Mathematical model for predicting handgrip strength in Quilombola children and adolescents: a cross-sectional study

Abstract

Objective: To determine which variables have the ability to predict hand grip strength in Quilombola children and adolescents. Methods: We calculated the sample seeking an R2 between 0.1 and 0.2 for a single dependent variable (handgrip strength), with 6 predictor variables (age, body mass, stature, BMI, fat, and lean mass), alpha of 0.05 and beta of 0.80; and included children and adolescents between the ages of 6 and 17 (n=82). We measured handgrip strength using the Jamar dynamometer and built a model and evaluated the association between the predictor variables (i.e., independent, x-axis) and the outcome variable (i.e., dependent, y-axis [dynamometry]) by analysis of variance of mathematically adjusted models (F-value >60, p <0.05). Results: We noted an increasing gain in strength over the years, although between the ages of 11 and 12 and between 13 and 14, there was an apparent loss of strength on the part of Quilombola adolescents, passing from 18.75 to 16.12 and from 23.5 to 19.83, respectively. We observed that the variables age, stature, and lean mass contributed significantly (p <0.05, β coefficient ranging from 3.050 to 3.844) to the performance of the built model (F [7.74] = 62.16, p <0.001; R2 = 0.84). Conclusion: Age, stature, and lean mass significantly contribute to the performance of the built model. Namely, 84% of the variation in the mean handgrip strength may be explained by the independent variables. Therefore, the predicted handgrip strength, in kg, corresponds to: -29.530 + 1.103 + 0.196 + 0.011 × (age [years] + stature [cm] + lean mass [kg]).