Abstract:
Body composition assessment of athletes is acknowledged as basic physiological
determinant of athletic health and performance. Assessment of body composition of male
Nigerian University athletes, using foreign-derived anthropometric regression equations
usually brings with it the issue of precision, accuracy and validity. Previous studies largely
focused on developing and validating commonly used equations among foreign athletes,
but little research efforts have been directed towards validating these equations on
Nigerian athletes using the Underwater Weighing (UWW) criterion. This study, therefore,
was carried out to validate the selected anthropometric equations.
The study was anchored to the Theory of Human Body Composition Assessment, while
the ex-post facto design was employed. The equations validated were Brozek and Keys
(BK) 1951, Sloan and Weir (SW) 1970, Sinning (SI) 1974, Forysth and Sinning (FS)
1975, and Jackson and Pollock (JP) 1979 in order to confirm or refute their respective
validity on body composition assessment of male university athletes in Southwestern
Nigeria. The multistage sampling procedure was used. Three first generation federal
universities in Ibadan, Lagos and Ile Ife) were enumerated, using the intact group of
endurance athletes (45), power athletes (45) and control group (45) in each University. The
instruments used were UnderWater Weighing equipment, spirometer, health-o-meter scale
and skinfold calipher. Underwater measurements were taken following Barton and
Cameroon (2009) procedure, while skinfold measurement of abdominal, chest, triceps,
subscapular, suprailiac, and thigh were taken following ISAK (2011) protocol. Data were
analysed using descriptive statistics, Pearson product moment correlation, t-test, and
multiple regression at 0.05 alpha level.
Participants’ age was 24.06±2.25 years. There was no significant difference in physical
characteristics of height, body weight and body density of endurance athletes, power
athletes and control group, but they significantly differed in percent fat and lean body
weight. There was moderate, positive, relationship between Body Density (BD) of UWW
and BD of BK, SW, SI, FS, but JP(r=0.77) was strongly significantly related. There was
positive, relationship between %bf of UWW and %bf of BK(r=0.27), SW(r=0.26),vi
SI(r=0.18), FS(r=0.38) except JP(r=0.72). There was significant difference between BD
determined by UWW and BD of BK(t=-12.33), SW(t=-16.21), SI(t=-11.58), FS(t=-7.75)
and JP(t=-2.92). There was also a significant difference between %bf of UWW and %bf of
BK(t=10.22), SW(t=14.95), SI(t=11.66), FS(t=6.34) and JP(t=8.00). Plotted against
validation criteria of Multiple Correlation Co-efficient (R), Constant Error (CE), Total
Error (TE), Standard Error of Estimate (SEE), the values obtained were BK(R2=.103,
CE=-0.02, TE=0.002, SEE=0.004), SW(R2=.103, CE=0.02, TE=.003, SEE=.006),
SI(R2=.138, CE=-0.02, TE=0.002, SEE=0.01), FS(R2=.209, CE=0.01, TE=0.002,
SEE=0.006), JP(R2 =.208, CE=-0.02, TE=0.002, SEE=0.002). All the examined equations
failed the validity test. As a credible alternative this equation was formulated:
BD=1.064+0.00392 (X1)+0.669 (X2)+0.07761 (X3).
The anthropometric regression equations of Brozek and Keys, Sloan and Weir, Sinning,
Forsyth and Sinning (1975), and Jackson and Pollock, have relationship with, but are
significantly different from underwater weighing. All the equations overestimated Body
Density and underestimated percent body fat in male University athletes in Southwestern
Nigeria. The validated prediction equations should be used with relative caution, while the
equation formulated needs to be adopted by Nigerian male athletes.