Our study provides specific QUS parameters that evaluate bone status (AD-SoS and UBPI) by sex and pubertal maturation and their association with physical fitness and body composition in Brazilian children and adolescents. It is the first research using phalange US and physical fitness in Brazilian adolescents.
Girls showed significantly higher QUS values than boys. Studies using DXA have shown that girls have higher BMD values than boys at trabecular sites [8, 21, 22]. The US method measures the distal portion of the proximal phalanges, which are rich in trabecular bone . Regarding the pubertal stage, the pubertal group showed higher values than the prepubertal and intrapubertal groups. Ribeiro et al.  evaluated 1356 black and white students aged 6–11 years and found significant differences between the prepubertal and pubertal groups for AD-SoS and UBPI. These results were expected with respect to age, sex, and pubertal stage. They show that girls who experience early puberty have more bone mass than do same-age boys. Thus, sex hormones are important modulators of bone mass , suggesting an effect of estrogen on trabecular bone [26, 27].
Several studies have investigated the association between the level of physical activity and bone mass in children and adolescents [5, 28–31]. However, few studies have investigated bone mass and its relation to physical fitness. Our study showed that the variables related to muscle strength (fat-free mass, abdominal strength, horizontal jumping) contributed positively to the QUS parameters, whereas BMI and %BF contributed negatively (Table 2). Hence, the level of muscle strength appears to influence the AD-SoS and UBPI, confirming the hypothesis that an increasing level of fitness improves bone quantity [4, 5] and quality, mainly in boys.
This effect on bone parameters related to physical strength variables is in accord with other studies. For instance, Ginty et al.  showed a positive association between the states of total and site-specific bone mineral, cardiorespiratory fitness, and muscle strength in male adolescents. Vicente-Rodriguez et al.  investigated the association between BMD with the physical fitness of 68 boys and girls. They found a direct association with cardiorespiratory fitness, muscular speed, and agility, suggesting that these results could have been due to the association of physical fitness and lean body mass. However, our study did not find a significant association between VO2peak and QUS parameters. In fact, the results suggested that the bone mass differences between males and females could probably be explained by differences in physical fitness and lean mass [34, 35]. The data in the literature regarding the association between physical fitness and bone status are still controversial. Although longitudinal studies have shown an increase in bone formation and reabsorption in adolescents caused by improved cardiorespiratory fitness [33, 36], other studies showed that during adolescence and youth only neuromotor (muscular) fitness [31, 37] was associated with BMD despite finding a significant correlation with cardiorespiratory fitness .
Regarding the variable fat, our study showed a negative correlation of AD-SoS and UBPI for %BF and BMI in boys and %BF for AD-SoS in girls. These results are consistent with those of other studies [38, 39], which also found a negative correlation of fat to bone. The physiological basis to explain the relation between weight, body fat distribution, and bone mass remains unclear, particularly when considering different racial groups . The results showing the adverse effect of increased fat mass on bone mass, along with significant positive associations of lean mass, corroborate the mechanostat theory described by some authors in which the geometry of the bone is adapted primarily by dynamic load imposed by muscle force—not to static loads represented by body weight .
The present study demonstrated that in a general regression analysis, sex and measurements related to growth (age, pubertal maturation, and height) were positive predictors for both AD-SoS and UBPI. In boys, the standing long jump and height were positive predictors for AD-SOS and pubertal maturation and height for UBPI. Conversely, for girls, age was a positive predictor for AD-SoS and UBPI but height only for Ad-SOS. An interesting finding was that for all groups the %BF was a negative predictor for both AD-SoS and UBPI. These findings are in accord with data from previous studies [22, 24, 40–42] and show that the body composition related to fat exerts a negative influence on bone mass in both sexes.
In our study, despite the lean mass having shown a relation with AD-SoS in boys and girls and with UBPI in boys, it did not appear as a predictor in the regression analysis. However, the horizontal jump as a positive predictor in boys showed that muscle strength positively influences AD-SoS, and the %BF has a negative influence on QUS parameters in both sexes.
The present study has some limitations. They include the large number of subjects lost from the original sample, self-evaluation of sexual maturation, no comparison of US data with DXA data, no implementation survey of fractures, and no evaluation of the ethnicity of the subjects.