Previous studies report greater postexercise heat loss responses during active recovery relative to inactive recovery despite similar core temperatures between conditions. Differences have been ascribed to nonthermal factors influencing heat loss response control since elevations in metabolism during active recovery are assumed to be insufficient to change core temperature and modify heat loss responses. However, from a heat balance perspective, different rates of total heat loss with corresponding rates of metabolism are possible at any core temperature. Seven male volunteers cycled at 75% of V̇o2peak in the Snellen whole body air calorimeter regulated at 25.0°C, 30% relative humidity (RH), for 15 min followed by 30 min of active (AR) or inactive (IR) recovery. Relative to IR, a greater rate of metabolic heat production (Ṁ − Ẇ) during AR was paralleled by a greater rate of total heat loss (ḢL) and a greater local sweat rate, despite similar esophageal temperatures between conditions. At end-recovery, rate of body heat storage, that is, [(Ṁ − Ẇ) − ḢL] approached zero similarly in both conditions, with Ṁ − Ẇ and ḢL elevated during AR by 91 ± 26 W and 93 ± 25 W, respectively. Despite a higher Ṁ − Ẇ during AR, change in body heat content from calorimetry was similar between conditions due to a slower relative decrease in ḢL during AR, suggesting an influence of nonthermal factors. In conclusion, different levels of heat loss are possible at similar core temperatures during recovery modes of different metabolic rates. Evidence for nonthermal influences upon heat loss responses must therefore be sought after accounting for differences in heat production.
- body heat storage
- heat production
- heat stress
- nonthermal factors
human body heat storage and changes in internal body temperature are a direct result of a thermal imbalance between the rate of heat production (i.e., metabolic rate minus work rate: Ṁ − Ẇ) and the rate of total heat dissipation to the surrounding environment (ḢL) (6). Most often, during exercise in a hot environment, heat production is estimated either from work activity/intensity tables, measurements of heart rate (HR), or occasionally by indirect calorimetry, and the potential for heat dissipation is estimated either from body mass loss or the measurement of heat loss responses such as skin blood flow and local sweat rate (LSR) (23). It is generally accepted that the only way to accurately examine dynamic heat balance is by performing simultaneous minute-by-minute measurements of the individual heat balance components by whole body calorimetry (27). As such, Ṁ − Ẇ is measured using the stoichiometric relationship of the products and reactants of oxidative metabolism (indirect calorimetry); while ḢL from the body is determined from the direct measurement of the rates of sensible (radiation, conduction, and convection) and insensible (evaporation from sweating and respiration) heat loss using a direct calorimeter.
It is well documented that following a bout of dynamic exercise, a sustained elevation of core body temperature above preexercise resting values is evident for at least 60 to 90 min of stationary upright seated recovery (13, 19). In parallel, sweating and skin blood flow have also been shown to rapidly reduce to preexercise resting values during the early stages of recovery, suggesting a rapid reduction in postexercise heat loss despite a residual body heat storage. This response has been related to the ensuing postexercise hypotension, which accompanies an inactive recovery, that is, a stationary upright seated position (13, 19). Several studies (2, 4, 16, 25, 30) have shown that an active (loadless pedaling) and a passive (assisted loadless pedaling) recovery mode, both of which maintain mean arterial pressure (MAP) via the augmentation of venous return through engagement of the skeletal muscle pump, elicit greater levels of skin blood flow (4, 16) and LSR (4, 16, 25, 30) relative to inactive recovery. Since no differences in core temperature have been observed between any of these recovery modes (4, 16, 25, 30), an adjustment (i.e., increased thermal sensitivity and/or shifted onset threshold) of the thermal control of sweating and/or skin blood flow has been ascribed as the underlying nonthermal mechanism (16), primarily because of a reduced postexercise baroreceptor unloading and hypotensive effect typically evident during inactive recovery (13, 19).
Most studies have not compared measurements of oxygen consumption between active and inactive recovery modes, assuming that any elevations in metabolism during active recovery relative to inactive recovery are insufficient to produce changes in hypothalamic temperature (16, 17, 30). Consequently, previous studies have ruled out the possibility that elevations in heat production are responsible for the modifications in skin blood flow or sweating response (16, 17, 30) during an active recovery. However, it is logical that an elevated thermogenesis will occur during loadless pedaling relative to inactive recovery due to the liberation of energy supplying the metabolic demands of contracting leg muscles. Furthermore, from a heat balance perspective, it is reasonable to suggest that given the greater heat loss responses during active recovery (and presumably a greater rate of heat dissipation), without a greater metabolic rate in parallel, different levels of body heat storage and probably core temperature would be observed between recovery modes. Nevertheless, no differences in core temperature response between active and inactive postexercise recovery modes are evident during a 5-min (4, 30) or a 30-min recovery (15, 17).
According to proportional control theory, the aim of the thermal controller is to establish a core body temperature at a given adjustable set point (7–9) that is dependent upon the effects of workload, environmental conditions, and blood flow distribution (29). It follows that for any given adjusted set point core temperature, a balance between rates of heat production, and total heat dissipation [i.e., a rate of body heat storage (Ṡ) of zero] must be attained (5). Therefore, if metabolic heat production during active postexercise recovery is greater than during inactive recovery, then a greater heat dissipation must occur for heat balance to be reached during active recovery. It could then be argued that the greater levels of heat loss responses during active recovery are a consequence of a greater metabolic load and not nonthermal factors. However, similar core temperatures previously reported between recovery modes (4, 16, 17, 30) possibly suggest similar changes in body heat content despite a higher metabolic heat production during active recovery. By definition, the change in body heat content is the sum of the minute-by-minute differences between heat production and heat loss. Therefore, because a smaller reduction in the level of total heat loss is required to achieve heat balance during active recovery, the relative rate at which postexercise heat loss decays must be correspondingly slower. Nonthermal factors during postexercise recovery may therefore only be revealed after accounting for differences in metabolic heat production between recovery modes.
The aim of the present study was to investigate postexercise thermoregulation from a heat balance perspective by directly measuring the individual components of heat production and heat loss using whole body calorimetry during a 15-min bout of high-intensity dynamic exercise followed by 30 min of 1) active recovery (loadless pedaling) and 2) inactive recovery (stationary upright seated). It was hypothesized that despite similar core temperatures between recovery modes, a significantly greater rate of metabolic heat production would be observed during active relative to inactive recovery, and the rate of total heat loss would be at a concomitantly greater level toward the end of active recovery to achieve heat balance, primarily via an enhanced evaporative heat loss. It was also hypothesized that changes in body heat content measured using calorimetry would be similar between recovery modes.
Following approval of the experimental protocol from the University of Ottawa Research Ethics Committee and obtaining written informed consent, seven healthy, nonsmoking normotensive male participants volunteered for the study. Mean characteristics of these participants were age, 22 years (SD 2); height, 1.77 m (SD 0.08); weight, 78.8 kg (SD 7.5); body fat, 16.23% (SD 6.69); body surface area, 1.95 m2 (SD 0.13); and peak oxygen consumption (V̇o2peak), 46.6 ml·kg−1·min−1 (SD 6.2).
Whole body calorimetry.
A modified Snellen direct air calorimeter was employed for the purpose of measuring the rate of evaporative (ḢE) and dry heat loss (ḢD), yielding an accuracy of ±2.3 W for the measurement of rate of total heat loss (ḢL)1 . A full peer-reviewed technical description of the fundamental principles and performance characteristics of the Snellen whole body calorimeter is available (24). The Snellen calorimeter has also been employed in previous studies (11, 12).
In summary, the calorimeter incorporates a semirecumbent constant load eddy current cycle ergometer. The ergometer pedals are located inside of the calorimeter and mechanically linked by chains to the resistance control unit, regulating rate of external work (Ẇ) at a predetermined level, outside of the calorimeter so that any heat generated by the unit does not enter the calorimeter. The calorimeter is housed within a climatic chamber slightly pressurized (+8.25 mmHg) to nullify potential air leakage through the calorimeter walls. Differential air temperature and humidity are measured over the calorimeter by sampling the influent and effluent air. The water content is measured using precision dew point thermometry (RH Systems model 373H, Albuquerque, NM), while the air temperature is measured using RTD high-precision thermistors (±0.002°C, Black Stack model 1560, Hart Electronics, American Fork, UT). Air mass flow through the calorimeter is estimated by differential thermometry over a known heat source (2 × 750 W heating elements) placed in the effluent air stream. Differential temperature over the heater is measured using a third aforementioned high-precision thermistor placed downstream from the heater. Air mass flow rate (kg air/min) is continuously measured during each trial. Data from the calorimeter were collected continuously at 8-s intervals throughout the trials. The real-time data were displayed and recorded on a personal computer (Dell OPTIPLEX GX270) with LabVIEW software (ver. 7.0, National Instruments, Austin, TX).
Rate of evaporative heat loss (ḢE) was calculated from the calorimetry data every minute using the following equation: (1) where mass flow is the rate of flow of air mass (kg air/s); (Humidityout − Humidityin) is the calorimeter inflow-outflow difference in absolute humidity (g water/kg air); and 2,427 is the latent heat of vaporization of sweat (J/kg sweat) at 30°C (28).
Rate of dry heat loss (ḢD) from radiation, convection, and conduction was calculated from the calorimetry data every minute using the following equation: (2) where mass flow is the rate of flow of air mass (kg/s); (Temperatureout − Temperaturein) is the calorimeter inflow-outflow difference in air temperature (°C), and 1,005 is the specific heat of air [J·(kg air·°C)−1].
A 6-liter fluted mixing box housed within the calorimeter was utilized for the concurrent measurement of metabolic energy expenditure (Ṁ). The indirect calorimetry open-circuit technique used expired gas samples drawn from the mixing box. Expired gas was analyzed for V̇o2 (error of ±0.01%) and V̇co2 (error of ±0.02%) using electrochemical gas analyzers (AMETEK model S-3A/1 and CD 3A; Applied Electrochemistry, Pittsburgh, PA). Expired air was recycled back into the calorimeter chamber to account for respiratory dry and evaporative heat loss. Before each session, gas mixtures of 4% CO2, 17% O2, and the balance N2 were used to calibrate the gas analyzers, and a 3-liter syringe was used to calibrate the turbine ventilometer (error of ±3%, typically <1%). Rate of metabolic energy expenditure (Ṁ) was calculated from minute-average values for V̇o2 and respiratory exchange ratio (22).
The calorimeter data were then used to calculate rate of body heat storage (Ṡ) and change in body heat content (ΔHb_cal) using the following equations: (3) (4) where Ṁ is the rate of metabolic heat production; ḢE is rate of evaporative heat loss; ḢD is rate of dry heat loss; and Ẇ is the rate of external work (all units in W).
HR was monitored using a Polar coded transmitter, recorded continuously, and stored with a Polar Advantage interface and Polar Precision Performance software (Polar Electro Oy, Finland).
Esophageal temperature (Tes) was measured by placing a pediatric thermocouple probe of ∼2 mm in diameter (Mon-a-therm Nasopharyngeal Temperature Probe, Mallinckrodt Medical, St. Louis, MO) through the participant's nostril, while they were asked to sip water through a straw. The location of the probe tip in the esophagus was estimated to be in the region bounded by the left ventricle and aorta, corresponding to the level of the eighth and ninth thoracic vertebrae (21). Temperature data were collected using a HP Agilent data acquisition module (model 3497A) at a sampling rate of 15 s and simultaneously displayed and recorded in spreadsheet format on a personal computer (IBM ThinkCentre M50) with LabVIEW software (ver. 7.0, National Instruments, Austin, TX).
LSR was measured using a 5.0 cm2 ventilated capsule placed over the medial inferior aspect of the trapezius muscle. Anhydrous compressed air from outside the calorimeter was passed through tubing and over the skin surface within the capsule (Brooks 5850, Mass Flow Controller, Emerson electric, Hatfield, PA) and then out of the calorimeter. The air within the LSR system was isolated from the calorimeter environment at all times. The vapor density of the effluent air was calculated from the RH, and temperature was measured using the Omega HX93 humidity and temperature sensor (Omega Engineering, Stamford, CT). LSR was calculated using the difference in water content between effluent and influent air and the flow rate. The flow rate through the capsule was 0.5 l/min. The sweat rate value was adjusted for skin surface area under the capsule and expressed in milligrams per minute per centimeter squared.
All participants volunteered for three separate testing days. On testing day 1, an incremental cycle ergometer V̇o2peak test was performed. On testing day 2 and 3, the calorimetry experimental exercise protocol was performed, once with active recovery, the other with inactive recovery. The presentation of recovery activities was balanced between participants so that the effect of order was avoided. Testing days were separated by a minimum of 72 h. All calorimeter trials were performed at the same time of day. Participants were asked to arrive at the laboratory after eating a small breakfast (i.e., dry toast and juice), but consuming no tea or coffee that morning, and also avoiding any major thermal stimuli on their way to the laboratory. Participants were also asked not to drink alcohol or exercise for 24 h prior to experimentation.
Following instrumentation, each participant entered the calorimeter regulated to an ambient air temperature of 25°C and 30% RH (water vapor pressure = 7.1 mmHg). The participant, in the upright seated position, rested for a 45-min habituation period, while a steady-state baseline resting condition was achieved. Subsequently, the participant cycled at 75% of their predetermined V̇o2peak for 15 min. This was immediately followed by 30 min of either 1) upright seated inactive recovery, or 2) active recovery of loadless pedaling at 60 revolutions/min, both while remaining inside the calorimeter.
For all experimentation, clothing insulation was standardized at ∼0.2 to 0.3 clo (i.e., cotton underwear, shorts, socks, and athletic shoes).
A two-way repeated-measures ANOVA was used to analyze the data from the 30-min postexercise recovery period using the repeated factors of recovery mode (levels: inactive and active) and time (levels: 0, 2, 5, 10, 15, 20, 25, and 30 min of postexercise recovery). The dependent variables were Ṁ − Ẇ, ḢL, ḢE, ḢD, Ṡ, ΔHb, ΔHR (from baseline rest), ΔTes (from baseline rest), and LSR. For ANOVA main effects, Huynh-Feldt corrected statistics are reported where the assumption of sphericity was not met. Paired sample t-tests were used to perform post hoc comparisons between stationary and active recovery. The level of significance was set at 0.05 and alpha level was adjusted during multiple comparisons so as to maintain the rate of type I error at 5% during the Bonferroni post hoc analysis (P ≤ 0.05 n−1, where n = number of comparisons). All analyses were performed using the statistical software package SPSS 15.0 for Windows (SPSS, Chicago, IL).
Rate of net metabolic heat production (Ṁ − Ẇ) was significantly greater during active recovery relative to inactive recovery [F(1,6) = 41.8, P = 0.001]. Differences between recovery modes were observed between 5 and 30 min postexercise (Fig. 1). Similarly, the rate of total heat loss (ḢL) was significantly greater during active recovery compared with inactive recovery [F(1,6) = 38.6, P = 0.001] with a difference in ḢL occurring between 10 and 30 min postexercise (Fig. 1). The subsequent rate of body heat storage, that is, Ṡ = [(Ṁ − Ẇ) − ḢL], was not different between recovery modes at any point following exercise [F(1,6) = 0.6, P = 0.483] (Fig. 2). Upon separating ḢL into the components of rates of evaporative (ḢE) and dry (ḢD) heat loss, both ḢE [F(1,6) = 19.6, P = 0.004] and ḢD [F(1,6)=85.4, P < 0.001] were greater during active recovery. A significant difference was observed between recovery modes after 10 to 30 min for ḢE (Fig. 3A) and between 2 min and 30 min for ḢD (Fig. 3B). However, the magnitude of difference between recovery modes for ḢD was approximately a quarter of that observed for ḢE.
Changes from baseline rest in esophageal temperature (ΔTes), heart rate (ΔHR), and change in body heat content measured with calorimetry (ΔHb_cal) are given in Table 1. No significant differences between recovery modes were observed for ΔHb_cal [F(1,6) = 1.3, P = 0.302], ΔHR [F(1,6) = 3.0, P = 0.134], or ΔTes [F(1,6) = 0.4, P = 0.549].
This study presents novel postexercise recovery data from a heat balance perspective. The main findings show that despite similar core temperatures between active and inactive recovery modes, differences in ḢL and its components (evaporative and dry heat loss) between modes, parallel the differences in Ṁ − Ẇ, with the thermoregulatory system approaching heat balance toward the end of recovery under both conditions. Because metabolic rate is greater during active recovery, ḢL is also greater, primarily via enhanced sudomotor activity and evaporation of sweat. However, despite the smaller postexercise reduction in the rates of heat production and heat loss, the changes in body heat content directly measured using calorimetry were similar between recovery modes.
The direct measurement of the individual avenues of human heat balance show that since the effect of the thermoregulatory system is to achieve heat balance, toward the end of recovery, ḢL is directly dependent upon the levels of Ṁ − Ẇ, irrespective of core temperature. This is demonstrated in the present data by similar elevations toward the end of recovery in both heat production and total heat loss during active recovery compared with inactive recovery. That is, the end-recovery difference between active and inactive recovery was 91 W (SD 26) for heat production and 93 W (SD 25) for total heat loss. Indeed, the individual variation observed in Ṁ − Ẇ between recovery modes was closely correlated with the individual variation in ḢL between modes (r = 0.70). It has been previously stated that no significant differences in metabolic heat production exist between active and inactive recovery because of the observation of similar core temperatures (16, 17, 30). However, the present data clearly demonstrate that different levels of total heat loss with corresponding rates of heat production occur at the same core temperature.
Nonthermal factors have been previously demonstrated to clearly influence postexercise sudomotor and vasomotor activity (12, 18) since different levels of skin blood flow and sweating have been shown using various inactive recovery interventions that elicit similar metabolic demand, such as head-down tilt (14, 18, 20) and lower-body positive pressure (10, 15). Even though factors such as central command, mechanoreceptors, and metaboreceptors may also be implicated, it is believed that the reduced baroreceptor unloading effect from the reversal of the postexercise hypotension typically observed during inactive recovery is the primary nonthermal factor responsible for the adjustment of thermoregulatory control during active recovery (13, 19). However, evidence for nonthermal factors upon postexercise thermoregulation has been typically sought by comparing the absolute values of heat loss responses between recovery conditions. In contrast, the present study demonstrates that any differences in metabolic rate between conditions will also directly influence postexercise levels of heat loss responses irrespective of core temperature. The question, therefore, remains, how can metabolic and nonthermal effects upon postexercise heat loss be separated when comparing conditions that require different levels of metabolism?
Changes in heat content of a body are directly determined by the overall net difference between the heat produced within the body and the heat exchange with the environment over a period of time (5). In the present study, greater rates of metabolic heat production and total heat loss were observed during active relative to inactive recovery (Fig. 1); however, the overall change in body heat content (ΔHb_cal) derived directly from the calorimetric measurements of heat balance was similar between recovery modes (Table 1). When viewed from a heat balance perspective, it is clear that relative to the total postexercise reduction in ḢL required to achieve heat balance, the rate of decline must be proportionally slower during active recovery compared with inactive recovery; otherwise, ΔHb_cal would be different between conditions. Since both core temperature and rate of body heat storage were similar between recovery modes, it is suggested that nonthermal factors also modulate the proportional control of sudomotor and vasomotor activity during postexercise recovery.
As previously mentioned, the primary nonthermal factor responsible for the adjustment of postexercise thermoregulatory control is believed to be baroreceptor activity. Because of technical limitations in the present study, it was not possible to measure MAP in the direct calorimeter; however, the HR and oxygen consumption (V̇o2) data support a greater MAP during active recovery. Using Fick's Principle, we know that, assuming no change in the arterial-venous difference in oxygen content between recovery modes, the greater oxygen consumption observed during active recovery is paralleled by a proportional increase in cardiac output. Since cardiac output is determined by the product of stroke volume and heart rate, and there was no difference between active and inactive recovery for HR (Table 1), it is suggested that a greater stroke volume occurred during active recovery. Previous studies have shown that an active loadless pedaling recovery elicits a greater venous return for cardiac filling via the activation of the muscle pump, and as such produces a concomitant increase in MAP (3).
Without the benefit of whole body calorimetry, human thermoregulation is rarely viewed in terms of heat balance, with discussion typically focused upon thermometric descriptions of homeostatic control. However, the data from the present study show that classical proportional temperature control theory (9) may also be interpreted as the thermoregulatory system regulating a balance between heat production and heat dissipation (Fig. 2). There is a constant dynamic heat flux within the body due to conductive and convective heat exchange between tissues of varying specific heat. However, for a given individual, the net heat storage accumulated (change in body heat content) throughout exercise and recovery is roughly proportional to the parallel changes in core body temperature, assuming that the temperature gradient from the body core across the intermediate tissue to the body shell is similar. This is supported by the fact that no differences occurred between recovery modes for the independent measures of change in body heat content (using whole body calorimetry) and change in core temperature (using esophageal temperature) (Table 1). Whether core/brain temperature or the rate of body heat storage itself is the stimulus for the thermoregulatory effector mechanisms of sweating and skin blood flow is a topic of much debate (1, 26). However, in the context of the present study, the identity of the regulated variable does not matter. The point of this study is that heat balance and body temperature change are unavoidably linked and when interpreting data from studies employing thermometry and the measurement of the heat loss responses of sweating and skin blood flow, heat balance must be considered. While this may be intuitive, such a link is often not identified. For example, the elevated heat loss responses previously reported during active relative to inactive recovery were considered to be solely a function of nonthermal factors, with metabolic heat production assumed to be the same since core temperatures were similar between recovery modes (16, 17, 30). In addition, Shibasaki et al. (25) measured postexercise oxygen consumption and found a significantly higher metabolic rate during an assisted loadless pedaling at ∼60 rpm (referred to as a “passive” recovery) compared with inactive recovery, while also finding a greater sweat rate in parallel. As esophageal temperature remained the same between conditions, the elevated sweat rate during this “passive” recovery was considered purely a nonthermoregulatory response. However, it is clear from a heat balance perspective that with an elevated rate of total heat loss, without a concomitant elevation in metabolic heat production, change in heat content, and therefore probably core temperature, would be different between recovery modes. On the contrary, the fact that core temperatures are the same between recovery modes despite an elevated heat production during active recovery is evidence of a possible nonthermal effect.
Logically, to facilitate the greater rate of total heat loss measured calorimetrically during active relative to inactive recovery, it is assumed that the heat loss responses of sweating and skin blood flow are also greater. Indeed, LSR of the upper back was measured in the present study, and a slower decay of LSR was observed during active recovery relative to inactive recovery (Fig. 4). This supports an increased sudomotor activity associated with active recovery independently of any effects of the increased convective heat loss that may be associated with pedaling relative to stationary recovery. ḢD was increased by ∼15 W during active recovery (Fig. 3B), which is thought to be primarily a consequence of an increased convection; however, this is minimal compared with the enhanced evaporative heat loss (Fig. 3A) during active recovery (∼75 W) and the combined data for ḢL clearly support the interpretation that heat loss via all avenues is regulated to balance Ṁ − Ẇ under both recovery modes (Fig. 2). Because of the technical limitations of performing trials in the direct calorimeter, skin blood flow was not measured; however, evaporation is by far the most dominant avenue for heat dissipation during heat stress, and it is likely that the greater sudomotor activity during active recovery is accompanied by a greater vasomotor activity (16).
Perspectives and Significance
Nearly all previous work studying the influence of nonthermal factors upon thermoregulatory control during and following exercise has used measurements of core temperature to determine whether any differences in metabolic heat production occur between conditions. We show that similar core temperatures are possible at different levels of postexercise heat production. Since the effect of the thermoregulatory system is to attain an equilibrium between rates of heat production and heat loss, the levels of postexercise heat loss responses during steady-state conditions are determined by metabolic rate. Therefore, when comparing between conditions of differing metabolism, evidence for nonthermal factors influencing thermoregulatory control must be sought by examining the rate of change of heat loss responses relative to the total change required to achieve heat balance. The significance of the present study is not only restricted to the influence of recovery mode upon postexercise thermoregulation. Because we clearly demonstrate that different rates of total heat loss (and heat loss responses) with corresponding rates of metabolism can exist at equal core temperatures, it is evident that any studies employing thermometry and the measurement of heat loss responses must also interpret their data considering the laws of heat balance.
This research was supported by the Natural Sciences and Engineering Research Council (Grant RGPIN-298159-2004 to G. Kenny) and the U.S. Army Medical Research and Material Command's Office of the Congressionally Directed Medical Research Programs (Grant DAMD17-02-2-0063 to G. Kenny). The provision of financial support does not in any way infer or imply endorsement of the research findings by either agency. Dr. Glen Kenny was supported by a University of Ottawa Research Chair Award.
We would also like to thank Candice Brown, Matthew Kennedy, and Andrea McCarthy for their assistance during data collection.
↵1 Rate of total heat loss (ḢL) is the sum of concurrent rates of evaporative (ḢE) and dry (ḢD) heat loss, i.e., ḢL = ḢE + ḢD.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society