Predictive equation for fat percentage based on body mass index for adolescents with down syndrome.
Life expectancy of patients with Down Syndrome (DS) has increased in recent decades (3) with an increased prevalence for being overweight and obese. Both factors increase the risk for developing various comorbidities, such as heart disease, type 2 diabetes, hypertension and chronic obstructive pulmonary disease (13). The development of an accurate method to assess body composition, in particular fat percentage (Fat%) would be an important tool for health professionals who wish to oversee nutritional strategies and exercise interventions in individuals with DS (13,17).
Dual energy X-ray absorptiometry (DEXA) is the gold standard for the assessment of bone mineral density. It also allows an accurate evaluation and excellent reproducibility for the determination of body composition (e.g., lean and fat percentage mass) (11). Although DEXA is a noninvasive method, it is expensive and difficult to apply to individuals with DS. This is especially the case when mild to severe intellectual disabilities exist, given the difficulty of maintaining the patient static throughout the evaluation (11,17,19). This is also a concern since the measurement of body fat by DEXA has been increasingly used as a reference standard for pediatric populations, particularly the evaluation of body composition (11,13).
Assessment of body composition using the skinfold protocol appears to be appropriate because of the low cost and high reproducibility (19). But, the assessment requires an experienced appraiser to grasp several folds of skin that must be repeated at least three times. The procedures increases the time of the evaluation (19), and it may result in slight discomfort to the individual with an intellectual disability (17).
The body mass index (BMI) is a simple and widely used procedure in epidemiological studies for the classification of general health as well as to predict cardiovascular diseases and adiposity (18-21). However, the issue with using BMI is that may be difficult to distinguish muscle tissue from fat mass and body fat distribution (i.e., visceral fat and subcutaneous fat) (8,9,15). Interestingly, in a study conducted by Sales et al. (19). BMI values were used to develop a predictive equation for Fat% in older women, which improved the specificity of BMI for indirect determination of Fat%. Otherwise, it is common to use BMI to predict the percentage of body fat (9). However, to date, it appears that no studies have developed an equation to determine Fat% based on BMI in patients with DS.
Thus, the purpose of this study was to evaluate the association between body fat percentage measured by DEXA and BMI in adolescents with DS, and then develop a predictive equation of Fat% for this population based on BMI.
This was a cross-sectional study that was conducted with 26 adolescents (14 girls and 12 boys) with Down Syndrome (DS Group). The control group (CTL Group) consisted of 15 children (7 girls and 8 boys) without DS. Sixteen subjects were the "n" minimum sample required to provide greater statistical power of 80% (Power = 0.80) with an alpha of 5%. The general characteristics of the subjects are shown in Table 1. Inclusion criteria were: (a) age between 9 and 18; (b) a medical report indicating Trisomy 21 in each subject of the DS Group; and (c) consent from the parents or responsible adults for each subject's participation and evaluation. Before data collection got underway, the subjects signed informed consent form. Then, the subjects were divided into two groups (DS and CTL) and submitted to the BMI measures and Fat% by the DEXA technique. The procedures were performed by the same examiner to ensure the reliability of the assessments. All procedures were approved by the Ethics Committee in Research of the Catholic University of Brasilia - UCB (No. 42088561.5.00000029).
Body Composition (BMI and DEXA)
The total body weight (kg) and height (m) were measured unshod and with appropriate clothing. An electronic balance (Filizola[R]) was used for the measurement of body weight, and a stadiometer (SECA[R] 214, USA) was used to measure height. The body mass index value (BMI) was calculated from the ratio of body weight in kilograms by the square of height obtained in meters (kg x [m.sup.-2]). The Fat% was performed by DEXA (DtX IQ LUNAR[R], IL, USA) according to the manufacturer's instructions. Following calibration of the equipment, the mass verification free of fat and fat mass was determined. The coefficient of variation and technician error reported by the laboratory corresponded to 2% and 0.4%, respectively. All procedures took place at the Laboratory of Physical Evaluation and Training of the Catholic University of Brasilia (LAFIT-UCB) with a trained technical assistance team to perform all the procedures.
Equation Development and Validation
The procedures for the development of the fat percentage prediction equation were similar to those reported by Sales et al. (19). A linear regression was developed from the BMI data and Fat% measured by DEXA (measured Fat%) for each group. Thus, an equation for estimating Fat% was generated (estimated Fat%) from the BMI data for each group. The equation generated for the CTL group was: Fat% = 2.199 x BMI - 20.86 (Figure 1A). For the DS group, the equation was: Fat% = 2.006 x BMI - 20.36 (Figure 1B).
The Shapiro-Wilk test was used to determine data normality. To determine the central tendency measures, mean and standard deviation were used. For comparison between the measured and estimated Fat%, the t test for independent samples was used. To check the level of association between the different methods (measured and estimated Fat%), Pearson correlation was used. The concordance between measured and estimated Fat% was carried out using the Bland-Altman test (4) and intraclass correlation coefficient (ICC). The significance level was set at 5% (P<0.05). All statistical analyses were performed using the Statistical Package for Social Sciences software (SPSS, Inc., Chicago, IL) 21.0, GPower 3.0.10 (University of Kiel, Germany) and GraphPad Prism 6.0 (GraphPad Software Inc., San Diego, CA).
Data related to age and BMI did not differ between groups (P>0.05). The CTL group had a mean of 12.9 [+ or -] 3.18 yrs of age and the DS group had a mean of 12.4 [+ or -] 2.68. For the BmI, the CTL group had a mean of 19.7 [+ or -] 4.5 kg x [m.sup.-2] while the DS group was 21.7 [+ or -] 4.64 kg x [m.sup.-2]. The estimated Fat% presented a mean of 22.42 [+ or -] 9.53% for the CTL group and 23 [+ or -] 9.24% for the DS group (Table 1).
The results of measured and estimated Fat% indicate that the equations for estimating the Fat% from BMI developed for both groups indicate a high level of agreement. There was no significant difference (P=0.999 for both groups) between Fat% measured by DEXA for the CTL group (22.4 [+ or -] 12.86%) and the DS group (23 [+ or -] 11.84%). The values for measured and estimated Fat% showed a strong positive association and a significant correlation in both groups (r = 0.75; P<0.001 for the CTL group and r = 0.78; P<0.0001 for the DS group) (Figures 1A and 1B, respectively).
[FIGURE 1 OMITTED]
There was also good agreement between the measured and estimated Fat% for both groups, which was analyzed graphically by the Bland-Altman plot and calculated by the intraclass correlation coefficient (ICC = 0.74 for the CTL group and ICC = 0.76 for DS group, P<0.05) (Figures 2A and 2B, respectively).
[FIGURE 2 OMITTED]
The primary finding of this study indicates a strong correlation between BMI and measured Fat% by DEXA in adolescents with and without DS. With this result, we performed the application of a linear regression to obtain a predictive equation for Fat% from BMI. The agreement between the predictive equations versus DEXA was analyzed by the method described by Bland-Altman and calculated by the intraclass correlation coefficient, which confirmed a high reliability between the Fat% measurement systems for adolescents with and without Down's syndrome.
Although the high prevalence of obesity in the general population is well-known, there is a little information on the prevalence of obesity in children with DS (2). A few reports indicate that young people with Ds present a 30% prevalence of overweight and obesity (5,6). In the present study, we found no differences in Fat% and BMI between the CTL group and the DS group. This finding is in agreement with Izquierdo-Gomez et al. (12) who studied adolescents with and without DS in both genders. They reported no significant differences in the Fat% between the groups. However, they did find a significantly better physical fitness level in the adolescents without DS.
Despite the fact that the term "obesity" is commonly used to describe individuals with DS, it is not always clear that children and adolescents with DS are in fact obese. In fact, Gonzalez-Aguero et al. (10) studied adolescents of both genders with DS and found no differences between the gender groups with and without DS in BMI, waist circumference, and Fat%. It is still too early to say that adolescent with DS are more obese than their peers without DS. In addition, DS with young growth patterns are different from the typical development growth patterns of children and adolescents without DS (1,16).
Our findings in individuals with DS strengthen the use of the BMI measure as an effective, practical, and easy method to apply to the general population (8,19-21). This highlights the importance of practical methods for the identification of overweight individuals with intellectual disabilities in order to prevent and treat potential comorbidities (14,17). Predictive equations for the determination of Fat% in different populations significantly decrease the cost and time spent in physical assessments. The equations also provide less discomfort for the subjects. The latter is important to consider, especially for individuals with some form of intellectual disability (2,7,19). Yet, interestingly, to our understanding there was no information regarding a predictive equation for Fat% based on BMI in individuals with DS.
Bandini and colleages (2) compared the BMI of adolescents with DS to Fat% measured by DEXA and found a moderate correlation (r = 0.49; P<0.01). However, in our study we found a strong correlation between Fat% measured by DEXA and BMI (r = 0.78; P<0.0001; Figure 1B). It is likely that this difference between our study and the findings of Bandini et al. (2) is the result of the subjects' age differences. They studied DS subjects who were ~5 yrs older than the subjects in the present study. This point may demonstrate the need for studies with specific age groups to develop age-specific predictive equations in body composition.
In the present study, the predictive equations Fat% developed for both groups showed a good level of agreement. This finding indicates differences between the averages of the measured and the estimated Fat% close to zero with low dispersion (i.e., limits of agreement relatively narrow). In addition, the intraclass correlation coefficient was calculated to confirm the correlation between the methods (ICC = 0.74 and 0.76; CTL group and DS group respectively; Figure 2A and 2B). Therefore, we believe that the use of BMI as a basis to estimate Fat% formulas is important, especially since many studies have used BMI in different populations (11).
We conclude that probably there are no differences in fat percentage of young individuals with DS relative to a control group without DS. Furthermore, this study presents equations to estimate Fat% developed from the BMI values for DS subjects [Fat% = 2.006 x BMI (kg x [m.sup.-2]) - 20.36] and adolescents without DS [Fat% = 2.199 x BMI (kg x [m.sup.-2]) - 20.86]. These equations appear to predict with accuracy the Fat% calculated by DEXA analysis. Of course it is necessary that additional studies are carried out with more subjects to determine the reproducibility of the proposed equations.
We thank Lincoln Vinicius Silva, Carlos Ernesto Santos Ferreira, Ronaldo Esch Benford Danielle Garcia de Araujo and Fernanda Rodrigues da Silva for technical support. Support Program for Graduate Institutions of Private Education / Higher Education Personnel Improvement Coordination (PROSUSP / CAPES) and the Research Support Foundation of the Federal District (FAPDF - 193 000 963/2015).
Address for correspondence: Milton Rocha Moraes, PhD, Catholic University of Brasilia, Brasilia, DF, Brazil, 71966-700, Email: email@example.com
(1.) Arnell H, Gustafsson J, Ivarsson SA, Anneren G. Growth and pubertal development in Down syndrome. Acta Paediatr. 1996; 85(9):1102-1106.
(2.) Bandini LG, Fleming RK, Scampini R, Gleason J, Must A. Is body mass index a useful measure of excess body fatness in adolescents and young adults with Down syndrome? J Intellect Disabil Res. 2013; 57(11):1050-1057.
(3.) Bittles AH, Glasson EJ. Clinical, social, and ethical implications of changing life expectancy in Down syndrome. Dev Med Child Neurol. 2004; 46(4):282-286.
(4.) Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1(8476):307-310.
(5.) Chumlea WC, Malina M, Rarick GL. Brachymesophalangia of the fifth finger, stature, and weight in children with Down syndrome. J Ment Defic Res. 1981; 25(Pt 1):7-10.
(6.) Cronk CE, Chumlea WC, Roche AF. Assessment of overweight children with trisomy 21. Am J Ment Defic. 1985; 89(4):433-436.
(7.) Deurenberg P, Weststrate JA, Seidell JC. Body mass index as a measure of body fatness: age- and sex-specific prediction formulas. Br J Nutr. 1991; 65(2):105-114.
(8.) Garn SM, Leonard WR, Hawthorne VM. Three limitations of the body mass index. Am J Clin Nutr. 1986; 44(6):996-997.
(9.) Garrow JS. Three limitations of body mass index. Am J Clin Nutr. 1988; 47(3):553.
(10.) Gonzalez-Aguero A, Ara I, Moreno LA, Vicente-Rodriguez G, Casajus JA. Fat and lean masses in youths with Down syndrome: Gender differences. Res Develop Disab. 2011 ; 32(5):1685-1693.
(11.) Horan M, Gibney E, Molloy E, McAuliffe F. Methodologies to assess paediatric adiposity. Ir J Med Sci. 2015; 184(1):53-68.
(12.) Izquierdo-Gomez R, Martinez-Gomez D, Tejero-Gonzalez CM, Cabanas-Sanchez V, Ruiz Ruiz J, Veiga OL. Are poor physical fitness and obesity two features of the adolescent with Down syndrome? Nutr Hosp. 2013; 28(4):1348-1351.
(13.) Loveday SJ, Thompson JM, Mitchell EA. Bioelectrical impedance for measuring percentage body fat in young persons with Down syndrome: Validation with dual-energy absorptiometry. Acta Paediatr. 2012; 101(11):e491-495.
(14.) Maiano C, Normand CL, Aime A, Begarie J. Lifestyle interventions targeting changes in body weight and composition among youth with an intellectual disability: A systematic review. Res Dev Disabil. 2014; 35(8):1914-1926.
(15.) Micozzi MS, Albanes D. Three limitations of the body mass index. Am J Clin Nutr. 1987; 46(2):376-377.
(16.) Myrelid A, Gustafsson J, Ollars B, Anneren G. Growth charts for Down's syndrome from birth to 18 years of age. Arch Dis Child. 2002; 87(2):97-103.
(17.) Nickerson BS, Esco MR, Bicard SC, Russell AR, Williford HN, Schaefer G. Validity of the body adiposity index in adults with Down syndrome. Res Dev Disabil. 2015; 38:9296.
(18.) Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. JAMA. 2006; 295(13):1549-1555.
(19.) Sales MM, Browne RAV, Moraes JFVN, Asano RY, Campbell CSG, Simoes HG. Indice de massa corporal estima percentual de gordura corporal calculado pela espessura de dobras cutaneas em mulheres adultas. Rev bras cienc mov. 2013; 21 (2):5-10.
(20.) Taylor HA, Jr., Coady SA, Levy D, Walker ER, Vasan RS, Liu J, et al. Relationships of BMI to cardiovascular risk factors differ by ethnicity. Obesity (Silver Spring). 2010; 18 (8):1638-1645.
(21.) Yusuf S, Anand S. Body-mass index, abdominal adiposity, and cardiovascular risk. Lancet. 2011; 378(9787):226-227; author reply 8.
The opinions expressed in JEPonline are those of the authors and are not attributable to JEPonline, the editorial staff or the ASEP organization.
Edilson Francisco Nascimento , Michel Kendy Souza , Thiago Santos Rosa , Geiziane Leite Rodrigues Melo , Brande Ranter Alves Sores , Francisco Eric Vale Sousa , Rodrigo Vanerson Passos Neves , Luiz Humberto Rodrigues Souza , Rafael Reis Olher , Loranny Raquel Castro Sousa, Tania Mara Viera Sampaio , Milton Rocha Moraes 
 Graduate Program on Physical Education, Catholic University of Brasilia, Distrito Federal, Brazil,  Sao Francisco College of Education, Maranhao, Brazil,  University of Bahia (UNEB) - DEDC / Campus XII,  Federal Institute of Education, Science and Technology of Goias, Goias, Brazil
Table 1. The Subjects' Baseline Characteristics. CTL DS n 15 26 Age (yrs) 12.9 [+ or -] 3. 18 12.4 [+ or -] 2.68 BMI (kg x [m.sup.-2]) 19.7 [+ or -] 4.5 21.7 [+ or -] 4.64 Fat% Measured (DEXA) 22.4 [+ or -] 12.86 23 [+ or -] 11.84 Fat% Estimated (Equation) 22.42 [+ or -] 9.53 23 [+ or -] 9.24 Mean values and standard deviation of the results for age, body mass index (BMI), body fat percentage (Fat%) measured by DEXA, and estimated by the respective equations through the body mass index for the CTL group and the DS group.
|Printer friendly Cite/link Email Feedback|
|Author:||Nascimento, Edilson Francisco; Souza, Michel Kendy; Rosa, Thiago Santos; Melo, Geiziane Leite Rodrig|
|Publication:||Journal of Exercise Physiology Online|
|Article Type:||Author abstract|
|Date:||Aug 1, 2016|
|Previous Article:||The use of the rating of perceived exertion to monitor and control the training load in Futsal.|
|Next Article:||Acute effects of an active static stretching class on arterial stiffness and blood pressure in young men.|