REF-1209 · Timpka et al., 2020 · Journal of Science and Medicine in Sport · Paper · PubMed

Does feeling unwell in the taper matter for race day?

In 245 World Championship athletes, a health complaint that worried them during the taper went with lower odds of peaking (endurance OR 0.38). Observational.

B+moderate evidenceMeasured on elite to world-class athletes

Timpka et al. asked whether what athletes report about their health, and the heat-stress prevention strategies they choose during the taper, predict whether they peak at a championship held in the heat.

Before the 2015 World Athletics Championships in Beijing, they collected data on health and heat-stress prevention; peaking was defined from each athlete's pre-competition ranking and final competition rank. Of 841 eligible athletes, 245 (29%) took part.

In both event groups a health complaint during the taper that had caused the athlete concern went with lower odds of peaking (sprint/power OR 0.33, P = 0.038; endurance/combined OR 0.38, P = 0.049).

Endurance/combined-event athletes who chose pre-cooling against the heat were less likely to peak (OR 0.35, P = 0.013), and sprint/power athletes who reported a sudden-onset injury complaint during the taper were more likely to peak (OR 4.47, P = 0.019), which the authors read as a possible benefit of rest ("appeared to benefit from rest").

They conclude that athletes' self-reported health should be monitored during the taper, concerns addressed, and heat-stress strategies tested individually before a hot championship.

Our reading: the only entry that links a taper-period signal to whether a championship athlete peaked; it supports "watch health in the taper, change nothing new", but an observational survey with low participation cannot say that the complaints or the pre-cooling caused the poorer results.

Protocol

Design: prospective cohort, data collected before the 2015 World Athletics Championships in Beijing, China (hot conditions); outcome from the championship.
Who: n = 245 participating athletes of 841 eligible (29%); split into sprint/power and endurance/combined events. Numbers per group, sex and age are not given in the abstract.
Exposures: self-reported health complaints during the taper (including whether they caused concern, and sudden-onset injury complaints) and the heat-stress prevention strategies chosen, such as pre-cooling.
Outcome: peaking, defined from the athlete's pre-competition ranking and final competition rank.
Analysis: baseline and endpoint data fitted into multiple logistic regression models ("multiple logic regression" in the abstract); covariates are not given in the abstract.
Not given in the abstract: the questionnaire items, how a "concern-causing" complaint was defined, the share of athletes reporting each exposure, adjustment variables, and funding.

Key Results

Exposure during the taperEvent groupOdds ratio for peaking (95% CI)P
Concern-causing health complaintSprint/power0.33 (0.11 to 0.94)0.038
Concern-causing health complaintEndurance/combined0.38 (0.14 to 1.00)0.049
Chose pre-coolingEndurance/combined0.35 (0.15 to 0.80)0.013
Sudden-onset injury complaintSprint/power4.47 (1.28 to 15.59)0.019
Participation: 245 of 841 eligible athletes (29%).
A number that is fragile: for endurance/combined athletes the upper confidence limit is 1.00 with P = 0.049, so the association sits on the edge of significance and could vanish with a small change in the data or the model.

Mechanisms

Authors' interpretation: health complaints that cause the athlete concern during the taper predict failure to peak; sprint/power athletes with an acute injury symptom may have benefited from the rest it forced; pre-cooling strategies need validation in real-world endurance and combined events.
Inference: athletes who chose pre-cooling may have been those most worried about or most affected by the heat, so the pre-cooling association may reflect who chose it rather than what it did; the same goes for health complaints and athletes with a heavier or more disrupted season.
Inference: the outcome is rank against expectation, not time; it depends on the field, the conditions and others' performance as well as the athlete's own peak.

Practical Application

Does feeling unwell in the taper matter for race day? In 245 World Championship athletes, a health complaint that worried them during the taper went with lower odds of peaking (endurance OR 0.38). Observational.
Ask about health every taper day, and act on a concern: this is the authors' recommendation, and it fits REF-1001 (B, elite track and field: weeks lost to injury and illness predicted failure at championships). The base's illness rules (SYNTH-36: fever or chest symptoms stop training; above-the-neck symptoms without fever, easy and short) apply in the taper as anywhere; this study adds that an unresolved worry is itself a warning for the peak (observational).
Nothing new in the taper, heat strategies included: endurance athletes who chose pre-cooling peaked less often (OR 0.35).This does not show pre-cooling hurts — laboratory pooling found precooling helped prolonged exercise in the heat (REF-111, A-) — but it supports the authors' advice to test any cooling strategy in training before a hot championship. SYNTH-06 currently advises cooling before and during a long hot effort on the strength of REF-111; this association does not overturn that, and part of it may come from the combined events, whose sprints and jumps REF-111 found precooling impairs (inference).
Do not read the sprint result across: a sudden-onset injury complaint went with more peaking in sprint/power athletes (OR 4.47, wide interval); it says nothing for endurance athletes.
Fit with the taper evidence: arriving overreached also blunts the taper gain (REF-1056, REF-1203); a taper protects a peak only if the athlete reaches it healthy.

Limitations

Abstract only. The full text (Elsevier) is not open access and the Linköping repository record has no file; questionnaire content, definitions, covariates and group sizes could not be read.
Observational and self-reported. Exposures were the athletes' own reports and choices, so confounding by heat sensitivity, prior illness, travel or season load cannot be excluded.
Low participation. 29% of eligible athletes took part; those who answer such surveys may differ from those who do not.
Several models and exposures were tested; one of the two health-complaint associations has a confidence interval reaching 1.00.
Setting. One championship, in the heat, in athletics; the endurance group includes combined-event athletes.
Funding and conflicts: not given in the abstract.
How this entry was checked and changed
Why this level

large observational cohort (n = 245), from the abstract. Associations only: 29% participation, self-reported exposures, a rank-based outcome, and an endurance confidence interval that reaches 1.00