Share this article on:

Physical training, lifestyle education, and coronary risk factors in obese girls


Medicine & Science in Sports & Exercise: January 1996 - Volume 28 - Issue 1 - p 19-23
Clinical Sciences: Clinical Investigations

The effects of supervised physical training (PT) and lifestyle education(LSE) on risk factors for coronary artery disease and non-insulin-dependent diabetes mellitus were compared in obese 7- to 11-yr-old black girls. The subjects were divided into two groups. The PT group (N = 12) completed a 5-d·wk-1, 10-wk, aerobic training program; and the LSE group participated in weekly lifestyle discussions to improve exercise and eating habits. The PT group showed a significant increase in aerobic fitness(P < 0.05) and decrease in percent body fat (P < 0.05), while the LSE group declined significantly more in dietary energy and percent of energy from fat (P < 0.05). Fasting insulin did not change significantly. The LSE group declined significantly more than the PT group in glucose (P < 0.05), and glycohemoglobin declined from baseline in both groups (P < 0.05). Lipid changes were similar in the two groups: total cholesterol/high density lipoprotein cholesterol(P < 0.01) and triglycerides (P < 0.05) declined, the low density lipoprotein (LDL)/apoproteinB ratio increased (which indicates a decrease in small dense LDL) (P < 0.05) and lipoprotein(a) increased (P < 0.05). Thus, the interventions were similarly effective in improving some diabetogenic and atherogenic factors, perhaps through different pathways; i.e., the PT improved fitness and fatness, while the LSE improved diet. Exercise and diet-induced changes in lipoprotein(a) require further investigation.

Georgia Prevention Institute, Department of Pediatrics; and Department of Medicine; Medical College of Georgia, Augusta, GA 30912

Submitted for publication October 1994.

Accepted for publication August 1995.

This work was supported by grants from the Medical College of Georgia Research Institute and Department of Pediatrics, and the American Heart Association-Georgia Affiliate.

Address for correspondence: Bernard Gutin, Ph.D., Professor of Pediatrics and Physiology, Georgia Prevention Institute, Medical College of Georgia, Augusta, GA 30912-3710; E-mail:

In both adults (3) and children(8), body fatness is related to risk factors for coronary artery disease (CAD) and non-insulin-dependent diabetes mellitus (NIDDM). Since regular exercise can play an important role in the treatment of childhood obesity (7), it is important to develop and evaluate different ways of helping obese children to exercise. One approach is supervised physical training (PT) without dietary intervention. Another approach is lifestyle education (LSE), which provides children with the knowledge and skills needed to incorporate exercise into their everyday lives; such education may also help the children to improve their eating habits(5). Black females are at especially high risk of obesity(18). Thus, the effects of supervised PT and LSE on CAD/NIDDM risk factors were investigated in obese black girls.

Back to Top | Article Outline


Obese black girls 7-11 yr of age (mean = 9.2 yr) were recruited via flyers sent to parents of children attending schools near Medical College of Georgia(MCG) and by advertisements in hospital newspapers. The main enticement was that they would receive without charge an intervention that might improve their body composition and health. In addition, the children were offered $100 for completion of all study requirements. All girls and parents gave informed consent. They agreed to accept assignment to either the formal PT program or to participate in weekly lifestyle discussions designed to improve their exercise and eating habits. Girls who attended the schools from which a bus could pick them up to transport them to the PT center were assigned to the PT group, and the others were assigned to the LSE group. All the schools were in similar socioeconomic neighborhoods.

Of the 13 girls who started the 5-d·wk-1, 10-wk PT program, one did not actively participate in the PT, despite constant prodding, and was exerting a disruptive influence on the other children. Thus, after 3 wk she was asked to discontinue attending. One girl lived outside of the bus area, but promised to provide her own transportation for 3 d·wk-1; she attended 82% of her scheduled sessions. Each of the 11 girls who took the bus had an attendance rate for the 5-d·wk-1 program that was greater than 93%. For the LSE sessions, 12 girls started, but two dropped out within the first 2 wk. For the other 10, the attendance was 95%. The PT group received no dietary information.

Before and after the 10-wk intervention period, the participants reported in a fasting state at 8 a.m. and a blood sample was drawn. Lipids, lipoproteins, apoproteins, insulin, and glucose were assayed as previously described (8). In the interests of parsimony, this study focused on several key indices of CAD risk; i.e., the ratio of total cholesterol (TC) to high density lipoprotein cholesterol (HDLC), which has been found to be a better predictor of CAD mortality than any other single lipoprotein measurement (6), and triglycerides (TG). Also investigated were the effects of the interventions on two other CAD-associated lipid factors whose relationship to activity and fitness is not well-known: the ratio of low density lipoprotein cholesterol (LDLC) to apoproteinB (apoB), an index of small dense LDL (21); and lipoprotein(a)(19). Lipoprotein(a) [Lp(a)] was determined by an enzyme linked immunosorbent assay (Strategic Diagnostics, Newark, DE). This kit included standards (6 levels) containing Lp(a) in human plasma in buffered solution, which were used to plot a standard curve from which controls and unknowns were calculated. An external source of Lp(a) controls (2 levels), assayed by Northwest Lipid Research Center in Seattle, WA, was also included in the controls. All Lp(a) analyses were done in batches within 1 month and all standards, controls, and plasma samples were run in duplicate. The reproducibility coefficient using a subsample (N = 19) was 0.93. In addition to fasting glucose and insulin values, glycohemoglobin (GHb) was measured to gain insight into the average blood glucose over the previous 4-6 wk (20). GHb was determined in duplicate by affinity chromatography using kits from Isolab (Akron, OH). Our normal range is 3.7-7.1%, with intra- and interassay variation of 3.0 and 4.2%, respectively.

Percent body fat was measured with dual x-ray absorptiometry (DXA) (Hologic QDR-2000, Waltham, MA). This involved a small radiation exposure of 0.5 mRem, about 33% of the amount received in a cross-country airplane trip. Since there are few published data on the DXA for measurement of body composition in children, reliability was determined in 42 children involved in another study(8). In replicate measurements, with repositioning between scans, the mean values were almost identical (and not significantly different), and the reproducibility coefficient was very high (r > 0.99). More complete information about the DXA measurement is provided elsewhere(9).

Diet was assessed with 3-d dietary records, including one weekend day. Only five and seven children returned complete baseline and post-intervention records for the LSE and PT groups, respectively; therefore, these results must be interpreted cautiously. The dietary records were analyzed using Nutritionist III software (N-Squared Computing, Salem OR)(16). The key dietary factors estimated were total energy intake and percent of energy from fat. Free-living physical activity (PA) was measured with a 7-d recall modified for use with children(23). Overall PA included the light, moderate, and hard categories of PA. To derive an estimate of vigorous PA, only the last two categories were used. The energy expenditure data were expressed per unit of body weight to take into account the fact that heavier children use more energy to move their bodies a given amount; thus, the values may be considered“movement units.”

Aerobic fitness was measured on a treadmill. The children walked at 2.5 mph for 3 min. Then the grade was raised to 2% for 1 min and to 4% for another 3 min period. Heart rate (HR) was measured during the last minute of these work stages with an electrocardiograph (Space Labs, Redmond, WA) and the values at the 0% and 4% grades were averaged to provide a stable index of aerobic fitness; a lower submaximal HR indicates better fitness. After the 4% workload, the grade was incremented by 2% every minute, encouragement was given to help elicit a maximal effort, and oxygen consumption was measured with indirect calorimetry. However, a clear maximal effort was not elicited in several children, as indicated by maximal HR less than 190 bpm or respiratory exchange ratios less than 1.00. Thus, submaximal HR, which has previously been shown to be a sensitive indicator of improved aerobic fitness in children(22), was used as the index of aerobic fitness.

The PT program was designed to exceed minimal criteria for a training effect: i.e., intensity above 60% of maximal HR, for 20 min, three times·wk-1 (1). The aerobic phase of the session was comprised of activities such as: group aerobics, cycle ergometry, minitramp, walking/jogging around the room or on a treadmill, circuit training with light weights, rope jumping, bench stepping, and using a cross-country ski machine. To assure that girls went beyond the minimal criteria, the girls were encouraged to keep their HR over 70% of maximal for the entire aerobic phase of the PT session (approximately 30 min), and to attend 5 d·wk-1. HR was monitored during two sessions·wk-1 for each girl, using a portable monitor (Polar Vantage XL, Port Washington, NY). The average time for which the girls kept their HR in their target zones was 28 min, and the average HR during this phase of the PT sessions was 163 bpm, well over the minimal levels prescribed. Thus, the PT group clearly obtained an adequate training stimulus.

The LSE group provided their own transportation and came to MCG one evening·wk-1 for 60-90 min. Behavioral techniques used for self-management of PA and eating habits were adapted for the 7-11 yr old girls(5). These included written record keeping, goal setting, relapse prevention, and reinforcement techniques.

Between group comparisons of preintervention values were made witht-tests for unpaired comparisons. Within groups analyses of variance(ANOVA) showed significant changes from pre- to post-intervention within the separate groups. Differences between group changes were compared with treatments by time, repeated measures ANOVAs; a significant treatment by time interaction indicated that the groups responded differently. When the interaction was not significant, a significant F-ratio for the time factor indicated that both groups taken together showed a significant change from baseline.

Back to Top | Article Outline


The groups were similar at baseline, as shown by the nonsignificant differences in baseline means for all variables except vigorous PA; this difference may suggest that some of the children who agreed to participate in the controlled PT may have been more inclined to vigorous PA at baseline.

Table 1 shows the baseline and post-intervention values for fitness, PA, fatness, and diet. Submaximal HR declined significantly in the PT group, indicating improved aerobic fitness, while it remained stable in the LSE group. Although overall PA increased somewhat in both groups, the changes did not achieve significance. However, both groups increased significantly in vigorous PA.

At baseline, the mean body mass indices (weight in kg·height-2 in meters) for the PT and LSE groups (27.9 and 26.8, respectively) were greater than the 95th percentile for black 10-yr-old girls(15). Percent body fat declined significantly in the PT group, but the decline in the LSE group was not significant. A complete description of total and regional body composition changes is provided elsewhere (9).

For those who provided dietary data, the LSE group declined significantly more than the PT group in intake of total energy and percent of energy from fat. No significant changes occurred in the percent of energy from saturated fat (data not shown).

Table 2 shows that fasting insulin concentrations did not change significantly and that glucose concentrations declined significantly more in the LSE group than in the PT group. GHb declined significantly in both groups taken together. None of the lipid changes differed significantly for the groups. For the groups taken together, significant reductions were found for TG and the TC/HDLC ratio. No significant change was seen for the separate components of the ratio, although the increase in HDLC approached significance (P = 0.08). The significant increase in the LDLC/apoB ratio indicated a reduction in small dense LDL. LP(a) increased significantly across both groups.

Back to Top | Article Outline


The PT group improved in aerobic fitness and body composition, suggesting that supervised PT was effective in improving these factors. Neither group increased overall PA significantly. For the PT group, this suggests that they may have compensated for the PT by reducing their activity at other times during the week. With respect to vigorous activity, both groups increased significantly and similarly. Any conclusions about changes in vigorous activity must be tempered in light of the higher baseline values for the PT group; i.e., it is possible that the change for the LSE group actually constituted a greater proportion of their baseline activity. Nonetheless, the lack of improvement in fitness and body composition for the LSE group implies that their increased PA may not have been sufficiently systematic to elicit a clear training effect in 10 wk. When the children were questioned about their free-living PA 4 wk after termination of the interventions, the PT subjects had dropped off rather sharply, implying that they had come to depend on the formal PT for their exercise, while the LSE subjects maintained the higher level; therefore, the long-term effect of these two approaches to increasing PA and fitness needs to be studied. Since the PT group received no instruction concerning diet while some of the lifestyle sessions did focus on diet, it is appropriate that the limited dietary data suggested that the LSE group decreased more in total energy and fat intake.

Compared with the normal range in our laboratory (7-179 pmol·l-1), the baseline values of insulin were quite high for these obese children, suggesting that they were insulin resistant(14). The fasting glucose values and GHb were well within the normal range, suggesting that the elevated insulin was maintaining normal glucose disposal. Contrary to our expectations, no significant changes occurred in fasting insulin values as a result of the interventions; in fact, the mean values increased slightly. Perhaps a longer period of PT(17) and/or a more substantial change in body composition(13) are required for reductions to be produced. The LSE group declined more than the PT group in fasting glucose values, perhaps because of the combination of diet change and increased informal PA. Taken together, both groups reduced GHb significantly from the upper portion to the middle of the normal range. Since this was accomplished without a significant change in insulin concentration, it suggests that insulin sensitivity was improved. Although little is known about the effects of PT or diet on Ghb in nondiabetic children, regular PT reduces GHb levels in insulin-dependent diabetics (20).

Both groups changed similarly in lipid profiles. The reductions in TC/HDLC and TG, along with the increase in the LDLC/apoB ratio, suggest an improved lipid profile. However, the increases in Lp(a) suggest an unfavorable change. Little is known about the relation of exercise and Lp(a). A recent paper(12) reported no association of aerobic fitness and body composition to Lp(a) in adults; we obtained a similar result in 7- to 11-yr-old children (11). One paper reported that Lp(a), measured 24 h after the last bout of exercise, was elevated in high-level athletes, perhaps because it plays a role in wound repair following vigorous exercise (4). However, an inverse relationship between aerobic fitness and Lp(a) in adolescents with insulin-dependent diabetes mellitus has been reported (2), and a letter to the editor(10) reported a decrease in Lp(a) after 8 d of cross-country skiing. More research is needed to clarify the acute and chronic effects of exercise on Lp(a).

There are several limitations to this preliminary study which require that conclusions be drawn very cautiously, including the small number of subjects and the incomplete dietary data. The fact that the subjects were paid to encourage compliance also limits the extent to which the results can be generalized to typical clinical settings in which payment is not provided.

In conclusion, it appears that supervised PT, without dietary intervention, helped the obese girls to improve their aerobic fitness and body composition, while lifestyle education helped the girls to improve their diets. Thus, the similar changes in the other CAD/NIDDM risk factors may have been elicited through different pathways, suggesting that a combination of PT and lifestyle education may be more effective than either alone.

Back to Top | Article Outline


1. American College of Sports Medicine. Guidelines for Exercise Testing and Prescription. Baltimore: Williams & Wilkins, 1995, pp. 153-176.
2. Austin, A., V. Warty, J. Janosky, and S. Arslanian. The relationship of physical fitness to lipid and lipoprotein(a) levels in adolescents with IDDM. Diabetes Care 16:421-425, 1993.
3. Björntorp, P. Visceral obesity: a “Civilization Syndrome”. Obesity Res. 1:206-222, 1993.
4. Cardoso G., C. Posadas, O. Orvananos, et al. Long distance runners and bodybuilders exhibit elevated plasma lipoprotein(a). Chem. Phys. Lipids 67-68:207-221, 1994.
5. Epstein, L. and R. Wing. Behavioral treatment of childhood obesity. Psych. Bull. 101:331-42. 1987.
6. Grover, S., C. Palmer, and L. Coupal. Serum lipid screening to identify high-risk individuals for coronary death. The results of the Lipid Research Clinics prevalence cohort. Arch. Intern. Med. 154:679-84, 1994.
7. Gutin, B. and T. Manos. Physical activity in the prevention of childhood obesity. In: Prevention and treatment of childhood obesity. Ann. N. Y. Acad. Sci. 699:115-126, 1993.
8. Gutin, B., S. Islam, T. Manos, N. Cucuzzo, C. Smith, and M. Stachura. The relations of percent body fat and maximal aerobic capacity to risk factors for atherosclerosis and diabetes in black and white seven- to eleven-year-old children. J. Pediatr. 125:847-852, 1994.
9. Gutin, B., N. Cucuzzo, S. Islam, C. Smith, R. Moffatt, and D. Pargman. Physical training improves body composition of black obese 7- to 11-year-old girls. Obesity Res. 3:305-312, 1995.
10. Hellsten, G., K. Boman, G. Hallmans, and G. Dahlen. Lipids and endurance physical activity (Letter). Atherosclerosis 75:93-94, 1989.
11. Islam, S., B. Gutin, C. Smith, F. Treiber, and M. Kamboh. Association of apolipoprotein(a) phenotypes in children with family history of premature coronary artery disease. Arterioscler. Thromb 14:1609-1616, 1994.
12. Israel, R., M. Sullivan, R. Marks, R. Cayton, and T. Chenier. Relationship between cardiorespiratory fitness and lipoprotein(a) in men and women. Med. Sci. Sports Exerc. 26:425-431, 1994.
13. Knip, M. and O. Nuutinen. Long-term effects of weight reduction on serum lipids and plasma insulin in obese children. Am. J. Clin. Nutr. 57:490-493, 1993.
14. Laakso, M. How good a marker is insulin level for insulin resistance? Am. J. Epidemiol. 137:959-965, 1993.
15. Must, W., G. Dallal, and W. Dietz. Reference data for obesity: 85th and 95th percentiles of body mass index (wt/ht2) and triceps skinfold thickness. Am. J. Clin. Nutr. 53:839-846, 1991.
16. Nieman, D., D. Butterworth, C. Nieman, and R. Lee. Comparison of six microcomputer dietary analysis systems with the USDA nutrient database for standard reference. J. Am. Diet. Assoc. 92:48-57, 1992.
17. Oshida, Y., K. Yamanouchi, S. Hayamizu, and Y. Sato. Long-term mild jogging increases insulin action despite no influence on body mass or VO2 max. J. Appl. Physiol. 66:2206-2210, 1989.
18. Pi-Sunyer, F. Obesity and diabetes in blacks.Diabetes Care 13:1144-1149, 1990.
19. Rhoads, G., G. Dahlen, K. Berg, N. Morton, and A. Dannenbert. Lp(a) lipoprotein as a risk factor for myocardial infarction.J.A.M.A. 256:2540-2544, 1986.
20. Rowland, T. Exercise and Children's Health. Champaign, IL: Human Kinetics, 1990, pp. 223-224.
21. Sniderman, A., S. Shapiro, D. Marpole, B. Skinner, B. Teng, and P. Kwiterovich. Association of coronary atherosclerosis with hyperapobetalipoproteinemia (increased protein but normal cholesterol levels in human low density (β) lipoprotein). Proc. Natl. Acad. Sci. USA 77:604-608, 1980.
22. Stewart K. and Gutin B. Effects of physical training on cardiorespiratory fitness in children. Res. Q. 47:110-120, 1976.
23. Wallace, J., T. McKenzie, and P. Nader. Observed vs recalled exercise behavior: a validation of a seven day exercise recall for boys 11 to 13 years old. Res. Q. Exerc. Sport 56:161-165, 1985.


©1996The American College of Sports Medicine