Thigh EMG in archers: Zanevskyy 2021 study

The study uses surface electromyography to measure the activity of rectus femoris, the muscle at the front of the thigh that flexes the hip, in 20 skilled male archers aged 23 to 28. Standing on a mechanical rig, each pushed as hard as possible to flex the thigh without moving it, at six hip positions: -15, 0, 20, 40, 50 and 70 degrees.

Mean peak activity was 6.59 mV at -15 degrees and 6.69 mV at 0, then fell to 5.69, 4.71 and 3.12 mV, down to 2.69 mV at 70 degrees: with the hip more flexed the muscle gave about two and a half times less signal. Differences between archers were large (at 0 degrees the range was 2.80 to 10.70 mV), with a standard deviation of about a quarter of the mean at every angle. By passing a curve through the six means, the authors place the maximum, 7.15 mV, at -8.9 degrees and the minimum, 1.44 mV, at 62.8 degrees. In the breakdown of variance the angle accounts for 60.9%, differences between people for 15.8% and the remainder for 23.3%.

For reliability the test was repeated five times at 16.2 degrees, a few minutes apart. The signal dropped from one trial to the next (5.87, 5.80, 5.62, 5.25 and 4.64 mV). On the raw data the reliability of a single trial was 0.698 and that of the mean of five 0.921; after removing this decline by calculation, the authors obtain 0.745 for one trial and 0.936 for the mean.

The instrument is an M-Test electroneuromyograph with adhesive Ag/AgCl electrodes: the active electrode sat over the muscle's innervation zone, the reference electrode towards the proximal tendon and the ground on the other leg. The peak potential during the effort was recorded in millivolts, without normalising it to a reference contraction. Testing took place in the afternoon, on the dominant leg, using a mechanical rig described in an earlier paper by the same group; the force produced is not reported.

To describe how activity changes with angle the authors fit a fifth-degree polynomial through the six means and derive a maximum, a minimum and an average angle (16.2 degrees), where the five repetitions were then done. Reliability is estimated with the one-way intraclass correlation coefficient, ICC(1,1) for one trial and ICC(1,k) for the mean of five. The decline across trials is described by a straight line (slope -0.3004 mV per trial, p = 0.017) and subtracted from the data before the indices are recalculated. For the archers we know age, height (179 ± 2.6 cm) and body mass (72 ± 1.9 kg): the discipline is not given and the level is described only as skilled.

The main limit is that shooting plays no part. The study measures a hip muscle in a maximal static effort, whereas on the line the legs do light, sustained work to hold posture; the link with an archer's stability is in the introduction, not in the data. The word archery in the title says only who the participants were, and the authors themselves write that the method applies to any population.

There are also problems of measurement and calculation. A fifth-degree polynomial through six points passes exactly through all of them, so the maximum at -8.9 degrees and the minimum of 1.44 mV at 62.8 degrees come from the shape of the curve between points, not from measurements: 1.44 mV is lower than any individual value recorded at 50 or 70 degrees. The drop in signal over the five closely spaced trials, which the authors call a learning effect, could also be fatigue; removing it with a straight line raises the ICC, but the trial-to-trial variability remains, and other forms of the index would handle a systematic decline separately without adjusting the data. The repetitions were in the same session, not on different days. The signal is not normalised, the electrode sits over the innervation zone, which is usually avoided because it makes amplitude less stable, and no filters or sampling rate are reported. The main table also has values that do not add up: at 50 degrees the mean (3.12 mV) is above the maximum (2.47) and at 70 degrees it is below the minimum (2.69 against 2.71); and at one point the text labels 0.745 as ICC(1,k), while elsewhere it is ICC(1,1).

For anyone who shoots, this study changes nothing: it does not measure shooting, does not compare archers with non-archers and does not say which leg muscles matter for stability on the line. It may interest people who use electromyography with archers, for example to monitor a physical preparation programme, because it shows three things about rectus femoris in a maximal effort: the signal changes a great deal with hip angle, varies considerably from person to person and, when the test is repeated a few minutes apart, falls from one trial to the next.

What follows is a caution rather than a finding. Comparing EMG values taken at different angles, or after several closely spaced repetitions, can lead to misleading conclusions unless both effects are taken into account, and the mean of several trials is more reliable than a single one.