SciELO - Scientific Electronic Library Online

vol.13 issue3Technical indexes corresponding to the critical speed and the maximal speed of 30 minutes in swimmers with different aerobic performance levelsDevelopment and validation of a new system for talent selection in female artistic gymnastics: the PDGO Battery author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links


Revista Brasileira de Medicina do Esporte

Print version ISSN 1517-8692On-line version ISSN 1806-9940

Rev Bras Med Esporte vol.13 no.3 Niterói May/June 2007 



Validation of anthropometric equations for the estimation of body density in professional soccer players



Paulo Henrique Santos da FonsecaI; João Carlos Bouzas MarinsII; Alexandre Tavares da SilvaIII

IPrograma de Mestrado em Educação Física/UFSC
IIDepartamento de Educação Física/UFV
IIIPrograma de Especialização em Educação Física/UFSM

Correspondence to




The objective of this study was to validate equations which estimate the body density in professional soccer players. Twenty-five soccer players were evaluated, aged 22.7 ± 4.4 years, body mass 73.9 ± 6.6 kg, and height 177.8 ± 5.5 cm, who played in the state championship of the Federação Gaúcha de Futebol in 2004. The validity of 11 anthropometric equations were analyzed by the process: Pearson correlation (r), dependent t test, constant error (EC), total error (ET) and estimated standard error (EPE), using hydrostatic weight as the "gold standard". In the current study of 11 equations, only the equations proposed by Jackson and Pollock (1978) are valid for estimating the body density of professional soccer players, since the other equations analyzed in this study present considerable errors in their estimation.

Keywords: Anthropometric. Validation. Equations.




High performance sports demand constant improvement of knowledge level on their intervenient variables (morphological, physiological, psychological, biomechanical, and cognitive among others). Soccer, due to its specific condition, involves a high number of athletes and is played in different climate conditions, with technical, tactical and physical varied alternatives, becoming hence a sport of high complexity of interpretation and study.

Under the perspective of this set of variables, body composition assessment represents one of the important identification elements of the soccer player profile.

Works published in journals which describe morphological characteristics of soccer athletes in its majority make use of a method called anthropometry. Such method uses measurement quantification in the study of size, shape, proportionality, composition and maturation of the human body(1). After obtaining the measurement values, these are applied in equations and, via calculations, it is possible to fractionate the human body composition in fat, muscular, bone and residual mass through the analysis of the relative value in percentage with the total body fat value. Such outcome is usually analyzed by technicians.

In order to make use of an equation in a given population, it is necessary that this equation has reached scientific criteria in the evaluated population, that is, it should be valid to measure what it is proposed to measure, or else, it can underestimate or hyper estimate values, adding an assessment error, which leads to misleading diagnosis, prescription and training control(2-5).

It is observed that in Brazil, both in professional soccer practice and in the academic field, equations coming from samples of foreign athletes or national generalized equations have been systematically used. In other words, equations which have used a wider sample for elaboration, which probably will imply in an evaluation error(2,4,6). Thus, soccer coaches should have as reference a specific equation which considers these athletes' characteristics.

Once this methodological issue has been identified, the study had as aim to analyze the validation of the anthropometric equations for body density estimation in professional soccer athletes.



The studied group consisted of professional soccer athletes (n = 25) mean age 22.7 ± 4.4 years and between 18 and 32 years, with a minimum of 2 months of training and 4 hours of daily training, comprised in the competition season promoted by the Soccer Federation of Rio grande do Sul State, year of 2004.

The anthropometric measurements were taken(7) with adaptations, respecting the measurement procedure concerning validation of the original equation, in which the analyzed measurements were: body mass (BM) and height (H), with the use of a scale and stadiometers (RIW 200, Welmy, Brazil). The measured perimeters were the ones from the forearm (FA P.) and abdomen (AB P.), with a measuring tape (Cescorf Científico, Cescorf, Brazil). For the skin-folds, the chosen registers were: triceps (TR. SF), biceps (BI. SF), subscapular (SS. SF), chest (C.SF), media underarm (MUA SF), supra- iliac (SI. SF), horizontal abdominal (ABh. SF), vertical abdominal (ABv SF), medium thigh (MT SF) and medium leg (ML SF), measured with the skinfold dividers (Cescorf Científico, Cescorf, Brazil).

The anthropometric procedures adopted were the following: a) for all measurements, the right side of the athlete was adopted; b) the measurement was conducted in a rotation system with three measurements, being the mean taken as final measurement. These measurements were taken by a single evaluator, c) the non-performance of training physical activities at least for 4 hours prior to the data collection, d) checking of the used instruments as well as the room temperature where collection would take place, which was standardized between 24ºC and 26ºC for all assessments. In addition to the previous procedures, e) the trustworthiness in the measurements of the single evaluator involved in the data collection was also tested.

The value of the hydrostatic weighing (HW), gold standard method, was obtained with the use of a 1 meter and 50 centimeters high tank and had the aid of two evaluators: on inside the tank, who helped and explained to the volunteer the testing methodology, and another who performed the scale reading (Filizola L, Filizola, Brazil) which had 5 grams resolution and 6 kg capacity. The volunteer was wearing a bathing suit during the measuring. Grouped position(8) was used as submersion position of the individual in water, and he was motivated to eliminate all the air kept in the lungs and air ways through expiration.

Breathing was kept blocked for approximately 5 to 10 seconds, for scale's stabilization, when the weighing reading was then recorded. This same procedure was repeated 6 to 10 times. The mean of the three last highest readings was used as hydrostatic weighing value. Whenever the values of the three last readings diverged in more than 50 grams, additional weighing was performed and all obtained values in the readings were recorded. Water temperature was standardized at 31ºC for all evaluations and measured with the aid of a thermometer (Incoterm, Brazil). Besides the temperature, it was tried to control whether the volunteers had performed physical activities and intaken ant kind of meal before the evaluation. The used instruments were also checked.

The residual volume (RV) was obtained through the equation(9):

Men Þ RV = 0.017 (age, years) + 0.027 (height, cm) - 3.477

After the hydrostatic weight and residual weight check, the values were applied in the equation which determines body density.

D = Body Density, in g/cm3;
W = Body Mass in kg on air;
Ww = weight in water in kg;
Wd = Water density (corrected by the temperature);
RV = residual volume, liters;
0.1 = gastrointestinal gas constant (100 ml).

The evaluation procedures were performed in the following stages: first, the aim of the study, the methodological procedures to be developed as well as the authorization of the Ethics in Research Committee were presented to the athletes. Later, the athletes read the consent form and after having agreed on all items, they signed an agreement form as research subjects. The evaluations were performed in the Cine Anthropometry, Measurements and Evaluation Laboratory, situated at the Santa Maria Federal University, in the Rio Grande do Sul State.

Data collection was first performed with the anthropometric measurements and later the individual was taken to the tank where the hydrostatic weighing occurred (HW).

Table 1 presents the list of the 11 equations proposed by several authors and which will be analyzed in this study. Table 1 also shows sample's age, correlation value and estimated standard error in the original validation studies.



Statistical analysis followed these procedures: first, the data normality was tested through the Shapiro-Wilk test, and it detected the need for parametric statistics(17). In order to analyze trustworthiness, which consists of results reproducibility from the part of the evaluator, a pre-test and post-test evaluation and their results were verified through the statistical process correlation moment by Pearson as well as dependent "t" test with significance level of p < 0.05. Descriptive analysis was used for validation of equations, which consists of mean results (X), of the standard deviation (s), of the minimum value (Min), of the maximum value (Max) and the variation coefficient (VC), besides the validation criteria(18): Pearson correlation (r), dependent "t" test, constant error (CE), total error (TE) and estimated standard error (ESE), being all analyzed in the statistical program SPSS 8.0 for Windows.



The outcomes obtained in this study are presented in tables. Table 2 presents the descriptive characteristics of the group of athletes studied, including the data of hydrostatic weighing, skinfolds and perimeter data. Table 3 presents the outcomes of the validation criteria of the 11 equations.





The test and re-test values showed high reliability from the part of the evaluator, both for the Pearson correlation and the dependent "t" test, being all values statistically significant (p < 0.05). These results guarantee the trustworthiness of the values obtained through anthropometry and densitometry, allowing hence the evaluation of the group of athletes studied.



The found results for body mass and height are normal when compared with the results of other national(6,19-25) and international athletes(4,26-33), being lower than in the study with athletes from Yugoslavia(34) and professional athletes from Iceland(35) and Germany(29).

Comparing the skinfold values with studies involving national athletes, these do not present remarkable difference when compared with athletes from Paraíba(6) and Santa Catarina states, and are similar to the athletes from Rio Grande do Sul state(20). Such similarity is probably due to the fact that the group studied is from the same region.

Body density values in most of the equations studied presented higher results than the ones from their original validation studies, except for the EQ2 and EQ5 equations, which presented similar values, and for EQ1, which obtained result below the results found in its original study.

When analyzing validation criteria, they reveal that the EQ6 equation hyper estimates body density of the subjects below the mean and underestimates body density of the subjects above the mean, once its values of standard deviation are lower than the ones of the hydrostatic weighing, an outcome similar to the one found in a study with soccer athletes from the sub-20 category(36). In the EQ1, 2, 3, 4, 5, 7, 8, 9, 10 and 11 equations the opposite fact occurs; they underestimate body density of the subjects below the mean and hyper estimate body density of the subjects above the mean, since its values of standard deviation are higher than the body density measured by hydrostatic weighing.

The second validation criterion, the "t" test, showed that only the EQ6, 10 and 11 equations do not differentiate from the standard measurement, while all the remaining presented statistical differences. The equations have reached moderate correlations and the highest value was obtained by an equation (EQ3) which had as validation group athletic individuals.

The correlation values behaved fairly below the original validation studies (table 1), and not even the EQ3 and EQ7 equations, which used athletes in their building process, repeated the same values.

Constant error (CE), which is the subtraction of the body density measurement estimated by the body density measured through hydrostatic weighing, shows that, except for the EQ11, which presents positive value, the remaining presented lower values of body density than the standard method. It is important to highlight that for the constant error (CE) and the estimated standard error (ESE), the lower its measurement; the higher the safety will be affected by the sample's variability.

When the CE is analyzed, one can note that the EQ4, 6, 10 and 11 equations obtained the lowest values, while for ESE an index around 0.0098 g/cm3 was chosen in relation to the mean, which corresponds approximately to 4% of body fat, as an accuracy threshold with which the body density values would be predicted through anthropometric variables(12). In the present study, all used equations presented measurements lower than this cutting point and it had its values lower than the values from original studies. A hypothesis for this result would be a small number of analyzed individuals as well as by the homogeneity of their measurements.



In an attempt to answer the aim of this study to analyze the validation of anthropometric equations in the estimation of body density in professional soccer athletes, it is concluded that:

In the present study, from the 11 equations analyzed, only the equations proposed by Jackson e Pollock (1978), EQ10 and EQ11, respectively from seven and three skinfolds, responded to the validation criteria, having moderate correlation with the hydrostatic weighing used as gold standard measurement, though.

The preference over the equation of three skinfolds is recommended in the performance of the body density assessment in professional soccer athletes, once it is practical and maximizes the evaluators' time.



1. Thorland WG, Johnson GO, Fagot TG, Tharp GD, Hammer RW. Body composition and somatotype characteristics of Junior Olympic athletes. Med Sci Sports Exerc. 1981;13(5):332-8.        [ Links ]

2. Jackson AS, Pollock ML. Prediction accuracy of body density, lean body weight, and total body volume equations. Med Sci Sports. 1977;9(4):197-201.        [ Links ]

3. Glaner MF, Rodrigues-Añez CR. Validação de procedimentos antropométricos para estimar a densidade corporal e percentual de gordura em militares masculinos. Revista Brasileira de Cineantropometria & Desempenho Humano. 1999;1(1):24-9.        [ Links ]

4. Vicente JGV, López JG, Pascual CM. Influencia de una pretemporada en el perfil cineantropométrico de futbolistas. Archivos de Medicina del Deporte. 2000;17(75): 9-20.        [ Links ]

5. Di Salvo V, Fagnani F, De Sanctis A. Functional assessment in football players. Österreichisches Journal für Sportmeidizin. 2001;2:13-6.        [ Links ]

6. Souza J. Variáveis antropométricas, metabólicas e neuromotoras de jogadores de futebol das categorias mirim, infantil, juvenil e júnior e em relação à posição de jogo: um estudo comparativo. Revista Treinamento Desportivo. 1999;4(3):43-8.         [ Links ]

7. Ross WD, Marfell-Jones MT. Kinanthropometry. In. MacDougall JD, Wenger HA, Green HJ, editores. Physiological testing of the high-performance. 2nd ed. Champaign, Illinois. Human Kinetics Publishers; 1991.        [ Links ]

8. Petroski EL, Pires Neto CS. Análise do peso hidrostático nas posições sentada e grupada em homens e mulheres. Resista Kinesis.1993;10:49-62.        [ Links ]

9. Goldman HI, Becklake MR. Respiratory function tests: normal values of medium altitudes and the prediction of normal results. Am Rev Respir Dis. 1959;79:457-67.        [ Links ]

10. Katch FI, McArdle WD. Prediction of body density from simple anthropometric measurements in college-age men and women. Hum Biol. 1973;45(3):445-55.        [ Links ]

11. 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;32:77-92.        [ Links ]

12. Lohman TG. Skinfolds and body density and their relation to body fatness: a review. Hum Biol. 1981;53(2):181-225.        [ Links ]

13. Thorland WG, Johnson GO, Tharp GD, Housh TJ, Cisar CJ. Estimation of body density in adolescent athletes. Hum Biol. 1984;56(8):439-48.         [ Links ]

14. Guedes DP. Estudo da gordura corporal através da mensuração dos valores de densidade corporal e da espessura de dobras cutâneas em universitários [dissertação]. Universidade Federal de Santa Maria, 1985.        [ Links ]

15. Petroski EL. Desenvolvimento e validação de equações generalizadas para a estimativa da densidade corporal em adultos [tese]. Universidade Federal de Santa Maria, 1995.        [ Links ]

16. Jackson AS, Pollock ML. Generalized equations for predicting body density of men. Br J Nutr. 1978;40:497-504.        [ Links ]

17. Vincent WJ. Statistics in kinesiology. 2nd ed. Champaign: Human Kinetics Publishers, 1999.        [ Links ]

18. Lohman TG. Advances in body composition assessment. Champaign: Human Kinetics Publishers, 1992.        [ Links ]

19. Silva SG, Pereira JL, Kaiss L, Kulaitis A, Silva M. Diferenças antropométricas e metabólicas entre jogadores de futebol das categorias profissional, junior e juvenil. Revista Treinamento Desportivo. 1997;2(3):35-9.        [ Links ]

20. Schwingel AC, Petroski EL, Velho NM. Análise morfológica de jogadores profissionais de futebol de campo. Revista da APEF. 1997;12(1):5-11.        [ Links ]

21. Sousa MSC, Sousa SJG, Santos JP, Torres MS, Gonçalves A. O percentual de gordura em atletas profissionais de futebol segundo diferentes métodos: ensaio envolvendo condições desportivas e da saúde. Revista Brasileira de Atividade Física & Saúde. 1999;4(3):63-74.        [ Links ]

22. Arruda M, Rinaldi W. Utilização da potência muscular no futebol: um estudo da especificidade em jogadores de diferentes posições. Revista Treinamento Desportivo. 1999;4(3):35-42.        [ Links ]

23. Osiecki RL, Gomes AC, Meira ALJ, Erichsen AO, Silva SG. Estudo comparativo dos aspectos funcionais e de composição corporal entre atletas de futebol de diferentes categorias. Revista Brasileira de Fisiologia do Exercício. 2002;1(1): 75-87.        [ Links ]

24. Balikian P, Lourenção A, Ribeiro LFP, Festuccia WTL, Neiva CM. Consumo máximo de oxigênio e limiar anaeróbio de jogadores de futebol: comparação entre as diferentes posições. Rev Bras Med Esporte. 2002;8(2):32-4.         [ Links ]

25. Fonseca PHS, Rech CR, Moura JAR, Zinn JL. Análise antropométrica e peso hidrostático de atletas de futebol com idade de 16 a 20 anos da cidade de Santa Maria. Anais XXI Simpósio Nacional de Educação Física. Pelotas/RS, Brasil; 2002.p. 25-30.         [ Links ]

26. Raven PB, Gettman LR, Pollock ML, Cooper KH. A physiological evaluation of professional soccer players. Br J Sports Med. 1976;10:209-16.         [ Links ]

27. Santos JAR. Estudo comparativo, fisiológico, antropométrico, e motor entre futebolistas de diferente nível competitivo. Revista Paulista de Educação Física. 1999;13(2):146-59.        [ Links ]

28. Edwards AM, Clark N, Macfadyen AM. Lactate and ventilatory thresholds reflect the training status of professional soccer players where maximum aerobic power is unchanged. Journal of Sports Science and Medicine. 2003;2:23-9.        [ Links ]

29. Mayer T, Ohlendorf K, Kindermann W. Konditionelle fähigkeiten deutscher spitzenfuBballer im längsschnitt. Deutsche Zeitschrift für Sportmedizin. 2000;51:271-7.        [ Links ]

30. Chmura J, Nazar K, Kaciuba-Uscilko H, Pilis W, Aust F, Wisnik P. Psychomotorischer leistungsfähigkeit und laktat-sowie adrenalin und noradrenalinschwelle bei fuBballspielern im verlauf ansteigender belastungsintensität. Österreichisches Journal für Sportmedizin. 2001;2:30.        [ Links ]

31. Helgerud J, Engen LC, Wisløff U, Hoof J. Aerobic endurance training improves soccer performance. Med Sci Sports Exerc. 2001;33(11):1925-31.        [ Links ]

32. Dauty M, Bryand F, Potiron-Josse M. Relation entre la force isocinétique, le saut et le sprint chez le footballeur de haut niveau. Science & Sport. 2002;17:122-7.        [ Links ]

33. Ferret JM, Mathian B, Dupuis JM, Martin G, Peretti E, David M. Variations des taux d'androgènes et de cortisol au cours de six saisons chez des footballeurs professionnels. Science & Sports. 2004;19:19-27.        [ Links ]

34. Ostojic SM, Mazic S. Effects of a carbohydrate-electrolyte drynk on specific soccer tests and performance. Journal of Sports Science and Medicine. 2002;1:47-53.        [ Links ]

35. Arnason A, Sigurdsson SB, Gudmunssonn A, Holme I, Engebretsen L, Bahr R. Physical fitness, injuries, and team performance in soccer. Med Sci Sports Exerc. 2004;36:278-85.        [ Links ]

36. Moura JAR, Rech CR, Fonseca PHS, Zinn JL. Validação de equações para a estimativa da densidade corporal em atletas de futebol categoria sub-20. Revista Brasileira de Cineantropometria & Desempenho Humano. 2003;5(2):22-32.        [ Links ]



Correspondence to:
Av. Medianeira, 1.879, apto. 603
97060-003 - Santa Maria, RS, Brazil

Approved in 13/10/06



All the authors declared there is not any potential conflict of interests regarding this article.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License