A Prediction Model For Estimating Individual Sweat Rate Using Heart Rate During Submaximal Running

PURPOSE: To develop a prediction model estimating an individual sweat rate (SR) using heart rate (HR) during submaximal running. METHODS: Ten men (24.9 ± 2.9 yrs, 174.6 ± 3.5 cm, 70.4 ± 7.0 kg) underwent a total of four testing sessions. For the first two sessions, they ran on a treadmill for 60 min at 23 ± 2 °C and 30 ± 10 % relative humidity, randomly either 45% (target HR of 130.5 ± 5.2 bpm) or 75% (165.8 ± 3.1 bpm) of their maximal HR calculated by using HR reserve and the Karvonen formula. After completion of the first two testing, they ran identically to the previous two tests, except running intensity, randomly either 55% (target HR of 142.1 ± 4.4 bpm) or 65% (153.9 ± 3.6 bpm) of maximal capacity. Before each running, their urine specific gravity was confirmed, maintaining between 1.003 and 1.030. Before and after each running, their weight with shorts, socks, and shoes were measured. No drinking was allowed during testing. SR was calculated by the difference between the two weights. RESULTS: The average HR and SR at 45, 55, 65, and 75% of maximal capacity was 129.0 ± 8.8, 143.2 ± 8.9, 153.9 ± 10.8, and 164.0 ± 6.5 bpm, and 0.74 ± 0.18, 0.90 ± 0.19, 1.04 ± 0.23, and 1.17 ± 0.25 L/h, respectively. The correlation between the pooled data of average HR and SR was r = 0.508 (p < 0.05). When the relationship between HR and SR was evaluated by everyone separately, the correlations were very high, ranging from r = 0.93 to r = 0.99 (p < 0.05). To verify whether individual SR can be predicted by individual HR responses during running, a regression line was established between 45% and 75% exercise intensities. And the differences between the predicted and the measured SR values of 55% and 65% exercise intensity was calculated. The mean error of two values was 30.6 ± 18.8 (3.4% difference from the measured), and 48.7 ± 23.1 ml/h (4.6% difference from the measured), respectively. CONCLUSIONS: If an individual HR and SR of two separate exercise intensity such as at 45% and 75% of maximal HR can be measured to establish a regression line, SR at other intensity can be predicted relatively in precision only by using HR measurement during running. It can be proposed that this SR prediction model is highly practical for an individual demand.