Journal Logo

BASIC SCIENCES: Epidemiology

Fatness, Fitness, and Cardiovascular Disease Risk Factors in Children and Adolescents


Author Information
Medicine & Science in Sports & Exercise: August 2007 - Volume 39 - Issue 8 - p 1251-1256
doi: 10.1249/MSS.0b013e318064c8b0
  • Free


It is well known that overweight is associated with cardiovascular disease (CVD) morbidity and mortality in adults (28). However, it has been suggested that cardiorespiratory fitness modifies the relationship between overweight and CVD morbidity and mortality (4,25). Indeed, Blair and colleagues have consistently shown that cardiorespiratory fitness reduces the health consequences of high levels of body fatness in adults (the fat-but-fit hypothesis) (4,25,35).

Given the current pediatric obesity epidemic (17) and concern for adverse CVD risk factors among overweight youth (15), we previously have examined the fat-fit hypothesis in adolescents (10,12). Both studies show some evidence that within the low- and high-body mass index (BMI) categories, a higher level of cardiorespiratory fitness was associated with a better CVD risk-factor profile. However, a major limitation in these studies was the use of the median split to cross-tabulate subjects into fat-fit groups. The median split approach does not allow for examination of the data by clinical cut point or recommendations (i.e., adverse versus normal), which may facilitate the utility of such cut points. In addition, the BMI was used as the index of adiposity, and we were unable to express cardiorespiratory fitness as maximal oxygen consumption (V˙O2max). In the present study, we attempted to overcome these limitations by using health-related cut points for estimated percent body fat and V˙O2max. Hence, the purpose of this study was to further explore the fat-but-fit hypothesis by using health-related cut points in a large, nationally representative sample of youth from the Australian Health and Fitness Survey. We hypothesized that a higher level of cardiorespiratory fitness would positively influence the CVD risk-factor profile within fatness groups, particularly those with higher levels of fatness.


Study design and subjects.

This is a secondary data analysis of the Australian Health and Fitness Survey (AHFS). The study design and procedures have been described in detail elsewhere (2,6) but will be summarized here. The AHFS was a cross-sectional population study of approximately 8500 schoolchildren 7-15 yr of age from 109 schools that assessed health-related fitness and self-reported health behaviors. The design of the survey was a two-stage probability sample. The first stage was the selection of schools, and the second stage consisted of a random sample of males and females of each year of age from the total school enrollment. School was not included in the analysis as a clustering factor, because the sampling procedure (five females and five males at each year level from each school) and the relatively large number of clustering units (schools) relative to sample size minimize the impact of clustering. The overall response rate was 67.5% for those participating in the field and technical-measures portion of the study. The study was approved by the local ethics committee, and parental informed written consent and child assent were obtained before participation in the study. All procedures were in accordance with those outlined by the Declaration of Helsinki. In the present analysis, 1615 (860 males and 755 females) 9-, 12-, and 15-yr-olds have been included on the basis of complete data for blood chemistries, 1.6-km-run time, and anthropometry. The analyses were restricted to these age strata because funding limitations prevented more costly procedures such as blood chemistry from being administered to the entire survey sample. The particular ages were chosen to represent prepubertal, peripubertal, and postpubertal stages of maturation.

Body fatness.

Skinfold thicknesses were measured by standard procedures (26) with Holtain calipers at five anatomical sites on the right side of the body. The following skinfolds were measured to the nearest 0.5 mm: triceps, biceps, subscapular, suprailiac, and midabdominal. The mean of triplicate trials at each site was used in analyses. Percent body fat (% fat) was calculated using the equations of Slaughter (33) from the triceps and subscapular sites. We have previously shown that this equation correlates highly with % fat measured by DXA (9). Waist circumference (WC) was measured in the standing position at the level of the umbilicus to the nearest 0.1 cm using a constant-tension tape. Data collectors were trained in anthropometric techniques by the same project director, and a member of the steering committee made site visits to ensure that the standardized protocol was being followed during data collection.

Cardiorespiratory fitness.

Cardiorespiratory fitness was determined by a 1.6-km run and used in the calculation of estimated maximal oxygen consumption (V˙O2max). The timed distance test consisted of having the subject run on a marked course. Subjects were advised on proper pacing strategy and motivated to give their best effort. The equation used to predict V˙O2max (mL·kg−1·min−1) from the 1.6-km run was based on work by Cureton et al. (8). The equation was based on a large sample of children and adolescents and uses chronological age, gender, body mass index, and time elapsed for the prediction of V˙O2max (r = 0.72, SEE = 4.8 mL·kg−1·min−1).

CVD risk factors.

Resting systolic and diastolic blood pressure (SBP and DBP, respectively) was measured in a seated position after 5 min of rest, and again after a further 5-min rest with a mercury sphygmomanometer according to protocols recommended at the time by the Heart Foundation of Australia (23). The means of the two trials were used in analyses. Mean arterial pressure (MAP) was calculated. After blood pressure measurement and a 12- to 14-h fast, a qualified nurse collected a venous blood sample into tubes with solid EDTA anticoagulant. Plasma lipid levels of total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), and triglycerides (TG) were determined by automated techniques at the Flinders Medical Center, Adelaide, South Australia, which, throughout the study, met the criteria specified by the World Health Organization Collaborating Center for Reference and Research in Blood Lipids at the Centers for Disease Control in Atlanta, GA. Low-density lipoprotein cholesterol (LDL-C) was derived from the Friedewald equation (14). The ratio of TC:HDL-C was calculated.

CVD risk score.

A CVD risk score was derived by first standardizing the individual CVD risk variables for age by regressing them onto age, age2, and age3 to account for any nonlinearity in age-related differences and then summing the age-standardized residuals (Z-scores) for WC, MAP, HDL-C, and TG. The age-standardized HDL-C was multiplied by −1 because it is inversely related to metabolic risk. Unfortunately, blood glucose or insulin were not available. These variables were chosen because they represent the same variables (with the exception of blood glucose) used in the adult clinical criteria for the metabolic syndrome (13). A higher score is indicative of a less favorable metabolic profile.

Data analysis.

Estimated V˙O2max and % fat were used to determine the initial classification of subjects into four cross-tabulated groups (low fat/high fit, low fat/low fit, high fat/high fit, and high fat/low fit) on the basis of the FITNESSGRAM cut points for the V˙O2max (42 mL·kg−1·min−1 for males and 35-39 mL·kg−1·min−1 (age dependent) for females) and % fat (25% for males and 32% for females). FITNESSGRAM is a comprehensive youth-fitness-assessment program that provides criterion-referenced standards for health-related physical fitness (29). This classification scheme resulted in a great majority (92% of males, 86% of females) of subjects being classified into the low-fat/high-fit group. On the basis of the inadequate sample size per stratum, the cut points were adjusted according to sample-specific values corresponding with the 75th percentile for % fat and the 25th percentile for V˙O2max, resulting in the following cut points for V˙O2max (48 mL·kg−1·min−1 for males and 42 mL·kg−1·min−1 for females) and % fat (18% for males and 30% for females). Given the normal age- and gender-associated variation in % fat and V˙O2max (27) and other expert opinions regarding % fat (18,19), these values seemed reasonable for health-related cut points. Although a majority of subjects were still classified into the low-fat/high-fit group (approximately 67%), this breakdown was used for the statistical analysis because reducing the cut points further would fall outside the intentions of analyzing the data by health-related cut points.

Differences across groups for individual CVD risk-factor variables were assessed by analysis of covariance (ANCOVA), controlling for chronological age, within each gender. Differences across groups for the CVD risk score were assessed by ANOVA; it was not necessary to control for age because the individual risk factors were regressed onto age. Post hoc analyses were analyzed by the Bonferroni multiple-comparison tests. Analyses were executed with the SPSS package (version 11.0).


Results comparing the individual CVD risk factors across the four cross-tabulated fatness and fitness groups are shown in Table 1. For blood pressure, there is a significant difference (or trend) across fat-fit groups. Among females in the high-fat category, those in the high-fit group have a significantly lower blood pressure compared with their low-fit counterparts. These differences are also apparent in males but are not statistically significant. In males, the SBP is significantly different among the extreme groups (low fat/high fit and high fat/low fit), and MAP is significantly different between the low- and high-fat groups with low fitness. In females, blood pressure levels are significantly different among the extreme groups. In males, all the lipid values are significantly different across the fat-fit categories, with significant differences among the extreme groups for TG, LDL-C, HDL-C, and the TC:HDL-C ratio. In females, the HDL-C and TC:HDL-C ratios are significantly different across the fat-fit categories, and these differences are significant at the extremes.

Differences in cardiovascular disease risk factors across body fat percentage and cardiorespiratory fitness groups among adolescents in the Australian study population.

Figure 1 shows the results for the CVD risk score. In general, there was a significant trend across groups in both genders (F = 60.6, P < 0.001 in males and F = 57.3, P < 0.001 in females). In males and females, the high-fat/low-fit group had the highest composite risk scores (2.22 and 2.23, respectively), which represents a poorer CVD risk profile, and the low-fat/high-fit group had the lowest score (−0.55 and −0.65, respectively), which represents the best metabolic profile. These differences between the two extreme groups were significant (P < 0.05). As shown, these scores were similar in both genders. There were also significant differences between the low- and high-fat subgroups within a fitness group in both genders. Within a fatness group, the higher-fitness group showed lower scores, except in the low-fatness group in males.

Results for the CVD risk score.


This paper represents the third report in which we have examined the fat-but-fit hypothesis in children and/or adolescents. We have previously included analyses to evaluate the relation of cardiorespiratory fitness to CVD risk factors in children and adolescents classified with high and low BMI using a median split (10,12). The present paper avoided the limitation of using the median split and attempted to use a health-related cut-points approach. Furthermore, adiposity was expressed as % fat, and cardiorespiratory fitness was expressed as V˙O2max. Although our primary approach using FITNESSGRAM cut points resulted in inadequate sample sizes per strata, the derived cut points are reasonable on the basis of normal age- and gender-associated variation in % fat and V˙O2max and other expert opinions regarding % fat; they clearly show that children and adolescents with high percent body fat and high cardiorespiratory fitness have a better CVD risk profile than do their high-fat/low-fit counterparts. This protective effect of fitness on elevated fatness has been well established in adults (35), and we now have shown it in three separate child and adolescent samples from various countries (Canada, United States, and Australia), using various methodologies for fatness (BMI and skinfolds) and cardiorespiratory fitness (treadmill time to exhaustion, 1.6-km run, and physical working capacity at a heart rate of 170 bpm). This finding has important implications in understanding the consequences of pediatric obesity and, perhaps, the development of type 2 diabetes, CVD, and the metabolic syndrome.

Recently, there has been increased interest in the genesis of the features of the metabolic syndrome in children and adolescence (7). Because there is no clear definition of the metabolic syndrome in children or adolescence, and the prevalence rate is relatively low, deriving a composite CVD risk-factor score (i.e., metabolic syndrome score) represents one way of expressing the clustering of the main components of the adult metabolic syndrome that is statistically more sensitive and less error prone than the dichotomous approach (5,30). Several papers have shown that the metabolic syndrome risk-factor variables track reasonably well from adolescence to adulthood (1,3,11,20,22,31,34,38), and it is now known that adults with the metabolic syndrome are at increased risk of total mortality and CVD mortality (24). Because overweight is associated with an increased risk of metabolic syndrome in adolescents (21), our finding that the CVD risk score is attenuated by fitness in the high-fat group has important implications for the treatment of pediatric obesity and the prevention of the metabolic syndrome in obese adolescence. More specifically, pediatricians and other health care providers should not assume an adverse CVD risk-factor profile among overweight youth. In addition, and more importantly, the promotion of a physically active lifestyle in overweight youth is important for reducing the risk of subsequent disease in the absence of weight loss. This finding is also consistent with previous work showing an inverse correlation between cardiorespiratory fitness and CVD risk factors in obese children and adolescence (16). Furthermore, our results show that the high-fat/low-fit group had the highest CVD risk score, clearly highlighting that this combination poses a substantial health hazard.

The results show some gender differences in that the CVD risk factors vary among groups between males and females. For example, all lipid classes were significantly different across groups in males, but only HDL-C and TC:HDL-C are significantly different across groups in females. In addition, among females in the high-fat group, those with high fitness had significantly lower blood pressure compared with their counterparts in the low-fit group. These differences in boys were not statistically significant. The reasons for these specific gender differences in CVD risk factors are unclear. It is possible that fatness and/or aerobic fitness or the interaction of fatness and fitness influences CVD risk factors differently in males and females.

The influence that cardiorespiratory fitness (V˙O2max) may have on attenuating the CVD risk-factor profile among fat subjects has not been investigated. However, it can be speculated that the oxidative capacity of skeletal muscle and mitochondrial function may contribute. Recently, it has been shown that rodents that have been bred for high aerobic capacity have a better CVD risk-factor profile compared with their low-aerobic capacity counterparts (36). Furthermore, there were significant differences in several mitochondrial proteins between the two groups, suggesting a link between mitochondrial function and the metabolic syndrome. The results were also apparent among rodents that were 5 wk old, which roughly corresponds with human adolescence. It is also possible that the fat-fit subjects are more physically active. Unfortunately, data on physical activity were not available in the current study. Finally, there may be genetic aspects of the fat-but-fit phenotype. It is plausible that the fat-fit phenotype may be partially attributable to "adverse" genetic polymorphisms for fatness genes and "beneficial" polymorphisms for cardiorespiratory fitness. To date, several genes have been identified for cardiorespiratory fitness (37) and fatness (obesity) (32).

The cross-sectional study design limits drawing causal inferences, and prospective studies on this topic are needed. Given the sample size per stratification and the relatively low prevalence of adverse CVD risk factors or the metabolic syndrome in children and adolescents, we were unable to conduct logistic regression to determine the risk of an adverse CVD risk factor(s) between fat-fit groups. In addition, the relatively high aerobic fitness of this group may have influenced the results. Related to the limitation in sample size is the nonsignificant difference between some groups. Additional work among a more diverse sample of youth is needed to further explore these issues. The cross-tabulation approach used in this study provides a useful model for future analyses because it allows for direct comparisons among youth in distinct categories.

In summary, this study provides evidence for the importance of considering both fitness and fatness in relation to CVD risk factors in young people. As expected, youth with combined high fatness and low fitness have the poorest metabolic profile, but the key finding is that CVD risk score is attenuated by aerobic fitness in the high-fat group.

All authors were involved in the conceptualization, data analysis and interpretation, and writing of the manuscript.


1. Andersen, L. B., and J. Haraldsdottir. Tracking of cardiovascular disease risk factors including maximal oxygen uptake and physical activity from late teenage to adulthood. An 8-year follow-up study. J. Intern. Med. 243:309-315, 1993.
2. The Australian Council for Health, P. E. a. R. Australian Health and Fitness Survey, 1985. Parkside, Australia: ACHPER, 1985.
3. Bao, W., S. R. Srinivasan, W. A. Wattigney, and G. S. Berenson. Persistence of multiple cardiovascular risk clustering related to syndrome X from childhood to young adulthood. Arch. Intern. Med. 154:1842-1847, 1994.
4. Barlow, C. E., H. W. Kohl, L. W. Gibbons, and S. N. Blair. Physical fitness, mortality and obesity. Int. J. Obes. Relat. Metab. Disord. 19:S41-S44, 1995.
5. Brage, S., N. Wedderkopp, U. Ekelund, et al. Features of the metabolic syndrome are associated with objectively measured physical activity and fitness in Danish children: the European Youth Heart Study (EYHS). Diabetes Care 27:2141-2148, 2004.
6. Coonan, W., and T. Dwyer. Recommended Guidelines and Protocols for the Establishment of a National Fitness, Health and Physical Performance Survey in Australian Schools. Adelaide, Australia: The Australian Council for Physical Education and Recreation, 1983.
7. Cruz, M. L., and M. I. Goran. The metabolic syndrome in children and adolescents. Curr. Diab. Rep. 4:53-62, 2004.
8. Cureton, K. J., M. A. Sloniger, J. P. O'Bannon, D. N. Black, and W. P. McCormack. A generalized equation for prediction of VO2peak from one-mile run/walk performance in youth. Med. Sci. Sports Exerc. 27:445-451, 1994.
9. Eisenmann, J. C., K. A. Heelan, and G. J. Welk. Assessing body composition among 3- to 8-year-old children: anthropometry, BIA, and DXA. Obes. Res. 12:1633-1640, 2004.
10. Eisenmann, J. C., P. T. Katzmarzyk, L. Perusse, A. Tremblay, J.-P. Despres, and C. Bouchard. Aerobic fitness, body mass index and CVD risk factors among adolescents: the Quebec Family Study. Int. J. Obes. 29:1077-1083, 2005.
11. Eisenmann, J. C., G. J. Welk, E. E. Wickel, and S. N. Blair. Stability of variables associated with the metabolic syndrome from adolescence into adulthood: the Aerobics Center Longitudinal Study. Am. J. Hum. Biol. 16:690-696, 2004.
12. Eisenmann, J. C., E. E. Wickel, G. Welk, and S. N. Blair. Cardiorespiratory fitness, fatness, and cardiovascular disease risk factors among adolescence: the Aerobics Center Longitudinal Study. Int. J. Pediatr. Obes. >In press.>
13. Expert Panel on Detection Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (adult treatment panel III). JAMA 285:2486-2497, 2001.
14. Friedewald, W. T., R. I. Levy, and D. S. Fredrickson. Estimation of the concentration of low density lipoprotein cholesterol in plasma without use of the preparative ultracentrifuge. Clin. Chem. 18:499-502, 1972.
15. Goran, M. I., G. D. C. Ball, and M. L. Cruz. Obesity and risk of type 2 diabetes and cardiovascular disease in children and adolescents. J. Clin. Endocrinol. Metab. 88:1417-1427, 2003.
16. Gutin, B., S. Islam, T. Manos, N. Cucuzzo, C. Smith, and M. E. Stachura. Relation of percentage of 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.
17. Hedley, A. A., C. L. Ogden, C. L. Johnson, M. D. Carroll, L. R. Curtin, and K. M. Flegal. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA 291:2847-2850, 2004.
18. Jackson, A. S., M. L. Pollock, and A. Ward. Generalized equations for predicting body density of women. Med. Sci. Sports Exerc. 12:175-182, 1980.
19. Jackson, A. S., and M. L. Pollock. Generalized equations for predicting body density of men. Br. J. Nutr. 40:497-504, 1978.
20. Katzmarzyk, P. T., L. Perusse, R. M. Malina, J. Bergeron, J. Despres, and C. Bouchard. Stability of indicators of the metabolic syndrome from childhood and adolescence to young adulthood: the Quebec Family Study. J. Clin. Epidemiol. 54:190-195, 2001.
21. Katzmarzyk, P. T., A. Tremblay, L. Perusse, J.-P. Despres, and C.Bouchard. The utility of the international child and adolescent overweight guidelines for predicting coronary heart disease risk factors. J. Clin. Epidemiol. 56:1-7, 2003.
22. Kemper, H. C. G., J. Snel, R. Verschuur, and L. Storm-van Essen. Tracking of health and risk indicators of cardiovascular diseases from teenager to adult: Amsterdam Growth and Health Study. Prev. Med. 19:642-655, 1990.
23. Kirkendall, N. M., A. C. Burton, F. H. Epstein, and E. D. Fries. Recommendations for human blood pressure determination by sphygmomanometers: report of a subcommittee of the postgraduate education committee. Circulation 36:980-988, 1967.
24. Lakka, H. M., D. E. Laaksonen, T. A. Lakka, et al. The metabolic syndrome and total and cardiovascular disease mortality in middle-aged men. JAMA 288:2709-2716, 2002.
25. Lee, C. D., S. N. Blair, and A. S. Jackson. Cardiorespiratory fitness, body composition, and cardiovascular disease mortality in men. Am. J. Clin. Nutr. 69:373-380, 1999.
26. Malina, R. M. Anthropometry. In: Physiological Assessment of Human Fitness, P. J. Maud and C. Foster (Eds.). Champaign, IL: Human Kinetics, pp. 205-219, 1995.
27. Malina, R. M., C. Bouchard, and O. Bar-Or. Growth, Maturation, and Physical Activity, 2nd ed. Champaign, IL: Human Kinetics, 2004.
28. Melanson, K. J., K. J. McInnis, J. M. Rippe, G. Blackburn, and P. F. Wilson. Obesity and cardiovascular disease risk: research update. Cardiol. Rev. 9:202-207, 2001.
29. Morrow, J., H. B. Falls, and H. W. Kohl (Eds.). FITNESSGRAM Technical Reference Manual. Dallas, TX: The Cooper Institute, 2001.
30. Raglund, D. R. Dichotomizing continuous outcome variables: dependence of the magnitude of association and statistical power of the cutpoint. Epidemiology 3:434-440, 1992.
31. Raitakari, O. T., K. V. Porkka, L. Rasanen, T. Ronnemaa, and J. S. Viikari. Clustering and six year cluster-tracking of serum total cholesterol, HDL-cholesterol and diastolic blood pressure in children and young adults. The Cardiovascular Risk in Young Finns Study. J. Clin. Epidemiol. 47:1085-1093, 1994.
32. Rankinen, T., A. Zuberi, Y. Chagnon, et al. The human obesity gene map: the 2005 update. Obesity 14:529-644, 2006.
33. Slaughter, M. H., T. G. Lohman, R. A. Boileau, et al. Skinfold equations for estimation of body fatness in children and youth. Hum. Biol. 60:709-723, 1988.
34. Stuhldreher, W. L., T. J. Orchard, R. P. Donahue, L. H. Kuller, M. F. Gloninger, and A. L. Drash. Cholesterol screening in childhood: sixteen-year Beaver County Lipid Study experience. J. Pediatr. 119:551-556, 1991.
35. Welk, G. J., and S. N. Blair. Physical activity protects against the health risks of obesity. In: Toward a Better Understanding of Physical Fitness and Activity, C. B. Corbin, R. Pangrazi, and D. Franks (Eds.). Scottsdale, AZ: Holcomb Hathaway Publishers, pp. 89-96, 2004.
36. Wisloff, U., S. M. Najjar, O. Ellingsen, et al. Cardiovascular risk factors emerge after artificial selection for low aerobic capacity. Science 307:418-420, 2005.
37. Wolfarth, B., M. S. Bray, J. M. Hagberg, et al. The human gene map for performance and health-related fitness phenotypes: the 2004 update. Med. Sci. Sports Exerc. 37:881-903, 2005.
38. Yong, L-C., and L. H. Kuller. Tracking of blood pressure from adolescence to middle-age: the Dormont High School Study. Prev. Med. 23:418-426, 1994.


©2007The American College of Sports Medicine