JOURNALOF
APPLIED
PHYSIOLOGY
Vol. 39, No. 6, December
1975.
Printed
in U.S.A.
Effect of exercise and thermal stress on plasma
volume
M. H. HARRISON, R. J. EDWARDS, AND D. R. LEITCH RAF Institute of Aviation Medicine, Farnborough, Hampshire, England
HARRISON, M. H., R. J. EDWARDS, AND D. R. LEITCH. E$ect of exercise and thermal stress on plasma volume. J. Appl. Physiol. 39(6) : 925-931. 1975.-S’ ix male subjects exercised for 50 min at 25y0 (light exercise) and 55y0 (moderate exercise) of their estimated aerobic capacities in environments of 42°C db, 35°C wb and 30°C db, 24°C wb, respectively. Alterations in the hematocrit, hemoglobin, and plasma protein concentrations, and in the activity of an injected aliquot of isotopically labeled albumin were each used to calculate the percentage change in plasma volume occurring during exercise and recovery. Changes in each measure were consistent with a reduction in plasma volume during exercise and a return to preexercise levels during recovery. There was no significant difference between the measures when exercising in the heat, but during the more severe exercise in the cooler environment disproportional changes in protein, hematocrit, and hemoglobin were observed. Disproportional changes were also seen during the recovery phase, when the hematocrit and hemoglobin concentration indicated a more rapid return of the plasma volume to preexercise levels than did either the plasma protein concentration or albumin activity. During moderate exercise and recovery there was a 1 y0 decrease in red cell volume. It is concluded that exercise accelerates the rate of protein movement from extravascular compartments to the intravascular compartment, leading to elevated plasma protein levels during recovery which favor the return of water to the intravascular space. Hemoglobin concentration is considered to be the most reliable measure of plasma volume change during exercise.
in plasma volume being dominated higher environmental temperature, more severe work load.
in the one by the in the other by the
METHODS
Experimental
Design
Six male subjects exercised for 50 min in a hot environment of 42°C dry bulb (Tdb), 35°C wet bulb (Twb; relative humidity (rh) 63 %), and again 3 mo later in a warm environment where Tdb = 30°C and T,b = 24°C (rh = 62 0. Air movement was 1.5 m . s- l. The exercise was preceded, and followed, by 50-min control and recovery periods, respectively, where Tdb = 30°C and Twb = 24°C. Exercise was performed on a Monark bicycle ergometer at a pedal speed of 50 rev l rein- l. The period of 3 mo between the two groups of experiments was to allow all radioactivity from the injected isotope (see below) to be cleared from the blood. It was considered that any effect of time on subject responses to heat and exercise would be minimal, and did not warrant a randomization of the experimental design. The subjects were all healthy volunteers from within the Institute, and none was acclimatized to heat. Details are given in Table 1. All experiments were carried out during the morning in the winter months. A light breakfast was hematocrit; hemoglobin; plasma protein; albumin taken 2 h before an experiment, after which no food or liquid was allowed until the end of the recovery period. Only shorts and shoes were worn. THE ALTERATIONS in the hematocrit and the plasma protein Subjects reported to the control climatic chamber 45 concentration which occur in resting subjects exposed to a min before the start of the 50-min control period, and high environmental temperature appear to be quantityrested quietly. They remained seated in a wheelchair tively representative of plasma volume change over a 2-h throughout the experiment except for weighings (which period (15). This may not be the case during exercise if, were carried out at the end of each 50-min control, exeras reported by several investigators, protein is transferred cise, and recovery period), and for the exercise period. from the extravascular space to the intravascular space Subjects were weighed nude after towelling down, and (16, 17, 20, 24, 26, 27). correction was made for blood taken. The present study examines the relationship between After the subject had been instrumented and weighed, measures of change in plasma volume based on red cell an intravenous catheter was introduced 7 cm into the antevolume (hematocrit and hemoglobin concentration), and cubital vein of the left arm under local anesthesia. The intravascular protein (total plasma protein concentration catheter was filled with heparinized saline (50 IU *ml-l) and radioiodinated human serum albumin (RISA) acafter taking each blood sample. tivity), under two conditions of exercise and climate, and At the end of the control period the subject was wheeled during recovery from that exercise. These conditions were into an adjacent climatic chamber, where he mounted a light exercise in a hot environment, and more severe bicycle, and exercised at a predetermined rate for 50 min. exercise in a cooler environment. Both were intended to This work rate represented 25 % of the subject’s estimated represent realistic work/environment situations, any change maximum aerobic capacity in the hot environment (light 925
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
926
HARRISON,
TABLE
AND
LEITCH
1. Subject data
25%i702 max;42°C Subj
Age, yr
w
kg
Work
load, W
~1 2 3 4 5 6(i)* 6 (ii) Mean
EDWARDS,
sf~ SE
27 30 31 36 39 41 35 33 &
AT,, = change of the experiment,
56.0 77.8 68.5 81.9 71.6 72.0 79.2 2
72.4
j= 3.3
in aural temperature; and was replaced
65.4 58.9 54.0 54.0 54.0 54.0
57.0
k
+2.06 +1.44 +1.53 +1.31 +l.lS +1.28
1.9
db, 35OC wb
ATac after 50 min, “C
+1.46
ABW = change by subject 6(ii).
-
-1.02 weight.
exercise), and 55 % of his estimated maximum aerobic capacity in the warm environment (moderate exercise; Table 1). Aerobic capacity was estimated using the technique described by Astrand and Rhyming (2). After completing the exercise the subject was returned to the control chamber for the 50-min recovery period. Measurements Blood samples were taken at 5-min intervals throughout the control, exercise, and recovery periods for the measurement of hematocrit (Hct), hemoglobin (Hb), total plasma protein (PP), plasma osmolarity, and plasma RISA activity. The average sample volume was 7 ml, and the total volume of blood taken did not exceed 250 ml. Microhematocrits were obtained from whole blood. Determinations were in quadruplicate, with no correction for trapped plasma (CF = 12,000 X s), or for differences between venous and whole body hematocrit (1, 10, 15). Plasma protein was measured by the method of Weichselbaum (30), and Hb by the cyanmethemoglobin method. Plasma osmolarity was determined by freezing point depression using a Fiske osmometer. All determinations were from which the precision of the methods in duplicate, could be calculated. This error, however, was generally exceeded by within-subject variability during the control periods, which was assessed from 120 observations as: hematocrit, =t 1.1 %; plasma protein, h2.7 %; hemoglobin, *1.5%; osmolarity, *1.2%. A measure of the rate of loss of plasma albumin from the intravascular space was obtained from a single injection of RISA (20 mg of labeled albumin of specific activity 0.20vein 0.25 PCi l mg-1, in 1 ml saline) into the antecubital of the right arm 15 min before the start of the control period. The plasma RISA activity during the three experimental periods was determined by counting three 0.5-ml plasma aliquots of each 7 ml blood sample using a gamma spectrometer. The counting error was Z& 1 .O %, and the mean variation between the three aliquots was & 1.5 %. The rate of loss of plasma albumin from the intravascular space was used to calculate the percentage change in plasma volume during exercise and recovery as described below. Heart rate (fn) and deep body temperature (T,,) were measured at 5-min intervals during exercise-the former
zk 0.20 * Subject
30°C db, 24°C wb %A BW after min
147.2 132.4 110.4 122.6 110.4
+1.33 +0.34 +0.49 +1 .OO +1.29
-1.08 -0.70 -0.75 -0.64 -0.54
147.2
+1.81
-0.62
load, W
128.4 6(i)
vO2maxi
ATac after 50 min, “C
Work
-1.93 -0.95 -0.81 -0.57 -0.73 -1.10
=t 0.13 in body
55%
%A BW after 50 min
rf= 6.8
was unable
+1.04
rt
to participate
0.23
-0.72
50
=t 0.08
in the second
phase
from an electrocardiogram (ECG), and the latter to within 0.05”C using a calibrated ear thermistor located in the auditory canal close to the tympanic membrane. The thermistor probe was thermally insulated from the external environment by packing the auditory canal with cotton wool, and positioning a thick wad of cotton wool over the surface of the ear, as described by Marcus (21). The body temperature so measured accurately reflects sublingual temperature, and is probably similar to that of the tympanic membrane (7). Sweat loss was estimated from change in body weight (*lo g>* Oxygen consumption during the 50-min exercise periods was measured from three paired Douglas bag collections of 2 min taken between 5 and 10 min, 30 and 35 min, and 40 and 45 min. Data Analysis Using the absolute values for the variables measured, the three experimental periods were compared by analysis of variance. Since the subjects did not differ significantly, the subject data were combined, and Table 2 gives mean observed values for each measure. The values were transformed into equivalent plasma volume change using the equations of van Beaumont, Greenleaf, and Juhos (5), and represented by linear regressions as shown in Table 3, and Figs. 1 and 2; subject variability is indicated by 95 % confidence limits. Values for the significance levels of the regressions refer to slopes different from zero. The rate of loss of RISA from the intravascular space during the control period was described, for each subject, by linear regression equations, and extrapolated back to give the RISA activity at the time of injection. This was compared with the activity of the standard to derive the resting plasma volume. The equations were also extrapolated forward over the exercise and recovery periods to estimate the plasma RISA activity which might have been expected had no exercise supervened. This required the assumption that the rate of loss of RISA would have remained essentially linear. The theoretical activities so calculated were subtracted from the actual activities observed during exercise and recovery, and these differences expressed as a percentage of the isotope activity at the time of injection.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA
VOLUME
DURING
927
EXERCISE
TABLE 2. Change in Hct, Hb, plasma protein, and RISA during light and moderate exercise, and during recovery from that exercise (mean of 6 subj; for environmental conditions see text) -_ Hct Condition
Time,
min
25oJob
Control Exercise
Recovery
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
(5) (10) (15) (20) (25) (30) (35) w (45) (50)
Hb, g. 100 ml-1 ssg
max
25%902msx
902 max
Plasma
550/o
902
Protein,
25% 902max
max
g. 100 ml-1 55%
RISA,
max
902
25%
counts - min-1. g-1 max
vO2
55% ---
.iTO2 max
42.5
40.8
14.66
13.73
7.06
6.74
1,055*
829”
44.1 44.3 44.4 44.4 45.1 45.3 45.5 45.6 45.8 46.0
43.6 43.6 43.8 43.8 44.0 44.0 43.7 43.6 43.7 43.6
15.20 15.29 15.32 15.35 15.49 15.59 15.76 15.75 15.96 16.04
14.78 14.75 14.91 14.90 15.02 15.00 14.85 14.87 14.80 14.93
7.52 7.59 7.61 7.70 7.74 7.87 7.89 7.86 7.91 7.94
7.29 7.35 7.44 7.49 7.48 7.58 7.59 7.51 7.52 7.49
1,126 1,123 1,140 1,141 1) 141 1,141 1,144 1,141 1,148 1,154
898 906 911 911 901 890 890 886 879
45.5 45.5 44.6 44.0 43.9 43.7 43.5 43.1 43.0 42.9
43.4 42.2 41.7 41.2 41 .o 41 .o 40.5 40.6 40.6 40.5
15.77 15.74 15.36 15.08 15.10 14.97 14.97 14.81 14.99 14.74
14.86 14.58 14.12 14.11 13.94 13.89 13.84 13.89 13.90 13.86
7.87 7.82 7.63 7.55 7.55 7.50 7.48 7.45 7.39 7.37
7.53 7.32 7.20 7.10 6.98 6.96 6.86 6.93 6.96 6.90
1,141 1,132 1,104 1,077 1,063 1,063 1,037 1,046 1,048 1,030
868 844 829 817 808 803 792 790 790 780
Control: mean of n = 60 observations before the start of exercise, and 65 min
over 50-min after injection
control period except: of the isotope.
* where
n = 6, and
is the plasma
RISA
activity
5 min
TABLE 3. Regression equations, derived from the data of Table 2, relating Hct, Hb, plasma protein, and RISA to change in plasma volume (calculated as mean of 6 subj) Percent Condition
25%VO2 42 ‘c&-,,
max
35 ‘Cwb
55% VOZ max 30 ‘c&, 24 OCwb X
= 5-50 min
Y
Hct
Decrease
in Plasma
Hb
r
Volume
Calculated
Plasma
protein
from r
RISA
Y
Exercise Recovery
5.63 12.13
+ -
0.1584x 0.2316~
0.99 -0.97
4.58 11.84
+ 0.1963x - 0.2296x
0.99 -0.91
5.61 9.94
+ -
0.1166x 0.1266x
0.97 -0.94
6.32 12.57
+ 0.1419x - 0.0956x
0.99 -0.79
Exercise Recovery
11.37 7.82
-
0.0014x 0.2199x
0.04 -0.90
12.49 10.24
-
0.27 -0.84
8.23 9.15
+ -
0.0553x 0.1682x
0.74 -0.88
8.18 8.05
+ -
0.66 -0.85
of exercise,
or recovery
from
exercise;
0.0166x 0.2178x
r = correlation
RESULTS
After 50 min of light exercise (25 % vo 2 max) mean fn was 158 beatsmin-1 (range 132-174), and mean T., was 38.58”C (range 38.28-39.10). After 50 min of moderate exercise (55 % vo2 max> mean fn was 170 beats emin(range 152-180), and mean T,, was 37.91”C (range 37.1038.79). The magnitude of the increase in T., (AT,,) is given for each subject in Table 1. Mean 02 consumption (voz) during light exercise was 1.06 lmin-l (range 0.87-1.42), and during moderate exercise was 2.04 1 l rein-1 (range 1.69-2.5 1). On the three occasions during exercise that 02 consumption was determined, mean values for 25 % and 55 % VOW max were 0.93, 1.11, 1.14, and 1.81, 2.10, and 2.22 lmin-l, respectively. Weight loss due to sweating averaged 0.69 kg during light exercise (range 0 047-l l 08), and 0.51 kg during moderate exercise (range 0.38-0.61); Table 1 relates these to body weight ( %ABW). There was no significant correlation between changes
0.0364x 0.0972x
coefficient.
in fn, T,,, vo2, or BW, of plasma volume. Plasma Volume During
and changes
in the four
measures
Light Exercise in the Hot Environment
Exercise. There was no statistically significant difference between change in plasma volume calculated from Hct, Hb, PP, or RISA. The regression equations describing each measure during exercise were significant (P < 0.001; Table 3). Concentration increases occurring in the first 5 min were equivalent to a reduction in plasma volume of 6 %; increases during the succeeding 45 min (Table 2) were attributable to a further 6-8 % decrease in plasma volume (Figs. 1 and 2). The mean PP level after 50 min indicated a change in plasma volume significantly less than did the Hct, Hb, and RISA levels (P < 0.01; Fig. 3A). Recovery. Hematocrit, Hb, PP, and RISA levels decreased during recovery (Table 2); the linear equations relating these to plasma volume change were significant at P < 0.001 except for RISA where P < 0.01 (Table 3). Both Hct and Hb indicated a return of plasma volume to near
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
HARRISON,
928
rA
+I
c z
T
+2 -
E
,
Ob,,
2
,
,
,
,
-
; 3 ;
-2-
E 3 c
-6 -
-4 -
‘*,,
-8-
.-C z
-10
-
:: L t: n
-12
-
3 E ZC >‘G
EDWARDS,
AND
LEITCH
-2-
-4oz E‘ we -620 0, E -8.-cs 0) z -10 3E 2 2 -12 olc z
-14 -
-14-
5 -2s -4 kna iia -6,oLa -8.5 E XtO -10 fii z 6
-12 -
*0
-14 -16, t
I
I
1
10
20
30
1
40
1I
50 /
1
1
60 70 (10) (20)
k-----Exercise-p+-
,
80 (30)
I
,
90 100 (40) $0)
Recovery---d lime
(min)
FIG. 2. Change in plasma volume during exercise, and during recovery from exercise calculated from A, the plasma protein conof isotopically labeled albumin. centration, and B, the plasma activity Details as Fig. 1 (note data in A from (14)).
1. Change in plasma volume during exercise, and during recovery from exercise calculated from A, the hematocrit (Hct), and B, the hemoglobin (Hb) concentration: 1) at 25y0 Voz max; 42°C db; 24”Cwb 35”Cwb (X-X); 2) at 55% J?o, max; 30°C db; Vertical bars represent 95% confidence limits about the (0 -0 -0). regression lines (Table 3) for mean of 6 subjects. Also, change in plasma volume calculated from Hct (- - -) for mean of 6 subjects resting at 50°C db, 38 ‘C wb (from (14)). FIG.
the preexercise level after 50 min (Fig. 1, A and B). Plasma protein, however, remained elevated at this time (Table Z), suggesting that plasma volume remained 4 % below the control level (Fig. 24). Labeled albumin activity after 50-min recovery was equivalent to a 9 % reduction in plasma volume compared with control (Fig. ZB). The Hct and Hb regression coefficients were significantly different from the PP and RISA coefficients (P < 0.01; Fig. 34 . Plasma Volume During Moderate Exercise in the Thermoneutral Environment Exercise. The increase in Hct and Hb observed after 5 min of exercise (Table 2) was equivalent to a 12 % decrease in plasma volume. Thereafter no further change occurred (Fig. 1, A and B) ; the regressions were not statistically significant (Table 3), the correlation coefficients reflecting random error about the regression line (i.e., within-subject variability and experimental error).
The initial decrease in plasma volume was only 8 % when calculated from increases in PP concentration and RISA activity. During the last 45 min of exercise PP levels continued to increase, and RISA activity, although decreasing, did so more slowly than during the control period. Thus, both measures suggested a continuing decrease in plasma volume (P < 0.05; Table 3). Because moderate exercise was associated with a considerable variance between subjects, statistical analysis revealed no significant difference between the four measures. However, a comparison of values after the first 5 min of exercise showed plasma volume calculated from PP and RISA to be significantly lower than plasma vol,ume calculated from Hct and Hb (P < 0.001; Fig. 3B). Recovery. Hematocrit values had fallen to below preexercise levels after 50-min recovery from exercise (Table Z), equivalent to an overall 3 % increase in plasma volume. The slope of the linear regression was significant (P < 0.001). Hemoglobin concentration had decreased to near control levels after 50 min (Table 2; P < 0.01). In both the exercise and recovery phases, the change in plasma volume calculated from Hct was consistently less than that calculated from Hb (Fig. 3B). Compared with the control period, the mean corpuscular hemoglobin concentration (MCHC) was significantly elevated (P < 0.001) during and after moderate exercise. This was not so during and after light exercise. Plasma protein had returned to near control levels after
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA
VOLUME
DURING
929
EXERCISE
lime
(min)
FIG. 3. Comparison of regression lines relating change in Hct plasma protein (- - - -), and RISA (-0 -0 -) t-j, Hb (- --), to change in plasma volume: A, at 25% VOW maxi 42°C db; 35°C wb; coefficients (r) B, at 55% VO, maxi 30°C db; 24°C wb. Correlation given in Table 3.
50 min (Table 2); RISA, however, indicated a 6 % reduction in plasma volume at this time (Fig. 2B). The regression lines were significant at P < 0.001 and P < 0.01, respectively, and there was a significant difference (P < 0.001) between RISA and the other three measures. Osmolardy A small, but consistent increase in osmolarity was observed during light exercise (3 mosm l-1, or 1 % ; P < O.OOl), and during moderate exercise (5 mosml-1, or 1.5 %; P < 0.001). Values returned toward control levels during recovery (P < 0.05). DISCUSSION
The use of labeled albumin for monitoring the behavior of the largest fraction of the plasma protein population provided an alternative to the less sensitive electrophoretic technique used by Senay and Christensen (28, 29) in their studies of dehydration, and its effect on plasma constituents. However, the rate of loss of RISA from the intravascular space is an exponential function of time (6), and its representation by a linear equation is, therefore, theoretically incorrect. In earlier studies the mean rate of loss of RISA over a 2-h period in a thermoneutral environment was found not to depart significantly from linearity (15), suggesting that for practical purposes the early part of the exponential can be treated as a straight line. In the present study the rate of loss of RISA during the 50-min control period was described by a linear regression because no statistically valid exponential could be derived. Nevertheless, it is clear that as the straight line is extrapolated forward, divergence from the exponential will increase, intraducing error. This error could not be detected in the
earlier study over 2 h, and is unlikely to be significant in the present 2.5-h study. In most cases linear regressions provided the only statistically valid representation of plasma volume change. However, during recovery from moderate exercise plasma volume change could also be described by quadratic and exponential functions. In such cases the linear form was preferred since this facilitated between-measure comparison. During light exercise in the hot environment Hct, Hb, PP, and RISA each indicated a similar rate of decrease in plasma volume (Fig. 3A). Plasma volume has also been shown to decrease in resting subjects exposed to a high environmental temperature (Figs. 1A and 2A; (13)). During moderate exercise in the cooler environment the four measures indicated little change in plasma volume after the initial hemoconcentration (Fig. 3B). The decrease during light exercise may, therefore, be attributed to the higher environmental temperature. This might be explained by the more rapid secretion of sweat hypotonic to plasma (25) in the hot environment establishing a greater osmotic gradient between extravascular and intravascular compartments than in the cooler environment. The hematocrit is generally regarded as a reliable indicator of change in blood volume because all available evidence suggests that the circulating red cell volume remains constant during exercise and thermal stress (3, 5, 10, 12, 22, 23). It is therefore considered unlikely that the proportionately greater increase of Hct and Hb observed at the start of moderate exercise (Fig. 3B) was the result of addition of red cells to the circulation from storage depots (e.g., the spleen), rather than a loss of protein. The similar magnitude of the four measures at the start of light exercise supports this view (Fig. 3A). Van Beaumont (4) failed to demonstrate any changes in MCHC during short maxi,mal exercise despite a 6 % rise in plasma osmolarity, and Costill et al. (8) have reported that variations in mean corpuscular volume (MCV) are unrelated to changes in plasma osmolarity during exercise. In the present study an increase in MCHC of 1 % after 50 min of moderate exercise was associated with a 1.5 % increase in plasma osmolarity; during recovery MCHC remained . significantly elevated, although osmolarity decreased to preexercise levels. A 1 % increase in osmolarity during light exercise was not associated with any change in MCHC. Dill, Yousef, and Nelson (13) have reported a 2 % decrease in cell volume after a 2-h desert walk; plasma osmolarity was not measured. However, it does appear that red cells may not be in osmotic equilibrium with their surroundings during exercise, and that the severity and duration of exercise are factors affecting water loss from the red cells. If water is lost from the red cells during exercise, the change in plasma volume calculated from Hct will be underestimated (9). Therefore, of the two chosen measures of volume change based on red cell volume, Hb may be considered to be the more reliable, and this will be used to consider the effect of exercise on protein movement between intravascular extravascular compartments. and Hemoglobin values indicated that plasma volume had returned to near preexercise levels within 50 min of ceasing exercise, a finding which supports the observations of Cul-
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
930
HARRISON,
lumbine and Koch (11). However, after light exercise PP and RISA values, and after moderate exercise RISA values, remained significantly elevated. Unless both Hct and Hb grossly exaggerated the rate at which plasma volume returned to normal it is clear that the total quantity of protein within the intravascular space was greater after exercise than before. This observation is consistent with an addition of protein to the intravascular space via the lymphatic system as proposed by Senay (26,!27). The present experiments also provide support for the suggestion put forward by Senay (26) that the protein content of the intravascular space during exercise may depend on the relative proportions of the cardiac output perfusing skin and “leaky” muscle capillaries. After light exercise in the hot environment the augmentation of intravascular protein was greater (P < 0.05) than after moderate exercise in the cooler environment. This is consistent with an increased perfusion of cutaneous vascular beds in the heat at the expense of the more protein-permeable skeletal muscle capillaries, thereby effectively reducing the rate of protein loss from the intravascular space. Unfortunately, however, the evidence for augmentation of intravascular protein in the present study is based almost entirely upon changes occurring during the period of recovery from exercise. Only a comparison of the PP and RISA values at the start and end of moderate exercise provides any support for the concept of an exercise-induced augmentation of intravascular protein. After 5 min of exercise both measures gave values for the decrease in plasma volume significantly lower than those indicated by Hct and Hb. At this time, therefore, the rate of protein loss from plasma was greater than the rate of water loss, which suggests an increase in capillary permeability to all protein fractions. Subsequently, both PP and RISA continued to increase during the exercise, although Hct and Hb indicated no change in plasma volumean observation consistent with a rate of protein return to the intravascular compartment exceeding the rate of loss. After 50 min of exercise the four measures each gave similar values for the decrease in plasma volume. Alternative explanations for the elevated protein levels during the recovery period are a direct return of protein in
EDWARDS,
AND
LEITCH
capillary filtrate across the capillary walls, or an increase in lymph flow. The former seems unlikely as it would involve movement against a concentration gradient (19). Although lymph flow is more likely to be reduced during recovery, since it is stimulated by the action of the muscle pump during exercise (18), the observed relationship between the labeled albumin and total plasma protein during recovery from both levels of exercise is consistent with a return of an albumin-rich fluid, such as lymph (19), to the intravascular compartment. An increase in intravascular protein concentration will raise plasma oncotic pressure (a), thereby helping to maintain blood volume by reducing water loss, and favoring water gain. In the present study 7r had increased by 5.3 and 4.2 Torr (or 20.4 and 17.8 %), after 50 min of light and moderate exercise, respectively. After 50 min recovery 7r was still respectively 6.9 and 3.7 % above preexercise values. The constancy of the plasma volume during the last 45 min of moderate exercise implies a balance between plasma oncotic and hydrostatic pressures. Clearly, however, no such balance was achieved in the hot environment, despite the lower hydrostatic pressures expected at the lower work load. To summarize, of the four measures of change in plasma volume considered, Hb was the most reliable. The hematocrit introduces error if water is lost from the red cells, as occurred during moderate exercise in the present study. Plasma protein, either as total protein, or as albumin, is unreliable because protein translocation may be increased by exercise, although clear evidence for this was only obtained in the postexercise period. Augmentation of protein within the intravascular compartment, by increasing oncotic pressure, favors water retention both during exercise and recovery. The authors are pleased to acknowledge the technical assistance of Mrs. S. E. Kemp; also Miss H. M. Ferres, Mrs. I. M. Cooke, and the Statistics Section, RAF Institute for Aviation Medicine, for the statistical analysis. I also thank the Director General of Medical Services, Royal Air Force, for permission to submit this paper for publication. Received
for publication
13 March
1975.
REFERENCES 1.
ALBERT, S. N., V. GRAVEL,
Y. TURMEL, AND C. A. ALBERT. volume measurement. Anesthesia Analgesia Current Res. 44 : 805-8 14, 1965. ASTRAND, P.-O., AND I. RHYMING. Physical fitness nomogram. J. Appl Physiol. 7 : 2 18-22 1, 1954. ASTRAND, P.-O., AND B. SALTIN. Plasma and red cell volume after prolonged severe exercise. J. A@. Physiol. 19: 829-832, 1964. BEAUMONT, W. VAN. Red cell volume with changes in plasma osmolarity during maximal exercise. J. A&d. Physiol. 35 : 37-50, 1973. BEAUMONT, W. VAN, J. E. GREENLEAF, AND L. JUHOS. Disproportional changes in hematocrit, plasma volume, and proteins during exercise and bed rest. J. Appl. Physiol. 33 : 55-6 1, 1972. CHIEN, S., AND M. I. GREGERSEN.Determination of body fluid volumes. In : Physical Techniques in Biological Research, edited by W. L. Nastuk. New York and London : Academic, 1962, vol. 4, chapt. 1. COOPER, K. E., W. I. CRANSTON, AND E. S. SNELL. Temperature in the external auditory meatus as an index of central temperature changes. J. Appl. Physiol. 19: 1032-1035, 1964.
Pitfalls
2. 3.
4.
5.
6.
7.
8.
in blood
COSTILL, D. L., L. BRANAM, D. EDDY, AND W.
FINK. Alterations cell volume following exercise and dehydration. J. A/$. Physiol. 37: 912-916, 1974. COSTILL, D. L., AND W. J. FINK. Plasma volume changes following exercise and thermal dehydration. J. A#. Physiol. 37: 521-525, 1974. COSTILL, D. L., AND B. SALTIN. Changes in the ratio of venous-tobody hematocrit following dehydration. J. A@. Physiol. 36: 608-610, 1974. CULLUMBINE, H., AND A. C. E. KOCH. The changes in plasma and tissue fluid volume following exercise. Quart. J. E$tl. Physiol. 35: 39-46, 1949. DILL, D. B., J. H. TALBOTT, AND H. T. EDWARDS. Studies in muscular activity. J. Physiol., London 69 : 267-305, 1930. DILL, D. B., M. K. YOUSEF,AND J. D. NELSON. Responses of men and women to two-hour walks in desert heat. J. A/@. Physiol. 35 : 23 l-235, 1973. HARRISON, M. H. Metabolic Effects of Centrifugation and Heat Stress (PhD thesis). London: University of London, 1972, p. 109. HARRISON, M. H. Plasma volume changes during acute exposure
in red
9.
10.
11.
12. 13.
14. 15.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA to high 1974.
VOLUME
DURING
environmental
931
EXERCISE
temperature.
J. A@Z.
PhysioZ.
37: 38-42,
16. JOHNSON, T. F., AND H. Y. C. WONG. Effect of training and competitive swimming on serum proteins. J. ApPZ. Physiol. 16 : 807809, 1961. 17. JOYE, H., AND J. POORTMANS. Hematocrit and serum proteins during arm exercise. 1Med. Sci. Sports 2 : 187-190, 1970. 18. KJELLMER, I. Studies on exercise hyperaemia. Acta Physiol. Stand. SufipZ. 244: 27, 1965. 19. LANDIS, E. M., AND J. R. PAPPENHEIMER. Exchange of substances through the capillary walls. In: Handbook of Physiology. Circulation. Washington, D.C. : Am. Physiol. Sot., 1963, sect. 2, vol. II, chapt. 29, p. 961-1034. 20. LANNE, R. DE, J. R. BARNES, AND L. BROUHA. Changes in concentration of plasma protein fractions during muscular work and recovery. J. AppZ. Physiol. 13 : 97-104, 1958. P. Some effects of radiant heating of the head on body 21. MARCUS, temperature measurements at the ear. Aerospace Med. 44: 403-406, 1973. L. G., AND S. ROBINSON. Immediate and delayed effects 22. MYHRE, of acute dehydration on plasma volume in man (Abstract). Physiologist 12 : 3 10, 1969. 23. NYLIN, G. The effect of heavy muscular work on the volume of
circulating 1947.
red
corpuscles
in man.
Am. J. Physiol.
149 : 180-184,
J. R. Serum protein determination during short 24. POORTMANS, exhaustive physical activity. J. A#. Physiol. 30: 190-192, 1971. S. AND A. H. ROBINSON. Chemical composition of 25. ROBINSON, sweat. Physiol. Rev. 34: 202-220, 1954. L. C., JR. Movement of water, protein, and crystal26. SENAY, loids between vascular and extra-vascular compartments in heatexposed men during dehydration, and following limited relief of dehydration. J. Physiol., London 2 10 : 617-635, 1970. and protein content 27. SENAY, L. C., JR. Changes in plasma volume during exposures of working men to various temperatures before and after acclimatization to heat; separation of the roles of the cutaneous and skeletal muscle circulation. J. Physiol., London 224: 61-81, 1972. 28. SENAY, L. C., JR., AND M. L. CHRISTENSEN. Changes in blood plasma during progressive dehydration. J. AppZ. Physiol. 20: 1136-l 140, 1965. 29. SENAY, L. C., JR., AND M. L. CHRISTENSEN. Variations of certain blood constituents during acute heat exposure. J. APPZ. PhysioZ. 24: 302-309, 1968. 30. WEICHSELBAUM, T. E. An accurate and rapid method for the determination of proteins in small amounts of blood serum and plasma. Am. J. CZin. Pathol. 10, Tech. Sect. : 4049, 1946.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
APPLIED
PHYSIOLOGY
Vol. 39, No. 6, December
1975.
Printed
in U.S.A.
Effect of exercise and thermal stress on plasma
volume
M. H. HARRISON, R. J. EDWARDS, AND D. R. LEITCH RAF Institute of Aviation Medicine, Farnborough, Hampshire, England
HARRISON, M. H., R. J. EDWARDS, AND D. R. LEITCH. E$ect of exercise and thermal stress on plasma volume. J. Appl. Physiol. 39(6) : 925-931. 1975.-S’ ix male subjects exercised for 50 min at 25y0 (light exercise) and 55y0 (moderate exercise) of their estimated aerobic capacities in environments of 42°C db, 35°C wb and 30°C db, 24°C wb, respectively. Alterations in the hematocrit, hemoglobin, and plasma protein concentrations, and in the activity of an injected aliquot of isotopically labeled albumin were each used to calculate the percentage change in plasma volume occurring during exercise and recovery. Changes in each measure were consistent with a reduction in plasma volume during exercise and a return to preexercise levels during recovery. There was no significant difference between the measures when exercising in the heat, but during the more severe exercise in the cooler environment disproportional changes in protein, hematocrit, and hemoglobin were observed. Disproportional changes were also seen during the recovery phase, when the hematocrit and hemoglobin concentration indicated a more rapid return of the plasma volume to preexercise levels than did either the plasma protein concentration or albumin activity. During moderate exercise and recovery there was a 1 y0 decrease in red cell volume. It is concluded that exercise accelerates the rate of protein movement from extravascular compartments to the intravascular compartment, leading to elevated plasma protein levels during recovery which favor the return of water to the intravascular space. Hemoglobin concentration is considered to be the most reliable measure of plasma volume change during exercise.
in plasma volume being dominated higher environmental temperature, more severe work load.
in the one by the in the other by the
METHODS
Experimental
Design
Six male subjects exercised for 50 min in a hot environment of 42°C dry bulb (Tdb), 35°C wet bulb (Twb; relative humidity (rh) 63 %), and again 3 mo later in a warm environment where Tdb = 30°C and T,b = 24°C (rh = 62 0. Air movement was 1.5 m . s- l. The exercise was preceded, and followed, by 50-min control and recovery periods, respectively, where Tdb = 30°C and Twb = 24°C. Exercise was performed on a Monark bicycle ergometer at a pedal speed of 50 rev l rein- l. The period of 3 mo between the two groups of experiments was to allow all radioactivity from the injected isotope (see below) to be cleared from the blood. It was considered that any effect of time on subject responses to heat and exercise would be minimal, and did not warrant a randomization of the experimental design. The subjects were all healthy volunteers from within the Institute, and none was acclimatized to heat. Details are given in Table 1. All experiments were carried out during the morning in the winter months. A light breakfast was hematocrit; hemoglobin; plasma protein; albumin taken 2 h before an experiment, after which no food or liquid was allowed until the end of the recovery period. Only shorts and shoes were worn. THE ALTERATIONS in the hematocrit and the plasma protein Subjects reported to the control climatic chamber 45 concentration which occur in resting subjects exposed to a min before the start of the 50-min control period, and high environmental temperature appear to be quantityrested quietly. They remained seated in a wheelchair tively representative of plasma volume change over a 2-h throughout the experiment except for weighings (which period (15). This may not be the case during exercise if, were carried out at the end of each 50-min control, exeras reported by several investigators, protein is transferred cise, and recovery period), and for the exercise period. from the extravascular space to the intravascular space Subjects were weighed nude after towelling down, and (16, 17, 20, 24, 26, 27). correction was made for blood taken. The present study examines the relationship between After the subject had been instrumented and weighed, measures of change in plasma volume based on red cell an intravenous catheter was introduced 7 cm into the antevolume (hematocrit and hemoglobin concentration), and cubital vein of the left arm under local anesthesia. The intravascular protein (total plasma protein concentration catheter was filled with heparinized saline (50 IU *ml-l) and radioiodinated human serum albumin (RISA) acafter taking each blood sample. tivity), under two conditions of exercise and climate, and At the end of the control period the subject was wheeled during recovery from that exercise. These conditions were into an adjacent climatic chamber, where he mounted a light exercise in a hot environment, and more severe bicycle, and exercised at a predetermined rate for 50 min. exercise in a cooler environment. Both were intended to This work rate represented 25 % of the subject’s estimated represent realistic work/environment situations, any change maximum aerobic capacity in the hot environment (light 925
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
926
HARRISON,
TABLE
AND
LEITCH
1. Subject data
25%i702 max;42°C Subj
Age, yr
w
kg
Work
load, W
~1 2 3 4 5 6(i)* 6 (ii) Mean
EDWARDS,
sf~ SE
27 30 31 36 39 41 35 33 &
AT,, = change of the experiment,
56.0 77.8 68.5 81.9 71.6 72.0 79.2 2
72.4
j= 3.3
in aural temperature; and was replaced
65.4 58.9 54.0 54.0 54.0 54.0
57.0
k
+2.06 +1.44 +1.53 +1.31 +l.lS +1.28
1.9
db, 35OC wb
ATac after 50 min, “C
+1.46
ABW = change by subject 6(ii).
-
-1.02 weight.
exercise), and 55 % of his estimated maximum aerobic capacity in the warm environment (moderate exercise; Table 1). Aerobic capacity was estimated using the technique described by Astrand and Rhyming (2). After completing the exercise the subject was returned to the control chamber for the 50-min recovery period. Measurements Blood samples were taken at 5-min intervals throughout the control, exercise, and recovery periods for the measurement of hematocrit (Hct), hemoglobin (Hb), total plasma protein (PP), plasma osmolarity, and plasma RISA activity. The average sample volume was 7 ml, and the total volume of blood taken did not exceed 250 ml. Microhematocrits were obtained from whole blood. Determinations were in quadruplicate, with no correction for trapped plasma (CF = 12,000 X s), or for differences between venous and whole body hematocrit (1, 10, 15). Plasma protein was measured by the method of Weichselbaum (30), and Hb by the cyanmethemoglobin method. Plasma osmolarity was determined by freezing point depression using a Fiske osmometer. All determinations were from which the precision of the methods in duplicate, could be calculated. This error, however, was generally exceeded by within-subject variability during the control periods, which was assessed from 120 observations as: hematocrit, =t 1.1 %; plasma protein, h2.7 %; hemoglobin, *1.5%; osmolarity, *1.2%. A measure of the rate of loss of plasma albumin from the intravascular space was obtained from a single injection of RISA (20 mg of labeled albumin of specific activity 0.20vein 0.25 PCi l mg-1, in 1 ml saline) into the antecubital of the right arm 15 min before the start of the control period. The plasma RISA activity during the three experimental periods was determined by counting three 0.5-ml plasma aliquots of each 7 ml blood sample using a gamma spectrometer. The counting error was Z& 1 .O %, and the mean variation between the three aliquots was & 1.5 %. The rate of loss of plasma albumin from the intravascular space was used to calculate the percentage change in plasma volume during exercise and recovery as described below. Heart rate (fn) and deep body temperature (T,,) were measured at 5-min intervals during exercise-the former
zk 0.20 * Subject
30°C db, 24°C wb %A BW after min
147.2 132.4 110.4 122.6 110.4
+1.33 +0.34 +0.49 +1 .OO +1.29
-1.08 -0.70 -0.75 -0.64 -0.54
147.2
+1.81
-0.62
load, W
128.4 6(i)
vO2maxi
ATac after 50 min, “C
Work
-1.93 -0.95 -0.81 -0.57 -0.73 -1.10
=t 0.13 in body
55%
%A BW after 50 min
rf= 6.8
was unable
+1.04
rt
to participate
0.23
-0.72
50
=t 0.08
in the second
phase
from an electrocardiogram (ECG), and the latter to within 0.05”C using a calibrated ear thermistor located in the auditory canal close to the tympanic membrane. The thermistor probe was thermally insulated from the external environment by packing the auditory canal with cotton wool, and positioning a thick wad of cotton wool over the surface of the ear, as described by Marcus (21). The body temperature so measured accurately reflects sublingual temperature, and is probably similar to that of the tympanic membrane (7). Sweat loss was estimated from change in body weight (*lo g>* Oxygen consumption during the 50-min exercise periods was measured from three paired Douglas bag collections of 2 min taken between 5 and 10 min, 30 and 35 min, and 40 and 45 min. Data Analysis Using the absolute values for the variables measured, the three experimental periods were compared by analysis of variance. Since the subjects did not differ significantly, the subject data were combined, and Table 2 gives mean observed values for each measure. The values were transformed into equivalent plasma volume change using the equations of van Beaumont, Greenleaf, and Juhos (5), and represented by linear regressions as shown in Table 3, and Figs. 1 and 2; subject variability is indicated by 95 % confidence limits. Values for the significance levels of the regressions refer to slopes different from zero. The rate of loss of RISA from the intravascular space during the control period was described, for each subject, by linear regression equations, and extrapolated back to give the RISA activity at the time of injection. This was compared with the activity of the standard to derive the resting plasma volume. The equations were also extrapolated forward over the exercise and recovery periods to estimate the plasma RISA activity which might have been expected had no exercise supervened. This required the assumption that the rate of loss of RISA would have remained essentially linear. The theoretical activities so calculated were subtracted from the actual activities observed during exercise and recovery, and these differences expressed as a percentage of the isotope activity at the time of injection.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA
VOLUME
DURING
927
EXERCISE
TABLE 2. Change in Hct, Hb, plasma protein, and RISA during light and moderate exercise, and during recovery from that exercise (mean of 6 subj; for environmental conditions see text) -_ Hct Condition
Time,
min
25oJob
Control Exercise
Recovery
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
(5) (10) (15) (20) (25) (30) (35) w (45) (50)
Hb, g. 100 ml-1 ssg
max
25%902msx
902 max
Plasma
550/o
902
Protein,
25% 902max
max
g. 100 ml-1 55%
RISA,
max
902
25%
counts - min-1. g-1 max
vO2
55% ---
.iTO2 max
42.5
40.8
14.66
13.73
7.06
6.74
1,055*
829”
44.1 44.3 44.4 44.4 45.1 45.3 45.5 45.6 45.8 46.0
43.6 43.6 43.8 43.8 44.0 44.0 43.7 43.6 43.7 43.6
15.20 15.29 15.32 15.35 15.49 15.59 15.76 15.75 15.96 16.04
14.78 14.75 14.91 14.90 15.02 15.00 14.85 14.87 14.80 14.93
7.52 7.59 7.61 7.70 7.74 7.87 7.89 7.86 7.91 7.94
7.29 7.35 7.44 7.49 7.48 7.58 7.59 7.51 7.52 7.49
1,126 1,123 1,140 1,141 1) 141 1,141 1,144 1,141 1,148 1,154
898 906 911 911 901 890 890 886 879
45.5 45.5 44.6 44.0 43.9 43.7 43.5 43.1 43.0 42.9
43.4 42.2 41.7 41.2 41 .o 41 .o 40.5 40.6 40.6 40.5
15.77 15.74 15.36 15.08 15.10 14.97 14.97 14.81 14.99 14.74
14.86 14.58 14.12 14.11 13.94 13.89 13.84 13.89 13.90 13.86
7.87 7.82 7.63 7.55 7.55 7.50 7.48 7.45 7.39 7.37
7.53 7.32 7.20 7.10 6.98 6.96 6.86 6.93 6.96 6.90
1,141 1,132 1,104 1,077 1,063 1,063 1,037 1,046 1,048 1,030
868 844 829 817 808 803 792 790 790 780
Control: mean of n = 60 observations before the start of exercise, and 65 min
over 50-min after injection
control period except: of the isotope.
* where
n = 6, and
is the plasma
RISA
activity
5 min
TABLE 3. Regression equations, derived from the data of Table 2, relating Hct, Hb, plasma protein, and RISA to change in plasma volume (calculated as mean of 6 subj) Percent Condition
25%VO2 42 ‘c&-,,
max
35 ‘Cwb
55% VOZ max 30 ‘c&, 24 OCwb X
= 5-50 min
Y
Hct
Decrease
in Plasma
Hb
r
Volume
Calculated
Plasma
protein
from r
RISA
Y
Exercise Recovery
5.63 12.13
+ -
0.1584x 0.2316~
0.99 -0.97
4.58 11.84
+ 0.1963x - 0.2296x
0.99 -0.91
5.61 9.94
+ -
0.1166x 0.1266x
0.97 -0.94
6.32 12.57
+ 0.1419x - 0.0956x
0.99 -0.79
Exercise Recovery
11.37 7.82
-
0.0014x 0.2199x
0.04 -0.90
12.49 10.24
-
0.27 -0.84
8.23 9.15
+ -
0.0553x 0.1682x
0.74 -0.88
8.18 8.05
+ -
0.66 -0.85
of exercise,
or recovery
from
exercise;
0.0166x 0.2178x
r = correlation
RESULTS
After 50 min of light exercise (25 % vo 2 max) mean fn was 158 beatsmin-1 (range 132-174), and mean T., was 38.58”C (range 38.28-39.10). After 50 min of moderate exercise (55 % vo2 max> mean fn was 170 beats emin(range 152-180), and mean T,, was 37.91”C (range 37.1038.79). The magnitude of the increase in T., (AT,,) is given for each subject in Table 1. Mean 02 consumption (voz) during light exercise was 1.06 lmin-l (range 0.87-1.42), and during moderate exercise was 2.04 1 l rein-1 (range 1.69-2.5 1). On the three occasions during exercise that 02 consumption was determined, mean values for 25 % and 55 % VOW max were 0.93, 1.11, 1.14, and 1.81, 2.10, and 2.22 lmin-l, respectively. Weight loss due to sweating averaged 0.69 kg during light exercise (range 0 047-l l 08), and 0.51 kg during moderate exercise (range 0.38-0.61); Table 1 relates these to body weight ( %ABW). There was no significant correlation between changes
0.0364x 0.0972x
coefficient.
in fn, T,,, vo2, or BW, of plasma volume. Plasma Volume During
and changes
in the four
measures
Light Exercise in the Hot Environment
Exercise. There was no statistically significant difference between change in plasma volume calculated from Hct, Hb, PP, or RISA. The regression equations describing each measure during exercise were significant (P < 0.001; Table 3). Concentration increases occurring in the first 5 min were equivalent to a reduction in plasma volume of 6 %; increases during the succeeding 45 min (Table 2) were attributable to a further 6-8 % decrease in plasma volume (Figs. 1 and 2). The mean PP level after 50 min indicated a change in plasma volume significantly less than did the Hct, Hb, and RISA levels (P < 0.01; Fig. 3A). Recovery. Hematocrit, Hb, PP, and RISA levels decreased during recovery (Table 2); the linear equations relating these to plasma volume change were significant at P < 0.001 except for RISA where P < 0.01 (Table 3). Both Hct and Hb indicated a return of plasma volume to near
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
HARRISON,
928
rA
+I
c z
T
+2 -
E
,
Ob,,
2
,
,
,
,
-
; 3 ;
-2-
E 3 c
-6 -
-4 -
‘*,,
-8-
.-C z
-10
-
:: L t: n
-12
-
3 E ZC >‘G
EDWARDS,
AND
LEITCH
-2-
-4oz E‘ we -620 0, E -8.-cs 0) z -10 3E 2 2 -12 olc z
-14 -
-14-
5 -2s -4 kna iia -6,oLa -8.5 E XtO -10 fii z 6
-12 -
*0
-14 -16, t
I
I
1
10
20
30
1
40
1I
50 /
1
1
60 70 (10) (20)
k-----Exercise-p+-
,
80 (30)
I
,
90 100 (40) $0)
Recovery---d lime
(min)
FIG. 2. Change in plasma volume during exercise, and during recovery from exercise calculated from A, the plasma protein conof isotopically labeled albumin. centration, and B, the plasma activity Details as Fig. 1 (note data in A from (14)).
1. Change in plasma volume during exercise, and during recovery from exercise calculated from A, the hematocrit (Hct), and B, the hemoglobin (Hb) concentration: 1) at 25y0 Voz max; 42°C db; 24”Cwb 35”Cwb (X-X); 2) at 55% J?o, max; 30°C db; Vertical bars represent 95% confidence limits about the (0 -0 -0). regression lines (Table 3) for mean of 6 subjects. Also, change in plasma volume calculated from Hct (- - -) for mean of 6 subjects resting at 50°C db, 38 ‘C wb (from (14)). FIG.
the preexercise level after 50 min (Fig. 1, A and B). Plasma protein, however, remained elevated at this time (Table Z), suggesting that plasma volume remained 4 % below the control level (Fig. 24). Labeled albumin activity after 50-min recovery was equivalent to a 9 % reduction in plasma volume compared with control (Fig. ZB). The Hct and Hb regression coefficients were significantly different from the PP and RISA coefficients (P < 0.01; Fig. 34 . Plasma Volume During Moderate Exercise in the Thermoneutral Environment Exercise. The increase in Hct and Hb observed after 5 min of exercise (Table 2) was equivalent to a 12 % decrease in plasma volume. Thereafter no further change occurred (Fig. 1, A and B) ; the regressions were not statistically significant (Table 3), the correlation coefficients reflecting random error about the regression line (i.e., within-subject variability and experimental error).
The initial decrease in plasma volume was only 8 % when calculated from increases in PP concentration and RISA activity. During the last 45 min of exercise PP levels continued to increase, and RISA activity, although decreasing, did so more slowly than during the control period. Thus, both measures suggested a continuing decrease in plasma volume (P < 0.05; Table 3). Because moderate exercise was associated with a considerable variance between subjects, statistical analysis revealed no significant difference between the four measures. However, a comparison of values after the first 5 min of exercise showed plasma volume calculated from PP and RISA to be significantly lower than plasma vol,ume calculated from Hct and Hb (P < 0.001; Fig. 3B). Recovery. Hematocrit values had fallen to below preexercise levels after 50-min recovery from exercise (Table Z), equivalent to an overall 3 % increase in plasma volume. The slope of the linear regression was significant (P < 0.001). Hemoglobin concentration had decreased to near control levels after 50 min (Table 2; P < 0.01). In both the exercise and recovery phases, the change in plasma volume calculated from Hct was consistently less than that calculated from Hb (Fig. 3B). Compared with the control period, the mean corpuscular hemoglobin concentration (MCHC) was significantly elevated (P < 0.001) during and after moderate exercise. This was not so during and after light exercise. Plasma protein had returned to near control levels after
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA
VOLUME
DURING
929
EXERCISE
lime
(min)
FIG. 3. Comparison of regression lines relating change in Hct plasma protein (- - - -), and RISA (-0 -0 -) t-j, Hb (- --), to change in plasma volume: A, at 25% VOW maxi 42°C db; 35°C wb; coefficients (r) B, at 55% VO, maxi 30°C db; 24°C wb. Correlation given in Table 3.
50 min (Table 2); RISA, however, indicated a 6 % reduction in plasma volume at this time (Fig. 2B). The regression lines were significant at P < 0.001 and P < 0.01, respectively, and there was a significant difference (P < 0.001) between RISA and the other three measures. Osmolardy A small, but consistent increase in osmolarity was observed during light exercise (3 mosm l-1, or 1 % ; P < O.OOl), and during moderate exercise (5 mosml-1, or 1.5 %; P < 0.001). Values returned toward control levels during recovery (P < 0.05). DISCUSSION
The use of labeled albumin for monitoring the behavior of the largest fraction of the plasma protein population provided an alternative to the less sensitive electrophoretic technique used by Senay and Christensen (28, 29) in their studies of dehydration, and its effect on plasma constituents. However, the rate of loss of RISA from the intravascular space is an exponential function of time (6), and its representation by a linear equation is, therefore, theoretically incorrect. In earlier studies the mean rate of loss of RISA over a 2-h period in a thermoneutral environment was found not to depart significantly from linearity (15), suggesting that for practical purposes the early part of the exponential can be treated as a straight line. In the present study the rate of loss of RISA during the 50-min control period was described by a linear regression because no statistically valid exponential could be derived. Nevertheless, it is clear that as the straight line is extrapolated forward, divergence from the exponential will increase, intraducing error. This error could not be detected in the
earlier study over 2 h, and is unlikely to be significant in the present 2.5-h study. In most cases linear regressions provided the only statistically valid representation of plasma volume change. However, during recovery from moderate exercise plasma volume change could also be described by quadratic and exponential functions. In such cases the linear form was preferred since this facilitated between-measure comparison. During light exercise in the hot environment Hct, Hb, PP, and RISA each indicated a similar rate of decrease in plasma volume (Fig. 3A). Plasma volume has also been shown to decrease in resting subjects exposed to a high environmental temperature (Figs. 1A and 2A; (13)). During moderate exercise in the cooler environment the four measures indicated little change in plasma volume after the initial hemoconcentration (Fig. 3B). The decrease during light exercise may, therefore, be attributed to the higher environmental temperature. This might be explained by the more rapid secretion of sweat hypotonic to plasma (25) in the hot environment establishing a greater osmotic gradient between extravascular and intravascular compartments than in the cooler environment. The hematocrit is generally regarded as a reliable indicator of change in blood volume because all available evidence suggests that the circulating red cell volume remains constant during exercise and thermal stress (3, 5, 10, 12, 22, 23). It is therefore considered unlikely that the proportionately greater increase of Hct and Hb observed at the start of moderate exercise (Fig. 3B) was the result of addition of red cells to the circulation from storage depots (e.g., the spleen), rather than a loss of protein. The similar magnitude of the four measures at the start of light exercise supports this view (Fig. 3A). Van Beaumont (4) failed to demonstrate any changes in MCHC during short maxi,mal exercise despite a 6 % rise in plasma osmolarity, and Costill et al. (8) have reported that variations in mean corpuscular volume (MCV) are unrelated to changes in plasma osmolarity during exercise. In the present study an increase in MCHC of 1 % after 50 min of moderate exercise was associated with a 1.5 % increase in plasma osmolarity; during recovery MCHC remained . significantly elevated, although osmolarity decreased to preexercise levels. A 1 % increase in osmolarity during light exercise was not associated with any change in MCHC. Dill, Yousef, and Nelson (13) have reported a 2 % decrease in cell volume after a 2-h desert walk; plasma osmolarity was not measured. However, it does appear that red cells may not be in osmotic equilibrium with their surroundings during exercise, and that the severity and duration of exercise are factors affecting water loss from the red cells. If water is lost from the red cells during exercise, the change in plasma volume calculated from Hct will be underestimated (9). Therefore, of the two chosen measures of volume change based on red cell volume, Hb may be considered to be the more reliable, and this will be used to consider the effect of exercise on protein movement between intravascular extravascular compartments. and Hemoglobin values indicated that plasma volume had returned to near preexercise levels within 50 min of ceasing exercise, a finding which supports the observations of Cul-
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
930
HARRISON,
lumbine and Koch (11). However, after light exercise PP and RISA values, and after moderate exercise RISA values, remained significantly elevated. Unless both Hct and Hb grossly exaggerated the rate at which plasma volume returned to normal it is clear that the total quantity of protein within the intravascular space was greater after exercise than before. This observation is consistent with an addition of protein to the intravascular space via the lymphatic system as proposed by Senay (26,!27). The present experiments also provide support for the suggestion put forward by Senay (26) that the protein content of the intravascular space during exercise may depend on the relative proportions of the cardiac output perfusing skin and “leaky” muscle capillaries. After light exercise in the hot environment the augmentation of intravascular protein was greater (P < 0.05) than after moderate exercise in the cooler environment. This is consistent with an increased perfusion of cutaneous vascular beds in the heat at the expense of the more protein-permeable skeletal muscle capillaries, thereby effectively reducing the rate of protein loss from the intravascular space. Unfortunately, however, the evidence for augmentation of intravascular protein in the present study is based almost entirely upon changes occurring during the period of recovery from exercise. Only a comparison of the PP and RISA values at the start and end of moderate exercise provides any support for the concept of an exercise-induced augmentation of intravascular protein. After 5 min of exercise both measures gave values for the decrease in plasma volume significantly lower than those indicated by Hct and Hb. At this time, therefore, the rate of protein loss from plasma was greater than the rate of water loss, which suggests an increase in capillary permeability to all protein fractions. Subsequently, both PP and RISA continued to increase during the exercise, although Hct and Hb indicated no change in plasma volumean observation consistent with a rate of protein return to the intravascular compartment exceeding the rate of loss. After 50 min of exercise the four measures each gave similar values for the decrease in plasma volume. Alternative explanations for the elevated protein levels during the recovery period are a direct return of protein in
EDWARDS,
AND
LEITCH
capillary filtrate across the capillary walls, or an increase in lymph flow. The former seems unlikely as it would involve movement against a concentration gradient (19). Although lymph flow is more likely to be reduced during recovery, since it is stimulated by the action of the muscle pump during exercise (18), the observed relationship between the labeled albumin and total plasma protein during recovery from both levels of exercise is consistent with a return of an albumin-rich fluid, such as lymph (19), to the intravascular compartment. An increase in intravascular protein concentration will raise plasma oncotic pressure (a), thereby helping to maintain blood volume by reducing water loss, and favoring water gain. In the present study 7r had increased by 5.3 and 4.2 Torr (or 20.4 and 17.8 %), after 50 min of light and moderate exercise, respectively. After 50 min recovery 7r was still respectively 6.9 and 3.7 % above preexercise values. The constancy of the plasma volume during the last 45 min of moderate exercise implies a balance between plasma oncotic and hydrostatic pressures. Clearly, however, no such balance was achieved in the hot environment, despite the lower hydrostatic pressures expected at the lower work load. To summarize, of the four measures of change in plasma volume considered, Hb was the most reliable. The hematocrit introduces error if water is lost from the red cells, as occurred during moderate exercise in the present study. Plasma protein, either as total protein, or as albumin, is unreliable because protein translocation may be increased by exercise, although clear evidence for this was only obtained in the postexercise period. Augmentation of protein within the intravascular compartment, by increasing oncotic pressure, favors water retention both during exercise and recovery. The authors are pleased to acknowledge the technical assistance of Mrs. S. E. Kemp; also Miss H. M. Ferres, Mrs. I. M. Cooke, and the Statistics Section, RAF Institute for Aviation Medicine, for the statistical analysis. I also thank the Director General of Medical Services, Royal Air Force, for permission to submit this paper for publication. Received
for publication
13 March
1975.
REFERENCES 1.
ALBERT, S. N., V. GRAVEL,
Y. TURMEL, AND C. A. ALBERT. volume measurement. Anesthesia Analgesia Current Res. 44 : 805-8 14, 1965. ASTRAND, P.-O., AND I. RHYMING. Physical fitness nomogram. J. Appl Physiol. 7 : 2 18-22 1, 1954. ASTRAND, P.-O., AND B. SALTIN. Plasma and red cell volume after prolonged severe exercise. J. A@. Physiol. 19: 829-832, 1964. BEAUMONT, W. VAN. Red cell volume with changes in plasma osmolarity during maximal exercise. J. A&d. Physiol. 35 : 37-50, 1973. BEAUMONT, W. VAN, J. E. GREENLEAF, AND L. JUHOS. Disproportional changes in hematocrit, plasma volume, and proteins during exercise and bed rest. J. Appl. Physiol. 33 : 55-6 1, 1972. CHIEN, S., AND M. I. GREGERSEN.Determination of body fluid volumes. In : Physical Techniques in Biological Research, edited by W. L. Nastuk. New York and London : Academic, 1962, vol. 4, chapt. 1. COOPER, K. E., W. I. CRANSTON, AND E. S. SNELL. Temperature in the external auditory meatus as an index of central temperature changes. J. Appl. Physiol. 19: 1032-1035, 1964.
Pitfalls
2. 3.
4.
5.
6.
7.
8.
in blood
COSTILL, D. L., L. BRANAM, D. EDDY, AND W.
FINK. Alterations cell volume following exercise and dehydration. J. A/$. Physiol. 37: 912-916, 1974. COSTILL, D. L., AND W. J. FINK. Plasma volume changes following exercise and thermal dehydration. J. A#. Physiol. 37: 521-525, 1974. COSTILL, D. L., AND B. SALTIN. Changes in the ratio of venous-tobody hematocrit following dehydration. J. A@. Physiol. 36: 608-610, 1974. CULLUMBINE, H., AND A. C. E. KOCH. The changes in plasma and tissue fluid volume following exercise. Quart. J. E$tl. Physiol. 35: 39-46, 1949. DILL, D. B., J. H. TALBOTT, AND H. T. EDWARDS. Studies in muscular activity. J. Physiol., London 69 : 267-305, 1930. DILL, D. B., M. K. YOUSEF,AND J. D. NELSON. Responses of men and women to two-hour walks in desert heat. J. A/@. Physiol. 35 : 23 l-235, 1973. HARRISON, M. H. Metabolic Effects of Centrifugation and Heat Stress (PhD thesis). London: University of London, 1972, p. 109. HARRISON, M. H. Plasma volume changes during acute exposure
in red
9.
10.
11.
12. 13.
14. 15.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
PLASMA to high 1974.
VOLUME
DURING
environmental
931
EXERCISE
temperature.
J. A@Z.
PhysioZ.
37: 38-42,
16. JOHNSON, T. F., AND H. Y. C. WONG. Effect of training and competitive swimming on serum proteins. J. ApPZ. Physiol. 16 : 807809, 1961. 17. JOYE, H., AND J. POORTMANS. Hematocrit and serum proteins during arm exercise. 1Med. Sci. Sports 2 : 187-190, 1970. 18. KJELLMER, I. Studies on exercise hyperaemia. Acta Physiol. Stand. SufipZ. 244: 27, 1965. 19. LANDIS, E. M., AND J. R. PAPPENHEIMER. Exchange of substances through the capillary walls. In: Handbook of Physiology. Circulation. Washington, D.C. : Am. Physiol. Sot., 1963, sect. 2, vol. II, chapt. 29, p. 961-1034. 20. LANNE, R. DE, J. R. BARNES, AND L. BROUHA. Changes in concentration of plasma protein fractions during muscular work and recovery. J. AppZ. Physiol. 13 : 97-104, 1958. P. Some effects of radiant heating of the head on body 21. MARCUS, temperature measurements at the ear. Aerospace Med. 44: 403-406, 1973. L. G., AND S. ROBINSON. Immediate and delayed effects 22. MYHRE, of acute dehydration on plasma volume in man (Abstract). Physiologist 12 : 3 10, 1969. 23. NYLIN, G. The effect of heavy muscular work on the volume of
circulating 1947.
red
corpuscles
in man.
Am. J. Physiol.
149 : 180-184,
J. R. Serum protein determination during short 24. POORTMANS, exhaustive physical activity. J. A#. Physiol. 30: 190-192, 1971. S. AND A. H. ROBINSON. Chemical composition of 25. ROBINSON, sweat. Physiol. Rev. 34: 202-220, 1954. L. C., JR. Movement of water, protein, and crystal26. SENAY, loids between vascular and extra-vascular compartments in heatexposed men during dehydration, and following limited relief of dehydration. J. Physiol., London 2 10 : 617-635, 1970. and protein content 27. SENAY, L. C., JR. Changes in plasma volume during exposures of working men to various temperatures before and after acclimatization to heat; separation of the roles of the cutaneous and skeletal muscle circulation. J. Physiol., London 224: 61-81, 1972. 28. SENAY, L. C., JR., AND M. L. CHRISTENSEN. Changes in blood plasma during progressive dehydration. J. AppZ. Physiol. 20: 1136-l 140, 1965. 29. SENAY, L. C., JR., AND M. L. CHRISTENSEN. Variations of certain blood constituents during acute heat exposure. J. APPZ. PhysioZ. 24: 302-309, 1968. 30. WEICHSELBAUM, T. E. An accurate and rapid method for the determination of proteins in small amounts of blood serum and plasma. Am. J. CZin. Pathol. 10, Tech. Sect. : 4049, 1946.
Downloaded from www.physiology.org/journal/jappl at Macquarie Univ (137.111.162.020) on February 15, 2019.
