Comparação entre métodos para análise da reprodutibilidade da arquitetura muscular do músculo vasto lateral por ultrassonografia

Abstract

The objective of the present study was to compare two protocols for the reproducibility of muscle architecture measurements [cross-sectional area (CSA), muscle thickness (MT), fascicle length (FL), and pennation angle (PA)] of the vastus lateralis (VL) using ultrasonography (US): (1) a single image acquisition followed by two quantifications with a 72-hour interval (protocol 1X); (2) two image acquisitions followed by two quantifications, both separated by 72 hours (protocol 2X). A sample of 54 individuals had their right and left legs assessed by US, totaling 104 experimental units. Quantification data for muscle architecture were obtained, along with their respective typical error (TE) and coefficient of variation (CV) values, and bias plots analyzed using the Bland-Altman method. The results demonstrated lower TE and CV values for protocol 1X (CSA: TE = 0.510 cm² and CV = 2.313%; MT: TE = 0.120 cm and CV = 6.225%; FL: TE = 0.205 cm and CV = 3.000%; PA: TE = 0.550° and CV = 3.684%) compared to protocol 2X (CSA: TE = 0.920 cm² and CV = 4.151%; MT: TE = 0.120 cm and CV = 6.225%; FL: TE = 0.480 cm and CV = 6.879%; PA: TE = 1.148° and CV = 7.823%). Data analyzed using the t-test and Wilcoxon test revealed a significant difference between the mean differences of 1X and 2X for FL (p < 0.001) and PA (p < 0.026), while no significant differences were observed for CSA (p < 0.895) and MT (p < 0.637). Levene's test indicated a significant difference (p < 0.001) between the variances of the mean differences in all muscle architecture parameters. In conclusion, our results suggest that the variability of vastus lateralis muscle architecture measurements is influenced by the reproducibility protocol adopted. The inclusion of two independent image acquisitions resulted in greater variability between measurements. These findings reinforce the importance of considering sources of error associated with the evaluator and biological variability when interpreting changes in muscle architecture.