Effects of exercise mode and intensity on postprandial thermogenesis in lean and obese men KAREN R. SEGAL, ALEXANDER CHUN, PILAR CORONEL, AND VERONICA VALDEZ Division of Pediatric Cardiology, Department of Pediatrics, Mount Sinai School of Medicine, New York, New York 10029 SEGAL,KARENR., ALEXANDERCHUN,PILARCORONEL,AND jects was greater when the meal was taken after a 30-min VERONICAVALDEZ. Effects of exercise mode and intensity on exercise bout than with no exercise and the meal-beforepostprandial thermog-enesis in lean and obese men. J. Appl. Physiol. 72(5): 1754-1763, 1992.-To characterize further the impact of exercise before a meal on thermogenesis, the effects of exercise intensity and mode and the duration of the effect of exercise on the thermic effect (TEF) of a 720-kcal mixed meal were compared in 10 lean and 10 obese men (16 ? 1 vs. 34 t 2% fat). In study A, TEF (kcal/3 h) was significantly greater for the lean than the obese men during rest and immediately after 1 h of cycling at 50 and 100 W. TEF was significantly greater after both exercise intensities than during rest for the obese men, but exercise had no effect on TEF in the lean men. In study B, TEF was significantly greater for the lean than the obese men during rest and immediately after 1 h of leg cycling at an 0, consumption of 1.09 l/min but only marginally different after 1 h of arm exercise at the same 0, consumption (P = 0.15). For the obese men, TEF was greater after arm than leg cycling and greater after leg cycling than at rest (P < O.Ol), but TEF was not different among the three conditions for the lean men. In study C, TEF was compared at rest and immediately and 24 h after 1 h of cycling at 100 W. TEF was greater for the lean than the obese men under all conditions (P < 0.05). For the obese but not the lean men, TEF was greater both immediately after and on the day after exercise than at rest (P < 0.01). Thus, acute exercise improves but does not normalize the blunted TEF in obesity; a minimally intense bout of exercise is needed to improve TEF; exercise mode alters thermogenesis in the obese men, even at a fixed intensity; and TEF in the obese men is enhanced for as long as 24 h after exercise. obesity; insulin resistance; indirect diture; euglycemic hyperinsulinemic
calorimetry; clamp
energy expen-
THERE IS CONSIDERABLECONFLICT regardingthether-
mic responses to various combinations of a meal and exercise. Some studies suggest that exercise enhances the thermic response to a meal in lean but not obese subjects, and others fail to confirm a potentiation of postprandial thermogenesis by exercise even in lean subjects (2,17,23, 31). These discordant findings may be due in part to differences in meal size an d composition , the timing of meal and exercise, intensity and duration of exercise , inadequate sample sizes, and confounding effects of imposing the thermic effects of food and exercise on diurnal changes in metabolic rate. Two studies demonstrated that in normal subjects the thermic effect of a meal was greater after exercise than at rest (17, 31), but these observations were not extended to obese subjects. We showed that the thermic effect of food for the obese sub1754
0161-7567/92 $2.00 Copyright
exercise sequence, although it was still blunted compared with the lean subjects (23). Although there is considerable evidence that exercise alters the thermic response to a subsequent meal, especially in obese subjects, characteristics of this exercise effect and the impact of factors such as the exercise intensity and the duration of the effect of exercise are unknown. In the present investigation, three models were used to compare the impact of exercise on thermogenesis in lean and obese individuals. In the first model, the effect of absolute work intensity was studied by comparing postprandial thermogenesis after leg cycling at two different work loads. Treadway and Young (26) reported that, in a heterogeneous group of five men and women ranging from 10 to 30% fat, the thermic effect of glucose was greater after high-intensity exercise than low, moderate, or no exercise. However, the impact of work intensity on the postexercise thermic response to a mixed meal has not been compared in lean and obese subjects. Variation in work intensity can be problematic, because as work intensity is manipulated, factors such as the total energy cost of exercise, degree of glycogen depletion, and body temperature are also affected, each of which may influence the thermic response to a subsequent meal. In the second model, the isolated effect of prior exercise intensity on the thermic response to a meal was assessed by comparing arm with leg ergometer exercise at the same 0, consumption (VO&: the former mode of exercise would elicit higher heart rate, blood pressure, plasma lactate concentration, and sympathetic stimulation, and thus, although absolute work rate is the same for the two modes of exercise, the relative stress is far greater for arm than leg work (20). Devlin and Horton (7) found that the thermic effect of infused glucose was increased on the day after glycogendepleting exercise, but the comparison of the immediate and longer-term effects of exercise has not been made. Furthermore the prolonged effect of exercise on the thermic response to a mixed meal has not been studied. Thus, in the third model, the immediate and longer-term (next day) effects of exercise on thermogenesis were compared. In summary, the aim of the present study was to assess quantitatively characteristics of the effect of physical exercise on thermogenesis in lean and obese subjects by I) comparing the impact of the same mode of exercise at two absolute intensities, 2) comparing two modes exer-
0 1992 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 12, 2019.
EFFECTS
OF
EXERCISE
cise that elicit the same energy cost but differ in relative intensity, and 3) determining the duration of the effect of an acute exercise bout on thermogenesis. Furthermore the relationship of the thermic response to a mixed meal to in vivo insulin action, determined by the euglycemic hyperinsulinemic clamp procedure, was examined.
ON
THERMOGENESIS
1755
Oral Glucose Tolerance Test
After an overnight (12-h) fast, a fasting blood sample was drawn, a 75-g glucose load was given, and venous blood samples were drawn at 30-min intervals for 2 h. The plasma was analyzed for glucose and insulin. The integrated areas under the glucose and insulin curves were calculated.
METHODS
Euglycemic Hyperinsulinemic Insulin Clamp
Subjects
Ten lean and 10 obese but otherwise healthy nonsmoking men, aged 25-35 yr, were recruited. All subjects were weight stable, had no personal or family history of diabetes, and were not taking any medications. The lean men were 48% fat and the obese group was ~28% fat by underwater weighing (see below). All subjects had normal glucose tolerance. The two groups were matched closely with respect to age, fat-free mass (FFM), and level of aerobic fitness. Men who engaged in regular aerobic exercise were not accepted into the study to eliminate level of cardiorespiratory fitness as a possible intervening variable. The subjects consumed a weight-maintenance diet containing 2300 g carbohydrate per day several days before and throughout the study. The protocol was approved by the Institutional Review Board of the Mount Sinai School of Medicine, and written informed consent was obtained from all subjects. Densitometry
Body fat content and FFM were determined by hydrodensitometry, according to the method described by Akers and Buskirk (1). Residual lung volume was estimated by means of the closed-circuit 0, dilution method of Wilmore (28). Percent body fat, fat weight, and fatfree weight were derived from body density. Aerobic Fitness Test
Maximal (peak VO,) and submaximal aerobic fitness were determined by a continuous multistage exercise test on an electromagnetically braked cycle ergometer. The subjects were familiarized with cycling exercise and all test procedures. The work rate was increased in 30-W increments every 2 min until volitional exhaustion was reached. Ventilatory measurements were made continuously by open-circuit respirometry with use of a Horizon Metabolic Measurement Cart (Sensormedics, Anaheim, CA). For each measurement, Vo2, CO, production (VCO,), minute ventilation (\73), and the ventilatory equivalent for 0, (vE/v02) were obtained. Criteria for an acceptable peak Vo2 were a peak heart rate within five beats of age-predicted maximum, a respiratory quotient (RQ = irco,liro,) > 1.15, and a leveling off of VO, such and last valthat the difference be tween the penultimate ues was 7O”C) to provide arterialized venous blood samples. Another catheter was placed in the antecubital vein of the other arm for infusion of glucose and insulin. A primed (25 &I) followed by continuous infusion (0.25 &ilmin) of [3-3H]glucose (New England Nuclear) was administered for 2 h to estimate hepatic glucose production in the basal state and for the following 2 h to determine hepatic glucose production during the clamp. Arterialized blood samples were obtained at 15min intervals for the first 90 min and every 5 min during the last 30 min of this control period to measure the plateau steady-state plasma glucose specific activity. After the 2-h control period, a primed continuous (40mU rnB2 . min-‘) infusion of regular human insulin (Humulin R, Eli Lilly, Indianapolis IN) was administered for 2 h to raise and maintain the plasma insulin at -100 pU/ml. The plasma glucose level was measured every 5 min and maintained at -90 mg/dl by adjusting the rate of infusion of a 20% dextrose in water solution with a servocontrol negative-feedback principle (6). Plasma [ 33H]glucose specific activity was measured at 15-min intervals for the first 90 min of the insulin infusion and at 5-min intervals during the last 30 min of the study. Indirect calorimetry was applied for the last 30 min of the baseline period and again during the last 30 min of the clamp to measure energy expenditure. l
Calculations
The rate of appearance (Ra) and disappearance (Rd) of glucose in the plasma was calculated from the [33H]glucose specific activity by use of the equations of Steele (24) in the steady-state form for the last 30 min of the basal period and in the non-steady-state form, as validated by Radziuk et al. (19), during the last 30 min of the clamp. The last 30 min of the euglycemic clamp study were used to calculate the rate of glucose disposal (Rd) by the entire body. The glucose distribution volume of 40 ml/kg body wt was assumed. The rate of total body glucose disposal during the clamp (Rd) was considered to represent the sum of the rates of glucose infusion,
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 12, 2019.
1756
EFFECTS
OF
EXERCISE
ON
THERMOGENESIS
corrected for the glucose space, and the residual hepatic glucose production, if any. The subjects voided before and after the test for measurement of urinary nitrogen. The nonprotein RQ was calculated from the respiratory exchange data and urinary nitrogen production rates. The rates of carbohydrate and lipid oxidation were calculated according to the procedure of Lusk (16), which is based on a nonprotein RQ of 0.707 for 100% fat oxidation and 1.000 for 100% carbohydrate oxidation. The thermic effect of glucose was estimated as the difference between the energy expenditure during the last 30 min of the baseline period and the last 30 min of the insulin/glucose infusion. Glucose storage was calculated as the difference between total glucose disposal and glucose oxidation.
a study duration of 3 h provides an unbiased estimate of the magnitude of the difference between the thermic response to a meal in lean and obese insulin-resistant subjects (22). Blood samples were drawn from an indwelling catheter for glucose, insulin, and lactate analysis. For the resting trials, blood samples were obtained at baseline and every 30 min for 3 h. For the exercise trials, blood samples were obtained at baseline, during the last minute of exercise, and every 30 min for 3 h. The integrated fasting and postprandial glucose and insulin areas were calculated. Blood samples were obtained at baseline and during the last minute of exercise for plasma lactate determination. Study B. The thermic effect of food at rest and after bouts of arm or leg cycling was determined by a protocol that consisted of six tests performed in random order on nonconsecutive days. Data for the first four treatments Thermogenesis Tests were taken from study A (not replicated): 1) PostabsorpRMR. 3) Postabsorptive, low Before the thermogenesis tests, the subjects visited the tive RMR. 2) Postprandial laboratory at least twice to become familiar with all proleg cycling. 4) Postprandial, low leg exercise. 5) Postabcedures. They reported to the laboratory at 8:00 A.M. in sorptive, arm exercise: Fasting metabolic rate was meathe postabsorptive state. Baseline metabolic rate was sured during 1 h of arm cycling at a VO, of -1 l/min and measured on each test day, after a 30-min rest period, to for 3 h after exercise. 6) Postprandial, arm exercise: The meal was given immediately after 1 h of arm cycling exerestablish the reliability of the measurement of resting cise, and postprandial metabolic rate was measured for 3 metabolic rate (RMR). We demonstrated that the measurement of the thermic effect of food in our laboratory is h after exercise. The absolute intensity of exercise was fixed at the highly reproducible, with a day-to-day intraindividual coefficient of variation
calorimetry; clamp
energy expen-
THERE IS CONSIDERABLECONFLICT regardingthether-
mic responses to various combinations of a meal and exercise. Some studies suggest that exercise enhances the thermic response to a meal in lean but not obese subjects, and others fail to confirm a potentiation of postprandial thermogenesis by exercise even in lean subjects (2,17,23, 31). These discordant findings may be due in part to differences in meal size an d composition , the timing of meal and exercise, intensity and duration of exercise , inadequate sample sizes, and confounding effects of imposing the thermic effects of food and exercise on diurnal changes in metabolic rate. Two studies demonstrated that in normal subjects the thermic effect of a meal was greater after exercise than at rest (17, 31), but these observations were not extended to obese subjects. We showed that the thermic effect of food for the obese sub1754
0161-7567/92 $2.00 Copyright
exercise sequence, although it was still blunted compared with the lean subjects (23). Although there is considerable evidence that exercise alters the thermic response to a subsequent meal, especially in obese subjects, characteristics of this exercise effect and the impact of factors such as the exercise intensity and the duration of the effect of exercise are unknown. In the present investigation, three models were used to compare the impact of exercise on thermogenesis in lean and obese individuals. In the first model, the effect of absolute work intensity was studied by comparing postprandial thermogenesis after leg cycling at two different work loads. Treadway and Young (26) reported that, in a heterogeneous group of five men and women ranging from 10 to 30% fat, the thermic effect of glucose was greater after high-intensity exercise than low, moderate, or no exercise. However, the impact of work intensity on the postexercise thermic response to a mixed meal has not been compared in lean and obese subjects. Variation in work intensity can be problematic, because as work intensity is manipulated, factors such as the total energy cost of exercise, degree of glycogen depletion, and body temperature are also affected, each of which may influence the thermic response to a subsequent meal. In the second model, the isolated effect of prior exercise intensity on the thermic response to a meal was assessed by comparing arm with leg ergometer exercise at the same 0, consumption (VO&: the former mode of exercise would elicit higher heart rate, blood pressure, plasma lactate concentration, and sympathetic stimulation, and thus, although absolute work rate is the same for the two modes of exercise, the relative stress is far greater for arm than leg work (20). Devlin and Horton (7) found that the thermic effect of infused glucose was increased on the day after glycogendepleting exercise, but the comparison of the immediate and longer-term effects of exercise has not been made. Furthermore the prolonged effect of exercise on the thermic response to a mixed meal has not been studied. Thus, in the third model, the immediate and longer-term (next day) effects of exercise on thermogenesis were compared. In summary, the aim of the present study was to assess quantitatively characteristics of the effect of physical exercise on thermogenesis in lean and obese subjects by I) comparing the impact of the same mode of exercise at two absolute intensities, 2) comparing two modes exer-
0 1992 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 12, 2019.
EFFECTS
OF
EXERCISE
cise that elicit the same energy cost but differ in relative intensity, and 3) determining the duration of the effect of an acute exercise bout on thermogenesis. Furthermore the relationship of the thermic response to a mixed meal to in vivo insulin action, determined by the euglycemic hyperinsulinemic clamp procedure, was examined.
ON
THERMOGENESIS
1755
Oral Glucose Tolerance Test
After an overnight (12-h) fast, a fasting blood sample was drawn, a 75-g glucose load was given, and venous blood samples were drawn at 30-min intervals for 2 h. The plasma was analyzed for glucose and insulin. The integrated areas under the glucose and insulin curves were calculated.
METHODS
Euglycemic Hyperinsulinemic Insulin Clamp
Subjects
Ten lean and 10 obese but otherwise healthy nonsmoking men, aged 25-35 yr, were recruited. All subjects were weight stable, had no personal or family history of diabetes, and were not taking any medications. The lean men were 48% fat and the obese group was ~28% fat by underwater weighing (see below). All subjects had normal glucose tolerance. The two groups were matched closely with respect to age, fat-free mass (FFM), and level of aerobic fitness. Men who engaged in regular aerobic exercise were not accepted into the study to eliminate level of cardiorespiratory fitness as a possible intervening variable. The subjects consumed a weight-maintenance diet containing 2300 g carbohydrate per day several days before and throughout the study. The protocol was approved by the Institutional Review Board of the Mount Sinai School of Medicine, and written informed consent was obtained from all subjects. Densitometry
Body fat content and FFM were determined by hydrodensitometry, according to the method described by Akers and Buskirk (1). Residual lung volume was estimated by means of the closed-circuit 0, dilution method of Wilmore (28). Percent body fat, fat weight, and fatfree weight were derived from body density. Aerobic Fitness Test
Maximal (peak VO,) and submaximal aerobic fitness were determined by a continuous multistage exercise test on an electromagnetically braked cycle ergometer. The subjects were familiarized with cycling exercise and all test procedures. The work rate was increased in 30-W increments every 2 min until volitional exhaustion was reached. Ventilatory measurements were made continuously by open-circuit respirometry with use of a Horizon Metabolic Measurement Cart (Sensormedics, Anaheim, CA). For each measurement, Vo2, CO, production (VCO,), minute ventilation (\73), and the ventilatory equivalent for 0, (vE/v02) were obtained. Criteria for an acceptable peak Vo2 were a peak heart rate within five beats of age-predicted maximum, a respiratory quotient (RQ = irco,liro,) > 1.15, and a leveling off of VO, such and last valthat the difference be tween the penultimate ues was 7O”C) to provide arterialized venous blood samples. Another catheter was placed in the antecubital vein of the other arm for infusion of glucose and insulin. A primed (25 &I) followed by continuous infusion (0.25 &ilmin) of [3-3H]glucose (New England Nuclear) was administered for 2 h to estimate hepatic glucose production in the basal state and for the following 2 h to determine hepatic glucose production during the clamp. Arterialized blood samples were obtained at 15min intervals for the first 90 min and every 5 min during the last 30 min of this control period to measure the plateau steady-state plasma glucose specific activity. After the 2-h control period, a primed continuous (40mU rnB2 . min-‘) infusion of regular human insulin (Humulin R, Eli Lilly, Indianapolis IN) was administered for 2 h to raise and maintain the plasma insulin at -100 pU/ml. The plasma glucose level was measured every 5 min and maintained at -90 mg/dl by adjusting the rate of infusion of a 20% dextrose in water solution with a servocontrol negative-feedback principle (6). Plasma [ 33H]glucose specific activity was measured at 15-min intervals for the first 90 min of the insulin infusion and at 5-min intervals during the last 30 min of the study. Indirect calorimetry was applied for the last 30 min of the baseline period and again during the last 30 min of the clamp to measure energy expenditure. l
Calculations
The rate of appearance (Ra) and disappearance (Rd) of glucose in the plasma was calculated from the [33H]glucose specific activity by use of the equations of Steele (24) in the steady-state form for the last 30 min of the basal period and in the non-steady-state form, as validated by Radziuk et al. (19), during the last 30 min of the clamp. The last 30 min of the euglycemic clamp study were used to calculate the rate of glucose disposal (Rd) by the entire body. The glucose distribution volume of 40 ml/kg body wt was assumed. The rate of total body glucose disposal during the clamp (Rd) was considered to represent the sum of the rates of glucose infusion,
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 12, 2019.
1756
EFFECTS
OF
EXERCISE
ON
THERMOGENESIS
corrected for the glucose space, and the residual hepatic glucose production, if any. The subjects voided before and after the test for measurement of urinary nitrogen. The nonprotein RQ was calculated from the respiratory exchange data and urinary nitrogen production rates. The rates of carbohydrate and lipid oxidation were calculated according to the procedure of Lusk (16), which is based on a nonprotein RQ of 0.707 for 100% fat oxidation and 1.000 for 100% carbohydrate oxidation. The thermic effect of glucose was estimated as the difference between the energy expenditure during the last 30 min of the baseline period and the last 30 min of the insulin/glucose infusion. Glucose storage was calculated as the difference between total glucose disposal and glucose oxidation.
a study duration of 3 h provides an unbiased estimate of the magnitude of the difference between the thermic response to a meal in lean and obese insulin-resistant subjects (22). Blood samples were drawn from an indwelling catheter for glucose, insulin, and lactate analysis. For the resting trials, blood samples were obtained at baseline and every 30 min for 3 h. For the exercise trials, blood samples were obtained at baseline, during the last minute of exercise, and every 30 min for 3 h. The integrated fasting and postprandial glucose and insulin areas were calculated. Blood samples were obtained at baseline and during the last minute of exercise for plasma lactate determination. Study B. The thermic effect of food at rest and after bouts of arm or leg cycling was determined by a protocol that consisted of six tests performed in random order on nonconsecutive days. Data for the first four treatments Thermogenesis Tests were taken from study A (not replicated): 1) PostabsorpRMR. 3) Postabsorptive, low Before the thermogenesis tests, the subjects visited the tive RMR. 2) Postprandial laboratory at least twice to become familiar with all proleg cycling. 4) Postprandial, low leg exercise. 5) Postabcedures. They reported to the laboratory at 8:00 A.M. in sorptive, arm exercise: Fasting metabolic rate was meathe postabsorptive state. Baseline metabolic rate was sured during 1 h of arm cycling at a VO, of -1 l/min and measured on each test day, after a 30-min rest period, to for 3 h after exercise. 6) Postprandial, arm exercise: The meal was given immediately after 1 h of arm cycling exerestablish the reliability of the measurement of resting cise, and postprandial metabolic rate was measured for 3 metabolic rate (RMR). We demonstrated that the measurement of the thermic effect of food in our laboratory is h after exercise. The absolute intensity of exercise was fixed at the highly reproducible, with a day-to-day intraindividual coefficient of variation
