Rodriguez, Gonzalez-Alonso & Coyle (1995) observed that both water and carbohydrate

Experimental Physiology (1998), 83, 419-430 Printed in Great Britain THE EFFECTS OF SUBSTRATE AND FLUID PROVISION ON THERMOREGULATORY, CARDIORESPIRATORY AND METABOLIC RESPONSES TO PROLONGED EXERCISE IN A COLD ENVIRONMENT IN MAN S. D. R. GALLOWAY* AND R. J. MAUGHAN Department ofenvironmental and Occupational Medicine, University Medical School, Aberdeen AB25 2ZD, UK (MANUSCRIPT RECEIVED 25 NOVEMBER 1997, ACCEPTED 8 JANUARY 1998) SUMMARY During prolonged exercise in a cold environment, fatigue is generally associated with a depletion of endogenous glycogen stores. This has lead many authors to hypothesize that the carbohydrate (CHO) content of fluids ingested in cool environments should be high, yet this hypothesis has not been specifically examined. In the present study, six healthy males cycled to exhaustion at approximately 80 % of their maximum oxygen consumption (Vo2,max) with either no drink (ND), a 15 % CHO-electrolyte drink (15 % CHO) or a 2 % CHO-electrolyte drink (2 % CHO). Dietary intake and exercise were replicated 2 days prior to each trial. Mean (S.D.) ambient temperature was 10.0 (0-3) C with a relative humidity of 72 (2) % and an air velocity of approximately 0-7 m s-5 on all trials. Weighted mean skin temperature was calculated, and rectal temperature and heart rate were recorded at rest, during exercise and at exhaustion. Venous samples were drawn before and during exercise and at exhaustion for determination of haemoglobin, haematocrit, blood metabolites and serum electrolytes and osmolality. Expired air was collected for calculation of VO and respiratory exchange ratio which were used to estimate rates of fuel oxidation. Ratings ot perceived exertion (RPE) were also obtained. Exercise capacity was not different (P = 0.49) between trials, with median (range) times to exhaustion of 90 6 (66 8-106 0), 97-7 (60*9-112 0) and 102.0 (77.1-170.7) min for the ND, 15 % CHO and 2 % CHO trials, respectively. The 15 % CHO drink significantly (P < 0.05) elevated blood glucose concentration and total CHO oxidation compared with the ND trial. The 2 % CHO drink significantly (P < 0.05) reduced the estimated change in plasma volume and resulted in a lower serum osmolality than the other two trials. No differences were observed in any thermoregulatory or cardiorespiratory responses between trials. These data suggest that maintenance of blood-borne substrate and/or plasma volume does not influence exercise capacity in the cold, as is commonly reported during this type and intensity of exercise in warmer environments. INTRODUCTION Fatigue in prolonged cycle exercise is generally attributed to a depletion of endogenous glycogen stores and/or problems associated with dehydration. It has previously been observed that drinks aimed primarily at either fluid or carbohydrate (CHO) provision can delay the onset of fatigue during exercise in the heat (for review, see Maughan, 1991). Below, Mora- Rodriguez, Gonzalez-Alonso & Coyle (1995) observed that both water and carbohydrate ingestion could improve cycling performance at 30 C and the effects were additive. Maughan, Bethell & Leiper (1996) found that even small amounts of ingested carbohydrate or plain water could improve exercise capacity at 21 'C. In addition, from our previous work (Galloway & Maughan, 1997) in which we observed that exercise capacity in a cold (11 C) * Corresponding author: s.d.r.galloway@strath.ac.uk 1685

420 S. D. R. GALLOWAY AND R. J. MAUGHAN environment was greater than exercise capacity in warmer (21 and 31 C) environments and thus total CHO oxidation was greater, we hypothesized that fatigue during prolonged exercise in a cold environment is most probably due to a depletion of endogenous CHO stores. This hypothesis suggests that substrate provision should be of benefit during exercise in a cold environment and reinforces the theory that the formulations of solutions ingested during exercise in the cold should be different from those ingested in warmer conditions. The effects of provision of CHO during prolonged exercise have been studied fairly extensively over the past 65 years. The early findings of Dill, Edwards & Talbott (1932) indicated that maintenance of blood glucose concentration could delay the onset of fatigue during exercise lasting 3-6 h. Since these early reports, many authors have observed beneficial effects of CHO provision during prolonged exercise (for review, see Coyle, 1992). The majority of studies indicate a beneficial effect of ingesting CHO, and some attribute the delay in onset of fatigue either to a partial sparing of endogenous muscle glycogen (Hargreaves, Costill, Coggan, Fink & Nishibata, 1984; Erickson, Schwarzkopf & McKenzie, 1987), maintenance of blood glucose (Coyle, Coggan, Hemmert & Ivy, 1986; Coggan & Coyle, 1987), or partial restoration of blood volume (Sawka, Hubbard, Francesconi & Horstman, 1983). The large majority of these studies have been performed at ambient temperatures above 20 'C. Thus, the one factor that has not been investigated is whether CHO provision is beneficial during exercise in a cold environment where substrate depletion is more likely to be a key factor in the cause of fatigue. The aim of the proposed work was, therefore, to investigate the effects of administering a concentrated CHO-electrolyte drink (to offset the effects of substrate depletion) and a dilute CHO-electrolyte drink (to offset the effects of hypohydration) on exercise capacity and the thermoregulatory, cardiorespiratory and metabolic responses to exercise on exposure to a low ambient temperature. The study should therefore determine whether CHO provision is an important factor during prolonged exercise in a cold environment. METHODS Six healthy male volunteers were studied. The mean (S.D.) physical characteristics of the subjects were: 26 (5) years of age, 71-7 (9-8) kg body mass, 177 (7) cm height, 1-89 (0-17) m2 body surface area and 4-21 (0-62) 1 min-' maximum oxygen consumption (Vo max) The subjects' Vo2max was recorded during an initial testing session and a maximal test verification was performed a few days later. The maximal test followed a discontinuous incremental protocol on an electrically braked cycle ergometer. All subjects were active, but none was specifically cycling trained. All subjects were given written information concerning the nature and purpose of the study and gave written informed consent. This study was approved by the Joint Ethical Committee of Aberdeen University and Grampian Health Board. Experimental protocol Each subject completed five rides to exhaustion on an electrically braked cycle ergometer at -80 % of Vo2max in a climatic chamber maintained at an ambient dry bulb temperature (Ta) of 10 'C (mean (S.D.), 10-0 (0-3) 'C), relative humidity of 72 (2) % and an air velocity of approximately 0.7 m s-1 on all trials. The first two of these tests served to familiarize the subjects with the experimental protocol. These familiarization trials also allowed evaluation of the power output to ensure that an intensity corresponding to approximately 800/ of Vo2max was attained and that the subjects were able to exercise for a duration of greater than 1 h. The final three rides were the experimental trials and were performed 1 or 2 weeks apart and at the same time of day. On these trials, subjects either ingested a 2 % dilute glucose-electrolyte drink (2 % CHO), a 15 % glucose polymer-sucrose-electrolyte drink (15 % CHO), or no drink (ND) as a control. The experimental conditions were administered in a cross-over randomization and the subjects were not informed of the drink composition during any of the trials. Subjects were instructed to record

SUBSTRATE INGESTION AND EXERCISE IN THE COLD 421 their dietary intake and physical activity for 2 days before the first experimental trial and were asked not to exercise heavily on the day before the trial; their dietary intake and physical activity were then replicated prior to the other two experimental trials. During the experimental period the subjects attended the lab in the morning following an overnight fast and emptied their bladder and defaecated if necessary prior to nude body mass being measured. The subjects then inserted a rectal thermistor 10 cm beyond the anal sphincter and rested in a sitting position for approximately 30 min in a standardized environment (26.4 (0.3) OC). Subjects were dressed in only shorts, socks and shoes on all trials. Thermistors (Comark, Kent, England) for recording of skin temperature were attached to the chest, triceps, thigh and calf (Ramanathan, 1964) and a Sport tester heart rate monitor (Polar PE3000, Bodycare, Kenilworth, UK) positioned. For the last 10 min of this seated rest period the subject's hand was immersed in hot (42 C) water, a venous cannula was then inserted into a lower forearm vein and two resting arterialized-venous blood samples (6.5 ml) were obtained 5 min apart. The cannula was kept patent by a slow (approximately 0.5 ml min-') saline drip on all trials and a glove was worn on the appropriate hand during all trials in an attempt to maintain arterialization of the venous blood samples. Baseline recordings of rectal (Tre) and skin (TAk) temperatures were made and resting heart rate recorded. Immediately following collection of the second resting blood sample, the subjects transferred to the climatic chamber where they either ingested a bolus of 7.14 ml (kg body mass)-1 of the assigned drink and then began exercising or, if on the no drink trial, immediately began exercising. During exercise on the trials where a drink was assigned, subjects ingested either 3.57 ml kg-' of the 2 % CHO drink or 1.79 ml kg-' of the 15 % CHO drink every 10 min. The drinks were maintained at a temperature of 14 C in a water bath. The subjects were asked to maintain a cadence of 60-70 r.p.m. throughout the test and exhaustion was defined as the point at which the subjects could no longer continue or could no longer maintain a cadence above 60 r.p.m. Blood samples (6.5 ml) were drawn during exercise at 15 min intervals and at exhaustion (END, immediately after the subject stopped). All blood samples were collected using a dry syringe; 2.5 ml of blood were immediately dispensed into a tube containing K3-EDTA (1.5 mg ml-'). Duplicate aliquots (100,ul) were then removed and were immediately deproteinized in 1 ml of ice cold 0-3 N perchloric acid (PCA) which was then centrifuged and the supernatant used for measurement of blood glucose, blood lactate and blood glycerol. The remaining 2.3 ml of blood in the EDTA tube was used for determination of haemoglobin and microhaematocrit. A 4 ml aliquot of blood was dispensed into a plain serum tube and the serum obtained was divided into two aliquots. One aliquot was immediately frozen and stored at -20 C until subsequent determination of serum free fatty acids (FFAs). The second aliquot was refrigerated and used for determination of serum electrolytes (Na+, K+ and Cl-) and serum osmolality. Blood glucose, haemoglobin and microhaematocrit were all determined within 5 h of sampling, and serum osmolality, Na+, K+ and Cl- within 2 days. The remaining plasma, serum and PCA supernatant were frozen and stored at -20 C until analysed. Expired gas was collected into Douglas bags over a 2 min period every 15 min and immediately analysed to determine minute ventilation (VE), oxygen uptake (VO), and respiratory exchange ratio (R), which was used to estimate rates of fuel oxidation (Consolazio, Johnson & Pecora, 1963). Where an expired gas collection clashed with a drink time the expired gas was collected early to allow drinks to be given on schedule (e.g. at the 30 min drink time expired air was collected from 28-30 min). Ratings of perceived exertion for both overall perception of exertion and for perceived exertion of the legs were recorded every 10 min throughout the test using the Borg category scale (Borg, 1982). The perceived exertion ratings were recorded immediately prior to the subjects receiving each drink. TWk, Tre, and Ta, heart rate and relative humidity were recorded every 5 min throughout exercise and at exhaustion. Time to exhaustion was noted in all trials and subjects were not informed of their elapsed time or total exercise duration at any stage during the study. Weighted mean skin temperature (TIk) was calculated using the equation of Ramanathan (1964). Following exercise, a final nude body mass was obtained and mass loss, corrected for fluid volume consumed, urine losses (1 subject only), respiratory water losses and losses through CO2-02 exchange (Mitchell, Nadel & Stolwijk, 1972), was taken to represent sweat loss. A postexercise urine sample was obtained to determine urine volume and urinary electrolyte losses. Assay methods Blood glucose was measured manually by the glucose oxidase method (Boehringer Mannheim Biochemica). Blood lactate was determined using a Roche Cobas Fara (automated centrifugal analyser)

422 S. D. R. GALLOWAY AND R. J. MAUGHAN and a method based on that of Maughan (1982). Blood glycerol was also determined by using a Roche Cobas Fara and a method based on that of Boobis & Maughan (1983). Serum FFAs were measured by the Roche Cobas Fara and an enzymatic colorimetric method (Boehringer Mannheim Biochemica, catalogue no. 1 383 175). Serum and urine Na+ and K+ were determined by flame photometry (Corning clinical flame photometer 410 C, Halstead, UK). Serum osmolality and urine osmolality were determined by freezing point depression (Gonotec osmomat 030, YSI Ltd, Farnborough, UK). Serum and urine C1- were determined by potentiometric titration (Jenway chloride meter, Dunmow, UK). Microhaematocrit and haemoglobin (cyanmethaemoglobin method) were measured on all samples for calculation of percentage changes in plasma volume (Dill & Costill, 1974). All analyses were performed in duplicate except for haematocrit which was measured in triplicate. All c.v. values were 2.0 % or below, except for serum K+ which had a cv. of 3.5 %. Data analysis All data are presented as means (S.E.M.) in figures and as means (S.D.) in text and tables. Where appropriate, following a Shapiro-Wilks test for normality of distribution, values are expressed as median (range). A two-way, two-factor analysis of variance (ANOVA) for repeated measures was applied to determine any treatment differences and time effects during the exercise protocol. Following observation of a main effect, ANOVA or Kruskal-Wallis tests were performed to determine at which time points an effect was observed. Post hoc analysis by Student's paired t test or the Wilcoxon test was performed to determine which trials were significantly different. ANOVA or Kruskal-Wallis tests were applied to determine any initial baseline differences in all variables. In all cases, significance was taken at P < 0*05. A power test revealed that for 90% statistical power with an expected pooled standard deviation in exercise time of about 20 min and a possible mean difference in exercise time (as a result of the intervention) of 30 min we would require 4.7 subjects in the design. We therefore planned the study based on a subject number of 6. RESULTS The mean (S.D.) ambient temperatures were 10.0 (0-3), 9.9 (0.3) and 10.0 (0.3) C for the ND, 15 % CHO and 2 % CHO trials, respectively, and no significant differences were observed between trials at any time. The drink formulations and compositions are presented in Table 1. There were no differences in the Na+ and K+ concentrations of the drinks but Clconcentration, osmolality and total volume consumed were significantly different (P < 0.05) between drinks. In general, the fluid volumes were well tolerated by the subjects but two of the six subjects did experience gastrointestinal discomfort with ingestion of the 15 % CHO drink which may have limited their exercise capacity. Exercise capacity was not different (P = 0.49) between any of the trials with median (range) times to exhaustion of 90.6 (66.8-106.0), 97.7 (60.9-112.0) and 102.0 (77.1-170-7) min for the ND, 15 % CHO and 2 % CHO trials, respectively. Cardiorespiratory and fuel oxidation responses No difference was observed in resting HR or in the HR response to exercise between trials. The largest median difference in HR of 6 beats min-' was observed at the 40 and 50 min sample times between the 15 % CHO trial and the 2 % CHO trial. Median (range) HR values at exhaustion were 164 (136-171), 164 (146-175) and 163 (134-172) beats min-' for the ND, 15 % CHO and 2 % CHO trials, respectively. No differences were noted in the VE, V 2 or R responses to exercise between trials. Significant time effects were observed for these variables in all trials. VE and Vo2 increased over the duration of exercise while R fell (all P < 0.05) over the duration of exercise in all trials. As a result of the elevation in V02 and the fall in R, estimated rates of fat oxidation were observed to increase and estimated rates of CHO oxidation were observed to decrease in all trials over the duration of exercise, but no differences were observed between trials at any time (Table 2). Despite the absence of any

SUBSTRATE INGESTION AND EXERCISE IN THE COLD 423 Table 1. Formulation of the 2 % CHO and 15 % CHO drinks Constituent 2 % drink 15 % drink Glucose (g 1-) 20.0 Sucrose (g l-l) 20-0 Glucose polymer (g 1-) 130.0 Na+ (mmol 1-) 67 (3) 65 (3) K+ (mmol 11) 0-7 (0-2) 0-6 (0.2) Osmolality (mosmol kg-') 240 (1)b 324 (6)- Cl- (mmol 1-1) 57 (1)b 51 (1)' Volume ingested (1) 2-88 (0-96)b 1-42 (0 24)c Values are means (S.D.). b and c indicate a significant difference (P < 0.05) from corresponding values with 15 % CHO and 2% CHO drinks, respectively. All drinks were made up in distilled water and low calorie lemon squash (125 ml 1-) was added for flavouring. Table 2. Estimated rates offat and carbohydrate (CHO) oxidation at 15 min intervals and total oxidation during trials with no drink (ND), a 15 % CHO drink and a 2 % CHO drink Exercise time (min) Total Oxidation Trial 15 30 45 60 75 90 oxidation (g) Fat (g min-') ND 0.36(0.24) 0.52(0.27)* 0.63(0.17)* 0.68(0.23)* 0.71(0.16)* 49.8(21-6) 15 % 0.28(0.13) 0.42(0 10)* 0.48(0 13)* 0.48(0 13)* 0.54(0.1 1)* 0.64(0.12)* 42.8(14.4) 2% 0-26(0.17) 0.41(0.18) 0.50(0.13)* 0.54(0. 10)* 0.57(0.18)* 0.68(0.12)* 52.0(23.1) CHO (g min-') ND 3.03(0.87) 2.75(0.87)* 2.55(0.70)* 2.52(0.74)* 2 41(0.74)* 210-0(48-4) 15 % 3 21(0 70) 3.01(0-68)* 2-95(0.70)* 2-92(0.58)* 2.79(0.45)* 2.70(0.46)* 262-6(62-3) 2% 3.22(0.61) 2.98(0.34) 2.84(0.44)* 2.81(0.44)* 2.82(0-57) 2.37(0.51)* 285.5(68-4) Values are means (S.D.). * Significant difference (P < 0.05) from 15 min values. significant difference in the CHO oxidation rates at each 15 min interval the oxidation of CHO over the first hour was significantly higher in the 15 % CHO trial (173-9 g h-1) than the ND trial (146.5 g h-1). In the 2 % CHO trial, estimated oxidation of CHO in the first hour was 163-2 g h-' and was not different from any other trial. No differences were observed in fat oxidation over the first hour or in total fat or total CHO oxidation between trials. Thermoregulatory responses There was no difference (P = 0.57) in Tre at rest or in the response to exercise between trials at any time. A significant elevation in Tre was observed in all trials between the resting value and all subsequent exercise values. Tre rose from 37-7 (0-2) C at rest to 39-6 (0.6) C at exhaustion in all trials. Similarly, there was no difference (P = 0.58) in TSk at rest or in the response to exercise between trials at any time. Between rest and all subsequent exercise values there was a significant reduction in TSk from 32-6 (0-5) to 26-5 (2-4) C in all trials. Estimated mean sweat rate, assessed from weight change and corrected for fluid volume consumed and urine losses where applicable (one subject), was not significantly different between trials. Sweat rates were 0 86 (0-24), 0.80 (0.26) and 0-75 (0-30) 1 h-' in the ND, 15 %

424 S. D. R. GALLOWAY AND R. J. MAUGHAN Table 3. Serum osmolality, Na +, K+ and Cl - responses at rest, during exercise and at exhaustion (END) in trials with no drink (ND), a 15 % CHO drink and a 2 % CHO drink Trial Exercise time (min) Rest I Rest2 15 30 45 60 75 END Osmolality ND 288(5) 288(6) 296(6)* 296(6)* 296(5)* 296(7)* 298(7)* 298(5)*e (mosmol kg-' )15% 290(3) 290(3) 298(3)* 298(3)*c 297(3)*c 297(3)*c 297(3)* 297(3)*c 2 % 287(2) 287(2) 294(4)* 292(5)*b 292(5)*b 289(4)b 289(3)* Na+ ND 140(2) 140(2) 142(2)* 143(3)* 142(2)* 142(1)* 143(2)* 142(2)* (mmol 1F') 15% 140(2) 140(2) 141(2) 141(2) 142(2) 142(2) 142(1) 141(2) 2 % 140(2) 139(2) 141(2) 140(2) 140(2) 141(1) 140(2) K+ ND 4-7(0.1) 4.8(0.2) 5.9(0.2)* 6.2(0-2)* 6.1(0-1)* 6.4(0.2)* 6.5(0 1)* 6.3(0-5)* (mmol 1') 15% 4-7(0.2) 4.6(0.2) 5-9(0-5)* 5.8(0.3)* 6.0(0.3)* 6.1(0.4)* 6.2(0.5)* 6.2(0.5)* 2% 4-9(0.3) 4.7(0.2) 5.8(0.4)* 6.0(0-2)* 6-0(0-3)* 6.1(0-3)* - 5-9(0-5)* Cl- ND 101(2) 100(3) 103(2)* 103(2)* 102(1)* 102(2)* 101(3) 101(2) (mmol I-) 15 % 100(3) 100(3) 102(2) 102(2) 100(2) 100(3) 102(2) 100(2) 2 % 100(2) 100(2) 101(1) 101(1) 100(1) 101(2) 99(3) Values are means (S.D). * Significant difference (P < 0-05) from resting values. a, b and c indicate a significant diffierence (P < 0.05) from corresponding values with no drink, 15 % CHO and 2 % CHO drinks, respectively. CHO and 2 % CHO trials, respectively. Combined estimated respiratory water loss and mass loss due to substrate exchange were 3*1 (0.3), 3*3 (0.2) and 3*4 (0.2) g min-' in the ND, 15 % CHO and 2 % CHO trials, respectively. Serum and urine electrolyte responses No differences were observed between trials in the serum Na+, K+ and Cl- response to exercise (Table 3). Serum K+ concentration was significantly elevated above that at rest during exercise in all trials. Serum Na+ and Cl- concentrations were significantly elevated above resting values during exercise in the ND trial only. Serum osmolality was significantly elevated from rest at the onset of exercise in all trials and remained elevated at all times in all trials except for at the 60 min time point in the 2 % CHO trial where values were not significantly different from rest. Significantly lower serum osmolality values were observed during exercise in the 2 % CHO trial than during the 15 % CHO trial at the 30, 45 and 60 min time points and at exhaustion. Post-exercise urine volume was significantly higher following ingestion of the 2 % CHO drink (167 (63-847) ml) than in the ND trial (60 (37-115) ml), but was not different from the 15 % CHO trial (70 (18-121) ml). Post-exercise urine osmolality was significantly lower following ingestion of the 2 % CHO drink (316 (112-838) mosmol kg-') than following the ND trial (787 (597-830) mosmol kg-'), but was not different from the 15 % CHO trial (718 (540-816) mosmol kg-'). No differences were observed between any trial in the post-exercise urine concentrations of Na+, K+, or CF. Metabolite responses In the ND trial, blood glucose concentration (Table 4) did not increase above resting values over the duration of exercise. Blood glucose concentration was, however, significantly elevated above resting values at exhaustion in the 15 % CHO trial and at the 45 min sample time in the 2 % CHO trial. Blood glucose concentration was significantly higher during the 2 % CHO trial

SUBSTRATE INGESTION AND EXERCISE IN THE COLD 425 Table 4. Blood glucose and lactate concentrations at rest, during exercise, and at exhaustion (END) in trials with no drink (ND), a 15 % CHO drink or a 2 % CHO drink Exercise time (min) Trial RestI Rest2 15 30 45 60 75 END Glucose ND 4.20(0.27) 4-30(0 31) 3-94(0-15)c 4.08(0.12)bC 4-22(0-39)bc 4.06(0.26)c 3.94(0-27) 4.20(0.52)b (mmol Il) 15 % 4.35(0.68) 4.37(0.67) 4.61(0.87) 4-65(0.67)8 4-80(0-70)8 4-58(0-67) 4-75(0-68) 5-11(0-62)*ac 2 % 4-49(0-21) 4.43(0-21) 4-61(0.45)& 4.58(0.35)8 4-73(0-27)*a 4-56(0.32) 4-56(0-43) Lactate ND 0-49 0-50 2-90* 2-42* 1-90* 1-73* 1-78* 1-96* (mmol 1-) (0.44-0-65) (0-42-0-70) (2-31-5-65) (1-80-5-01) (1-56-4.33) (1-38-3-95) (1-38-3-82) (1-63-4.48) 15% 0.48 0.48 3-14* 2-35* 1-72* 2-28* 2-16* 2-23* (0-37-0-86) (0-36-0-87) (2-52-5-28) (1-92-4-91) (1-65-4-55) (1-51-4-41) (1-37-3-63) (1-31-3-34) 2 % 0-56 0-57 3-17* 2-66* 2-05* 1-80* 1-56* (0-47-0-81) (0-48-0.82) (2-43-5-13) (1-98-5-10) (1-83-4-53) (1-49-3-57) (1-17-3-64) Values are means (S.D.) or median (range) where appropriate. * Significant difference (P < 0.05) from resting values. a, b and c indicate a significant difference (P <0.05) from corresponding values with no drink, 15% CHO and 2% CHO drinks, respectively. than during the ND trial at the 15, 30, 45 and 60 min sample times. Blood glucose concentration was also significantly higher during the 15 % CHO trial than during the ND trial at the 30 and 45 min sample times and at exhaustion (END). In addition, the blood glucose concentration was significantly higher at exhaustion in the 15 % CHO trial than in the 2 % CHO trial. Blood lactate concentration (Table 4) was significantly elevated above resting values during exercise in all trials, but no differences were observed between trials at any time. Serum FFA concentration (Fig. 1) dropped significantly from resting values over the first 15 min of exercise on all trials. In the ND and 2 % CHO trials, serum FFA concentration then increased over the duration of exercise and was above resting values by the 45 and 30 min sample times, respectively. In the 15 % CHO trial, serum FFA concentration remained below resting values until the 60 min sample time, at which time there was a gradual increase in serum FFA concentration. As a result of the different responses, serum FFA concentration was significantly higher during the ND trial than during the the 15 % CHO trial at the 30, 45, 60 and END time points. In addition, a significantly higher serum FFA concentration was observed at the END time point in the 2 % CHO trial compared with the 15 % CHO trial. Blood glycerol concentration (Fig. 1) increased significantly above resting values in all trials; a significant increase occurred within the first 15 min in the ND and 2 % CHO trials, but not until the 45 min sample time in the 15 % CHO trial. The blood glycerol concentration was significantly higher in the ND trial than in the 15 % CHO trial at the 45 min, 60 min and END time points. In addition, blood glycerol concentration at the END time point was significantly higher in the 2 % CHO trial than in the 15 % CHO trial. Blood glycerol concentrations (median (range)) at exhaustion (END) were 0f26 (0.19-0.46), 0* 18 (0.10-0.20), and 0.27 (0-14-0.32) mmol F' for the ND, 15 % CHO and 2 % CHO trials, respectively. Plasma volume responses A significant reduction in plasma volume (Fig. 2) was observed on all trials between rest and the 15 min exercise sample time. Plasma volume then remained significantly reduced below resting values at all times in the ND and 15 % CHO trials. In the 2 % CHO trial, plasma

426 S. D. R. GALLOWAY AND R. J. MAUGHAN -J -i.5 E U- E en cn 1.2 1.1 1.01 0.9 0.8 0.7 0.6 0.5 0.4 - ND -- 15% -o-- 2% { I b 0.3 0.2 nv.1ii N. 0 15 30 45 60 - Time (mins) END 0.35 -J 0.30 0.25 ND -1 15% -o- 2% E 1-0 0 ro 0.20 0.15 0.10 0.05 n-nn w W w w B 0 15 30 45 60 Time (mins) END Fig. 1. Serum FFA (top) and blood glycerol (bottom) concentrations at rest, during exercise, and at exhaustion (END) with no drink (ND), a 15 % CHO drink and a 2 % CHO drink. Serum FFA values are means. Group mean errors are shown to indicate the variance at each time point for clarity. Median values for blood glycerol are shown. Range values have been omitted for clarity but are indicated in the text. Differences from resting values are indicated in the text. a, b and c indicate a significant difference (P < 0.05) from corresponding values with no drink, 15 % CHO drink, or 2 % CHO drink, respectively. volume was partly restored over the duration of exercise, and by the 60 min sample time was not significantly below that at rest. Significantly smaller reductions in plasma volume were observed in the 2 % CHO trial than in the other two trials from the 30 min sample time onwards and at exhaustion (END). No other differences were observed between trials at any time.

SUBSTRATE INGESTION AND EXERCISE IN THE COLD 427 E0 a E -5 *ND -15 15% -- 2%h -20 0 15 30 45 60 END - Time (mins)i- Fig. 2. Change in plasma volume during the first 60 min of exercise and at exhaustion (END) with no drink (ND), a 15 % CHO drink and a 2 % CHO drink. Values are means. Group mean errors are shown to indicate the variance at each time point for clarity. Significance levels are as indicated in Fig. 1. Subjective responses Overall and leg ratings of perceived exertion (RPE) increased over the duration of exercise in all trials but no significant differences were observed between trials at any time for either of these variables. Leg RPE was generally higher than overall RPE with mean scores ranging from 11-18 for overall RPE and 12-20 for leg RPE throughout the exercise duration in all trials. DISCUSSION We have previously observed that exercise capacity in the cold (11 C) is greater than that observed in warmer (21 and 31 C) environments (Galloway & Maughan, 1997), and, due to the longer exercise time, total CHO oxidation is also much greater at 11 C, thus implicating depletion of endogenous CHO stores as a possible main factor in the cause of fatigue at 11 C. This is supported by the muscle glycogen data of Febbraio, Parkin, Baldwin, Zhao & Carey (1996b), who observed that muscle glycogen concentrations (mean (S.E.M.)) at exhaustion from exercise at 70 % of VO2max in ambient temperatures of 40, 20 and 3 'C were 327 (31), 169 (40) and 153 (28) mmol (kg dry wt)-', respectively. Many studies have examined the effects of water provision and carbohydrate provision on endurance exercise performance or exercise capacity in neutral or warm environments (Below et al. 1995; Maughan et al. 1996)

428 S. D. R. GALLOWAY AND R. J. MAUGHAN and have found beneficial and additive effects of ingesting substrate and fluid but none have examined the potential effects in a cold environment. In cold environments, where the thermoregulatory stress is low, it is commonly suggested that drinks administered during exercise should contain a higher CHO concentration as the requirement for fluid replacement is less vital (Hargreaves, 1991; Maughan, 1991; Maughan & Noakes, 1991). Despite these recommendations that CHO provision should predominate in the cold this does not appear to have been studied. The acceptance of CHO provision in the cold is based solely on substrate oxidation data from studies where no fluid was ingested or on anecdotal information from athletes (Astrand & Rodahl, 1986). Exercise capacity in the cold The results indicate that provision of substrate, whether as a concentrated CHO solution (15 % CHO) or a dilute CHO (2 % CHO) solution, does not improve exercise capacity at 10 C under the conditions of the present study. This result is surprising since the higher blood glucose concentration observed following ingestion of the 15 % CHO drink would be expected to result in a delay in the onset of fatigue, as has been observed during exercise at this intensity in a warmer environment (Coyle, Hagberg, Hurley, Martin, Ehsani & Holloszy, 1983). One possible discrepancy between previous work and the present study may be that in the ND trial none of the present subjects became hypoglycaemic (blood glucose below 3-5 mmol F') and thus it could be argued that blood-borne substrates did not not limit exercise capacity at this temperature during this intensity of exercise in the current subject group. Subjects involved in previous CHO supplementation studies in which a delay in the onset of fatigue during exercise at the same intensity as used in the present study was consistently observed typically became hypoglycaemic in the placebo trials (Coyle et al. 1986). The exercise duration in the studies in which this fall in blood glucose was observed were generally much longer (2-4 h). Coyle et al. (1983) observed that subjects who did not become hypoglycaemic during the placebo trials did not exercise for longer following ingestion of CHO. This may explain the lack of a significant effect of ingestion of CHO on exercise capacity in the present study. Alternatively, it may be that there is a larger variability in response to exercise in a cold environment than in warmer conditions. The relatively large variability in exercise capacity in the present study (pooled S.D. of 24*3 min with mean exercise duration of 96.1 min) is somewhat greater than expected. We would expect a pooled S.D. of around 15-20 min with this exercise duration. This larger variability will reduce the chance of observing significant differences in exercise capacity if the differences are small. In the present study, the higher blood glucose concentration and higher rate of CHO oxidation in the first hour of exercise during the 15 % CHO trial than in the ND trial suggests that the 15 % CHO drink was effectively emptied from the stomach and absorbed and may have contributed a portion of the total substrate oxidized. Therefore, the lack of any difference in exercise capacity should not be attributable to poor availability of substrate from the ingested drink. However, two of the subjects did experience some gastrointestinal distress in the form of discomfort from a full stomach when ingesting the 15 % CHO drink which probably resulted in an early termination of exercise for these subjects. Gastrointestinal distress suggests that the drink was not effectively emptied from the stomach or was sitting in the intestine. Despite this discomfort, these two subjects had an elevated blood glucose concentration in the 15 % CHO trial, suggesting that a substantial portion of the drink was emptied and absorbed. No other subjects felt that their exercise capacity was limited by extreme discomfort when drinking the 15 % CHO drink.

SUBSTRATE INGESTION AND EXERCISE IN THE COLD The dilute CHO drink (2 % CHO) was not effective in delaying the point of fatigue in the present study, despite the partial restoration of blood and plasma volume compared with the other trials. This finding is in direct conflict with the results of previous work examining the effects of fluid provision on exercise capacity in warmer environments (Maughan, 1991; Coyle, 1992). The partial restoration of plasma volume indicates that the high fluid volume ingested in the 2 % CHO trial was effective in replacing intravascular fluid losses. The absence of any significant effects of the fluid replacement on heart rate or thermoregulatory responses suggests that there was not a sufficient thermal stress to warrant the large fluid replacement. However, the pattern of exercise times indicates a tendency for a longer exercise duration with the 2 % CHO drink. Although the effect of drink ingestion was not significant in the present study, it suggests that provision of a small amount of CHO is as beneficial as supplementation with larger amounts of CHO during exercise in the cold. The present findings are similar to those of Febbraio et al. (1996a), who observed that ingestion of a 7 % CHO solution during prolonged exercise at 5 C was more beneficial than ingestion of a 14 % CHO solution. This may indicate that the rate of exogenous oxidation of CHO from these very different drinks was similar. This claim would have to be substantiated by examining the oxidation of exogenous substrate from these drinks during exercise in the cold. If the rate of oxidation of exogenous substrate was similar in the two drink trials then it would suggest that the composition of a drink designed for a cold environment does not necessarily have to be high in carbohydrate. An important factor in the present study is that the effects of CHO content and fluid volume alone have not been examined. In this study, therefore, the effects of a large fluid intake or a large CHO intake were observed and it cannot be stated whether fluid volume or CHO intake is likely to be the more important factor during exercise in the cold or whether a compromise between the two extremes would be beneficial. It seems that future studies should examine the effects of CHO content alone or fluid volume alone on exercise capacity in the cold. In addition, further studies should determine whether high CHO contents are necessary in cold conditions based on the proportion of exogenous CHO oxidized from drinks with a range of CHO contents. Conclusion Despite observing effects on blood glucose and plasma volume which have both previously been implicated with the delay in onset of fatigue during prolonged exercise at the intensity described in this study, no differences in exercise capacity were observed. This suggests that, in our experiments, exercise capacity was limited by some other factor in the cold (10 C) environment. A possible explanation would be that a selective glycogen depletion of fibres is occurring in the active muscles, resulting in an inability to maintain the required power output. This may be particularly important to the exercise model adopted in the present work where constant load exercise is performed until exhaustion. With no decline in pedal cadence below 60 r.p.m. and no decline in power output the specific fibres recruited may become depleted. This is a speculative explanation since muscle glycogen data were not obtained in the present work; however, the decline in CHO oxidation observed during the latter stages of each trial suggests that there was a depletion of endogenous muscle glycogen resulting in a decreased CHO utilization. The authors would like to thank SmithKline Beecham Consumer Healthcare for supporting this work. 429

430 S. D. R. GALLOWAY AND R. J. MAUGHAN REFER ENCES ASTRAND, P.-O. & RODAHL, K. (1986). Textbook of Work Physiology, 3rd edn, chap. 11, pp. 564. McGraw-Hill, Singapore. BELOW, P. R., MORA-RODRIGUEZ, R., GONZALEZ-ALONSO, J. & COYLE, E. F. (1995). Fluid and carbohydrate ingestion independently improve performance during 1 h of intense exercise. Medicine and Science in Sports and Exercise 27, 200-210. BOOBIS, L. H. & MAUGHAN, R. J. (1983). A simple one-step enzymatic fluorimetric method for the determination of glycerol in 20,l of plasma. Clinical Chimica Acta 132, 173-179. BORG, G. A. (1982). Physiological bases of perceived exertion. Medicine and Science in Sports and Exercise 14, 377-381. COGGAN, A. R. & COYLE, E. F. (1987). Reversal of fatigue during prolonged exercise by carbohydrate infusion or ingestion. Journal ofapplied Physiology 63, 2388-2395. CONSOLAZIO, C. F., JOHNSON, R. E. & PECORA, L. J. (1963). Physiological measurements of metabolic functions in man. McGraw-Hill, New York. COYLE, E. F. (1992). Carbohydrate supplementation during exercise. Journal ofnutrition 122, 788-795. COYLE, E. F., HAGBERG, J. M., HURLEY, B. F., MARTIN, W. H., EHSANI, A. A. & HOLLOSZY, J. 0. (1983). Carbohydrate feedings during prolonged strenuous exercise can delay fatigue. Journal of Applied Physiology 55, 230-235. COYLE, E. F., COGGAN, A. R., HEMMERT, M. K. & Ivy, J. L. (1986). Muscle glycogen utilisation during prolonged strenuous exercise when fed carbohydrate. Journal ofapplied Physiology 61, 165-172. DILL, D. B., EDWARDS, H. T. & TALBOTT, J. H. (1932). Studies in muscular activity. VII. Factors limiting the capacity for work. Journal of Physiology 77, 49-62. DILL, D. B. & COSTILL, D. L. (1974). Calculation of percentage changes in volumes of blood, plasma, and red cells in dehydration. Journal ofapplied Physiology 37, 247-248. ERICKSON, M. A., SCHWARZKOPF, R. J. & MCKENZIE, R. D. (1987). Effects of caffeine, fructose and glucose ingestion on muscle glycogen utilization during exercise. Medicine and Science in Sports and Exercise 19, 579-583. FEBBRAIO, M. A., MURTON, P., SELIG, S. E., CLARK, S. A., LAMBERT, D. L., ANGUS, D. J. & CAREY, M. F. (1996a). Effect of CHO ingestion on exercise metabolism and performance in different ambient temperatures. Medicine and Science in Sports and Exercise 28, 1380-1387. FEBBRAIO, M. A., PARKIN, J. M., BALDWIN, S., ZHAO, S. & CAREY, M. F. (1996b). Effect of ambient temperature on metabolic indices of fatigue during prolonged exercise. Medicine and Science in Sports and Exercise 28, S 180. GALLOWAY, S. D. R. & MAUGHAN, R. J. (1997). Effects of ambient temperature on the capacity to perform prolonged exercise in man. Medicine and Science in Sports and Exercise 29, 1240-1249. HARGREAVES, M. (199 1). Carbohydrates and exercise. Journal of Sports Sciences 9, 17-28. HARGREAVES, M., COSTILL, D. L., COGGAN, A. R., FINK, W. J. & NISHIBATA, I. (1984). Effect of carbohydrate feedings on muscle glycogen utilisation and exercise performance. Medicine and Science in Sports and Exercise 16, 219-222. MAUGHAN, R. J., BETHELL, L. R. & LEIPER, J. B. (1996). Effects of ingested fluids on exercise capacity and on cardiovascular and metabolic responses to prolonged exercise in man. Experimental Physiology 81, 847-859. MAUGHAN, R. J. & NOAKES, T. D. (1991). Fluid replacement and exercise stress: a brief review of studies on fluid replacement and some guidelines for the athlete. Sports Medicine 12, 16-31. MAUGHAN, R. J. (1991). Fluid and electrolyte loss and replacement in exercise. Journal of Sports Sciences 9, 117-142. MAUGHAN, R. J. (1982). A simple, rapid method for the determination of glucose, lactate, pyruvate, alanine, 3-hydroxybutyrate and acetoacetate on a single 20-,u1 blood sample. Clinical Chimica Acta 122, 231-240. MITCHELL, J. W., NADEL, E. R. & STOLWIJK, J. A. J. (1972). Respiratory weight losses during exercise. Journal ofapplied Physiology 32, 474-476. RAMANATHAN, N. L. (1964). A new weighting system for mean surface temperature of the human body. Journal ofapplied Physiology 19, 531-533. SAWKA, M. N., HUBBARD, R. W., FRANCESCONI, R. P. & HORSTMAN, D. H. (1983). Effects of acute plasma volume expansion on altering exercise-heat performance. European Journal of Applied Physiology 51,303-312.