Is There a Correct Body Mass Assessment for Active Children: Implications for Predicting

Is There a Correct Body Mass Assessment for Active Children: Implications for Predicting Childhood Diseases

Michael S. Laymon, DPTSc, PT, OCS, CCD

Jan S. Kodat, DPTSc, PT, CCD

Wendy L. Chung, DPTSc, PT

Azusa Pacific University, Department of Physical Therapy, Azusa, California

KEY WORDS: body composition, body mass index, infrared, bioelectrical impedance, pre-adolescent

Abstract

Objective: To compare measurements of body mass index (BMI), bioelectrical impedance assessment (BIA), and infrared (IR) body composition measurements in this pre-adolescent population and determine if they can be used interchangeably to predict childhood disease.

Design: One hundred fourteen pre-adolescent basketball players ages 8 to 13 years old were the subjects of this study (49 male and 65 female). Height, weight, IR body composition, BIA, and BMI were determined for each subject on the same day.

Results: Mean body composition for females between the ages of 8 and 13 years was 18.8% for BMI, 18.3% for BIA, and 24.2% for IR. Mean body composition for males between the ages of 9 and 12 years was 22.6% for BMI, 13.5% for BIA, and 26.3% for IR. There was no significant correlation for body composition between BMI, BIA, or IR measurements for boys or IR for girls. There was significant correlation for girls between BMI and BIA measurements. BIA values for boys were consistently lower across the population as compared with BMI and IR results.

Conclusions: IR, BMI, and BIA are not correlated to each other and cannot be used interchangeably when reporting or monitoring body composition. BMI should not be used as a measure of body composition for adolescent males involved in recreational sports. BMI, IR, and BIA are not indicators of body composition on an individual basis and should not be used as a basis for prediction of childhood disease. There continues to be no accurate, cost-effective means to assess individual body composition by a rapid, noninvasive methodology.

INTRODUCTION

Body fat composition is universally accepted as an indicator for general health. Obesity (excess body fat) has become a worldwide epidemic with a detrimental impact on health and therefore represents an enormous economic burden. Diseases associated with obesity include cardiovascular disease1,2; stroke; type II diabetes mellitus3; hypertension; dyslipidemia; cancers of the breast, endometrium, prostate, and colon4,5; gallbladder disease; osteoarthritis6-8; respiratory problems, including asthma9 and sleep apnea10; and depression.11,12

It has been reported that juvenile obesity has grown to affect over 20% of children in the United States, especially girls.13 As a result of the continued rise in the childhood obesity population, the importance of monitoring body composition in children has widespread implications. Body composition of children can act as a predictor of future health care needs as a larger population of obese children develop into adults.14 Freedman et al. and Sinaido et al. identified children with high fasting glucose levels and were overweight (as measured by body mass index [BMI]) as having an increased risk for type 2 diabetes.15,16 Cruz et al. showed a high correlation between increased body fat and diabetes in obese Hispanic children with a family history of type 2 diabetes.17 This onset of early diabetes is also associated with bone loss, even in children as young as 17 months.18

Aside from the health and economic costs associated with obesity, treatment generally incorporates increases in physical activity. Aerobic capacity and the ability to perform physical activities has also been reported as being adversely affected for those who are overweight or obese.19,20 This underscores the importance of an accurate body composition measurement when clinically recommending or developing exercise programs designed to facilitate weight loss. Likewise, many health fairs are now beginning to recognize body composition as an important component of health promotion and screening activities. Therefore, an accurate measure of body fat, as reported by a measure of true body composition, needs to be incorporated that is easily used, inexpensive, and reproducible.

Body mass index is the most commonly used measure for classifying overweight and obesity in children. BMI as a predictor of body fat is generally used because it requires no special instrumentation or training. BMI is determined by dividing body weight (in kilograms) by the square of height (in meters). In adults, overweight is defined for both genders as a BMI of ≥25.0 kg/m2, obesity as a BMI of ≥30.0 kg/m2, and morbid obesity as a BMI of ≥35.0 kg/m2. One study demonstrated how BMI was an effective predictor of percent body fat mass in a small portion of the pre-adolescent population, but it was not useful as a reference on an individual basis.21 There does not appear to be a true correlation between BMI and actual body fat in quantitative terms, especially when referring to childhood obesity.22,23 Because BMI is easily calculated using height and weight measurements that are part of all childhood medical records, national statistics reporting the prevalence of adolescent obesity and overweight based on BMI has been accepted by default. Recognizing the limitation of BMI as a predictor for accurate individual body composition, many researchers have sought alternative methods to reliably measure individual body composition.

Several studies implicate the use of bioelectrical impedance (BIA) as a truer measure of percent body fat as compared with BMI or skinfold thickness (SFT).24-26 BIA relies on the conductivity of body tissues to predict the amount of total body fat based on the premise that fat acts as an electrical current insulator. One study suggests that BIA appears to effectively measure fat free mass and is consistent with triceps SFT in predicting body fat in non-obese pre-adolescent girls.27 There is currently a plethora of BIA measurement devices available to healthcare providers and the general public with relatively limited published data targeted to specific populations or the obese.

Another available option that is readily accessible and easy to use is infrared (IR). Infrared measurement is based on the principle that human fat absorbs light at specific wavelengths in the near-infrared portion of the spectrum. These units emit that frequency and measure how much of the energy emitted is absorbed. Calculation of percent body fat is determined by integrating height and weight with the IR absorption rate. Like with BMI and BIA, there are conflicting studies about the use of IR in accurately determining percent body fat in the pre-adolescent population.28,29 There is an increasing availability of IR body composition units that are relatively inexpensive and easy to use.

Most of the studies used to validate use of these various methods of body composition have been limited by the use of small sample sizes. The purpose of this study was to compare measurements of BMI, BIA, and IR body composition in this pre-adolescent population and to determine if they can be used interchangeably to predict childhood disease.

METHODS

Our subjects were 49 male and 65 female
(n = 114) pre-adolescent basketball players ages 8 to 13 years (Table 1). Permission to participate in this study was obtained by the legal guardian of each individual subject. Height, weight, IR body composition, bioelectrical impedance body composition, and BMI were determined for each subject on the same day of registration for the basketball league.

Body mass index was determined by dividing the subject's weight in kilograms by the square of the height in meters. BIA was collected using RJL Systems (Clinton, MI) model 1PG-104 plethysmograph and body composition calculated using the Cyprus Body Composition Analysis Program (RJL Systems) version 1.4. Two electrodes (2.3 cm ¥ 2.5 cm Ag/AgCl and clear adhesive hydrogel; Kendall Meditrace 710) were placed on the anterior distal portion of the right wrist and at the posterior/lateral portion of the left lateral malleolus. Infrared body composition was determined using the Futrex-1100 Personal Body Fat Tester (Hagerstown, MD). The Futrex unit was zeroed using an optical standard. The subject's height and weight were entered into the unit and a reading was taken at the midpoint of their relaxed biceps brachii.

SPSS version 11 (Chicago, IL) was used for all statistical calculations. Subjects with negative values for BIA or IR were eliminated from the study. Subjects were screened and in good health with no metabolic or cardiovascular disorders. A one-way analysis of variance was used to determine if there were any correlation between percent body fat collect by BMI, BIA, or IR. A Tukey's correlation was used as the post hoc statistic to determine the relationship between each set of groups. All subjects have participated in at least 2 different recreational activities within the year.

RESULTS

Table 2 summarizes body composition by gender and method. Mean body composition for females between the ages of 8 and 13 years was 18.8% for BMI, 18.3% for BIA, and 24.2% for IR. Mean body composition for males between the ages of 9 and 12 years was 22.6% for BMI, 13.5% for BIA, and 26.3% for IR. There was no significance correlation for body composition between BMI, BIA, or IR measurements for boys. There was significant correlation for girls between BMI and BIA measurements. BIA values for boys were consistently lower across the population as compared with BMI and IR results. The percentages for overweight (BMI >85% for age and gender) and obese (BMI >95% for age and gender) subjects are listed in Table 3. Overweight and obese classifications for BIA and IR are based on the 85% and 95% BMI rates, respectively.30 Overall obesity for girls calculated by BMI is 8% of the subject population versus 23% for BIA and 59% for IR. The percentage of boys classified as obese is 35% calculated from BMI, 2% calculated by BIA, and 68% when calculated by IR. BMI demonstrated the least variation within groups with BIA having greatest variation for girls and IR for boys.

Discussion

There is great concern over the trend of increasing adolescent obesity. Overweight and obesity in adolescence might persist into adulthood and increase the risk for chronic diseases later in life. Body mass index has been the tool used to assess body composition among adults and children. This method is chosen because it is simple to use and requires no special training or equipment. With the growing concern over BMI's ability to accurately measure individual body compensation, many public health professionals are looking to other noninvasive, economical, and simple ways to measure body composition accurately. Our data shows that there is no consistency on an individual basis between BMI, BIA, or IR. Therefore, these measurements might not be compared or referenced with each other.

In comparing our BMI data with other studies, we find that our female population showed a much lower percentage of obesity than the national average (8% vs 11%). This gives support to promoting recreational sports to help reduce obesity because all our subjects participate regularly in recreational sports. On the other hand, there is a great discrepancy in the male population between the national average of 11% obesity and our 35% measured by BMI. One possible explanation for this massive difference might also be explained by the fact that our subjects regularly participate in physical activity. BMI does not consider relative bulk or percentages of various tissue types; it merely uses height and weight. If we compare the average body composition in the male population of 22.6% for BMI with 13.5% for BIA, one has to attribute such a difference to the mechanism of measurement. This discrepancy between measurement methods was not seen in the female population. Because muscle weighs more than adipose, the authors concluded that the BIA measurement would be a truer measure of body composition for the male population. Even though external sex characteristics were not apparent within our subject population, metabolic changes might already be taking place.

Infrared devices have been identified as an alternate method to calculate body composition. Our data indicates that IR measurements were more variable and gave an overall 18% higher body composition measure than BMI. This is the direct opposite of data reported by Thomas et al. who found IR to give a 15% underestimate of total body fat as compared with other methods. The difference between the finding of our study and the Thomas study might be because the Thomas study used adult subjects. Our data further shows that of the 3 modalities used to determine body composition, IR produced the highest percent body fat in both genders and through all ages. We think the lack of correlation of IR body composition measures in our subjects with BMI or BIA might be because the device itself and not from an erroneously based concept. In the device we used, the examiner enters the height and weight of the individual into the IR unit and then places the unit over the subject's biceps muscle to record a measurement. It was found that if you kept the same height and weight entered in the unit from one individual and then measured a very different-sized biceps on another individual whose body composition was much different, the reading would often be the same as the initial subject, thus leading the authors to believe the formula used by the IR unit is similar to simply calculating BMI. We used 2 different units and found the same scenario with both units.

In reviewing the results of the BIA unit, the authors recognize the influences that can play a role in getting an accurate reading such as dehydration, amount of recent activity, and electrolyte balance. Although the recordings were taken before the actual practice or "try out" activity, it was not recorded as to the individual's activity before arriving for the registration. Also, because all the measurements were taken in 1 day, the subjects arrived intermittently throughout the day and thus would have had a variety of different activities and fluid intake depending on their arrival time. The computer software program used to determine the percent body fat with the BIA readings also makes a gender distinction. However, because secondary sex characteristics have not developed yet within this population of adolescents, the authors think there should not be a gender bias in the composition calculation. The software also had difficulty calculating using some of the lower ranges of height and weight, thus giving a negative value for body fat percentage. Although, these subjects were not included in our data analysis, this demonstrates a limitation of the Cyprus Body Composition Analysis Program, which relies on a 2-component system to determine body composition. Sun et al. have reported greater precision when using BIA with a multicomponent model analysis.31 BIA costs with the type of unit used in our study in comparison to IR or BMI are also greater as a result of the need to use electrodes. Initial investigation of non-electrode units (either handheld or integrated with a standing scale through bare feet) proved to be too inconsistent to consider using in this study.

As the issue of obesity and associated long-term health risks in adolescents becomes an increased focus of attention, methods to decrease the prevalence of adolescent obesity have grown in interest. Recently, several studies suggest gastric bypass as a safe and effective tool in the treatment of adolescent obesity.32,33 The 1991 National Institutes of Health Consensus Conference criteria used to justify bariatric surgery in adults have been applied to the obese adolescent population. That criteria sets eligibility for bariatric surgery if the subject has a BMI >35 kg/m2 with or >40 kg/m2 without obesity comorbidity. This study demonstrates that the BMI-only criteria for pre-adolescent males might provide a false-positive evaluation leading to surgical intervention. In light of the nature of bariatric surgery, we suggest that BMI alone not be used to justify surgical intervention within the adolescent population and that more direct body composition measurement methods and the existence of comorbidities be the standard. This brings into question the ability to apply body composition measurement assumptions shown to be valid in the adult population to adolescents. It is clear that we do not use the same BMI measurement to define obesity in adults versus adolescents. In adults, we classify someone as obese when their BMI is >30. In children and adolescents, we use BMI of greater than the 95th percentile for age and gender.

Further investigations need to be completed to determine an effective and useful tool to accurately determine percent body fat in the juvenile population. Software packages need to be updated to include lower level values for children, and companies developing these tools need to establish good data specific to children by including children in the research instead of assuming a linear regression correlation for children using the adult data. When using BIA, more data should be collected to develop a standardized testing protocol that addresses activity and fluid/food intake before testing. Future studies might include further comparisons with dual energy x-ray absorptiometry (DEXA) to determine the most accurate method of measure because these systems directly measure fat composition.

Conclusions

IR, BMI, and BIA are not correlated to each other when measuring body composition in this pre-adolescent population and cannot be used interchangeably to predict childhood disease. BMI should not be used as a measure of body composition for adolescent males involved in recreational sports. BMI, IR, and BIA are not indicators of body composition in the pre-adolescent population on an individual basis. There continues to be no accurate, cost-effective means to assess individual body composition by a rapid, noninvasive methodology.

References

1. Hubert HB, Feinleib M, McNamara PM, Castelli WP: Obesity as an independent risk factor for cardiovascular disease: a 26-year follow-up of participants in the Framingham Heart Study. Circulation 67:968-977, 1983.

2. Manson JE, Colditz GA, Stampfer MJ: A prospective study of obesity and risk for coronary heart disease in women. N Engl J Med 322:882-918, 1990.

3. Hu FB, Manson JE, Stampfer MJ, et al: Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med 345:790-797, 2001.

4. Stoll BA: Perimenopausal weight gain and progression of breast cancer precursors. Cancer Detect Prev 23:31-36, 1999.

5. Wolff GL: Body weight and cancer. Am J Clin Nutr 45:168-180, 1987.

6. Hartz AJ, Fischer ME, Bril G, et al: The association of obesity with joint pain and osteoarthritis in the HANES data. J Chron Dis 39:311-319, 1986.

7. Felson DT, Anderson JJ, Naimark A, et al: Obesity and knee osteoarthritis: the Framingham study. Ann Intern Med 109:18-24, 1988.

8. Davis MA, Ettinger WH, Neuhaus JM: Obesity and osteoarthritis of the knee: evidence from the National Health and Nutrition Examination Survey (NHANES I). Semin Arthritis Rheum 20:34-41, 1990.

9. Chen Y, Dales R, Tang M, Krewski D: Obesity may increase the incidence of asthma in women but not in men: longitudinal observations from the Canadian National Population Health Surveys. Am J Epidemiol 155:191-197, 2002.

10. Young T, Peppard PE, Gottlieb DJ: Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med 165: 1217-1239, 2002.

11. Roberts RE, Kaplan GA, Shema SJ, Strawbridge WF: Are the obese at greater risk for depression? Am J Epidemiol 152:163-170, 2000.

12. Friedman KE, Reichmann SK, Costanzo PR, Musante GJ: Body image partially mediates the relationship between obesity and psychological distress. Obes Res 10:33-41, 2002.

13. Bar-Or O: Juvenile obesity, physical activity, and lifestyle changes. Phys Sportsmed 28:51-58, 2000.

14. Dietz W: Health consequences of obesity in youth: childhood predictors of adult disease. Pediatrics 101:518-525, 1998.

15. Freedkman DS, Dietz WH, Srinivasan SR, Berenson GS: The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics 103:1175-1182, 1999.

16. Sinaiko AR, Jacobs DR Jr, Steinberger J, et al: Insulin resistance syndrome in childhood: associations of the euglycemic insulin clamp and fasting insulin with fatness and other risk factors. J Pediatr 139:700-707, 2001.

17. Cruz ML, Bergman RN, Goran MI: Unique effects of visceral fat on insulin sensitivity in obese Hispanic children with a family history of type 2 diabetes. Diabetes Care 25:1631-1636, 2002.

18. Shore RM, Chesney RW, Mazess RB, Rose PG, Bargman GJ: Osteopenia in juvenile diabetes. Calcif Tissue Int 33:455-457, 1981.

19. Mattsson E, Larson UE, Rossner S: Is walking for exercise too exhausting for obese women? Int J Obes Relat Metab Disord 21:380-386, 1997.

20. Tsuritani I, Honda R, Noborisaka Y, et al: Impact of obesity on musculoskeletal pain and difficulty of daily movements in Japanese middle-aged women. Maturitas 42:23-30, 2002.

21. Schaefer F, Georgi M, Wuhl E, Scharer K: Body mass index and percentage fat mass in healthy German school children and adolescents. Int J Obes Relat Metab Disord 22:461-469, 1998.

22. Ellis K, Shypailo R, Abrams S, Wong W: The reference child and adolescent models of body composition. A contemporary comparison. Ann NY Acad Sci 904:374-382, 2000.

23. Wells J: A critique of the expression of paediatric body composition data. Arch Dis Child 85:67-72, 2001.

24. Nagaya T, Yoshida H, Takahashi H, Matsuda Y, Kawai M: Body mass index (weight/height²) or percentage body fat by bioelectrical impedance analysis: which variable better reflects serum lipid profile? Int J Obes Relat Metab Disord 23:771-774, 1999.

25. Vienna A, Capucci E, De Stefano GF, Hauser G: Bioelectrical impedance analysis and anthropometry in Ecuadorian children of African ancestry. Coll Antropol 22:433-446, 1998.

26. Lohman T, Caballero B, Himes J, et al: Body composition assessment in American Indian children. Am J Clin Nutr 69(suppl):764S-766S, 1999.

27. Bandini L, Vu D, Must A, Dietz W: Body fatness and bioelectrical impedance in non-obese pre-menarcheal girls: comparison to anthropometry and evaluation of predictive equations. Eur J Clin Nutr 51:673-677, 1997.

28. Fuller NJ, Dewit O, Wells JC: The potential of near infra-red interactance for predicting body composition in children. Eur J Clin Nutr 55:967-972, 2001.

29. Thomas DW, Ryde SJ, Ali PA, et al: The performance of an infra-red interactance instrument for assessing total body fat. Physiol Meas 18:305-315, 1997.

30. Troiano RP, Flegal KM: Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics 101:497-504, 1998.

31. Sun SS, Chumlea WC, Heymsfield SB, et al: Development of bioelectrical impedance analysis prediction equations for body composition with the use of a multicomponent model for use in epidemiologic surveys. Am J Clin Nutr 77:331-340, 2003.

32. Stanford A, Gloscock JM, Eid GM, et al: Laparoscopic Roux-en-Y gastric bypass in morbidly obese adolescents. J Pediatr Surg 38:430-433, 2003.

33. Sugerman HF, Sugerman EL, DeMaria EJ, et al: Bariatric surgery for severely obese adolescents. J Gastrointest Surg 7:102-108, 2003.

Table 1. Demographics of Subjects

Age

8 9 10 11 12 13 Total

Gender

Female 5 14 16 15 7 8 65

Male 0 6 8 9 26 0 49

Table 2. Percent Body Composition by Method

N BMI BIA IR

Gender

Female 65 18.8 ± 3.3 18.4 ± 8.5 24.2 ± 5.1

Male 49 22.6 ± 2.6 13.5 ± 3.9 26.3 ± 5.2

Total 114 20.4 ± 3.6 16.3 ± 7.3 25.1 ± 5.3

BMI, body mass index; BIA, bioelectrical impedance assessment; IR, infrared body composition.

Table 3. Percentage of Subjects Classified and Overweight and Obesity Calculated by Method

BMI BIA IR

Overweight Obese Overweight Obese Overweight Obese

Gender

Female 12% 8% 12% 23% 15% 59%

Male 45% 35% 4% 2% 6% 80%

Total 26% 19% 9% 14% 11% 68%
BMI, body mass index; BIA, bioelectrical impedance assessment; IR, infrared body composition.