Implications of the Onset of Sweating on the Sweat Lactate Threshold

he relationship between the onset of sweating (OS) and sweat lactate threshold (sLT) assessed using a novel sweat lactate sensor remains unclear. We aimed to investigate the implications of the OS on the sLT. Forty healthy men performed an incremental cycling test. We monitored the sweat lactate, blood lactate, and local sweating rates to determine the sLT, blood LT (bLT), and OS. We de ned participants with the OS during the warm-up just before the incremental test as the early perspiration (EP) group and the others as the regular perspiration (RP) group. Pearson's correlation coef cient analysis revealed that the OS was poorly correlated with the sLT, particularly in the EP group (EP group, r = 0.12; RP group, r = 0.56). Conversely, even in the EP group,

OS during the warm-up just before the incremental test as the early perspiration (EP) group and the others as the regular perspiration (RP) group. Pearson's correlation coef cient analysis revealed that the OS was poorly correlated with the sLT, particularly in the EP group (EP group, r = 0.12; RP group, r = 0.56). Conversely, even in the EP group, the sLT was strongly correlated with the bLT (r = 0.74); this was also the case in the RP group (r = 0.61). Bland-Altman plots showed no bias between the mean sLT and bLT (mean difference: 19.3 s). Finally, in ve cases with a later OS than bLT, the sLT tended to deviate from the bLT (mean difference, 106.8 s). The sLT is a noninvasive and continuous alternative to the bLT, independent of an early OS, although a late OS may negatively affect the sLT. Keywords: lactate threshold; sweat rate; exercise testing; incremental exercise; sweating; perspiration; body temperature regulation; sports; physiology 1. Introduction The visualization of exercise tolerance to optimize daily training is encouraged by athletes and supporters. In particular, monitoring metabolic responses, such as the anaer- obic threshold (AT), during exercise enables athletes to evaluate their aerobic capacity in real-time [1–4], which leads to a de ned relative workload intensity [4,5]. Previous studies have investigated the ventilatory threshold (VT), measured using an expiration gas analyzer, and the lactate threshold (LT) as good indicators of AT [3,6]. VT testing is costly and requires expertise; thus, it is not easily accessible to recreational or younger athletes without favorable facilities. LT testing requires frequent blood sampling while interrupting strenuous exercise, meaning it does not re ect usual (continuous) exercise. In addition, contamination with other substances, such as sweat, makes the assessment dif cult. In recent years, wearable sensing technology has been the focus of more precise evaluations of physiological responses in the body. We have developed a method to visualize the lactate dynamics of sweat during exercise in a noninvasive, simple, and real-time manner [7,8]. Furthermore, the sweat LT (sLT), assessed using sweat lactate dynamics, is consistent with the LT calculated from

cult. In recent years, wearable sensing technology has been the focus of more precise evaluations of physiological responses in the body. We have developed a method to visualize the lactate dynamics of sweat during exercise in a noninvasive, simple, and real-time manner [7,8]. Furthermore, the sweat LT (sLT), assessed using sweat lactate dynamics, is consistent with the LT calculated from blood samples (bLT) and the VT [7]. However, unlike blood lactate levels, changes in sweat lactate levels may be affected by sweating dynamics. Sensors2023,23, 3378.

Sensors2023,23, 3378 2 of 11 During exercise, sweating occurs with a rise in body temperature that re ects the increased workload [9–11]. Previous reports have shown a negative correlation between the local sweating rate and sweat lactate level during exercise [12–14], suggesting that increased sweating dilutes the sweat lactate concentration. However, the relationship between the onset of sweating (OS) and sLT remains unclear. If the relationship between the bLT and sLT is strongly dependent on the OS, LT determination using sweat instead of blood may require careful attention to environmental conditions, such as the temperature and humidity, based on the subject's condition. We aimed to investigate the relationship between the OS and sLT and the effect of the OS on the relationship between the sLT and bLT during incremental exercise. 2. Materials and Methods 2.1. Participants Volunteers were recruited from December 2020 to August 2021, and 40 healthy men aged 18–37 years participated in this study. The exclusion criteria were as follows: (1) the presence of comorbidities such as active cardiopulmonary disease, hypertension, or di- abetes within two weeks; (2) elite athletes; (3) smokers and those taking medication or performance-enhancing drugs. The study protocol was conducted in compliance with the ethical guidelines for medical and health research involving human subjects and was approved by the Ethics Committees of Keio University School of Medicine (Approval No. 20180357). Written informed consent was obtained from the study participants for the participation and publication of the ndings before enrollment. 2.2. Procedures and Protocols All experiments were performed at sports facilities under similar conditions (24.8 1.9 C temperature, 42.0 9.3% relative humidity). Prior to the exercise test, the body weight (kg) and height (cm) of each participant were measured. The body mass index (BMI) and body surface area (BSA) were calculated using the formula used by DuBois (body surface area (m 2 ) = 0.007184 height (cm) 0.725 weight (kg) 0.425 ) [15]. The participants were instructed to avoid drinking alcohol or caffeine for 12 h [16]. They were also required to be well hydrated and refrain from vigorous exercise for 3 h before the

body surface area (BSA) were calculated using the formula used by DuBois (body surface area (m 2 ) = 0.007184 height (cm) 0.725 weight (kg) 0.425 ) [15]. The participants were instructed to avoid drinking alcohol or caffeine for 12 h [16]. They were also required to be well hydrated and refrain from vigorous exercise for 3 h before the exercise test. Each participant completed incremental exercise using an electromagnetically braked ergometer (POWER MAX V3 Pro; Konami Sports Co., Ltd., Tokyo, Japan). The ergometer seat was adjusted to a favorable height. Two minutes of rest was set to measure the resting outcomes. Immediately after the warm-up for 4 min with a load of 20 W, the exercise test was started at 50 W, increasing by 25 W every minute. The pedaling cadence range was set to 70–80 revolutions per minute (rpm). Each test was terminated owing to (a) subjective exhaustion or (b) reaching 70 rpm for 10 s. 2.3. Measurements during the Exercise Test The sweat rate and sweat lactate concentration were continuously monitored during incremental exercise. The sweat lactate was measured using a sweat lactate sensor chip (Grace Imaging Inc., Tokyo, Japan), which we developed and applied in several studies (Figure S1) [7,8,17,18]. It is a type of electrochemical sensor that detects the potential generated by the redox reaction between the lactate and lactate oxidase. The special protective membrane structure of the sensor allows the aforementioned reaction to last for 30 to 60 min. As a result, the lactate concentration in sweat can be continuously measured throughout the exercise test. Further detailed information regarding the composition and fabrication of the sensor chip is available in our previous study [7]. This sweat lactate sensor chip connected to a wearable sweat lactate sensor was placed on the upper arm [7]. The installation area was carefully cleaned using an alcohol-free cloth to prevent contamination of the sweat sample. The sensor chip was rmly xed with tape to prevent detachment from the skin. Real-time sweat lactate values were automatically recorded in

lactate sensor was placed on the upper arm [7]. The installation area was carefully cleaned using an alcohol-free cloth to prevent contamination of the sweat sample. The sensor chip was rmly xed with tape to prevent detachment from the skin. Real-time sweat lactate values were automatically recorded in

Sensors2023,23, 3378 3 of 11 a connected application device (Grace Imaging Inc., Tokyo, Japan) via Bluetooth at 1 Hz. The sweat lactate value was quanti ed as the current value because the chip reacts with sweat lactate and generates an electric current [7,8,18,19]. The sLT was de ned as the rst signi cant increase in the lactate concentration in sweat above the baseline based on graphical plots [7,8,18,19]. The sweat rate was measured using a pre-calibrated perspiration meter (SKN-200M; SKINOS Co., Ltd., Ueda, Japan) on the upper arm [17] and a Fitbit Inspire HR (Fitbit Inc., San Francisco, CA, USA) was attached to the left wrist, two- nger widths above the ulnar styloid process, to measure the heart rate. The heart rate at rest was measured once at the onset of the warm-up. The sweat rate was recorded at 1 Hz and expressed in milligrams per square centimeter per minute (mg/cm 2 /min). The baseline of the local sweat rate was calculated as the average of the sweat volume in the rest period. The OS was de ned as the rst signi cant increase in sweat rate above the baseline based on graphical plots (Figure).Sensors 2023, 23, x FOR PEER REVIEW 3 of 11 protective membrane structure of the sensor allows the aforementioned reaction to last for 30 to 60 min. As a result, the lactate concentration in sweat can be continuously measured throughout the exercise test. Further detailed information regarding the composition and fabrication of the sensor chip is available in our previous study [7]. This sweat lactate sensor chip connected to a wearable sweat lactate sensor was placed on the upper arm [7]. The installation area was carefully cleaned using an alcohol-free cloth to prevent contamination of the sweat sample. The sensor chip was firmly fixed with tape to prevent detachment from the skin. Real-time sweat lactate values were automatically recorded in a connected application device (Grace Imaging Inc., Tokyo, Japan) via Bluetooth at 1 Hz. The sweat lactate value was quantified as the current value because the chip reacts with sweat lactate and generates an electric current

sweat sample. The sensor chip was firmly fixed with tape to prevent detachment from the skin. Real-time sweat lactate values were automatically recorded in a connected application device (Grace Imaging Inc., Tokyo, Japan) via Bluetooth at 1 Hz. The sweat lactate value was quantified as the current value because the chip reacts with sweat lactate and generates an electric current [7,8,18,19]. The sLT was defined as the first significant increase in the lactate concentration in sweat above the baseline based on graphical plots [7,8,18,19]. The sweat rate was measured using a pre-calibrated perspiration meter (SKN-200M; SKINOS Co., Ltd., Ueda, Japan) on the upper arm [17] and a Fitbit Inspire HR (Fitbit Inc., San Francisco, CA, USA) was attached to the left wrist, two-finger widths above the ulnar styloid process, to measure the heart rate. The heart rate at rest was measured once at the onset of the warm-up. The sweat rate was recorded at 1 Hz and expressed in milligrams per square centimeter per minute (mg/cm 2 /min). The baseline of the local sweat rate was calculated as the average of the sweat volume in the rest period. The OS was defined as the first significant increase in sweat rate above the baseline based on graphical plots (Figure 1). Figure 1. Representative data of the local sweat rate, blood lactate, and sweat lactate throughout the exercise protocol, including the rest and warm-up periods. The onset of sweating (OS), blood lactate threshold (bLT), and sweat lactate threshold (sLT) are indicated by the dotted lines. The OS preceded both the bLT and sLT (35 out of 40 cases, 88%). To measure blood lactate concentration, a blood sample was obtained from the auricle at rest and every minute during exercise. The lactate concentration in the blood was immediately measured using a lactate analyzer (Lactate Pro2 LT-1730, ARKRAY, Inc., Kyoto, Japan). The blood lactate values were graphically plotted in millimoles per liter (mmol/L). The bLT was determined using graphical plots [3]. Figure 1. Representative data of the local sweat rate, blood lactate, and sweat lactate throughout the exercise protocol, including the rest and warm-up

in the blood was immediately measured using a lactate analyzer (Lactate Pro2 LT-1730, ARKRAY, Inc., Kyoto, Japan). The blood lactate values were graphically plotted in millimoles per liter (mmol/L). The bLT was determined using graphical plots [3]. Figure 1. Representative data of the local sweat rate, blood lactate, and sweat lactate throughout the exercise protocol, including the rest and warm-up periods. The onset of sweating (OS), blood lactate threshold (bLT), and sweat lactate threshold (sLT) are indicated by the dotted lines. The OS preceded both the bLT and sLT (35 out of 40 cases, 88%). To measure blood lactate concentration, a blood sample was obtained from the auricle at rest and every minute during exercise. The lactate concentration in the blood was immediately measured using a lactate analyzer (Lactate Pro2 LT-1730, ARKRAY, Inc., Kyoto, Japan). The blood lactate values were graphically plotted in millimoles per liter (mmol/L). The bLT was determined using graphical plots [3]. 2.4. Statistical Analysis The OS, bLT, and sLT were determined visually by three researchers independently in accordance with previous reports [7,17]. We rst de ned subjects with the OS during the resting period or warm-up as the early perspiration (EP) group, and those with the OS during incremental exercise as the regular perspiration (RP) group. The relationships between the OS, sLT, and bLT were examined using Pearson's correlation coef cient and a linear regression analysis. For each combination, a one-samplet-test on the difference in time was used to examine the xed error, and a linear regression analysis on the mean and

Sensors2023,23, 3378 4 of 11 difference in time was used to rule out the proportional error. Furthermore, to investigate the effect of an early OS on the approximation between the bLT and sLT, we compared mean bLT–sLT (s) values between the EP and RP groups using an independentt-test. Based on the results of the Kolmogorov–Smirnov test, either an independentt-test or Mann–Whitney U test was applied to the age, height, weight, BMI, body water, BSA, room temperature, and relative humidity to compare each variable between the groups. All analyses were performed using IBM SPSS Statistics version 27 (IBM Corp., Armonk, NY, USA), with the statistical signi cance set at 0.05. 3. Results 3.1. Participants Characteristics and Physiological Results All participants completed all procedures and were eligible for the analysis. During exercise, continuous negligible responses were initially detected in the sweat rate, sweat lactate, and blood lactate (Figure). Subsequently, they rapidly increased and were de ned as the OS, sLT, and bLT, respectively. For all participants, these time points could be clearly de ned. Based on the relationship between these points, 17 participants (43%) were categorized into the EP group and 23 participants (57%) into the RP group. Furthermore, in ve participants (13%), the OS was later than the bLT. The descriptive data of the participants in the EP and RP groups are presented in Table. No signi cant differences were observed in any of the characteristics between the groups (p> 0.05). Table 1.Descriptive data of the participants. Total (n= 40) Early Perspiration Group (n= 17) Regular Perspiration Group (n= 23) p-Value Age (years) 21.8 4.0 22.4 4.6 21.4 3.5 0.71 BMI 22.4 2.1 22.7 1.9 22.2 2.3 0.48 BSA (m 2 ) 1.8 0.1 1.8 0.1 1.8 0.1 0.86 Body water (%) 57.8 5.4 56.8 5.6 58.6 5.3 0.31 Body fat ratio (%) 17.1 5.0 18.6 5.4 16.0 4.5 0.11 Muscle mass (kg) 53.6 5.8 53.0 5.3 54.0 6.3 0.61 RT ( C) 24.1 1.9 24.2 1.8 24.0 2.0 0.91 RH (%) 42.5 9.3 42.3 8.6 42.6 9.9 0.92 All values are presented as means standard deviations orn(%). Based on the

(%) 57.8 5.4 56.8 5.6 58.6 5.3 0.31 Body fat ratio (%) 17.1 5.0 18.6 5.4 16.0 4.5 0.11 Muscle mass (kg) 53.6 5.8 53.0 5.3 54.0 6.3 0.61 RT ( C) 24.1 1.9 24.2 1.8 24.0 2.0 0.91 RH (%) 42.5 9.3 42.3 8.6 42.6 9.9 0.92 All values are presented as means standard deviations orn(%). Based on the pre-performed Kolmogorov– Smirnov test, the independentt-test or Mann–Whitney U test was applied for intra-group comparisons. BSA, body surface area; RT, room temperature; RH, relative humidity. Table level, and local sweat rate. As in previous reports, the heart rate increased with the exercise intensity and the sweat rate increased after the onset of the incremental exercise. The sweat lactate levels were stable at rst and then rapidly increased from the sLT to the end. Table 2.Physiological data of the participants. Baseline At the Warm-Up Onset At the Incremental Exercise Onset At the Sweat Lactate Threshold At the End of Incremental Exercise Load (watt) 0.0 0.0 20.0 0.0 50.0 0.0 131.7 48.5 261.2 43.6 Heart rate (bpm) - 79.3 11.1 94.4 14.0 137.0 23.1 172.9 13.8 Sweat lactate ( A) 4.0 1.1 3.9 1.2 3.7 1.2 3.9 1.4 9.6 4.6 Local sweat rate (mg/cm 2 /min) 0.04 0.11 0.01 0.09 0.08 0.17 0.18 0.20 0.84 0.46 The baseline is the average of data in the rest period. All values are presented as means standard deviations.

Sensors2023,23, 3378 5 of 11 3.2. Relationship between OS and sLT Figurea shows the relationship between the OS and sLT. A poor correlation was observed in the total cohort (r = 0.38,p< 0.05), and no correlation was observed in the EP group (r = 0.12,p= 0.61). In contrast, in the RP group, the OS moderately correlated with the sLT (r = 0.56,p< 0.05). In the ve patients with the OS later than the bLT, the OS showed a strong correlation with the sLT (r = 0.82,p= 0.09). The Bland-Altman plot revealed that that the mean difference between the OS and sLT was large in total, especially in the EP group (total, 183.7 s, EP group, 389.6 s, RP group, 85.2 s), which validates the inconsistency between these thresholds (Figureb). Figureb also shows a positive xed error (p< 0.05, 95% CI: 155.1–274.1) and a proportional error (p< 0.05) in total. The xed error indicated that the sLT occurred after the OS in all the cases.Sensors 2023, 23, x FOR PEER REVIEW 5 of 11 Table 2. Physiological data of the participants. Baseline At the Warm-Up Onset At the Incremental Exercise Onset At the Sweat Lactate Threshold At the End of Incremental Exercise Load (watt) 0.0 ± 0.0 20.0 ± 0.0 50.0 ± 0.0 131.7 ± 48.5 261.2 ± 43.6 Heart rate (bpm) - 79.3 ± 11.1 94.4 ± 14.0 137.0 ± 23.1 172.9 ± 13.8 Sweat lactate (µA) 4.0 ± 1.1 3.9 ± 1.2 3.7 ± 1.2 3.9 ± 1.4 9.6 ± 4.6 Local sweat rate (mg/cm 2 /min) 0.04 ± 0.11 0.01 ± 0.09 0.08 ± 0.17 0.18 ± 0.20 0.84 ± 0.46 The baseline is the average of data in the rest period. All values are presented as means ± standard deviations. 3.2. Relationship between OS and sLT Figure 2a shows the relationship between the OS and sLT. A poor correlation was observed in the total cohort (r = 0.38, p < 0.05), and no correlation was observed in the EP group (r = 0.12, p = 0.61). In contrast, in the RP group, the OS moderately correlated with

This study investigates the implications of the onset of sweating on the sweat lactate threshold.