Brain and visual memory in elite archers: Seo 2012

Twenty high-level Korean women archers and 23 women who had never competed in archery did the same task inside a functional MRI scanner: they looked at a column of 11 tilted lines and counted how many had the same slope as a reference line. It is a test of visuospatial working memory with nothing to do with shooting, chosen precisely to see whether years of competition leave a mark outside the sport.

On the answers themselves the two groups did not differ: the archers were correct on 47.91% of trials and the others on 40.62% (P = 0.29), with mean response times of about 8.4 seconds against 9.1 (P = 0.23). The task was hard for everyone, with fewer than half the answers correct.

What differed was how the brain went about it. Both groups used the parietal and frontal areas typical of this kind of task; the archers showed more activity in the middle frontal gyrus, the supplementary motor area and the dorsolateral prefrontal cortex, and a larger drop, compared with the rest condition, in parts of the so-called default mode network (the paracentral cortex and precuneus, and the anterior and posterior cingulate). The more the frontal areas rose, the more the precuneus fell, and the percentage of correct answers went up with middle frontal activation and with precuneus deactivation. The authors read this as a different strategy for the same task, not as a greater ability.

The archers belonged to the Korean University Archery Association, and 10 of them were also on the national team; they had been shooting for more than ten years (15 on average) and held official scores from international competitions. The comparison group was made up of women who had never taken part in a competition. Age (28.9 and 26.3 years) and years of education (15.5 and 15.4) were similar, and all were right-handed.

In the scanner (3 tesla) the task was split into blocks: four rest blocks and four task blocks of 30 seconds each, about four minutes in all. At rest the participant only had to confirm that 7 vertical lines matched the reference; in the active blocks she counted, among 11 lines tilted 15 to 75 degrees to the left or right, those matching the reference, and answered with one of four buttons. Images were processed in SPM5. Within each group the maps were corrected for multiple comparisons over the entire brain; the comparison between groups, on the other hand, was run inside regions chosen in advance (for activation, the dorsolateral prefrontal cortex, the middle frontal gyrus, the precuneus and the supplementary motor area), with a correction limited to those areas. The links between areas, and between areas and answers, are Pearson correlations on percentage signal change.

None of the differences between the groups survives correction for multiple comparisons across the whole brain: the authors say so themselves, and the differences appear only inside the regions chosen before the analysis. The sample is small and made up only of Korean women, and the task lasts a few minutes. The fact that the scores did not differ significantly does not mean the abilities are the same: the archers were about 7 percentage points ahead, and with 20 and 23 people a gap of that size can slip past the test.

The study is cross-sectional, so it cannot say whether the differences come from years of shooting or from traits the archers already had before they started. Reading the pattern as a different strategy is also an interpretation: the authors set aside the opposite reading, that more activation signals more effort, by argument rather than by measurement. The paper calls the comparison group "novices", but they were women who had never competed, not beginner archers; the correlation coefficients appear only in the figures, and the paper does not say whether they were computed on both groups together. Nothing was measured on the shot.

For anyone who shoots, the study changes nothing concrete: it does not measure aiming, scores or any part of technique, and it does not show that archery training improves visuospatial memory. What it says is narrower: archers with many years of competition, faced with a spatial task unrelated to shooting, get roughly the same results as people who do not shoot, but they get there with different brain activity, more concentrated in the frontal attention areas and with a sharper drop in the areas that are busy at rest.

In the registry it counts as one piece of the debate about experts: in this study years of competition gave no measurable edge on a basic cognitive task, even though the brain went about it differently. There are no mental training exercises to be drawn from it, and the claims about "the archer's brain" sometimes made on its back rest on 43 people and on differences that emerged only in regions chosen in advance.