Comparison of regional patch collection vs. whole body washdown for measuring sweat sodium and potassium loss during exercise

Abstract

This study compared simultaneous whole body washdown (WBW) and regional skin surface (REG) sweat collections to generate regression equations to predict WBW sweat Na+ concentration ([Na+]) and K+ concentration ([K+]) from single- and five-site REG sweat patch collections. Athletes (10 men, 10 women) cycled in a plastic chamber for 90 min in the heat. Before exercise, the subject and bike were washed with deionized water. After the onset of sweating, sterile patches were attached to the forearm, back, chest, forehead, and thigh and removed on saturation. After exercise, the subject and bike were washed with ammonium sulfate solution to collect all sweat electrolyte loss and determine the volume of unevaporated sweat. All individual patch sites and five-site REG (weighted for local sweat rate and body surface area) were significantly ( P = 0.000) correlated with WBW sweat [Na+]. The equation for predicting WBW sweat [Na+] from five-site REG was y = 0.68 x + 0.44 [ r = 0.97, intraclass correlation coefficient (ICC) = 0.70] and did not differ between sexes. There were sex differences in the regression results between five-site REG and WBW sweat [K+] (men: y = 0.74 x + 0.30, r = 0.89, ICC = 0.73; women: y = 0.04 x + 3.18, r = 0.03, ICC = 0.00). Five-site REG sweat [Na+] and [K+] significantly overestimated that of WBW sweat (59 ± 27 vs. 41 ± 19 meq/l, P = 0.000 and 4.4 ± 0.7 vs. 3.6 ± 0.7 meq/l, P = 0.000, respectively). For both sexes, the best sites for predicting WBW sweat [Na+] and [K+] were the thigh (1 ± 8 meq/l < WBW, P = 1.000, y = 0.75 x + 11.37, r = 0.96, ICC = 0.93) and chest (0.2 ± 0.3 meq/l > WBW, P = 1.000, y = 0.76 x + 0.55, r = 0.89, ICC = 0.87), respectively. In conclusion, regression equations can be used to accurately and reliably predict WBW sweat [Na+] and [K+] from REG sweat collections when study conditions and techniques are similar to that of the present protocol.