Greater fluid loss does not fully explain the divergent hemodynamic balance mediating postexercise hypotension in endurance-trained men

Abstract

Following exercise, mean arterial pressure (MAP) is reduced ~5–10 mmHg from preexercise baseline. In nonendurance-trained males, postexercise hypotension results from peripheral vasodilation not offset by increased cardiac output (CO). By contrast, postexercise hypotension occurs through a reduction in CO from preexercise baseline in endurance-trained males. The reason(s) explaining these divergent responses remain unknown. Exercise at fixed percentage of peak oxygen consumption (V̇o_2peak) is associated with a greater rate of metabolic heat production in trained individuals and therefore elevated sweat rates, both when compared with untrained individuals. We hypothesized that greater fluid loss would explain the postexercise reduction in CO of endurance-trained males. Twelve endurance-trained males (Trained: V̇o_2peak, 64 ± 5 ml O₂·kg⁻¹·min⁻¹) cycled for 60 min at 60% V̇o_2peak (Trained_60%). On separate days, 12 nonendurance trained males (Untrained: V̇o_2peak, 49 ± 3 ml O₂·kg⁻¹·min⁻¹) cycled at 1) 60% V̇o_2peak (Untrained_60%), and 2) a rate of heat production equivalent to that achieved by the Trained group (Untrained_Matched). Fluid loss was similar between Trained_60% (−1.32 ± 0.20 kg) and Untrained_Matched (−1.32 ± 0.23 kg; P = 0.99) but was greater in these conditions relative to Untrained_60% (−0.95 ± 0.11 kg; both P < 0.01). During the final 30 min of postexercise supine recovery, MAP was similarly reduced by 5 ± 2 mmHg in all three conditions (P = 0.91). The reduction in MAP was mediated by a 0.5 ± 0.3 l/min reduction in CO from baseline in Trained_60% (P = 0.01). In contrast, CO returned to baseline following exercise during Untrained_Matched and Untrained_60% (both P ≥ 0.30). These data demonstrate that greater fluid loss does not fully explain the divergent postexercise hemodynamic responses observed in trained relative to untrained males.

NEW & NOTEWORTHY Even when matched for exercise-induced fluid loss, cardiac output was decreased in trained males but returned to baseline following exercise in their untrained counterparts. However, as per our hypothesis, reductions in stroke volume were similar between groups. This suggests that exercise-induced fluid loss is an important determinant of the stroke volume response during recovery but factors affecting heart rate such as exercise intensity and/or heat stress are also important determinants of postexercise hemodynamics.

INTRODUCTION

Recovery from dynamic exercise is associated with profound cardiovascular adjustments, highlighted by a marked and sustained reduction in mean arterial pressure of ~5–10 mmHg from resting values (12, 16–18, 20, 27, 29–31, 37). In sedentary individuals, the arterial baroreflex is reset to lower operating pressures following exercise (21), resulting in attenuated vasoconstrictor outflow for a given blood pressure relative to preexercise (13, 22). These alterations in conjunction with decreased sensitivity of the peripheral vasculature to sympathetic stimulation (21), increased nitric oxide bioavailability (19), and time-dependent histamine receptor activation (27, 30, 31) cause sustained systemic vasodilation. Cardiac output increases but not enough to offset the reduced peripheral resistance, and the resultant hemodynamic imbalance culminates in hypotension (17, 20).

In contrast to sedentary males (and endurance-trained females), the postexercise balance between vascular resistance and cardiac output differs in endurance-trained males. Rather than being mediated through systemic vasodilation, postexercise hypotension in these individuals manifests through a reduction in cardiac output from baseline values while vascular resistance returns to, and is maintained at, preexercise levels (12, 29, 37). Although the underlying physiological mechanisms remain unknown, it is possible that the postexercise reduction in cardiac output occurs secondary to greater exercise-induced fluid loss in endurance-trained males compared with their untrained counterparts (24, 29).

To date, only one study has directly compared the postexercise balance in vascular resistance and cardiac output between endurance-trained and untrained individuals (37). In that study, exercise was performed at a fixed percentage of peak aerobic capacity (V̇o_2peak). While this approach elicited similar relative levels of cardiovascular strain during exercise, it likely resulted in greater sweating-induced fluid loss in endurance-trained individuals (not quantified) (24). During exercise in a given environment, whole body sweat rate is determined by rate of metabolic heat production (14, 32). Since exercise at a fixed percentage of V̇o_2peak results in a greater rate of metabolic heat production in endurance-trained individuals, it has been surmised that greater sweating-induced fluid loss could explain the postexercise reduction in cardiac output within this population (24). In support of this hypothesis, Lynn et al. (29) found that fluid replacement during exercise prevented the postexercise reduction in cardiac output in endurance-trained males. However, these observations were not compared with those of untrained individuals. Thus it remains unknown if greater fluid loss during exercise can explain the divergent postexercise balance between vascular resistance and cardiac output in trained and untrained males.

The purpose of this study was to examine if greater fluid loss during exercise explains the divergent postexercise balance between vascular resistance and cardiac output between endurance-trained and untrained males. Following exercise at a fixed percentage of V̇o_2peak, it was hypothesized that greater fluid loss in endurance-trained males would be paralleled by postexercise reductions in cardiac output below baseline values whereas cardiac output would be maintained at baseline values in untrained males. It was further hypothesized that postexercise cardiac output would be reduced below baseline values in both endurance-trained and untrained males when fluid loss during exercise was similar between groups.

METHODS

Ethical approval.

The current study was approved by the University of Ottawa Health Sciences and Science Research Ethics Board and conformed to the Declaration of Helsinki. Verbal and written informed consent was obtained from all volunteers before their participation in the study.

Participants.

Twelve endurance-trained (Trained; age: 24 ± 4 yr) and twelve nonendurance-trained (Untrained; age: 24 ± 4 yr) males participated in the study. Endurance-trained males were highly active (≥60-min training sessions for competitive running, cycling, or rowing, 4–6 days/wk), while untrained males were habitually active (≤30 min of structured physical activity; typically, recreational team sports, 2–4 days/wk). Of note, the latter group could still be considered aerobically fit (confirmed by V̇o_2peak data; see results) but was not specifically endurance-exercise trained. All participants were free of cardiovascular, metabolic, renal, or respiratory diseases and were not taking prescription medication. Only males were recruited for this study given that Senitko et al. (37) showed that the mechanisms underpinning postexercise hypotension in both endurance-trained and untrained females were similar to those observed in untrained males.

Experimental procedures.

Endurance-trained participants completed one screening session and one experimental session, whereas untrained participants completed the screening session as well as two experimental sessions. All participants were asked to abstain from strenuous and/or prolonged physical activity 24 h before each session. Furthermore, participants were instructed to eat a small meal (i.e., dry toast and juice) no less than 2 h before arriving to the laboratory and to consume 500 ml of water the night before as well as 2 h before arriving to the laboratory on the day of each session. Body height, mass, surface area, and density, as well as V̇o_2peak were measured during the screening session. Body height was measured using an eye-level physician stadiometer (model 2391; Detecto Scale, Webb City, MO), while body mass was measured using a digital weight scale (model CBY150X; Mettler Toledo, Schwerznbach, Switzerland) with a high-performance weighing terminal (model IND560; Mettler Toledo). Measurements of body height and mass were subsequently used to calculate body surface area (11). Body density was measured using the hydrostatic weighing technique and used to estimate body fat percentage (38). V̇o_2peak was determined during a maximal incremental cycling protocol on an upright cycle ergometer (Corival Upright; Lode, Groningen, The Netherlands). The starting workload was set at 100 W and increased at a rate of 20 W/min until participants reached volitional fatigue or could no longer maintain a pedaling cadence of >50 rpm. Breath-by-breath oxygen consumption was measured by an automated indirect calorimetry system (Medgraphics Ultima; Medical Graphics, St. Paul, MN), and V̇o_2peak was taken as the highest average oxygen consumption recorded over 30 s.

Upon arrival to the laboratory on the day of the experimental session(s), participants provided a urine sample for the measurement of urine specific gravity. Participants then rested in a supine position for a ~60 min instrumentation period, after which 30 min of baseline measurements were taken. Thereafter, participants were transferred to the upright cycle ergometer where they performed 60 min of cycling exercise. At the end of exercise, participants were transferred back into the supine position for a 60-min recovery period. The entire protocol was performed within a room maintained at ~24°C. Both endurance-trained (Trained_60%) and untrained (Untrained_60%) individuals completed an experimental session during which exercise was performed at 60% V̇o_2peak. On a separate day, untrained participants returned to the laboratory to complete an experimental session during which exercise was performed at a rate of metabolic heat production equivalent to that achieved by the endurance-trained group at 60% V̇o_2peak (Untrained_Matched). To do so, endurance-trained and untrained individuals were matched in pairs. All sessions for a given pair were performed at the same time of day. For untrained participants, the two experimental sessions were performed in a counterbalanced manner and separated by ≥72 h. No fluid intake was allowed during the sessions. After each session, maximum skin blood flow for the chest, forearm, and calf was measured (see Measurements).

Measurements.

A pediatric thermocouple probe (Mon-a-therm; Mallinckrodt Medical, St. Louis, MO) was used for the continuous measurement of esophageal temperature. Skin temperature was measured at four sites using 0.3-mm diameter T-type (copper/constantan) thermocouples (Concept Engineering, Old Saybrook, CT) affixed to the skin with surgical tape. Mean skin temperature was calculated using the following weightings: quadriceps, 30%; calf, 30%; chest, 20%; and bicep, 20% (36). All temperature data were sampled at a rate of 15 s and simultaneously displayed and recorded on a personal computer with LabVIEW software (version 7.0; National Instruments, Austin, TX).

Local sweat rate was measured continuously on the chest using a ventilated sweat capsule covering a surface area of 2.8 cm², affixed to the skin with a double-sided adhesive ring and topical skin glue (Collodion HV; Mavidon; Medical Products, Lake Worth, FL). Dry compressed air was supplied to the capsule at a rate of 0.75 l/min while the water content of the effluent air from the capsule was measured with a high-precision dew point hygrometer (RH Systems, Albuquerque, NM). Local chest sweat rate was subsequently calculated from the water content of the effluent air multiplied by the flow rate of air through the capsule and normalized to the skin surface area under the capsule (mg·min⁻¹·cm⁻²).

Local skin blood flow was assessed at the chest, forearm, and calf on the right side of the body using laser-Doppler flowmetry (PeriFlux System 5000; Perimed, Stockholm, Sweden). Each laser‐Doppler probe was placed in an area of skin that, visually, did not appear overly vascularized and housed inside a local heating unit affixed to the skin using adhesive rings. Skin blood flow at each site was recorded at baseline before exercise and throughout recovery. Skin blood flow data are presented as a percentage of maximum values at each site, obtained by first setting the local heater to 42°C for 20 min and then to 44°C for an additional 25 min.

Systolic and diastolic blood pressures were measured at the brachial artery using an automated blood pressure monitor (Tango+; SunTech Medical, Morrisville, NC). Mean arterial pressure was subsequently calculated as follows: diastolic blood pressure + 1/3 × pulse pressure. Cardiac output was measured noninvasively using a carbon dioxide rebreathing system (Applied Electrochemistry, Brastrop, TX) in eight matched pairs of participants. Because of technical issues, cardiac output was measured using an inert gas rebreathing system in the other four pairs of participants (Innovision, Odense, Denmark). Heart rate was recorded continuously and stored at a sampling rate of 15 s using a Polar-coded WearLink transmitter, Polar RS400 interface, and Polar Trainer 5 software (Polar Electro, Oy, Finland). Venous blood samples were taken at baseline as well as 30 and 60 min following exercise using an indwelling catheter inserted into the antecubital vein of the right arm. Blood samples (~10 ml) were collected without stasis into a K2 EDTA vacutainer (BD Vacutainer, Franklin Lakes, NJ) and immediately analyzed for hemoglobin concentration and hematocrit content. Changes in plasma volume were subsequently calculated using the formula proposed by Dill and Costill (10).

Expired oxygen and carbon dioxide concentrations were measured using electrochemical gas analyzers (AMETEK model S-3A/1 and CD3A; Applied Electrochemistry, Pittsburgh, PA), while minute ventilation was measured with a turbine ventilometer. A gas mixture of known concentrations (~17% oxygen, ~4% carbon dioxide, balance nitrogen) and a 3-liter syringe were used to calibrate the gas analyzers and turbine ventilometer, respectively, ~20 min before the start of baseline data collection. Participants wore a full-face mask (model 7600 V2; Hans-Rudolph, Kansas City, MO) attached to a two-way T-shape nonrebreathing valve (model 2700; Hans-Rudolph). Metabolic rate during exercise was calculated from measurements of oxygen uptake and respiratory gas exchange ratio obtained every 30 s. Metabolic heat production was calculated as metabolic rate minus external work (14, 25).

Data analysis.

Values for esophageal and mean skin temperatures as well as local sweat rate and heart rate were calculated as an average of the final 5 min for each time period (preexercise, end exercise, and at 10-min intervals postexercise). Preexercise baseline values for mean arterial pressure and cardiac output were calculated as an average of three measurements taken at 10-min intervals (i.e., 10, 20, and 30 min of baseline resting). Postexercise, these variables are presented as a measurement taken every 10 min. Total peripheral resistance (mean arterial pressure ÷ cardiac output) and stroke volume (cardiac output ÷ heart rate) were subsequently calculated from the measurements of mean arterial pressure, cardiac output, and heart rate. Finally, at the 30th and 60th min of postexercise recovery, the change from preexercise baseline values was computed for thermal (i.e., esophageal and mean skin temperatures as well as local chest sweat rate) and cardiovascular (i.e., mean arterial pressure, cardiac output, total peripheral resistance, heart rate, and stroke volume) variables.

Statistical analysis.

The primary analysis of interest was to examine the postexercise balance between total peripheral resistance and cardiac output within each condition. Hemodynamic responses (i.e., mean arterial pressure, cardiac output, total peripheral resistance, heart rate, and stroke volume) were compared before and following exercise within each condition using a one-way ANOVA with the repeated factor of time (7 levels: preexercise and 10, 20, 30, 40, 50, and 60 min postexercise). Likewise, thermal responses (i.e., esophageal and mean skin temperatures as well as local chest sweat rate) were compared within each condition using a one-way ANOVA with the repeated factor of time (8 levels: preexercise, end exercise, and 10, 20, 30, 40, 50, and 60 min postexercise). Furthermore, the change in each variable from baseline was evaluated between conditions with a two-way mixed model ANOVA with the repeated factor of time (30 and 60 min of recovery for hemodynamic variables and end exercise, and 30 and 60 min of recovery for thermal variables) and nonrepeated (for Trained_60% and Untrained_Matched) or repeated (for Untrained_60% and Untrained_Matched) factor of condition. The change in plasma volume from baseline was assessed within each condition with a one-way ANOVA (repeated factor of time). Post hoc comparisons were carried out with using paired or independent samples t-tests (where appropriate), which were also employed for between-condition comparisons for physical characteristics, urine specific gravity, and changes in body weight and plasma volume. The level of significance was set at P ≤ 0.05. Statistical analyses were completed using the software package SPSS 23.0 for Windows (IBM, Armonk, NY). Values are presented as means ± 95% confidence interval unless otherwise indicated.

RESULTS

Participant characteristics.

Age (P = 0.96), height (P = 0.23), body mass (P = 0.23), body surface area (P = 0.88), and body fat percentage (P = 0.12) were similar between groups. By design, V̇o_2peak was greater in the endurance-trained group (64.4 ± 4.6 ml O₂·min⁻¹·kg⁻¹) relative to the untrained group (49.0 ± 3.2 ml O₂·min⁻¹·kg⁻¹_, P < 0.01; Table 1).

Table 1.

Participant characteristics

	Trained	Untrained
Age, yr	24 ± 4	24 ± 4
Height, m	1.81 ± 0.08	1.77 ± 0.08
Mass, kg	73.0 ± 6.7	77.0 ± 7.2
Body surface area, m2	1.94 ± 0.12	1.94 ± 0.13
V̇o2peak, mlO2·min−1·kg−1	64.4 ± 4.6	49.0 ± 3.2*
Body fat, %	12.3 ± 3	15.1 ± 3.3

All values are expressed as means ± SD; both groups n = 12 males. V̇o_2peak, rate of peak oxygen consumption as determined during a maximal incremental cycling protocol. Body fat (%) was assessed using the hydrostatic weighing technique.

^*P ≤ 0.05 vs. endurance trained.

Metabolic heat production.

By design, rate of metabolic heat production during exercise was greater in Trained_60% (405 ± 16 W/m²; equivalent to 60 ± 1% V̇o_2peak) and Untrained_Matched (403 ± 17 W/m²; equivalent to 73 ± 3% V̇o_2peak) compared with Untrained_60% (334 ± 12 W/m², both P < 0.01; equivalent to 61 ± 0% V̇o_2peak). Importantly, metabolic heat production was similar between Trained_60% and Untrained_Matched (P = 0.90).

Fluid balance.

Participants were adequately hydrated (1) before each session, and there were no differences in urine specific gravity between conditions (Trained_60%: 1.016 ± 0.005; Untrained_60%: 1.017 ± 0.014; Untrained_Matched: 1.017 ± 0.005; all P ≥ 0.89). A greater reduction in body mass was observed in Trained_60% (−1.32 ± 0.20 kg) and Untrained_Matched (−1.32 ± 0.23 kg) relative to Untrained_60% (−0.95 ± 0.08 kg; both P < 0.01) but no differences were observed between Trained_60% and Untrained_Matched (P = 0.99). Plasma volume was reduced from preexercise at 30 min (Trained_60%: −4 ± 2%; Untrained_Matched: −4 ± 1%; Untrained_60%: −3 ± 1%; all P ≤ 0.01) and 60 min (Trained_60%: −6 ± 2%; Untrained_Matched: −4 ± 2%; Untrained_60%: −4 ± 2%; all P ≤ 0.01) postexercise recovery. However, no differences in the change in plasma volume were observed between conditions (all P ≥ 0.29).

Body temperature and heat loss responses.

In all conditions, esophageal and mean skin temperatures were elevated from baseline values at the end of exercise (all P < 0.01) as well as throughout postexercise recovery (all P ≤ 0.05; Table 2). Chest sweat rate was increased from baseline at end exercise (all P < 0.01) but returned to preexercise values during the postexercise period (all P ≥ 0.06; Table 2). Chest and forearm skin blood flow returned to baseline in Trained_60% (all P ≥ 0.08) but remained elevated in Untrained_60% and Untrained_Matched (all P ≤ 0.05; except 40 and 50 min of recovery in Untrained_Matched, both P ≥ 0.07; Table 2). In all conditions, calf skin blood flow was similar to preexercise values throughout recovery (all P ≥ 0.09; Table 2).

Table 2.

Absolute body temperature and heat loss responses within each experimental condition

			Postexercise
	Preexercise	End Exercise	10 min	20 min	30 min	40 min	50 min	60 min
Trained60%
Esophageal temperature, °C	36.67 ± 0.13	38.07 ± 0.17*	37.35 ± 0.17*	37.26 ± 0.14*	37.18 ± 0.15*	37.11 ± 0.15*	37.04 ± 0.14*	36.95 ± 0.14*
Mean skin temperature, °C	32.40 ± 0.39	33.88 ± 0.37*	33.78 ± 0.50*	33.29 ± 0.56*	32.91 ± 0.57*	32.82 ± 0.55*	32.83 ± 0.45*	32.79 ± 0.51*
Chest sweat rate, mg·min−1·cm−2	0.09 ± 0.02	1.59 ± 0.27*	0.70 ± 0.40*	0.44 ± 0.32*	0.22 ± 0.16*	0.13 ± 0.06*	0.10 ± 0.04*	0.10 ± 0.03*
Chest skin blood flow, %max	11 ± 7		20 ± 12*	14 ± 8	11 ± 6	11 ± 6	12 ± 6	12 ± 7
Forearm skin blood flow, %max	11 ± 4		24 ± 8*	16 ± 7*	14 ± 8	13 ± 8	14 ± 8	14 ± 8
Calf skin blood flow, %max	9 ± 6		14 ± 7	10 ± 5	8 ± 4	8 ± 3	8 ± 3	7 ± 3
Untrained60%
Esophageal temperature, °C	36.58 ± 0.14	37.96 ± 0.21*	37.22 ± 0.18*	37.13 ± 0.16*	37.03 ± 0.14*	36.95 ± 0.14*	36.87 ± 0.14*	36.77 ± 0.15*
Mean skin temperature, °C	32.62 ± 0.35	33.89 ± 0.30*	33.74 ± 0.17*	33.51 ± 0.31*	33.36 ± 0.28*	33.28 ± 0.28*	33.23 ± 0.22*	33.24 ± 0.24*
Chest sweat rate, mg·min−1·cm−2	0.08 ± 0.01	1.21 ± 0.21*	0.52 ± 0.30*	0.28 ± 0.21*	0.09 ± 0.02*	0.08 ± 0.01*	0.07 ± 0.01*	0.08 ± 0.01*
Chest skin blood flow, %max	8 ± 4		15 ± 5*	12 ± 4*	10 ± 4*	10 ± 4*	10 ± 4*	11 ± 4*
Forearm skin blood flow, %max	8 ± 3		20 ± 9*	14 ± 4*	12 ± 4*	12 ± 4	13 ± 5*	14 ± 5*
Calf skin blood flow, %max	7 ± 1		13 ± 6	11 ± 4	10 ± 3	9 ± 3	9 ± 3	10 ± 3
UntrainedMatched
Esophageal temperature, °C	36.63 ± 0.13	38.54 ± 0.31*	37.50 ± 0.21*	37.34 ± 0.23*	37.27 ± 0.20*	37.19 ± 0.19*	37.13 ± 0.18*	37.07 ± 0.19*
Mean skin temperature, °C	32.73 ± 0.21	34.34 ± 0.34*	34.17 ± 0.34*	33.89 ± 0.37*	33.59 ± 0.39*	33.33 ± 0.41*	33.28 ± 0.39*	33.23 ± 0.36*
Chest sweat rate, mg·min−1·cm−2	0.08 ± 0.01	1.52 ± 0.18*	0.86 ± 0.35*	0.50 ± 0.28*	0.25 ± 0.22*	0.20 ± 0.2*	0.17 ± 0.19*	0.13 ± 0.08*
Chest skin blood flow, %max	8 ± 3		20 ± 10*	16 ± 8*	14 ± 8*	14 ± 9*	13 ± 10	14 ± 9*
Forearm skin blood flow, %max	10 ± 7		27 ± 7*	19 ± 6*	16 ± 5*	14 ± 5	15 ± 7*	16 ± 8*
Calf skin blood flow, %max	6 ± 5		13 ± 8	9 ± 5	7 ± 4	8 ± 5	8 ± 6	9 ± 9

Values are expressed as means ± 95% confidence intervals. Values indicate an average of the final 5 min of each time point. Endurance-trained males (n = 12) performed 60 min of cycling at 60% V̇o_2peak (Trained_60%). Individually matched untrained males (n = 12) completed 2 separate sessions consisting of 60 min of cycling at 1) 60% V̇o_2peak (Untrained_60%), and 2) a rate of metabolic heat production equivalent to that achieved by their endurance-trained counterparts (Untrained_Matched). Skin blood flow data were not collected during exercise.

^*P ≤ 0.05 vs. preexercise.

At end exercise, a greater increase in esophageal temperature from baseline was observed in Untrained_Matched relative to Trained_60% (P < 0.01) but responses were similar at the 30- and 60-min time points of recovery (both P ≥ 0.10; Table 3). Furthermore, changes in mean skin temperature and chest sweat rate from baseline did not differ between conditions (all P ≥ 0.18; Table 3). Increases in chest skin blood flow from baseline were greater in Untrained_Matched compared with Trained_60% at the 30-min time point of recovery only (P = 0.04) but similar at 60 min (P = 0.10; Table 3). No differences in the skin blood flow responses of the forearm and calf were observed between conditions (all P ≥ 0.11; Table 3).

Table 3.

Comparison of change in body temperature and heat loss responses from baseline between Trained_60% and Untrained_Matched

		Postexercise
	End Exercise	30 min	60 min
Esophageal temperature, °C
Trained60%	1.40 ± 0.21	0.51 ± 0.13	0.28 ± 0.13
UntrainedMatched	1.91 ± 0.28*	0.64 ± 0.12	0.44 ± 0.13
Mean skin temperature, °C
Trained60%	1.48 ± 0.32	0.50 ± 0.30	0.39 ± 0.26
UntrainedMatched	1.61 ± 0.28	0.86 ± 0.39	0.50 ± 0.43
Chest sweat rate, mg·min−1·cm−2
Trained60%	1.50 ± 0.26	0.13 ± 0.17	0.00 ± 0.03
UntrainedMatched	1.44 ± 0.17	0.17 ± 0.22	0.05 ± 0.08
Chest skin blood flow, %max
Trained60%		0 ± 4	1 ± 2
UntrainedMatched		5 ± 6*	5 ± 6
Forearm skin blood flow, %max
Trained60%		3 ± 6	3 ± 6
UntrainedMatched		6 ± 6	6 ± 6
Calf skin blood flow, %max
Trained60%		−1 ± 5	−3 ± 5
UntrainedMatched		1 ± 2	3 ± 6

Values are expressed as means ± 95% confidence intervals and represent the change from baseline at each time point. Endurance-trained males (n = 12) performed 60 min of cycling at 60% V̇o_2peak (Trained_60%). Individually matched untrained males (n = 12) completed 60 min of cycling at a rate of metabolic heat production equivalent to that achieved by their endurance-trained counterparts (Untrained_Matched). Skin blood flow data were not collected during exercise.

^*P ≤ 0.05 vs. Trained_60%.

A comparison of thermal responses within the untrained group (Table 4) revealed greater increases in esophageal temperature during exercise in Untrained_Matched vs. Untrained_60%, which persisted through the 30- and 60-min time points of recovery (all P ≤ 0.02; Table 4). No differences in the mean skin temperature response were observed between conditions (all P ≥ 0.17; Table 4). However, the increase in chest sweat rate from baseline was greater in Untrained_Matched compared with Untrained_60% at end exercise and 30 min of recovery (both P ≤ 0.03; Table 4). The responses of chest, forearm, and calf skin blood flow were similar between conditions (all P ≥ 0.11; Table 4).

Table 4.

Comparison of change in body temperature and heat loss responses from baseline between Untrained_60% and Untrained_Matched

		Postexercise
	End Exercise	30 min	60 min
Esophageal temperature, °C
Untrained60%	1.31 ± 0.26	0.45 ± 0.14	0.19 ± 0.13
UntrainedMatched	1.91 ± 0.28*	0.64 ± 0.12*	0.44 ± 0.13*
Mean skin temperature, °C
Untrained60%	1.27 ± 0.47	0.73 ± 0.48	0.62 ± 0.44
UntrainedMatched	1.61 ± 0.28	0.86 ± 0.39	0.50 ± 0.43
Chest sweat rate, mg·min−1·cm−2
Untrained60%	1.13 ± 0.21	0.00 ± 0.02	−0.01 ± 0.02
UntrainedMatched	1.44 ± 0.17*	0.17 ± 0.22	0.05 ± 0.08
Chest skin blood flow, %max
Untrained60%		2 ± 1	2 ± 3
UntrainedMatched		5 ± 6*	5 ± 6
Forearm skin blood flow, %max
Untrained60%		4 ± 2	5 ± 3
UntrainedMatched		6 ± 6	6 ± 6
Calf skin blood flow, %max
Untrained60%		2 ± 3	2 ± 3
UntrainedMatched		1 ± 2	3 ± 6

Values are expressed as mean ± 95% confidence intervals and represent the change from baseline at each time point. Untrained males (n = 12) completed 2 separate sessions consisting of 60 min of cycling at 1) 60% V̇o_2peak (Untrained_60%), and 2) a rate of metabolic heat production equivalent to that achieved by their endurance-trained counterparts (Untrained_Matched). Skin blood flow data were not collected during exercise.

^*P ≤ 0.05 vs. Untrained_60%.

Hemodynamic responses.

Throughout postexercise recovery, mean arterial pressure was reduced from preexercise values in Trained_60% (all P ≤ 0.05; Table 5). This was accompanied by a reduction in cardiac output from baseline through the final 30 min of recovery (P < 0.01), while total peripheral resistance was unchanged (P ≥ 0.19; Fig. 1). Furthermore, heart rate was elevated whereas stroke volume was reduced from preexercise values throughout recovery (all P < 0.01; Fig. 1).

Table 5.

Mean arterial pressure responses

		Postexercise
	Preexercise	10 min	20 min	30 min	40 min	50 min	60 min
Trained60%	88 ± 3	86 ± 4	85 ± 3	85 ± 3*	84 ± 3*	85 ± 2*	85 ± 3*
Untrained60%	85 ± 4	83 ± 7	82 ± 4*	79 ± 5*	79 ± 4*	79 ± 4*	79 ± 3*
UntrainedMatched	86 ± 5	83 ± 5	81 ± 5	80 ± 4*	79 ± 5*	80 ± 5*	80 ± 4*

Values are expressed as means ± 95% coincidence intervals in mmHg. Endurance-trained males (n = 12) performed 60 min of cycling at 60% V̇o_2peak (Trained_60%). Individually matched untrained males (n = 12) completed 2 separate sessions consisting of 60 min of cycling at 1) 60% V̇o_2peak (Untrained_60%), and 2) a rate of metabolic heat production equivalent to that achieved by their endurance-trained counterparts (Untrained_Matched).

^*P ≤ 0.05 vs. preexercise.

Fig. 1.

Cardiac output (top left), total peripheral resistance (top right), heart rate (bottom left), and stroke volume (bottom right) before (Pre) and following (recovery) exercise in endurance-trained males (n = 12) who performed 60 min of cycling at 60% V̇o_2peak. (Trained_60%).Values presented as means ± 95% confidence intervals. *P ≤ 0.05, different from preexercise.

In Untrained_60%, mean arterial pressure was reduced from preexercise throughout recovery (all P < 0.01; Table 5). Cardiac output returned to baseline values (all P ≥ 0.07) while total peripheral resistance was reduced from preexercise (Fig. 2B; all P ≤ 0.04; except at 30 min, P = 0.06; Fig. 2). Heart rate was elevated (all P < 0.01) and stroke volume was reduced (all P ≤ 0.02; except at 60 min, P = 0.07) compared with baseline (Fig. 2).

Fig. 2.

Cardiac output (top left), total peripheral resistance (top right), heart rate (bottom left), and stroke volume (bottom right) before (Pre) and following (recovery) 60 min of cycling exercise at 60% V̇o_2peak in untrained males (n = 12) (Untrained_60%). Values presented as means ± 95% confidence intervals. *P ≤ 0.05, different from preexercise.

Like Untrained_60%, the postexercise reduction in mean arterial pressure in Untrained_Matched (all P ≤ 0.04; Table 5) was due to reduced total peripheral resistance from preexercise (all P ≤ 0.04) while cardiac output was unchanged (all P ≥ 0.14; Fig. 3). Likewise, heart rate was elevated and stroke volume was reduced from preexercise levels (all P < 0.01; Fig. 3).

Fig. 3.

Cardiac output (top left), total peripheral resistance (top right), heart rate (bottom left), and stroke volume (bottom right) before (Pre) and following (recovery) 60 min of cycling exercise in untrained males (n = 12) performed at a rate of metabolic heat production equivalent to that performed by trained males (Untrained_Matched). Values presented as means ± 95% confidence intervals. *P ≤ 0.05, different from preexercise.

The reduction in mean arterial pressure from preexercise baseline was similar between Trained_60% and Untrained_Matched at the 30-min (−4 ± 3 vs. −6 ± 4 mmHg) and 60-min (−3 ± 3 vs. −6 ± 4 mmHg) time points of recovery (both P ≥ 0.29). Furthermore, there was a greater change in cardiac output from baseline in Trained_60% (both P ≤ 0.03) while the difference in total peripheral resistance from preexercise was greater in Untrained_Matched (both P ≤ 0.01; Fig. 4). Finally, the difference in the postexercise cardiac output response was mediated through a greater increase in heart rate in Untrained_Matched vs. Trained_60% (both P < 0.01), as the stroke volume responses were similar between conditions (both P ≥ 0.87; Fig. 4).

Fig. 4.

Change (∆) in cardiac output (top left), total peripheral resistance (top right), heart rate (bottom left), and stroke volume (bottom right) from baseline during postexercise recovery in Trained_60% (gray bars) and Untrained_Matched (black bars). Values presented as means ± 95% confidence intervals. *P ≤ 0.05, different from Trained_60%.

Similar reductions in mean arterial pressure from pre- to postexercise were observed between Untrained_60% and Untrained_Matched at both the 30-min (−5 ± 3 vs. −6 ± 4 mmHg) and 60-min (−5 ± 3 vs. −6 ± 3 mmHg) time points (both P ≥ 0.58). Likewise, the cardiac output and total peripheral resistance responses were not different between conditions (all P ≥ 0.20; Fig. 5). However, heart rate was elevated in Untrained_Matched compared with Untrained_60% at both time points (both P < 0.01) whereas stroke volume was reduced in Untrained_Matched at the 60-min (P = 0.04) but not 30-min time point (P = 0.17; Fig. 5).

Fig. 5.

Change (∆) cardiac output (top left), total peripheral resistance (top right), heart rate (bottom left), and stroke volume (bottom right) from baseline during postexercise recovery in Untrained_60% (gray bars) and Untrained_Matched (black bars). Values presented as means ± 95% confidence intervals. *P ≤ 0.05, different from Untrained_60%.

DISCUSSION

The primary finding of this study is that, even when exercise-induced fluid loss is matched, the postexercise balance between cardiac output and vascular resistance differs between endurance-trained and untrained males. The ~5 mmHg reduction in mean arterial pressure occurred secondary to reduced cardiac output from baseline values in endurance-trained individuals following dynamic exercise at 60% V̇o_2peak (Trained_60%; Fig. 1). By contrast, in untrained males this response was mediated via decreased peripheral resistance regardless of whether exercise was performed at 60% V̇o_2peak (Untrained_60%; Fig. 2) or at an exercise intensity that elicited a similar whole body fluid loss to endurance-trained males (Untrained_Matched; Fig. 3). It is important to highlight that similar reductions in stroke volume were observed in Trained_60% and Untrained_Matched. However, comparatively greater postexercise increases in heart rate from preexercise were observed in the latter group, leading to the divergent cardiac output responses. Regardless, our findings are consistent with previous reports demonstrating that postexercise hypotension primarily arises from decreased cardiac output in trained males (12, 29, 37) and systemic vasodilation in their untrained counterparts (17, 20, 37).

Building on the study by Senitko et al. (37), Lynn et al. (29) suggested that fluid status is an important determinant mediating postexercise hypotension in endurance-trained males. In their study, endurance-trained males consumed water during exercise in a thermoneutral environment to offset fluid losses from sweating. Although fluid consumption did not alter the magnitude of postexercise hypotension relative to a control (no fluid) condition, it did prevent cardiac output from falling below baseline values (29). It is important to note, however, that Lynn et al. (29) did not compare the responses of endurance-trained men to a group of untrained individuals. As such, it remained unknown whether greater fluid loss is responsible for the divergent mechanisms mediating postexercise hypotension between endurance-trained and untrained men. In a given environment, whole body sweat rate during exercise and thereby fluid loss are dependent on rate of metabolic heat production (14, 32). When exercise is performed at a fixed percentage of V̇o_2peak, such as in previous studies that have compared postexercise hemodynamics between trained and untrained individuals (16–18, 20, 27, 29–31, 37), endurance-trained individuals exercise at a greater rate of metabolic heat production and therefore incur greater fluid loss from sweating.

As proposed by Kenny and Gagnon (24), if greater fluid loss during exercise contributes to the divergent hemodynamic balance mediating postexercise hypotension in endurance-trained males, one might expect that hemodynamic balance would be similar between trained and untrained individuals when fluid loss is matched. This response would likely occur through reduced stroke volume secondary to decreased central venous pressure. Indeed, stroke volume was reduced by ~20 ml in both Trained_60% and Untrained_Matched (Fig. 4). That said, despite differences in fluid loss the reductions in plasma, volumes were similar between conditions, suggesting that the observed differences in fluid loss were from extravascular sources. Regardless, the current and previous data (29, 37) fit with the idea that the postexercise reduction in stroke volume is influenced, at least partially, by the magnitude of exercise-induced fluid loss.

Despite similar magnitudes of fluid loss when metabolic heat production was matched during exercise, the postexercise reduction in stroke volume was offset by a greater increase in heart rate within untrained individuals such that cardiac output returned to baseline levels in Untrained_Matched (Fig. 2) but remained attenuated throughout recovery in Trained_60% (Fig. 1). The reasons underlying the divergent hemodynamic responses cannot be discerned from the current data but may be due to the relative intensity at which exercise was performed (~73% V̇o_2peak in Untained_Matched vs. 60% V̇o_2peak in Trained_60%).

Increasing exercise intensity is associated with dose-dependent elevations in heart rate and indexes of cardiac sympathetic activity and autonomic function (i.e., systolic time intervals, heart rate variability) that persist into recovery (33, 34), which may explain the greater postexercise cardiac output response in Untrained_Matched relative to Trained_60%. Parenthetically, an exercise intensity-dependent effect on heart rate and thereby cardiac output is also supported by the fact that heart rate was elevated from baseline by 7–11 beats/min (Fig. 4) and 7–13 beats/min (Fig. 5) during recovery in Trained_60% and Untrained_60%, respectively. Moreover, we observed a greater increase in esophageal temperature in Untrained_Matched compared with Trained_60%. This may have also contributed to the elevated heart rate in Untrained_Matched given that heat stress is associated with elevated sympathetic activity and/or enhanced cardiac function (6–9, 35, 40). In support of this notion, Lynn et al. (29) reported that, despite elevated fluid loss, the postexercise reduction in cardiac output in trained individuals was abolished when exercise was performed in a hot environment (similar to the observation made when fluid was provided to offset sweat losses). Thus the postexercise cardiac output response appears sensitive to intensity- and/or heat-stress-induced augmentation of heart rate, in addition to body fluid status.

Considerations.

A primary limitation of this study was that measures of central hemodynamics (i.e., central venous pressure, blood volume, pulmonary wedge pressure, indexes of cardiac function, etc.) were not feasible. Furthermore, our design did not allow us to fully determine the precise interactions among exercise intensity, heat load, and fluid loss (among other factors) on blood pressure regulation. While potential mechanisms explaining the observed responses are discussed in preceding paragraphs, other factors beyond the scope of this study should also be considered. For example, alterations in heart rate and stroke volume, and the resulting influence on cardiac output, likely do not occur independently. Despite the large difference in relative exercise intensity (73 and 60% V̇o_2peak), an ~300-ml greater reduction in body weight was observed in Untrained_Matched (~1.3 kg or 1.7% of body weight) compared with Untrained_60% (~1.0 kg or 1.3% of body weight). Nonetheless, the change in plasma volume from pre- to postexercise was comparable between groups/conditions. That said, the difference in fluid loss between conditions was presumably consistent with that observed between the trained and untrained males in the study by Senitko et al. (37) given that the difference in V̇o_2peak between groups (~13 ml·min⁻¹·kg⁻¹) was similar to the current study (~15 ml·min⁻¹·kg⁻¹). Furthermore, as highlighted in the above sections, differences in sympathetic activation stemming from between-group differences in relative exercise intensity and/or hyperthermia may have also influenced responses independently of changes in fluid status. Clearly, future work is required to separate the influence of exercise intensity-related factors and fluid loss, with diuretic-induced dehydration, for example, on postexercise hemodynamics.

Perspectives.

Exploration of the mechanisms underpinning postexercise physiological control is an important and on-going area of study. While in most circumstances altered cardiovascular function following exercise is nonsymptomatic, reduced orthostatic tolerance can be experienced during recovery from exercise. This response may exacerbated by exercise-induced dehydration, especially when performed in a warm environment (15, 20, 23, 28), and may also be linked to an abrupt suppression of the heat loss responses (i.e., cutaneous vasodilation and sweating) following exercise cessation resulting in prolonged elevations in body temperature (26). Moreover, as highlighted in a recent review by Luttrell and Halliwill (28), the postexercise period can be seen as a “crystal ball,” in that the recovery of certain physiological systems (e.g., blood pressure and heart rate) provides important insight into autonomic function (2, 3, 39) and clinically relevant risk factors (4, 5). The findings of the current study therefore add to a growing body of research describing the risks and benefits of the dynamic postexercise period.

Conclusion.

In the current study, we report divergent hemodynamic mechanisms underpinning postexercise hypotension in endurance-trained and untrained males when exercise-induced fluid loss is similar. Similar reductions in stroke volume were observed between groups, which were compensated for by elevated heart rate in untrained but not trained males, leading to divergent cardiac output responses. Additional studies are needed to further examine the mechanisms mediating the imbalance between postexercise stroke volume and heart rate in endurance-trained males.

GRANTS

This study was supported by the Natural Sciences and Engineering Research Council Discovery Grant RGPIN-06313-2014 and Discovery Grants Program-Accelerator Supplements Grant RGPAS-462252-2014 (to G. P. Kenny). G. P. Kenny is supported by a University of Ottawa Research Chair Award. R. D. Meade is supported by a Natural Sciences and Engineering Research Council of Canada Alexander Graham Bell Graduate Scholarship.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

G.P.K. conceived and designed research; R.D.M. performed experiments; R.D.M. analyzed data; R.D.M., C.G.C., D.G., and G.P.K. interpreted results of experiments; R.D.M. prepared figures; R.D.M. drafted manuscript; R.D.M., C.G.C., D.G., and G.P.K. edited and revised manuscript; R.D.M., C.G.C., D.G., and G.P.K. approved final version of manuscript.