Graded cutaneous vascular responses to dynamic leg exercise

0 downloads 0 Views 2MB Size Report
Each subject performed two leg dynamic exercise on ... sounds superimposed on a pressure inflation/deflation cycle curve. Criteria for selection of appropriate sounds have been ..... slope as a fractional change, rather than as an absolute one.
Graded cutaneous vascular responses to dynamic leg exercise W. FRED

TAYLOR,

JOHN

M. JOHNSON,

Departments of Physiology and Pharmacology, San Antonio, Texas 78284 W. FRED, JOHN M. JOHNSON, WOJCIECH A. KoC. M. KWAN. Graded cutaneous vascular responses to dynamic leg exercise. J. Appl. Physiol. 64(5): 1803-1809, 1988.-The cutaneous vascular conductance-esophageal temperature (CVC-T,,) relationship was examined at five work loads (75-200 W) in each of four men to find whether there is a role for exercise intensity in the control of skin blood flow (SkBF). Several factors contributed to our evaluation of the CVC-T,, relationship during work. Laser-Doppler velocimetry (LDI?) provided a continuous measure of SkBF that is not influenced by underlying muscle blood flow. Local warming to 39°C at the site of measurement of SkBF provided a consistent skin temperature and facilitated observation of changes in LDF. Mean arterial pressure was measured noninvasively once per minute to calculate CVC. Supine exercise minimized baroreceptor-induced cutaneous vasoconstriction. Our major finding was that the internal temperature at which CVC began to rise during exercise (CVC threshold) was graded with work load beyond 125 W (P < 0.05). In that range the CVC threshold increased by 0.16’6 for every increment of 25 W. The CVC threshold was never reached at the highest work load in three of the four subjects. There was no consistent effect of work load on the slope of the CVC-T, relationship or on the internal temperature at which sweating began during exercise (sweat rate threshold). We conclude that the level of work beyond 125 W affects the CVC-T,, relationship in a graded fashion, principally through shifts in threshold. TAYLOR, SIBA, AND

skin blood flow; human; laser-Doppler velocimetry; vascular conductance; regional blood flow

cutaneous

OF SKIN BLOOD FLOW (SkBF)is a major element in the reflex regulation of body temperature. SkBF can also be affected by reflexes which are not directly involved in temperature regulation. For example, the baroreceptors are generally regarded as exerting reflex control over SkBF in both normothermic (1, 23) and hyperthermic (7, 15) conditions. There are also data consistent with the notion that reflexes associated with the onset of dynamic exercise cause cutaneous vasoconstriction (6,17,25,31). It is less clear to what extent this cutaneous vasoconstriction is sustained when exercise is prolonged beyond the first 23 min and internal temperature begins to rise. The heat production attendant to exercise evokes a reflex for thermoregulatory cutaneous vasodilation. This drive for vasodilation competes with the vasoconstrictor reflexes associated with exercise. However, available evidence is inconsistent as to whether the cutaneous vasoconstrictor CONTROL

0161-7567/88

$1.50

Copyright

WOJCIECH

The University

A. KOSIBA,

AND

C. M. KWAN

of Texas Health Science Center,

response to exercise is sustained. A major reason for questioning whether there is a sustained cutaneous vasoconstrictor response to exercise derives from conflict among studies that examined responses to graded exercise. Although there is clearly a graded vasoconstrictor effect of exercise on the splanchnit and renal circulations (11,24), data on the cutaneous circulation are less conclusive. Findings consistent with a graded response to the initiation of exercise have been reported for the human forearm (31) and finger (6, 12). However, Wenger et al. (29) and Johnson (14) failed to find a graded effect of exercise on the control of SkBF. The highest work loads used in those investigations were 70% of maximal O2 consumption and 150 W, respectively. If the lack of a graded cutaneous vascular response to exercise could be extrapolated to the resting condition, it would suggest little reflex role for exercise in the control of SkBF. However, Johnson and Park (16) found that, with respect to rest, exercise elevated the internal temperature threshold for the onset of cutaneous vasodilation at high skin temperature. We felt it worthwhile to reexamine this question with respect to whether there is a graded effect of exercise on this threshold. We used a combination of methods and procedures to optimize our examination of the cutaneous vascular responses to exercise. Earlier studies used plethysmography (which measures total forearm blood flow) to evaluate the SkBF response to exercise. In this study, we used laser-Doppler velocimetry to provide a continuous linear index of SkBF that is uninfluenced by the blood flow to underlying forearm muscle (3,20). We also raised local skin temperature at the site of SkBF measurement to 39°C to facilitate the observation of changes in SkBF with exercise (25). Rather than rely solely on SkBF, we measured arterial blood pressure to calculate cutaneous vascular conductance (CVC). Finally, exercise was performed in the supine position to minimize baroreceptormediated sympathetic activity to skin (16, 18). The general question which this investigation addressed was: Does exercise alter the vasomotor state of the cutaneous circulation in a graded manner? Specifically, is the threshold internal temperature at which cutaneous vasodilation occurs increased with exercise intensity? METHODS

AND

PROCEDURES

Four moderately active men consented to participate in this institutionally approved investigation after being

0 1988 the American

Physiological

Society

1803

1804

EXERCISE

AND SKIN

BLOOD

FLOW

thoroughly informed of all methods and procedures used. maintain T1, at 39OC. Previous work in this laboratory Each subject was given a physical examination and was showed that a T1, of -39OC facilitates the observation of in excellent health. Physical characteristics (mean t SD) changes in SkBF attendant to exercise (25). were height 183.4 t 4.4 cm, weight 77.4 t 6.8 kg, and Arterial blood pressure was measured noninvasively age 33 t 6 yr. Subjects were attired in shorts and shoes. once per minute with an electrosphygmomanometer Ambient temperature was 21.1 t 0.5°C, and relative (Narco Bio-Systems, PE-300). Systolic and diastolic arhumidity was 50%. terial pressures were determined by analysis of Korotkoff Each subject performed two leg dynamic exercise on sounds superimposed on a pressure inflation/deflation an electrically braked bicycle ergometer while in the cycle curve. Criteria for selection of appropriate sounds supine position with the legs slightly elevated above heart have been published previously (10, 22, 25). There is a level. Five levels of work were selected for each particihigh correlation between this method and direct intrapant, each performed on a separate day at the same hour. arterial measurements of blood pressure during exercise Heart rate (HR) was continuously measured from the (13). Mean arterial pressure (MAP) was calculated as electrocardiogram throughout each procedure. HR values diastolic blood pressure + ‘/3 of pulse pressure. Cutaneous before exercise and during the 6th min of work, at each vascular conductance (CVC) was calculated as 100 x of the five work loads, are presented in Table 1. Work (LDF/MAP), and when combined with data from other loads were selected for each subject to evoke similar studies, it was normalized for the individual value before responses in HR at each of five levels of exercise. In this exercise. That is, CVC values are expressed as percentway the relative load for each subject was similar, alages of the average value before the onset of exercise for though absolute work loads varied slightly among subeach experiment. jects. Each increment in work load produced an increase Sweat rate (SR) was continuously measured by dewin HR of -15 beats/minute over the previous level of point hygrometry (5). A SR detection capsule was placed work. The duration of exercise varied with the intensity over 6 cm2 of forearm skin, distal to the LDF probe. T1, of the work, since the goal was to achieve a moderate under the capsule was maintained at 34 t O.l"C by increase in internal temperature to induce cutaneous heating dry N2 before it circulated through the capsule vasodilation. This required a longer duration of exercise (21) at low work loads (low relative heat production) and a Hk, T,,, Tl,, LDF, and SR were each sampled once shorter exercise bout at the higher work intensities (high per second by analog-to-digital converters and a laborarelative heat production). The highest level of work could tory computer. Average values for each 20-s period were only be sustained for 6-8.5 min. Esophageal temperature (T,,) was monitored as an calculated. The standard protocol is shown in Fig. 1. Each experindex of internal temperature. This was accomplished by iment commenced with 4 min of control measurements placement of a copper-constantan thermocouple catheter while the subject was supine. At 4 min, T1, at the site of in the esophagus at the level of the atria. The location LDF measurement was increased to 39°C and was mainof the thermocouple was verified by analysis of the tained at that level for the remainder of the procedure. electrocardiogram recorded from the catheter tip (4). SkBF was measured by laser-Doppler velocimetry (3, Exercise was initiated after LDF had increased to a stable 20), which is uninfluenced by underlying muscle blood level (usually at 42 min). After exercise was completed, flow (Laserflo blood perfusion monitor, TSI). This an occlusion cuff was placed about the upper arm and method provides a continuous measurement of forearm inflated to 280 mmHg. This arterial occlusion allowed SkBF from an area of -1 mm2 and -1 mm depth. observation of the LDF signal at zero blood flow. Exercise Although the laser-Doppler flow signal (LDF) does not was also performed during this occlusion period to ensure provide absolute units of blood flow, it does provide a that movement attendant to the exercise did not cause reliable index of relative changes in SkBF (20). an artifactual increase in the LDF signal. The LDF Local skin temperature (T1,) at the site of LDF measoutput was always 0 V during this period of forearm urement on the forearm was monitored with a thermoarterial occlusion and exercise (see Fig. 1). couple situated between the skin surface and a heating To evaluate whether exercise per se affected the conelement. The heating element (4 cm diam) held the LDF trol of CVC, the relationship between T,, and CVC probe in place against the skin surface and was used to during exercise was examined. Two parameters with regard to this relationship are of particular interest: the TABLE 1. Average heart rate at rest and during T,, at which cutaneous vasodilation began (CVC threshthe sixth minute of exercise and difference in heart rate old, see Fig. 3) and the CVC-T,, slope after CVC threshfor each of the five levels of exercise old had been attained. The internal temperature at the Level HR Rest HR Exercise AHR onset of sweating (SR threshold) was also noted. The roles for the level of exercise in the threshold for cuta1 59*7 102k6 4321 2 57t5 112,t3 56k2 neous vasodilation, the threshold for sweating, and in 3 58&6 12823 72k3 the CVC-T,, slope were tested by analysis of variance for 4 58k7 146&l 88t6 a randomized complete block design. Additional statislOlk3 5 62k7 163t5 tical analysis included regression analysis of the relationValues are means t SE, in beats/min. Note that each increment in ship of CVC threshold, CVC-T,, slope, and SR threshold exercise intensity caused heart rate (HR) to increase by -15 beats/ to the level of work (30). min.

EXERCISE

AND

SKIN

BLOOD

1805

FLOW

RESULTS

Typical patterns of the responses to the standard protocol are shown in Fig. 1. T1, was raised to 39°C at 4 min and was maintained at that level throughout the procedure. Elevations in MAP, T,, and HR are evident during the exercise period. A characteristic decrease in LDF is apparent at the onset of exercise. As exercise continued and T, rose, the direction of change in LDF was reversed and LDF increased. The pattern of response of CVC was usually more pronounced than for LDF. Figure 2 shows CVC, LDF, and MAP responses to exercise in one subject. Note that the cutaneous vasoconstrictor response is more obvious in CVC than in LDF. The threshold for cutaneous vasodilation and the slope of the CVC-T, relationship were obtained from the CVC values for each study. The CVC threshold was taken as the T,, at which CVC began a sustained rise with internal temperature during exercise. The responses in %CVC (percentage of preexercise values) during exercise at all five work loads for one subject are shown in Fig. 3. This figure illustrates the method of determination of CVC threshold noted on each panel. CVC and T,, typically decreased at the beginning of exercise. As exercise continued, T,, rose while CVC showed a further decrease or no change until the threshold was reached (indicated by ~7 in Fig. 3). Thereafter, T, and CVC rose together. This subject (as well as 2 others) did not reach threshold at the highest work load (200 W). Instead, CVC continued to fall throughout the 100 HR, bpm

37--

-

-

MAP, mmHg

100 75 -J0

20

40

I 60

I 80

Minutes 1. Typical protocol followed in this study. Local temperature (T& at site of laser-Doppler flow (LDF) measurement was elevated to 39°C at 4 min. Increases in heart rate (HR), esophageal temperature (T-), and mean arterial pressure (MAP) can be seen during exercise (100 W). Note initial decrease in LDF with onset of exercise followed by an increase as T, increased. Drop in LDF to 0 at -70 min was caused by arterial occlusion of forearm. Exercise at same level was performed during last minute of occlusion. Note that there is essentially no motion artifact in LDF record. FIG.

175

+ Watts’

LDFJ

;J

CVC, WOO

mmHg

MAP, mmHg

I

IO

1

I

1

1

20

30 40 60 Minutes FIG. 2. Laser-Doppler flow (LDF), cutaneous vascular conductance (CVC), and mean arterial pressure (MAP) responses from 1 subject. Note that at onset of exercise (175 W) LDF showed a transient small increase, whereas CVC showed a marked reduction. This marked cutaneous vasoconstriction was made apparent only through evaluation of cvc.

exercise period. In those cases, CVC threshold was estimated (conservatively) as the T,, at the end of exercise. Figure 4 shows results from one of those studies. All CVC threshold values from the four subjects are shown in association with the respective work loads in Fig. 5, top. It is apparent from these data that at work loads above 125 W the CVC threshold increased with the intensity of exercise. The equation for the linear regression analysis above 125 W is CVC threshold = 6.50 x 10D3 (work load) + 36.03 (r = 0.81); i.e., the CVC threshold rose ~0.16OC for each 25-W increment in work load. The slope of the CVC threshold-work-load relationship is significantly different from zero (P < 0.05). These data show a graded effect of exercise; i.e., the internal temperature at the initiation of cutaneous vasodilation rises with increasing exercise intensity above work loads of 125 W. For each study, the slope of the CVC-T,, relationship was measured from the CVC threshold throughout the remainder of the exercise period, unless the relationship deviated obviously from linearity. Figure 3 shows those regressions for one subject. The average CVC slopes at exercise Zeueb 1-4 are listed in Table 2. Individual values are shown in Fig. 5, middle. At exercise level 5, when heat production was highest, CVC continued to decrease throughout exercise in three of four subjects, and these data are not included in Table 2 or Fig. 4. There was no significant relationship between CVC slope and work

1806

EXERCISE

200

1 y = 281.0 (T -

200 R-J

36.84)

+ 98.4b

AND

SKIN

BLOOD

FLOW

I = 0.85

y = 107.5 (1 - 36.62) + 91.46 r = 0.91 0.0 f

150

34

I

I

1 44

I

1 54

Minutes

I

100 watts

36.62O C

I 200

1 y = 280.7

1

(T - 36.07)

FIG. 4. Cutaneous vascular conductance (CVC) responses from 1 subject at his highest work load (200 W). Note that CVC continued to decrease throughout exercise period. CVC threshold for cutaneous vasodilation was not achieved.

+ 40.59

.I

7

r = 0.97

150 36.87O C

0 > 0

100 50

ol-

200 R ‘1; > c)

1 y = 293.3

150 Watts I

1

0Q qL$T 88pa

(T - 37.15) + 75.42 I = 0.99

150 100

O0

50 I

t

37&

0 200

I

175 Watts

c

I

I

1

1

g

150 1

>0 0

100

37.390 c

@

50

0

0000

00000

1 00 0 200 Watts

0 36.0

I

I

36.5

37.0 Ls

I

37.5

1

38.0

(“c>

FIG. 3. Percent cutaneous vascular conductance (CVC) (% of preexercise control value) vs. esophageal temperature (T,,) for 1 subject at all 5 work loads. In each panel pathway of response pattern at beginning of exercise is indicated by G . Threshold temperature for cutaneous vasodilation is indicated by 7 or 1. Regression lines are shown for T,,CVC relationship (at levels of T,, above threshold) for first 4 work loads. Regression analysis was not performed at highest work load, since cutaneous vasodilation was not achieved. Equations describing CVC-T,, relationship and correlation coefficients are shown for each of 4 lowest work loads.

load (P > 0.05). The average levels of T,, at which SR began during each of the exercise levels are also presented in Table 2. The individual data points for the SR threshold are shown in Fig. 5, bottom. It is apparent that there was no graded effect of work load on SR threshold (P > 0.05). DISCUSSION

With the onset of exercise, there is an increase in sympathetic nerve activity which results in regional vaso-

constriction and a redistribution of blood flow from splanchnic (24), renal (ll), cutaneous (17, 25, 31), and inactive muscle (2, 31) circulations to the metabolically vasodilated active muscle. The response of the cutaneous circulation to exercise is unique in that in addition to participating in exercise reflexes (16, 17, 25), skin also subserves thermoregulatory reflexes (23). As exercise proceeds and heat is generated, the cutaneous vasoconstriction is reversed by the thermoregulatory drive for vasodilation. Johnson and Park (16) noted that the principal effect of exercise (relative to rest) on the control of SkBF was to elevate the internal temperature at which cutaneous vasodilation began. In the present study we found that above 125 W, exercise intensity competes with the reflex thermoregulatory drive by elevating the internal temperature threshold for cutaneous vasodilation in a graded manner. This effect of exercise is seen at work loads higher than those used by either Johnson (14) or Wenger et al. (29). Zelis et al. (31) also reported a sustained cutaneous vasoconstriction at their most strenuous work load. Christensen et al. (6) noted that the vasoconstriction of the finger attending the onset of exercise was greatest at the highest work load. A recent study by Hirata et al. (12) reported graded responses in hand cutaneous vascular tone. Other investigations (14, 29), however, have not confirmed a graded role for exercise in the control of the cutaneous circulation. The studies by Wenger et al. (29) and Johnson (14) found no effect of work load on the SkBF-T,, relationship. We feel that in the present study the 1) use of supine exercise, 2) expression of CVC (rather than SkBF), 3) local warming of the skin, 4) examination of the skin circulation (rather than that of the total forearm) 5) emphasis on the effects of exercise on the CVC threshold, and 6) use of a wide range of work loads each contributed to our ability to discover a graded effect of exercise on the control of the cutaneous circulation. Johnson and Park (16) noted effects of posture and exercise on the T,, threshold. With regard to exercise they reported that both supine and upright exercise elevated the T,, threshold for cutaneous vasodilation. In the current study, we found this effect of exercise on the threshold for vasodilation to be graded with the level of exercise, at least above 125 W. Although an earlier study

EXERCISE

AND

37.5 -

=

2. Cutaneous vascular conductance threshold and slope and sweat rate threshold cvc

0

Level 0

37.0 3

m

1 2 3 4 5

A 0 A

=

/

B

A

. A

. . . . . . . . . . . . . 7 . . . . . . . . . 0 50 100 150 200 25 10

400 -

I

A

3 200 -

n

I

.

300 -

A

A A

A n

.

100 -

0

A

fi

0

n

. A 0

.‘..~T~..l...-17~-.1~~~=1 0 50 100

200

150

250

37,s 1 0 .

0 A n

37.0 -

A

n

q l

I

A

0 A

A A

0

0

n

d

.

36.5

1807

FLOW

A

.

36.5

BLOOD

TABLE A

.

.

SKIN

A

. . . . I r v . . 1 . . . . 1 . . . . 1 . . 1 . 1 50 100 150 200 250 WORKLOAD

(WAlTS)

FIG. 5. Top: cutaneous vascular conductance (CVC) threshold vs. work load for all subjects. Each subject is represented by a different v&Z. Heavy line is overall linear regression between 125 and 200 W. Over this range there was a significant (P < 0.05) positive relationship between CVC threshold and work load. Three data points overlie each other at 75 W and 36.84”C, 2 at 100 W and 36.62”C, and 2 at 175 W and 37.15”C. Middle: slopes of CVC-esophageal temperature relationship for each study from 4 subjects. Three negative slopes at highest work loads were excluded. Two values obscure each other at 175 W and 92%/‘C. Linear regression analysis showed no significant relationship between this slope and work intensity (P > 0.05). Bottom: sweat rate (SR) threshold vs. work load from 4 subjects at each level of exercise. There was no significant trend in SR threshold with respect to work load (P > 0.05). Symbols for each subject are consistent among punezs.

suggested that exercise reduced the slope of the SkBFT, relationship (18),a more recent study found the slope to not differ between rest and exercise (16). Similarly, we did not find the level of work to be a significant factor in the slope of the CVC-T,, relationship.

Threshold, “C

36.82kO.02 36.80k0.12 36.81kO.03 37.08kO.05 37.29kO.04

cvc Slope, % of control/“C

SR Threshold, “C

243k60

37.17kO.06

195t60

37.13*0.09

217k74 142*59 rt

37.06kO.09 37.1HO.13 37.03kO.16

Values are averages t SE for internal temperature at which cutaneous vasodilation occurred (CVC threshold), slope of the cutaneous vascular conductance-esophageal temperature (CVC-T,) relationship (CVC slope), and internal temperature at which sweating began (SR threshold) for all 5 work levels. * CVC slope at highest work load is not reported because 3 of 4 subjects did not increase skin blood flow. In these subjects, CVC threshold was estimated conservatively as level of T, at end of exercise.

The upright posture itself contributes to i.ncreased sympathetic vasoconstrictor activity to skin and increases the threshold for cutaneous vasodilation (16). To minimize this effect all exercise was performed in the supine posture. Controlling the level of T1, at the site of LDF measurement also facilitated our observation of changes in SkBF. An earlier investigation from this laboratory (25) found the changes in both SkBF and CVC with exercise to be most apparent at a T1, of -39*C. In this investigation we applied a T lot of 39°C to provide a consistent temperature at the site of measurement and to allow better visualization of changes in LDF and CVC. We chose to use LDF (3, 20) rather than venous occlusion plethysmography to evaluate responses in SkBF. This method has the advantage of being a continuous index of SkBF and is not confounded by changes in muscle blood flow. In preliminary studies, using venous occlusion plethysmography ‘, we often observed an anticipatory increase in forearm blood flow before exercise. This abrupt change in blood flow, presumably in forearm muscle (26), made it difficult to draw conclusions regarding the partition between skin and muscle in the change in blood flow with the onset of exercise. This ambiguity was resolved as the LDF signal is specific to skin. There were several assumptions regarding the use of LDF in reaching our conclusions which merit consideration. Our first assumption was that the pattern of change in SkBF during exercise was reflected by similar changes in LDF. It should be recalled that LDF represents blood flow to a rather small volume of tissue (-1 mm3). Our earlier comparison with plethysmographic estimates of changes in forearm SkBF indicated that the pattern of increase in LDF was quite similar to that for forearm SkBF (20). Thus we feel confident that the CVC threshold derived from thi .s method adequately reflects that for s#kin of the entire forearm. Our previous work also showed that the absolute value of LDF at a given level of forearm SkBF varied among studies, although there was a highly linear relationship between LDF and forearm SkBF for each comparison (20).It appears likely that the reason for this observation is that the number of cutaneous vessels contributing to the LDF signal

1808

EXERCISE

AND

varies from site to site, but the pattern of response is similar among them. Consequently, we normalized the LDF signal for each study to the value immediately before exercise as a means whereby a quantitative comparison of the rise in CVC with respect to internal temperature could be achieved. Thus we expressed that slope as a fractional change, rather than as an absolute one. A third assumption with respect to the use of LDF is that its ratio with MAP is a reasonable indicator of true CVC. To be absolutely correct, this use requires not only a linear relationship between SkBF and LDF but an intercept at the origin as well. The design of this series of experiments precluded deriving slopes and intercepts for this relationship for each study, so we do not know how far this intercept (LDF at 0 SkBF) may have deviated from the origin, despite the observation (with the instrument used in the current study) of zero output from the LDF derive during occlusion. The potential error from this unknown intercept depends on several factors, including the level of LDF. For example, if LDF = a X SkBF + b, then SkBF = (LDF - b)/a. For CVC, SkBF/MAP = (LDF - b)/(a x MAP). The range of b in the calculation of CVC must be between zero (our simplifying assumption) and the lowest level of LDF observed before local warming (a value cleariy above the true intercept). If one assumes, conservatively, that the lowest value during normothermic rest is equal to the intercept, our error in ignoring it in our calculation of CVC at threshold would average 3.6 t 2% over the 20 studies. The true error would thus be between 0 and 3.6%. At lower levels of LDF (e.g., before local warming) the potential error would be higher, however. CVC is a more sensitive indicator of the cutaneous vasomotor state than is SkBF. This is illustrated in Fig. 2 where, with the onset of work, there is an increase in MAP but little change in LDF. Using only LDF, one would miss the impressive cutaneous vasoconstriction apparent in the CVC record. At the highest level of work the CVC threshold was never achieved in three of four subjects (Fig. 4). In those instances the slope of the CVC-T,, relationship was negative throughout exercise, and CVC was lower at the end than at the beginning of exercise. These data at level five are similar to the results obtained by Zelis et al. (31). The exercise at level 5 approached a maximum effort as it resulted in exhaustion within 6-8.5 min. At that level of work the exercise-induced vasoconstriction is apparently so marked that it overpowered the thermoregulatory drive for reflex cutaneous vasodilation. The observation that the SR thresholds did not vary with work load does not deny that exercise could alter sweating at a given level of T,,. Observations from this laboratory (19) and others (9, 27) support the notion that exercise promotes sweating beyond its effects via internal temperature. In those studies (9, 19, 27), sweating was initiated by prior heat stress, and exercise served to modify the preexisting level of SR. We cannot say from these data whether exerciseinduced vasoconstrictor control of skin differs substantially from that of the splanchnic or renal circulations. Above a certain level of work, the influence of exercise

SKIN

BLOOD

FLOW

on each of these regional circulations increases with intensity (11, 23, 24). However, the level of work above which this influence becomes measureable appears to be higher for the cutaneous circulation than for the viscera (11, 23, 24,28). We do not know whether this difference arises from differential vasomotor control, the sensitivity of methods used or from competing thermoregulatory responses specific to skin. The CVC threshold is defined here as the level of internal temperature at which cutaneous vasodilation begins. This event could be altered by exercise through an effect on central thermoregulatory centers or could result from a net summation of vasomotor drives to the cutaneous arterioles. In the former case (i.e., a central effect of exercise), the integration of drive from exercise and internal temperature would occur centrally such that thermoregulatory vasodilation would be initiated at higher central temperatures. Alternatively, exercise could have its competitive effects via activation of cutaneous vasoconstrictor pathways. Internal temperature and the vasodilator activity required to overcome those vasoconstrictor effects would be higher during exercise and would increase with exercise intensity. The CVC threshold would reflect this net balance. The authors and to TSI for appreciation to This study Institute Grant Received

2 June

are grateful to Dr. D. S. O’Leary for his contributions the loan of the Laserflo monitor. We also express our Kim Morren for secretarial support. was supported by National Heart, Lung, and Blood HL-20663. 1986; accepted

in final

form

13 November

1987.

REFERENCES 1. BEISER, G. D., R. ZELIS, S. E. EPSTEIN, D. T. MASON, AND E. BRAUNWALD. The role of skin and muscle resistance vessels in reflexes mediated by the baroreceptor system. J. Clin. Inuest. 49: 225231,197O. B. S., AND J. T. SHEPHERD. Regulation of the circu2. BEVEGARD, lation during exercise in man. Physiol. Reu. 47: 178-213, 1967. 3. BONNER, R. F., T. R. CLEM, P. D. BOWEN, AND R. L. BOWMAN. Laser-Doppler continuous real-time monitor of pulsatile and mean blood flow in tissue microcirculation. In: Scattering Techniques Applied to SU~FUITI&CU~UF and Non-Equilibrium Systems, edited by S. H. Chen, B. Chu, and R. Nossal. New York: Plenum, 1981, p. 685-702. G. L., J. M. JOHNSON, AND P. A. HONG. Electro4. BRENGELMANN, cardiographic verification of esophageal temperature probe position. J. Appl. Physiol. 47: 638-642, 1979. 5. BRENGELMANN, G. L., M. MCKEAG, AND L. B. ROWELL. Use of dew-point detection for quantitative measurement of sweating rate.

J. Appl. Physiol. 39: 498-500,1975. 6. CHRISTENSEN, E. H., M. NIELSEN, AND B. HANNISDAHL. Investigations of the circulation in the skin at the beginning of muscular work. Actu Physiol. Stand. 4: 162-170,1942. 7. CROSSLEY, R. J., A. D. M. GREENFIELD, G. L. PLASSARAS, AND D. STEPHENS. The interrelation of thermoregulatory and baroreceptor reflexes in the control of blood vessels in the human forearm. J. Physiol. Lord. 183: 628-636,1966. 8. EDHOLM, 0. G., R. H. Fox, AND R. K. MACPHERSON. Vasomotor control of the cutaneous blood vessels in the human forearm. J. Physiol. Land. 139: 455-465,1957. 9. GISOLFI, C., AND S. ROBINSON. Central and peripheral stimuli regulating sweating during intermittent work in men. J. Appl. Physiol. 29: 761-768,197O. 10. GLASSER, S. P., AND M. R. RAMSEY. An automated system for blood pressure determination during exercise. Circulation 63: 348353,198l. 11 &&.GRIMBY, G. Renal clearances during prolonged supine exercise at

EXERCISE

AND SKIN

different loads. J. Appl. Physid. 20: 1294-1298,1965. 12. HIRATA, K., T. NAGASAKA, A. HIM, M. HIRASHITA, AND T. TAKAHATA. Cutaneous vascular tone during heat load modified by exercise intensity. J. Therm. Bid. 9: 117-120,1984. 13. HOSSACK, K. F., B. W. GROSS, J. B. RITTERMAN, F. KUSUMI, AND R. A. BRUCE. Evaluation of automated blood pressure measurements during exercise testing. Am. Heart J, 104: 1032-1038,1982. 14. JOHNSON, J. M. Responses of forearm blood flow to graded leg exercise in man. J. Appl. Physid. 46: 457-462,1979. 15. JOHNSON, J. M., M. NIEDERBERGER, L. B. ROWELL, M. M. EISMAN, AND G. L. BRENGELMANN. Competition between cutaneous vasodilator and vasoconstrictor reflexes in man. J. Appl. Physiol. 35: 798-803,1973. 16. JOHNSON, J. M., AND M. K. PARK. Effect of upright exercise on threshold for cutaneous vasodilation and sweating. J. Appl. Physiol. 50: 814-818,1981. 17. JOHNSON, J. M., AND M. K. PARK. Effect of heat stress on cutaneous vascular responses to the initiation of exercise. J. Appl. Physid 53: 744-749,1982. 18. JOHNSON, J. M., L. B. ROWELL, AND G. L. BRENGELMANN. Modification of the skin blood flow-body temperature relationship by upright exercise. J. Appl. Physiol. 37: 880-886,1974. 19. JOHNSON, J. M., W. F. TAYLOR, D. S. O’LEARY, AND M. K. PARK. Does exercise play a non-thermal role in the regulation of sweat rate? (Abstract) Federation Proc. 41: 1592,1982. 20. JOHNSON, J. M., W. F. TAYLOR, A. P. SHEPHERD, AND M. K. PARK. Laser-Doppler measurement of skin blood flow: comparison with plethysmography. J. Apple Physiol. 56: 798-803,1984. 21. KRANING, K. K., E. G. SCHIBLI, S. T. MOMII, AND E. A. MERRICK. A gas-stream heater to control local skin temperature under evap-

BLOOD

FLOW

1809

orative sweat capsules. J. Appl. Physiol. 44: 981-985,1978. 22. ROBERTS, L. M., J. R. SMILEY, AND G. W. MANNING. A comparison of direct and indirect blood-pressure determinations. Circulation 8: 232-242,1953. 23. ROWELL, L. B. Human cardiovascular adjustments to exercise and thermal stress. Physiol. Rev. 54: 75-159,1974. 24. ROWELL, L. B., J. R. BLACKMON, AND R. A. BRUCE. Indocyanine green clearance and estimated hepatic blood flow during mild to maximal exercise in upright man. J. Clin. Invest. 43: 1677-1690, 1964. 25. TAYLOR, W. F., J. M. JOHNSON, D. S. O’LEARY, AND M. K. PARK. Modification of the cutaneous vascular response to exercise by local skin temperature. J. Appl. Physid. 57: 1878-N&1984. 26. OVNAS, B. The vasodilator nerves. An. Heart J. 62: 277-285,196l. 27. VAN BEAUMONT, W., AND R. W. BULLARD. Sweating exercise stimulation during circulatory arrest. Science Wash. DC 152: 15211523,1966. 28. WADE, 0. L., B. COMBES, A. W. CHILDS, H. 0. WHEELER, A. COURNAND, AND S. E. BRADLEY. The effect of exercise on the splanchnic blood flow and splanchnic blood volume in normal man. Clin. Sci. Land. 15: 457-463,1956. 29. WENGER, C. B., M. F. ROBERTS, J. A. J. STOLWIJK, AND E. R. NADEL. Forearm blood flow during body temperature transients produced by leg exercise. J. Appl. Physiol. 38: 58-63,1975. 30. ZAR, J. H. Biostatisticul Analysis. Englewood Cliffs, NJ: PrenticeHall, 1974, p. 198-227. 31. ZELIS, R., D. T. MASON, AND E. BRAUNWALD. Partition of blood flow to the cutaneous and muscular beds of the forearm at rest and during leg exercise in normal subjects and in patients with heart failure. Circ. Res. 24: 799-806, 1969.