Physiological responses to cold stress during prolonged intermittent low

Physiological intermittent

responses to cold stress during low- and high-intensity walking

prolonged

A. S. WELLER,l C. E. MILLARD,2 M. A. STROUD,2 P. L. GREENHAFF,l AND I. A. MACDONALD1 and Pharmacology, University of Nottingham Medical School, IDepartment of Physiology Queen’s Medical Centre, Nottingham NG7 2UH; and 2Physiology Department, Centre for Human Sciences, Defence Evaluation and Research Agency, Farnborough GU14 6TD, United Kingdom Weller, A. S., C. E. Millard, M. A. Stroud, P. L. Greenhaff, and I. A. Macdonald. Physiological responses to cold stress during prolonged intermittent low- and high-intensity walking. Am. J. Physiol. 272 (Regulatory Integrative Comp. PhysioZ. 41): R2025-R2033,1997.-In a previous study [Am. J. Physiol. 272 (Regulatory Integrative Comp. Physiol. 41): R226-R233,1997], the physiological responses to 240 min of intermittent low-intensity walking exercise in a cold (+5”C), wet, and windy environment (Cold) may have been influenced by a 120-min preceding phase of intermittent higherintensity exercise. Furthermore, the physiological responses observed during this latter phase may have been different if it had been more prolonged. To address these questions, active men attempted a 360-min intermittent (15 min of rest, 45 min of exercise) exercise protocol in Cold and a thermoneutral environment (+15”C, Neutral) at a low (0% grade, 5 km/h; Low; n = 14) and a higher (10% grade, 6 km/h; High; n = 10) intensity. During Low, rectal temperature was lower in Cold than in Neutral, whereas O2 consumption, carbohydrate oxidation, plasma norepinephrine and epinephrine, and blood lactate were higher. During High, Cold had a similar but less marked influence on the thermoregulatory responses to exercise than during Low. In conclusion, the physiological responses to Low are similarly influenced by Cold whether or not they are preceded by High. Furthermore, during intermittent exercise up to an intensity of -60% of peak 02 consumption, a cold, wet, and windy environment will influence the physiological responses to exercise and potentially impair performance. exercise; rectal temperature; shivering; substrate plasma norepinephrine; plasma epinephrine

oxidation;

WATER HAS A MUCH GREATER thermal capacity than air at

an equivalent temperature. Consequently, heat production is usually sufficient to offset heat loss during exercise in a cold dry environment, whereas hypothermia may result during exercise in cold water (14). However, when a low ambient temperature is combined with rain and wind, reduced clothing and air insulation will increase heat loss during exercise and potentially lead to hypothermia. Indeed, Pugh (15, 16) described hypothermic casualties in hill walkers and climbers exposed to cold (NO-5OC), wet, and windy conditions. Clothing insulation was inadequate, and casualties were characterized by mental impairment and an overwhelming feeling of extreme fatigue and exhaustion (15, 16). In an attempt to elucidate this syndrome, Pugh (17, 18) simulated these environmental conditions in the laboratory and reported that at a low exercise intensity [02 consumption (+02) -1.0 Ymin] ?02 was enhanced by -5O%, rectal temperature (T,,) 0363-6119/97

$5.00

Copyright

o 1997

was reduced by .- 1°C and the thermal gradient extended deeper into the thigh muscle than in a cold, dry, and wind-still condition. However, at a higher exercise intensity (V02 -2.2 l/min) wind and wet clothing did not influence VO,, and T,, was only -0.4OC lower than in the control condition (18). Pugh (18) described a VO, “cutoff’ point, above which individuals exercising in a cold, wet, and windy environment would not experience any influence on the physiological responses to exercise but below which there would be an obligatory increase in energy expenditure and subnormal T,, and muscle temperatures. This cutoff point is likely to be dependent on factors such as clothing insulation, body morphology and mass, and body fatness (23) and may account for the random nature of the hypothermic casualties described by Pugh (15,16). Pugh’s work has important implications to recreational and military activities in a cold, wet, and windy environment. However, it should be recognized that the conclusions were based on only three subjects, the mode of exercise was cycling, and more detailed metabolic responses were not investigated. With regard to the latter, reduced body temperatures per se and/or the physiological responses to cooling (peripheral vasoconstriction and shivering thermogenesis) during exercise in a cold environment may influence exercise metabolism (for reviews see Refs. 3 and 27). We recently reported the influence of a cold, wet, and windy environment on the physiological responses to a 360-min intermittent treadmill-walking protocol (26). Compared with a thermoneutral, windy condition, T,, and the metabolic responses to an initial 120-min phase of exercise at -60% of peak vo2 (90apeak) were not influenced by cold stress. However, during a subsequent 240-min phase of exercise at -30% Vo2ppak, T,, was -0.6OC lower, whereas the following were higher: $02 (-25%), the proportion of carbohydrate oxidized, and the venous concentrations of lactate, glucose, norepinephrine (NE), and epinephrine (Epi). It is possible that the cold-induced differences observed during this lower-intensity phase were influenced by the preceding higher-intensity phase. Furthermore, it is possible that the cold, wet, and windy environment would have influenced the physiological responses during the intermittent higher-intensity exercise, if this phase had been more prolonged. Therefore, the aim of the present study was to establish the effects of a cold, wet, and windy environment on the physiological responses to prolonged intermittent walking, when the exercise periods were maintained at a low or a higher intensity. It was hypothesized the American

Physiological

Society

R2025

RZO26 that the cold-induced responses to intermittent be due to an insufficient Furthermore, it was would not influence the prolonged intermittent

COLD

STRESS

influences on low-intensity rate of heat hypothesized physiological high-intensity

AND

LOW-

AND

the physiological walking would production per se. that cold stress responses to more walking.

METHODS Subjects Fourteen active men participated in the study, which was reviewed and approved by the Ethics Committee of the Defence Evaluation and Research Agency, Centre for Human Sciences. Each subject had been screened for medical history and normal cardiovascular function, and the nature, purpose, and possible risks associated with the study were explained before written consent was obtained. The study consisted of four conditions: a 360-min intermittent walking protocol at two contrasting exercise intensities (Low and High, see below), each in two different environments (Neutral and Cold). AI1 14 subjects attempted Low, whereas 10 of the subjects attempted High. During Low/Cold, five subjects withdrew from the experiment at various stages before the scheduled completion, and one subject was withdrawn after 133 min when T,, had fallen below 35°C. Four subjects withdrew from the experiment within 143 min of starting High/Neutral or High/Cold, and the remaining six subjects completed 240-360 min. The physical characteristics of the eight subjects who completed 360 min in Low were as follows: age 27 + 1 (SE)yr; height, 1.76 t 0.03 m; body mass, 77.27 t 3.72 kg;body surface area, 1.94 * 0.06 m2 (4); percent body fat (%BF), 15.4 t 1.6. The physical characteristics of the six subjects who completed 240 min in High were as follows: age, 27 ? 1 yr; height, 1.72 t 0.03 m; body mass, 72.16 2 2.49 kg; body surface area, 1.85 ? 0.05 m2; %BF, 12.6 ? 1.5. The methods and procedures used in this study were similar to those previously reported (26). Consequently, a detailed description of methodology is not duplicated here. Study Protocol After a 3-MJ standard meal, body density and %BF were estimated by skinfold thicknesses (5). Subjects were then weighed nude and instrumented in a temperature-controlled room (25°C). This included placement of a rectal probe, skin thermistors, electrocardiogram electrodes, and a venous cannula. The subjects were then dressed in a cotton combat uniform, and prechamber baseline (Base) measurements were obtained. The subjects entered the cold chamber and started the resting phase of the exercise protocol 134 t 2 (Low) and 133 ? 4 min (High) after first reporting to the laboratory. Each subject completed a 360-min intermittent treadmillwalking exercise protocol (15 min of rest, 45 min of exercise) in pairs. In Low the six exercise periods (L1-L6) consisted of horizontal treadmill walking at a velocity of 5 km/h, whereas in High the four work periods (HI-H4) consisted of inclined (10%) treadmill walking at a velocity of 6 km/h. Unlike our previous study (26), vogpeak was not measured in this study. However, inasmuch as five of the present subjects participated in the previous study and the other subjects were drawn from a similar population, it is not unreasonable to assume that Low and High equated to -30 and -60% of vo 2 peak (26). Although high-intensity exercise usually refers to an intensity greater than that used in the present study, High is used in this study merely to distinguish it from the lower-intensitv exercise (Low). In a thermoneutral (Neutral)

HIGH-INTENSITY

WALKING

environment the target ambient temperature was + 15°C and the clothing was not wetted. In the Cold environment the target ambient temperature was +5”C and the clothing was wetted at regular intervals. In both environments the subjects faced a direct air current of -5 m/s. The mass of water intake and urine output were recorded. At the end of the experiments the subjects were rewarmed slowly if required and then reweighed nude. The time between each condition was 6.3 t 0.2 days in Low and 5.8 t 1.5 days in High. Measurements Temperatures, heart rate, and fluid Loss. Temperatures [T,e, skin (T&, and ambient] and heart rate (HR) measurements were averaged over 15-min blocks. Mean Tsk (T&) was estimated by the formula of Ramanathan (19). Fluid loss (g/min) was determined by subtracting urine output and the estimated loss of metabolic mass from the change in nude body weight (pre- vs. postexperiment) and adding fluid intake. ThermaL balance. All thermal balance data (rates) are expressed in watts per square meter, The rate of heat gained or lost (heat debt) from storage (2s) was computed during each exercise period from the equation of Burton and Edholm (1). The rate of energy expenditure (M) was calculated from the formula of Jequier and co-workers (12), and in Low, where no significant external work was performed, this equatedto the rate of internal heat production (HP). During High, HP was calculated by subtracting the external work produced in walking on the inclined treadmill. The rate of heat loss (HL). during.each exercise period was estimated by subtracting S from HP. The total amount of heat lost from the body mass (-S in kJ/m2) was also calculated at the end of each condition in relation to the initial thermal state. The average thermal insulation of the tissues (Insti) at the end of each exercise period was calculated by the formula used by Burton and Edholm. Respiratory gas exchange. Expired air samples were obtained at Base and at 15 and 30 min into each exercise period using the Douglas bag technique. v02, CO2 production, respiratory exchange ratio (RER), and the ventilatory equivalent for VOW were calculated for all samples. The amount and rate of carbohydrate and fat oxidized were estimated from nonprotein VOW and RER data using the formulas detailed elsewhere

(26) Bkood sampling and analysis. Blood samples were obtained from subjects in a semireclined position at Base and upright during the last 5 min of each exercise period. Samples were suitably handled for the subsequent determination of blood glucose, lactate, glycerol, and P-hydroxybutyrate; plasma free fatty acids, NE, and Epi; and hematocrit. Because of the considerable diminution of peripheral blood flow in Cold, not all scheduled blood samples were obtained. However, all the blood samples withdrawn during Base and at the end of L2, L4, and L6 were obtained in six subjects in Low. In High a complete data set was secured in five subjects during Base, at the end of H2, and at the termination of the exercise protocol. Data Analysis

and Statistics

Values are means t SE. Data were analyzed by repeatedmeasures analysis of variance (ANOVA) with two withinsubjects factors (condition and time) using SuperANOVA software (version 1.11, Abacus Concepts, Berkeley, CA). ANOVA results were corrected by the Huynh-Feldt E-adjusted degrees of freedom when the sphericity test was significant and significant condition and condition-time interactions were identified. The exact level of significance at each time point was calculated using Fisher’s protected least&n&ant

COLD

STRESS

AND

LOW-

AND

HIGH-INTENSITY

NEUTRAL COLD

difference post hoc analysis, which was also used to identify within-condition differences between the initial values (Base or 0 min) and the end of the first exercise period. Student’s paired t-tests were used when the comparison of only two means was required. The degree of association between variables was assessed by calculating Pearson product-moment correlations. Simple regression analysis was also used to assess the relationship between selected variables. Statistical significance was set at P 5 0.05 for all statistical tests. RESULTS

Exercise Duration and Ambient Conditions All subjects completed Low/Neutral, whereas the mean duration in Low/Cold was 294 t 23 min. In High there was no significant difference in the time endured in Neutral (303 2 23 min) and Cold (237 ? 32 min). The mean wet and dry bulb temperatures in the environmental chamber in each condition were +10.67 5 0.27 and +15.00 t 0.12OC in Low/Neutral, +4.27 5 0.17 and +5.02 ? 0.12OC in Low/Cold, +ll.Ol t 0.29 and +14 75 -+ 0 12OC in High/Neutral, and +4.12 t 0.05 and ‘+5.16 ? 0.13OC in High/Cold. The mean wind velocities were 5.5 and 4.9 m/s in Low and High, respectively.

l

R2027

W&KING

36.5 -

tt** tt** tt$$ tt$tt**tt** ** ** * J11 vuI I1 JJ JJ RI

R2 -

Ll

R3

L2 60

0

-R4

L3

R5

L4 180

120

Time

JJ

R6

L5 240

L6 300

1 360

(min)

+ u

NEUTRAL COLD

38.5

38

Thermal and HR Responses T,,. In Low there was a significant condition (P 5 0.01) and condition-time interaction (P 5 0.0001) in T,, (Fig. 1A). T,, was lower in Cold than in Neutral (apart from 210 min) from 180 min to the cessation of the exercise protocol (P 5 0.05). T,, at 60 min (37.53 t 0.09 and 37.56 t 0.09”C in Neutral and Cold, respectively) was not different from that at 360 min in Neutral but had decreased to 36.80 t 0.15OC in Cold (P 5 0.01). In High the condition-time interaction in T,, just attained significance (P = O.OS), and T,, was lower in Cold than in Neutral at 240 min (Fig. 1B). T,, at 60 min (38.07 2 0.13 and 38.23 t O.lO°C in Neutral and Cold, respectively) was not different from that at 240 min in Neutral but had decreased to 37.65 t 0.22”C in Cold (P 5 0.01). TSk.Tsk at the end of each exercise period are listed in Table 1. In Low and High, Tsk was lower in Cold than in Neutral except at 0 min (condition effect, P 5 0.001; condition-time interaction, P 5 0.0001). If individual Tsk sites in the Cold condition are considered at the end of Low, chest temperature (Tch, 21.74 t 1.02OC) was higher than thigh (Tth, 15.70 t 0.47”C, P 5 O.OOl), arm U’ arm2 16.57 t 0.94”C, P 5 O.Ol), and small toe temperature (Ttoe, 15.78 t 0.60°C, P 5 0.001). At the end of High in Cold, there was no difference among Tch (21.16 t 0.42”C), Tar, (19.85 t 0.88OC), and Ttoe (21.80 t 2.15”C), although Tth (18.18 t 0.56”C) was lower than Tch (P 5 0.001). Thermal balance. The total heat debt (-S) at the end of each condition was -212 t 27 kJ/m2 in Low/Neutral, -1,290 t 83 kJ/m2 (-1,203 t 79 kJ/m2 after 240 min) in Low/Cold, -122 -+ 69 kJ/m2 in High/Neutral, and -951 + 68 kJ/m2 in High/Cold. The rate of heat debt C-S), l?P, and l!IL during each exercise period is shown in Fig. 2. Compared with High, despite an -5.5-fold

37.5

37

36’5i

Rl

R2

Hl

-R3 H2

-R4 H3

H4 I

0

60

120

Time

180

I

240

(min)

Fig. 1. Rectal temperature plotted against time during low-intensity (Low, A) and higher-intensity (High, B) intermittent exercise protocol in thermoneutral (15”C, Neutral) and cold (5”C), wet, and windy (Cold) environments. Values are means + SE; n = 8 and 6 for Low and High, respectively. Rl-R6, rest periods 1-6; Hl-H4 and Ll-L6, work periods during High and Low. Significant betweenand withincondition differences (assessed by Fisher’s least-significant difference) are denoted by single (P 5 0.05) and double (P 5 0.01) symbols. * Difference between Neutral and Cold. Within-Neutral differences: T between Base and subsequent values; 8 between LZ / HZ and subsequent values. Within-Cold differences: *between Base and subsequent values; s between L2 / H2 and subsequent values.

greater cold-induced increase in I!IP during Low (mean 27 and 5% over the exercise periods in Low and High, respectively), the larger heat debt at this exercise intensity was associated with an -2.5-fold greater cold-induced increase in HL (mean 47 and 19% in Low and High, respectively). Ins,i. In the average Insti at the end of each exercise period was greater in Cold than in Neutral (condition effect, P 5 0.0001): mean 0.046 t 0.003 and 0.076 t 0.003 W-l. m2* “C in Neutral and Cold, respectively, over the exercise periods. In High, Insti was also higher in Cold than in Neutral (condition effect, P 5 0.01; condition-time interaction, P 5 0.05): mean

LOW

1

R2028

COLD

STRESS

AND

LOW-

AND

Table 1. Mean skin temperature at the start of the exercise protocol and at the end of each work period in Neutral and Cold during Low and High

HIGH-INTENSITY

W&KING

A 600

Neutral

Cold

32.6 30.0 29.3 29.2 29.3

n

S-NEUTRAL 0 q HP- "

q

HL-

#

”

32.7 24.0 20.8 19.8 18.8

HL-

”

**JJ

**JJ

JJ **

2 0.2 + 0.2a9c k 0.3a,c +_ 0.4a,C 5 O.isaJ

S-COLD

HP- ”

1

m ta

Low 0 min Ll L2 L3 L4 L5 L6

•,

2 0.3 2 0.6b + 0.5f,g +_ 0.5f,g 2 oAfpg

29.1+0.4a,c

18.6+o.4fyg

29.3 + 0.4a9c

18.5 + 0.4fpg

High 0 min Hl H2 H3 H4

33.0 k 0.2

33.2 25.8 22.8 22.2 20.6

31.3 +0.4"yb 29.7 It 0.7a,c 29.7 + 0.5a9”J 29.2 5 0.5a9c,e

2 0.2 + 0.8f + 0.8f9g _+ 0.6f9g 2 0.6fpg

Values are means 4 SE in “C; n = 8 (Low) and 6 (High). Neutral, thermoneutral (15°C); Cold, cold, wet, and windy (5°C); Low and High, intermittent low- and high-intensity walking. Level of significance was calculated using Fisher’s protected least-significant difference post hoc analysis. aSignificantly different from Cold: P < 0.01. Significantly different from 0 min within Neutral: bP < 0.05; “P < 0.01; significantly different from subsequent values within neutral: dP < 0.05; eP < 0.01. fSignificantly different from 0 min within Cold: P < 0.01. Qignificantly different from subsequent values within Cold: P < 0.01.

0.026 2 0.002 and 0.042 2 0.001 W-l m2 “C in Neutral and Cold, respectively. Fluid balance and HR. In Low the overall rate of fluid loss was greater in Neutral than in Cold (2.23 t 0.15 and 1.56 t 0.14 g/min, P 5 0.01); this was also the case in High (9.80 t 0.99 and 4.50 t 0.75 g/min in Neutral and Cold, respectively, P 5 0.001). HR at the end of each exercise period are given in Table 2; it is evident that the environmental conditions did not affect this variable during Low or High. l

-600 ,+ 60

180 240 Time (min)

120

300

360

S-NEUTRAL 0 HP- ”

B

mm ”

S-COLD

q

HP- ”

H

HL-

”

600. *;I .

400 .

E 8

200.

21

0.

l

-600.

I

60

120

180

Time

I

I

240

(min)

Fig. 2. Rate of heat storage (2S>, heat production @P) and heat loss (HL) during each exercise period in Low (A) and High (B). See Fig. 1 legend for explanation of abbreviations and symbols.

Metabolic Responses Po2. In Low, VO, was higher in Cold than in Neutral (condition effect, P 5 0.0001; condition-time interaction, P 5 0.0001; Fig. 3A). Differences were located from the end of L2 (P 5 0.01, least significant difference), and the mean of all 90~ values was -25% higher in Cold than in Neutral. From the value at the end of LI (14.8 t 0.5 and 15.9 t 0.6 mlkg-l l rnin-l in Neutral and Cold, respectively), *02 at the end of the final exercise period had not changed in Neutral but had increased -30% to 20:8 t 0.9 ml*kg-l*min-l in Cold (P 5 0.01). In High, VO, was higher in Cold than in Neutral (condition effect, P 5 0.01; Fig. 3B). However, the mean VO, over all time points was only 6% higher in Cold, and the post hoc analysis did not detect any between- or within-condition differences (HI; 33.6 t 1.6 and 35.4 t 1.4 ml.kg-l*minl in Neutral and Cold, respectively). Substrate oxidation. There were no between- or within-condition differences in the ventilgtory equivalent for VOWin Low or High; therefore, VO, and RER were used to estimate rates of substrate oxidation. In

Table 2. Mean heart rate at the start of the exercise protocol and at the end of each work period in Neutral and Cold during Low and High Neutral

Cold

Low 0 min Ll L2 L3 L4 L5 L6

69.0 96.3 94.9 93.8 96.7 98.5 99.8

t 2.6 +_ 3.8 + 4.2 + 4.0 t 4.7 2 5.0 +_ 5.0

75.155.6 94.7 92.8 93.4 96.5 98.0 100.7

2 2.7 5 1.8 + 1.7 2 3.0 + 2.6 +_ 2.8

69.6 137.9 138.3 143.4 141.3

+ 2 + k +

High 0 min Hl H2 H3 H4

67.8 + 4.5 138.12 5.8

139.355.7 139.7 !I 4.7

139.624.9

n = 8 (Low) Values are means + SE in beats/min; There were no significant betweenand within-condition

3.4 5.7 5.6 7.6 8.5

and 6 (High). differences.

COLD

STRESS

A

II **; 0 ++

I ; 98

L

;;

II

+m

AND

LOW-

0

NEUTRAL

0

COLD

AND

11 ** !! ;H’ 00 I)+

i@

m

NHJTRAL

cl

COLD

40-

HIGH-INTENSITY

R2029

W&KING

cise. The carbohydrate oxidation rates (CI!IO,X) in Low and High are shown in Fig. 4, A and B, respectively. In Low, CHO,, was higher in Cold than in Neutral (condition effect, P 5 0.0001; condition-time interaction P 5 0.001). Compared with the value at the end of LI (‘0.55 t 0.08 and 0.66 -+ 0 08 g/min in Neutral and Cold, respectively), by the end of L6 &IO,, had decreased to 0.27 t_ 0.04 g/min in Neutral (P 5 0.01) but had not changed in Cold. In High, CHO,, was not different between conditions, although within both conditions ClkO,, had decreased from the value at the end of H1 to that at the end of H4 (P 5 0.05). Venous metabolites and catecholamines. There were no between- or within-condition differences in hematocrit during Low and High; therefore, any changes in plasma volume would have been small. Consequently, circulating concentrations have not been corrected for hemoconcentration. Venous metabolites are listed in Table 3. In Low and High, there were no betweencondition differences in blood glucose, glycerol, and

A

0

NEUTRAL

0

COLD

330.

220. 1.5. 1.

10,

0.5 0

is O s

60

120

Time

180

240

U

(min)

3 O

Fig. 3. 02 consumption at baseline (Base) and during low-intensity