http://orcid.org/0000-0002-0724-1961
Figures
Abstract
Introduction
In sports related to low body weight, such as classical ballet, the assessment of body composition is important for monitoring performance and health status. This study aimed to cross-validate anthropometry and bioelectrical impedance (BIA) predictive equations for estimating body composition of non-professional classical ballet dancers, using dual-energy-X-ray absorptiometry (DXA) as a reference method.
Materials and methods
Thirty-seven female non-professional classical dancers (median age of 19 years), at intermediate/advanced level, were evaluated in a cross-sectional study. Body composition was evaluated by DXA, anthropometry and tetrapolar BIA. Twenty different predictive equations of anthropometry (n = 8) and BIA (n = 12) were used to estimate Body Fat (BF) and Fat-Free Mass (FFM), testing their validity against DXA using the Bland-Altman statistics.
Results
For BF estimated by anthropometry equations, just one equation showed agreement with DXA (r = 0.852, p < 0.0005; p = 0.600 for one sample T-test). According to the Bland-Altman analysis, this equation also showed validity, with the absence of proportional bias. Regarding the predictive BIA equations tested, none were valid for our study group.
Conclusion
Only one of the anthropometric equations, the one proposed by Durnin and Womerley (1974), but none of the BIA equations analyzed, was valid for the evaluation of body composition of the studied classical dancers. Our results emphasize the importance of previous cross-validation of existing equations or the development of specific equations for body composition assessment in specific populations.
Citation: Leal LLA, Barbosa GSL, Ferreira RLU, Avelino EB, Bezerra AN, Vale SHdL, et al. (2019) Cross-validation of prediction equations for estimating body composition in ballet dancers. PLoS ONE 14(7): e0219045. https://doi.org/10.1371/journal.pone.0219045
Editor: Cherilyn N. McLester, Kennesaw State University, UNITED STATES
Received: November 22, 2018; Accepted: June 16, 2019; Published: July 2, 2019
Copyright: © 2019 Leal et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Classical ballet is a sport that performing requires good physical training, muscular strength, and endurance [1–4]. The practice of ballet also requires, at different levels of training, a light body with a low percentage of body fat (BF) [5,6] that is placed as necessary for and related to the quality of the dance movements [7–9].
Few girls who practice classical ballet become professionals as adults. Nevertheless, studies have shown that even practitioners of this activity at a non-professional level suffer from pressures for an aesthetic body shape suitable for dancing, which brings impacts on body composition and self-steem [8,10,11]. Thus, studying body composition in ballet dancers is important not only for a better performance, but for the promotion of a healthy nutritional status. Monitoring body composition in dancers can also restrict unnecessary weight loss and the side effects resulting from nutritional, metabolic, and musculoskeletal disorders, such as amenorrhea, eating disorders, and osteoporosis [5,12–14].
In the last decades, dual-energy-X-ray absorptiometry (DXA) analysis has gained credibility and acceptance for body composition assessment, being considered a reference method [15]. Though not commonly available to athletes, DXA has been used to evaluate body composition for different sport modalities, such as basketball, judo, handball, football [16–19] and dance [20–22].
Methods more commonly available to all types of athletes include anthropometry and bioelectrical impedance (BIA) prediction equations, which are used to estimate fat mass (FM) and fat-free mass (FFM) [23,24]. Athletes vary greatly in physique depending on their sport and this is pertinent given that precision errors should be specific to the population studied [23,25]. Additionally, using different prediction equations in populations not similar to that of the development of the equation may give incorrect estimates. For classical ballet dancers, the existing prediction equations were developed using professional dancers of specific ethnicities, one for Greek [26] and the other for American dancers [8]. Thus, the direct application of these equations may give inaccurate results if applied to other populations.
The aim of the present study was to cross-validate BF and FFM anthropometry and BIA prediction equations for body composition assessment, using DXA as a reference method, in non-professional classical ballet dancers. We hypothesized that not all existing anthropometry and BIA prediction equations would be applicable to the study population, considering DXA as a reference method.
Materials and methods
Ethics, design and study population
According to the Helsinki Declaration, all participants received detailed information about the study and were invited to give written consent, which was given in the presence of their parents or guardians when necessary. The study was approved by the Ethics Committee of the Onofre Lopes University Hospital, of the Federal University of Rio Grande do Norte (CAAE protocol 38086214.2.0000.5292, acceptance number 925.040).
In the initial screening of the female classical ballet dancers population in Natal-RN/Brazil, in February 2016, we found 50 non-professional classical ballet dancers training in intermediate/advanced level (considered as minimum training of 6 hours per week, using point shoes for at least one year) [27]. These dancers were included in the study from June 2016 to April 2018 according to their availability and acceptance of participating in data collection. Participants answered a questionnaire about their personal data, including self-referred sexual maturation [28], use of medications, and physical activity practice. None of the participants presented any disease or medication use on a regular basis that could interfere with the results, such as antidepressants, antipsychotics, anticonvulsant and hypoglycemic drugs, besides not practicing any other physical activity that exceeded the practice of classical ballet.
Body composition assessment
Body composition measurements were performed at the Onofre Lopes University Hospital (HUOL-UFRN). All measurements were performed on the same morning for each dancer, after 8–10 hours of overnight fasting. Participants were asked not to perform physical exercises or drink alcohol in the 12 hours before the exams. As specific preparation for BIA, participants were asked to ingest at least 2 L of water in the previous day, besides not ingesting coffee and avoiding the use of diuretics (including teas) within 12 hours prior to the examination. To control body water retention, data collection was not performed during the menstrual period of the evaluated subjects. For DXA, as well as BIA, participants were required to remove accessories with metals.
DXA was performed using Lunar DPX L/GE X-ray system (Madison, WI, USA) with the addition of a pediatric software. The examination was performed with the participant lying immobile in dorsal decubitus for evaluation of the entire body, with knees and ankles immobilized with a soft velcro tape. The evaluation determined BF and FFM in percentiles/scores and in pounds, which were converted to kilograms (Kg).
To evaluate the accuracy of DXA scans, the DXA equipment was calibrated daily using a soft tissue phantom, according to the manufacturer’s instructions, allowing a coefficient of variation of ± 3%. The DXA equipment was also routinely evaluated for precision after every 100 patients scan. The Least Significant Change (LSC) was calculated, using the precision calculation tool from the International Society for Clinical Densitometry (ISCD), as recommended (https://www.iscd.org/resources/calculators/precision-calculator/). The LSC calculation was done using repeated measurements in 15 patients scanned three times. Only one technologist performed the assessments and precision assessments. Additionally, assessments were performed two times for dancers from our sample. In these dancers, coefficients of variation (CV) were calculated for FFM and FM and they varied from 0.21% to 1.21%, with mean 0.80%.
For the anthropometric evaluation, an electronic P200C anthropometric scale (Líder) was used, with a capacity of 200 Kg and a reported accuracy of ± 100 g for weight measurements and accuracy of ± 0.1 mm for height measurements. Body Mass Index (BMI) was calculated and classified as proposed by the World Health Organization for adults [29]. The z-scores of BMI-for-age and Height-for-age indicators were calculated with the Anthroplus program for adolescents [30].
To measure the skinfolds, a Lange adipometer (Beta Technology Inc., Houston, Texas) was used, with an acuracy of ± 0.1 mm, at the following cutaneous sites: subscapular, tricipital, bicipital, medial axillary, suprailiac, abdominal, thigh and calf. The procedure used for the measurement of the skinfolds followed the recommendations and anatomical sites considered in the literature [31]. Measurements were performed alternately on the participant’s right side and in triplicate. The value used for the calculations was the mean of the triplicates. The maximum difference accepted in the procedures for a new measurement was ± 10% of the value of each measure. Two trained evaluators were responsible for the measurement of skinfolds. Technical error of measurement (TEM) was assessed and intra-evaluator TEM was within ± 5.0% and inter-evaluator TEM within ± 6.0% [32].
For BIA evaluation, the Quantum II tetrapolar tool (RJL Systems, Michigan, USA) was used, following the method described by Lukaski et al. [33]. Resistance (Ω) and reactance (Ω) were measured with the subject lying supine, with four surface self-adhesive spot electrodes and a standard conduction current of 800 PA and 50 kHz. Two electrodes were placed on the dorsal surface of the right hand, and two electrodes were placed on the dorsal surface of the right foot as recommended.
Age, sex and anthropometry data (weight, height, skinfolds) and BIA (resistance and reactance) were used in predictive equations to estimate body composition. The predictive equations of BF and FFM were selected by research in PubMed, Scielo and Portal Periódicos Capes, a virtual library available in Brazil, using the following keywords and Boolean operators: body composition AND (predictive OR estimate) AND equation AND validation AND bioimpedance OR anthropometry OR skinfolds). We excluded studies with equations involving different age groups of the present study, equations validated only for males and those developed for populations of specific athletes different from the studied population or for populations with specific diseases. Only studies conducted with similar equipment were used, resulting in the selection of eight predictive anthropometry equations and twelve predictive BIA equations. Of these, three were developed for dancers, as shown in Table 1.
Body fat value (Kg) was obtained by subtracting the FFM value from the total mass value. Equations were applied in the present study population according to the corresponding age group: For each equation, we assessed only participants within the same age range originally proposed by the equation. The Siri equation [48] was used to calculate the BF (%) from body density.
Statistical analysis
The Shapiro-Wilk test and visual inspection of histograms were used to verify the normality of the variables in the study. The variables with normal distributions were presented in table and figures as mean (SD), while the other variables were presented as median (Q1-Q3). All variables were tested for outliers, evaluating box-plots.
Correlation between BF or FFM results generated by the prediction equations and DXA was performed using Pearson’s correlation (r). This analysis was done once the results generated by the equations presented normal distributions.
To give more robustness to the analysis, the one-sample t-test was used to verify if the mean differences between the results of the equations and the results of DXA were significantly different from zero. This test is commonly used for one measured variable and a theoretical expectation of what the mean should be under the null hypothesis. In this analysis, a significant p-value indicated that the tested predictive equation did not present a good agreement with DXA.
Equations with correlation (r with p < 0.05) and agreement (one-sample t-test with p > 0.05) with DXA, were also assessed by the Bland-Altman analysis for cross-validation. The Bland-Altman plots were constructed by placing on the x-axis the mean between the results of the equation and DXA and on the y-axis the difference between the result of the equation and DXA [49]. A central trend line representing the mean of the differences between the equation and DXA was added, and the lines of the minimum and maximum limits were the standard deviations multiplied by ± 1.96. Then, a simple linear regression analysis was performed. This was used to test the presence of proportional bias between the tested equations and DXA, considering the data present in the Bland-Altman plot. The differences between the equation and DXA values were considered as the dependent variable and the mean between the equation and DXA the independent variable. The presence of proportional bias was assumed when a significant p-value (< 0.05) was found, and the equation was not considered valid with DXA as a reference method. For the simple linear regression, the adjusted R2, the beta coefficient, which determines whether the bias is positive or negative, the standard error of estimate (SEE) and the p-value were reported.
Statistical analyses were performed with the Statistical Package for Social Sciences version 22.0 (SPSS Inc. Chicago, IL) and the Graph Pad Prism version 3.0 (Graph Pad Software, San Diego, CA) programs.
Results
In the present study, 38 classical dancers from the initial 50 elegible were evaluated, considering the availability to participate in the study. One dancer was excluded from the sample because of self-referenced prepubescent sexual maturation, which was not compatible with the other participants, totalizing 37 dancers. The median age was 19 (16–24) years, with a range of 14 and 49 years old, which included 17 adolescents and 20 adults who practiced ballet for 10 (5–15) years, with a median of 9 (6–18) hours a week. The results of anthropometry, BIA and DXA are in Table 2. Adult dancers had a BMI considered within the eutrophy range, as well as the adolescents, who presented z-scores with age-appropriate limits. Mean BF was of 28.37 (7.01)% and median FFM of 68.50 (61.69–72.89)%, using DXA. When evaluated separately, the adolescent dancers’ BF of 28.15 (6.55)% was similar to the adults value of 28.61 (7.70)%. None of the studied variables presented extreme values (more than 3 box-lengths from the edge of the box).
Among the anthropometry equations for BF (%) prediction, only the equation proposed by Durnin and Womersley (1974) [34] showed a significant correlation with DXA (r = 0.852, p <0.0005) (Table 3) and no significant difference in the BF (%) estimate when compared to DXA (one-sample t-test, p = 0.600) (Fig 1A). All other equations tended to underestimate BF (%) when compared to DXA, generating negative differences (Fig 1A). In the case of the twelve BIA equations for FFM (%) prediction, some presented correlation with DXA (Table 4), but all presented differences that were significantly higher or lower than zero for the results obtained from FFM against DXA (Fig 1B).
Error bars represent ± 1.96 x standard deviations of the mean differences. *Equation presented one-sample t-test with p = 0.600, no significant difference for results obtained from BF by the equation against DXA. The other predictive equations presented p values < 0.05 for the one-sample t-test. The doted line represents the reference value expected in the one-sample t-test (zero).
Therefore, for results from the Durnin and Womersley [34] equation, the Bland-Altman plot was constructed and a simple linear regression was performed to verify the cross-validity of the results against those obtained by DXA. The plot revealed that the equation did not underestimate or overestimate BF, with the mean difference of 0.57%. No proportional bias between DXA and the equation was found, once the regression did not explain the difference found in the two methods of BF estimation (R2 = -0.055, p = 0.929) (Fig 2).
The doted lines represent 95% limits of agreement (±1.96 SD). The red line represents the regression line between mean BF by the equation and BF by DXA and mean differences between BF equation and BF DXA. Beta coeficient = 0.021 and SEE = 4.91. Curved lines represent the 95% confidence interval for the regression line.
Discussion
Classical ballet is a sport with performance intrinsically related to body composition and shape [50]. The present study aimed to evaluate cross-validity of eight predictive equations of anthropometry and twelve BIA equations to estimate BF and FFM in classical ballet dancers. To our knowledge, no studies have performed this type of evaluation, using DXA as a reference method and considering non-professional dancers.
The weekly training volume found in the present study was lower than that found in other studies, showing that our results are compatible with a non-professional intermediate/advanced practicing [21,26,51]. Although the dancers of the study presented a eutrophic BMI, they presented a higher BF% when compared to other studies in professional dancers, with percentages ranging from 17.5% to 24.6% [26,52]. Not only time and level of practice can be considered the causes for the higher BF% observed in our study. Ethnicity and higher age and may also be determining variables for the different results [24,53]. As an example, Eliakim et al. [20] found in Caucasian adolescent ballet dancers a lower BF% than that found in the present study [21].
Few studies have analyzed body composition of Brazilian ballet dancers [54–57], one of which studied pre-professional ballet dancers [54] and none used DXA. These studies reported mean ages of 13.1 to 36.8 years and mean BF% values of 15.2% to 21.0%. Most of the studies used skinfolds measurements with Slaughter’s et al. equation [40], and only one used BIA [56] with Houtkooper’s et al. equation [45]. These studies reinforce that the availability and use of DXA in Brazil is still limited for dancers, and it is necessary to validate more accessible methods to evaluate this population.
The Durnin & Wormersley’s [34] equation was the only applicable to the population in our study, once the final Bland-Altman analysis revealed no significant difference in measurements or proportional bias. This equation was developed from research conducted in England in 1974 with 209 males and 272 females, mostly sedentary, aged between 16 and 72 years old.
Interestingly, our analysis showed that the three equations originally developed with dancers [8,26] were not applicable in the evaluated sample. We did not find other studies cross-validating these three equations. The equation proposed by Hergenroeder et al. [8] was developed using 112 professional ballet dancers in the Houston Ballet Academy, with a mean age of 14.4 (1.3) years, dancing 16 (12) hours per week, and validated with the TOBEC method. The authors concluded that the simplicity of the developed equation, using only weight (Kg), could facilitate studies and clinical practice.
Yannakoulia et al. [26] developed two equations validated with 42 dancers of a professional ballet school in Athens. One of the equations was considered the best because it needed only BIA data in its formula. The second equation needed BIA and anthropometry data, which was recognized by the authors more challenging to apply. The equation was also validated with DXA, the ballet dancers had a mean age similar (21 years) to that found in our study (19 years) and were not considered professionals. Dancers trained more hours per week (28.3 hours) than those in the present study, and this may justify the non-applicability of the equations of Yannakoulia et al. [26] in the population herein analyzed.
Studies have reported a correlation between BF and FFM estimates using anthropometry and BIA predictive equations with DXA in several populations, including dancers [21,26,58]. However, no study has sought to cross-validate the existing anthropometry or BIA predictive equations for dancers using DXA or any other method as a reference method.
One limitation of our study is the difference in the standardization for the skinfolds measurements in the different equations. To decrease bias and enable data collection, we obtained anthropometric data using a standardized protocol [31], compatible with most of the selected equations. In addition, most of the mathematical models found in the literature were developed with the hydrostatic weighing method as a reference, as well as methods such as total body electrical conductivity and isotope dilution. These methods differ from DXA because they have different principles of body composition, and this may be a limitation. Our small sample size, determined by the fact we used a very homogeneous specific population of athletes, might also be a limitation to further generalization.
Of importance, there is a tendency in clinical practice to individually follow up some skinfolds to assess the evolution of body composition [23]. This follow up is a problem, especially in evaluating people practicing sports related to aesthetics, since it makes a body analysis impossible, and may induce unnecessary anxiety and fixation about particular sites of the body, inherent to the shape of an individual. Embracing this kind of assessment also limits the quality of collective planning measures and consensuses for actions to improve body composition of populations.
On the other hand, if an anthropometry or BIA predictive equation are going to be used in clinical care to estimate body composition, professionals should be aware to carefully observe the equations, since criteria such as gender, ethnicity, age, and type of physical activity may be important variables that influence the results. Our study showed that not all anthropometry and BIA prediction equations tested would apply to the study population, accepting our initial hypothesis. Thus, previous cross-validation of predictive equations from the literature for a specific population, as we have done in the present study, or developing new equations for specific populations, such as classical ballet practitioners, are of relevance once it can improve the assessment of that population.
Conclusion
The results of the present study showed that from the twenty body composition predictive equations tested, only the Durnin and Womersley [34] equation was applicable to the estimation of body composition in the studied classical ballet dancers. Our data reinforce the importance of cross-validating existing predictive equations to estimate body composition of specific populations or developing equations, considering the characteristics of the population of interest.
Acknowledgments
The authors thank all the ballet dancers who took part in the study and the Endocrinology Sector of the Onofre Lopes University Hospital for the contribution and technical assistance given to the dual-energy-X-ray absorptiometry (DXA) tests performed. The authors declare that there are no potential conflicts of interest regarding the research, authorship and / or publication of this article.
References
- 1. Speranza E, Contaldo F, Pasanisi F. Resting Energy Expenditure, Body Composition and Phase Angle in Anorectic, Ballet Dancers and Constitutionally Lean Males 2019. pmid:30818800
- 2. Wyon M. Cardiorespiratory Training for Dancers. J Danc Med Sci 2005;9.
- 3. Brown MA, Howatson G, Quin E, Redding E, Stevenson EJ. Energy intake and energy expenditure of preprofessional female contemporary dancers. PLoS One 2017;12:1–13. pmid:28212449
- 4. Civil R, Lamb A, Loosmore D, Ross L, Livingstone K, Strachan F, et al. Assessment of Dietary Intake, Energy Status and Factors Associated With RED-S in Vocational Female Ballet Students 2019;5:1–11. pmid:30687712
- 5. Srdic B, Stokic E. Body mass index, body fat mass and the occurrence of amenorrhea in ballet dancers 2005;20:195–9. pmid:16019361
- 6. Beck KL, Mitchell S, Foskett A, Conlon CA, von Hurst PR. Dietary Intake, Anthropometric Characteristics, and Iron and Vitamin D Status of Female Adolescent Ballet Dancers Living in New Zealand. Int J Sport Nutr Exerc Metab 2015;25:335–43. pmid:25386731
- 7. Caroline E, Meereis W. Breve Relato—Movimentos do balé clássico e alterações da postura corporal: corporal um breve relato. Rev Educ Física 2016;85:269–73.
- 8. Hergenroeder AC, Brown B, Klish WJ. Anthropometric measurements and body composition in ballet dancers. Med Sci Sport Exerc 1993;25:145–50.
- 9. Koutedakis Y, Jamurtas A. The Dancer as a Performing Athlete. Sport Med 2004;34:651–61.
- 10. García A, Amado D, Sánchez-miguel PA, Sánchez R. Factors Dancers Associate with their Body Dissatisfaction. Body Image 2018;25:40–7. pmid:29475190
- 11. Lim S-N, Chai J-H, Song JK, Seo M-W, Kim H-B. Comparison of nutritional intake, body composition, bone mineral density, and isokinetic strength in collegiate female dancers. J Exerc Rehabil 2015;11:356–62. pmid:26730387
- 12. Hincapi CA, Cassidy JD. Disordered eating, menstrual disturbances, and low bone mineral density in dancers: A systematic review. Arch Phys Med Rehabil 2010;91:1777–1789.e1. pmid:21044726
- 13. Byrne S, McLean N. Eating disorders in athletes: A review of the literature. J Sci Med Sport 2001;4:145–59. pmid:11548914
- 14. Wewege MA, Ward RE. Bone mineral density in pre-professional female ballet dancers: A systematic review and meta-analysis. J Sci Med Sport 2018:6–11. pmid:29526411
- 15. Shepherd JA, Ng BK, Sommer MJ, Heymsfield SB. Body composition by DXA. Bone 2017. pmid:28625918
- 16. Bilsborough JC, Greenway K, Opar D, Livingstone S, Cordy J, Coutts AJ. The accuracy and precision of DXA for assessing body composition in team sport athletes. J Sports Sci 2014;32:1821–8. pmid:24914773
- 17. López-taylor JR, González-mendoza RG, Gaytán-gonzález A, Jiménez-alvarado JA, Villegas-balcázar M, Jáuregui-ulloa EE, et al. Accuracy of Anthropometric Equations for Estimating Body Fat in Professional Male Soccer Players Compared with DXA. J Sports Med 2018;2018:1–7. pmid:29736402
- 18. Silva B, Clemente FM, Bezerra P, Cam M. Body Composition Evaluation Issue among Young Elite Football Players: DXA Assessment. Sports 2017;5:1–9. pmid:29910377
- 19. Cavedon V, Zancanaro C, Milanese C. Anthropometric prediction of DXA- measured body composition in female team handball players. PeerJ 2018. pmid:30515356
- 20. Gunter K, Ph D, Russo AC, Ph D, Sklar SE. Bone Mineral Density and Body Composition of Collegiate Modern Dancers. J Danc Med Sci 2011;15:31–6.
- 21. Eliakim A, Ish-Shalom S, Giladi A, Falk B, Constantini N. Assessment of body composition in ballett dancers: Correlation among anthropometric measurements, bio-electrical impedance analysis, and dual-energy X-ray absorptiometry. Int J Sports Med 2000;21:598–601. pmid:11156282
- 22. Van Marken Lichtenbelt WD, Fogelholm M, Ottenheijm R, Westerterp KR. Physical activity, body composition and bone density in ballet dancers. Br J Nutr 1995;74:439–51. pmid:7577885
- 23. Santos DA, Dawson JA, Matias CN, Rocha PM, Minderico CS, Allison DB, et al. Reference values for body composition and anthropometric measurements in athletes. PLoS One 2014;9:1–11. pmid:24830292
- 24. Thibault R, Genton L, Pichard C. Body composition: Why, when and for who? Clin Nutr 2012;31:435–47. pmid:22296871
- 25. Buehring B, Krueger D, Libber J, Heiderscheit B, Sanfilippo J, Johnson B, et al. Dual-Energy X-Ray Absorptiometry Measured Regional Body Composition Least Significant Change: Effect of Region of Interest and Gender in Athletes. J Clin Densitom 2014;17:121–8. pmid:23643968
- 26. Yannakoulia M, Keramopoulos A, Tsakalakos N, Matalas AL. Body composition in dancers: the bioelectrical impedance method. Med Sci Sports Exerc 2000;32:228–34. pmid:10647554
- 27.
Sampaio F. Ballet passo a passo: história, técnica, terminologia. Fortaleza: Expressão Gráfica e Editora; 2013.
- 28.
Tanner J. Growth at adolescence. Oxford: Blackwell; 1962.
- 29.
WHO. Physical status: the use and interpretation of anthropometry. Report of a WHO Expert Committee. World Health Organ Tech Rep Ser 1995;854:1–452. pmid:8594834
- 30. De Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. Development of a WHO growth reference for school-aged children and adolescents. Bull World Heal Organ 2007;85:812–9.
- 31.
Ross W, Carr R, Carter J. Antropometry illustrated: a browser ased interactive textbook and learning system. 1st ed. The human animal series.; 2000.
- 32.
Pederson D, Gore C. Error en la medición antropométrica. In: Norton K, Olds T, editors. Antropometrica., Biosystem; 2000.
- 33. Lukaski HC, Bolonchuk WW, Hall CB, Siders WA. Validation of tetrapolar bioelectrical impedance method to assess human body composition. J Appl Physiol 1986;60:1327–32. pmid:3700310
- 34. Durnin JVGA, Womersley J. Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 Years. Br J Nutr 1974;32:77–97. pmid:4843734
- 35. Jackson AJ, Pollock ML, Ward A. Generalized equations for predicting body density of woman. Med Sci Sports Exerc 1980;12:175–82. pmid:7402053
- 36. Guedes D. Estudo da gordural corporal através da mensuração dos valores de densidade corporal e da espessura de dobras cutâneas em universitários. Kinesis 1985;1:183–212.
- 37. Petroski EEL, Pires-Neto CS, Neto CP. Validação de Equações Antropométricas para estimativa da densidade Corporal em Mulheres. Rev Bras Atividade Física e Saúde 1995;2:65–73.
- 38. Sloan AW, Burt JJ, Blyth CS. Estimation of body fat in young women. J Appl Physiol 1962;17:967–70. pmid:13989261
- 39. Pollock ML, Hickman T, Kendrick Z, Jackson A, Linnerud AC, Dawson G. Prediction of body density in young and middle-aged women. J Appl Physiol 1975;40:300–4. pmid:1141107
- 40. Slaughter MH, Lohman TG, Boileau RA, Horswill CA, Stillman RJ, Van Loan MD, et al. Skinfold equations for estimation of body fatness in children and youths. Hum Biol 1988;60:709–23. pmid:3224965
- 41. Chumlea WC, Guo SS, Kuczmarski RJ, Flegal KM, Johnson CL, Heymsfield SB, et al. Body composition estimates from NHANES III bioelectrical impedance data. Int J Obes 2002;26:1596–609. pmid:12461676
- 42. Segal KR, Van Loan M, Fitzgerald PI, Hodgdon JA, Van Itallie TB. Lean body mass estimation by bioelectrical impedance analysis : a four-site cross-Validation. Am J Clin Nutr 1988;47:7–14. pmid:3337041
- 43. Gray D, Bray G, Gemayel N, Kaplan K. Effect of obesity on bioelectrical impedance. Am J Clin Nutr 1989;50:255–60. pmid:2756912
- 44. Deurenberg P, Weststrate AJ, Seidell CJ. Body mass index as a measure of body fatness: age-and sex- specific prediction formulas. Br Journul Oj Nutr 1991;65:105–14.
- 45. Houtkooper LB, Going SB, Lohman TG, Roche aF, Van Loan M. Bioelectrical impedance estimation of fat-free body mass in children and youth: a cross-validation study. J Appl Physiol 1992;72:366–73. pmid:1537738
- 46. Kyle UG, Genton L, Karsegard L, Slosman DO, Pichard C. Single prediction equation for bioelectrical impedance analysis in adults aged 20–94 years. Nutrition 2001;17:248–53. pmid:11312069
- 47. Sun S, Chumlea W, Heymsfield S, Lukaski H, Schoeller D, Friedl K, et al. Development of bioelectrical impedance analysis prediction equations for body composition with the use of a multicomponent model for use in epidemiological surveys. Am J Clin Nutr 2003;77:331–40. pmid:12540391
- 48.
Siri WE. Body composition from fluid space and density: analysis of method. In: J B, A. H, editors. Tech. Meas. body Compos., Washington: National Academy of Science and Natural Resource Council; 1961.
- 49. Bland U, Giavarina D. Lessons in biostatistics 2015;25:141–51.
- 50. Grochowska-Niedworok E, Kardas M, Fatyga E, Piórkowska-Staniek K, Muc-Wierzgoń M, et al. Study of top ballet school students revealed large deficiencies in their body weight and body fat. Acta Paediatr 2017. pmid:29284180
- 51. Łagowska K, Kapczuk K. Testosterone concentrations in female athletes and ballet dancers with menstrual disorders. Eur J Sport Sci 2016;16:490–7. pmid:25951882
- 52. Liiv H, Wyon M, Jurimae T, Purge P, Saar M, Maestu J, et al. Anthropometry and somatotypes of competitive DanceSport participants: A comparison of three different styles. HOMO- J Comp Hum Biol 2014;65:155–60. pmid:24182600
- 53. Roman EP, Filho A de AB. Diferenças no crescimento e na composição corporal entre escolares de origem germânica e brasileira 2007;25:227–32.
- 54. Da Silva CC, Goldberg TBL, Soares-Caldeira LF, Oliveira RDS, De Paula Ramos S, Nakamura FY. The Effects of 17 Weeks of Ballet Training on the Autonomic Modulation, Hormonal and General Biochemical Profile of Female Adolescents. J Hum Kinet 2015;47:61–71. pmid:26555850
- 55. Amaral R, Pacheco R, Navarro F. Perfil Nutricional E Antropométrico De Praticantes De Ballet. Rev Bras Nutr Esportiva 2008;64:199–212.
- 56. Ribeiro LG, Da Veiga G V. Risk behaviors for eating disorders in Brazilian dancers. Int J Sports Med 2010;31:283–8. pmid:20148375
- 57. Silva CV Da Fayh APT. Perda de peso e consumo de bebidas durante sessões de treinamento de ballet clássico. Rev Bras Ciências Do Esporte 2011;33:495–506.
- 58. Aguirre C, Salazar G. Evaluation of simple body composition methods: assessment of validity in prepubertal Chilean children. Eur J Clin Nutr 2014;69:269–73. pmid:25097002