INSERM U 317, Laboratoire de Pharmacologie Médicale et Clinique, Faculté de Médecine, 37 Allées Jules Guesde, 31073 Toulouse Cedex, France. E-mail: email@example.com
Objective: The aim of this study was to determine how training modifies metabolic responses and lipid oxidation in overweight young male subjects.
Research Methods and Procedures: Eleven overweight subjects were selected for a 4-month endurance training program. Before and after the training period, they cycled for 60 minutes at 50% of their Vo2max after an overnight fast or 3 hours after eating a standardized meal. Various metabolic and endocrine parameters, and respiratory exchange ratio values were evaluated.
Results: Exercise-induced plasma norepinephrine concentration increases were similar before and after training in fasted or fed conditions. After food intake, exercise promoted a decrease in plasma glucose and a higher increase in epinephrine than in fasting conditions. The increase in epinephrine after the meal was more marked after training (264 ± 32 vs. 195 ± 35 pg/mL). Training lowered the resting plasma nonesterified fatty acids. During exercise, changes in glycerol were similar to those found before training. Lipid oxidation during exercise was higher in fasting than in fed conditions (15.5 ± 1.4 vs. 22.3 ± 1.7 g/h). Training did not significantly increase fat oxidation when exercise was performed in fed conditions, but it did in fasting conditions (18.6 ± 1.4 vs. 27.2 ± 1.8 g/h).
Discussion: Endurance training decreased plasma nonesterified fatty acids, cholesterol, and insulin concentrations. Training increased lipid oxidation during exercise, in fasting conditions, and not when exercise was performed after the meal. During exercise in overweight subjects, the fasting condition seems more suited to oxidizing fat and maintaining glucose homeostasis than a 3-hour wait after a standard meal.
The rapidly increasing prevalence of overweight and obesity in developed countries confirms that obesity has become a major threat to public health. Regular physical exercise, even without decrease in body mass, elicits pronounced metabolic improvement (1). The actual concept is that lipids, particularly nonesterified fatty acids (NEFAs),1 play an important role as muscle substrate during exercise exerted at one-half or less of the maximal aerobic capacity, whereas greater efforts mainly depend on glucose and muscle glycogen as substrates (2)(3). Lipid mobilization during exercise is mainly dependent on an increase in sympathoadrenal activity and a decrease in insulin secretion (3)(4)(5). The impact of endurance training on fat oxidation and energy expenditure has been mainly explored in lean subjects (6)(7). A recent paper devoted to obese men suggests that low-intensity exercise training increases fat oxidation during exercise (8). However, endocrine modifications, catecholamine changes, and the incidence of nutritional status have not been explored. A previous study has shown that during submaximal exercise performed 4 hours after a carbohydrate meal, the decrease in plasma glucose concentration was significantly marked in trained compared with untrained lean subjects (9). This suggests that it could be important to investigate the effect of training on endocrine and metabolic responses promoted by exercise according to the nutritional status (fed or fasted) in overweight subjects.
To improve the therapeutic role of exercise, it is necessary to determine the nutritional situation (fed or fasted) that can be proposed to overweight or obese patients aiming to enter a training program. From this point of view, the nutritional conditions proposed to these subjects will focus on lipid oxidation during exercise. The purpose of our study was to evaluate the effect of training on lipid use and endocrine responses during submaximal aerobic exercise in overweight subjects. Before training, the subjects performed standardized exercise at 50% of peak O2 consumption (Vo2max) on 1 day in fasted conditions, and 1 week later, 3 hours after a standardized meal (fed). Four months after the training program, they exercised in a similar schedule at 50% of their new Vo2max.
Research Methods and Procedures
Subjects and Exercise Training Protocol
Eleven overweight untrained men (age, 25.6 ± 1.4 years; body mass index, 27.7 ± 0.2 kg/m2) participated in the study. All were drug-free, and their weight had remained stable for at least 3 months before the beginning of the study. They had all given their written informed consent before the experiments began. The studies were performed according to the Declaration of Helsinki and approved by the Ethical Committee of Toulouse Hospital, France. The Vo2max of the subjects was determined during a cycling test. The training program of aerobic exercise (1 hour per day, 5 days per week) consisted of running (3 days per week) and cycling (2 days per week), for 4 months under the control of a physical exercise coach. The exercise intensity and duration were progressively increased. The subjects exercised at a target heart rate corresponding to 50% to 85% of their Vo2max. Heart rate was monitored with a Polar Accurex Plus Cardiometer Monitor (BFM, La Varenne, St. Hilaire, France). Compliance with training was good, as checked by a training diary including day-to-day activities. Food and calorie intake were uncontrolled throughout the protocol.
Design of the Study
Before and after the training program, the following measurements and tests were carried out. Plasma concentrations of glucose, glycerol, NEFA, triglycerides, cholesterol, insulin, leptin, and catecholamines were evaluated. The amount of food consumed was recorded over 3 days in a food intake diary, and the diaries were analyzed with the Regal table and Profile Nutrition software (ACIM, Saint Doulchard, France). Fat mass and fat-free mass (bone and muscle) were measured using a total body DXA (Hologic Inc., Waltham, MA). For maximum aerobic capacity, the Vo2max was assessed using a graded test on an electromagnetically braked bicycle ergometer (Ergometrics 800s Ergoline, Jaeger, Germany). To ensure that the volunteers were untrained, their Vo2max was determined during a cycling test: an initial workload of 60 W was followed by a sequential increase in workload of 30 W every 3 minutes until exhaustion. Verbal encouragement was given to attain maximal performance. Heart rate was continuously monitored. The highest Vo2 achieved was taken as Vo2max, and the workload corresponding to 50% of each subject's Vo2max was calculated. Two criteria assessed that the subjects achieved their true Vo2max. Maximal heart rate measured at exhaustion (188.6 ± 2.6 beats/min) was not different from their age-predicted maximal heart rate (189 ± 2.8 beats/min), and the maximal respiratory exchange ratio (RER) measured at exhaustion was near 1.1. The mean Vo2max before training was 34.3 ± 1.3 mL/kg per minute. After the Vo2max determination, the subjects were randomly assigned to start the first day of the experimental protocol during the following week. The time before the second experimental protocol was 7 days. After the Vo2max test and during the whole experimental period, the subjects were asked to maintain their usual diet.
The subjects were investigated at 8:00 am for 2 days separated by 1 week according to a randomized crossover procedure. The evening preceding every trial, the subjects were given a standard meal consisting of 3350 kJ (55% carbohydrate, 30% fat and 15% protein). After an overnight fast in one investigation, the subjects remained in the fasted condition throughout the investigation period, and in the second, the subjects ate a usual standardized breakfast (2962 kJ; 83% carbohydrate, 7% fat and 10% protein) provided by the investigators. Then, after a 3-hour rest period, the subjects started a 1-hour bout of exercise on the bicycle ergometer at a workload corresponding to 50% of their individual Vo2max. The mean power developed was 88.2 ± 5.8 W. Then, the subjects rested in the semirecumbent position for 60 minutes. The heart rate was continuously monitored with a Polar Accurex Plus Cardiometer Monitor during the exercise. Water intake was allowed ad libitum during exercise and recovery periods.
Before exercise and every 15 minutes during exercise and recovery, 5 mL of blood was collected from an indwelling polyethylene catheter inserted into an antecubital vein for glycerol, NEFA, insulin, and glucose determinations. Every 30 minutes, an additional volume of 5 mL of blood was collected for catecholamine determinations. The blood was collected on 50 μL of an anticoagulant and antioxidant cocktail (Immunotech SA, Marseille, France) to prevent oxidation of catecholamines and processed immediately in a refrigerated centrifuge. The plasma was stored at −80 °C until analysis. Before each blood sample was taken, the subjects wore a mask for 6 minutes to measure respiratory gas exchanges. RER values (RER = CO2 production/Vo2) were calculated on-line through an open-circuit indirect calorimetry device using an Oxycon Pro apparatus (Jaeger, Germany) coupled to a computer for RER calculation. Certified calibration gases were used to calibrate analyzers every day before the beginning of the assay and before the exercise. From the gas exchange measurements, the proportion of lipid oxidized during exercise and recovery was evaluated according to the formula described by Ferrannini (10).
For the posttraining test, the experimental design was identical. The workload corresponded to 50% of the new individual Vo2max, and the mean power developed was 110.0 ± 5.6 W. The subjects continued training between the 2 experimental days (separated by 1 week). Nevertheless, the subjects did not train the day preceding the investigation to avoid the consequences of the last physical activity bout on the metabolic responses analyzed.
Drugs and Biochemical Determinations
Plasma glucose and NEFA concentrations were determined with a glucose-oxidase technique (Biotrol kit; Biomerieux, Marcy l'Etoile, France) and by an enzymatic procedure (Wako kit; Oxoid, Dardilly, France), respectively. Glycerol in plasma was analyzed with an ultrasensitive radiometric method (11). Plasma leptin levels were assayed using a radioimmunoassay kit from Linco (St. Charles, MO). Plasma insulin concentrations were measured using radioimmunoassay kits from Biorad (Marnes la Coquette, France). Plasma epinephrine and norepinephrine were assayed in 1-mL aliquots of plasma by high-performance liquid chromatography using electrochemical (amperometric) detection. The detection limit was 20 pg/sample. Day-to-day variability was 4% and within-run variability was 3%.
A statistical comparison of the curves was performed using multivariance analysis (with quadruple-, triple-, or double-multivariate repeated measure design) with the nutritional status, period (fed or fasting), and training as factors of the analysis. Values were considered statistically significant when p < 0.05. Statistical analyses were performed using software packages (Statview 4.5 and SuperAnova 1.11; Abacus Concepts Inc., Berkeley, CA).
Effect of Training on Basal Parameters and on the Metabolic Response to a Meal at Rest
After 4 months of training, weight, body mass index, and percentage of fat mass decreased (87.6 ± 1.6 vs. 89.5 ± 1.6 kg, p < 0.05; 26.9 ± 0.31 vs. 27.7 ± 0.23, p < 0.01; and 21.7 ± 1.1 vs. 22.8 ± 0.9%, p < 0.01, respectively). Vo2max values were significantly increased after training (39.90 ± 1.43 vs. 34.30 ± 1.30 mL/kg per minute before training, p < 0.001). During the exercise test, the mean power developed at 50% Vo2max was significantly higher after the 4-month training than before (p < 0.001). Resting heart rate was 65.8 ± 0.3 beats/min before and 58.7 ± 0.3 beats/min after training (p < 0.001). Systolic and diastolic blood pressure remained unchanged. Endocrine and metabolic values measured at 8:00 am are reported in Table 1. Plasma insulin, leptin, cholesterol, and NEFA levels were decreased after the 4-month training. At rest and in fasted conditions, lipid oxidation was increased after training. As found with food diary analysis, the qualitative amount and the quantitative composition of food intake were unchanged during training (Table 2).
Table 1. Endocrine and metabolic parameters measured before and after a 4-month training period in overweight subjects
Data are expressed as means ± SEM.
NS, not significant.
493 ± 76
249 ± 24
84 ± 6
103 ± 9
1.59 ± 0.28
1.18 ± 0.20
2.29 ± 0.13
2.02 ± 0.11
5.21 ± 0.16
4.93 ± 0.11
5.24 ± 0.41
3.55 ± 0.30
5.74 ± 0.78
4.37 ± 0.58
287 ± 26
263 ± 28
27 ± 4
34 ± 7
Lipid oxidation (g/15 min)
1.61 ± 0.09
1.75 ± 0.07
Table 2. Calorie and food intake measured during the 1st and the 15th week of training in overweight subjects
Diaries were analyzed with the Regal table and Profile Nutrition Software (Acim, France). Data are expressed as means ± SEM.
NS, not significant.
10979 ± 680
10260 ± 733
17.6 ± 1.0
16.6 ± 1.0
39.8 ± 1.9
39.7 ± 1.8
42.6 ± 1.6
43.7 ± 2.3
Table 3 depicts plasma glucose, insulin, NEFA, and glycerol concentrations in response to the meal before and after training. Values at 0 minutes were those measured before the meal. After training and before the meal, plasma glucose, insulin, and NEFA concentrations were found to be lower than before training; glycerol concentrations were higher. The meal induced similar changes in plasma glucose before and after training, and the increase in plasma insulin concentrations was found to be lower after training; the changes differed significantly (p < 0.01).
Table 3. Resting plasma insulin, glucose, NEFA, and glycerol response to the pre-exercise meal before and after training
Significantly different when compared to time 0 min (before the meal).
Significantly different when compared to before training.
Data are expressed as means ± SEM. The insulin, NEFA, and glycerol responses to a meal before and after training were different (p < 0.02, p < 0.06, and p < 0.02, respectively), but not the glucose response.
Effect of Training on Endocrine and Metabolic Responses during Exercise
Plasma Catecholamines, Glucose, and Insulin
Plasma norepinephrine concentrations were increased to a similar extent after 30-minute exercise with fasting or 3 hours after food intake and did not change until the end of exercise (Figure 1). The changes were similar before and after training. However, plasma epinephrine concentrations steadily increased until the end of the exercise bout both before and after training (Figure 2). The major results concerned the effect of the nutritional status on plasma epinephrine; the increase was higher in fed conditions, before (p < 0.04) and after training (p < 0.004), than in the fasted condition. In addition, it was noticed that after feeding, the increase in epinephrine concentration was greater after training, and after 60-minute exercise, the values differed significantly from those measured before training (p < 0.03). In fasted conditions, no modification of plasma glucose levels occurred during exercise (Figure 3). After feeding, an early decrease in plasma glucose level occurred 15 minutes after the beginning of the exercise. The changes in plasma glucose during the exercise period differed in fed and fasted conditions before (p < 0.001) and after (p < 0.002) training. Before exercise, insulin levels were significantly higher after feeding than in the fasted status (Figure 3). As expected, plasma insulin levels decreased during exercise in fed and in fasted subjects. It was noted that during exercise, plasma insulin concentrations were lower in fasting conditions and that in addition, after training, the insulin concentrations were significantly lower in fed or fasted conditions than before training.
Plasma NEFA and Glycerol
The profile of plasma NEFA concentration changes during exercise differed according to the nutritional status of the subjects. In fasted conditions, NEFA concentrations initially decreased for 15 minutes and then steadily increased during exercise, but the values remained close to those found in the pre-exercise situation. We observed an increased NEFA concentration 15 minutes after recovery and then a decrease to values different from those found in basal conditions (Figure 4). Plasma NEFA concentrations greatly decreased during food intake and progressively increased during exercise. Consequently, the changes induced by exercise differed according to the nutritional status. Training lowered plasma NEFA concentrations in fasted or fed conditions at rest, but the time-course of the exercise-induced changes in both nutritional conditions were very similar to those found before training. The evolution of exercise-induced changes differed in fasted and fed conditions before (p < 0.002) and after (p < 0.0005) training. Three hours after food intake, the plasma glycerol concentration was lower than the corresponding fasted values (Figure 5). In addition, the time-course of plasma glycerol concentration changes during exercise did not differ according to the nutritional status of the subject. Plasma glycerol levels increased during exercise and steadily decreased during recovery. The changes induced by exercise were not different in fasted or fed conditions before or after training.
Effect of Training on Lipid Use
At rest, before or after training, calculated fat oxidation was significantly higher (p < 0.05) in fasted conditions than 3 hours after food intake (1.68 ± 0.07 vs. 0.78 ± 0.11 g lipids and 1.73 ± 0.07 vs. 1.17 ± 0.13 g lipids per 15 minutes before and after training, respectively). Before training, the subjects derived 45.5 ± 3.2% and 31.4 ± 3.28% of calories from fat oxidation in fasted and fed conditions, respectively. After training, the percentage of calories from fat oxidation was significantly increased (50.2 ± 2.69%; p < 0.05) in fasting conditions, but not in fed conditions (34.1 ± 2.1%).
Before training, lipid oxidation was higher in fasted than in fed conditions but did not reach a significant difference. After training, exercise-induced lipid oxidation was significantly higher in fasted than fed conditions (Table 4). In addition, training did not increase lipid oxidation when exercise was performed after feeding, whereas lipid oxidation was significantly increased when exercise was performed in fasted conditions.
Table 4. Energy expenditure (kilojoules) and lipid oxidation during 60-minute exercise in fed and fasted conditions before and after training in overweight subjects
Energy provided by lipids
Lipid oxidation (g)
p < 0.05 when comparing fed to fasted conditions.
p < 0.05 when compared to before training in similar nutritional conditions.
This study was designed to investigate whether training modifies the responses to exercise according to the nutritional status in overweight subjects. It shows that fasting conditions are more suited to promoting lipid oxidation and to maintaining glucose homeostasis than a 3-hour wait after a meal.
The proportion of energy derived from lipids or carbohydrates is highly variable during exercise and influenced by numerous factors, including hormonal milieu, nutritional status, and exercise intensity (2)(12)(13)(14). Studies have shown that obese subjects exhibit alterations of lipid mobilization (15)(16) and probably have a constitutive impairment of lipid oxidation, which could play a role in the etiology of obesity. Postobese subjects have been shown to exhibit levels of exercise-induced lipid mobilization similar to controls, but they exhibit a low rate of fat oxidation (17)(18).
Physical activity programs are currently proposed in association with dietary prescriptions to improve physical fitness and prevent or treat obesity. Nevertheless, it is important to define the optimal nutritional conditions for overweight or obese subjects who plan to be engaged in a training program. To our knowledge, in overweight subjects before and after endurance training, no longitudinal investigations concerning the effects of nutritional conditions on the metabolic and endocrine changes during exercise have been conducted.
The increase in sympathoadrenal activity and the decrease in insulin secretion are the most important autonomic neuroendocrine responses during exercise (19). The results obtained in this study show that the sympathoadrenal activity clearly differs according to the nutritional status of the subjects. The sympathetic activity (as seen by the increased plasma norepinephrine levels) involving nerve impulses to sympathetic centers from motor centers in the brain and muscle contraction (5) was not modified by feeding. On the other hand, when the exercise bout was performed after food intake, a decrease in plasma glucose levels could explain the enhanced epinephrine secretion. Even a small decrease in plasma glucose is considered to be able to enhance epinephrine secretion during exercise (4). This point has previously been well delineated by Galbo et al. (20). Insulin secretion in response to a meal has the potential to decrease plasma glucose concentrations during exercise and to increase carbohydrate oxidation, and thus, decrease lipid oxidation. Glycerol levels were lowered after food intake, but the exercise-induced glycerol changes occurring during the exercise bout were identical. This suggests that lipid mobilization from fat deposits was unaffected by plasma insulin, the levels of which remained higher at the onset of exercise after the meal. In addition, the higher exercise-induced concentrations of epinephrine, acting on the human adipocyte, can counteract the antilipolytic effect of insulin (5)(21). In contrast to glycerol, the changes in plasma NEFA promoted by exercise differed profoundly. After an overnight fast, plasma NEFA levels decreased during the initial 15 minutes of exercise and then remained near pre-exercise values, as we have previously reported (16). One explanation is that the onset of muscular exercise is associated with a rapid uptake of NEFA that exceeds the amount provided by lipolysis (22). In the fed situation, plasma NEFA concentration gradually increased over the pre-exercise levels. The difference in the NEFA concentration is probably not caused by the rate of lipolysis because there is no change in the glycerol concentration. NEFA uptake and re-esterification should also be induced (23). Our results suggest that the use of NEFA as muscle substrate during the exercise bout is higher in fasted than in fed conditions and are congruent with the fact that carbohydrate ingestion regulates and limits fatty acid oxidation during exercise (24)(25).
The most striking effect of 4 months of training was the reduction of plasma NEFA and the weak increment in plasma glycerol concentrations. This suggests that, although spontaneous lipolysis was weakly increased, training enhanced NEFA use in the fasted condition (assessed by the significant spontaneous increase in lipid oxidation at rest). Training promoted an increase in Vo2max. The exercise test was adjusted to the new Vo2max and performed at a relative power identical to that imposed before training; the power developed was higher. The fact that, after training, the exercise was done at the same relative intensity as before was confirmed by the same norepinephrine level. These results support previously reported experiments demonstrating that the norepinephrine response to exercise in humans, performed at the same relative intensity before and after endurance training, increases only for exercise bouts performed at above 60% Vo2max (26). The increase in plasma glycerol levels was quite similar in fasted and fed states. This result fits with a previous study showing that the rate of whole body lipolysis (appearance of glycerol) during exercise was the same in trained and untrained subjects (27). In addition, it has been shown that training induces no changes in lipolysis and blood flow response in adipose tissue to epinephrine infusion in healthy men (28). After training, the greater increase in plasma epinephrine after feeding, which was more marked than before training, could be explained by plasma glucose values attained at the onset of exercise. In a previous investigation, Montain et al. (9) showed that trained men exhibited a marked decrease in plasma glucose concentration during exercise after a meal occurring 4 and 6 hours before exercise. The fall in plasma glucose concentration observed during exercise after the meal could be due both to a reduction in glucose appearance and an increase of glucose uptake by the muscle (11). It could also be related to an improvement of insulin sensitivity (29)(30). The fact that postmeal plasma insulin concentrations were lower than those measured before training for a quite similar decline in plasma glucose concentrations supports this comment. It has been shown that plasma insulin levels return to fasting levels within 2 hours after a meal in trained nonobese subjects (31). However, in that study, a drop in plasma glucose levels was observed when the exercise was performed 4 hours after a meal, and the authors suggest that the effects of the insulin elevation persist in certain tissues for some time after the plasma insulin levels have returned to normal. In summary, the increase in adrenal medulla secretion observed after a meal is probably linked to the fall in plasma glucose concentration related to a posttraining improvement of insulin sensitivity.
As found in a study conducted on trained and untrained nonobese subjects (32), our study shows that exercise tended to induce more marked lipid use in the fasted state than after feeding in overweight patients. We also found that lipid oxidation was higher after the training period. This result also fits with the results of Bergman and Brooks (32), showing that trained nonobese subjects oxidize more fat in the fasted condition than 3 hours after a breakfast. Interestingly, it was observed in our study that in fed conditions, training did not increase fat oxidation during exercise, whereas when exercise was performed after an overnight fast, training significantly improved lipid oxidation.
In summary, as expected, training improved biochemical and endocrine blood parameters. As previously reported in trained subjects (33), an important benefit of training is the lowering of circulating NEFA levels, which have been described as a predictive risk factor for sudden death (34). In addition, our data obtained from overweight male subjects show that when aiming at increasing oxidation of excess lipid by exercise, fasting conditions are more suitable than fed conditions. Our results also show that, in fasting conditions, overweight subjects easily carried out the exercise at an intensity of 50% Vo2max, and, moreover, no disruption in glucose homeostasis occurred. These results should be taken into account in informing obese subjects entering a training program in order to obtain a favorable effect of exercise on physical fitness and metabolic parameters.
This study was supported by the Commission of the European Communities RDT program and grants from Merck Laboratories and the Fondation pour la Recherche Médicale. The investigation was carried out in the Center of Clinical Investigation of Toulouse Hospital. The authors thank M. T. Canal and M. A. Marques for their contribution to the study.