A Case-Series Observation of Sweat Rate Variability in Endurance-Trained Athletes

dequate uid replacement during exercise is an important consideration for athletes, however sweat rate (SR) can vary day-to-day. The purpose of this study was to investigate day-to-day variations in SR while performing self-selected exercise sessions to evaluate error in SR estimations in similar temperature conditions. Thirteen endurance-trained athletes completed training sessions in a case-series design 1x/week for a minimum 30 min of running/biking over 24 weeks. Body mass was recorded pre/post-training and corrected for uid consumption. Data were split into three Wet-Bulb Globe Thermometer (WBGT) conditions: LOW (<10 C), MOD (10–19.9 C), HIGH (>20 C). No signi cant differences existed in exercise duration, distance, pace, or WBGT for any group (p> 0.07). Signi cant differences in SR variability occurred for all groups, with average differences of: LOW = 0.15 L/h; MOD = 0.14 L/h; HIGH = 0.16 L/h (p< 0.05). There were no signi cant differences in mean SR between LOW-MOD (p> 0.9), but signi cant differences between LOW-HIGH and MOD- HIGH (p< 0.03). The assessment of SR can provide useful data for determining hydration strategies. The signi cant differences in SR within each temperature range indicates a single assessment may not accurately represent an individual's typical SR even in similar environmental conditions. Keywords:hypohydration; hyperhydration; hyponatremia; uid loss; uid balance 1. Introduction The lean tissue of the

between LOW-HIGH and MOD- HIGH (p< 0.03). The assessment of SR can provide useful data for determining hydration strategies. The signi cant differences in SR within each temperature range indicates a single assessment may not accurately represent an individual's typical SR even in similar environmental conditions. Keywords:hypohydration; hyperhydration; hyponatremia; uid loss; uid balance 1. Introduction The lean tissue of the human body is composed of approximately 73% water [1]. Variations in body fat will result in individual body water levels ranging from ~50–70% of total body mass. This water is critically important for cardiovascular function and thermoregulation. When the body has adequate uid intake to match uid losses indi- viduals are considered to be in a state of euhydration. When uid intake is in excess of uid loss individuals can become hyperhydrated and when uid loss exceeds uid intake individuals become hypohydrated. Hypohydration is the term used to describe a state of suboptimal body water. At rest, internal factors that in uence the body's water status are mainly body composition, hormonal activity, and sweating [2]. External factors with the greatest in uence on body water levels include uid intake, medications, medical conditions, physical activity, envi- ronmental conditions, and clothing [3]. Many athletes, from youth to professional, initiate training in hypohydrated states [4–6], and dehydration through sweat loss with insuf cient uid replacement will exacerbate this hypohydrated condition. As a result of hypohydra- tion there is increased cardiovascular strain [7], thermal strain [8], perceived exertion [9], and reduced oxygen and nutrient delivery to the exercising muscle [8,10,11]. Due to these physiological responses to reductions in body water, exercise performance has been shown to be diminished with as little as 2% hypohydration [12–18]. Hyperhydration is the term used to describe a state of overhydration. At rest excess uid consumption typically leads to increased urine output allowing for the maintenance of a euhydrated state. In individuals with compromised kidney function, unnecessary Nutrients2021,13, 1807.

of overhydration. At rest excess uid consumption typically leads to increased urine output allowing for the maintenance of a euhydrated state. In individuals with compromised kidney function, unnecessary Nutrients2021,13, 1807.

Nutrients2021,13, 1807 2 of 12 increases in uid intake above uid loss can lead to uid retention [19]. Similarly, exercis- ing individuals can experience uid retention due to the increased actions of antidiuretic hormone and aldosterone upregulation during exercise [20,21]. Studies have demonstrated hyperhydration does not aid in exercise performance or heat tolerance [22–26]. While hy- perhydration does not bene t exercise it can be detrimental to health. Uncompensable uid intake can lead to dilution of electrolytes, particularly sodium, leading to hyponatremia, and if untreated can result in cerebral or pulmonary edema leading to death [8]. The evaporation of sweat can remove ~580 kcal/L and serves as a valuable tool to dissipate heat produced through metabolic processes [27]. As we exercise or perform physical work, metabolic heat production increases to match the increased work output. Sweat rate also increases in an effort to combat the increases in body temperature associated with increased metabolic rate. Unfortunately, sweat that drips from the body and does not evaporate does not provide a signi cant source of heat loss. In environments with higher humidity, sweat rates can be elevated signi cantly without fully corresponding to estimated metabolic heat production due to reductions in evaporative cooling as a result of moisture in the air [28]. There have been numerous studies exploring average sweat rates and the impact of intensity, duration, environmental conditions, and clothing [29,30]. Typical sweat rates are reported between 0.5–2.0 L/h during activity [31], although due to the large number of variables in uencing sweat rate there is signi cant variability in the sweat rates reported across and within sports. It has been previously reported that about 2% of athletes have sweat rates that can exceed 3 L/h with the highest reported sweat rate during exercise is 5.73 L/h [29]. Within the extremes of the sweat rate range, little to no sweat is produced as a result of conditions such as hypohidrosis and anhidrosis, while hyperhidrosis can lead to extremely high sweat rates [32]. Reporting an average sweat rate when such large ranges exist can result in athletes incorrectly using and applying the

rate during exercise is 5.73 L/h [29]. Within the extremes of the sweat rate range, little to no sweat is produced as a result of conditions such as hypohidrosis and anhidrosis, while hyperhidrosis can lead to extremely high sweat rates [32]. Reporting an average sweat rate when such large ranges exist can result in athletes incorrectly using and applying the information as the foundation for their individual hydration needs. The most frequently used method to assess sweat loss in both laboratory and eld settings is through pre- and post-exercise body mass changes during exercise. This tech- nique is recommended as a viable method of assessing exercise sweat rate by the American College of Sports Medicine [31] and National Athletic Trainers Association [8]. Limited research has explored the within-subject variability of sweat rate, with day-to-day variation reported to be 5–7% in well-controlled settings. Due to the ease of measuring changes in body mass pre- and post-exercise this technique is commonly recommended to profes- sional and recreational athletes. While athletes have been exposed to this technique, many athletes use it without the full understanding of the controls in place for laboratory-based studies likely resulting in increases in sweat rate variation on a day-to-day basis. Therefore, the purpose of this study was to determine the variability observed in day-to-day sweat rate within endurance trained individuals carrying out regular training without arti cially controlled preparation, environmental, and exercise guidelines. 2. Materials and Methods 2.1. Study Participants Individuals training as recreational runners and triathletes along with collegiate cross- country runners were recruited from the local area to explore the variability in sweat rate throughout multiple seasons. Data collection trials began in September and concluded in February as a result of COVID-19 restrictions. Thirteen endurance-trained males (n= 3) and females (n= 10) were included in the present study and were currently running a minimum of 120 min per week for the previous three months. All participants provided written consent prior to participating. This study was approved by the Institutional Review Board at Mississippi State University.

(n= 3) and females (n= 10) were included in the present study and were currently running a minimum of 120 min per week for the previous three months. All participants provided written consent prior to participating. This study was approved by the Institutional Review Board at Mississippi State University.

Nutrients2021,13, 1807 3 of 12 2.2. Experimental Design Participants completed training sessions once per week for a minimum duration of 30 min in a case-series design. Sessions included either running or biking at a self-selected pace and intensity between 5:30 and 9:30 am. Duration and distance were measured using the athlete's GPS watches, pace was then calculated from these values. Environmental conditions were recorded using a Wet-Bulb Globe Thermometer (WBGT; QUESTemp 32, 3M, St. Paul, MN, USA) at exercise initiation and every 15 min during the training. Sweat rates were calculated from the change in body mass measured (Defender 3000, OHAUS, Parsippany, NJ, USA) before and after training, with correction for uid intake. Immediately prior to the initiation of exercise athletes weighed dry exercise clothes and shoes alone, were asked to void their bladder, and then athlete body mass was collected while wearing dry exercise clothes and shoes. To account for sweat trapped by clothes [33], immediately post-exercise athletes' body mass was collected while wearing exercise clothes and shoes, athletes then changed to allow sweaty clothes and shoes to be weighed alone. Sweat rate was calculated by: Sweat Rate= (CBWPRE CBWPOST)+(CSWET CSDRY)+(FBPRE FBPOST) Time(h) (1) where, CBW = clothed body weight (kg); CS = exercise clothing and shoes (kg); FB = food/beverage (kg). Due to the short duration of exercise no adjustments were made to account for respi- ratory uid losses. Data collection took place outdoors in natural environmental conditions over 7 months except during times of precipitation. The natural environmental conditions were separated into three WBGT ranges: LOW (less than 10 C), MOD (between 10–20 C) and HIGH (above 20 C). 2.3. Statistical Analysis All data were analyzed using SPSS v26 statistical software (IBM, Armonk, NY, USA). Participant comparisons were included if a minimum of two sessions were completed for LOW, MOD, or HIGH conditions, and the highest and lowest sweat rates were used for analysis, as well as the mean sweat rates. A Shapiro-Wilks test of normality was conducted, there were no outliers, and the signi cance was above an alpha level of 0.05, therefore the data is

were included if a minimum of two sessions were completed for LOW, MOD, or HIGH conditions, and the highest and lowest sweat rates were used for analysis, as well as the mean sweat rates. A Shapiro-Wilks test of normality was conducted, there were no outliers, and the signi cance was above an alpha level of 0.05, therefore the data is normal and is parametric. Not all participants completed multiple training sessions in each range, therefore independent sample T-tests were used for comparisons between WBGT, duration, distance, and pace for each temperature range. Analysis of variance (ANOVA) tests with Tukey's Honestly Signi cant Difference (HSD) pairwise comparisons were used to analyze differences in mean sweat rates between temperatures. The participants who completed multiple training sessions in all three temperature ranges (n= 4) were included for subsequent analysis. A Friedman's Rank test was performed to test for differences between temperature ranges. Post-hoc analysis with Wilcoxon Signed- Rank test was conducted with a Bonferroni correction applied. Signi cance was set a priori atp< 0.05. 3. Results Participants displayed no signi cant change between rst and last training session (68.2 14.7 kg vs. 68.4 14.9 kg respectively;p= 0.61). Participant sessions were split into WBGT ranges for analysis. There were no signi cant differences in duration, distance, pace, and WBGT for any of the groups (p> 0.07). These data are shown in Table. There were signi cant differences in sweat rate variability for all groups. LOW WBGTs demonstrated an average difference of 0.15 L/h in sweat rate between highest and lowest recordings (p< 0.01). MOD WBGTs showed an average difference of 0.14 L/h in sweat rate (p< 0.05). HIGH WBGTs revealed an average difference in sweat rate of 0.16 L/h (p< 0.01). Individual sweat rates are shown in Table

Nutrients2021,13, 1807 4 of 12 rates for the four participants who completed training sessions within each WBGT range are represented in Table. Table 1. Differences in duration, distance, pace, and Wet-Bulb Globe Thermometer between temperature ranges. (mean stdev). Range WBGT ( C) Duration (Minutes) Running Distance (Miles) Running Pace (Min/Mile) Cycling Distance (Miles) Cycling Velocity (Miles/Hour) <10 C 5.7 3.2 42.67 13.33 (n= 9) 6.05 4.67 (n= 8) 8.40 2.39 (n= 8) 21.04 0.03 (n= 1) 17.43 0.16 (n= 1) 10–20 C 13.5 2.2 40.63 11.29 (n= 6) 5.19 1.82 (n= 6) 8.07 1.10 (n= 6) >20 C 22.5 1.4 43.29 14.11 (n= 11) 6.84 5.60 (n= 10) 7.62 1.67 (n= 10) 24.41 4.53 (n= 1) 19.34 0.83 (n= 1) Table 2. Participant sweat rates (L/h) separated into highest, lowest, and mean values. 2A denotes Wet-Bulb Globe Thermometer values below 10 C, 2B values between 10–20 C, and 2C values above 20 C. 2A. <10 C Subject Low High Mean Relative Mean High-Low Maximal Variation from Mean (L/h) (L/h) (L/h) (mL kg 1 h 1 ) (L/h) (%) 1 0.84 0.90 0.87 13.5 0.06 3.4 2 0.43 0.52 0.48 8.5 0.09 9.4 3 0.41 0.56 0.48 7.2 0.15 15.5 4 0.65 0.95 0.84 8.7 0.30 17.8 5 0.97 1.00 0.99 9.6 0.03 1.5 8 0.54 0.61 0.57 9.1 0.07 6.2 9 0.57 0.75 0.66 10.2 0.18 13.7 11 0.39 0.52 0.45 5.2 0.13 14.4 12 0.67 0.90 0.78 12.6 0.23 14.8 2B. 10–20 C Subject Low High Mean Relative Mean High-Low Maximal Variation from Mean (L/h) (L/h) (L/h) (mL kg 1 h 1 ) (L/h) (%) 1 1.13 1.23 1.18 11.6 0.10 4.2 2 0.43 0.68 0.60 9.0 0.25 21.0 3 0.58 0.78 0.64 11.9 0.20 15.5 4 0.83 0.84 0.84 10.2 0.01 0.6 6 0.46 0.79 0.57 9.2 0.33 29.1 10 0.49 0.49 0.49 8.5 0.00 0.0 2C. >20 C Subject Low High Mean Relative Mean High-Low Maximal Variationfrom Mean (L/h) (L/h) (L/h) (mL kg 1 h 1 ) (L/h) (%) 1 1.32 1.35 1.33 13.44 0.03 1.2 2 0.81 1.10 0.90 13.9 0.29 16.1 3 0.75 0.96 0.85

0.84 10.2 0.01 0.6 6 0.46 0.79 0.57 9.2 0.33 29.1 10 0.49 0.49 0.49 8.5 0.00 0.0 2C. >20 C Subject Low High Mean Relative Mean High-Low Maximal Variationfrom Mean (L/h) (L/h) (L/h) (mL kg 1 h 1 ) (L/h) (%) 1 1.32 1.35 1.33 13.44 0.03 1.2 2 0.81 1.10 0.90 13.9 0.29 16.1 3 0.75 0.96 0.85 16.1 0.21 12.3 4 1.16 1.16 1.16 13.0 0.00 0.0 5 1.62 1.70 1.67 23.0 0.08 2.4 6 0.63 0.81 0.72 12.5 0.18 12.5 7 0.76 0.91 0.82 13.8 0.15 9.2 8 0.64 0.74 0.69 11.9 0.10 7.2 9 1.01 1.22 1.11 17.6 0.21 9.4 10 0.58 0.59 0.59 10.4 0.01 0.9 13 0.79 0.81 0.80 11.6 0.02 0.1

Nutrients2021,13, 1807 5 of 12 Table 3. Participant sweat rates (L/h) for the four participants that completed multiple training sessions within each Wet-Bulb Globe Thermometer range. Sweat rate values shown as highest, lowest, and mean values for each range. Subject Temperature Low High Mean 1 <10 C 0.84 L/h 0.90 L/h 0.87 L/h 10–20 C 1.13 L/h 1.23 L/h 1.18 L/h >20 C 1.32 L/h 1.35 L/h 1.33 L/h 2 <10 C 0.43 L/h 0.52 L/h 0.48 L/h 10–20 C 0.43 L/h 0.68 L/h 0.60 L/h >20 C 0.81 L/h 1.10 L/h 0.90 L/h 3 <10 C 0.41 L/h 0.56 L/h 0.48 L/h 10–20 C 0.58 L/h 0.78 L/h 0.64 L/h >20 C 0.75 L/h 0.96 L/h 0.85 L/h 4 <10 C 0.65 L/h 0.95 L/h 0.84 L/h 10–20 C 0.83 L/h 0.84 L/h 0.84 L/h >20 C 1.16 L/h 1.16 L/h 1.16 L/h Pairwise comparisons revealed no signi cant differences in sweat rate between LOW and MOD temperatures (p> 0.9), but signi cant sweat rate differences were found between LOW and HIGH (p< 0.03) and between MOD and HIGH (p< 0.01). These differences are represented in Figure. In the four participants who completed sessions in each range, there was a statistically signi cant difference in mean sweat rate depending on temperature condition, 2(2) = 6.500,p= 0.039. Post hoc analysis revealed changes in temperature conditions did not elicit a signi cant change in mean sweat rate (Low-Mod,p= 0.144; Mod- High,p= 0.68; Low-High,p= 0.68). However it should be noted the differences in mean sweat rate mirrored those seen in the group totals. The differences between temperature ranges are shown in Figure.Nutrients 2021, 13, x FOR PEER REVIEW 6 of 13 >20 °C 0.75 L/h 0.96 L/h 0.85 L/h 4 <10 °C 0.65 L/h 0.95 L/h 0.84 L/h 10–20 °C 0.83 L/h 0.84 L/h 0.84 L/h >20 °C 1.16 L/h 1.16 L/h 1.16 L/h Pairwise comparisons revealed no significant differences in sweat rate between LOW and MOD temperatures (p > 0.9), but significant sweat rate differences were found be- tween LOW and HIGH (p < 0.03) and between MOD and HIGH (p

0.65 L/h 0.95 L/h 0.84 L/h 10–20 °C 0.83 L/h 0.84 L/h 0.84 L/h >20 °C 1.16 L/h 1.16 L/h 1.16 L/h Pairwise comparisons revealed no significant differences in sweat rate between LOW and MOD temperatures (p > 0.9), but significant sweat rate differences were found be- tween LOW and HIGH (p < 0.03) and between MOD and HIGH (p < 0.01). These differ- ences are represented in Figure 1. In the four participants who completed sessions in each range, there was a statistically significant difference in mean sweat rate depending on temperature condition, χ 2 (2) = 6.500, p = 0.039. Post hoc analysis revealed changes in tem- perature conditions did not elicit a significant change in mean sweat rate (Low-Mod, p = 0.144; Mod-High, p = 0.68; Low-High, p = 0.68). However it should be noted the differences in mean sweat rate mirrored those seen in the group totals. The differences between tem- perature ranges are shown in Figure 2. Figure 1. Mean sweat rates (absolute: L/h; relative: mL∙kg −1 ∙h −1 ) within each temperature range. a = significantly different from LOW, b = significantly different than MOD. Figure 1. Mean sweat rates (absolute: L/h; relative: mL kg 1 h 1 ) within each temperature range. a = signi cantly different from LOW, b = signi cantly different than MOD.

This study investigates day-to-day variations in sweat rate among endurance-trained athletes.