Table 2

Overview of fitness criterion validity studies carried out within the framework of the ALPHA study

AuthorParticipantsAge (years)TestGold standardMain outcomeConclusion
Cardiorespiratory fitness
 Ruiz et al6Boys = 122, girls = 7113–1920mSRTPortable gas analyser during 20mSRTLeger's equation: % error=17.13; SEE=4.27 ml/kg/min; mean difference=4.9 ml/kg/minANN-equation: % error=7.38; SEE=2.84 ml/kg/min; mean difference=0.5 ml/kg/minAll the measurements of error were lower in the ANN-based equation (which includes speed, gender, age, weight and height) compared with the equation reported by Leger
 Ruiz et al7Boys = 26, girls = 2213–1920mSRTPortable gas analyser during 20mSRTLeger's equation: r=0.587, SEE=6.5 ml/kg/min, mean difference=5.5 ml/kg/minMatsuzaka's equation: r=0.736, SEE=5.5 ml/kg/min, mean difference=3.2 ml/kg/minBarnett's (a): r=0.757, SEE=5.3 ml/kg/min, mean difference=2.9 ml/kg/minBarnett's (b): r=0.725, SEE=5.6 ml/kg/min, mean difference=1.3 ml/kg/minRuiz's equation: r=0.758, SEE=5.3 ml/kg/min, mean difference=3.7 ml/kg/minBarnet (b) and Ruiz's equation seem to be the best ones to estimate VO2max in the present sample of adolescents
 Castro-Pinero et al8Boys = 34, girls = 328–171-mile run/walkGas analyser in maximal treadmill testR2=0.52, SEE=3.2 ml/kg/min, % error=32.2, mean difference=10.01 ml/kg/minCureton's equation is not accurate for estimating VO2peak in endurance trained children
 Castro-Piñero et al9Boys = 44, girls = 426–171/2-mile run/walkGas analyser in maximal treadmill testR2=0.44, SEE=4.4 ml/kg/min, % =13.6 ml/kg/min, mean difference= −0.4 ml/kg/min for the new equationSEE=7.1 ml/kg/min, %=50.4 ml/kg/min, mean difference=18.1 ml/kg/min for Fernhall's equationThe new regression equation is valid for estimating VO2peak, and is more accurate than Fernhall's equation in the sample studied
Musculoskeletal fitness (muscular strength)
 España-Romero et al15Boys = 31, girls = 3512–16Jamar, DynEx, TKK dynamometersKnown weightsSystematic bias of −1.92, −1.43 and 0.49 kg for the Jamar, DynEx and TKK dynamometers, respectively (all p<0.05). 95% limits of agreement of 19.2, 3.56 and 1.32 for the Jamar, DynEx and TKK, respectivelyThe TKK seems to be the most appropriated dynamometer to assess handgrip strength in this particular population
 España-Romero et al13Boys = 123, girls = 706–12Handgrip strengthThe equation relating grip span as a function of hand span in boys is formulated as y=x/4+0.44 and in girls as y=0.3x−0.52, where x is the hand span (maximal width between first and fifth fingers) and y is the optimal grip spanThere is an optimal grip span to which the dynamometer should be adjusted when measuring hand grip strength in children
 Castro-Piñero et al17Boys = 49, girls = 456–17Standing broad jump, vertical jump, squat jump, countermovement jump, throw basketball, push-up, isometric strengthThe standing broad jump test was strongly associated with other lower body muscular strength tests (R2=0.829–0.864), as well as with upper body muscular strength tests (R2=0.694–0.851)The standing broad jump test might be therefore considered a general index of muscular fitness in youth. This test is practical, time-efficient, and low in cost and equipment requirements
 Artero et al16Boys = 74, girls = 5212–16Handgrip strength, bent and extended arm hang, standing broad jump, squat jump, countermovement jump, Abalakov jumpGymnex Iso-2 dynamometer with angular velocities of 90°/s for upper body and 60°/s for lower bodyThe handgrip strength and the standing long jump tests showed the highest associations with the isokinetic parameters (0.61≤r≤0.87; 0.39≤R2≤0.76)The handgrip strength and the standing broad jump tests seem to be the most valid tests when compared to isokinetic strength in youth
Musculoskeletal fitness (flexibility)
 Castro-Piñero et al18Boys = 45, girls = 426–17Sit and reach, modified sit and reachGoniometerThe SRT was associated with hamstring flexibility in both children (β=1.089, R2=0.281, p=0.001) and adolescents (β=0.690, R2=0.333, p=0.004). The MSRT was also associated with hamstring flexibility in both children (β=1.296, R2=0.298, p<0.001) and adolescents (β=0.588, R2=0.243, p=0.027)The validity of the SRT and the MSRT for estimating hamstring flexibility is weak, and the MSRT is not a more valid method than the SRT in children and adolescents
 Chillón et al19Boys = 87, girls = 5112–16Back-saver sit and reach, sit and reachAngular kinematic analysisThe hip angle independently explained a 42% (p<0.001) of the variance, the lumbar angle additionally explained a 30% (p<0.001), and the dorsal angle added a 4% (p<0.001) of the variance in the BSSRT. The inter-method mean difference between BSSRT and SRT measures (BSSRT − SRT) was 0.41 cm (p=0.21)Hip flexibility is the main determinant of the BSSRT score in adolescents, followed by lumbar flexibility. The BSSRT can be therefore considered an appropriate and valid test for assessing hip and low-back flexibility at these ages. The BSSRT and SRT provide rather comparable values
  • ANN, artificial neural network; BSSRT, Back-saver sit and reach test; MSE, mean square error; MSRT, modified sit and reach; RMSE, root mean square error; SEE, standard error estimate; SSE, sum of squared errors; SRT, sit and reach test; VO2max, maximum oxygen uptake; VO2peak, peak oxygen uptake; %, percentage error; 20mSRT, 20 m shuttle run test.