EMG and shooting distance in archery: Clarys 1990

This is not one study but four, carried out over three years at the Vrije Universiteit Brussel with the same electromyography system and gathered into a single paper. The common thread is a single question: can electromyography tell anything apart in archery? The answer that emerges from the four experiments is that muscle activation patterns change little, intensity does change, and what separates archers is their ability to repeat themselves.

In the first experiment, eight competitive archers shot nine arrows per distance at 18 and 25 metres indoors and at 50, 70 and 90 metres outdoors, with six muscles instrumented on both sides. Overall intensity rose with distance, but the only significant step was between 25 and 50 metres: there was no difference between the two indoor distances or among the three outdoor ones. Each muscle's contraction pattern stayed the same at every distance, with the trapezius firing first on both the bow and the drawing side. Precision fell with distance, but not significantly. The same experiment compared the archers' own bows with an imposed standard bow, finding no difference in either activity or precision, and found no relation between an archer's arm pulling force and the draw weight of their bow.

The second experiment tested four string grips on nine internationally competitive archers who knew only the vertical three-finger grip. Averaged over all muscles, activity was 59.26% of maximal contraction with the three-finger grip, 62.40% with the two-finger grip, 56.70% with the reversed grip and 53.40% with the thumb grip, the most economical of the four. On precision, measured as the contour of the group in centimetres, the result reverses: 14.1 centimetres with three fingers and 11.1 with two, against 42.0 with the thumb and 53.2 with the reversed grip. Those last two figures, though, are the end point of three attempts: the thumb grip went from 175.7 to 75.7 to 42.0 centimetres over nine arrows in all, the reversed grip from 179.4 to 97.5 to 53.2. There was no correlation between muscle activity and precision.

The third experiment compares three Olympic archers, six Belgian national competitors and six beginners with a few months' practice, over ten muscles and eighteen arrows at 18 metres, also measuring arrow speed. Neither activation pattern nor intensity nor arrow speed separated the three groups. Repeatability did: in the Olympic archers 92% of the patterns were classified as identical from one arrow to the next, in the national archers 41%, and in the beginners 79% of the patterns fell into the category furthest from exact repetition. The Olympic archers reproduced exactly the same arrow speed on nine or ten shots out of eighteen. The authors put that consistency down to the control of three muscles in particular: the trapezius at the start of the draw, the biceps while aiming and the extensor digitorum at the release.

The fourth experiment had seven skilled archers shoot nine arrows with stabilisers and nine without, at 70 and at 90 metres, in random order and with their own bows. The patterns did not change. Intensity did: without stabilisers activity was significantly lower, with one exception, the brachioradialis of the bow arm, which was more active without them. Precision, however, did not differ between the two conditions, and no relation emerged between muscle activity, aiming time, drawing time and precision.

All four experiments use the same apparatus, described in the paper piece by piece: active bipolar surface electrodes with two 15 mm silver chloride plates 10 mm apart, a pre-amplifier in the electrode itself, an external amplifier, a frequency-modulated tape recorder on seven channels, and then digitisation on a personal computer with a 12-bit converter. The skin was shaved, cleaned with a chemical agent and abraded with sandpaper, and the electrodes were strapped on and filled with electrolyte paste. The raw signal was rectified and averaged to give a linear envelope, and normalised to the maximal voluntary contraction.

The analysis is of two kinds. The quantitative one is integrated electromyography, the area under the curve, expressed as a percentage of maximal contraction. The qualitative one compares the shape of the curves of consecutive shots with a three-category system, IDANCO: identical when the reproduction is perfect, analogue when it is very similar, conform when the pattern is recognisable but with shifts in the timing and amplitude of the slopes and peaks. In the third experiment the authors also introduce a muscle economy index, built as the difference between the areas of two shots divided by the number of muscles.

In all four experiments, performance is not the FITA score but grouped precision, the contour of the holes on the target measured in centimetres; in the first experiment the contour of the three best arrows of each end of nine is used. In the second and third experiments arrow speed was also measured with a dedicated device. The samples are four and separate: eight competitive archers for the distances, nine archers with international experience for the grips, fifteen people across three levels for the comparison between levels, seven skilled archers for the stabilisers.

The grip protocol deserves a detail, because the archers did not know the grips: fifteen minutes of trial and error without a target to get used to them, then three ends of three arrows at the target to settle the anchor, and finally the recording over three shots. In the stabiliser experiment the nine arrows with and the nine without were mixed, using the same personal bow. The comparisons use two-way analysis of variance, with the threshold at p < 0.05.

The number of arrows is the biggest limitation and applies to all four experiments. Nine arrows per distance in the first, three arrows recorded per grip in the second, eighteen arrows in the third, nine per condition in the fourth: on samples this small, grouped precision is an unstable measure, and the authors themselves report non-significant differences almost throughout. The third experiment rests on three Olympic archers, and the percentage of identical patterns attributed to Olympic archers comes from those three.

The comparison between grips has a design flaw the authors state openly: the archers had used the three-finger grip for years and the other three for a quarter of an hour. What is measured, then, is not the economy of a technique but the economy of a technique just encountered, and the comparison says nothing about what those grips would do after months of practice. The conclusion about the thumb grip is explicitly the authors' hypothesis, and they call for a test after a long period of training. In the copy consulted, moreover, the figure for the grip experiment carries the horizontal axis of distances from the previous experiment.

In the fourth experiment none of the seven archers had ever shot or competed without stabilisers, and the conclusion that lower muscle activity leads to more precision and delays fatigue is stated by the authors as their belief rather than as a result: fatigue was not measured and precision did not change. The table of precision scores with and without stabilisers is given without units, and in the table of muscle activity two muscles are assigned to the drawing arm where the methods text and the discussion speak of the bow arm; the same table carries a note flagging incomplete data on some rows.

Some things that would be measured today are missing. Surface electromyography says how electrically active a muscle is, not how much force it produces, and the paper uses it to infer economy and efficiency. The IDANCO system sorts curves into three categories by judgement, not with a measured distance between curves. Trunk inclination is mentioned in the conclusions of the first experiment but does not appear among the measures described. Precision is always the grouping, never the score, so nothing the paper finds is tied to a competition result. Finally its age: the work is from 1990, the equipment and the electronics are those of the time, and the stabiliser comparison concerns the systems of thirty years ago.

The most cited thing in this paper is also the firmest: across fifteen archers at three levels, the muscle activation pattern, the intensity and the arrow speed did not say who was the Olympian and who the beginner. What said it was the resemblance between one arrow and the next, with 92% of patterns classified as identical in the three Olympic archers against 41% in the national competitors, and with arrow speed reproduced exactly on nine or ten shots out of eighteen in the best of them. In today's language this is repeatability, and this is one of the first papers to put it at the centre.

On distance, the practical point is that muscular load does not rise in proportion to the metres: between 50, 70 and 90 metres there was no measurable difference, and the step came between indoors and outdoors. In the same experiment, individual archers' arm pulling force bore no relation to the draw weight of the bow they had chosen, which the authors remark on as an arbitrary choice even among archers at that level.

The grips are the most fragile part and also the most curious. The thumb grip demanded less muscle than any other, but the groups were four times wider than those shot with the fingers, and over three attempts of three arrows they went from 175.7 to 42.0 centimetres. An archer reading that figure is not looking at a comparison between techniques but at the learning curve of strong archers trying a grip for the first time.

On stabilisers the study says something narrow: over nine arrows per condition, taking the stabilisers off lowered measured muscle activity without changing the grouping, in archers who had never shot without them. It does not say that shooting without stabilisers is better, and the idea that lower activity leads to points or to less fatigue is stated by the authors as their expectation. The one muscle that worked harder without stabilisers was the brachioradialis of the bow arm, the forearm that holds the bow, which the authors read as compensating for the missing external stabilisation.