Human Genomics Laboratory, Pennington Biomedical Research Center, 6400 Perkins Road, Baton Rouge, LA 70808. E-mail: email@example.com
Objective: To determine whether “low-intensity” exercise (walking) and “high-intensity” exercise (aerobic dance), when added to a weight loss diet, have different effects on coronary heart disease (CHD) risk factors and physical fitness.
Research Methods and Procedures: Ninety obese women were divided into diet only (DO), diet plus walking (DW), and diet plus aerobic dance (DA) groups. DXA was used to evaluate segmental body composition. Leg-extension strength and maximal oxygen uptake (V̇o2max) were the indicators of physical fitness. Blood pressure, lipoproteins, and fasting glucose were used as indices for CHD risk factors. These items were measured before and after a 14-week intervention period.
Results: Whole-body plus all segmental fat masses were significantly reduced (p < 0.001). Reductions in whole-body and lower-limb fat- and bone-free masses were significantly less (p < 0.01) in the DA group (−1.5 and −0.1 kg, respectively) compared with the DO (−2.1 and −0.4 kg, respectively) and DW (−2.5 and −0.5 kg, respectively) groups. Improvements in leg-extension strength and V̇o2max were significantly greater (p < 0.05) in the DA group compared with the DO group. The CHD risk factors clearly improved (p < 0.05) within each group. Reductions in low density lipoprotein-cholesterol and fasting glucose were significantly greater (p < 0.05) in the DA group compared with the DO and DW groups.
Discussion: Adding higher intensity aerobic dance to a weight-loss diet program may help maintain fat- and bone-free mass and may be more effective in improving CHD risk factors compared with low-intensity walking.
Because weight-loss treatment benefits the health of obese individuals (1, 2, 3), obese patients with risk factors for coronary heart disease (CHD)1 should be treated by an appropriate weight-loss program. Unfortunately, losing weight through diet alone includes a decline in fat-free mass during the intervention period (3) and induces an attenuation of fat oxidation after the intervention period (4), which may contribute to weight regain.
To prevent weight regain after weight loss, it has been recommended that obesity be treated with a combination of physical exercise and reduced energy intake (5). Walking is an aerobic exercise that is accessible to many segments of the population because it is convenient, easy and low cost; because exercise intensity during walking is generally below the lactate threshold, walking is considered a typical low-intensity exercise. Aerobic dance, a popular exercise, particularly among women, is representative of moderate- to vigorous-intensity exercise. We confirmed the various effects and safety of combining aerobic dance with diet in the treatment of obesity (6).
Leutholtz et al. (7) investigated the effects of exercise intensity (40% vs. 60% of heart-rate reserve) on body composition during energy restriction. Because those investigators found that exercising at 60% of the heart-rate reservewas no better at changing the body composition than exercising at 40% of the heart-rate reserve when the total volume of exercise training was controlled, the authors concluded that exercise intensity did not influence body composition changes. To our knowledge, four prospective studies without weight-loss treatments (8, 9, 10, 11) have demonstrated the effects of exercise intensity on plasma lipoproteins, cardiovascular events, CHD, and risk of type 2 diabetes. Of these studies, three showed that exercise intensity was associated with reduced risks (8, 9, 10), and a study by Kraus et al. (11) revealed that improvement in plasma lipoproteins were related to the amount of physical activity and not to the exercise intensity. However, few details have been reported on the effects of exercise intensity on improving CHD risk factors and physical fitness during a weight-loss program.
With these points in mind, we examined in this study whether there would be any differences of effects between additive walking (“low-intensity” exercise) and aerobic dance (“high-intensity” exercise) to diet on CHD risk factors, physical fitness, segmental body composition, and abdominal fatness in moderately obese and middle-aged Japanese women. To clarify the effects of exercise itself, a diet-only group was set as the control. Based on the findings of Janssen and Ross (3), the primary hypothesis of our study was that diet plus exercise would be more effective than diet alone for improving these parameters, particularly body fat change. We also hypothesized, based on the findings of Leutholtz et al. (7), that no difference would be observed in any measurement variables between the walking and aerobic dance groups.
Research Methods and Procedures
One hundred twenty sedentary women were recruited through advertisements in local newspapers. Of this group, 90 sedentary women, ages 34 to 66 years, were chosen as subjects if they met the following criteria for intra-abdominal fat obesity: 1) exercise-induced energy expenditure <60 min/wk; 2) BMI >25 kg/m2; and 3) intra-abdominal fat area (IFA) at the level of the umbilicus >100 cm2. These criteria were set according to the guidelines for intra-abdominal fat obesity (obesity disease) in the Japan Society for the Study of Obesity (12). We determined through medical history and physical examination that none of the subjects smoked, had concomitant renal, hepatic, or cardiac disease, or were being treated with drugs such as β-blockers, which could affect the variables of the study. To increase subject adherence to the weight-loss programs during the intervention period, subjects were assigned to one of three groups taking personal lifestyle (occupations, daily schedules, etc.) into account: 1) a diet-only (DO) group (n = 35); 2) a diet and walking (DW) group (n = 22); or 3) a diet and aerobic dance (DA) group (n = 33). All subjects completed the study requirements without dropping out, and assays and measurements were carried out before and after the 14-week (98-day) intervention period. Examiners were blinded to the treatment group. The aim and design of the study were explained to each subject before they gave their written informed consent.
Body weight was measured to the nearest 0.1 kg using a digital scale, height was measured to the nearest 0.1 cm using a wall-mounted stadiometer, and BMI was calculated as weight (in kilograms) divided by height squared (meters squared). Waist circumference was measured to the nearest 0.1 cm at the level of the umbilicus with subjects in the standing position.
Body Composition by DXA
DXA (DPX-L, Lunar, Madison, WI) was used to evaluate segmental body composition (upper limbs, trunk, lower limbs, and whole body), which consisted of fat mass (FM) and fat- and bone-free mass (FBFM). Transverse scans were used to measure FM and FBFM, and pixels of soft tissue were used to calculate the ratio of mass attenuation coefficients at 40 to 50 keV (low energy) and 80 to 100 keV (high energy), using software version 1.3Z.
Abdominal Adipose Tissue Area by Computed Tomography
The IFA and subcutaneous fat area (SFA) were measured at the level of the umbilicus (L4–L5) using computed tomography (CT) scans (SCT-6800TX, Shimadzu, Tokyo, Japan) performed on subjects in the supine position. The IFA and SFA were calculated using a computer software program (FatScan, N2system, Osaka, Japan) (13). First, a region of the subcutaneous fat (SF) layer was defined by tracing its contour on each scan, and the range of CT values (in Hounsfield units) for fat tissue was calculated. Total fat area was determined by delineating the surface having a mean CT value plus or minus 2 standard deviations, and the IFA was measured by drawing a line within the muscle wall surrounding the abdominal cavity. The SFA was then calculated by subtracting the IFA from the total fat area, and the IFA to SFA ratio was determined. The intra-class correlation for repeated IFA determinations in our laboratory was 0.99.
Systolic and diastolic blood pressures were taken from the right arm using a mercury manometer after the subjects rested at least 20 minutes in a sitting position. Cuff sizes were selected based on upper-arm girth and length.
Biochemical Assays of Blood
A blood sample of ∼10 mL was drawn from each subject after an overnight fast. Serum total cholesterol and triglycerides were determined enzymatically, serum high-density lipoprotein-cholesterol (HDL-C) was measured by the heparin-manganese precipitation method and fasting plasma glucose (FPG) was assayed by a glucose oxidase method. Serum low-density lipoprotein-cholesterol (LDL-C) was estimated according to the equation of Friedewald et al. (14): LDL-C = total cholesterol − (HDL-C + triglycerides/5).
The subjects held a dynamometer (GRIP-D5101, Takei, Niigata, Japan) in one hand with the arm extended at the side and squeezed with maximum force, which was measured in kilograms. Two trials were performed on both right and left hands with a brief rest between trials. The average of all four trials was used as the measure of a subject's performance.
Leg-extension strength was measured while the subject leaned back in a chair, pulling against a strain gauge with legs extended (GF-300, Yagami, Aichi, Japan). The best of two trials was recorded in kilograms.
Maximal Oxygen Uptake
Maximal oxygen uptake (V̇o2max) was determined during a graded exercise test using a cycling ergometer (818E, Monark, Stockholm, Sweden). After a warm up, the subject started with a workload of 90 kg · m/min, which was increased by 90 kg · m/min each minute until volitional exhaustion occurred. Pulmonary ventilation and gas exchange were measured breath-by-breath with an online data acquisition system (Oxycon alpha System, Mijnhardt, Breda, The Netherlands).
Diet and Exercise Regimens
The subjects were instructed to take a well-balanced supplemental food product (MicroDiet, SunnyHealth, Nagano, Japan) every day. The diet was developed for very-low-energy diets (170 kilocalories per package) and comprised protein, carbohydrates, fat, various amino acids, vitamins, and minerals. Two other meals per day consisting of 240 kcal of protein, 480 kcal of carbohydrates, and 240 kcal of fat were allowed. The subjects also kept daily food diaries during the 14-week intervention period and learned about proper daily nutrition through weekly lectures and counseling by dieticians.
Aerobic Dance Protocol
In addition to restricting energy intake, the subjects from the DA group performed a bench-stepping exercise 3 days/wk for 45 min per session supervised in the hospital by two or three physical trainers. The bench-stepping exercise was a combination exercise of low-impact aerobic dance and stepping with a step bench (10 to 20 cm high). The exercise started with basic steps at the beginning of the intervention period and progressed to more advanced lunge steps for the latter half of the period (15, 16). Subjects were instructed to perform the aerobic dance at a level that raised their heart rate 70% to 85%, corresponding to their V̇o2max. The target Borg's scale (ratings of perceived exertion) (17) ranged from 13 (fairly hard) to 17 (very hard). The oxygen costs of the bench-stepping exercise were measured three times (at the first, middle, and last month) during the study by the aforementioned metabolic measurement system.
In addition to restricted energy intake, the subjects from the DW group were instructed to walk every day around their houses for more than 30 minutes per session. In a pilot study, we determined that energy expenditure during the aerobic dance program was ∼350 kcal per session (1050 kcal per week) (data not shown). In short, daily energy expenditure, on average, was estimated as 150 kcal (1050 divided by 7). A 30-minute walk was determined as corresponding with the 150-kcal energy expenditure when the following equation was used:
The target exercise intensity was set at a level that raised subjects’ heart rate 40% to 50%, corresponding to their V̇o2max, or 9 (very light) to 11 (light) of the Borg's scale (17). Subjects measured their heart rates by palpation while walking and recorded the duration (in minutes) and intensity (heart rate or the Borg's scale) of the walking each session. They were instructed how to measure their heart rates by palpation while walking every week. At that time, portable heart-rate monitors monitored the subjects’ heart rates. Hence, the subjects could check the validity of their heart rates by palpation. According to subjects’ diary entries (body weight, heart rate, and walking time), energy expenditure while walking was calculated every day by each subject using the equation above.
Values are expressed as mean ± SD in Tables 1 and 2 and as mean ± SE in Figure 1. Differences among the DO, DW, and DA groups before the intervention were examined using one-way ANOVA, and a Scheffe's post hoc test was applied when these results were significant (p < 0.05). Paired t tests were used to assess changes within groups for all variables, and Bonferonni adjustments (p < 0.0167, p = 0.05/number of groups) were used to interpret all t test results. An analysis of covariance (ANCOVA) was used, with absolute changes acting as dependent variables, and changes in body weight and energy expenditures during exercises acting as the covariates. The mean values obtained from the ANCOVAs were used to assess treatment differences. Data were analyzed with the Statistical Analysis System (Release 6.12; SAS, Cary, NC).
Attendance at the exercise programs (40 sessions) averaged 92% (range, 83% to 100%) for the subjects from the DA group. The average maximal heart rate was 169 ± 7 beats/min during the graded exercise test in the combined groups. No difference was found in maximal heart rate among the three groups. The average V̇o2 during the bench-stepping exercise was 19.5 ± 4.7 mL/kg/min (first month, 17.3 ± 2.6 mL/kg/min; second month, 19.8 ± 3.1 mL/kg/min; last month, 21.3 ± 3.6 mL/kg/min), which corresponded to 81.5 ± 14.7% (first month, 72.4 ± 9.8%; second month 82.9 ± 10.3%; last month, 89.1 ± 11.2%) of V̇o2max. The frequency of walking was 5.6 ± 1.4 day/wk. Walking duration averaged 187 ± 130 min/wk. Because we asked subjects to walk for more than 210 minutes per week (more than 30 minutes every day), the level of adherence was estimated as 89% (187 of 210 minutes). The average heart rate during walking was 115 ± 12 beats/min, which corresponded to 46.5 ± 4.4% of V̇o2max. The exercise intensity (percentage of V̇o2max) was significantly higher in the DA group than in the DW group. The mean energy expenditures during the walking and the bench-stepping exercise were calculated as 849 ± 354 kcal/wk and 1166 ± 130 kcal/wk, respectively.
Table 1 shows subjects’ measurement variables by treatment group at the beginning of the study. No difference was found in any variables among the three groups. Table 2 presents changes in the measurement variables by treatment group, with results of paired t tests and ANCOVA shown on the right side of the table.
Significant (p < 0.001) reductions were observed in all variables (body weight, BMI, and waist circumference) within all groups. For all variables, reductions were greater in the DA group compared with the DO group.
The FM, FBFM, and percent FM were significantly reduced (p < 0.001) in each body segment, with the exception of the lower-limb FBFM in the DA group (p > 0.0167). Results of the ANCOVA revealed that reductions in trunk and whole-body FMs were significantly greater (p < 0.01) in the DA (−4.3 ± 1.4 and −8.0 ± 2.7 kg) and DW (−3.7 ± 1.3 and −6.6 ± 2.2 kg) groups compared with the DO (−2.6 ± 1.4 and −5.0 ± 2.3 kg) group. Moreover, reductions in whole-body and lower-limb FBFMs were significantly less (p < 0.01) in the DA (−1.5 ± 1.2 and −0.1 ± 0.5 kg) group compared with the DO (−2.1 ± 1.1 and −0.4 ± 0.4 kg) and DW (−2.5 ± 1.4 and −0.5 ± 0.6 kg) groups. Proportions of the whole-body FM and FBFM that accounted for the reduction in body weight were 64.9% and 24.7% in the DO group, 72.5% and 25.3% in the DWgroup, and 78.4% and 13.7% in the DA group (see Figure 1). The proportion of FBFM reduction was smaller in the DA group compared with the DO and DW groups, but statistical significance was not found.
Significant (p < 0.001) reductions were observed in IFA, SFA, and the IFA-to-SFA ratio within all groups. However, there was no difference among the groups in the absolute change of any variable.
Hand grip strength did not change in any group, but there was significant improvement (p < 0.01) in leg-extension strength for the DA group (17.3%) and V̇o2max for all groups (10.9%, 17.0%, and 24.7% for DO, DW, and DA, respectively). Improvement of V̇o2max was significantly larger (p < 0.05) in the DA group compared with the DO group.
Blood Pressure and Blood Assay
Except for HDL-C, the CHD risk factors clearly improved (p < 0.05) within all groups. Reductions in LDL-C and FPG were significantly greater (p < 0.05) in the DA group compared with the DO and DW groups, even after adjusting for energy expenditure during exercise and change in body weight.
Recently, several studies have indicated that aerobic exercises (e.g., walking, stair climbing, stationary cycling, etc.), which are prescribed as treatments for obesity, do not always play an important role in body-weight reduction (18, 19, 20, 21). In some studies (20, 21), only a 1% (∼1 kg) per year weight reduction was observed during an intervention period consisting of 1500 to 2000 kcal per week of aerobic exercise without diet change. Mertens et al. (21) reported that energy intake increased during the intervention period by 8% compared with before the intervention. Leon et al. (22) observed a reduction in physical activities other than the prescribed exercise program. These authors concluded that the exercise program without diet failed to improve CHD risk factors (22).
In response to a 7.7- to 10.2-kg weight loss, segmental body composition and CHD risk factors were clearly changed in all groups. We found that 1) reductions in truncal and whole-body FMs were significantly greater in the combination groups (DW and DA) compared with the DO group, 2) no significant difference in FM was found between the combination groups, 3) reductions in FBFMs of the lower limbs and whole-body segments were significantly less in the DA group compared with the DO and DW groups, 4) increases in leg-extension strength and V̇o2max were significantly greater in the DA group compared with the DW and DO groups, and 5) improvements in LDL-C and FPG were significantly greater in the DA group compared with the DO and DW groups.
In a weight-loss intervention program for obese and middle-aged women, Janssen and Ross (3) observed that the reduction in total FM was greater in the aerobic exercise group (24%) than in the DO group (19%). van Aggel-Leijssen et al. (4) verified that, although a combination of low-intensity exercise and diet offered no advantages for FM reduction over a DO weight-loss program, adding exercise to a reduced-calorie diet counteracts the decline in fat oxidation that occurs in response to weight loss. The reduction in FM seen in the DA and DW groups may be attributed to the restraint on the decline in the fat oxidation in the combination groups.
Our second finding of no significant difference in FMs between the two exercise groups supports two previous studies. Leutholtz et al. (7) found that exercise intensity was not associated with whole-body FM reduction during energy restriction. Grediagin et al. (23) compared two groups performing either low-intensity (50% of V̇o2max) or high-intensity (80% of V̇o2max) exercises without diet modification and found no significant difference in whole-body FM change between the two groups. These results may indicate that, independent of diet, exercise intensity does not influence FM changes.
Our third finding of a considerably reduced loss of FBFM in the DA group suggests that high-intensity exercise contributes to the maintenance of FBFM. This finding seems to be emphasized by the data in Figure 1, despite the fact that statistical significance was not achieved. Growth hormone was reportedly secreted when exercise was performed at >50% of V̇o2max (24), which means that walking (46.5 ± 4.4% of V̇o2max) may be of insufficient intensity to induce growth hormone secretion. In addition, the exercise intensity may influence motor-unit activation (25): only slow-twitch fibers (type I) were activated when exercise was performed at <40% of V̇o2max, but at >80% of V̇o2max, fast-twitch fatigable fibers (type II B) were also mobilized (25). Type II B fibers hypertrophy selectively and predominantly compared with type I fibers (26). Consequently, the bench-stepping exercise (81.5 ± 14.7% of V̇o2max) may have helped maintain FBFM, particularly in the lower limbs, which were used the most during the bench stepping.
Despite the significant reduction in lower-limb FBFM (−3.7%) of the DW group, leg-extension strength (+4.0%) was unchanged. The leg-extension strength might be maintained by increasing in both nerve impulse frequency (action potential) and synchronicity of motor unit activation through walking programs; however, the physiological mechanisms cannot be explained by the data from this study. Several studies have confirmed that aerobic capacity improves through walking (+24%) (21) or a combination of walking and various other aerobic exercises (+8% to +21%) (20, 27, 28), which suggests that a low-intensity exercise, such as walking, could effectively improve an obese individual's aerobic capacity. In addition, according to the American College of Sports Medicine's (6th Edition) guideline (29), low-fit or deconditioned individuals can increase their cardio-respiratory fitness with exercise intensity of 40% to 49% of heart-rate reserve. In this study, the exercise intensity while walking was 46.5% of V̇o2max, and an increase rate of V̇o2max in this group was 17% (baseline, 24.7 mL/kg/min; posttreatment, 28.9 mL/kg/min). These results agree with previous studies. However, it should be noted that there was no difference in the V̇o2max improvement between the DW and DO groups. Hence, we could not conclude that cardio-respiratory fitness was significantly improved through walking in subjects in the DW group.
Our fifth finding that LDL-C and FPG had a significantly greater improvement in the DA group suggests that slowing the loss of FBFM is important for improving FPG and LDL-C during a weight-loss program. Further studies on how maintaining segmental FBFM with a combination of diet and resistance training affects changes in LDL-C, FPG, fasting insulin, glucose, and insulin responses during oral glucose tolerance tests are needed to validate this finding.
There are some limitations of this study. First, sample size was relatively small. Insufficient sample size may be concomitant with a type II error. Second, the age range of subjects was wide (34 to 66 years), and subjects included both premenopausal and postmenopausal women. Because estrogens may independently influence CHD risk factors (30), these factors might partly preclude our definitive conclusions. Third, to increase subject adherence to the weight-loss programs during the intervention period, subjects were not randomized to the treatments. However, no difference was found in any variables among the three groups at baseline (see Table 1). This suggests that, if at all, influence of subject assignment without randomization on the measurement variables would be little. Fourth, mode of exercise and frequency of exercise, as well as intensity of exercise, varied. The DA subjects took steps without advancing, although the bench-stepping exercise sessions and their motions (mode of exercise) were very similar to walking. With regard to variation of exercise frequency, it is known that the amount of physical activity (energy expenditure) is more important for improving CHD risk factors than the frequency of exercise (29, 31). In this study, although there was no significant difference in the energy expenditure per week between the DW and DA groups, a 30% larger energy expenditure was estimated in the DA group compared with the DW group. Because our data were analyzed using ANCOVA with energy expenditure as a covariate, influence of difference in energy expenditure on CHD risk factors was excluded in the results of this study.
In summary, our first hypothesis is supported by the significantly greater decreases in FM and percent FM seen in the two diet-plus-exercise groups compared with the DOgroup. However, the DA group had more favorable effects with regard to improving physical fitness and CHD risk factors and maintaining FBFM compared with the DW group, which is in opposition to our second hypothesis. Our data raises the possibility that additive exercise performed at a relatively higher intensity (aerobic dance) contributes to the maintenance of FBFM and may be more effective for improving physical fitness and CHD risk factors during weight loss compared with walking. This information would be helpful for developing an exercise program that takes the patient's clinical and physical status, as well as her/his exercise preference into consideration.
This study was supported, in part, by Human Beings in the Ecosystem, Tsukuba Advanced Research Alliance (TARA), and the University of Tsukuba.