Evidence of subtle gray-matter pathologic changes in healthy elderly individuals with nonspecific white-matter hyperintensities

Objective: To investigate whether additional "occult" tissue changes can be detected in the normal-appearing white matter and gray matter of otherwise normal elderly individuals with nonspecific white-matter hyperintensities on conventional magnetic resonance images of the brain. Methods: Conventional and magnetization transfer magnetic resonance images were obtained from 12 otherwise normal elderly subjects with white-matter hyperintensities and 11 age- and sex-matched normal individuals. After automatic tissue segmentation, image coregistration, and masking of T2-visible lesions, we obtained magnetization transfer ratio histograms of the normal-appearing white matter and gray matter. For each histogram, the average magnetization transfer ratio, the peak height, and the peak position were measured. We also calculated the percentages of gray-matter and white-matter volumes normalized over the total volume of the intracranial content and the total normalized brain volumes. Results: Average magnetization transfer ratio (P = .03) and mean peak position (P = .01) of the gray-matter histograms from elderly individuals with white-matter hyperintensities were significantly lower than the corresponding quantities from those without white-matter hyperintensities. The normalized percentages of gray and white matter and normalized brain volume did not differ between the 2 groups. The average gray-matter magnetization transfer ratio was correlated with the average lesion magnetization transfer ratio (r = 0.68; P < .01). Conclusions: This study shows that brain abnormalities in otherwise normal elderly subjects with nonspecific white-matter hyperintensities extend beyond the macroscopic white-matter lesions visualized on conventional magnetic resonance images.