Recently, there has been concern about obesity and its co-morbidities among football linemen from the youth to professional level. The purpose of this study was two-fold: (1) to examine the association between body mass index (BMI) and percentage fat determined by air-displacement plethysmography (ADP) and hydrodensitometry (HW) in high school football linemen and (2) to compare ADP with HW in high school football linemen to determine the feasibility and accuracy of ADP compared with HW in this group. Thirty-one high school football linemen were recruited from local high schools and were measured for height, body mass, and percentage fat estimated from ADP and HW. BMI (kg/m ²) was calculated from the measured height and body mass.

A majority of subjects were at-risk of being overweight (32%) or overweight (55%), and 6.5% had BMI ≥ 35 kg/m². The mean percentage fat approximated 20% for both ADP and HW measures. BMI was significantly correlated with percentage body fat for ADP (r=0.68) and HW (r=0.76). BMI was significantly correlated with fat mass (FM) and fat-free mass (FFM) for both ADP (FM r=0.83, FFM r=0.49) and HW (FM r=0.88, FFM r=0.57). When analyzed by weight status (non-overweight or overweight), HW was significantly correlated with BMI for the non-overweight group (r=0.72) and overweight group (r=0.54). FM was significantly correlated with BMI for both ADP and HW, respectively, in the nonoverweight (r=0.60 and r=0.75) and overweight (r=0.64 and r=0.78) groups. Significant correlations were also noted between FFM and BMI for both ADP (r=0.68) and HW (r=0.56) but only in the overweight group. Percentage fat based upon ADP and HW were significantly correlated (r=0.68) and the mean difference between HW and ADP was -1.4%. Group analysis revealed a significant correlation in body fat percentage between ADP and HW (r=0.78) only in the non-overweight group. BMI showed a high sensitivity (1.0) for identifying subjects with high body fat as overweight and a moderately high specificity (0.6-0.7) for identifying subjects with lower body fat as normal weight, i.e. suggesting some tendency to misclassify subjects with lower body fat as overweight.

In conclusion, although the BMI correlated reasonably well with fatness in this group and can be used as a first step in the screening process for overweight and obesity, caution should be used due to the tendency that individuals with lower body fat values may be misclassified as overweight about half the time. These two methods of body composition analysis showed that ADP tended to overestimate percentage body fat by ∼1.4% compared to HW; however, there was considerable variation in the error among the sample. Furthermore, ADP showed a slightly greater tendency to overestimate body fat percentage as the body size of subjects increased. Although there are several limitations to both procedures, the advantage of ADP in this group was its greater ease of measurement and acceptability compared with HW.