This study determined whether cutaneous blood flow during exercise is different in endurance-trained (Tr) compared with untrained (Untr) subjects. Ten Tr and ten Untr men (62.4 ± 1.7 and 44.2 ± 1.8 ml ⋅ kg− 1 ⋅ min− 1, respectively; P < 0.05) underwent three 20-min cycling-exercise bouts at 50, 70, and 90% peak oxygen uptake in this order, with 30 min rest in between. The environmental conditions were neutral (∼23–24°C, 50% relative humidity, front and back fans at 2.5 m/s). Because of technical difficulties, only seven Tr and seven Untr subjects completed all forearm blood flow and laser-Doppler cutaneous blood flow (CBF) measurements. Albeit similar at rest, at the end of all three exercise bouts, forearm blood flow was ∼40% higher in Tr compared with Untr subjects (50%: 4.64 ± 0.50 vs. 3.17 ± 0.20, 70%: 6.17 ± 0.61 vs. 4.41 ± 0.37, 90%: 6.77 ± 0.62 vs. 5.01 ± 0.37 ml ⋅ 100 ml− 1 ⋅ min− 1, respectively; n = 7; all P < 0.05). CBF was also higher in Tr compared with Untr subjects at all relative intensities (n = 7; all P < 0.05). However, esophageal temperature was not different in Tr compared with Untr subjects at the end of any of the aforementioned exercise bouts (50%: 37.8 ± 0.1 vs. 37.9 ± 0.1°C, 70%: 38.1 ± 0.1 vs. 38.1 ± 0.1°C, and 90%: 38.8 ± 0.1 vs. 38.6 ± 0.1°C, respectively). We conclude that a higher CBF may allow Tr subjects to achieve an esophageal temperature similar to that of Untr, despite their higher metabolic rates and thus higher heat production rates, during exercise at 50–90% peak oxygen uptake.
- body temperature regulation
- forearm blood flow
during prolonged exercise at a given relative intensity [i.e., percent maximal oxygen uptake (V˙o 2 max)], endurance-trained (Tr) individuals have a higher metabolic rate [i.e., higher oxygen uptake (V˙o 2)] and, (therefore, produce more heat than do untrained (Untr) counterparts. Despite this higher heat production, Tr achieve a core temperature similar to that of Untr subjects (1, 24), indicating that they are also able to dissipate more heat. Hence, it is logical to hypothesize that, during exercise at a given relative intensity, cutaneous blood flow (CBF), one of the main heat loss responses, is higher in Tr compared with Untr subjects.
Evidence that the cutaneous vasculature may adapt to exercise training was observed by Roberts et al. (22), who reported that, after 10 days of endurance exercise training, cutaneous vasodilation during exercise (at the pretraining intensity) starts at a lower core temperature. However, the hypothesis that CBF may be higher in Tr compared with Untr subjects during exercise at a given relative intensity has not been well investigated. In the only study that directly addressed this hypothesis, Tankersley et al. (28) studied elderly men during exercise at 65–70% V˙o 2 max, observing a nonsignificant trend for a higher CBF [estimated as forearm blood flow (FBF)] in trained compared with sedentary older men.
Therefore, our main purpose was to test the hypothesis that, during exercise at a given relative exercise intensity, CBF is higher in Tr compared with Untr subjects. An additional purpose of this study was to compare the CBF response in Tr vs. Untr men during exercise at a given absolute intensity (i.e., a givenV˙o 2, normalized by body weight) when metabolic rate and therefore heat production and heat dissipation are similar in Tr compared with Untr subjects.
Ten Tr men (regularly performing endurance training) and ten active Untr men (not regularly performing endurance training) were selectively recruited for their ability to exercise at 90% peakV˙o 2(V˙o 2 peak) for at least 20 min. Tr subjects demonstrated a 40% higherV˙o 2 peak while cycling compared with Untr (i.e., 62.4 ± 1.7 vs. 44.2 ± 1.8 ml ⋅ kg− 1 ⋅ min− 1, respectively; P < 0.001). However, mean age, weight, height, surface area (6), and maximal heart rate were not significantly different between Tr and Untr subjects [25.4 ± 1.9 vs. 26.8 ± 1.1 (SE) yr, 72.9 ± 3.3 and 81.1 ± 3.3 kg, 1.80 ± 0.02 and 1.80 ± 0.02 m, 1.9 ± 0.1 and 2.0 ± 0.1 m2, and 183.8 ± 3.3 and 188.4 ± 2.7 beats/min, for Tr and Untr, respectively]. Because of technical difficulties, only seven Tr and seven Untr subjects completed all FBF and laser-Doppler CBF measurements.V˙o 2 peak was determined in all subjects by using a standard incremental cycle-ergometer protocol 2–7 days before the beginning of the experimental trial.
Protocol and experimental design.
All subjects performed three 20-min cycle-ergometer bouts at a work rate adjusted to elicit 50, 70, and 90% of theirV˙o 2 peak in this order, with 30 min of rest in between. These bouts were performed on the same day. Bouts of 20 min of exercise were selected because it was the minimum time needed to allow CBF to reach a stable value for several minutes. Bouts longer than 20 min would have compromised the conclusion of the 90%V˙o 2 peak bout. All three exercise bouts were performed on the same day to maintain the CBF probe attached to the same site on the subject for all three exercise intensities, because variable responses have been observed with changes in sampling sites (11). A rest interval of 30 min with fan cooling was interposed between exercise bouts to allow esophageal temperature to return to preexercise values and to remain at resting values for at least 10 min. Environmental conditions were maintained at ∼23–24°C and 50 ± 5% relative humidity. Convective cooling was provided by two fans (2.5 m/s wind speed) located 80 cm away from the subject's chest and back.
On arrival to the laboratory, subjects voided their bladder and had their nude body mass recorded. Subsequently, they were instrumented for esophageal temperature, mean skin temperature, blood pressure, heart rate, and CBF measurements. Then, subjects sat on the cycle ergometer (model 819, Monark), and, after resting data collection, they started the first of the three exercise bouts. These 20-min exercise bouts were performed at a work rate adjusted to elicit 50, 70, and 90% of the subject's V˙o 2 peak, in this order. The cycle ergometer cadence was preselected by the subjects (range 65–85 rpm) and maintained for all three bouts. Five and 7 ml/kg of a 6% carbohydrate-electrolyte solution (Gatorade) were ingested 20 min before the second and third exercise bouts, respectively, to prevent dehydration. Immediately after the experiment, subjects towel dried, voided their bladder, and had their final nude body mass recorded.
CBF was assessed by using both FBF and laser-Doppler CBF. FBF was measured by venous occlusion plethysmography according to the procedures outlined by Whitney (30). Briefly, the occlusion cuff was inflated to 60 mmHg, and blood flow to the hand was restricted. FBF was measured as the average of 6–10 values obtained at rest and during the 16- to 18-min period of each exercise bout. FBF values were used as an index of forearm CBF (9, 25). CBF was also estimated continuously by laser-Doppler flowmetry (ALF 21) on the dorsal side of the left forearm (2 cm distal from the FBF strain gauge). CBF is a measurement specific to the skin surface and is not influenced by blood flow to underlying skeletal muscle (25). CBF is reported as a percentage of the initial resting value (i.e., obtained before the 50%V˙o 2 peak exercise bout). Exercise time to the onset of cutaneous vasodilation and exercise time to a stable CBF were determined from the CBF-time graph by an experienced observer, blind to the subject and exercise intensity being analyzed. Esophageal temperature at the onset of cutaneous vasodilation (threshold temperature for cutaneous vasodilation) and esophageal temperature at the time a stable CBF was achieved were determined from the CBF-esophageal temperature graph by an experienced observer, blind to the subject and exercise intensity being analyzed.
Esophageal temperature was measured with an esophageal thermistor (YSI 491, Yellow Springs Instruments, Yellow Springs, OH) placed through the nasal passage into the esophagus at a distance equal to one-quarter of the standing height (16). Four skin thermistors (YSI 409A) were attached to the chest, upper arm, thigh, and calf to estimate mean skin temperature by using the weighting method of Ramanathan (21). All thermistors were connected to a temperature display (YSI 2100). Esophageal temperature was recorded continuously at rest and during exercise. Skin temperatures were recorded at rest and every 5 min during exercise. The maximal rate of change in esophageal temperature (°C/min) was calculated as the largest increase in esophageal temperature during a 4-min interval as follows. Esophageal temperature increases were listed every minute, 4-min running averages were calculated, and the maximal rate of change in esophageal temperature was defined as the highest running average.
Blood pressure and heart rate.
Blood pressure was measured at rest and every minute during exercise by auscultation by using an automatic monitor (model STBP-680, Colin Medical Instruments, South Plainfield, NJ). Mean arterial pressure (MAP) was calculated from systolic (SBP) and diastolic (DBP) blood pressure [MAP = (2DBP + SBP)/3]. Heart rate was displayed continuously on the same monitor, but only 19-to 20-min values on each bout were recorded.
V˙o2 and heat production.
V˙o 2 and CO2production were determined continuously during exercise. Briefly, subjects breathed through a one-way Daniels valve, connected to a dry-gas volume meter (model CD4, Parkinson-Cowan) and to a mixing chamber. Expired air was continuously sampled from the mixing chamber and analyzed for O2 (model S-3A/I, Ametek) and CO2 (model CD-3A, Ametek) concentrations. Both gas analyzers and the dry-gas meter were interfaced to a laboratory computer. The analyzers were calibrated before and after each exercise bout by using gases of known concentration.V˙o 2 is reported as the average of the last 5 min of each exercise bout. Heat production (J ⋅ kg− 1 ⋅ min− 1) was estimated by calculating energy expenditure usingV˙o 2 and respiratory exchange ratio and then subtracting the work rate. (To express heat production in W/kg, divide the heat production values by 60.)
Body mass and whole body sweating.
Nude body mass was measured on a platform scale (model FW 150 KAI, Acme Scale, CA; accuracy ±20 g). Whole body sweating was determined for the whole trial as the difference in body mass plus the fluid consumed minus the urine losses, corrected for respiratory water and carbon losses (20).
Data were analyzed with a two-factor ANOVA (exercise intensity by training level) with repeated measures over exercise intensity, by using SPSS 6.0 for Windows statistical software. When significance was found, individual statistical differences were identified by using Newman-Keuls post hoc comparisons. Significant deviations from sphericity were tested with the Mauchly sphericity test, and, when significance was found, Newman-Keuls post hoc comparisons were performed by using individual error terms (31). One-way comparisons between Tr and Untr were assessed by using one-way ANOVA. Probability of making a type I error was set at P < 0.05.
CBF at rest and during prolonged exercise.
FBF was similar at rest (2.30 ± 0.16 and 2.42 ± 0.22 ml ⋅ 100 ml− 1 ⋅ min− 1in Tr and Untr, respectively) and 47, 40, and 35% higher in Tr compared with Untr subjects, at the end of exercise at 50, 70, and 90%V˙o 2 peak, respectively (P < 0.05; Fig. 1). A direct relationship between absoluteV˙o 2(ml ⋅ kg− 1 ⋅ min− 1) and FBF was observed (R = 0.703, P < 0.05; see also Fig. 2). When comparing the present FBF values to other studies, keep in mind that in the present study, two fans provided convective cooling during exercise. Laser-Doppler flow and thus CBF was also significantly higher in Tr compared with Untr subjects at the end of exercise at 50, 70, and 90% V˙o 2 peak(P < 0.05; Fig. 3). All subjects reached a stable laser-Doppler flow value before 15 min at all exercise intensities except for one Tr subject at 90%V˙o 2 peak, who appeared to reach a stable value at 18–19 min. During exercise at 90%V˙o 2 peak, the stable laser-Doppler flow value was observed to be higher in Tr compared with Untr subjects (P < 0.05), despite a continuing increase in core temperature (i.e., a plateau in the CBF-core temperature relationship; see Fig. 4).
V˙o2, cardiovascular variables, and heat production.
V˙o 2(%V˙o 2 peak, l/min and ml ⋅ kg− 1 ⋅ min− 1) and heat production (J ⋅ kg− 1 ⋅ min− 1) are reported in Table 1. Differences in heat production in Tr compared with Untr subjects were proportional to the differences in V˙o 2.V˙o 2(ml ⋅ kg− 1 ⋅ min− 1) was ∼40% higher (P < 0.05) and heat production was ∼38% higher (P < 0.05) in Tr compared with Untr subjects at all three relative exercise intensities (i.e., 50, 70, and 90%V˙o 2 peak).
Esophageal and mean skin temperatures.
Esophageal temperatures at the end of the exercise bout were higher at 70% compared with 50%V˙o 2 peak and at 90% compared with 70%V˙o 2 peak in both Tr and Untr subjects (P < 0.05; see Fig.5). Skin temperature is shown in Table2. At given relative exercise intensity, when absoluteV˙o 2 peakvalues (ml ⋅ kg− 1 ⋅ min− 1) were ∼40% higher in Tr vs. Untr, final esophageal temperatures were not different in Tr compared with Untr subjects (i.e., 50%: 37.8 ± 0.1 vs. 37.9 ± 0.1°C, 70%: 38.1 ± 0.1 vs. 38.1 ± 0.1°C, and 90%: 38.8 ± 0.1 vs. 38.6 ± 0.1°C, respectively). Conversely, when absolute V˙o 2(ml ⋅ kg− 1 ⋅ min− 1) was similar (i.e., Tr at 50% vs. Untr at 70% V˙o 2 peak), esophageal temperature was lower in Tr compared with Untr subjects (i.e., 37.8 ± 0.1 vs. 38.1 ± 0.1°C; P < 0.05).
Maximal rate of change in esophageal temperature.
The maximal rate of change in esophageal temperature was highly correlated with heat production (R = 0.89, slope = 3.03 kJ ⋅ kg− 1 ⋅ °C− 1,P < 0.05). The maximal rate of change in esophageal temperature was higher in Tr compared with Untr subjects at 50, 70, and 90% V˙o 2 peak (P< 0.05); it also was higher at 70% compared with 50%V˙o 2 peak and at 90% compared with 70%V˙o 2 peak in both Tr and Untr subjects (Table 3).
CBF at the exercise onset.
The threshold esophageal temperature for the onset of cutaneous vasodilation was not significantly different between Tr and Untr subjects at any given relative intensity (Table 3). The threshold increased from 50 to 70%V˙o 2 peak (P < 0.05) but not from 70 to 90%V˙o 2 peak in both Tr and Untr subjects. When absolute heat production was similar (i.e., Tr at 50% compared with Untr at 70%V˙o 2 peak), the threshold esophageal temperature for cutaneous vasodilation was lower in Tr compared with Untr subjects (37.4 ± 0.2 vs. 37.9 ± 0.1°C, respectively; P < 0.05). Exercise time to the onset of cutaneous vasodilation was always shorter in Tr compared with Untr subjects at all relative intensities (P < 0.05), and it also decreased from 50 to 70%V˙o 2 peak, and from 70 to 90% V˙o 2 peak(P < 0.05). However, when absolute heat production was similar (i.e., Tr at 50% compared with Untr at 70%V˙o 2 peak), exercise time to cutaneous vasodilation was similar in Tr compared with Untr subjects (8.14 ± 1.34 vs. 8.29 ± 0.99 min; P = not significant).
Whole body sweating.
Whole body sweating from the beginning of the 50%V˙o 2 peak bout to the end of the 90%V˙o 2 peak bout (i.e., including the 50 and 70%V˙o 2 peak resting periods; 2 h total) was higher in Tr compared with Untr subjects (15.7 ± 1.9 vs. 9.1 ± 1.6 g/kg; P < 0.05).
The main finding of the present study is that, during exercise at a given relative exercise intensity (i.e., at 50, 70, and 90%V˙o 2 peak) and relative heat production, CBF is higher in Tr compared with Untr subjects. Another interesting observation is that, at a given absolute exercise intensity (i.e., Tr at 50% compared with Untr at 70%V˙o 2 peak) and absolute heat production, CBF is similar in Tr compared with Untr subjects.
The observation that CBF during exercise can be affected by endurance training, albeit not a surprising finding, is one of the first direct observations of this phenomenon. Previously, Tankersley et al. (28) compared FBFs at a given relative intensity between Tr and Untr older subjects, reporting a nonsignificant trend for higher FBFs in the Tr compared with the Untr group. Furthermore, Roberts et al. (22) reported that, at a given relative exercise intensity, endurance training decreased the esophageal temperature at which cutaneous vasodilation occurred without modifying the slope of the CBF-esophageal temperature relationship. An extrapolation of the data from Roberts et al. would suggest that, during exercise at a given core temperature (and thus relative exercise intensity), CBF might be higher after endurance training. Ho et al. (7) reported similar findings in a cross-sectional sample of Tr and Untr elderly subjects. Therefore, extrapolation of data from previous studies (7, 22, 28) supports our direct observation that, after reaching a stable value, CBF during exercise is higher in Tr compared with Untr subjects during exercise at a given %V˙o 2 peak.
Another novel observation of this study, to our knowledge, was related to the plateau in the CBF-core temperature relationship (or the abolition of increases in CBF during upright exercise after core temperature surpasses ∼38°C; see Refs. 2, 3, 8, 12-15, 17,18, 26, 29). We observed that, at 90%V˙o 2 peak, the plateau in CBF occurs at a higher CBF value in Tr compared with Untr subjects.
Our protocol, consisting of three different relative exercise intensities, also allowed us to compare Tr and Untr subjects at one given absolute intensity and level of heat production (i.e., Tr 50% vs. Untr 70%). When compared at a given absolute exercise intensity and level of heat production, CBF was observed to be similar in Tr compared with Untr subjects. In summary, when the findings of this and previous paragraphs are combined, it seems that the CBF response is proportional to the heat dissipation response. That is, when at a given %V˙o 2 peak, Tr individuals produce and dissipate more heat compared with Untr and they display a proportionally higher CBF. However, when at a similar absolute V˙o 2(ml/kg), Tr and Untr individuals produce and dissipate a similar amount of heat, and they display a similar CBF (see Fig. 2).
During exercise at a given relative intensity (i.e., at 50, 70, and 90% V˙o 2 peak), esophageal temperature was similar, despite a 40% higher metabolic rate and 38% higher heat production in Tr compared with Untr subjects.Åstrand (1), Saltin and Hermansen (24), and Davies et al. (5) previously reported that relative exercise intensity determines core temperature between 30 and 80%V˙o 2 max. Our observation extends these previous findings (1, 5, 24) to 90%V˙o 2 peak. At a given relative intensity, we observed an earlier cutaneous vasodilation in Tr compared with Untr subjects. However, in contrast to Roberts et al. (22), we did not observe a significantly lower threshold esophageal temperature for cutaneous vasodilation in Tr compared with Untr subjects during exercise at a given %V˙o 2 peak. Our failure to reproduce the results of Roberts et al. may be due to the higher variability inherent to our cross-sectional design and also to the thermoregulatory effects of a previous exercise bout on subsequent bouts (4, 10). An interesting observation of the present study was that, during exercise at a given absolute intensity, cutaneous vasodilation in trained subjects started at a similar time but at a much lower (threshold) esophageal temperature in Tr compared with Untr subjects.
Our design also allowed us to compare the changes in CBF when intensity increased from 50 to 70%V˙o 2 peak with the changes in CBF when intensity increased from 70 to 90%V˙o 2 peak. CBF increased significantly from 50 to 70%V˙o 2 peak but generally failed to increase significantly from 70 to 90%V˙o 2 peak, despite large increases in core temperature (∼0.7°C). Both Tr and Untr subjects demonstrated this response. Our findings clearly indicate that further increases in CBF are attenuated or abolished at high exercise intensities (i.e., from 70 to 90%V˙o 2 peak), despite much higher core temperatures. Similar conclusions were reached by studies using incremental upright exercise models (19, 26). This attenuation in the CBF response at high exercise intensities, despite increasing core temperature, has been attributed to effects of high exercise intensities on the cutaneous circulation (19, 26, 27), yet the mechanism of effect of exercise intensity on the cutaneous circulation is not clear. Nadel and co-workers (17) reported no effect of increasing exercise intensity on the CBF response to a given core temperature. However, Brengelmann et al. (3) and other studies (12, 14,15, 17, 18) reported an attenuation in the CBF response during upright exercise without increases in exercise intensity. In summary, during prolonged upright exercise, increases in CBF are attenuated or abolished when core temperature surpasses ∼38°C, independent of whether exercise intensity increases (19, 26) or not (3, 12, 14, 15,17, 18).
Whole body sweating over the entire experiment, including the rest periods, was higher in Tr compared with Untr subjects. It is possible that a larger sweat rate at a given relative intensity contributed to the larger heat dissipation in Tr compared with Untr subjects. The time pattern of sweating observed in other experiments during exercise and recovery (23) suggests that most of the sweat observed in this protocol was produced during exercise. The placement of the 90%V˙o 2 peak exercise bout (i.e., the bout that could produce the highest sweat rate during the subsequent rest period) at the end of the protocol and body weight determination soon after exercise ensured the lowest possible influence of rest period sweating on overall body weight losses.
A potential limitation in the study design was the bias introduced in the experiment by ordering the exercise bouts (i.e., 50, 70, and 90%V˙o 2 peak). We decided to order the exercise bouts because of the following reasons. First, during pilot experiments, we did not observe differences in final core temperature when the 50, 70, or 90%V˙o 2 peak exercise bouts were performed in ascending order. Second, pilot experiments suggested that, after a 90%V˙o 2 peak exercise bout, ∼60–90 min were needed for esophageal temperature to return to resting values. Moreover, in one case, esophageal temperature never returned to its resting value. Another alternative, performing each exercise bout on different days, would have precluded the comparisons of CBF across different exercise intensities, because CBF comparisons among trials are valid only when the laser-Doppler probe remains in its original place (11). A second limitation of the present study is that all exercise bouts were performed on the same day (to maintain the laser-Doppler probe in the same place). We cannot discount the possibility that, at 70 and 90%V˙o 2 peak, the threshold temperatures for cutaneous vasodilation and the time to cutaneous vasodilation could have been influenced by the previous exercise bout, as previously observed for sweating thresholds (4). Finally, the design of the present study did not allow us to differentiate between training per se, compared with a higherV˙o 2 peak. Therefore, we have to recognize the possibility that the results of the present study might be due to the higherV˙o 2 peak, by itself, in the Tr compared with the Untr group and not to the effect of training, per se.
In summary, we observed that 1) during upright exercise at a given relative exercise intensity, Tr individuals have higher CBFs while reaching similar core temperatures to that of Untr individuals, and 2) during upright exercise at a given absolute exercise intensity, Tr individuals have similar CBFs and reach lower core temperatures compared with Untr individuals.
The cooperation of the subjects involved and the assistance of José González-Alonso, Ricardo Mora-Rodrı́guez, Jeff Horowitz, Paul Below, Doug Ellett, Brad McDonald, and Shelley Capehart is gratefully appreciated.
Address for reprint requests and other correspondence: E. F. Coyle, Human Performance Laboratory, Dept. of Kinesiology and Health Education, The Univ. of Texas at Austin, Austin, TX 78712 (E-mail:).
This study was supported in part by a student grant from the Gatorade Sports Science Institute.
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. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society