Agreement Between MRI, Ultrasound, and Histology in Detecting Size Changes of the Vastus Lateralis Following Resistance Training

Abstract

HypothesisWe hypothesize that MRI‐determined vastus lateralis (VL) cross‐sectional area (CSA) will increase following resistance training and that changes in VL size determined by ultrasonography and histology methods will agree with and/or be related to MRI‐determined changes in VL size.MethodsTwelve healthy, untrained, college‐aged males (20±1 y; BMI: 26.9±5.4 kg/m2) had VL assessments via MRI, B‐mode ultrasound (thickness and cross‐sectional area) prior to (PRE) and after 10 weeks of resistance training (POST). Muscle biopsies were obtained at the site of imaging at PRE and POST for mean fiber cross‐sectional area (fCSA) determination. Agreement between MRI mCSA (mCSAMRI) and panoramic‐based ultrasound mCSA (mCSAUS) at PRE and POST as well as change scores in these variables were examined using Bland‐Altman plots for mean bias and 95% limits of agreement (LOA) as well as Lin’s concordance correlation coefficients (CCC). The relationship between fCSA with mCSAMRI, mCSAUS, and B‐mode ultrasound thickness (VLThick), as well as VLThick and mCSAMRI were tested with Pearson’s correlations.ResultsBoth mCSAMRI and VLThick significantly increased following training (28.1±6.1 to 31.1±6.9cm2 (p=0.005) and 2.35±0.35 to 2.63±0.39cm (p=0.001), respectively). mCSAUS and fCSA did not significantly increase with training (26.3±6.4 to 28.3±7.5cm2 (p=0.064), and 4896±901.3 5443.2±1449.3µm2 (p=0.12), respectively). Bland‐Altman analysis revealed a small mean bias in mCSAUS relative to MRI (‐2.29 m2, 95% LOA: ‐8.74, 4.15 cm2). Concordance between mCSAUS and mCSAMRI taken at PRE and POST time points were excellent (CCC=0.83). Bland‐Altman analysis revealed good agreement between change scores in mCSAUS and mCSAMRI (‐1.08cm2, 95% LOA: ‐5.49, 3.33 cm2). There was a moderate concordance between mCSAUS and mCSAMRI change scores (CCC=0.70). A positive, moderate correlation was found between percent change in both VLThick and mCSAMRI as well as VLThick and mCSAUS (r=0.39, p=0.071 and r=0.52, p=0.080, respectively). A negative correlation was found between percent changes in fCSA and mCSAMRI (r= ‐0.569, p=0.054). Finally, correlations were not significant between percent changes in VLThick and fCSA (r=0.053, p=0.871) or fCSA and mCSAUS (r= ‐0.271, p=0.394).ConclusionThese data suggest that, relative to MRI, panoramic ultrasound is a reliable method to track mCSA changes of the VL. However, fCSA changes do not associate with tissue level changes, and this discordance warrants further research.