RSNA 2014 

Abstract Archives of the RSNA, 2014


SSA15-05

Periventricular vs. Deep White Matter Hyperintensity in Normal Aging

Scientific Papers

Presented on November 30, 2014
Presented as part of SSA15: Neuroradiology (The Aging Brain & Neurodegenerative Diseases)

Participants

Soham Banerjee BS, Presenter: Nothing to Disclose
Kevin S. King MD, Abstract Co-Author: Nothing to Disclose
Roderick McColl PhD, Abstract Co-Author: Nothing to Disclose
Anthony R. Whittemore MD, Abstract Co-Author: Nothing to Disclose
Keith Hulsey, Abstract Co-Author: Nothing to Disclose
Ronald M. Peshock MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To map the spatial distribution of white matter hyperintensity (WMH) related to normal aging in the absence of hypertension, diabetes, and hypercholesterolemia.

METHOD AND MATERIALS

MRI brain images were acquired from a population based study. An automated algorithm generated each participant’s WMH distribution registered onto the MNI-152 standard template. A cohort of 827 participants (age range 18-83; 337 males, 490 females) without hypertension, diabetes, or hypercholesterolemia was identified. This cohort was divided into two groups: 702 participants <=55 years and 125 participants >55 years, based on previous studies demonstrating a tenfold increase in WMH volume after age 55. Logistic regression was performed at every voxel determining the voxel WMH probability. The two age groups predicted probabilities were compared, to create a probability ratio map representing the age effect on WMH at each voxel. The distance of each WMH voxel from the lateral ventricles was calculated and compared with its probability ratio using double log linear regression analysis.

RESULTS

Individual analysis was performed for each of the 216694 voxels that comprised the entire population’s WMH distribution. On average, the older age group had predicted probabilities 6.2 times greater than the younger group at each voxel. Double log linear regression revealed a positive association (r2=0.23 with >75000 data points, slope p<0.0001) between probability ratio and distance from the lateral ventricles, suggesting that aging is associated with WMH further away from the ventricles. In comparison, in voxels directly adjacent to the lateral ventricles, age was only associated with a 1.06 times increased probability.

CONCLUSION

Normal aging had the highest impact on the prevalence of deep WMH and lowest impact on periventricular WMH.

CLINICAL RELEVANCE/APPLICATION

The disparate impact of normal aging on periventricular and deep WMH may assist in evaluating the relative contributions aging and treatable cardiovascular pathology make towards WMH in these regions.

Cite This Abstract

Banerjee, S, King, K, McColl, R, Whittemore, A, Hulsey, K, Peshock, R, Periventricular vs. Deep White Matter Hyperintensity in Normal Aging.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14015568.html