Linear and nonlinear relationship between chronological age and maturity status with handgrip strength in children...

Changes in hand grip strength in relation to chronological age and maturity status could improve the predictive power through nonlinear models in schoolchildren.

To determine whether nonlinear models provide more accurate and higher predictions of hand grip strength (HGS) in children and adolescents, taking into account chronological age and maturational status.

A descriptive cross-sectional (correlational) study was designed in schoolchildren aged 6-16 years. The sample selection was non-probabilistic (accidental). A total of 1,048 schoolchildren from 03 public schools were selected (562 males and 482 females). Weight, height and body mass index (BMI) were assessed using standardized anthropometric techniques. The maturational state was estimated by the age at which the maximum velocity of stature (APHV) is reached, using Moore's technique. Handgrip strength (HGS) was evaluated for both hands using a digital dynamometer.

The nonlinear (cubic) relationships showed better explanatory power and fit than the classic linear model in both sexes and hands (HGS). In male schoolchildren, the coefficient of determination (R²) of the nonlinear model was 2% to 3% higher than that of the linear model. Meanwhile, in females the R² was higher from 1%-4%, both for chronological age and maturational state. The cubic model showed a better fit of the SEE (in males it ranged between 3.44 and 5.32 and in females between 3.36 and 3.57), large effect sizes (f² > 0.35), evidencing a greater precision and ability to capture the variability of the HGS in both hands in relation to the classical linear model (in males 4.81-5.55 and in females 3.37-5.08).

The results of the study have shown that the cubic model provides a better explanation and fit of the relationship between chronological age and maturity status with HGS than the classical linear model. It was also determined that fluctuations in HGS are more accurately described by maturity status than by chronological age. These results suggest including nonlinear models and controlling for maturity status. This facilitates the design of interventions according to the different stages of maturational development.