- Email: [email protected]
Aerobic fitness and resting energy expenditure in young adult males
Aerobic Fitness and Resting Energy Expenditure Eric T. Poehlman,
Christopher
L. Melby,
Stephen
in Young Adult Males
F. Badylak,
and Jorge Calles
Ambiguous findings have been reported in previous studies regarding the relationships among aerobic fitness, resting metabolic rate (RMR), and the thermic effect of a meal (TEMj. We reexamined the association among these variables in young nonobese men who exhibited a wide range of aerobic fitness levels. RMR was measured after an overnight fast and TEM was assessed for three hours after ingestion of a liquid meal. Preprandial and postprandial plasma levels of insulin, glucose, and thyroid hormones (total T, and T,) were measured. Daily energy intake was estimated from three-day food diaries, body composition from underwater weighing, and aerobic fitness from a test of VD, max. Data were analyzed with linear and curvilinear regression analysis, as well as with ANDVA to test for differences among subjects classified by fitness level (ie, untrained, moderately, and highly trained). A significant correlation was found between RMR and VD, max (r = .77, P < .Ol). RMR adjusted for body weight and FFW was higher in highly trained men when compared to moderately and untrained individuals. However, a curvilinear relationship was found between TEM and VO, max (P < .05). Subjects who exhibited moderate levels of fitness showed the highest TEM, whereas a lower TEM was noted in untrained and highly trained men. These findings were observed in the absence of differences in plasma concentrations of total T, and T, among fitness levels. These findings suggest that highly trained men have a high RMR adjusted for their metabolic size. TEM varied in a curvilinear manner with fitness level, with the highest thermic response found in subjects with moderate levels of aerobic fitness. Thes,e data suggest that aerobic fitness in young. nonobese men may be a factor contributing to individual variation in RMR and postprandial thermogenesis. D 1989 by Grune & Stratton, inc.
I
N SEVERAL STUDIES differences have been found in resting metabolic rate (RMR)lm4 and the thermic and response to a meal (TEM)2.‘.5-” between exercise-trained untrained individuals. These findings imply that chronic endurance exercise may alter energy expenditure during nonactive times. However, the direction and magnitude of the relationship between RMR, TEM, and aerobic fitness is unclear. Some investigators have found a higher RMR in endurance-trained men,2*3.8 whereas others6~“*” have reported no significant difference between trained and untrained individuals. Studies that have examined the possible relationship between TEM and fitness level have also yielded ambiguous results. TEM has been found to increase with improved fitness,“,6 whereas other investigators have reported an inverse relationship.2.3,7.9-” The lack of concordant results among studies could be due to several methodological factors including the following: (1) failure to examine a wide range of fitness levels; most studies have classified subjects into two discreet fitness groups (ie, trained and untrained), (2) the use of different criteria to define trained and untrained individuals, and (3) insufficient sample sizes to detect differences in RMR and TEM among individuals varying in aerobic fitness. These considerations have prompted a more extensive investigation of the potential associations among RMR, TEM, and aerobic fitness in young, nonobese males who exhibited a wide range of fitness levels. Futhermore, we measured plasma levels of glucose, insulin, total thyroxine (T,) and triiodothyronine (T,) in the fed and fasted states to identify possible relationships between hormonal and substrate responses with variations in RMR and TEM and aerobic fitness. MATERIALS
AND METHODS
Subjects
Twenty-eight nonobese written
consent
to participate
men 19 to 36 years of age gave their in the study. Subjects were recruited
Metabolism, Vol38, No 1 (JanuaryI. 1989: pp 85-90
from the university community. It was our intent to recruit individuals with a wide range of physical activity levels and maximal aerobic capacities (range of 40 mL - kg-’ - min’ to 80 mL . kg-’ . min-‘). Subjects were classified into fitness levels (ie, untrained, moderately, and highly trained) after a test of maximal aerobic capacity (ie, ‘v’0, max). Highly trained men were competitive runners who reported a training distance of 80 to 120 km/wk and a training frequency of 6 to 7 times per week. Moderately trained men were recreational athletes who reported running 3 to 4 times per week and frequent participation in sporting activities. Untrained men reported no regular participation in any physical activity or sport. All subjects were nonsmokers and had been weight-stable for at least 6 months prior to the study. Subjects were instructed to maintain their normal dietary habits, which included a minimum of 250 g of carbohydrate per day for three days prior to testing. The protocol was approved by the Institutional Review Board of Purdue University.
Body Composition,
VO, max and Energy
intake
Body fat was estimated from body density by underwater weighing using the Siri equation. ‘* Residual lung volume was assessed by the method of Wilmore et al.” Fat-free weight (FFW) was estimated as total body weight minus fat weight. Maximal aerobic capacity (ir0, max) was assessed by a progressive and continuous test to exhaustion on a motor-driven treadmill.’ Daily energy and macronutrient intake was estimated from a three-day food diary (two weekdays, one weekend day).14
From the Division of Endocrinology, Metabolism, and Nutrition. Department of Medicine, University of Vermont, Burlington. VT and the Exercise Physiology Laboratory and Hillenbrand Biomedical Engineering Center, Purdue University, West Lafayette, IN. Dr. Poehlman is supported in part by NIH Gram No. AGO7857 and GCRC Grant No. RR1 09. Address reprint requests to E.T. Poehlman. PhD, Endocrinology, Metabolism, and Nutrition, Department of Medicine, University of Vermont, Burlington, VT 05405. o 1989 by Grune & Stratton, Inc. 00260495/89/3801-0014$03.00/0
85
86
POEHLMAN ET AL
Resting Metabolic Rate and Thermic Effect of a Meal At least 24 hours after the last exercise session, RMR and TEM were determined for each subject by indirect calorimetry using a ventilated hood system (SensorMedics; Anaheim, CA).’ Briefly, each subject was acquainted with the hoed system one day before measuring metabolic rate. At 6:30 AM the next day after a 12-hour fast, the subject assumed a supine position and a plastic catheter was inserted into an antecubital vein for blood sample collection. The Plexiglas ventilated hood was placed over the subject’s head and a 30-minute habituation period was allowed. Baseline RMR was then measured for 30 minutes. A blood sample was obtained at time 0 minutes. Thereafter, the hood was removed and the subjects consumed a liquid meal (Sustacal; Mead Johnson; Evansville, IN) with an energy content of 10 kcal . kg FFW-’ and a nutrient composition of 24% protein, 55% carbohydrate, and 21% fat. Each subject consumed the liquid meal within 10 minutes, after which he was allowed to void if necessary. The subject was then returned to the hood and all measures were continued. TEM was measured continuously for the next 180 minutes and the results averaged for 30minute time periods. The laboratory was maintained at 24OCduring the measurement period. Blood sample collection was repeated every 30 minutes for measurement of plasma levels of glucose and insulin. Blood samples for total plasma T, and T, concentrations were collected before the meal (time 0 minutes) and 90 and 180 minutes postprandially. Blood pressure was taken at 30-minute intervals using a standard cuff and sphygmomanometer. Heart rate was monitored by continuous electrocardiography during the test. Energy expenditure was calculated using the Weir formula.” The reliability of RMR measurement under our laboratory conditions has been previously reported.’ Plasma Determinations Plasma samples were assayed for insulin, total T,, and T, using radioimmunoassay (RIA) kits (Diagnostic Products Corporation, Los Angeles). Plasma glucose was assessed with the glucose oxidase technique, using Reagent Set (Bio-dynamics/bmc, Indianapolis). Statistical Analysis Linear and nonlinear regression analysis was first used to assess relationships among RMR, TEM, ir0, max, energy intake, and plasma hormones. I6Data were also analyzed with subjects classified into highly trained, moderately trained, and untrained levels, based on commonly accepted classifications of 90, max.” Differences among fitness levels were analyzed by ANOVA. When the F ratio was significant at P < .05, multiple comparisons among group means were evaluated by a Duncan multiple range test corrected for unequal subject number in the groups. The mean of the postprandial Table 1. Physical Characteristics Untrained In - 9)
Characteristic Age tyr)
22.9
Ht (m)
* 1.7
1.8 + 0.01
of Untrained,
Fig 1. The relationship between capacity fV0, max) in 29 males.
RMR and maximal
aerobic
measurements was used to represent the overall effect of feeding. All values are reported as mean f SEM. RESULTS
Resting Metabolic Rate
Physical characteristics are shown in Table 1. Highly trained men had a lower body weight (P < .05) than the untrained group. Percent body fat was lower (P-C .Ol) in highly trained men than the other two groups. No differences in age, height, body mass index (wt/h*), and FFW were found among groups. As illustrated in Fig 1, a significant correlation (P < .Ol) was noted between 60, max and RMR . kg-’ - h-‘. This persisted when RMR and VO, max were adjusted for FFW (r = S5, P < .Ol; not shown in figure). In Table 2 with subjects classified into fitness levels, the highest RMR was observed in highly trained as compared to moderately trained and untrained men. Thermic Effect of a Meal
Using curvilinear regression, data were analyzed as continuous variables to test for a nonlinear relationship between 90, max and TEM (Fig 2). A significant curvilinear relationship was found (P -C .05), indicating that the higher TEM was noted in the mid-range of 90, max values, whereas a lower TEM was observed in the extremes of (low Moderately
Trained, and Highly Trained Men
Moderatelytrained In- 11) 22.7
+ 0.8
1.8 + 0.02
Highlytrained (n - 8) 26.3
+ 2.3
1.8 f 0.02
Wt (kg)
79.8
zt 3.2
74.1
f. 2.2
70.8
+ 1.5
Body mass index (wt/ht’)
24.1
f 0.8
23.4
+ 0.8
21.9
+ 0.4
% body fat
14.8 f 1.5
13.6 f 1.3
FFW kg)
68.1
+ 3.0
64.0
f 1.8
65.9
45.6
+ 0.7
53.3
+ 0.7
70.9 * 2.1
6.9 + 0.8
P
NS NS H < U; <.05 NS H < U; <.Ol H < M; <.Ol
\iO, max (mL
.
kg-’
.
min-‘)
+ 1.4
NS
Abbreviations: U, untrained; M, moderately trained; H, highly trained; NS, not significant. Values are the mean + SEM. Differences between \iO, max among groups were not statistically analyzed because it was used to classify fitness level of subjects.
a7
RESTING ENERGY EXPENDITURE AND AEROBIC FITNESS
Table 2. Resting Metabolic
Rate of Untrained,
Untrained Variable
RMR (kcal RMR (kcal
- min-‘l . kg-’ .
In -
h-‘)
9)
Moderately
Moderately (n-
Trained 11)
Trained,
and Highly Trained Men Highly Trained (n -
1.24 + 0.05
1.13 f 0.03
1.31 * 0.05
0.93
0.92
1.11 2 0.03
f 0.02
* 0.02
P
9)
H > M: <.Ol H > U: <.Ol H > M; <.05
RMR (kcal
.
FFW-’
.
h-‘1
1.09 * 0.03
1.06 f 0.03
1.20 * 0.04
t-l z u; <.05 H > M; ~05
Values are the mean + SEM.
(0.97 + 0.05) when compared to the untrained (0.83 + 0.04, P < .05) and highly trained (0.77 k 0.04, P < .Ol) men. Hormones, Substrates, and Cardiovascular Parameters
I
.
Fig 2. The curvilinear relationship between $0, max and TEM in 28 male nonobese subjects. A significant curvilinear relationship was found in which the equation is: TEM (kcal - kg-’ - 180 min.‘) = -0.909 - 0.065 (VO, max. mL - kg-’ - min-‘) O.ooO5881 (VO, max mL . kg-’ . min-‘1’ (P c .OSl.
and high) QO, max. A similar relationship was obtained when TEM was adjusted for FFW. No significant association between VO, max and TEM (r = - .23) was noted (not shown in figure form). The time course of changes of TEM in untrained, moderately trained, and highly trained subjects are illustrated in Fig 3. After meal consumption an elevation of metabolic rate was noted in all subjects. Total postprandial energy expenditure (TEM) varied according to the level of aerobic fitness (inset Fig 3). The postprandial response (kcal . kg-’ . 180 min-‘) was higher in the moderately trained group
Preprandial and postprandial plasma insulin, glucose, and heart rate responses are shown in Table 3. As expected, fasting and postprandial plasma insulin levels were lower in highly trained as compared to moderately and untrained men. In spite of differences in plasma insulin, no differences in plasma glucose were noted among the three groups before or after meal consumption. Identical results were found when insulin and glucose responses were analyzed by surface area under the response curves. Preprandial and postprandial heart rate was lower in the highly trained as compared with the other two groups. No differences were noted among groups in the preprandial and postprandial blood pressure and respiratory quotient (RQ). Plasma T, (ng/lOO mL) concentrations at time 0 minutes (ie, before meal ingestion) were not different among untrained (117 k 5.3), moderately trained (117 k 5.3), and highly trained men (118 + 10.4). Plasma T, (pg/lOO mL) concentrations were also not different at time 0 minutes in untrained (6.5 -L0.3), moderately trained (6.4 2 0.5), and highly trained men (5.6 2 0.3). No significant changes were noted in plasma T, or T, after meal ingestion. As shown in Table 4, daily energy intake was higher in the highly trained men in comparison with the other two groups. Highly trained men consumed a higher intake (P < .05) of protein than moderately trained and untrained men. The percent contribution of the macronutrients to daily energy intake was not different among the three groups.
10 t
Fig 3. Increase in poatprandial energy expenditure (TEMI for untrained, moderately trained, and highly trained male subjects after ingestion of a liquid meal.
T
88
POEHLMAN ET AL
Table 3. Plasma Insulin, Plasma Glucose, Heart Rate, and Respiratory Quotient (RQ) Before and After Meal Consumption in Untrained. Moderatelv Trained. and Hiahlv Trained Men Untrained (n - 9)
Variable Insulin (0 min) (/dJ/mL)
25.2
ModeratelyTrained (n- 111
+ 3.4
25.1
HighlyTrained (n - 6)
+ 3.6
14.2 ? 3.0
P
U > H: <.05 M > H; 1.05
Insulin (30-180
min) (pU/mL)
Glucose (0 mini (mg/lOO mL) Glucose (30-l 80 min) (mg/lOO mL)
86.6
f 7.4
67.4
+ 6.9
50.2
* 8.3
90.4
f 2.1
91.4
+ 2.4
88.7
+ 2.4
101.4
f 3.4
97.3
f 2.8
93.1
+ 2.5
Heart rate (0 min) (beats/min)
61 f 2.4
Heart rata (30-l 80 min)
70 i
55 * 2.1
1.6
65 + 2.9
U > H: <.Ol NS NS
51 * 2.9
U > H; <.Ol
54 ? 2.6
M > H; <.Ol
RQ (0 min)
0.81
+ .02
0.80
f .Ol
0.81
+ .02
NS
RQ (30-l 80 min)
0.89
f .Ol
0.90
f .Ol
0.87
+ .Ol
NS
All values are the mean + SEM; 0 min = time before meal ingestion: 30-l 80 min = mean postprandial response.
Significant linear correlations were noted between RMR . kg-’ . h-’ and daily energy intake (r = S4, P < .Ol) and between RMR a kg-’ - h-’ and basal levels of total plasma T, (r = .37, P -c .05; not shown in figure form).
trained men.3 However, LeBlanc et al did not find an elevated RMR in highly trained females as compared with moderately trained and untrained women.9 It is possible that RMR response to exercise training is more sensitive in males who have a higher proportion of FFW to total body weight. Hill et al observed a -9% higher RMR per kg FFW in trained v untrained men.6 This difference was similar to that observed in the present study, but a small sample size in their study probably precluded statistical significance. Thus, the highly trained male may achieve high levels of energy expenditure by two ways: (1) the direct energy cost of aerobic exercise and (2) a high RMR. In the present study it is unlikely that the high RMR in highly trained men is due to the prolonged effects of the last exercise bout. At least 24 hours separated the last exercise session and RMR measurement. Bahr et al” reported no prolonged effect of heavy exercise (80 minutes, 70% ir0, max) on basal oxygen consumption 24 hours postexercise in male subjects. Hill et al6 noted a higher RMR in trained men 36 hours postexercise, which along with our data, suggests that the elevated RMR in highly trained men may be a chronic adaptation to the exercise-trained state. The mechanism for the high RMR in highly trained men remains to be established. However, preliminary evidence suggests that RMR could be affected when the caloric turnover is high, concurrent with the maintenance of a state of energy balance. This would be accomplished by matching a high level of food intake with a high level of exerciseinduced energy expenditure, which is characteristic of the highly trained men in this study. The association among energy intake, energy expenditure (90, max), and RMR in
DISCUSSION
Resting Metabolic Rate
The potential relationships among fitness level, RMR, and TEM were examined in young males who exhibited a wide range of aerobic fitness levels. An elevated RMR (-10%) was found at high fitness levels when adjusted for differences in body weight and FFW. Hence, it appears that the highly trained male has a high RMR for his metabolic size. This is consistent with previous reports*’ but in apparent contradiction with others.5-7.9The discrepancy between our results with those of previous studies may be partially attributed to two factors: (1) differences in the definition of the highly trained individual among investigators and (2) smaller sample sizes in previous studies. Concerning the former point, other investigators who have not observed a high RMR in trained individuals have examined subjects with an aerobic capacity approaching -60 mL - kg-’ - min-‘.5-7S9Present results and those of others***support a high RMR in male subjects who engage in heavy exercise and achieve very high maximal aerobic capacities (-70 ml - kg-’ - min-‘). This may suggest that very high levels of aerobic fitness must be achieved before they impact upon RMR. This notion is consistent with the results of Tremblay et al who found a high RMR . kg FFW-’ in highly trained men by subdividing their trained group into highly trained and moderately
Table 4. Nutrient Intakes as Determined From 3-Day Food Diaries for Untrained, Moderately Trained, and Highly Trained Men Untrained Nutrient
(n -
9)
Moderately In-
Trained 11)
Highly
Trained
(n = 9)
P
Energy intake (kcal/d)
3036
f 214
3153
* 191
4688
r 448
H > M; <.Ol
Protein (g/d)
119.0
+ 12.7
121.8
+ 7.7
167.0
+ 21.3
H > M; <.05
(15.6
f 1.8)
(15.9
* 1.3)
(16.1
+ 1.9)
NS
419.0
+ 48.8
388.5
+ 47.4
572.0
f 93
NS
(54.7
* 4.3)
(48.6
+ 4.3)
(52.9
* 4.0)
NS
103.1
+ 10.0
110.5
* 16.6
152.3
+ 27.8
NS
(30.8
+ 3.01)
(31.2
f 4.2)
(31.1
f 3.3)
NS
H > L; <.Ol H > L; <.05 Carbohydrates (g/d)
Fat (g/d)
Values are the mean + SEM. Values in parentheses indicate percentage contribution of the macronutrient to total daily intake.
RESTING ENERGY EXPENDITURE
AND AEROBIC
FITNESS
this study is supported by correlations between RMR and energy intake (r = .54, P -C .Ol) and between RMR and VO, max (r = .77, P < .Ol; Fig 1). Relevant to this discussion are the findings of Woo et al who created a high caloric turnover condition by overfeeding six males, but they maintained a state of energy balance by a proportionate increase in aerobic exercise.” RMR was higher after exercise training, but no changes were found in plasma thyroid hormones. This is consistent with our findings in which no differences in thyroid hormones were observed between individuals varying in aerobic fitness, but an elevated RMR was noted in the highly trained men. Thermic Effect of a Meal
In the present study we found that TEM varied with fitness level in a curvilinear manner, ie, moderate levels of aerobic fitness were associated with the higher TEM than the extremes of low and high fitness. This finding is supported by two analyses: (1) linear and nonlinear regression analysis to examine the relationships between TEM and VO, max without prior classification, and (2) difference testing with subjects classified into commonly accepted fitness levels using ANOVA. This approach reduced the possibility of identifying misleading physiologic relationships from only evaluating subjects classified into discreet fitness levels. Linear regression analysis revealed a low order correlation between VO, max and TEM (r = .23, NS), whereas by using nonlinear regression we found a significant curvilinear relationship (Fig 2). This provides evidence that moderate levels of fitness were associated with the highest TEM, whereas a more efficient postprandial response (ie, lower TEM) was observed at the extremes of the fitness continuum. This observation may partially account for the linear relationship between VO? max and TEM reported in previous studies when, in fact, trained subjects were similar in aerobic fitness to the moderately trained, high TEM responders in this study. 5.6When the highly trained group was excluded from our data analysis, a linear correlation was observed (r = .47, P C: .OS). Thus, the positive association between VO, max and TEM observed in the untrained and moderately trained subjects is similar to previous investigationssJ’ that examined fewer subjects and a small range of fitness levels. Thus, we would propose that a curvilinear relationship between VO, max and TEM would be only observed in an experimental design, which examined a wider range of fitness levels and a larger sample size. The enhanced energy expenditure due to a high TEM could partially account for the observation that some individuals undergoing exercise programs lose more weight than expected from the calories expended during the exercise.*‘,” We also examined the differences between TEM and VO, max by classifying subjects into commonly accepted aerobic fitness categories (Fig 3), as previously performed.3~6~7-” TEM was higher in the moderately trained as compared to umrained and highly trained individuals. This supports the notion that moderate levels of exercise training may enhance postprandial energy expenditure. The hypothesis of a curvilinear relationship between aer-
89
obic fitness and TEM has been suggested,** and the present study, to our acknowledge, provides the first evidence of this relationship. It is unknown whether similar hormonal and/or substrate mechanisms regulate the blunted TEM in untrained and highly trained men. A lower TEM has been found in highly trained men and women2,7*‘o*”and in nonobese, untrained men.6 Although it cannot be directly determined from our data, the low TEM in untrained men may be associated with a lower rate of insulin-mediated glucose uptake and lower rates of glucose storage (ie, obligatory thermogenesis). This is suggested by their higher postprandial levels of plasma insulin. This explanation seems unlikely to account for the low TEM in highly trained men who are noted for their enhanced insulin sensitivity. An alternative possibility may be found in the sympathetic activation of thermogenesis. Sympathetic activity has been shown to be positively correlated to the size of TEM.23 Low levels of plasma catecholamines have been reported in highly trained men, which is consistent with their low TEM.’ We noted a lower postprandial heart rate in highly trained men, which indirectly supports the lower sympathetic tone in this group as compared to other individuals. One may consider if differences in the phenotypic expression of RMR and TEM among individuals are the result of training and/or genetic influences. We reported a large variation in changes of RMR and TEM to short-term exercise training among six pairs of monozygotic twins, whereas a coherent response pattern was noted within twin pairs. This suggested that biologic variation in RMR and TEM in response to exercise training is partially influenced by the genetic background.24 However, it is premature to draw conclusions regarding the degree of heritability of RMR and TEM in response to exercise. The aforementioned experiment was not designed to assess the contribution of heredity to variation in RMR and TEM in the population but rather to explore the phenomenon of the genotype-environment interaction in man. Furthermore, biologic traits (ie, FFW) that are known to covary with the phenotypic expression of RMR and TEM have also exhibited a genotype dependency in response to chronic exercise.25 Our studies point out the need to prospectively examine subjects undergoing long-term aerobic training, resulting in large changes in fitness to critically examine the association between changes in aerobic fitness and thermogenesis. In summary, our data suggest that differences in fitness level are associated with variations in resting energy expenditure. High levels of aerobic fitness are associated with an elevated RMR, whereas moderate levels of fitness are related to an enhanced postprandial thermogenesis. ACKNOWLEDGMENT
The authors thank Dr. G. Bottoms, Hormonal
Assay Laboratory, Purdue University, for his cooperation in the hormonal analysis, and to Drs. E. Danforth, Jr, and E.A.H. Sims for their constructive criticisms of the manuscript. REFERENCES 1. Lennon DF, Nagle F, Stratman F, et al: Diet and exercise training effects on resting metabolic rate. Int J Obes 9:39-47, 1985
90
2. Poehlman ET, Melby CL, Badylak SF: Resting metabolic rate and postprandial thermogenesis in highly trained and untrained males. Am J Clin Nutr 47:793-798,1988 3. Tremblay A, Fontaine E, Nadeau A: Contribution of postexercise increment in glucose storage to variations in glucose-induced thermogenesis in endurance athletes. Can J Physiol Pharmacol 63:1165-1169,1985 4. Tremblay A, Fontaine E, Poehlman ET, et al: The effect of exercise-training on resting metabolic rate in lean and moderately obese individuals. Int J Obes 10:51 l-517,1986 5. Davis JR, Tagliaferro AR, Kertzer R, et al: Variations in dietary-induced thermogenesis and body fatness with aerobic capacity. Eur J Appl Physiol50:319-329, 1983 6. Hill JO, Heymsfield SB, McMannus C, et al: Meal size and thermic response to food in male subjects as a function of maximum aerobic capacity. Metabolism 33:743-749, 1984 7. LeBlanc J, Diamond P, C&e J, et al: Hormonal factors in reduced postprandial heat production of exercise-trained subjects. J Appl Physiol56:772-776, 1984 8. LeBlanc J, Jobin M, C&C J, et al: Enhanced metabolic response to caffeine in exercise-trained human subjects. J Appl Physiol59:832-837, 1985 9. LeBlanc J, Mercier P, Samson P: Diet-induced thermogenesis with relation to training state in female subjects. Can J Physiol Pharmacol62:334-337, 1984 10. Poehlman ET, Desprds JP, Bessette H, et al: Influence of caffeine on the resting metabolic rate of exercise-trained and inactive subjects. Med Sci Sports Exert 17:689-694, 1985 11. Tremblay A, C&t J, LeBlanc J: Diminished dietary thermogenesis in exercise-trained human subjects. Eur J Appl Physiol 52:1-4, 1983 12. Siri WE: The gross composition of the body. Adv Biol Med Phys 4:239-280, 1956 13. Wilmore JH, Vodak PA, Parr RB, et al: Further simplifica-
POEHLMAN ET AL
tion of a model for determination of residual long volume. Med Sci Sports Exert 12:216-218, 1980 14. Poehlman ET, Arciero PJ, Melby CL, et al: Resting metabolic rate and postprandial thermogenesis in vegetarians and nonvegetarians. Am J Clin Nutr 48:209-213, 1988 15. Weir JB de V: New methods for calculating metabolic rate with special reference to protein. J Physiol (Lond) 109:1-9, 1949 16. Snedecor GW, Cochran WG: Statistical Methods (ed 6). Ames, IA, Iowa State University Press, 1973, pp 447-459 17. Sinning WE, Experiments and demonstrations in exercise physiology. Philadelphia, Saunders, 1975 18. Bahr R, Ingnes I, Vaage 0, et al: Effect of duration of exercise on excess postexercise 0, consumption. J Appl Physiol 62:485-490, 1987 19. Woo R, O’Connell M, Horton ES, et al: Changes in resting metabolism with increased intake and exercise. Clin Res 33:712A, 1985 (abstr) 20. Bjiirntorp P, Berchtold P, Lindholm B, et al: Effects of physical training on glucose tolerance, plasma insulin, and lipids and on bcdy composition in men after myocardial infarction. Acta Med Stand 192:439-443,1972 21. Oscai LB, Williams BT: Effect of exercise on overweight middle-aged males. J Am Geriatr Sot 16:794-797, 1968 22. Segal KR, Xavier Pi-Sunyer F: Exercise, resting metabolic rate, and thermogenesis, in DeFronzo RA (ed): Diabetes/Metabolism Reviews, Vo12, New York, Wiley, 1986, pp 20-34 23. Schwartz RS, Jaeger LF, Silberstein S, et al: Sympathetic nervous system activity and the thermic effect of feeding in man. Int J Obes 11:141-149, 1987 24. Poehlman ET, Tremblay A, Nadeau A, et al: Heredity and changes in hormones and metabolic rates with short-term training. Am J Physiol250:E711-E717,1986 25. Poehlman ET, Tremblay A, Marcotte M, et al: Heredity and changes in body composition and adipose tissue metabolism after short-term exercise training. Eur J Appl Physiol56:398-402, 1987
Christopher
L. Melby,
Stephen
in Young Adult Males
F. Badylak,
and Jorge Calles
Ambiguous findings have been reported in previous studies regarding the relationships among aerobic fitness, resting metabolic rate (RMR), and the thermic effect of a meal (TEMj. We reexamined the association among these variables in young nonobese men who exhibited a wide range of aerobic fitness levels. RMR was measured after an overnight fast and TEM was assessed for three hours after ingestion of a liquid meal. Preprandial and postprandial plasma levels of insulin, glucose, and thyroid hormones (total T, and T,) were measured. Daily energy intake was estimated from three-day food diaries, body composition from underwater weighing, and aerobic fitness from a test of VD, max. Data were analyzed with linear and curvilinear regression analysis, as well as with ANDVA to test for differences among subjects classified by fitness level (ie, untrained, moderately, and highly trained). A significant correlation was found between RMR and VD, max (r = .77, P < .Ol). RMR adjusted for body weight and FFW was higher in highly trained men when compared to moderately and untrained individuals. However, a curvilinear relationship was found between TEM and VO, max (P < .05). Subjects who exhibited moderate levels of fitness showed the highest TEM, whereas a lower TEM was noted in untrained and highly trained men. These findings were observed in the absence of differences in plasma concentrations of total T, and T, among fitness levels. These findings suggest that highly trained men have a high RMR adjusted for their metabolic size. TEM varied in a curvilinear manner with fitness level, with the highest thermic response found in subjects with moderate levels of aerobic fitness. Thes,e data suggest that aerobic fitness in young. nonobese men may be a factor contributing to individual variation in RMR and postprandial thermogenesis. D 1989 by Grune & Stratton, inc.
I
N SEVERAL STUDIES differences have been found in resting metabolic rate (RMR)lm4 and the thermic and response to a meal (TEM)2.‘.5-” between exercise-trained untrained individuals. These findings imply that chronic endurance exercise may alter energy expenditure during nonactive times. However, the direction and magnitude of the relationship between RMR, TEM, and aerobic fitness is unclear. Some investigators have found a higher RMR in endurance-trained men,2*3.8 whereas others6~“*” have reported no significant difference between trained and untrained individuals. Studies that have examined the possible relationship between TEM and fitness level have also yielded ambiguous results. TEM has been found to increase with improved fitness,“,6 whereas other investigators have reported an inverse relationship.2.3,7.9-” The lack of concordant results among studies could be due to several methodological factors including the following: (1) failure to examine a wide range of fitness levels; most studies have classified subjects into two discreet fitness groups (ie, trained and untrained), (2) the use of different criteria to define trained and untrained individuals, and (3) insufficient sample sizes to detect differences in RMR and TEM among individuals varying in aerobic fitness. These considerations have prompted a more extensive investigation of the potential associations among RMR, TEM, and aerobic fitness in young, nonobese males who exhibited a wide range of fitness levels. Futhermore, we measured plasma levels of glucose, insulin, total thyroxine (T,) and triiodothyronine (T,) in the fed and fasted states to identify possible relationships between hormonal and substrate responses with variations in RMR and TEM and aerobic fitness. MATERIALS
AND METHODS
Subjects
Twenty-eight nonobese written
consent
to participate
men 19 to 36 years of age gave their in the study. Subjects were recruited
Metabolism, Vol38, No 1 (JanuaryI. 1989: pp 85-90
from the university community. It was our intent to recruit individuals with a wide range of physical activity levels and maximal aerobic capacities (range of 40 mL - kg-’ - min’ to 80 mL . kg-’ . min-‘). Subjects were classified into fitness levels (ie, untrained, moderately, and highly trained) after a test of maximal aerobic capacity (ie, ‘v’0, max). Highly trained men were competitive runners who reported a training distance of 80 to 120 km/wk and a training frequency of 6 to 7 times per week. Moderately trained men were recreational athletes who reported running 3 to 4 times per week and frequent participation in sporting activities. Untrained men reported no regular participation in any physical activity or sport. All subjects were nonsmokers and had been weight-stable for at least 6 months prior to the study. Subjects were instructed to maintain their normal dietary habits, which included a minimum of 250 g of carbohydrate per day for three days prior to testing. The protocol was approved by the Institutional Review Board of Purdue University.
Body Composition,
VO, max and Energy
intake
Body fat was estimated from body density by underwater weighing using the Siri equation. ‘* Residual lung volume was assessed by the method of Wilmore et al.” Fat-free weight (FFW) was estimated as total body weight minus fat weight. Maximal aerobic capacity (ir0, max) was assessed by a progressive and continuous test to exhaustion on a motor-driven treadmill.’ Daily energy and macronutrient intake was estimated from a three-day food diary (two weekdays, one weekend day).14
From the Division of Endocrinology, Metabolism, and Nutrition. Department of Medicine, University of Vermont, Burlington. VT and the Exercise Physiology Laboratory and Hillenbrand Biomedical Engineering Center, Purdue University, West Lafayette, IN. Dr. Poehlman is supported in part by NIH Gram No. AGO7857 and GCRC Grant No. RR1 09. Address reprint requests to E.T. Poehlman. PhD, Endocrinology, Metabolism, and Nutrition, Department of Medicine, University of Vermont, Burlington, VT 05405. o 1989 by Grune & Stratton, Inc. 00260495/89/3801-0014$03.00/0
85
86
POEHLMAN ET AL
Resting Metabolic Rate and Thermic Effect of a Meal At least 24 hours after the last exercise session, RMR and TEM were determined for each subject by indirect calorimetry using a ventilated hood system (SensorMedics; Anaheim, CA).’ Briefly, each subject was acquainted with the hoed system one day before measuring metabolic rate. At 6:30 AM the next day after a 12-hour fast, the subject assumed a supine position and a plastic catheter was inserted into an antecubital vein for blood sample collection. The Plexiglas ventilated hood was placed over the subject’s head and a 30-minute habituation period was allowed. Baseline RMR was then measured for 30 minutes. A blood sample was obtained at time 0 minutes. Thereafter, the hood was removed and the subjects consumed a liquid meal (Sustacal; Mead Johnson; Evansville, IN) with an energy content of 10 kcal . kg FFW-’ and a nutrient composition of 24% protein, 55% carbohydrate, and 21% fat. Each subject consumed the liquid meal within 10 minutes, after which he was allowed to void if necessary. The subject was then returned to the hood and all measures were continued. TEM was measured continuously for the next 180 minutes and the results averaged for 30minute time periods. The laboratory was maintained at 24OCduring the measurement period. Blood sample collection was repeated every 30 minutes for measurement of plasma levels of glucose and insulin. Blood samples for total plasma T, and T, concentrations were collected before the meal (time 0 minutes) and 90 and 180 minutes postprandially. Blood pressure was taken at 30-minute intervals using a standard cuff and sphygmomanometer. Heart rate was monitored by continuous electrocardiography during the test. Energy expenditure was calculated using the Weir formula.” The reliability of RMR measurement under our laboratory conditions has been previously reported.’ Plasma Determinations Plasma samples were assayed for insulin, total T,, and T, using radioimmunoassay (RIA) kits (Diagnostic Products Corporation, Los Angeles). Plasma glucose was assessed with the glucose oxidase technique, using Reagent Set (Bio-dynamics/bmc, Indianapolis). Statistical Analysis Linear and nonlinear regression analysis was first used to assess relationships among RMR, TEM, ir0, max, energy intake, and plasma hormones. I6Data were also analyzed with subjects classified into highly trained, moderately trained, and untrained levels, based on commonly accepted classifications of 90, max.” Differences among fitness levels were analyzed by ANOVA. When the F ratio was significant at P < .05, multiple comparisons among group means were evaluated by a Duncan multiple range test corrected for unequal subject number in the groups. The mean of the postprandial Table 1. Physical Characteristics Untrained In - 9)
Characteristic Age tyr)
22.9
Ht (m)
* 1.7
1.8 + 0.01
of Untrained,
Fig 1. The relationship between capacity fV0, max) in 29 males.
RMR and maximal
aerobic
measurements was used to represent the overall effect of feeding. All values are reported as mean f SEM. RESULTS
Resting Metabolic Rate
Physical characteristics are shown in Table 1. Highly trained men had a lower body weight (P < .05) than the untrained group. Percent body fat was lower (P-C .Ol) in highly trained men than the other two groups. No differences in age, height, body mass index (wt/h*), and FFW were found among groups. As illustrated in Fig 1, a significant correlation (P < .Ol) was noted between 60, max and RMR . kg-’ - h-‘. This persisted when RMR and VO, max were adjusted for FFW (r = S5, P < .Ol; not shown in figure). In Table 2 with subjects classified into fitness levels, the highest RMR was observed in highly trained as compared to moderately trained and untrained men. Thermic Effect of a Meal
Using curvilinear regression, data were analyzed as continuous variables to test for a nonlinear relationship between 90, max and TEM (Fig 2). A significant curvilinear relationship was found (P -C .05), indicating that the higher TEM was noted in the mid-range of 90, max values, whereas a lower TEM was observed in the extremes of (low Moderately
Trained, and Highly Trained Men
Moderatelytrained In- 11) 22.7
+ 0.8
1.8 + 0.02
Highlytrained (n - 8) 26.3
+ 2.3
1.8 f 0.02
Wt (kg)
79.8
zt 3.2
74.1
f. 2.2
70.8
+ 1.5
Body mass index (wt/ht’)
24.1
f 0.8
23.4
+ 0.8
21.9
+ 0.4
% body fat
14.8 f 1.5
13.6 f 1.3
FFW kg)
68.1
+ 3.0
64.0
f 1.8
65.9
45.6
+ 0.7
53.3
+ 0.7
70.9 * 2.1
6.9 + 0.8
P
NS NS H < U; <.05 NS H < U; <.Ol H < M; <.Ol
\iO, max (mL
.
kg-’
.
min-‘)
+ 1.4
NS
Abbreviations: U, untrained; M, moderately trained; H, highly trained; NS, not significant. Values are the mean + SEM. Differences between \iO, max among groups were not statistically analyzed because it was used to classify fitness level of subjects.
a7
RESTING ENERGY EXPENDITURE AND AEROBIC FITNESS
Table 2. Resting Metabolic
Rate of Untrained,
Untrained Variable
RMR (kcal RMR (kcal
- min-‘l . kg-’ .
In -
h-‘)
9)
Moderately
Moderately (n-
Trained 11)
Trained,
and Highly Trained Men Highly Trained (n -
1.24 + 0.05
1.13 f 0.03
1.31 * 0.05
0.93
0.92
1.11 2 0.03
f 0.02
* 0.02
P
9)
H > M: <.Ol H > U: <.Ol H > M; <.05
RMR (kcal
.
FFW-’
.
h-‘1
1.09 * 0.03
1.06 f 0.03
1.20 * 0.04
t-l z u; <.05 H > M; ~05
Values are the mean + SEM.
(0.97 + 0.05) when compared to the untrained (0.83 + 0.04, P < .05) and highly trained (0.77 k 0.04, P < .Ol) men. Hormones, Substrates, and Cardiovascular Parameters
I
.
Fig 2. The curvilinear relationship between $0, max and TEM in 28 male nonobese subjects. A significant curvilinear relationship was found in which the equation is: TEM (kcal - kg-’ - 180 min.‘) = -0.909 - 0.065 (VO, max. mL - kg-’ - min-‘) O.ooO5881 (VO, max mL . kg-’ . min-‘1’ (P c .OSl.
and high) QO, max. A similar relationship was obtained when TEM was adjusted for FFW. No significant association between VO, max and TEM (r = - .23) was noted (not shown in figure form). The time course of changes of TEM in untrained, moderately trained, and highly trained subjects are illustrated in Fig 3. After meal consumption an elevation of metabolic rate was noted in all subjects. Total postprandial energy expenditure (TEM) varied according to the level of aerobic fitness (inset Fig 3). The postprandial response (kcal . kg-’ . 180 min-‘) was higher in the moderately trained group
Preprandial and postprandial plasma insulin, glucose, and heart rate responses are shown in Table 3. As expected, fasting and postprandial plasma insulin levels were lower in highly trained as compared to moderately and untrained men. In spite of differences in plasma insulin, no differences in plasma glucose were noted among the three groups before or after meal consumption. Identical results were found when insulin and glucose responses were analyzed by surface area under the response curves. Preprandial and postprandial heart rate was lower in the highly trained as compared with the other two groups. No differences were noted among groups in the preprandial and postprandial blood pressure and respiratory quotient (RQ). Plasma T, (ng/lOO mL) concentrations at time 0 minutes (ie, before meal ingestion) were not different among untrained (117 k 5.3), moderately trained (117 k 5.3), and highly trained men (118 + 10.4). Plasma T, (pg/lOO mL) concentrations were also not different at time 0 minutes in untrained (6.5 -L0.3), moderately trained (6.4 2 0.5), and highly trained men (5.6 2 0.3). No significant changes were noted in plasma T, or T, after meal ingestion. As shown in Table 4, daily energy intake was higher in the highly trained men in comparison with the other two groups. Highly trained men consumed a higher intake (P < .05) of protein than moderately trained and untrained men. The percent contribution of the macronutrients to daily energy intake was not different among the three groups.
10 t
Fig 3. Increase in poatprandial energy expenditure (TEMI for untrained, moderately trained, and highly trained male subjects after ingestion of a liquid meal.
T
88
POEHLMAN ET AL
Table 3. Plasma Insulin, Plasma Glucose, Heart Rate, and Respiratory Quotient (RQ) Before and After Meal Consumption in Untrained. Moderatelv Trained. and Hiahlv Trained Men Untrained (n - 9)
Variable Insulin (0 min) (/dJ/mL)
25.2
ModeratelyTrained (n- 111
+ 3.4
25.1
HighlyTrained (n - 6)
+ 3.6
14.2 ? 3.0
P
U > H: <.05 M > H; 1.05
Insulin (30-180
min) (pU/mL)
Glucose (0 mini (mg/lOO mL) Glucose (30-l 80 min) (mg/lOO mL)
86.6
f 7.4
67.4
+ 6.9
50.2
* 8.3
90.4
f 2.1
91.4
+ 2.4
88.7
+ 2.4
101.4
f 3.4
97.3
f 2.8
93.1
+ 2.5
Heart rate (0 min) (beats/min)
61 f 2.4
Heart rata (30-l 80 min)
70 i
55 * 2.1
1.6
65 + 2.9
U > H: <.Ol NS NS
51 * 2.9
U > H; <.Ol
54 ? 2.6
M > H; <.Ol
RQ (0 min)
0.81
+ .02
0.80
f .Ol
0.81
+ .02
NS
RQ (30-l 80 min)
0.89
f .Ol
0.90
f .Ol
0.87
+ .Ol
NS
All values are the mean + SEM; 0 min = time before meal ingestion: 30-l 80 min = mean postprandial response.
Significant linear correlations were noted between RMR . kg-’ . h-’ and daily energy intake (r = S4, P < .Ol) and between RMR a kg-’ - h-’ and basal levels of total plasma T, (r = .37, P -c .05; not shown in figure form).
trained men.3 However, LeBlanc et al did not find an elevated RMR in highly trained females as compared with moderately trained and untrained women.9 It is possible that RMR response to exercise training is more sensitive in males who have a higher proportion of FFW to total body weight. Hill et al observed a -9% higher RMR per kg FFW in trained v untrained men.6 This difference was similar to that observed in the present study, but a small sample size in their study probably precluded statistical significance. Thus, the highly trained male may achieve high levels of energy expenditure by two ways: (1) the direct energy cost of aerobic exercise and (2) a high RMR. In the present study it is unlikely that the high RMR in highly trained men is due to the prolonged effects of the last exercise bout. At least 24 hours separated the last exercise session and RMR measurement. Bahr et al” reported no prolonged effect of heavy exercise (80 minutes, 70% ir0, max) on basal oxygen consumption 24 hours postexercise in male subjects. Hill et al6 noted a higher RMR in trained men 36 hours postexercise, which along with our data, suggests that the elevated RMR in highly trained men may be a chronic adaptation to the exercise-trained state. The mechanism for the high RMR in highly trained men remains to be established. However, preliminary evidence suggests that RMR could be affected when the caloric turnover is high, concurrent with the maintenance of a state of energy balance. This would be accomplished by matching a high level of food intake with a high level of exerciseinduced energy expenditure, which is characteristic of the highly trained men in this study. The association among energy intake, energy expenditure (90, max), and RMR in
DISCUSSION
Resting Metabolic Rate
The potential relationships among fitness level, RMR, and TEM were examined in young males who exhibited a wide range of aerobic fitness levels. An elevated RMR (-10%) was found at high fitness levels when adjusted for differences in body weight and FFW. Hence, it appears that the highly trained male has a high RMR for his metabolic size. This is consistent with previous reports*’ but in apparent contradiction with others.5-7.9The discrepancy between our results with those of previous studies may be partially attributed to two factors: (1) differences in the definition of the highly trained individual among investigators and (2) smaller sample sizes in previous studies. Concerning the former point, other investigators who have not observed a high RMR in trained individuals have examined subjects with an aerobic capacity approaching -60 mL - kg-’ - min-‘.5-7S9Present results and those of others***support a high RMR in male subjects who engage in heavy exercise and achieve very high maximal aerobic capacities (-70 ml - kg-’ - min-‘). This may suggest that very high levels of aerobic fitness must be achieved before they impact upon RMR. This notion is consistent with the results of Tremblay et al who found a high RMR . kg FFW-’ in highly trained men by subdividing their trained group into highly trained and moderately
Table 4. Nutrient Intakes as Determined From 3-Day Food Diaries for Untrained, Moderately Trained, and Highly Trained Men Untrained Nutrient
(n -
9)
Moderately In-
Trained 11)
Highly
Trained
(n = 9)
P
Energy intake (kcal/d)
3036
f 214
3153
* 191
4688
r 448
H > M; <.Ol
Protein (g/d)
119.0
+ 12.7
121.8
+ 7.7
167.0
+ 21.3
H > M; <.05
(15.6
f 1.8)
(15.9
* 1.3)
(16.1
+ 1.9)
NS
419.0
+ 48.8
388.5
+ 47.4
572.0
f 93
NS
(54.7
* 4.3)
(48.6
+ 4.3)
(52.9
* 4.0)
NS
103.1
+ 10.0
110.5
* 16.6
152.3
+ 27.8
NS
(30.8
+ 3.01)
(31.2
f 4.2)
(31.1
f 3.3)
NS
H > L; <.Ol H > L; <.05 Carbohydrates (g/d)
Fat (g/d)
Values are the mean + SEM. Values in parentheses indicate percentage contribution of the macronutrient to total daily intake.
RESTING ENERGY EXPENDITURE
AND AEROBIC
FITNESS
this study is supported by correlations between RMR and energy intake (r = .54, P -C .Ol) and between RMR and VO, max (r = .77, P < .Ol; Fig 1). Relevant to this discussion are the findings of Woo et al who created a high caloric turnover condition by overfeeding six males, but they maintained a state of energy balance by a proportionate increase in aerobic exercise.” RMR was higher after exercise training, but no changes were found in plasma thyroid hormones. This is consistent with our findings in which no differences in thyroid hormones were observed between individuals varying in aerobic fitness, but an elevated RMR was noted in the highly trained men. Thermic Effect of a Meal
In the present study we found that TEM varied with fitness level in a curvilinear manner, ie, moderate levels of aerobic fitness were associated with the higher TEM than the extremes of low and high fitness. This finding is supported by two analyses: (1) linear and nonlinear regression analysis to examine the relationships between TEM and VO, max without prior classification, and (2) difference testing with subjects classified into commonly accepted fitness levels using ANOVA. This approach reduced the possibility of identifying misleading physiologic relationships from only evaluating subjects classified into discreet fitness levels. Linear regression analysis revealed a low order correlation between VO, max and TEM (r = .23, NS), whereas by using nonlinear regression we found a significant curvilinear relationship (Fig 2). This provides evidence that moderate levels of fitness were associated with the highest TEM, whereas a more efficient postprandial response (ie, lower TEM) was observed at the extremes of the fitness continuum. This observation may partially account for the linear relationship between VO? max and TEM reported in previous studies when, in fact, trained subjects were similar in aerobic fitness to the moderately trained, high TEM responders in this study. 5.6When the highly trained group was excluded from our data analysis, a linear correlation was observed (r = .47, P C: .OS). Thus, the positive association between VO, max and TEM observed in the untrained and moderately trained subjects is similar to previous investigationssJ’ that examined fewer subjects and a small range of fitness levels. Thus, we would propose that a curvilinear relationship between VO, max and TEM would be only observed in an experimental design, which examined a wider range of fitness levels and a larger sample size. The enhanced energy expenditure due to a high TEM could partially account for the observation that some individuals undergoing exercise programs lose more weight than expected from the calories expended during the exercise.*‘,” We also examined the differences between TEM and VO, max by classifying subjects into commonly accepted aerobic fitness categories (Fig 3), as previously performed.3~6~7-” TEM was higher in the moderately trained as compared to umrained and highly trained individuals. This supports the notion that moderate levels of exercise training may enhance postprandial energy expenditure. The hypothesis of a curvilinear relationship between aer-
89
obic fitness and TEM has been suggested,** and the present study, to our acknowledge, provides the first evidence of this relationship. It is unknown whether similar hormonal and/or substrate mechanisms regulate the blunted TEM in untrained and highly trained men. A lower TEM has been found in highly trained men and women2,7*‘o*”and in nonobese, untrained men.6 Although it cannot be directly determined from our data, the low TEM in untrained men may be associated with a lower rate of insulin-mediated glucose uptake and lower rates of glucose storage (ie, obligatory thermogenesis). This is suggested by their higher postprandial levels of plasma insulin. This explanation seems unlikely to account for the low TEM in highly trained men who are noted for their enhanced insulin sensitivity. An alternative possibility may be found in the sympathetic activation of thermogenesis. Sympathetic activity has been shown to be positively correlated to the size of TEM.23 Low levels of plasma catecholamines have been reported in highly trained men, which is consistent with their low TEM.’ We noted a lower postprandial heart rate in highly trained men, which indirectly supports the lower sympathetic tone in this group as compared to other individuals. One may consider if differences in the phenotypic expression of RMR and TEM among individuals are the result of training and/or genetic influences. We reported a large variation in changes of RMR and TEM to short-term exercise training among six pairs of monozygotic twins, whereas a coherent response pattern was noted within twin pairs. This suggested that biologic variation in RMR and TEM in response to exercise training is partially influenced by the genetic background.24 However, it is premature to draw conclusions regarding the degree of heritability of RMR and TEM in response to exercise. The aforementioned experiment was not designed to assess the contribution of heredity to variation in RMR and TEM in the population but rather to explore the phenomenon of the genotype-environment interaction in man. Furthermore, biologic traits (ie, FFW) that are known to covary with the phenotypic expression of RMR and TEM have also exhibited a genotype dependency in response to chronic exercise.25 Our studies point out the need to prospectively examine subjects undergoing long-term aerobic training, resulting in large changes in fitness to critically examine the association between changes in aerobic fitness and thermogenesis. In summary, our data suggest that differences in fitness level are associated with variations in resting energy expenditure. High levels of aerobic fitness are associated with an elevated RMR, whereas moderate levels of fitness are related to an enhanced postprandial thermogenesis. ACKNOWLEDGMENT
The authors thank Dr. G. Bottoms, Hormonal
Assay Laboratory, Purdue University, for his cooperation in the hormonal analysis, and to Drs. E. Danforth, Jr, and E.A.H. Sims for their constructive criticisms of the manuscript. REFERENCES 1. Lennon DF, Nagle F, Stratman F, et al: Diet and exercise training effects on resting metabolic rate. Int J Obes 9:39-47, 1985
90
2. Poehlman ET, Melby CL, Badylak SF: Resting metabolic rate and postprandial thermogenesis in highly trained and untrained males. Am J Clin Nutr 47:793-798,1988 3. Tremblay A, Fontaine E, Nadeau A: Contribution of postexercise increment in glucose storage to variations in glucose-induced thermogenesis in endurance athletes. Can J Physiol Pharmacol 63:1165-1169,1985 4. Tremblay A, Fontaine E, Poehlman ET, et al: The effect of exercise-training on resting metabolic rate in lean and moderately obese individuals. Int J Obes 10:51 l-517,1986 5. Davis JR, Tagliaferro AR, Kertzer R, et al: Variations in dietary-induced thermogenesis and body fatness with aerobic capacity. Eur J Appl Physiol50:319-329, 1983 6. Hill JO, Heymsfield SB, McMannus C, et al: Meal size and thermic response to food in male subjects as a function of maximum aerobic capacity. Metabolism 33:743-749, 1984 7. LeBlanc J, Diamond P, C&e J, et al: Hormonal factors in reduced postprandial heat production of exercise-trained subjects. J Appl Physiol56:772-776, 1984 8. LeBlanc J, Jobin M, C&C J, et al: Enhanced metabolic response to caffeine in exercise-trained human subjects. J Appl Physiol59:832-837, 1985 9. LeBlanc J, Mercier P, Samson P: Diet-induced thermogenesis with relation to training state in female subjects. Can J Physiol Pharmacol62:334-337, 1984 10. Poehlman ET, Desprds JP, Bessette H, et al: Influence of caffeine on the resting metabolic rate of exercise-trained and inactive subjects. Med Sci Sports Exert 17:689-694, 1985 11. Tremblay A, C&t J, LeBlanc J: Diminished dietary thermogenesis in exercise-trained human subjects. Eur J Appl Physiol 52:1-4, 1983 12. Siri WE: The gross composition of the body. Adv Biol Med Phys 4:239-280, 1956 13. Wilmore JH, Vodak PA, Parr RB, et al: Further simplifica-
POEHLMAN ET AL
tion of a model for determination of residual long volume. Med Sci Sports Exert 12:216-218, 1980 14. Poehlman ET, Arciero PJ, Melby CL, et al: Resting metabolic rate and postprandial thermogenesis in vegetarians and nonvegetarians. Am J Clin Nutr 48:209-213, 1988 15. Weir JB de V: New methods for calculating metabolic rate with special reference to protein. J Physiol (Lond) 109:1-9, 1949 16. Snedecor GW, Cochran WG: Statistical Methods (ed 6). Ames, IA, Iowa State University Press, 1973, pp 447-459 17. Sinning WE, Experiments and demonstrations in exercise physiology. Philadelphia, Saunders, 1975 18. Bahr R, Ingnes I, Vaage 0, et al: Effect of duration of exercise on excess postexercise 0, consumption. J Appl Physiol 62:485-490, 1987 19. Woo R, O’Connell M, Horton ES, et al: Changes in resting metabolism with increased intake and exercise. Clin Res 33:712A, 1985 (abstr) 20. Bjiirntorp P, Berchtold P, Lindholm B, et al: Effects of physical training on glucose tolerance, plasma insulin, and lipids and on bcdy composition in men after myocardial infarction. Acta Med Stand 192:439-443,1972 21. Oscai LB, Williams BT: Effect of exercise on overweight middle-aged males. J Am Geriatr Sot 16:794-797, 1968 22. Segal KR, Xavier Pi-Sunyer F: Exercise, resting metabolic rate, and thermogenesis, in DeFronzo RA (ed): Diabetes/Metabolism Reviews, Vo12, New York, Wiley, 1986, pp 20-34 23. Schwartz RS, Jaeger LF, Silberstein S, et al: Sympathetic nervous system activity and the thermic effect of feeding in man. Int J Obes 11:141-149, 1987 24. Poehlman ET, Tremblay A, Nadeau A, et al: Heredity and changes in hormones and metabolic rates with short-term training. Am J Physiol250:E711-E717,1986 25. Poehlman ET, Tremblay A, Marcotte M, et al: Heredity and changes in body composition and adipose tissue metabolism after short-term exercise training. Eur J Appl Physiol56:398-402, 1987