Purpose: In many sporting and occupational settings, measuring core temperature using invasive methods is not practical. In these circumstances, ingestible temperature sensor capsules are a promising alternative. To date, no researcher has investigated the validity of intestinal temperature (T(int)) during free running or examined the between-trial reproducibility of this technique. Therefore, in two investigations, we examined the validity and reliability of Tint during prolonged intermittent shuttle running.
Methods: In investigation A, 10 male games players completed 60 min of exercise while their rectal temperature (T(rec)) and T(int) were monitored. In investigation B, Tint was measured while nine males undertook two 90-min bouts of exercise, separated by 7 d.
Results: A mean systematic bias of -0.15 degrees C (95% CI, 0.10-0.20) was found between T(int) and T(rec) during exercise. This bias for T(int) to record higher temperatures than T(rec) was uniform through the range of measurements, such that the exercise-mediated changes in body temperature were similar between methods of measurement. The 95% limits of agreement were found to be +/- 0.22 degrees C (95% CI, 0.11-0.33) and correlations were high (r > 0.85), suggesting that random error between methods was acceptably small. In investigation B, the mean change between repeated trials was a negligible 0.01 degrees C (95% CI, -0.02 to 0.05). The within-subjects SD was 0.08 degrees C (95% CI, 0.05-0.15). Random error was uniform through the measurement range and was deemed acceptable on the basis of statistical power calculations.
Conclusion: These findings suggest that the test-retest variability of Tint is acceptably small during intermittent shuttle running. The small amount of random measurement error and similar thermal responses to exercise suggest that Tint is as appropriate for use in exercise physiology research as T(rec), provided that the consistent bias between these measurement methods is allowed for.