Sweat Lactate Response Between Males With High

Eur J Appl Physiol (2004) 91: 1–6 DOI 10.1007/s00421-003-0968-2

O R I GI N A L A R T IC L E

J. M. Green Æ R. C. Pritchett Æ T. R. Crews J. R. McLester Jr. Æ D. C. Tucker

Sweat lactate response between males with high and low aerobic fitness

Accepted: 18 August 2003 / Published online: 9 October 2003
Springer-Verlag 2003

Abstract Sweat lactate indirectly reflects eccrine gland metabolism. However the potential influence of aerobic fitness on sweat lactate is not well-understood. Six males with high aerobic fitness [peak oxygen consumption (V_ O2peak): 61.6 (2.5) mlÆkg)1Æmin)1] and seven males with low aerobic fitness [V_ O2peak: 41.8 (6.4) mlÆkg)1Æmin)1] completed a maximal exertion cycling trial followed on a different day by 60 min of cycling (60 revÆmin)1) in a 30C wet bulb globe temperature environment. Intensity was individualized at 90% of the ventilatory threshold (V_ E/V_ O2 increase with no concurrent V_ E/V_ CO2 increase). Sweat samples were collected from the lumbar region every 10 min and analyzed for lactate concentration. Sweat rate (SR) was significantly greater (p<0.05) for subjects with a high [1445 (254) mlÆh)1] versus a low [1056 (261) mlÆh)1] fitness level. Also, estimated total lactate excretion (SR·mean sweat lactate concentration) was marginally greater (p=0.2) in highly fit males. However, repeated measures ANOVA showed no significant differences (p>0.05) between groups for sweat lactate concentration at any time point. Current results show highly fit (vs. low fitness level) males have a greater sweat rate which is consistent with previous literature. However aerobic fitness and subsequent variations in SR do not appear to influence sweat lactate concentrations in males. Keywords Eccrine glands Æ Lactic acid Æ Sweat composition

J. M. Green (&) Æ R. C. Pritchett Æ T. R. Crews J. R. McLester Jr. Æ D. C. Tucker Department of Physical Education and Recreation, Western Kentucky University, 1 Big Red Way, Bowling Green, KY 42101, USA E-mail: [email protected] Tel.: +1-270-7456035 Fax: +1-270-7456043

Introduction Lactate in sweat indirectly reflects eccrine gland metabolism (Gordon et al. 1971; Green et al. 2000a; Sato 1977; Wolfe et al. 1970). Adenosine triphosphate (ATP) production is required within heat-activated ‘‘eccrine’’ sweat glands for myoepithelial cell contraction, which supports the expulsion or secretion of sweat (Sato 1977). While eccrine glands contain an abundant number of mitochondria and produce ATP primarily aerobically (Sato 1977), the presence of lactate is indirect evidence that some of the ATP results from oxygenindependent metabolism (Gordon et al. 1971; Green et al. 2000a; Sato 1977; Wolfe et al 1970). Although concentrations of various electrolytes in sweat have been rigorously investigated, studies utilizing sweat lactate measurements to examine the metabolic characteristics of eccrine glands are comparatively sparse. Although less specific than actual ATP measures, this constituent provides an indirect method for assessing variations in the metabolic tendencies of sweat glands. For example, in a study by Green et al. (2000b) sweat rates (SR) were significantly greater in males, with no significant differences between males and females for sweat lactate concentrations. Because females possess a greater number of heat-activated sweat glands (Bar-Or et al. 1968) it was concluded that there may be potentially different metabolic tendencies between eccrine glands in males and females (Green et al. 2000b). Based on similar SR and sweat lactate concentrations, it was also speculated that younger and middle-aged males with similar relative peak oxygen consumption (V_ O2peak) values demonstrate similar metabolic tendencies in eccrine glands (Green et al. 2001). Few studies have compared sweat lactate responses relative to aerobic fitness. Lamont (1987) found significantly lower SR and significantly greater sweat lactate concentrations in females with lower versus higher V_ O2peak values. Higher sweat lactate concentrations were attributed in part to a less dilute sweat, the result of

2

a lower SR. For combined subjects in that study, a significant (p<0.01) inverse relationship (r=)0.79) was found between SR and sweat lactate concentration. Potential fitness-associated differences in sweat lactate in males have not been investigated. Because of SR differences and the typically lower density of heat-activated sweat glands in males (Bar-Or et al. 1968), it is uncertain whether the results of Lamont (1987) will hold true for both genders. Additionally, in Lamont (1987) a single overall sweat lactate measurement was taken which did not allow an evaluation of the sweat lactate response across time. This may be important because not only are SR greater but the onset of sweating occurs more quickly in fitter subjects (Yukitoshi et al. 1997), which may alter the dilution of sweat and consequently sweat lactate concentrations particularly in the early stages of an exercise bout. Serial measurements would allow potential time-dependent changes in sweat lactate concentrations to be evaluated from the beginning of exercise. This paradigm provides a more thorough examination of sweat lactate response compared to relying on a single overall measurement. Because previous research has been limited to females, the current study compared sweat lactate concentrations between males with high and low aerobic fitness levels cycling in a hot [30C wet bulb globe temperature (WBGT)] environment. Also, in order to detect potential differences across time, multiple independent sweat samples were collected at 10-min intervals and analyzed for lactate concentration.

Methods Subjects/descriptive data College-aged males (n=13) were recruited for participation. Prior to data collection, each subject completed a written informed consent form outlining requirements for participation. All procedures were approved by the Human Subjects Review Board at Western Kentucky University. Each subject arrived at the laboratory with instructions to be well-hydrated, at least 3 h post-prandial and to have abstained from caffeine and alcohol for a minimum of 24 h. Age (years), height (cm) and mass (kg) were measured and recorded. Body fat percentage was estimated using Lange skinfold calipers (Cambridge, Md., USA) and a three-site method (chest, abdomen, thigh) (Jackson and Pollock 1985).

V_ O2peak trial Following descriptive data collection, each subject completed a maximal exertion cycling trial to determine V_ O2peak. Seat height was appropriately set for each individual on a Monark cycle Ergometer (Varberg, Sweden) and handlebars were adjusted based on individual preference. Subjects were instructed to maintain a cadence of 60 revÆmin)1 using a Franz XB 700 metronome (Franz, New Haven, Conn., USA). Subjects were fitted with an appropriately-sized air-cushioned face-mask (Vacu-med, Ventura, Calif., USA) and with a Polar heart rate (HR) monitor transmitter (Stamford, Conn., USA) at the level of the sternum. Subjects pedaled at the required cadence for 3 min at 0 W. At the conclusion of this stage the workload of the ergometer was increased by 25 WÆmin)1 until subjects achieved volitional exhaustion or could no longer maintain the required 60 revÆmin)1. Metabolic data

[V_ O2, carbon dioxide output (V_ CO2), respiratory exchange ratio (RER), minute ventilation (V_ E)] were collected using a Vacumed Vista mini-cpx (silver) metabolic measurement system (Vacu-med, Ventura, Calif., USA). Turbofit software, designed for use with the metabolic system, was set to report mean metabolic data over 15-s time periods. The system was calibrated prior to each test with a gas of known composition. A 3-l syringe (Hans Rudolph, Kansas City, Mo., USA) was used to calibrate the system for measurement of ventilation. HR response was collected using a receiver interfaced with the computer and Turbofit software. Ratings of perceived exertion (RPE) were collected during the last 15 s of each stage using the original Borg category (6–20) scale. Criteria for achievement of V_ O2 peak were: (1) RPE‡18, (2) RER‡1.1, (3) plateau of V_ O2 with increased workload and (4) 85% of age-predicted maximum HR (Maud and Foster 1995) Each subject met at least two of these four criteria. Following V_ O2peak trials subjects having a V_ O2peak ‡58 mlÆkg)1Æmin)1 (n=6) were placed in the high aerobic fitness group. Subjects having a V_ O2 peak £ 49 mlÆkg)1Æmin)1 (n=7) were placed in the low aerobic fitness group. Subjects having a V_ O2peak between these values were excluded from participation in the study.

Cycling at 30C WBGT Within 7 days of their V_ O2peak trial, male subjects with high and those with low aerobic fitness reported to the laboratory with instructions to be well-hydrated, at least 3 h post-prandial and not having consumed caffeine or alcohol for a minimum of 24 h. In the environmentally controlled room subjects were fitted with a Polar HR monitor at the level of the sternum for evaluation of the HR response. Rectal temperature (Tre) was assessed during this trial using a rectal thermistor inserted 8 cm beyond the rectal sphincter. The thermistor was connected to a Pysitemp Thermalert TH-8 (Physitemp, Clifton, N.J., USA) which measured Tre to the nearest 0.1C. Within 5 min following this preparation, subjects pedaled at 60 revÆmin)1 for a total of 60 min in an environmentally controlled room [30 (1)C WBGT, (wet bulb: 27.4 (1.0), dry bulb: 36.0 (1.0), globe: 36.7 (1.0)]. The intensity for each individual was established using data from V_ O2peak cycling trials. The ventilatory threshold was estimated by a minimum of two experienced investigators. The point where V_ E/V_ O2 demonstrated an abrupt increase with no concurrent increase in V_ E/V_ CO2 was identified for each subject (Caiozzo et al. 1982). The resistance associated with 90% of this point was utilized for 60-min cycling trials. HR and Tre were recorded at 10, 20, 30, 40, 50 and 60 min. Initial Tre at 0 min was measured.

Sweat collection/analysis Sweat samples were collected using a modified method of Brisson et al. (1991). An 8 cm · 9 cm piece of impermeable parafilm (American Can, Greenwich, Conn., USA) was cut and placed on the adhesive side of a 10 cm · 14 cm Opsite Wound Dressing (Smith and Nephew, Largo, Fla., USA). This collection device was positioned on the lower lumbar region of each subject just prior to the beginning of the 60-min trial. At each specified time interval the device was peeled from the skin at the top and the desired volume of uncontaminated sweat was collected using microhematocrit capillary tubes (Chase Scientific Glass, Rockwood, Tenn., USA). After each sample was collected the Opsite device was removed, the skin of the lower back was dried with a clean cotton towel and a new sweat collection device was positioned in the same location. This method has been utilized in previous studies (Green et al. 2000a, 2000b, 2001) to eliminate contamination of serial samples. Also, because the parafilm is impermeable, sweat samples are altered minimally by evaporation. Each sweat sample was analyzed for lactate concentration using a YSI 1500 Sport Lactate Analyzer (Yellow Springs Instruments, Yellow Springs, Ohio, USA). Prior to

3 each test the analyzer was calibrated with a 5 mmolÆl)1 standard and checked for linearity using 30 mmolÆl)1. Because the analyzer range was 35 mmolÆl)1 and initial sweat lactate concentrations (prior to 20 min) may exceed this concentration, dual samples were taken at the 10 min time point. If an initial analysis was out of range, a sample was diluted with equal parts of de-ionized water, analyzed using the YSI 1500 Sport and the result was doubled. The efficacy of this procedure was verified by diluting and analyzing multiple, simultaneously collected samples within the range of the analyzer. Pre- and post-test body mass (to the nearest 0.1 kg) were recorded and used to calculate SR. SR values were corrected for respiratory water loss and metabolic water loss using the formula: respiratory water loss = 0.019(V_ O2)·(44)ambient water vapor pressure)·time with V_ O2 expressed in l min)1, ambient water vapor pressure expressed in mmHg, and time expressed in minutes (Mitchell et al. 1972). No corrections were made for fluid consumption, as subjects were not allowed to consume fluids during testing. Also, no corrections were required for urinary water loss.

Statistical analyses Descriptive characteristics were compared between groups using a multivariate analysis of variance (MANOVA) with a Bonferroni follow-up procedure. SR was compared using an independent Ttest. HR and sweat lactate concentrations were compared separately using a 2 (group) · 6 (time) repeated measures analysis of variance (ANOVA) for each variable. Tre was analyzed using a 2 (group) · 7 (time) repeated measures ANOVA. V_ O2 at 90% of the ventilatory threshold and the workload (watts) to achieve this workload were compared between groups using an independent samples T-test for each variable. Results were considered significant at p £ 0.05. Total sweat lactate excretion was calculated by multiplying sweat rate (l min)1) by mean sweat lactate concentration (for six time points). This was compared between groups using an independent T-test.

group (Table 1). There was no significant main effect for sweat lactate concentration between groups (Fig. 1). Similarly, for HR (Fig. 2) and Tre (Fig. 3) there were no significant main effects. Visual inspection of Tre responses between groups (Fig. 3) suggested that the core temperature of highly fit subjects increased more overall than that of subjects with a low fitness level. Therefore DTre (Tre at 60 min minus Tre at 0 min) was calculated for each individual and mean values between groups were compared using an independent T-test. DTre was significantly greater in highly fit subjects [1.42 (0.26)] compared to subjects with a low fitness level [0.91 (0.37)]. SR (mlÆh)1) was significantly greater in male subjects with a high versus those with a low fitness level (Fig. 4). Total lactate excretion was higher for highly fit subjects [15.6 (2.3) mmol] compared to subjects with low aerobic fitness [13.6 (3.6) mmol] but the difference only approached significance (p=0.2).

Discussion Current literature lacks information regarding potential fitness-associated differences in sweat lactate concentration

Results Descriptive data for subjects with high (n=6) and low (n=7) aerobic fitness levels are presented in Table 1. Age (years), height (cm) and body fat (%) were not significantly different. Weight was significantly lower and V_ O2peak was significantly greater in males with a high fitness level. V_ O2 at 90% of the ventilatory threshold and workload at 90% of the ventilatory threshold were significantly greater for the highly fit

Fig. 1 Mean sweat lactate concentrations (mmolÆl)1) between males with high and with low fitness levels across time. No significant differences (p>0.05) between groups were found

Table 1 Descriptive characteristics of subjects with high and with low fitness levels. Values are means and standard deviations Variable

Age (years) Height (cm) Mass (kg) Body fat (%) V_ O2peak (mlÆkg)1Æmin)1) 90% of Ventilatory threshold (mlÆkg)1Æmin)1) Workload at 90% VT *p<0.05 between groups

High fitness level (n=6)

Low fitness level (n=7)

Mean

SD

Mean

SD

25.2 179.0 72.3* 9.7 61.6* 37.0*

3.7 5.9 6.5 3.5 2.5 7.0

22.4 182.9 89.9 12.4 41.8 21.3

0.8 9.0 10.4 5.8 6.4 2.5

210*

27

144

22

Fig. 2 Mean heart rate (beatsÆmin)1) between males with high and with low fitness levels across time. No significant differences (p>0.05) between groups were found

4

Fig. 3 Mean rectal temperature (ºC) between males with high and with low fitness levels across time. No significant differences (p>0.05) between groups were found

in males. The results of the present study indicate that no significant differences in sweat lactate concentration exist between males with high and low aerobic fitness (Fig. 1). Lactate concentrations in sweat were similar irrespective of the significantly greater SR in highly fit males (Fig. 4). Current results for SR (Fig. 4) parallel those of Lamont (1987). Lamont (1987) found significantly lower total sweat lactate secretion and sweat lactate concentration in females with a high fitness level compared to females with a lower one. It was concluded that significantly greater sweat rates in fitter females accounted for the lower lactate concentration. Conversely, sweat lactate concentrations in the current study were not significantly different between the two groups (Fig. 1). Additionally, total lactate excretion was marginally greater (p=0.2) in subjects with high versus those with a low fitness level owing to the greater SR. These discrepancies may center on methodological differences.

Fig. 4 Mean sweat rates (mlÆh)1) between males with high and with low fitness levels. *p<0.05 between groups

Lamont (1987) utilized a whole body wash-down technique while the current study collected and analyzed serial samples in impermeable sweat collection devices (Brisson et al. 1991) and then calculated overall sweat lactate excretion as a function of SR. Patterson et al. (2000) suggested that the wash-down technique may be inadequate for assessing the whole body sweat lactate concentration. Additionally, a single measurement from the wash down technique does not permit changes across time to be evaluated. Sweat collection devices tightly control for evaporation and allow multiple sampling. However, regional differences in sweat composition go undetected unless multiple sites are tested. Sweat lactate concentrations provide indirect information regarding metabolic ATP production supporting eccrine sweat secretion (Gordon et al. 1971; Green et. al.2000a; Wolfe et al. 1970). While this study did not measure intraglandular ATP production directly, it can be speculated that the similar lactate concentrations in the two groups (Fig. 1) indicate that the relative aerobic and oxygen-independent ATP contributions may have been similar in aerobically fit and unfit subjects. This, however, must be interpreted in light of the significantly greater SR in highly fit subjects (Fig. 4). It is possible that, with improved aerobic fitness, sweat glands systematically alter aerobic and oxygen-independent ATP contributions, resulting in similar sweat lactate concentrations. A second possible explanation is that the highly fit subjects simply possessed a larger density of heatactivated sweat glands with minimal differences in metabolic tendencies. With heat acclimation a greater number of glands are incorporated and sweat output per gland increases as well (Yukitoshi et. al. 1997). While the state of acclimation was not evaluated, data for the current investigation were collected during very warm months (May through September in the Southeast United States) and it can be safely assumed that greater SR in fitter subjects resulted in part from chronic exercise in a warm environment. Acclimation has been shown to decrease sweat lactate concentration Kuono (1956); however, the mechanism behind this change remains unclear. Considering the current results it is likely that a combination of the above hypotheses may account in part for the similarities in sweat lactate concentrations between males of high and low aerobic fitness. Circulating catecholamines, which are intensity dependent, influence sweat production (Sato 1977) and consequent changes in glandular metabolism and sweat lactate concentrations might be expected. Although this possibility was not investigated in the current study, the fact that there were no significant between-group differences in HR response (Fig. 2) or Tre (Fig. 3) indicates that the two groups of subjects experienced similar relative cardiovascular and thermal strain. As pointed out in the Results, the change in Tre was significantly greater in highly fit subjects likely resulting from greater absolute heat production because a higher workload was required in this group [210 (27) versus 144 (22) W] to

5

achieve the desired 90% of ventilatory threshold. Conclusions regarding the potential influence of circulating catecholamines on sweat lactate responses in the current study would only be speculative. However, this possibility deserves attention. Other potential factors influencing sweat lactate concentration should also be mentioned. The lactate concentration in sweat reflects secreted lactate. It is possible that some lactate is reabsorbed in the duct of the eccrine glands prior to reaching the epidermis. The lactate concentration in sweat is unaffected by increases in blood lactate (Green et al. 2000a). Because sweat lactate concentrations are consistently greater than blood lactate concentrations, lactate transport from blood to sweat would necessitate movement against a concentration gradient. However, lactate movement from the duct of the eccrine gland into the plasma compartment could occur by passive diffusion. Little is known regarding ductal reabsorption of lactate. It is possible, however, that differences exist between those with high and low aerobic fitness. The potential fate of lactate in a sweat gland may also be affected by mitochondrial uptake of lactate. Eccrine glands contain mitochondria (Sato 1977) responsible for aerobic ATP production. Lactate is likely a viable fuel source for these structures. No study to date has determined whether heat acclimation or enhanced aerobic fitness increases the size and density of mitochondria within eccrine glands similar to training-induced skeletal muscle mitochondrial adaptations. It is plausible to believe that excessive sweating might provide an adequate stimulus for increasing the oxidative capacity of eccrine glands via mitochondrial adaptations. Changes in mitochondrial density could influence lactate uptake in the eccrine glands. Such an alteration would go undetected when sampling in vivo at the level of the skin. Skin blood flow represents an influential factor affecting sweat lactate concentration. Blood flow occlusion decreases SR and increases sweat lactate concentration (Astrand 1963; Weiner and Van Heyningen 1952). This is not surprising as eccrine glands rely primarily on the oxidative metabolism of circulating glucose for fuel (Sato 1977). Decreasing oxygen and fuel delivery attenuates the capacity to produce ATP which limits sweat secretion. This relative hypoxia stimulates greater oxygen-independent ATP production and more resulting lactate (Astrand 1963). Additionally, decreased sweat production creates a less dilute sweat subsequently elevating the lactate concentration. Fritzsche and Coyle (2000) showed that cutaneous blood flow was higher in endurance-trained subjects. Greater blood and oxygen delivery to the eccrine coil could theoretically enhance aerobic ATP production possibly altering sweat lactate concentration. The current results show a consistently (but not significantly) lower sweat lactate concentration for highly fit males particularly at the onset (initial 10 min) of exercise (Fig. 1). The enhanced cardiovascular system typical of highly fit subjects may have more effectively perfused the cutaneous tissue and eccrine glands with oxygen especially at the onset of exercise.

Such a paradigm could be responsible for the consistently lower mean lactate concentrations in highly fit subjects (Fig. 4). Skin blood flow was beyond the scope of this investigation but would strengthen this argument. As mentioned, multiple sweat samples allow the evaluation of sweat lactate concentration responses across time. Consistent with other studies (Ament et al. 1997; Green et al. 2000a, 2000b, 2001), the sweat lactate concentration in the current study was initially high and reached a plateau within 30 min (Fig. 4) regardless of fitness. The reasons for this pattern of sweat lactate are not well-understood. Sweat lactate production or reabsorption at the duct may differ during the preliminary period of sweating resulting in variations in lactate concentration in secreted sweat. Additionally, a finite time is required for the glands to begin producing significant volumes of sweat. During this time, lower relative sweat production brings about less dilute sweat. This, coupled with potentially different lactate production and/or reabsorption early during sweating, creates a twofold explanation for initially high sweat lactate concentrations. An evaluation of the current results lends support to the above hypotheses. Figure 1 shows a convergence of lactate concentrations between subjects with high and with low fitness levels across time. Sweat lactate concentrations were somewhat higher in subjects with low fitness levels, particularly at the onset of exercise. However, values became almost identical at 30 min. In this group it is plausible that eccrine glands follow a different course in realizing a plateau in sweat lactate concentration. Additionally, highly fit subjects are known to have a lower threshold for sweating (Yukitoshi et al. 1997), meaning that sweat volume is not only greater but is initiated more quickly. Thus, highly fit subjects may have had more dilute sweat due to greater initial sweat volume. The relative metabolic tendencies of eccrine sweat glands in males with high and low aerobic fitness exercising at similar relative intensities in a 30C WBGT environment may be comparable. However more indepth evidence (direct measurements of ATP) is necessary to make definitive conclusions. The current results show males with a high fitness level have a greater SR than males that are less fit, which is consistent with previous literature. However, aerobic fitness and subsequent variations in SR do not appear to influence sweat lactate concentrations in males. Acknowledgements The authors wish to thank Smith and Nephew for their generous donation of the Opsite Wound Dressings used in the current study. Also, appreciation is extended to the Faculty Scholarship Committee at Western Kentucky University for support of the current project.

References Ament W, Huizenga JR, Mook GA, Gips CH, Verkerke GJ (1997) Lactate and ammonia concentration in blood and sweat during incremental exercise. Int J Sports Med 18:35–39

6 Astrand I (1963) Lactate content in sweat. Acta Physiol Scand 58:359–367 Bar-Or O, Lundegren HM, Magnusson LI, Buskirk ER (1968) Distribution of heat-activated sweat glands in obese and lean men and women. Human Biol 40:235–248 Brisson GR, Boisvert P, Peronnet F, Perrault H, Boisvert D, Lafond JS (1991) A simple and disposable sweat collector. Eur J Appl Physiol 50:405–411 Caiozzo VJ, Davis JA, Ellis JF, Azus JL, Vandagriff CA, Prietto CA, McMaster WC (1982) A comparison of gas exchange indices used to detect the anaerobic threshold. J Appl Physiol 53:1184–1189 Fritzsche RG, Coyle EF (2000) Cutaneous blood flow during exercise is higher in endurance-trained humans. J Appl Physiol 88:738–744 Gordon RS Jr, Thompson RH, Muenzer J, Thrasher D (1971) Sweat lactate in man is derived from blood glucose. J Appl Physiol 31:713–716 Green JM, Bishop PA, Muir IH, McLester JR Jr, Heath HE (2000a) Effects of high and low blood lactate concentrations on sweat lactate response. Int J Sports Med 21:1–5 Green JM, Bishop PA, Muir IH, Lopeak RG (2000b) Gender differences in sweat lactate. Eur J Appl Physiol 82:230–235 Green JM, Bishop PA, Muir IH, Lopeak RG (2001) Lactate-sweat relationships in younger and middle-aged men. J Aging Phys Act 9:67–77

Jackson AS, Pollock ML (1985) Practical assessment of body composition. Phys Sports Med 13:76–90 Kuono Y (1956) Human perspiration. Thomas, Illinois Lamont L (1987) Sweat lactate secretion during exercise in relation to women
s aerobic capacity. J Appl Physiol 62:194–198 Maud PJ, Foster C (eds) (1995) Physiological assessment of human fitness. Human Kinetics, Champaign, Ill. Mitchell JM, Nadel ER, Stolwijk JAJ (1972) Respiratory weight loss during exercise. J Appl Physiol 32:474–476 Patterson MJ, Galloway SDR, Nimmo MA (2000) Variations in regional sweat composition in normal human males. Exp Physiol, 85:869–875 Sato K (1977) The physiology, pharmacology, and biochemistry of the eccrine sweat gland. Rev Physiol Biochem Pharmacol 79:51–131 Weiner JS, Van Heyningen RE (1952) Observations on lactate content of sweat. J Appl Physiol 4:733–744 Wolfe S, Cage G, Epstein M, Tice L, Miller H, Gordon RS (1970) Metabolic studies of isolated human eccrine sweat glands. J Clin Inv 49:1880–1884 Yukitoshi A, McLellan TM, Shephard RJ (1997) Interactions of physical training and heat acclimation. Sports Med 23:173–210