Are anthropometric characteristics powerful markers to predict the Cooper Run Test? Actual Caucasian data

Introduction

Cardiorespiratory fitness (CRF) is the overall capacity of the cardiovascular and respiratory systems that provide a measure of the body’s ability to deliver and use oxygen as energy source to support muscle activity during physical activity (PA) and exercise (Tomkinson & Olds, 2008). CRF is considered a powerful marker of cardiovascular health especially in youth to avoid long-term poor medical outcomes and to enhance disease prevention (Ortega et al., 2008). Regarding this, the accuracy of CRF measurement is considered relevant for researchers and sport specialists to assess the level of physical fitness and to predict health status. Usually, CRF is assessed through the measurement of or estimation of maximum oxygen uptake (VO₂ max) during a maximal graded exercise test. Even if the accuracy of this procedure is well recognized, this method is restricted to well-equipped laboratories due to the expensive, complicated and difficult experimental protocols (Fox, 1973).

On the other hand, several studies demonstrated the validity and reliability of specific field tests to determine CRF that provide a single measure and estimate the physiological responses required to perform prolonged exercise (Ruiz et al., 2009; Ortega et al., 2015). In fact, field-based fitness tests are easy to administer, involve minimal equipment, low cost and a larger number of participants can be evaluated in a relatively short period of time (España-Romero et al., 2010).

Nevertheless, there are several indirect field tests for the prediction and the evaluation of CRF (Castro-Piñero et al., 2010), Cooper’s 12-min Run Test (CRT) is considered appropriate from childhood to maturity (Ainsworth et al., 2000; Ayán et al., 2015; Penry, Wilcox & Yun, 2011; Giuriato et al., 2020) and for this reason is widely used. In particular, CRT discovers similar results to Multistage Run Tests (p > 0.05), as reported by Penry, Wilcox & Yun (2011) and a value of bias correction factor (Cb) very high about accuracy of distance between test and retest (Cb = 0.994) and an effect size coefficient (ES) low as 0.059, which indicates the good repeatability of the test (Alvero-Cruz, Giráldez Garcia & Carnero, 2016).

Previously, some authors reported that 12-min run distances ranging from approximately 1,800 m to 2,500 m in adolescents (11–14 years old) and decreased in young that are overweight and obese (distance not exceed 1,500 m; Penry, Wilcox & Yun, 2011; Weisgerber et al., 2009) while Lovecchio et al. (2013) found an opposite trend between boys and girls. Boys showed constant performance from 11 to 13 years old (CRT distance average 2,026 m) where height and weight slightly increased with a peak at 14 years old where height and weight increased 5% and 11% respectively. In contrast, girls showed a higher performance from 11 to 12 years old (CRT distance average 1,850 m) where height and weight slightly increased and a lower performance from 13 to 14 years old (CRT distance average 1,700 m) where height and mass increased 5% and 11% respectively. More in depth, Giuriato et al. (2020), through an allometric analysis of the same test within the same aged samples suggested that height positively influenced the CRF performance.

Also Kemper (1985) did not find improvements in CRT among 12 to 14-year-old girls (Kemper, 1985) and this phenomenon can be explained by the delay in growth of about two years between girls and boys (Peak of Height Velocity in Girls 11/13 years vs Peak of Height Velocity in Boys 13/15 years) (Malina et al., 2021; Giuriato et al., 2021). Furthermore, a study conducted by Kirchengast (2010) suggests that boys generally have a higher amount of lean body mass compared to girls at prepuberal age (p < 0.001), which allows boys to reach better cardiovascular fitness levels than girls (James et al., 2019).

Usually, field-based evaluations are performed by trainers or PE teachers with a common opinion that taller children are faster and lower children are stronger (Lovecchio & Zago, 2019). However, anthropometric measures have often been used to normalize performance (Rowland, Vanderburgh & Cunningham, 1997; Welsman & Armstrong, 2000) since recent authoritative reports have disclaimed individual normalizations based on weight or height (Armstrong & Welsman, 2019a) suggesting that performance also depends on growth.

Indeed to compare and evaluate CRT performance, we have to consider the divergence among ethnicity and mostly the anthropometrics characteristics such as height and mass that are specific human growth factors that affect the performance (Lovecchio & Zago, 2019; Lloyd et al., 2014): unfortunately, to the best of our knowledge, in Italy there are no findings of CRT among healthy adolescents considering the anthropometric characteristics that could be relevant also for other Caucasian countries (Lovecchio et al., 2015, 2019, 2020).

As several studies report, considering activities in the gym or on the track, anthropometric characteristics are strongly recommended to guide training, predict performance, and identify talents (Lombardi & Piacentini, 2019; Buchheit & Mendez-Villanueva, 2014; Norton & Olds, 1996). For these reasons, the aim of this study was to evaluate possible correlations between biometric measurements and athletic performance in middle school students and the magnitude of the influence of anthropometric characteristics on performance.

Materials and Methods

Results

The average age of the participants was 12.38 ± 1.03 years and most participants were boys (52%); the age most represented was 12 years (2,811 subjects) and the least was 14 years (1,612 subjects) as represented in Table 2 (Yıldırım Usta et al., 2019).

Distribution of the attributes were analyzed, keeping track of the sex covariate by representing data with different colors (Fig. 1); while the Shapiro-Wilk test was run to further validate normality, providing very small p-values (p < 0.001 both for boys and girls), the effect size (0.031 for boys and 0.022 for girls) was small enough that the correction on this parameter allows a practical assumption of normality for the distributions. Analogous results were achieved via Lilliefors test, and graphical results are attached despite of the large and comparable size of the cohort for both boys and girls (Fig. 2), CRT results can be well fitted with Gaussian distributions, with boys showing an average higher level of performance than girls (i.e., 2,096.36 m vs 1,807.37 m); on the other hand, a smaller SD (i.e., 371.12 m vs 282.00 m) can be observed for girls, as represented in Fig. 3.

Parameters were ranked for information provided through univariate regression (Table 3, girls first column, boys second column) or R-Relief-F (third column) when including sex as covariate. Notably, as shown in Table 3, VO₂ for girls is more informative than BMI, underlying a possible causality related to different level of metabolic/muscular performance rather than structural feature like BMI. Linear regressions were performed using the four more representative covariates: BMI (Fig. 4); VO₂ Peak (Fig. 5); mass (Fig. 6) and age at peak height velocity (Fig. 7) despite the first is already dependent on mass and height, while the last on height. As shown in Figs. 4–7, all the R2 values were below 0.5, highlighting the absence of correlation or any sort of dependence.

Homoskedasticity was verified graphically, by plotting normalized residuals against CRT results (Fig. 8). Finally, the independence of the observations was implicitly verified by the experimental design.

Discussion

The purpose of this study was to evaluate, in a sample of middle school children, the correlation between cardiorespiratory fitness and anthropometrics characteristics, in accordance with biometric measurements and performance, as a predictor of mean results.

As represented in Fig. 2, CRT results for both sexes are fitted in the Gaussian Curve, it indicates that our sample was composed of athletic average-prepared young people, in fact we used strict inclusion criteria. We confirmed a mean better CRT result in boys vs girls, as observed in other studies (Freitas et al., 2002; Lovecchio et al., 2013); however, we discovered a smaller SD in girls than boys, suggesting a more homogeneous aerobic status for girls that can be explained by the fact that they do less physical activity on average than boys in Italy, as described in the WHO Regional Office for Europe’s report (Inchley et al., 2020). It is also very interesting that the Gaussian Distribution of CRT results suggests that aerobic capacity in adolescents (11–14 years old) is set in the middle, showing a normal distribution of physical activity status in Italian students of that age.

According to BMI, that is considered a simple resume of anthropometric characteristics (Romero-Corral et al., 2008) also in children (Ceccarelli et al., 2020), a homoscedastic distribution both for boys and girls is visually evident (Fig. 3); thus, showing difficult reasoning of inference according to height and mass, we can state the same result for the variable mass alone. In addition, analyzing linear correlation coefficients, we can affirm there is a weak negative correlation for both BMI and mass variables with respect to the CRT results. We can explain this finding with the possible body composition differences when analyzing mass (and other related anthropometric parameters) in weight bearing performances, as suggested by other studies (Artero et al., 2010).

Figure 4 showed a visual homoscedastic distribution for VO₂ peak (calculated through an equation using height and mass), with a positive, but low linear correlation coefficient, leading to an accuracy reduction of indirect formulas to predict VO₂ peak in average adolescents. The only visual heteroscedastic distribution was observed in regression between distance in CRT and age at peak of height velocity (PHV), in addiction linear correlation coefficient was near 0 (0.01 for boys and −0.03 for girls) with a countertrend between sexes, according to the theory in which testosterone production determines improvements of performance (Handelsman, Hirschberg & Bermon, 2018), so we can affirm that height and mass are just indicators subjected to uncontrollable variations. Our study states further, for an unbiased population, the same that a recent paper finds that age at PHV is not a significant predictor in endurance performance prediction of basketball girls adolescents (Gryko et al., 2022).

Based on a non-linear analysis, Armstrong & Welsman (2019a) suggested that a sex-specific morphological covariate is more appropriate to investigate performance avoiding bias due to the maturity effects. Suggesting that the cross-sectional analyses in young adolescents show that there is no significant gender difference in maximal stroke volume once the fat-free mass has been correctly checked through multilevel modeling (Armstrong & Welsman, 2019b).

Giuriato et al. (2020) proposed a multiplicative allometric model to investigate CRF with CRT as dependent variable, showing that CRT performance during growth may not be predictable with only one anthropometry. Further, the results found by Giuriato et al. (2020) showed in both sexes an improvement in CRT distance. This trend using the scaling method, associated with anthropometrics (height, body mass), suggests that slim subjects correspond to the optimal height-to-body mass for CRF, highlighting that body mass (probably fat mass) influenced negatively the CRF performance, on the contrary height influenced the CRF performance positively. These results are in line with Nevill et al. (1998), who suggested that proportional oxygen absorption peak and oxygen supply/use is facilitated by an increase in fat-free mass, thus promoting a lighter weight and penalizing heavier young people. Indeed, normalization only per body mass or height is conflicting, because growth does not follow a linear trend. In fact, Armstrong & Welsman (2019a, 2019b) suggested that the VO₂ peak (deriving by the meter run) increases with simultaneous changes in morphological covariates specific to gender, age, and maturity, with the timing of these individual-specific changes. Further, Lovecchio et al. (2023), using BMI-adjusted analyses, found a small to moderate decline in CRF between 1984 and 2010, proposing that these trends were probably affected by increases in sedentary behaviors and subsequently declines in vigorous physical activity levels.

Anthropometric variations during growth are non-linear and therefore confusing, indeed Giuriato et al. (2020) found a different correlation between body mass and CRF in adolescent boys and girls. Even Scandinavian data, collected by Santtila et al. (2006), did not use weight gain as a reason for explaining the decline in CRF, despite the subjects being 19-year-old boys, and therefore out of the tumultuous period of growth. Thus, during growth body mass accentuates its instability making its use only unreliable.

Conclusions

In conclusion, CRT results showed a normal distribution for boys and girls independently of their anthropometric measures and a very weak or absent correlation for all the parameters and leading to the assertion that CRT results cannot be predicted from anthropometric data.

Supplemental Information