Peripheral quantitative computed tomography is a valid imaging technique for tracking changes in skeletal muscle cross-sectional area

Abstract

Peripheral quantitative computed tomography (pQCT) has recently expanded to quantifying skeletal muscle, however its validity to determine muscle cross-sectional area (mCSA) compared to magnetic resonance imaging (MRI) is unknown. Eleven male participants (age: 22 ± 3 y) underwent pQCT and MRI dual-leg mid-thigh imaging before (PRE) and after (POST) 6 weeks of resistance training for quantification of mid-thigh mCSA and change in mCSA. mCSA agreement at both time points and absolute change in mCSA across time was assessed using Bland-Altman plots for mean bias and 95% limits of agreement (LOA), as well as Lin's concordance correlation coefficients (CCC). Both pQCT and MRI mCSA increased following 6 weeks of resistance training (∆mCSA_pQCT: 6.7 ± 5.4 cm², p < 0.001; ∆mCSA_MRI: 6.0 ± 6.4 cm², p < 0.001). Importantly, the change in mCSA was not different between methods (p = 0.39). Bland-Altman analysis revealed a small mean bias (1.10 cm², LOA: −6.09, 8.29 cm²) where pQCT tended to overestimate mCSA relative to MRI when comparing images at a single time point. Concordance between pQCT and MRI mCSA at PRE and POST was excellent yielding a CCC of 0.982. For detecting changes in mCSA, Bland-Altman analysis revealed excellent agreement between pQCT and MRI (mean bias: −0.73 cm², LOA: −8.37, 6.91 cm²). Finally, there was excellent concordance between pQCT and MRI mCSA change scores (CCC = 0.779). Relative to MRI, pQCT imaging is a valid technique for measuring both mid-thigh mCSA at a single time point and mCSA changes following a resistance training intervention.