Nicotine and archery performance: Hung 2021 study

Eleven male college recurve archers, national level, nonsmokers, recruited in 2018 from two universities in Taiwan. Randomised placebo-controlled crossover: a chewing gum containing 2 milligrams of nicotine or a flavour-matched gum without it, chewed for thirty minutes, then the cognitive tests, then a simulated game of 36 arrows.

With nicotine the three outcomes in the first table all improve. The mean time for correct rejections in a figure-matching test falls from 8.23 ± 0.98 to 7.29 ± 0.87 (p = 0.038). The time to fill the grooved pegboard goes from 53.41 ± 4.05 to 48.76 ± 3.18 seconds (p = 0.003). The score rises from 290.58 ± 10.09 to 298.05 ± 8.56 out of 360 (p = 0.009), that is from 8.07 to 8.28 points per arrow: seven and a half points over 36 arrows.

On the physiological side, heart rate while shooting is higher with nicotine, 101.70 ± 8.04 beats per minute against 95.41 ± 5.33 (p < 0.01). Salivary alpha-amylase, which the authors use as an indicator of sympathetic activation, falls after the nicotine gum (39.64 ± 15.97 against 51.82 ± 19.09 units per millilitre, p < 0.01) and rises immediately after the shooting (72.36 ± 36.66 against 50.74 ± 20.44 in the placebo arm, p < 0.01). Salivary cotinine, measured to confirm the gum had worked, reads 10.3 ± 1.8 nanograms per millilitre after thirty minutes and is undetectable both before and after the shooting test.

The authors conclude that nicotine improves performance by acting on cognitive function and stimulating the sympathetic system. The table of heart rate variability parameters, however, says less than that: the condition effect is significant for heart rate alone (p = 0.028), while pNN50 stops at 0.229, LF and HF at 0.371 and the LF over HF ratio at 0.655.

Recruitment by advertisement and word of mouth from the archery teams of the National Taiwan University of Sport and the National Chung Cheng University. Criteria: recurve bow, national-level performance, at least three years of continuous practice with a minimum of two hours three times a week, no history of cardiac disease, no current injury. Nobody was excluded. Mean age 20.3 ± 0.3 years, weight 71.1 ± 9.2 kilograms, height 174.4 ± 4.5 centimetres. The sample size was calculated with G*Power for the heart rate variability outcome, starting from a partial eta squared of 0.34 taken from an earlier study, with alpha at 5 per cent and power at 80: six participants came out as the requirement.

The protocol runs through five moments, from before the gum to immediately after the shooting. On arrival a Polar V800 chest strap and five minutes seated, then saliva collection, then thirty minutes of chewing following the manufacturer's instructions (chew until the taste is strong, park the gum between cheek and gum, chew again when the taste fades), a second saliva collection, the cognitive tests, the simulated game, a third saliva collection. The trials took place at the same time of day, with abstention from coffee and energy drinks for the preceding twelve hours, from alcohol and from exercise on the day before and on the day itself, while habitual caffeine was kept so as not to introduce a withdrawal, and with the diet of the first visit replicated.

The two cognitive tests are the Cognitrone from the Vienna Test System, in which an abstract figure is matched against a model for seven minutes and the mean time for correct rejections is recorded, and the grooved pegboard, twenty-five holes with randomly oriented slots, done with the dominant hand and with the other, the mean of the two times serving as the outcome.

The shooting test took place outdoors under national competition rules, with each archer's own equipment, after a warm-up and a sight check: 72 arrows a day in two games of 36, six arrows at a time with four minutes allowed, fifteen minutes of rest between the two games, arrows pulled and scored always by the same person, no technical input during the test. The same test was repeated on three separate days for each of the two conditions, and the score used is the mean.

Heart rate variability was computed from five-minute recordings at one millisecond resolution, with maximum entropy spectral analysis and coefficients estimated by the Burg algorithm, the low frequency component between 0.04 and 0.15 hertz and the high frequency one between 0.15 and 0.40, both expressed in normalised units. Saliva was collected unstimulated, after rinsing the mouth, over ten minutes seated; cotinine by enzyme-linked immunoassay and alpha-amylase by kinetic assay, in triplicate, with intra-assay coefficients of variation of 5 and 4 per cent.

The analysis uses a two-way repeated measures analysis of variance, condition by time, with a Tukey test, and paired t tests for the comparisons between nicotine and placebo, reporting partial eta squared as the measure of effect. Ethics approval from the committee of Jen-Ai Hospital in Taichung, funding from Taiwan's ministry of science and technology, no competing interests declared.

The physiological conclusion rests on a different comparison from the one it appears to. When the abstract says nicotine reduced heart rate variability, the numbers behind it are the changes from baseline within the nicotine arm, where thirty minutes after the gum heart rate, LF and the LF over HF ratio rise and pNN50 and HF fall, while in the placebo arm nothing moves at the same moment. That part holds. The comparison a reader expects from such a sentence, nicotine against placebo, is a different column of the table, and there, among the five parameters, significance reaches heart rate alone, at p = 0.028: pNN50 sits at 0.229, LF and HF at 0.371, the LF over HF ratio at 0.655. And during the shooting, which is the moment that counts, that ratio is lower with nicotine than with placebo, 4.73 ± 2.04 against 5.28 ± 3.31, in the opposite direction to the proposed mechanism. What nicotine moved, reading the authors' own numbers, is heart rate.

Those five parameters, moreover, are not five measures. LF and HF in normalised units are complementary: in every cell of the table they add to exactly 100, and they carry identical p and eta squared values in all three columns. The LF over HF ratio is their quotient. Two independent measures remain, heart rate and pNN50, presented as five.

The time for correct rejections is printed in milliseconds: 7.29 and 8.23. Nobody matches two figures on a screen and responds in seven thousandths of a second, so the unit is a drafting error and the comparison between the two arms holds, but the number cannot be quoted as it stands.

Cotinine runs the opposite way to how it should. The authors measure it precisely because it lasts longer than nicotine, and then find it at 10.3 nanograms per millilitre thirty minutes after the gum and undetectable after the shooting test, an hour or more later. Cotinine accumulates as nicotine is metabolised, it does not vanish first: one of the two readings does not add up, and with it the only check that the gum delivered what it was meant to.

Blinding is stated but never tested. The control gum matched the flavour, yet two milligrams of nicotine in someone who has never smoked gives a recognisable burn in the throat, and nobody was asked which gum they thought they had chewed. On an outcome the archer watches take shape, arrow by arrow, expectation weighs; and it is not said whether the person recording the scores knew which condition the day belonged to.

The crossover is missing its order. The study is described as randomised, but neither the sequence, nor how it was allocated, nor a washout interval between the two arms is reported, and all of it sits inside one week with six days of 72 arrows. Without the order, a learning or fatigue effect across those six days cannot be told apart from the effect of the gum.

The sample size calculation concerns heart rate variability, with an effect borrowed from a study on strength, not the score: on the outcome that gives the paper its title there is no power calculation at all. Nor is there any correction for multiple comparisons across the three paired tests of the first table, the five cardiac parameters and the alpha-amylase.

The distance of the simulated game is not reported, so those 290 and 298 points over 36 arrows cannot be set beside the scores of other studies. The sample is eleven men, all nonsmokers, all recurve, all national level, with a single dose and a single route, and the authors list this among their own limits. Seven and a half points over 36 arrows are also less than the distance between one archer and another, which in the two arms amounts to 8.6 and 10.1 points of standard deviation; in the study's favour, though, each score is the mean of six games rather than one.

One tension remains that the paper does not try to resolve. The idea that sympathetic activation helps shooting runs against the larger part of the archery literature, where a lower heart rate and parasympathetic predominance in the seconds before release go with the better arrows. And the authors' second sympathetic indicator moves on its own account: alpha-amylase falls after the gum, while heart rate rises.

What the study puts on the table is a clear result on an outcome that matters: eleven national-level archers, six games in each of the two conditions, seven and a half points more over 36 arrows with the nicotine gum, together with a faster hand on the pegboard and a heart rate six beats higher while shooting.

What the study does not allow anyone to say starts with the explanation. The authors put the gain down to sympathetic activation, but in their own table the only autonomic parameter that changes between the two conditions is heart rate, and during the shooting the ratio between the two components of variability goes the other way. So a higher score remains without a demonstrated mechanism, in eleven people, at a single dose, within a week whose order of conditions is not reported.

Then there is the rest of the matter, which lies outside the numbers. Nicotine is addictive, and it is the authors themselves who write so in their closing pages: they recall the eight million deaths a year linked to tobacco and the possibility that an athlete who does not use it may become dependent. On the regulatory side, nicotine is not among the substances prohibited by the world anti-doping agency, which has nonetheless kept it in its monitoring programme since 2013: that is not a ban, it is an observation of what athletes consume.

The registry keeps this work for what it is: a placebo crossover on eleven archers in which the score rises, in which the cognitive effect is consistent with what is known about nicotine more generally, and in which the autonomic reading the authors give does not survive a comparison with their own table. It draws no recommendation from it: a substance that creates dependence is not a technical detail of preparation, and a decision about oneself is not taken on the strength of a gain of two tenths of a point per arrow measured once in eleven people.