KEY WORDS: Exercise intensity, substrate utilization, obesity, lactate threshold.

The ability to mobilize and utilize fat as a fuel is important for a variety of populations. For endurance athletes the strong relationship between the capacity to oxidize fatty acids and exercise performance is of interest (Holloszy and Coyle, 1984; Jansson and Kaijser, 1987). For overweight and obese subjects an increased rate of fat oxidation might be beneficial in order to reduce body weight (Jeukendrup and Achten, 2001). Exercise training programmes at the intensity eliciting a maximal rate of fat oxidation are therefore helpful to treat and prevent obesity and the metabolic syndrome and to increase the capacity of endurance athletes to oxidize fat.

Endurance training is known to increase the rate of fat oxidation during submaximal exercise at a given workload (Hurley et al., 1986; Martin et al., 1993; Phillips et al., 1996). Dériaz and colleagues (2001) in humans and Weber and colleagues (1993) in animals reported a positive correlation between maximal aerobic power (VO₂ max) and the highest rate of fat oxidation. These findings suggest that the ability to oxidize fatty acids is related to high levels of cardiorespiratory fitness. In addition, results of biopsy studies of both rat (Wolfe et al., 1990) and human muscle (Kiens et al., 1993) indicated that training induced increases in free fatty acid (FFA) binding proteins and mitochondrial density enhance the ability for FFA oxidation. The evidence is that endurance trained athletes are able to oxidize more fat at a given exercise intensity compared with untrained subjects.

In endurance trained people the rate of fat oxidation increases from low to moderate intensities (Romijn et al., 1993) and declines at exercise intensities of approximately 80 to 85% VO₂ max (Astorino, 2000). The intensity associated with the highest rate of fat oxidation is between 55 and 75% VO₂ max, shown in several recent studies (Romijn et al., 1993; Astorino, 2000; Romijn et al., 2000; Van Loon et al., 2001; Achten et al., 2002; Knechtle et al., 2004). This wide range of exercise intensities may have been a consequence of different study protocols, subject groups or type of exercise.

The highest rate of fat oxidation in sedentary, obese subjects is not well documented. Several lines of evidence indicate that obese subjects may have an impaired capacity to oxidize fat (Kim et al., 2000; Pérez-Martin et al., 2001) compared with trained individuals. However, Steffan and co-workers (1999) compared rates of fat oxidation in obese and normal weight women with similar VO₂ max values (ml·kg^-1·LBM^-1·min^-1) and found no difference in substrate use between the two groups. Also Ranneries and colleagues (1998) found no difference in fat oxidation between formerly obese women and normal weight women at 50% VO₂ max. Thus, it seems cardiorespiratory fitness level (defined as VO₂ max), rather than body composition influences the rate of fat oxidation.

Recommended training intensity at submaximal intensities is often given by percentages of maximal oxygen uptake (%VO₂ max) or heart rate (%HRmax). In athletes (Meyer et al., 1999; Weltman et al., 1999) and in obese subjects (Byrne and Hills, 2002) the proportion of peak or maximal cardiorespiratory capacity corresponds with wide ranges of exercise intensity as defined by individual lactate threshold (LT). Furthermore Achten and colleagues (2002) found a large between-subject variation for the maximal rate of fat oxidation expressed either in %VO₂ max or %HRmax. Consequently, relying on exercise intensity described by specific percentages of VO₂ max or HRmax, some individuals will be working well below and others well above the intensity that elicits the highest rate of fat oxidation. Thus, an individual determination of exercise intensities associated with a maximal rate of fat oxidation will ensure a more targeted and thus effective approach for exercise prescription (Jeukendrup and Achten, 2001).

A well known and established marker of an individual submaximal exercise criterion is the LT (Casaburi et al., 1995). Training at an intensity near the LT seems correlated with high rates of fat oxidation in athletes. Recently Knechtle and colleagues (2004) found in endurance athletes a relationship between the highest rate of fat oxidation and the LT in cycling but not in running. Achten and Jeukendrup (2004) showed a significant correlation between the intensity at which lactate concentration increased above baseline and the maximal rate of fat oxidation in endurance trained athletes. However, no previous study has investigated the relationship between the highest rate of fat oxidation and the LT in obese subjects. Due to the fact that studies concerning the relationship between fat oxidation and LT have only been performed with athletes, we included a group of highly trained athletes as a control group for our obese subjects in order to compare our results with the literature.

Therefore, a primary aim of this study was to determine the exercise intensity associated with the highest rate of fat oxidation in sedentary obese subjects. A secondary purpose was to compare the VO₂ at LT with the VO₂ (L·min^-1) at the intensity that elicits a maximal rate of fat oxidation in athletes and obese subjects using the same exercise protocol.

Subjects
Twenty endurance trained athletes (AT; 10 women, 10 men) and twenty sedentary, obese subjects (OB; 10 women, 10 men) were included in the study. The athletes were recruited from advertisements in a national sports journal, the obese subjects were recruited from circulated flyers and advertisements in the newsletters from the Swiss Foundation of Obesity. All of the athletes were either active triathletes (6 women, 7 men) or cyclists (4 women, 3 men) at either national or international level with a training background of at least five years. All obese subjects were sedentary, whereby sedentary was defined as exercising less than once per week for the previous 6 months. Obesity was defined as a body mass index (BMI) greater or equal than 30 kg·m^-2.

Prior to all testing procedures, trained and sedentary subjects completed a screening questionnaire regarding their medical and exercise histories. None of them were following either an energy-restricted diet, or using medications that affected energy metabolism. Metabolic and endocrine disorders were excluded by measuring fasting plasma lipoprotein lipids (triglycerides, total cholesterol, HDL cholesterol) and fasting blood glucose and insulin levels. The following criteria lead to exclusion: a) total blood cholesterol > 5.7 mmol·L^-1; b) triglycerides > 2.0 mmol·L^-1; c) glucose > 6.1 mmol·L^-1. After analysis of the screening questionnaire and the fasting blood samples, five obese subjects were excluded. From the 42 athletes who responded to the advertisement, 20 were selected who most closely matched the OB with respect to age. Prior to testing all participants gave their written informed consent after explanations of the experimental procedures and possible risks and benefits. All procedures were approved by the local ethics committee.

Maximal Exercise Testing
Initially, height and body weight were measured and BMI (kg·m^-2) was calculated. The body fat percentage (BF %) was determined according the equation of Deurenberg (1991). To assess VO_2peak, the subjects performed an incremental exercise test on a stationary cycle ergometer (ergoline 900^®, ergoline, Bitz, Germany). Women were tested without respecting their menstrual cycle. Obese subjects started at 40 W, athletes at 100 W. Workload was increased by 30 W at 3 min intervals until cessation of the test. During exercise, oxygen uptake (VO₂) and carbon dioxide production (VCO₂) were measured continuously (Oxycon Pro, Jaeger, Würzburg, Germany). Gas analyzers were calibrated prior to each test. Heart rate was recorded continuously by an electrocardiogram. At the end of every 3 min step, blood samples from the earlobe were collected in a 20 µl glass capillary to measure the concentration of lactate by an enzymatic method (Super GL ambulance, Ruhrtal Labor Technik, Möhnesee, Germany). Before each measurement of lactate the analyzer was calibrated with a 10 mmol·L^-1 lactate standard solution.

Determination of lactate threshold (LT)
The LT was determined in the maximal exercise test. According to Coyle and colleagues (1983), LT was identified as the VO₂ at which lactate increased 1 mmol·L^-1 above baseline, since this could be objectively determined in all subjects.

Submaximal testing protocol
Subjects completed submaximal exercise protocols, requiring them to exercise at three different workloads of 55, 65 and 75% VO_2peak in randomized order spread over two separate days (either one or two exercise bouts per day). Volunteers were advised to follow their normal diet, to avoid strenuous exercise the day before the test and to abstain from eating for 10 h before the submaximal tests. Each stage of the submaximal test lasted 20 min and was separated by at least 15 min of passive recovery. Oxygen consumption (VO₂) was measured continuously throughout the 20 min exercise bouts. Workload was adjusted in the first five minutes to reach the preset percentage of VO₂. Heart rate was measured continuously (Polar M52^®, Kempele, Finland). At the beginning and at the end of each stage, the concentration of blood lactate was measured. To ensure that gas exchange was stable at the onset of exercise and subjects had recovered from the previous bout of exercise, subjects rested until their RER was maintained around 0.80 and lactate concentration reached baseline values.

The reproducibility of the submaximal exercise stages was tested pre-study. Five healthy female and male volunteers (age: 24.3 ± 2 years, BMI: 23.6 ± 1.2 kg·m^-2) performed the maximal test and the submaximal exercise protocol twice within one week. The VO₂ at LT and the ventilatory responses at the three exercise intensities (VO₂ and VCO₂) did not differ between the two tests (Student's t-test). The coefficients of variation (CV) for RER during each of the tested intensities were respectively 2.6, 2.3 and 3.9%. The reliability of LT was assessed by the CV for VO₂ at LT. The CV was found to be 2.9%.

Indirect calorimetry and calculations
VO₂ and VCO₂ measures from the last 5 min of each exercise intensity (55, 65, 75% VO_2peak) were used to calculate rates of fat and carbohydrate oxidation. Fat and carbohydrate oxidation and energy expenditure were calculated using the stochiometric equations of Frayn (1983), which defined oxidation of carbohydrates (g·min^-1) as 4.55 x VCO₂ - 3.21 x VO₂ - 2.87 n and oxidation of fat (g·min^-1) as 1.67 x VO₂ - 1.67 x VCO₂ - 1.92 n. Nitrogen excretion rate (n) was assumed to be 135 µg·kg^-1·min^-1 in accordance with Carraro and colleagues (1990). Energy expenditure from fat and carbohydrate were converted to kJ·min^-1 by multiplying the oxidation rate of fat by 37 and the oxidation rate of carbohydrate by 16 using the Atwater (1909) general conversion factor. VO_2peak was expressed per kg body weight and kg lean body mass (LBM). Fat and carbohydrate oxidation rates were expressed as kJ·min^-1 per kg body mass (Figures 1 and 2).

Statistical analysis
All data from the maximal exercise test were reduced to group means. Comparisons between the athlete and the sedentary obese group for a single measurement (Age, BMI, LBM, VO_2peak, LT expressed in %VO_2peak, HR_max, and RER_max) were made with the Student's t-test for independent samples. ANOVAs with repeated measures were performed to detect statistically significant differences between intensity for each metabolic variable (rate of carbohydrate oxidation per kg body weight, rate of fat oxidation per kg body weight, percent fat oxidation of total energy oxidation, and total energy oxidation/consumption).

The strength of the relationship between VO₂ at lactate threshold and VO₂ eliciting the maximal rate of fat oxidation was assessed using the Pearson product moment correlation coefficient. All calculations were performed with SYSTAT (SYSTAT, Inc., Evanston, Illinois). Statistical significance was set at p < 0.05.

Subject characteristics and VO_2peak test
The anthropometric data of the subjects and physiological measures from the VO_2peak test are shown in Table 1 and 2 respectively. AT showed a significantly lower body mass (p < 0.01) and BMI (p<0.01) compared with OB. VO_2peak values expressed relative to body mass and expressed per kg of lean body mass were significantly higher in athletes compared with obese (p < 0.01).

Energy expenditure
The rate of energy expenditure increased with increasing intensity in AT and OB (p < 0.01) and showed significant differences between AT and OB in both genders (p < 0.01) (Figures 1 and 2). Absolute energy expenditure (kJ·min^-1) at a defined intensity was significantly higher in men than in women (p < 0.01). Total energy expenditure expressed per kg of lean body mass showed no statistical difference between sexes in both groups. Relative rates of fat oxidation expressed in percent of total energy oxidation were not significantly different between the three tested intensities (55, 65 and 75% VO_2peak) in either sex (women: 36.2 ± 15.2, 35.5 ± 12.9, 31.4 ± 12.3%; men: 32.3 ± 14.5, 34.5 ± 15.3, 27.8 ± 11.1%). At 55% VO_2peak female AT attained a significantly higher percentage of fat to total energy expenditure at 75% VO_2peak (range 24-38%) than obese women (p = 0.02), whereas men showed no significant difference (AT, 28.6 ± 11.8 %; OB, 26.9 ± 11.9%).

Substrate oxidation
Within the tested intensities, AT (men and women) demonstrated their highest rate of fat oxidation at an intensity of 75% VO_2peak, sedentary obese subjects reached a maximal rate of fat oxidation at 65% VO_2peak. Female athletes showed a significantly higher absolute rate of fat oxidation (kJ·min^-1) than obese women (p < 0.01). Also normalized rate of fat oxidation, expressed in kg body mass (kJ·min^-1·kg^-1), was found to be significantly higher for female AT compared with OB (Figure 1). There was no significant difference between obese men and male athletes in fat utilization and normalized rate of fat oxidation at the intensity that elicited highest rates of fat oxidation (AT: 19.72 ± 8.05 kJ·min^-1·kg^-1 at 75% VO_2peak; OB: 17.57 ± 8.24 kJ·min^-1·kg^-1 at 65% VO_2peak). Carbohydrate utilization per kg body mass increased in athletes and obese with increasing workload (Figure 1 and 2). Between men, AT had a significantly higher absolute (p < 0.01) and normalized rate of carbohydrate oxidation (p < 0.01) than OB (Figure 2). Women AT and OB showed no significant difference in absolute and normalized carbohydrate utilization (Figure 1).

Lactate threshold (LT)
Female and male AT reached LT at a significantly higher intensity expressed in %Wmax (p < 0.01) and in %VO_2peak (p < 0.01) than obese women and men (Figure 3). Expressed in %Wmax, LT was near 35% Wmax in OB and similar for women and men. AT had their LT at approximately 74 and 77% VO_2peak, women and men respectively. There was a modest relationship between VO₂ at LT and VO₂ (L·min^-1) at a work rate eliciting a maximum rate of fat oxidation in OB (women: r = 0.43, ns; men: r = 0.32, ns). In contrast for AT there was a strong relationship between VO₂ at LT and VO₂ (L·min^-1) at a work rate eliciting a high rate of fat oxidation (women: r = 0.67, p<0.05; men: r = 0.75, p=0.01).

• Within the tested intensities of 55, 65 and 75% VO_2peak athletes reached higher rates of fat oxidation at higher relative work rates compared with obese subjects.
• We found in obese women and men the intensity of the highest rate of fat oxidation at 65% VO_2peak.
• Between the lactate threshold and the intensity eliciting a high rate of fat oxidation a significant correlation was found in athletes but not in obese subjects.