Testing the white matter retrogenesis hypothesis of cognitive aging

Abstract

The retrogenesis hypothesis postulates that late-myelinated white matter fibers are most vulnerable to age- and disease-related degeneration, which in turn mediate cognitive decline. While recent evidence supports this hypothesis in the context of Alzheimer's disease, it has not been tested systematically in normal cognitive aging. In the current study, we examined the retrogenesis hypothesis in a group (n = 282) of cognitively normal individuals, ranging in age from 7 to 87 years, from the Brain Resource International Database. Participants were evaluated with a comprehensive neuropsychological battery and were imaged with diffusion tensor imaging. Fractional anisotropy (FA), radial diffusivity (RD), and axial diffusivity (DA), measures of white matter coherence, were computed in 2 prototypical early-myelinated fiber tracts (posterior limb of the internal capsule, cerebral peduncles) and 2 prototypical late-myelinated fiber tracts (superior longitudinal fasciculus, inferior longitudinal fasciculus) chosen to parallel previous studies; mean summary values were also computed for other early- and late-myelinated fiber tracts. We examined age-associated differences in FA, RD, and DA in the developmental trajectory (ages 7-30 years) and degenerative trajectory (ages 31-87 years), and tested whether the measures of white matter coherence mediated age-related cognitive decline in the older group. FA and DA values were greater for early-myelinated fibers than for late-myelinated fibers, and RD values were lower for early-myelinated than late-myelinated fibers. There were age-associated differences in FA, RD, and DA across early- and late-myelinated fiber tracts in the younger group, but the magnitude of differences did not vary as a function of early or late myelinating status. FA and RD in most fiber tracts showed reliable age-associated differences in the older age group, but the magnitudes were greatest for the late-myelinated tract summary measure, inferior longitudinal fasciculus (late fiber tract), and cerebral peduncles (early fiber tract). Finally, FA in the inferior longitudinal fasciculus and cerebral peduncles and RD in the cerebral peduncles mediated age-associated differences in an executive functioning factor. Taken together, the findings highlight the importance of white matter coherence in cognitive aging and provide some, but not complete, support for the white matter retrogenesis hypothesis in normal cognitive aging.

Figure 1

Fractional anisotropy skeleton (in green) representing the major WM tracts for all subjects is overlayed on the MNI standard brain. Sections of the tract skeleton representing the prototypical early- and late-myelinated fiber bundles used in the study are marked: early posterior limb of the internal capsule (PLIC) in blue and cerebral peduncles (CP) in yellow; late superior longitudinal fasciculus (SLF) in red and inferior longitudinal fasciculus (ILF) in pink. Images are in radiological orientation R- right, L- Left.

Figure 2

Mean FA, RD, and DA (upper, middle, and lower panels, respectively) values across four regions-of-interest in the entire cohort. Values are means adjusted for age and sex. Error bars are standard errors. All pairwise comparisons (for all three metrics) are significant (p<0.001).

Figure 2

Mean FA, RD, and DA (upper, middle, and lower panels, respectively) values across four regions-of-interest in the entire cohort. Values are means adjusted for age and sex. Error bars are standard errors. All pairwise comparisons (for all three metrics) are significant (p<0.001).

Figure 2

Mean FA, RD, and DA (upper, middle, and lower panels, respectively) values across four regions-of-interest in the entire cohort. Values are means adjusted for age and sex. Error bars are standard errors. All pairwise comparisons (for all three metrics) are significant (p<0.001).

Figure 3

Scatterplots of association between age and mean in early- (left column) and late-myelinated (right column) fiber tracts. The top 6 plots are for FA; the middle 6 plots are for RD; and the bottom 6 plots are for DA. Separate linear trend lines are superimposed for age-associated differences in the Younger Group (solid black line) and the Older Group (dotted line).

Figure 3

Scatterplots of association between age and mean in early- (left column) and late-myelinated (right column) fiber tracts. The top 6 plots are for FA; the middle 6 plots are for RD; and the bottom 6 plots are for DA. Separate linear trend lines are superimposed for age-associated differences in the Younger Group (solid black line) and the Older Group (dotted line).

Figure 3

Scatterplots of association between age and mean in early- (left column) and late-myelinated (right column) fiber tracts. The top 6 plots are for FA; the middle 6 plots are for RD; and the bottom 6 plots are for DA. Separate linear trend lines are superimposed for age-associated differences in the Younger Group (solid black line) and the Older Group (dotted line).