White matter organization in cervical spinal cord relates differently to age and control of grip force in healthy subjects

Abstract

Diffusion tensor imaging (DTI) can be used to elucidate relations between CNS structure and function. We hypothesized that the degree of spinal white matter organization relates to the accuracy of control of grip force. Healthy subjects of different age were studied using DTI and visuomotor tracking of precision grip force. The latter is a prime component of manual dexterity. A regional analysis of spinal white matter [fractional anisotropy (FA)] across multiple cervical levels (C2-C3, C4-C5, and C6-C7) and in different regions of interest (left and right lateral or medial spinal cord) was performed. FA was highest at the C2-C3 level, higher on the right than the left side, and higher in the lateral than in the medial spinal cord (p < 0.001). FA of whole cervical spinal cord (C2-C7) was lower in subjects with high tracking error (r = -0.56, p = 0.004) and decreased with age (r = -0.63, p = 0.001). A multiple regression analysis revealed an independent contribution of each predictor (semipartial correlations: age, r = -0.55, p < 0.001; tracking error, r = -0.49, p = 0.003). The closest relation between FA and tracking error was found at the C6-C7 level in the lateral spinal cord, in which the corticospinal tract innervates spinal circuitry controlling hand and digit muscles. FA of the medial spinal cord correlated consistently with age across all cervical levels, whereas FA of the lateral spinal cord did not. The results suggest (1) a functionally relevant specialization of lateral spinal cord white matter and (2) an increased sensitivity to age-related decline in medial spinal cord white matter in healthy subjects.

Figure 1.

Cervical spinal cord white matter in relation to age and force control. A, Left, Medial (M) and lateral (L) example ROIs on axial T2-weighted MR image at C3 level. Right, Example of axial FA maps at C3 in a young (22 years) and old (64 years) subject. Although FA differences between subjects are often difficult to detect visually (Beaulieu, 2009), slightly reduced FA can be visualized in medial spinal cord in the older subject. B, Left to right, Sagittal T2-weighted image, ADC map, and FA map in the young and old subject (same subjects as in A, left and right, respectively). The older subject did not have any signs of stenosis of the spinal cord. C, Left, Example ROIs at the C6 level. Right, Example of axial FA maps at C6 in one subject with low (64 years) and one subject with high (51 years) force-tracking error. Tendency for reduced FA can be visualized in lateral spinal cord in the subject with high tracking error. D, Single-trial force-tracking examples: target trace is shown in gray and actual force in black. Arrow indicates region of increased error during the ramp to the 3 N target force level in the subject with higher mean error across trials (same subjects as in C, left and right, respectively).

Figure 2.

Spinal cervical FA. A, FA was significantly higher in the lateral SC than in medial SC ROIs (mean ± SE) (F_(1,23) = 25.2, p < 0.001, η_p² = 0.52). B, FA was significantly higher on the right side than on the left (F_(1,23) = 13.9, p = 0.001, η_p² = 0.38). C, FA decreased significantly from higher to lower cervical levels (F_(2,46) = 29.7, p < 0.001, η_p² = 0.56).

Figure 3.

Spinal cervical FA as a function of age. A, Pearson's correlation between FA of the medial cervical spinal cord and age (n = 24, r = −0.68, p = 0.003). B, Pearson's correlation between FA of the lateral cervical spinal cord and age (n = 24, r = −0.37, p = 0.07).

Figure 4.

Spinal cervical FA as a function of tracking error. A, Pearson's correlation between FA of the medial cervical spinal cord and tracking error during the ramp up to 3 N (n = 24, r = −0.39, p = 0.06). B, Pearson's correlation between FA of the lateral cervical spinal cord and tracking error during the ramp up to 3 N (n = 24, r = −0.63, p = 0.0009).

Figure 5.

FA as a function of age, tracking error, and spinal cervical level. A–C, FA of the medial spinal cord. Significant Pearson's correlations for age but not for tracking error at each cervical level. D–F, FA of the lateral spinal cord. y-axis shows mean normalized values for age and RampE 3N (i.e., values of 1 indicate mean). The Pearson's correlations between the lateral spinal cord FA and the tracking error increased successively from the C2–C3 (nonsignificant), to C3–C4 (weakly significant), to C6–C7 (highly significant). In contrast, no significant correlation was found with age.