To capture patterns of normal age-associated atrophy, we previously used a multivariate statistical approach applied to voxel based morphometry that identified age-associated gray and white matter covariance networks (Brickman et al. [2007]: Neurobiol Aging 28:284-295). The current study sought to examine the stability of these patterns by forward applying the identified networks to an independent sample of neurologically healthy younger and older adults. Forty-two younger and 35 older adults were imaged with standard high-resolution structural magnetic resonance imaging. Individual images were spatially normalized and segmented into gray and white matter. Covariance patterns that were previously identified with scaled subprofile model analyses were prospectively applied to the current sample to identify to what degree the age-associated patterns were manifested. Older individuals were also assessed with a modified version of the Mini Mental State Examination (mMMSE). Gray matter covariance pattern expression discriminated between younger and older participants with high optimal sensitivity (100%) and specificity (90.5%). While the two groups differed in the degree of white matter pattern expression (t (75) = 5.26, P < 0.001), classification based on white matter expression was relatively low (sensitivity = 80% and specificity = 61.9%). Among older adults, chronological age was significantly associated with increased gray matter pattern expression (r (32) = 0.591, P < 0.001) but not with performance on the mMMSE (r (31) = -0.314, P = 0.085). However, gray matter pattern expression was significantly associated with performance on the mMMSE (r (31) = -0.405, P = 0.024). The findings suggest that the previously derived age-associated covariance pattern for gray matter is reliable and may provide information that is more functionally meaningful than chronological age.
To capture patterns of normal age‐associated atrophy, we previously used a multivariate statistical approach applied to voxel based morphometry that identified age‐associated gray and white matter covariance networks (Brickman et al. [2007]: Neurobiol Aging 28:284–295). The current study sought to examine the stability of these patterns by forward applying the identified networks to an independent sample of neurologically healthy younger and older adults. Forty‐two younger and 35 older adults were imaged with standard high‐resolution structural magnetic resonance imaging. Individual images were spatially normalized and segmented into gray and white matter. Covariance patterns that were previously identified with scaled subprofile model analyses were prospectively applied to the current sample to identify to what degree the age‐associated patterns were manifested. Older individuals were also assessed with a modified version of the Mini Mental State Examination (mMMSE). Gray matter covariance pattern expression discriminated between younger and older participants with high optimal sensitivity (100%) and specificity (90.5%). While the two groups differed in the degree of white matter pattern expression (t (75) = 5.26, P < 0.001), classification based on white matter expression was relatively low (sensitivity = 80% and specificity = 61.9%). Among older adults, chronological age was significantly associated with increased gray matter pattern expression (r (32) = 0.591, P < 0.001) but not with performance on the mMMSE (r (31) = −0.314, P = 0.085). However, gray matter pattern expression was significantly associated with performance on the mMMSE (r (31) = −0.405, P = 0.024). The findings suggest that the previously derived age‐associated covariance pattern for gray matter is reliable and may provide information that is more functionally meaningful than chronological age. Hum Brain Mapp 2008. © 2007 Wiley‐Liss, Inc.