Hearing your heartbeat while shooting: Takai 2009 study

Thirty-six Japanese archers shot six arrows at 50 metres while listening to the sound of their own heartbeat through earphones, and on another day the same six arrows hearing nothing. With the heartbeat in their ears the mean score over six arrows rose from 42.3 to 44.4 out of 60, the shooting time became more consistent (coefficient of variation 12.5% against 16.7%) and the arrows released during ventricular diastole, the phase in which the heart is not ejecting blood, went from 102 to 139 out of 216.

The length of the shot itself did not change: what changed was its consistency. And in both conditions the arrows released in diastole were worth more on average than those released in systole: 7.3 against 6.8 points without the heartbeat, 7.6 against 6.9 with it.

On mood, nothing. The six indicators collected on a visual scale (tense, sad, angry, lively, weary, confused) showed no difference between the conditions, and the comments gathered afterwards point in opposite directions: twelve archers out of thirty-six reported pleasant sensations ("it felt as though I was shooting with more rhythm than usual", "listening to the sound, my shooting tempo got better"), while six found it unpleasant ("the sound bothers me and I cannot concentrate", "it irritates me"). The R-R interval shortened from the initial rest to the shooting in both conditions, from about 700 to 713 milliseconds down to about 638 to 648.

A second experiment serves as a control. Twelve archers shot the same six arrows while listening to a metronome at 90 beats per minute, chosen because a preliminary test gave a heart rate of around 90 at the moment of the shot. The metronome changed nothing: not the score, not the shooting time, not its variability, not the distribution of arrows across the two phases of the cardiac cycle. All that remains is that here too the arrows in diastole were worth more than those in systole. From this the authors conclude that the result of the first experiment comes from hearing one's own heartbeat, not from any external rhythm.

The thirty-six participants in the first experiment were Japanese archers aged 16 to 24 (27 men and 9 women, mean age 19.5), among them high school students, university students and working adults, none of whom had taken part in an experiment of this kind before. Testing took place in the first half of May 2006 at the ranges each of them normally used for training, under clear or cloudy skies, with a mean temperature of 20.5 degrees, mean humidity of 60.7% and mean wind of 1.10 metres per second.

The design is within-subject: one condition per day, two consecutive days, with the order counterbalanced across participants. The distance was 50 metres, chosen with the less experienced archers in mind, and the arrows numbered six, as in a competition end. Each day was divided into three parts: one minute of rest standing with the eyes open, four minutes of shooting and another minute of rest. The four minutes come from the Japanese federation's rules, which allowed forty seconds per arrow; everyone shot within that time, and those who finished early stood still.

The heartbeat came from an interface that derives a sound from the electrocardiographic signal, presented through earphones proportionally and with immediate feedback: the sound started 30 milliseconds after the R wave, was fixed at a frequency of 300 hertz, and the volume was set by each archer so as not to interfere with shooting. In both conditions the archers wore the earphones, and the difference was that in one no sound came through. Shooting time was measured with an accelerometer taped to the flat of the riser on the side opposite the clicker, about 5 centimetres above the grip, and defined as the interval between the vibration of the set-up and the vibration of the release; the moment of release is that vibration itself. Everything was amplified and converted at 1 kilohertz. Ventricular systole is defined from the peak of the R wave to the end of the T wave, and diastole from the end of the T wave to the next R wave. Mood was collected before and after on a visual analogue scale marked on 10-centimetre lines, one item for each of the six scales of the short Japanese POMS.

Six arrows per condition is very little on which to measure a score. The jump from 42.3 to 44.4 points out of 60 amounts to less than half a point per arrow, measured outdoors at 50 metres on two different days, with wind and temperature changing: a difference of that size sits inside the variability of a single end, and the study has no group shooting both days with no sound at all against which to check it.

Nor is the unit of analysis the archer. The degrees of freedom in the comparisons of per-arrow score (1 and 134 in the first experiment, 1 and 140 in the second) indicate that the counting was done on arrows rather than people, and with six arrows from each of thirty-six archers, arrows from the same archer are not independent observations. This inflates the significance of the comparisons on score and cardiac phase.

The comparison between diastole and systole is the paper's most repeated finding, since it appears in both conditions and both experiments, but it remains an association: nobody assigned the archers a moment at which to release. The explanation the authors offer, that the body sways less in diastole because there is no ejection of blood, comes from other papers they cite and is not measured here, where there is no force platform and no measurement of sway.

The second experiment is on twelve people only, all male university students, so its absence of effect should also be read as a small sample. The archers knew which condition they were in, and there is no blinding of any kind. The metronome was fixed at 90 beats for everyone, with no adjustment to each archer's own rate. The paper is in Japanese, with an English abstract, and the journal is national in scope.

One thing belongs at the end, and the authors add it themselves in a note: from 1 August 2006, a few months after these tests, the Japanese federation's rules banned headphones, earphones and communication devices beyond the waiting line. The method measured here cannot be used in competition, and the authors point to the study of what remains after a period of training with the heartbeat in the ears as the next step.

This study has survived in citations as the evidence that releasing in diastole gives more points, and it is worth being precise about what it actually says. It says that, across thirty-six archers shooting six arrows each, arrows that left during the phase in which the heart is not ejecting blood were worth half a point more on average than those that left in the other phase, and that this showed up in all four sets of data collected. It is an association measured on arrows, not an experiment in which anyone decided when to release, and the explanation about body sway is not measured here.

On the rest the finding is more fragile than it sounds. Listening to one's own heartbeat through earphones made the duration of the shot more consistent, and brought more arrows into diastole, with a score two points higher out of sixty: two different days, six arrows at a time, outdoors at 50 metres. That the metronome produced nothing is the most interesting piece of information in the paper, because it separates hearing one's own heartbeat from having an external rhythm to lean on.

For an archer today the practical question is closed by the rules. Headphones and earphones are not allowed beyond the waiting line, and were not allowed within a few months of these tests: anyone wanting to use this method could only do so in training, and what would remain of it in competition nobody has measured. The participants' reactions are worth noting too, because they do not run in one direction: twelve of the thirty-six found the sound pleasant, six found it a nuisance, and on the mood measures there was no difference at all. This is not a device that works the same way for everyone.