Magnetic resonance imaging (MRI) and dual-energy X-ray absorptiometry (DXA) were used to assess changes in thigh lean mass in septuagenarian men and women during a 5-year longitudinal study. Twenty-four older individuals participated in the study (10 men: 71.6 ± 4.1 years; 14 women: 71.3 ± 3.2 years at baseline). Thigh MRI and whole-body DXA scans were used to estimate changes in thigh lean mass. Both MRI and DXA showed that thigh lean mass was reduced by approximately 5% (P = 0.001) over the 5-year period in both men and women. The percentage loss of muscle mass determined with MRI and DXA showed moderate correlation (R2 = 0.466; P < 0.001). Bland-Altman analysis showed that the average change over 5 years of follow-up measured by DXA was only 0.18% greater than MRI, where the limits of agreement between DXA and MRI were ± 10.4%. Baseline thigh lean mass did not predict the percentage loss of thigh lean mass over the 5-year period (R2 = 0.003; P = 0.397), but a higher baseline body fat percentage was associated with a larger loss of thigh muscle mass in men (R2 = 0.677; P < 0.003) but not in women (R2 = 0.073; P < 0.176). In conclusion, (1) DXA and MRI showed a similar percentage loss of muscle mass over a 5-year period in septuagenarian men and women that (2) was independent of baseline muscle mass, but (3) increased with higher baseline body fat percentage in men.
Keywords: Ageing; Body composition; DXA; MRI; Muscles; Validity.