Age-related changes in human skeletal muscle microstructure and architecture assessed by diffusion-tensor magnetic resonance imaging and their association with muscle strength

Abstract

Diffusion-tensor magnetic resonance imaging (DT-MRI) offers objective measures of muscle characteristics, providing insights into age-related changes. We used DT-MRI to probe skeletal muscle microstructure and architecture in a large healthy-aging cohort, with the aim of characterizing age-related differences and comparing these to muscle strength. We recruited 94 participants (43 female; median age = 56, range = 22-89 years) and measured microstructure parameters-fractional anisotropy (FA) and mean diffusivity (MD)-in 12 thigh muscles, and architecture parameters-pennation angle, fascicle length, fiber curvature, and physiological cross-sectional area (PCSA)-in the rectus femoris (RF) and biceps femoris longus (BFL). Knee extension and flexion torques were also measured for comparison to architecture measures. FA and MD were associated with age (β = 0.33, p = 0.001, R² = 0.10; and β = -0.36, p < 0.001, R² = 0.12), and FA was negatively associated with Type I fiber proportions from the literature (β = -0.70, p = 0.024, and R² = 0.43). Pennation angle, fiber curvature, fascicle length, and PCSA were associated with age in the RF (β = -0.22, 0.26, -0.23, and -0.31, respectively; p < 0.05), while in the BFL only curvature and fascicle length were associated with age (β = 0.36, and -0.40, respectively; p < 0.001). In the RF, pennation angle and PCSA were associated with strength (β = 0.29, and 0.46, respectively; p < 0.01); in the BFL, only PCSA was associated with strength (β = 0.43; p < 0.001). Our results show skeletal muscle architectural changes with aging and intermuscular differences in the microstructure. DT-MRI may prove useful for elucidating muscle changes in the early stages of sarcopenia and monitoring interventions aimed at preventing age-associated microstructural changes in muscle that lead to functional impairment.

Keywords: aging; diffusion tensor imaging; muscle strength; sarcopenia; skeletal muscle fibers; thigh.

FIGURE 1

Representative tractography results (top row), and whole‐thigh fractional anisotropy (FA, middle row) and mean diffusivity (MD, bottom row) measures versus age. Dashed lines link the tractography examples to their corresponding FA and MD data points, which are highlighted by colored circles on the scatter plots. Tractography results are presented as anterior coronal views of the thigh showing, in clockwise order from the top left, muscle regions‐of‐interest (ROIs), tracts colored by their average FA and MD, respectively, and fiber tract directionality (red, left–right; green, anterior–posterior; blue, superior–inferior). In the scatter plots below, standardized linear regressions demonstrate associations between FA and age and MD and age, and regression lines, 95% confidence intervals, and regression statistics are shown.

FIGURE 2

Boxplots showing fractional anisotropies (FA, top) and mean diffusivities (MD, bottom) derived from diffusion‐tensor MRI. Median region‐of‐interest (ROI) measures are shown for 12 muscles of the thigh: adductor longus (AL), adductor magnus (AM), biceps femoris long head (BFL), biceps femoris short head (BFS), gracilis (G), rectus femoris (RF), sartorius (S), semimembranosus (SM), semitendinosus (ST), vastus intermedius (VI), vastus lateralis (VL), and vastus medialis (VM). Boxplots represent median values by thick lines, with hinges indicating 25th and 75th percentiles and notches denoting confidence intervals around the median. The underlying raw data are represented by dots. Boxplots are sorted by increasing FA, highlighting a pattern of differences across the thigh where the quadriceps muscles tend to have lower FAs and higher MDs.

FIGURE 3

Density plots showing representative muscle architecture parameters in the rectus femoris (RF, top row) and biceps femoris longus (BFL, bottom row) muscles of the upper leg in two volunteers (both male, age = 30 and 78 years, BMI = 25.1 kg/m² in both cases). Coronal tractography images, inset, highlight the structure of each muscle for both participants. In the RF, pennation angle and fascicle length tend to be lower in the older participant, while curvature is similar between participants. In the BFL, fascicle lengths are again lower in the older participant, but both pennation angle and curvature are higher. In both muscles, pennation angle appears to show a bimodal distribution, perhaps indicating different muscle compartments.

FIGURE 4

Scatter plots showing muscle architecture parameters versus age (top panel) and peak concentric, isokinetic knee extension, and flexion torques, obtained with an angular velocity of 30°/s (bottom panel). Results are shown separately for the rectus femoris (RF) and biceps femoris longus (BFL) muscles, with knee extension torques being reported for the former and flexion torques for the latter. Simple, unadjusted linear regressions show significant associations between curvature, fascicle length, and physiological cross‐sectional area (PCSA) and age, particularly in the RF. Positive associations are seen between pennation angle and torque and PCSA and torque, in both muscles. No clear trends are apparent between fiber curvature or fascicle length and torque. Regression lines and 95% confidence intervals are shown, along with unadjusted model statistics. See Table S3 for multiple regression statistics including sex as a covariate.

FIGURE 5

Schematic showing the diffusion tensor magnetic resonance imaging (DT‐MRI) data‐processing pipeline for this study. Representative axial spin‐echo echo planar imaging (SE‐EPI) DT‐MRI and Dixon data are shown for an 89‐year‐old male subject. The top row shows processing of the DT‐MRI raw data, including denoising and distortion correction, while the bottom row describes region‐of‐interest (ROI) drawing on Dixon images, registration of these ROIs to the DT‐MRI data, and the resulting volumetry. Tractography and fiber architecture determination is performed on the processed DT‐MRI data using muscle volumes from volumetry (dashed line) as seed regions and regions of high tract density (Oudeman, Mazzoli, et al., 2016) or high‐fat fraction as termination regions.