Sex differences in postsynaptic sweating and cutaneous vasodilation

Ethical approval.

The current experimental protocol was approved by the University of Ottawa Health Sciences and Science Research Ethics Board, as well as the Institutional Review Boards of the University of Texas Southwestern Medical Center and Texas Health Presbyterian Hospital, Dallas. Written informed consent was obtained from all volunteers prior to their participation in the study.

Participants.

Reported sex differences in sweating during pharmacological stimulation (21) were used to calculate (β = 0.8, α = 0.05) a minimum sample size of 11 participants in each group, assuming an effect of 30% and standard deviation of 25%. Twenty-four participants, 12 males and 12 females, volunteered for the study. Fourteen of the participants (7 males, 7 females) performed the experiments at the Institute for Exercise and Environmental Medicine (IEEM) at Texas Health Presbyterian Hospital Dallas and University of Texas Southwestern Medical Center, while the remaining 10 participants (5 males, 5 females) performed the experiments at the Human and Environmental Physiology Research Unit (HEPRU) at the University of Ottawa. To minimize the influence of differences in hormonal status across the menstrual cycle, female participants performed each experimental session within the first and tenth day after the onset of their self-reported menses. Female participants taking oral contraceptives (n = 4) performed the experimental sessions during the no pill/placebo phase of oral contraceptive use. Hormonal status was confirmed by taking a venous blood sample on the day of each experimental session. None of the experimental sessions for female participants were withdrawn or repeated based on blood sample results. Participants were healthy, nonsmoking, and free of any known cardiovascular, metabolic, and respiratory diseases. Participant characteristics are presented in Table 1.

Measurements.

Local sudomotor activity was measured on the dorsal forearm using plastic capsules attached to the skin with adhesive rings and topical skin glue (Collodion HV; Mavidon Medical products, Lake Worth, FL). The sweat capsule at site 2 also housed the laser-Doppler flow probe (see details below), allowing for the simultaneous measurement of local sweat rate and cutaneous blood flow. Anhydrous compressed nitrogen was passed through each capsule at a known flow rate (HEPRU: 1 l/min; IEEM: 0.3 l/min). Water content of the effluent air was measured using high precision dew point mirrors (HEPRU: RH systems, Albuquerque, NM) or capacitance hygrometers (IEEM: Vaisala, Woburn, WA). Local sweat rate was calculated using the difference in water content between effluent and influent air multiplied by the flow rate and normalized for the skin surface area under the capsule.

The number of active sweat glands was measured following the 25 min period of continuous infusion at the greatest dose employed during experimental session 1, as well as at the end of passive heating during experimental session 2 using the modified iodine-paper technique with computer-assisted analysis (7, 9). The sweat output per gland was calculated by dividing sweat rate at the corresponding measurement period by the number of active sweat glands.

Local cutaneous blood flow was estimated using laser-Doppler velocimetry (PeriFlux System 5000; Perimed, Stockholm, Sweden). Prior to the start of the experimental trial, laser-Doppler flow probes (integrating probe 413; Perimed, Stockholm, Sweden) were affixed with an adhesive ring to the forearm in a site that demonstrated cardiosynchronous pulsatile activity. Cutaneous vascular conductance was subsequently calculated as the ratio of cutaneous blood flow perfusion units to mean arterial pressure and expressed as a percentage of maximum, as determined during combined local heating and SNP infusion (experimental session 1) or during local heating alone (experimental session 2).

Esophageal temperature was measured with a general purpose thermocouple temperature probe (Mallinckrodt Medical, St-Louis, MO). The probe was inserted 40 cm past the entrance of the nostril while the participants sipped water (100–300 ml) through a straw. Skin temperature was measured at 10 (HEPRU) or 6 (IEEM) sites using thermocouples attached to the skin with surgical tape. Mean skin temperature was subsequently calculated using a 10- or 6-point weighting of the regional proportions determined by Hardy and DuBois (14). In 10 participants (5 males and 5 females), the thermocouples were integrated into heat flow sensors (HEPRU: Concept Engineering, Old Saybrook, CT) to estimate rate of dry heat exchange. Data were either collected with a data acquisition system at 50 Hz (IEEM: Biopac System, Santa Barbara, CA), or with a HP Agilent data acquisition module (HEPRU: model 3497A) at a rate of one sample every 15 s and simultaneously displayed and recorded in spreadsheet format on a personal computer with LabVIEW software (HEPRU: version 7.0; National Instruments, TX).

Systolic and diastolic blood pressures were measured using an automated blood pressure monitor (Tango+; SunTech Medical, Morrisville, NC) and used to calculate mean arterial pressure (diastolic blood pressure + 1/3 × pulse pressure).

Urine specific gravity was determined in duplicate using a handheld refractometer (HEPRU: TS400, Reichter, Depew, NY; IEEM: PAL-10S, Atago, Japan).

On the day of each experimental session, a venous blood sample (10 ml) was obtained from female participants to confirm that the session occurred in the follicular/low hormone phase of the menstrual cycle. The blood samples were collected with a SST vacutainer (BD Vacutainer; Franklin Lakes, NJ) for the determination of plasma 17β-estradiol and progesterone. Plasma concentrations of 17β-estradiol and progesterone were quantified by independent external laboratories (HEPRU: Gamma-Dynacare Medical Laboratories, Ottawa, Ontario, Canada; IEEM: ARUP Laboratories, Salt lake City, UT).

Experimental design.

Participants volunteered for one preliminary and two experimental sessions. For both experimental sessions, participants reported to the laboratory between 0700 and 1000 and provided a urine sample for the measurement of urine specific gravity. The participants were asked to drink 500 ml of water the night prior to, as well as the morning of, each experimental session and to refrain from alcohol, caffeine, and exercise 24 h prior to experimentation. The two experimental sessions were performed on separate days, separated by a minimum of 24 h. Possible differences in acclimatization status between sexes were not taken into account, and all experimentation occurred between the months of September and April. Both experimental sessions were performed at an ambient room temperature of 22–24°C.

During the preliminary session, training history, body height, mass, and density, as well as maximum oxygen uptake were determined. Since the cholinergic sensitivity of the sweat gland has been associated with an individual's training status (2), we recruited males and females with a similar training history, which was assessed by having the participants quantify their physical activity levels using the quantitative (3 mo) and 7 day physical activity recall questionnaires proposed by Kohl et al. (20). Body height was determined using a stadiometer, while body mass was measured using a scale. Body surface area was subsequently calculated from the measurements of body height and mass (8). Body density was measured using the hydrostatic weighing technique and used to calculate body fat percentage (27). Maximum oxygen uptake was determined by indirect calorimetry (HEPRU: Moxus system, Applied Electrochemistry, Pittsburgh, PA; IEEM: Parvo medics system, Parvo Medics, Sandy, UT) during a progressive incremental exercise protocol (4) performed on a treadmill (HEPRU) or cycle ergometer (IEEM).

For the first experimental session, participants were placed in a semirecumbent position on a bed, following which three microdialysis membranes were inserted by advancing a 25 gauge needle 15 to 20 mm through the dermal layer of dorsal forearm skin. The microdialysis probe was subsequently threaded through the lumen of the needle, which was subsequently withdrawn, leaving in place a 1 cm dialysis membrane within the dermal layer (Bioanalytical Systems, West Lafayette, IN). All microdialysis probes were first perfused with lactated Ringer solution (Baxter, Deerfield, IL) at a rate of 4 μl/min via a perfusion pump (HEPRU: model 400, CMA Microdialysis, Solna, Sweden; IEEM: Harvard Apparatus, Holliston, MA) for 120 min while hyperemia associated with insertion trauma subsided. During this time, the area directly over one microdialysis membrane was instrumented for the measurement of local sweat rate (site 1), a second was instrumented for the dual measurement of sweat rate and cutaneous vasodilation (site 2), while the third was instrumented for the measurement of cutaneous vasodilation (site 3). Dose-response curves for sweating were subsequently assessed at sites 1 and 2 to incremental doses of MCh and ACh, respectively (Sigma-Aldrich, St-Louis, MO). Both MCh and ACh were infused in 10-fold increments, from 1 × 10⁻⁶ M to 1 M. Dose-response curves for cutaneous vasodilation were assessed at sites 2 and 3 to incremental doses of ACh and SNP, respectively (Hospira, Lake Forest, IL). SNP was infused in 10-fold increments, from 5 × 10⁻⁸ M to 5 × 10⁻² M. Each dose was initially primed through the microdialysis membrane at an infusion rate of 102 μl/min for 1 min, following which each dose was infused for 10 min at a rate of 4 μl/min. The highest dose (i.e., 1 M ACh and MCh and 5 × 10⁻² M SNP) was infused for an additional 15 min (total of 25 min) to ensure a steady-state response to the greatest concentration employed. At the end of the experimental protocol, maximum cutaneous blood flow at the site receiving ACh (site 2) was determined by infusing SNP (5 × 10⁻² M), while the local heater at the site receiving SNP (site 3) was set to 44°C with continuous infusion of SNP for an additional 20 min.

For the second experimental session, the participants donned a liquid-perfusion garment (Allen-Vanguard, Ottawa, ON, Canada) which covered the entire body except for the feet, hands, and the forearm from which local sweat rate and cutaneous vasodilation were measured. Participants were then placed in a semirecumbent position on a bed while water at 34°C was perfused through the suit for a 30 min instrumentation period. Following instrumentation, a 20 min baseline period was performed with water at 34°C circulating through the suit, following which the temperature of the water was increased to 48°C and heating continued until an increase in esophageal temperature of ∼1.4°C. At the end of heating, the temperature of the water was decreased to cool the participant while maximum cutaneous blood flow was determined by increasing the temperature of the heater housing the laser-Doppler probes to 44°C for 30 min. For this experimental session, clothing beneath the liquid-perfusion garment was standardized to shorts only for males and to shorts and a sports bra for female participants.

Data analysis.

For each participant, dose-response curves were generated by plotting sweat rate/cutaneous vascular conductance over the log concentration of the agonist and fitted using a nonlinear regression analysis to determine minimal and maximal values of the relationship, as well as the log concentration of the agonist causing 50% of the maximal response (log EC₅₀). All dose-response analyses were performed using a standardized slope of 1 (6, 19). During the passive heating session, local sudomotor activity and skin blood flow were measured at two sites, the average of both being used for the statistical analyses. For all variables, minute averages were performed for the statistical analyses. To account for the relative influence of core and skin temperatures on the activation of heat loss responses (15, 22, 30), mean body temperature was calculated as: 0.8 × esophageal temperature + 0.2 × mean skin temperature (29). The onset threshold and thermosensitivity of local sudomotor activity and cutaneous vascular conductance were determined using the linear portion of each response plotted against mean body temperature and analyzed using segmented regression analysis (5).

Statistical analysis.

All dependent variables were compared between groups (males vs. females) within each experimental session. An extra-sum-of-squares F test was used to determine whether the best fit values of the dose-response curves differed between males and females. Independent samples T-tests were used for single comparisons between groups, while a two-way mixed model analysis of variance was used for multiple comparisons between groups using the repeated factor of agonist concentration (experimental session 1) or change in mean body temperature (experimental session 2) and the nonrepeated factor of sex. When a significant main effect was observed, post hoc comparisons were carried out with the Holm-Bonferonni approach. The level of significance for all analyses was set at an alpha level of P ≤ 0.05. Statistical analyses were performed using commercially available statistical software (SPSS 19.0 for Windows, SPSS, Chicago, IL). Dose-response and segmented regression analyses were performed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA). Participant characteristics are presented as mean ± standard deviation, while all variables are reported as mean ± 95% confidence intervals. Confidence intervals were calculated as 1.96 × standard error of the mean.

Participant characteristics.

Males had a significantly greater height (P = 0.01), weight (P = 0.003), body surface area (P = 0.002), and maximum oxygen consumption both in absolute values (P ≤ 0.001) as well as relative to body mass (P = 0.02). However, there was no significant difference between groups in maximum oxygen consumption expressed relative to fat free mass (P = 0.11).

Training history.

During the 3 mo prior to their participation in the study, males reported engaging in physical activities long enough to work up a sweat on average 5 ± 2 (range: 1–10) times per week, compared with 5 ± 1 (range: 2–7) for females (P = 0.81). On average, males exercised 7 ± 3 (range: 3–14) h/wk, while females exercised 6 ± 3 (range: 2–11) h/wk (P = 0.60). During the 7 days prior to their participation in the study, males and females reported performing vigorous physical activity for 4 ± 3 (range: 0–8) and 4 ± 3 (range: 1–10) h, respectively (P = 0.93). Males and females also reported performing moderate physical activity for 2 ± 2 (range: 0–5) and 4 ± 3 (range: 0–10) h, respectively (P = 0.10).

Postjunctional sweating and cutaneous vasodilation.

On the day of the experimental session, plasma concentrations of 17β-estradiol and progesterone for the female participants averaged 154 ± 102 (range: 46–406) pmol/l and 1.1 ± 0.4 (range: 0.6–1.8) nmol/l, respectively. Urine specific gravity averaged 1.019 ± 0.006 in males and 1.012 ± 0.007 in females (P = 0.01).

There were no significant differences in baseline sweat rate between males and females at the MCh and ACh sites (P > 0.1). Sweating increased as a function of increasing concentrations of MCh and ACh (P ≤ 0.001, Fig. 1). Furthermore, increases in sweating as a function of increasing MCh (P = 0.046) and ACh (P = 0.008) concentrations significantly differed between males and females (agonist concentration × sex interaction). For both agonists, sweat rate did not differ between sexes at the lower concentrations, but was greater in males at the greatest concentration employed (P ≤ 0.05, see Fig. 1). The best-fit values of the dose-response curves for sweating (Fig. 2) significantly differed between males and females at both the MCh [F_(3,162) = 6.685, P < 0.001] and ACh [F_(3,162) = 11.43, P < 0.001] sites. Minimum values (P > 0.1, Table 2), and the log EC₅₀ (P > 0.1, Table 2) did not differ between sexes. In contrast, maximum values were lower in females compared with males at both the MCh (P = 0.05, Table 2) and ACh (P = 0.04) sites. The number of active sweat glands at the greatest concentration employed was significantly greater in females (56 ± 7 glands/cm²) compared with males (47 ± 5 glands/cm², P = 0.03). As such, sweat rate at the greatest concentration employed was associated with a lower sweat gland output in females (7.75 ± 1.68 μl/min) relative to males (12.24 ± 2.30 μl/min, P ≤ 0.001).

Baseline cutaneous vascular conductance was lower in females at the ACh (P ≤ 0.001) and SNP (P = 0.06) sites. Cutaneous vascular conductance increased with incremental concentrations of both ACh and SNP (P ≤ 0.001), and these increases were similar between sexes at both sites (P > 0.1, Fig. 3). The best-fit values of the dose-response curves for cutaneous vasodilation (Fig. 4) did not significantly differ between males and females at both the ACh [F_(3,162) = 1.545, P = 0.21] and SNP [F_(3,106) = 1.320, P = 0.27] sites. Although minimum values were lower in females for ACh (P = 0.07) and SNP (P = 0.04), maximum values and the log EC₅₀ did not significantly differ between sexes (all P > 0.1, Table 3).

Sweating and cutaneous vasodilation during passive heating.

The day of the experimental session, plasma concentrations of 17β-estradiol and progesterone for the female participants averaged 153 ± 90 (range: 46–334) pmol/l and 1.4 ± 1.3 (range: 0.6–5.1) nmol/l, respectively. Urine specific gravity did not significantly differ between groups (males: 1.013 ± 0.007 vs. females: 1.016 ± 0.008, P = 0.52). The heating period increased mean skin temperature by 4.24 ± 0.39°C and 4.34 ± 0.38°C in males and females, respectively (P = 0.91), which resulted in a similar dry heat gain between sexes (P = 0.39). Passive heating increased esophageal temperature by 1.39 ± 0.10°C over 71 ± 8 min in males and 1.46 ± 0.13°C (P = 0.71) over 61 ± 12 min (P = 0.18) in females. The increases in mean skin and esophageal temperatures were associated with an increase of at least 1.75°C in mean body temperature for all participants.

Sweat rate increased as a function of increases in mean body temperature (P ≤ 0.001) and was significantly different between males and females (P = 0.013, Fig. 5). Although sweat rate did not initially differ between males and females, it became significantly greater in males at an increase in mean body temperature of 1.25°C (P < 0.01). The number of active sweat glands at the end of the heating period did not differ between males (75 ± 8 glands/cm²) and females (83 ± 12 glands/cm², P = 0.22). As such, the lower sweat rate in females was associated with a lower sweat gland output (9.93 ± 2.73 μl/min) relative to males (14.63 ± 3.22 μl/min, P = 0.03). The lower sweat rate in females was paralleled by a lower thermosensitivity of the response (females: 1.29 ± 0.34 mg·min⁻¹·cm⁻²·°C⁻¹ vs. males: 1.83 ± 0.33 mg·min⁻¹·cm⁻²·°C⁻¹, P = 0.03) without any differences in the mean body temperature onset threshold (females: 0.85 ± 0.19°C vs. males: 0.67 ± 0.13°C, P = 0.10).

Cutaneous vascular conductance increased as a function of increases in mean body temperature (P ≤ 0.001) and was similar between males and females (P = 0.41, Fig. 6). The onset threshold (males: 0.56 ± 0.14°C vs. females: 0.73 ± 0.14°C, P = 0.07) and thermosensitivity (males: 65 ± 17%/°C vs. females: 76 ± 14%/°C, P = 0.31) of cutaneous vascular conductance also did not differ between males and females.

Considerations.

Although the results of the current study provide evidence that sex differences in sudomotor activity during heat stress are mediated in part through a peripheral component, they do not rule out the possibility that they are also mediated centrally. To rule out any potential central modulation of sudomotor activity would imply measuring sudomotor-specific skin sympathetic nerve activity. Even if such a measurement could be obtained, the interpretation of skin sympathetic nerve activity is limited when used between independent groups, as bursts do not display synchronicity, have variable amplitudes, may contain multiple peaks, and the measurement may not be reproducible even within the same individual (31). Nonetheless, recent findings suggest that some components of the skin sympathetic nerve activity signal can be reliably used when assessing cutaneous vasodilation (18). Therefore, future studies are warranted to reliably investigate potential sex differences in skin sympathetic nerve activity to determine whether the lower sudomotor activity in females is also modulated by a central component of temperature regulation.

It may be argued that the observed sex differences in sudomotor activity are related to the differences in maximum oxygen consumption between groups. Although some studies report a greater sweating response to cholinergic stimulation in individuals with greater maximum oxygen consumption (2, 3, 21), it is most likely that such observations are associated with the beneficial adaptations that come with regular exercise training, as opposed to a direct effect of maximum oxygen consumption per se. In fact, Buono and Sjoholm (2) report a similar sweat response to cholinergic stimulation between males and females of similar training status, despite females having lower maximum oxygen consumption values. It has also been shown that the cholinergic sensitivity of the sweat gland is improved with exercise training (3), yet we did not observe any differences in the log EC₅₀ of the sweating response between males and females. In the current study, males and females were therefore matched for training status, as opposed to maximum oxygen consumption values. The similar training status between sexes, combined with the similar maximum oxygen consumption values expressed as a function of lean body mass, suggest that the lower absolute maximum oxygen consumption values in females were simply due to differences in the amount of metabolically active tissue, as opposed to differences in the benefits of regular exercise training which may confer adaptations for temperature regulation.