EEG in national team archers: Zhu 2021 study

Thirty-six archers from the Chinese national team, 22.7 years old on average and with 9.2 years of training behind them, had their electroencephalogram recorded an hour after training, across three phases of winter preparation with different loads. Thirty-six university students served as a comparison group.

The three phases are described as follows: low load and low intensity, with technique and general physical preparation and no more than two hundred real arrows a day without scoring; high load and low intensity, with five hundred to eight hundred real arrows a day and scoring; and a third combination, low load and high intensity, which the article names but does not define.

In the high-load phase the power of the alpha band in the parieto-occipital region is lower: in O1 from 162.77 to 117.78 square microvolts, in O2 from 168.46 to 115.27, in P3 from 88.32 to 56.45, in P4 from 82.44 to 55.16, returning to baseline values in the third phase. In the beta band the frontal values rise in the third phase, from 23.37 to 40.85 in F3 and from 22.15 to 42.29 in F4.

Using a technique that estimates six neurotransmitters from ultra-slow fluctuations in the electroencephalogram, the author reports a rise in serotonin (from minus 3.96 to 7.96 in the high-load phase and to 6.17 in the third) and two opposite values for dopamine: 19.39 in one table and 1.63 in the other, from a baseline of 10.74. He reads the whole as central fatigue linked to arrow volume, and concludes that recovery should be planned as part of training.

The instrument is a Chinese-made Solar 1848 electroencephalograph, with silver electrodes placed according to the international ten twenty standard. Recording took place an hour after the end of the training session. The analysis of ultra-slow fluctuations, which the article calls SET, was done with a system developed by a single research group, and rests on the idea that particular fluctuation frequencies correspond to individual neurotransmitters: an inhibitor, serotonin, acetylcholine, dopamine, noradrenaline and an excitatory mediator.

Two sets of indices are reported. The power in the alpha and beta bands, measured in square microvolts across sixteen derivations, from frontal to occipital. And the estimate of the six neurotransmitters across the whole brain, compared between the low-load phase and the other two.

The comparisons are made with multiple analyses between phases and t tests on the means before and after the special training session in the main phases, in SPSS 14.0. The control group appears only once, in the comparison of the whole-brain neurotransmitter distribution between archers and non-athletes in the low-load phase, where no difference emerges; in the load phases the archers are compared only with themselves.

For the two hundred, five hundred and eight hundred shots a day the article does not say how they were spread through the day or how many weeks each phase lasted, and it reports no measure of performance: there are no scores, no fatigue questionnaires, no sleep data and no blood values.

Neither of the two tables of electroencephalographic power contains a single asterisk, although each carries the note defining asterisks as marking a significant difference from the low-load phase. The fall in parieto-occipital alpha, which is the main finding, is therefore described as significant in the text without any comparison being marked in the tables, and the standard deviations are as large as the means: 117.78 plus or minus 91.32 in O1.

The statistical picture does not hold. In the two neurotransmitter tables the F ratios and the p values are mutually incompatible: F 0.363 with p = 0.000, F 1.689 with p = 0.150, F 0.599 with p = 0.059. The two tables also carry the same column heading although they are meant to describe different comparisons, and dopamine rises to 19.39 in one and falls to 1.63 in the other, while the discussion states that in the high-load phase it is lower than in the low-load phase. The third load phase is not described in the methods section, which defines two.

On the method for estimating neurotransmitters it should be said plainly that this is not a chemical measurement: it is an inference from an electrical signal, with a technique used in a narrow literature, and the frequency band given in the article, 1 to 255 megahertz, is not an electroencephalographic band. The group of students serves a single comparison. The link between serotonin and central fatigue is taken from studies on long-distance running and swimming, not measured here.

Above all, nothing about the archer is measured: no score, no accuracy, no subjective scale. The conclusion that the national team's load sits in a reasonable range rests on no performance outcome at all. The article runs to three pages, was received on 29 July 2021 and accepted on 18 August, and the author contribution statement mentions performing surgeries, which has no bearing on this study.

The most solid thing in this source is not neurological but organisational: it describes what it meant, in the winter preparation of a national team, to move from a technical phase with no more than two hundred arrows a day and no scoring to a volume phase with five hundred to eight hundred arrows a day with scoring. Anyone who shoots finds there a measure of what a loading period looks like at national level.

On the rest the source will not bear the weight one might want to put on it. The idea that a high volume of arrows leaves a trace of central fatigue detectable on the electroencephalogram remains the author's reading, resting on tables in which the differences are not marked, the statistics do not add up and dopamine moves in two opposite directions. There is no score anywhere, so it cannot even be said whether shooting was worse in the volume phase. In the registry the source counts as an early observation on an idea widespread among coaches, to be looked for again in better data.