Structural brain imaging correlates of general intelligence in UK Biobank

Abstract

The associations between indices of brain structure and measured intelligence are unclear. This is partly because the evidence to-date comes from mostly small and heterogeneous studies. Here, we report brain structure-intelligence associations on a large sample from the UK Biobank study. The overall N = 29,004, with N = 18,426 participants providing both brain MRI and at least one cognitive test, and a complete four-test battery with MRI data available in a minimum N = 7201, depending upon the MRI measure. Participants' age range was 44-81 years (M = 63.13, SD = 7.48). A general factor of intelligence (g) was derived from four varied cognitive tests, accounting for one third of the variance in the cognitive test scores. The association between (age- and sex- corrected) total brain volume and a latent factor of general intelligence is r = 0.276, 95% C.I. = [0.252, 0.300]. A model that incorporated multiple global measures of grey and white matter macro- and microstructure accounted for more than double the g variance in older participants compared to those in middle-age (13.6% and 5. 4%, respectively). There were no sex differences in the magnitude of associations between g and total brain volume or other global aspects of brain structure. The largest brain regional correlates of g were volumes of the insula, frontal, anterior/superior and medial temporal, posterior and paracingulate, lateral occipital cortices, thalamic volume, and the white matter microstructure of thalamic and association fibres, and of the forceps minor. Many of these regions exhibited unique contributions to intelligence, and showed highly stable out of sample prediction.

Keywords: Brain; Cortex; Intelligence; Subcortical; White matter.

Fig. 1

Indicative overlap between initial cognitive measures in the imaging visit (VNR; verbal numerical reasoning), MRI measures, and the Enhanced Cognitive Battery (Matrix Reasoning, Symbol-Digit and Trail-Making Part B) among the 29,004 participants included in the current analysis. For ease of illustration, the MRI numbers are based on grey matter volume (highest N among global MRI measures), and the Enhanced Battery numbers are based upon Trail Making Part B (highest N among the Enhanced Battery). A total of 18,426 have MRI and at least one cognitive test. There are slight variations in missingness among Enhanced Battery and MRI measures (see Table 1).

Fig. 2

Brain imaging regions of interest according to the Harvard-Oxford Atlas (top: cortical regions) and AutoPtx (bottom: white matter tracts; adapated from Cox et al., 2019).

Fig. 3

Associations between global brain MRI measures and g. Panel a) shows associations with total brain volume, and panel b) shows tissue-specific brain MRI measures accounting for 6.54% of the variance in g. Standardised estimates are reported; grey dashed paths are non-significant. Indicators are all corrected for age, sex, with imaging data also corrected for scanner head position coordinates. MRI residual correlations are shown in Table S4.

Fig. 4

Associations between white matter tract-specific microstructure and g. Tabulated results also reported in Tables S6 and S7. Top panel shows left and right lateral and superior views of the white matter tracts of interest, heatmapped according to association magnitude. Lower panel displays the association magnitudes sorted by tract class and then from strongest to weakest (based on the average of FA and MD), with 95% CIs; MD valences have been flipped to aid visual comparison.

Fig. 5

Associations between regional cortical volumes and g with 95% CIs. Left and right associations are shown separately (left hand regions appear first). Association magnitudes are also reported in Table S5.

Fig. 6

Out of sample prediction (Manchester to Newcastle) of g from regional MRI data. Left panel shows the spatial distribution of standardised beta weights. Right panel shows the associations (standardised estimates and 95% confidence intervals) between g and the weighted composite scores (derived using those weights) in the training and test samples.