Breathing and imagery in archers: Keskin 2025

Thirty-six novice archers aged 14 to 18 were split into three groups of twelve for eight weeks: one group added fifteen minutes of alternate-nostril breathing to their training, a second fifteen minutes of guided imagery, and the third did the training alone.

The score, measured over three ends of three arrows at 18 metres, rose by 2.25 points in the breathing group (p = 0.003) and by 4.92 points in the imagery group (p < 0.001), while in the controls it fell by 0.42 points (p = 0.51). The comparison across the three groups on score is the study's clearest result (F = 17.37, p < 0.001), and in the pairwise comparison all three differences hold up, including the one between breathing and imagery.

On the other measures the picture is less clean, because the control group improved too. Postural sway, measured with an application in five positions with the eyes closed on a 0 to 100 scale, improved in all three groups (p < 0.001 in the first two, p = 0.008 in the controls), and in the between-group comparison the two interventions sit ahead of the controls but cannot be told apart from each other. Attention, measured with the Bourdon test, improved in all three; reaction time shortened in all three, by 49.7 milliseconds in the breathing group and 71.4 in the imagery group; grip strength rose in all three, with no difference between the two interventions.

On breathing, forced vital capacity rose in the two intervention groups but not in the controls, while forced expiratory volume in one second rose in all three; the ratio between them and peak expiratory flow show no between-group differences. Among the imagery scores, the only one on which the imagery group improves is motivational specific imagery, where the controls improved as well.

The sample size was calculated in advance with G*Power, with a type I error rate of 5%, an effect size of 0.25 and 80% power, referencing an earlier study: at least 33 were needed, and allowing for 10% dropout the target was set at 36. Of the 42 assessed, six were excluded: four for lack of time, one for pulmonary disease and one for an upper-limb asymmetry on the Y balance test. Nobody dropped out along the way.

To take part an archer had to be actively practising the sport and to pass two field tests designed to reduce the risk of injury during the sessions: the upper extremity stability test, in which touches with the opposite hand are counted over fifteen seconds in a push-up position, and the upper quarter Y balance test. Musculoskeletal disorders, neurological or vestibular disease, pulmonary disease, hearing loss and previous orthopaedic surgery were exclusions.

The measures: a portable spirometer following the American and European respiratory society standards; balance with the Sway Medical Balance App, in five positions with the eyes closed, on a 0 to 100 score; imagery with a sport mental training inventory of five subscales and twenty items on a five-point Likert scale, with an internal consistency of 0.87 in this sample; attention with the Bourdon test, in which set letters are marked within a time limit; visual reaction time with the web-based Human Benchmark test, as the average of five trials; grip strength with a Jamar dynamometer on the dominant hand, as the average of three trials; and score over three ends of three arrows at 18 metres under federation rules.

All the sessions took place at an archery training centre under the supervision of a licensed physiotherapist and a certified coach. Each group trained twice a week for eight weeks, an hour per session, with the same warm-up and cool-down; the two intervention groups added fifteen minutes. Imagery was done in a quiet, dimly lit room guided by a recording, in five three-minute cycles with thirty-second breaks, seated with the eyes closed, mentally rehearsing the twelve standard steps of the technique. The breathing was based on the Nadi Shodhana method, in a room at 25 degrees, with inspiration and expiration of equal length, five repetitions per nostril repeated over fifteen minutes; single-nostril breathing for the first two weeks, alternating from week three to week eight.

The analysis used the paired t-test or Wilcoxon within groups, and one-way or Welch analysis of variance or the Kruskal-Wallis test between groups, with Tukey or Mann-Whitney as pairwise comparisons and a Bonferroni correction that brings the pairwise threshold to 0.017.

The score is calculated from nine arrows at 18 metres. Over so few arrows a difference of two or five points sits inside the normal variation of a beginner from one series to the next, and that applies to the study's main result. The distance is the indoor one and the population is teenagers aged 14 to 18 who are still learning: nothing here concerns trained archers or outdoor distances, despite the title referring to Olympic archers.

There is a contradiction about how the groups were formed. The abstract and the methods paragraph say the allocation was random; the protocol paragraph says the participants were allocated by a sequential assignment method, chosen to keep the groups the same size while keeping the procedure simple; the discussion goes back to saying the allocation was random. The three statements do not sit together, and with sequential assignment the design is not randomised.

The authors themselves acknowledge baseline differences between the groups on some respiratory and cognitive measures, and that the small sample and the violation of normality assumptions ruled out the analyses that adjust for the starting value: they therefore accept that part of the between-group difference may come from there. There is no follow-up, and nobody was kept unaware of the allocation: an archer doing imagery knew they were doing it.

The two intervention groups also trained fifteen minutes longer per session. On postural sway, attention, reaction time and grip strength the control group improved as well, so on these measures the advantage attributable to the intervention is smaller than the within-group improvement suggests.

Some passages of the text do not match the tables. The discussion states that the superiority of the imagery group over the controls was not statistically significant, while the between-group table gives p = 0.001 for score between those two groups. The table notes give thresholds of 0.005 and 0.17, where the methods text says 0.05 and 0.017. The acknowledgements declare the use of a language model for language editing.

The study concerns teenagers who are learning to shoot, and its main result is that adding fifteen minutes of breathing or imagery to an hour-long session went together with better scores over nine arrows at 18 metres, while those who only shot did not improve. The gain was larger with imagery, about five points over nine arrows, than with breathing, about two.

The size of the measurement is worth keeping in mind, though. Nine arrows are few for assessing a beginner, and in a group of this age, over eight weeks, normal variation is wide. The two groups that improved were also training a quarter of an hour longer than everyone else, and the authors admit they could not adjust for the baseline differences between groups.

On the other measures the picture is more interesting read the other way round. Balance, attention, reaction time and grip strength improved in all three groups, controls included. In other words, over eight weeks of archery a beginner improves on these things anyway, and the part attributable to breathing or imagery is an addition that separates from the controls in the pairwise comparisons but not from each other.

One methodological detail is worth noting for what it says about the movement: in the imagery group the archers mentally rehearsed the twelve steps of the technique, from stance through release to follow through, with the same recorded guidance for everyone. This was not imagining the target: it was rehearsing the sequence.

Nothing here says what would happen to a trained archer, or that either practice is preferable to the other: on score imagery did better, on the other measures the two cannot be told apart, and the study runs for eight weeks with no follow-up.