Validity of the 20-m MSRT as a predictor of VO2peak in Lisbon elementary school children

Dissertação de mestrado em Exercício e SaúdeSeveral regression equations have been developed to predict cardiovascular fitness (CVF) based on the 20-m multistage shuttle-run test (MSRT) for children and youth. Purpose: The aim of this study was to compare the validity of the FITNESSGRAM's MSRT software reports and previously reported equations as predictors of CVF in Lisbon healthy elementary school children, by measuring the VO2peak during the actual 20-m MSRT performance. Methods: Ninety subjects, 45 boys and 45 girls aged 8-10 years, performed the MSRT and their VO2peak was measured in field using a portable gas analyzer (K4b2, Cosmed, Italy). Heigth and weigth were measured and %Fat was estimated from skinfold thickness. The total laps (TL) completed were recorded and used to calculate maximal speed (MS), and used as predictors in the equations. Results: Using Bland-Altman analytical approach, there were differences between measured and predicted VO2peak. The FITNESSGRAM's reports had the smallest mean difference (d)(ml.kg-1.min-1) between measured and the estimated values (1.84), but also had the highest range (ml.kg-1.min-1) between upper and lower limits of agreement (LA) (28.6) and the highest slope (1.20) meaning that the equation overpredicted VO2peak in unfit participants and underpredicted VO2peak in fit participants. The lowest slope (0.44) and the lowest range of the LA (22.23) were observed using the equation of Fernhall et al (1998). The comparision using the equation of Leger et al (1998) provided a reasonably large d (4.69), one of the highest ranges (27.08) and the second uppermost slope (1.04). The equation of Matsuzaka at al (2004) including MS showed higher d(5.38) than the equation using TL (d=4.23). From the 3 equations of Barnett et al (1993), the equation including skinfold and MS showed the highest d(6.07) but the smallest range (24.09). Conclusion: Previously reported equations might be unsuitable for Lisbon healthy elementary school children. We suggest that the FITNESSGRAM's software reports or the equation of Fernhall et al (1998) may be the most appropriate. However, they still yield unsatisfactory accuracy. Our findings advocate that published equations predicting VO2peak from MSRT may provide reasonable group estimates, but poor individual predictions of VO2peak