion is that brain size indirectly has an overall impact on an individual’s intelligence. The basic idea is that an increased quantity in all physical components of the brain such as ‘mitotic divisions’, ‘cortical columns’ and ‘stem cells’ allows for a more efficient processing power to increase rate of performance on IQ tests. This was the fundamental idea around the link between intelligence and the brain.

In contradiction, Andreasen et al.[1] believe that it is ‘complexity of circuitry, dendritic expansion, number of synapses, thickness of myelin, metabolic efficiency, or efficiency of neurotransmitter production, release and reuptake’ that influence the quality of the response and that ‘The greater volume of grey matter can be postulated to reflect a greater number of nerve cell bodies and dendritic expansion; a greater number of neuronal connections presumably enhances the efficiency of the computational processing in the brain.’ Therefore, Andreasen believes that it is the quality of features, such as brain architecture, that matters not their quantity. The idea was that certain physical functions were enhanced by the effectiveness of more complex or detailed components of the brain therefore causing a greater cognitive function and processing power. His inclusion of myelination is also interesting as it depends on nutrition during brain development and is directly related to the speed of information processing.

Another view was expressed by Raz et al.[31] who emphasise the role of brain asymmetry in the development of intelligence as they claim that ‘leftward volume asymmetry may reflect either a greater number of processing elements or more extensive connectivity in the left hemisphere.’

According to Wickett et al.[36]The brain size-IQ correlation of r = 0.395 clearly indicates that either there are many more variables to be introduced in an attempt to explain intelligence’. However, they suggest that perhaps certain aspects of the brain may be relevant to intelligence, for example neuronal quantity or myelination.[8]

Egan et al.[11] believe ‘that small differences in brain volume translate into millions of excess neurons for some individuals, accounting for their higher IQ.’ Therefore, an increase in neurons due to a greater brain volume relates directly to an increased intelligence. However, there is no direct evidence for an increase in neuron number with volume increase. Furthermore, comparative studies between primates and birds show that, in animals, a vastly different brain size does not translate in additional neurons. Certain birds have very similar numbers of neurons to primates of a much larger brain size.[25] According to Herculano-Houzel et al.[16] the key is that all brains are not in fact made the same way. For example, primate brains always have more neurons than a rodent brain of the same size. With an average number of neurons of about 86 billion, the human brain does not have the largest number of neurons. For instance, the African elephant has on average 257 billion neurons.[16] However, the human brain has 16 billion neurons located in the cerebral cortex alone. This is the largest number of neurons in any mammalian cortex. This study suggests that the number of cortical neurons may be the main seat of intelligence in the human brain.[17]

Finally, the evolutionary explanation asserts, without a specific mechanism, that brain size is a major contributing factor to intelligence between different species. The problem is that measuring intelligence in animals is much harder than for humans. However, well studied activities in certain primates such as self-control show that an increase in brain size correlates with their cognitive function as demonstrated in two studies, MacLean, E. L. et al.[21] and Benson-Amram et al.[4] MacLean’s paper concludes: ‘These results suggest that increases in absolute brain size provided the biological foundation for evolutionary increases in self-control’.

Beyond the morphological features, it has long been accepted that high intelligence IQ test scores also greatly depend on the level of education and training.[27]

5. Conclusion

There are three main conclusions that we can draw from this analysis:


  1. There is a correlation between brain morphology and intelligence at the population level. On average, at the population level, certain physical features of the brain explain the variability of intelligence.
  2. The correlation coefficients are too small to be used as a predictive tool on an individual basis.
  3. There is no well-established mechanism that explains an increase in intelligence with an increased physical factor of the brain

The overall conclusion is that brain morphology does have an impact on intelligence. Various analyses of data sets, including the one carried out in this work, show conclusively that there is a significant correlation between brain morphology and intelligence. However, this correlation is very small, and much smaller than what earlier studies suggested. This leads us to conclude that while the morphology of the brain is a contributing factor to an individual’s intelligence, there are many other non-morphological factors which determine complete intelligence. Examples of this can be nurture, brain architecture, nutrition, education. The low-correlation coefficient also indicates that brain volume cannot be used as a valid predictor of intelligence. While the trend exists at a population level, no meaningful comparison can be made between two individuals, as noted already by Charles Darwin in the Descent of Man:[7]No one supposes that the intellect of any two animals or of any two men can be accurately gauged by the cubic contents of their skulls.”


  1. Andreasen, N., M Flaum, V. Swayze, D.S. O’Leary, R Alliger, G Cohen, J Ehrhardt and W. T. Yuh. 1993. “Intelligence and brain structure in normal individuals.” American Journal of Psychiatry, 150, 130–134.
  2. Ankney, C. Davidson. 1992. “Sex differences in relative brain size: The mismeasure of woman, too?” Intelligence, 16, 392–336
  3. Aydin, K, A Ucar, K. K. Oguz, O. O. Okur, A. Agayev, Z. Unal, S. Yilmaz and C. Oztuk. 2007. “Increased Gray Matter Density in the Parietal Cortex of Mathematicians: A Voxel-Based Morphometry Study.” AJNR Am J Neuroradiol, 28, 1859–1864.
  4. Benson-Amram, S., B. Dantzer, G. Stricker, E. M. Swanson and K. E. Holekamp, 2016. “Brain size predicts problem-solving ability in mammalian carnivores.” Proceedings of the National Academy of Sciences, 113, 2532–2537.
  5. Carter, R. 2019. The Brain Book: An Illustrated Guide to its Structure, Functions, and Disorders. London: DK.
  6. Chollet F. 2019. “The Measure of Intelligence.” preprint arXiv:1911.01547.
  7. Darwin, Ch. 1875. The Descent of Man, and Selection in Relation to Sex. London: John Murray Pub.
  8. Deary, I.J. 2000. Looking Down on Intelligence. Oxford: Oxford University Press. 255.
  9. Deary, I.J. 2001. Intelligence: A Very Short Introduction. Oxford: Oxford University Press.
  10. Duncan, J., Burgess, P., Emslie, H. 1995. “Fluid intelligence after frontal lobe lesions.” Neuropsychologia, 33, 261–268.
  11. Egan, V., A. Chiswick, C. Santosh, K. Naidu, J. E. Rimmington and J. J. K. Best. 1994. “Size isn’t everything: a study of brain volume, intelligence and auditory evoked potentials.” Personality and Individual Differences, 17, 357–367.
  12. Fields, R. D. 2008. “White matter in learning, cognition and psychiatric disorders.” Trends in Neurosciences, 31, 361–370.
  13. Gignac, G., P. A. Vernon and J. C. Wickett. 2003. “Factors influencing the relationship between brain size and intelligence.” In The scientific study of general intelligence, edited by H Nyborg 93–106. Amsterdam: Pergamon.
  14. Gignac, G. E. and T. C. Bates. 2017. “Brain volume and intelligence: The moderating role of the intelligence measurement quality.” Intelligence, 64, 18–29.
  15. Harvey, I., R. Persaud, M. A. Ron, G. Baker and R. M. Murray (1994). “Volumetric MRI measurements in bipolarism compared with schizophrenics and healthy controls.” Psychological Medicine, 24, 689–699.
  16. Herculano-Houzel, S., Kamilla Avelino-de-Souza, Kleber Neves, Jairo Porfirio, Débora Messeder, Larissa Mattos Feijó, José Maldonado, and Paul R. Manger. 2014. “The elephant brain in numbers.” Frontiers in Neuroanatomy, 8, 46.
  17. Herculano-Houzel, S. 2016. The human advantage. Cambridge: MIT Press.
  18. Jordan, D. 2020. “State of the art in magnetic resonance imaging.” Physics Today, 73, 34-40.
  19. Jung, R. E., and R.J. Haier. 2007. “The Parieto-Frontal Integration Theory (P-FIT) of intelligence: converging neuroimaging evidence.” Behavioral and Brain Sciences, 30, 135–154.
  20. Legg S. and M. Hutter. 2007. “Universal intelligence: A definition of machine intelligence.” Minds and machines, 17(4), 391–444.
  21. MacLean, E. L , Brian Hare, Charles L. Nunn, Elsa Addessi, Federica Amici, Rindy C. Anderson, Filippo Aureli, Joseph M. Baker, Amanda E. Bania, Allison M. Barnard, Neeltje J. Boogert, et al. 2014. “The evolution of self-control.” Proceedings of the National Academy of Sciences USA, 111, E2140–E2148.
  22. McDaniel, M. A. 2005. “Big-brained people are smarter: A meta-analysis of the relationship between in vivo brain volume and intelligence.” Intelligence, 33, 337–346.
  23. Molina, D. K. and V. DiMaio. 2012. “Normal Organ Weights in Men: part ii- the brain, lungs, liver, spleen and kidneys.” The American Journal of forensic medicine and pathology, 33, 368–372.
  24. Nave, G., W. H. Jung, R. K. Linner, J. W. Kable and P.D. Koellinger. 2019. “Are bigger brains smarts? Evidence from a large-scale reregistered study.” Psychological science, 30, 42–54.
  25. Olkowicz, S., Kocourek, M.,  Lučan, R. C., Porteš, M., Fitch, W.T., Herculano-Houzel, S. and  Němec, P., 2016. “Birds have primate-like numbers of neurons in the forebrain.” Proceedings of the National Academy of Sciences USA, 113, 7255–7260.
  26. Paradiso, S., N.C. Andreasen, D. S. O’Leary, S. Arndt, and Robinson, R. G. 1997. “Cerebellar size and cognition: correlations with IQ, verbal memory and motor dexterity.” Neuropsychiatry, Neuropsychology and Behavioural Neurology, 10, 1–8.
  27. Pearson, K. 1906 “On the Relationship of Intelligence to Size and Shape of Head, and to Other Physical and Mental Characters.” Biometrika, 5, No. 1/2 (Oct): 105-146.
  28. Pietschnig, J., L. Penke, J. M. Wicherts, and M. Zeiler. 2015. “Meta-analysis of associations between human brain volume and intelligence differences: How strong are they and what do they mean?” Neuroscience and Biobehavioural Reviews, 57, 411–432.
  29. Posthuma, D., E. De Geus, W. F. Baare, H. E. Hulshoff Pol, R. S Kahn, and D. I. Boomsma. 2002. “The association between brain volume and intelligence is of genetic origin.” Nature Neuroscience, 5, 83.
  30. Rassmusen L. J. H., A. Caspi, A. Ambler , J M. Broadbent, H J Cohen, T d’Arbeloff, M Elliot, R J. Hancox, H. Harrington, S. Hogan, et al. 2019. “Association of Neurocognitive and Physical Function With Gait Speed in Midlife.” Jama Network Open, 2, e1913123–e1913123.
  31. Raz, N., I. T. Torres, W. D. Spencer, D. Millman, J. C. Baertschi, and G. Sarpel. 1993. “Neuroanatomical correlates of age-sensitive and age-invariant cognitive abilities: an in vivo MRI investigation.” Intelligence, 17, 407–422.
  32. Richardson M. W. 2019. “How Much Energy does the Brain Use?”
  33. Rohatgi, A, WebPlotDigitizer, Version: 4.2, April, 2019.
  34. Rushton, J. P. and C. D. Ankney. 1996. “Brain size and cognitive ability: correlations with age, sex, social class, and race.” Psychonomic Bulletin and Review, 3, 21–36.
  35. Stott, D. H. 1983. “Brain size and ‘intelligence’.“ British Journal of Developmental Psychology, 1, 279–287.
  36. Wickett, J. C., P. A. Vernon, and D. H. Lee. 1994. “In vivo brain size, head perimeter, and intelligence in a sample of healthy adult females.” Personality and Individual Differences, 16, 831–838.
  37. Willerman, L., R. Schultz, J. N. Rutledge, and E. D. Bigler. 1991. “In Vivo Brain Size and Intelligence.” Intelligence, 15, 223–228.
  38. Whitwell, J. L., W. R. Crum, H. C. Warr, and N. C. Fox. 2001. “Normalization of Cerebral Volumes by Use of Intracranial Volume: Implications for Longitudinal Quantitative MR Imaging.” American Journal of Neuroradiology, 22, 1483–1489.

About The Author

Zéphyr Goriely lives in Oxford, United Kingdom and attends the Cherwell School. Born in the US, he moved to the UK when he was six years old.
He loves biology and plans to study it at university. When not focussed on academics, he takes pride in playing volleyball where he captained the England National team and plays for his national champion club.

Leave a Reply

Your email address will not be published. Required fields are marked *