IC-P2-080: Brain structure change is associated with clinical decline of elderly controls: A longitudinal study using the ADNI dataset

Abstract

Background: Automated segmentation and measurement of subregional brain volumes from MRI data has allowed high-throughput quantification of regional brain atrophy. Such measures predict subsequent clinical decline in patients with mild cognitive impairment (MCI). Quantitative analysis of brain structure change over time may be even more sensitive to the earliest effects of neurodegenerative disease, including that which may be present in elderly controls who have passed clinical screening. Methods: We examined whether longitudinal measures of brain change in healthy elderly controls is associated with decline on clinical and neuropsychological measures. We analyzed longitudinal MRI data from 70 control subjects of the Alzheimer’s Disease Neuroimaging Initiative (ADNI) who had 6and 12-month follow-up imaging at the time of the analysis. 6and 12-month subregional change was quantified with automated anatomical segmentation using FreeSurfer software and non-linear image registration. Subjects were sorted based upon the degree of change in the temporal horn of the lateral ventricle and dichotomized at the median, creating a group with a high-degree of anatomical change and a group with a low-degree of anatomical change. In addition, atrophy was analyzed as a continuous measure, plotting change in the temporal horn against 12-month change in MMSE. Results: Healthy elderly control subjects showing a high degree of anatomical change in bilateral temporal horn of the lateral ventricle declined more on MMSE over 12 months. This effect was significant both for 6-month change and for 12-month change in temporal horn volumes (each p .01, two-tailed). Average age of the high atrophy group (76.6 4.9) did not differ from the age of the low atrophy group (75.9 5.1). Correlation between change in temporal horn volume and MMSE decline was significant at p .02 for 6-month change and at p .005 for 12-month change. Conclusions: Change in temporal horn volume is associated with clinical decline of healthy elderly controls. Such changes may reflect processes of normal aging, but the rate of atrophy may rather suggest the presence of subclinical neurodegenerative disease. No patients converted to MCI or Alzheimer’s disease. Continued follow-up of the high atrophy cohort will determine whether such measures predict later clinical conversion.