Pediatric Exercise Science, 2007, 19, 82-92 © 2007 Human Kinetics, Inc.

Aerobic-Exercise Training Improves Ventilatory Efficiency in Overweight Children Christopher Kaufman, Aaron S. Kelly, Dan R. Kaiser, Julia Steinberger, and Donald R. Dengel The objective of this study was to investigate the effect of an 8-week aerobicexercise training program on ventilatory threshold and ventilatory efficiency in overweight children. Twenty overweight children (BMI > 85th percentile) performed a graded cycle exercise test at baseline and were then randomly assigned to 8 weeks of stationary cycling (n = 10) or a nonexercising control group (n = 10). Ventilatory variables were examined at ventilatory threshold (VT), which was determined via the Dmax method. After 8 weeks, significant improvements occurred in the exercise group compared with the control group for oxygen uptake at VT (exercise = 1.03 ± 0.13 to 1.32 ± 0.12 L/min vs. control = 1.20 ± 0.10 to 1.11 ± 0.10 L/min, p < .05) and ventilatory equivalent of carbon dioxide (VE/VCO2) at VT (exercise = 32.8 ± 0.80 to 31.0 ± 0.53 vs. control = 30.3 ± 0.88 to 31.7 ± 0.91, p < .05). Aerobic-exercise training might help reverse the decrements in cardiopulmonary function observed over time in overweight children. Key Words: pediatric, fitness, Dmax method

In the past 2 decades there has been an increased awareness of the negative consequences of high caloric intake, poor nutrition, and lack of physical activity in Americaʼs youth. These factors, along with others, have resulted in an increased prevalence of overweight and obese children (2). Currently, 22% of children are overweight (2) and 31.5% of all children in the United States are at risk for being overweight (5). It is widely accepted that physical fitness in overweight children is worse (or less) than in normal-weight children (3). Indeed, obesity is frequently associated with decreased physical capacity and premature fatigue. Obesity can cause extrinsic compression of the chest and impairment of its mechanics at rest and during exercise, especially in children whose physical development of pertinent respiratory muscles and pulmonary mechanics are not fully realized. Inselma et al. (7) noted significant decrements in parameters of ventilation (VE; e.g., ventilatory equivalent of oxygen [VE/VO2] and ventilatory equivalent of Kaufman and Dengel are with the School of Kinesiology, Kelly and Steinberger are with the Dept. of Pediatrics, and Kaiser is with the Dept. of Medicine, University of Minnesota, Minneapolis, MN 55455.

82

09Kaufman(82).indd 82

1/7/07 4:49:23 PM

Ventilatory Efficiency in Overweight Children

83

carbon dioxide [VE/VCO2] during a graded exercise test in obese as compared with normal-weight children). Others, however, have failed to find differences between groups (11,12). It is physiologically plausible to presume that if overweight children are limited in ventilatory function and efficiency during an acute bout of exercise (whether because of the actual excess weight on their chest and trunk or simply because of a deconditioned state), exercise training might help improve that function. However, the extent to which chronic exercise training in overweight children can improve ventilatory threshold (VT) and ventilatory-efficiency parameters (e.g., VE/VCO2 at VT and slope of relationship of VE and VCO2 throughout exercise test) during exercise has not been previously studied. Therefore, the primary aim of this study was to describe in detail the effect of 8 weeks of aerobic cycling exercise on ventilatory parameters in overweight children and to compare these changes with an age-matched group of overweight children who remained inactive over the same 8-week period. The secondary aim was to examine the usefulness and accuracy of the Dmax method for determining VT in overweight children. Despite the aforementioned negative effects of obesity, we hypothesized that aerobic exercise would improve VT and ventilatory-efficiency parameters.

Methods Participant Population A total of 20 children were recruited from the greater Minneapolis–St. Paul, Minnesota, area to participate in the study. The criterion for overweight was an age-adjusted body-mass index (BMI) greater than the 85th percentile (17). Each group had 3 participants with a BMI between the 85th and 95th percentile and 7 participants with a BMI greater than the 95th percentile. Overweight participants were randomized into two groups (exercise group, n = 10, and control group, n = 10) after baseline testing and were matched according to age, gender, and BMI. Participants must have been healthy overweight (BMI > 85th percentile) children between the ages of 8 and 14 years. “Healthy” was defined as being free of any major pulmonary, cardiovascular, or metabolic disease states, such as hypertension or diabetes. All of the children and their parents or guardians provided written informed assent and consent. The study procedures were reviewed and approved by the University of Minnesotaʼs institutional review board.

Graded Exercise Test Participants arrived in a fasted state in the morning at the University of Minnesota Laboratory of Physiological Hygiene and Exercise Science. They were informed to not exercise strenuously the day before testing. Furthermore, they were told not to consume caffeine during the 8 hr before testing. Verification of adherence to the aforementioned study requirements was obtained before each testing session. If a participant did not adhere to the requirements, the visit was rescheduled. All participants adhered to the requirements for all visits, so no visits were rescheduled. Cardiorespiratory fitness was assessed using a graded-intensity test on an

09Kaufman(82).indd 83

1/7/07 4:49:24 PM

84

Kaufman et al.

electromagnetically braked stationary cycle ergometer (ERG 401, Dimeq Corporation, Berlin, Germany) starting at 20 W and increasing 20 W every 2 min until volitional exhaustion. Expired oxygen and carbon-dioxide concentrations and volumes were collected and analyzed using a MedGraphics CPX-D metabolic cart (Medical Graphics Corp., St. Paul, MN). Before the graded exercise test, each participantʼs body composition was determined by dual-energy X-ray absorptiometry (DXA; Prodigy, 3M, Madison, WI; software version 6.7) at the University of Minnesota General Clinical Research Center. The scans were performed using a fast transverse speed mode. The scanner was calibrated monthly with known phantoms, and no machine drift was noted during the period of study. Height and weight were measured with a stadiometer and standard electronic scale (Model 5002, Scale-Tronix Inc., Wheaton, IL), respectively. Tanner growth stage for pubertal development was determined by a trained pediatrician. Repeat measurements were performed on all participants after the 8-week period.

Ventilatory-Threshold and Ventilatory-Efficiency Measurement VT was obtained from the graded-exercise-test data using the Dmax method. We calculated the point that yielded the maximal perpendicular distance from a line drawn from the lowest and highest data points to a curve representing ventilation (on y axis) as a function of oxygen consumption (VO2; on x axis), as outlined by Cheng et al. (1). An example is shown in Figure 1. In addition, we examined all 40 exercise tests (20 at baseline and 20 after 8 weeks) to determine whether the Dmax method, as compared with the ventilatory-equivalent method (VT = increase

Figure 1 — Example of ventilatory-threshold determination via Dmax method using breath-by-breath data.

09Kaufman(82).indd 84

1/7/07 4:49:24 PM

Ventilatory Efficiency in Overweight Children

85

in VE/VO2 without concomitant increase in VE/VCO2) and V-slope method (VT = loss of linear relation between VCO2 and VO2), was indeed a valid method of determining an individualʼs VT. Ventilatory efficiency during exercise was determined by examining the ventilatory equivalent for carbon dioxide (VE/VCO2) at VT and the slope of the relationship between VE and VCO2 from the first minute after the start of the graded exercise test to VT. The VE-to-VCO2 slope was calculated as the slope (a) of VE (L/min) versus VCO2 (L/min) with the equation VE = a VCO2 + b. VE (y axis) and VCO2 (x axis) were graphed in Microsoft Excel® (Microsoft Inc., Seattle, WA), and the linear-regression function was used to provide the VE equation. The researcher determining VT and ventilatory efficiency was originally blinded to the group and time status to reduce the potential for bias.

Exercise-Training Protocol The exercise-training protocol and its effect on vascular function and other metabolic variables has been described previously (9). Briefly, the training protocol was carried out in a supervised and structured environment (Laboratory of Physiological Hygiene and Exercise Science). Participants in the exercise group trained on a stationary cycle ergometer four times per week starting at 50–60% of VO2peak (as determined from baseline graded exercise test) for 30 min per session. Intensity or duration of exercise was increased weekly until participants were exercising at 70–80% of VO2peak for 50 min during the final 2 weeks. Exercise intensity was monitored by heart rates corresponding to the desired percentage of VO2peak. Attendance rate was recorded for each training session and at the conclusion of the study; all participants had attended at least 80% of all scheduled training sessions. Participants in the control group were instructed to maintain current levels of physical activity and did not participate in a structured exercise program.

Statistical Analyses All statistical procedures were carried out using SPSS version 13.0 (SPSS Inc., Chicago, IL). Comparisons of variables between groups before and after 8 weeks were analyzed using a two-way repeated-measures analysis of variance (ANOVA), which produced p values for the effects of group status (exercise vs. control), time (baseline vs. post), and interaction (Group × Time). Results were reported as means ± standard error of the mean (SEM). Repeated-measures ANOVA was used, and intraclass correlations were calculated to compare the Dmax method of determining VT with the ventilatory equivalent and V-slope methods in all participants at baseline and postintervention. Bland–Altman plots were generated to calculate the agreement between methods by examining the differences between two of the three VT-determination methods as compared with the average VT (from the same two methods used in calculation of differences) using GraphPad Prism version 4.0 (GraphPad Software Inc., San Diego, CA). An alpha level of .05 was used to denote statistical significance.

09Kaufman(82).indd 85

1/7/07 4:49:24 PM

86

Kaufman et al.

Results Baseline characteristics of each group are presented in Table 1. No significant differences existed between groups at baseline. After 8 weeks of aerobic cycle training, the exercise group had significant improvements compared with the control group in VO2peak (exercise = 1.66 ± 0.13 to 1.84 ± 0.14 L/min vs. control = 1.83 ± 0.12 to 1.60 ± 0.13 L/min; Group × Time = p < .05) and VO2 at VT (Figure 2). No significant differences were noted, however, for the group and time effect (p > .05) for VO2peak and VO2 at VT. When VT was expressed as a percentage of VO2peak there was no significant difference between groups across time (exercise = 62.1 ± 5.6 to 71.9 ± 2.6% VO2peak vs. control = 65.7 ± 4.4 to 69.4 ± 3.8% VO2peak; p > .05). No significant differences were found within or between groups across time for height and Tanner score. Furthermore, no significant changes in weight (exercise = 67.9 ± 5.3 to 69.0 ± 5.5 kg vs. control = 73.4 ± 8.8 to 74.3 ± 8.9 kg; p > .05), BMI (exercise = 29.6 ± 2.0 to 29.6 ± 2.1 kg/m2 vs. control = 30.4 ± 2.3 to 30.4 ± 2.2 kg/m2; p > .05), and percentage body fat (exercise = 43.6 ± 2.0 to 43.5 ± 2.1% vs. control = 44.6 ± 2.5 to 45.3 ± 2.4%; p > .05) were observed within or between groups over the 8-week period. The 8-week training period resulted in significant improvements in the exercise group as compared with the control group for the measures of ventilatory efficiency. The slope of the relationship between VE and VCO2 during the exercise test to VT improved significantly for the exercise group as compared to the control group (Exercise = 0.11 ± 0.01 to 0.086 ± 0.01 vs. control = 0.09 ± 0.01 to 0.11 ± 0.01, p < .05). No significant differences were noted for the group and time effect for

Table 1 Baseline Characteristics of Exercise (n = 10) and Control Groups (n = 10)

Age (years)

Exercise

Control

p

10.8 ± 0.6

11.0 ± 0.7

.83

Gender(M/F)

4/6

4/6

Tanner Stage

1.9 ± 0.3

2.3 ± 0.4

.46

Height (cm)

151.2 ± 4.10

152.7 ± 3.80

.80

Weight (kg)

67.9 ± 5.3

73.4 ± 8.8

.60

2

BMI (kg/m )

29.6 ± 2.0

30.4 ± 2.3

.77

Body fat (%)

43.6 ± 2.0

44.6 ± 2.5

.75

VO2peak (L/min)

1.66 ± 0.14

1.83 ± 0.12

.36

VO2 at VT (L/min)

1.02 ± 0.1

1.2 ± 0.1

.31

VE/VO2 at VT

30.8 ± 0.9

29.7 ± 1.4

.55

VE/VCO2 at VT

31.0 ± 0.5

31.7 ± 0.9

.48

Note. Data are presented as M ± SEM. BMI = body-mass index; VT = ventilatory threshold; VE/VO2 = ventilatory equivalent of oxygen; VE/VCO2 = ventilatory equivalent of carbon dioxide.

09Kaufman(82).indd 86

1/7/07 4:49:25 PM

Ventilatory Efficiency in Overweight Children

87

the VE and VCO2 slope (p < .05). Furthermore, VE/VCO2 at VT improved in the exercise group as compared with the control group (Figure 3). Because the exercise group had significant improvements in ventilatory efficiency and volume of O2 at VT attributed to the training effect, their cardiopulmonary systems were able to accommodate increased energy demands with decreased VE (42.3 ± 4.1 to 30.7 ±

Figure 2 — Changes (M ± SEM) in oxygen consumption at ventilatory threshold (VT) over time at baseline (solid bars) and posttraining (open bars).

Figure 3 — Changes (M ± SEM) in ventilatory equivalent of carbon dioxide at ventilatory threshold (VT) over time at baseline (solid bars) and posttraining (open bars).

09Kaufman(82).indd 87

1/7/07 4:49:25 PM

88

Kaufman et al.

4.6 L/min, p < .05) and VCO2 (1.3 ± 0.13 to 1.01 ± 0.15 L/min, p < .05) at VT as compared with the control group (VE = 34.1 ± 3.7 to 36.1 ± 2.96 L/min; VCO2 = 1.09 ± 0.12 to 1.17 ± 0.10 L/min). When comparing the Dmax method with the ventilatory equivalent and V-slope methods for determining VT, we found no statistically significant differences, and intraclass correlations ranged from r = .94 to .98. Bland–Altman plots for comparisons between methods are presented in Figure 4. When comparing the Dmax and

Figure 4 — Bland–Altman plot of ventilatory-threshold (VT) determination with (A) Dmax method vs. ventilatory-equivalent method (V.Equiv.) and (B) Dmax method vs. V-slope method.

09Kaufman(82).indd 88

1/7/07 4:49:26 PM

Ventilatory Efficiency in Overweight Children

89

ventilatory-equivalent methods, the mean difference and standard deviation were –0.02 ± 0.06 L/min, with the 95% limit of agreement being –0.13 to 0.09 L/min (N = 34). The mean difference between the Dmax and V-slope method was 0.003 ± 0.09, with 95% limit of agreement being –0.16 to 0.17 L/min (N = 34). However, we were not able to determine the VT in 8 of the 40 exercise tests (20%) using the conventional methods (ventilatory equivalent and V-slope). No differences were noted in the effort or performance during the exercise tests between children whose VT could be determined and those whose VT could not be determined with the ventilatory-equivalent and V-slope methods.

Discussion The overall findings of this study show that 8 weeks of aerobic cycle training improves cardiopulmonary function and ventilatory efficiency in overweight children as compared with participants in an overweight control group who maintained their sedentary lifestyles. Numerous studies have compared cardiopulmonary function and ventilatory efficiency between normal-weight and overweight children (12,14,15,21), and the overall findings indicate reduced cardiovascular function and ventilatory efficiency (increased VE, VCO2, and slope of the VE–VCO2 curve) throughout exercise in overweight children. The improvements in ventilatory parameters at VT as a result of exercise training that were observed in our study are consistent with those in prior studies that examined normal-weight children (6,13). Although our study lacked a normal-weight control group, the results are promising because we showed the ability of aerobic-exercise training to help reverse further decrements in ventilatory efficiency caused by sustained sedentary behavior. Indeed, it is worth noting that the control group had consistently negative changes (i.e., decreased VO2peak and VO2 at VT, as well as increase in the slope of VE–VCO2 curve) for almost all variables studied over the 8-week period. Therefore, the interaction term in the two-way repeated-measures ANOVA was significant without having a significant time or group effect. We further analyzed the data for the determination of the delta change (post–baseline) for each outcome variable in each participant using independent t tests (exercise vs. control) and found significant differences between groups for VO2peak, VO2 at VT, VE and VCO2 slope, and VE/VCO2 at VT. We contend that these findings show the benefit of regular aerobic exercise in reversing the continued decline in ventilatory efficiency parameters, such as those observed in the control group. A factor that might explain the negative physiologic changes for the control group was the seasonal time of the study. The baseline testing occurred at the end of the school year, and the 8-week period was during the summer months when the children were not attending school. It is our speculation, based on conversations with the children and their parents, that the children in the control group might have became more sedentary during the summer months because no organized physical education classes or school activities were being held on a regular basis. If this is indeed what occurred, it likely explains much of the decrements in the variables studied. This is very speculative, however, and further study documenting changes in daily physical activity during interventional periods for control groups is needed.

09Kaufman(82).indd 89

1/7/07 4:49:26 PM

90

Kaufman et al.

The analyses comparing Dmax VT determination with other conventional methods indicated no differences between methods and good agreement. Previous studies in active adults (1,8) found no statistical differences between the Dmax method and the conventional linear-regression method. Other methods of determining VT (e.g., VE/VO2 or VE/VCO2 plotted against each other or over VO2) have inherent subjective estimations, and, therefore, the Dmax method appears to be more valid in identifying each individualʼs VT. Its primary advantage over conventional methods was that it provided an objective and reliable method for threshold determination. Considering that several studies (4,12,14) have reported difficulties using conventional methods to determine VT (ventilatory-equivalent or V-slope method) in children and adolescents, the Dmax method appears to be a useful alternative method for measuring of VT. Our analyses support the usefulness of the Dmax method for determining VT in children and show its benefit for instances when conventional methodology is limited. Whipp and Davis (20) described the effect of obesity limiting proper ventilation as (a) increased metabolic cost and, therefore, ventilatory requirement to do work; (b) interference of normal chest-wall motion as a result of abdominal obesity and a higher metabolic cost of ventilation as a result of higher breathing frequency; and (c) ventilatory insufficiency as a result of those two factors. Our findings support the notion that despite the negative effects of obesity on breathing mechanics, overweight children can improve ventilatory efficiency through a moderate aerobic-training program. More interesting is that our data suggest that these changes occurred despite no change in body fatness. This suggests that some of the limitations of obesity on ventilatory efficiency might not be solely attributable to the additional body weight but could be a manifestation of the untrained state that is typically associated with obesity. Although not directly measured, the possible physiologic mechanisms involved in the training effect in the exercise group resulting in higher VO2peak, VO2 at VT and improved ventilatory efficiency would be, but are not limited to, improved cardiac output (18); oxygen-transport capacity (hemoglobin concentration); and peripheral vascular adaptations, such as muscle-capillary density (19), O2 diffusion and O2 extraction (16), and pulmonary diffusion capacity and alveolar-ventilation-to-perfusion ratios (10). Consequently, the exercise group was able to utilize a larger portion of O2 before becoming anaerobic (e.g., before VT), so their pulmonary efficiency increased because they required less ventilation, possibly as a result of enhanced CO2 buffering. A training effect was observed over a rather short period of time (8 weeks), further supporting the concept that a structured exercise program carried out over a longer period of time could substantially help reverse the cardiopulmonary decrements noted in overweight children. The study lacked a proper normal-weight control group, and thus the interpretations of the results of this study are limited. Furthermore, there might have been a confounding effect on the results because the exercise group trained using the same mode of exercise (cycling) that was used during the baseline and posttesting sessions. The training cycle ergometers and the test ergometer were different models, however, and as a result of the controlled nature of this mode of exercise, any increase in mechanical efficiency because of training would have been minimal. The relatively small sample (N = 20) could potentially limit the statistical power, and, therefore, additional studies with larger sample sizes are needed.

09Kaufman(82).indd 90

1/7/07 4:49:27 PM

Ventilatory Efficiency in Overweight Children

91

Acknowledgments This research was supported in part by Minnesota Obesity Center Grant #: I P30 DK 50456-08 (D.R.K.), American Heart Association Pre-Doctoral Grant #: 0315213Z (A.S.K.), and GCRC: M01-RR00400, General Clinical Research Center Program, NCRR/NIH.

References 1. Cheng, B., H. Kuipers, A.C. Snyder, H.A. Keizer, A. Jeukendrup, and M. Hesselink. A new approach for the determination of ventilatory and lactate thresholds. Int. J. Sports Med. 13:518-522, 1992. 2. Goran, M.I. Metabolic precursors and effects of obesity in children: a decade of progress. Am. J. Clin. Nutr. 73:158-171, 2000. 3. Goran, M.I., D.A. Fields, G.R. Hunter, S.L. Herd, and R.L. Weinsier. Total body fat does not influence maximal aerobic capacity. Int. J. Obes. Relat. Metab. Disord. 24:841-848, 2000. 4. Hebestreit, H., B. Staschen, and A. Hebestreit. Ventilatory threshold: a useful method to determine aerobic fitness in children? Med. Sci. Sports Exerc. 32(11):1964-1969, 2000. 5. Hedley, A.A., C.L. Ogden, C.L. Johnson, M.D. Carroll, L.R. Curtin, and K.M. Flegal. Overweight and obesity among US children, adolescents, and adults, 1999–2002. JAMA. 291:2847-2850, 2004. 6. Hoffor, A.S., A.C. Harrison, and P.A. Kirk. Anaerobic threshold alterations caused by interval training in 11-year-olds. J. Sports Med. Phys. Fitness. 30(1):53-56, 1990. 7. Inselma, L., A. Milanese, and A. Deurloo. Effect of obesity on pulmonary function in children. Pediatr Pulmonol. 16:130-137, 1993. 8. Kara, M., H. Gokbel, and C.S. Bediz. A combined method for estimating ventilatory threshold. J. Sports Med. Phys. Fitness. 39(1):16-19, 1999. 9. Kelly, A.S., R.J. Wetzsteon, D.R. Kaiser, J. Steinberger, A.J. Bank, and D.R. Dengel. Inflammation, insulin, and endothelial function in overweight children and adolescents: the role of exercise. J. Pediatr. 145(6):731-736, 2004. 10. Koch, G., and B.O. Eriksson. Effect of physical training on anatomical R-L shunt at rest and pulmonary diffusing capacity during near-maximal exercise in boys 11–13 years old. Scand. J. Clin. Lab. Invest. 31(1):95-103, 1973. 11. Lazarus, R., G. Colditz, C.S. Berkey, and F.E. Speizer. Effects of body fat on ventilatory function in children and adolescents: cross-sectional findings from a random population sample of school children. Pediatr. Pulmonol. 24:187-194, 1997. 12. Marinov, B., S. Kostianev, and T. Turnovska. Ventilatory efficiency and rate of perceived exertion in obese and non-obese children performing standardized exercise. Clin. Physiol. Funct. Imaging. 22(4):254-260, 2002. 13. Nourry, C., F. Deruelle, C. Guinhouva, et al. High-intensity intermittent running training improves pulmonary function and alters exercise breathing pattern in children. Eur. J. Appl. Physiol. 94(4):415-423, 2005. 14. Reybrouck, T., L. Mertens, D. Schepers, J. Vinckx, and M. Gewillig. Assessment of cardiorespiratory exercise function in obese children and adolescents by body massindependent parameters. Eur. J. Appl. Physiol. 75:478-483, 1997. 15. Reybrouck, T., M. Weymans, J. Vinckx, H. Stijns, and M. Vanderschueren-Lodeweyckx. Cardiorespiratory function during exercise in obese children. Acta Paediatr. Scand. 76(2):342-348, 1987. 16. Roca, J., A.G. Agusti, A. Alonso, et al. Effects of training on muscle O2 transport at VO2max. J. Appl. Physiol. 73(3):1067-1076, 1992.

09Kaufman(82).indd 91

1/7/07 4:49:27 PM

92

Kaufman et al.

17. Rosner, B., R. Prineas, J. Loggie, and S.R. Daniels. Percentiles for body mass index in US children 5 to 17 years of age. J. Pediatr. 132(2):211-222, 1998. 18. Saltin, B., G. Blomqvist, J.H. Mitchell, R.L. Johnson, Jr., K. Wildenthal, and C.B. Chapman. Response to exercise after bed rest and after training. Circulation. 38(5 Suppl):VII, 1-78, 1968. 19. Shono, N., H. Urata, B. Saltin, et al. Effects of low intensity aerobic training on skeletal muscle capillary and blood lipoprotein profiles. J. Atheroscler. Thromb. 9(1):78-85, 2002. 20. Whipp, B.J., and J.A. Davis. The ventilatory stress of exercise in obesity. Am. Rev. Respir. Dis. 129(2 Pt 2):S90-S92, 1984. 21. Zanconato, S., E. Baraldi, P. Santuz, et al. Gas exchange during exercise in obese children. Eur. J. Pediatr. 148(7):614-617, 1989.

09Kaufman(82).indd 92

1/7/07 4:49:27 PM