O1‐01‐05: Brain Changes on Structural MRI from the Eighth to Eleventh Decades of Life

Zixuan Yang,Wei Wen,Jiyang Jiang,John D. Crawford,Simone Reppermund,Charlene Levitan,Melissa Jane Slavin,Nicole Kochan,Robyn Richmond,Henry Brodaty,Julian Trollor,Perminder S. Sachdev

DOI: https://doi.org/10.1016/j.jalz.2015.07.031

2015-01-01

Abstract:Although brain aging in the younger old (YO) has been well studied, relatively less is known about the brains of people ≥ 90 years – the oldest old (OO). We aimed to use high resolution magnetic resonance imaging (MRI) to cross-sectionally examine brain morphological changes in the non-demented OO population. 317 community dwelling participants (71-103 years, 277 non-demented and 40 demented) from the Sydney Memory and Ageing Study and the Sydney Centenarian Study comprised the sample. Dementia was diagnosed by consensus according to DSM-IV criteria. Age-associated changes on MRI-derived measurements, of cortical and subcortical regions as well as accumulation of white matter hyperintensities (WMHs), were examined by general linear model adjusted for confounding factors. The non-demented aging pattern was further compared to dementia-related changes on structural MRI in YO (70-89) and OO (90+) respectively. In the non-demented sample, total grey matter (GM) volume declined at a rate of 5.06% per decade in a linear pattern. The mean age-related cortical thinning was 2.97% per decade, with the greatest decline in medial temporal, parietal and occipital cortices. The volumes of subcortical structures were estimated to decline at rates of 3.34 – 8.98% per decade, with most “shrinkage” in the hippocampus, putamen and pallidum. While the decline in white matter (WM) decelerated with advanced age, the accumulation of WMHs increased with age exponentially. In those with dementia, relative to their non-demented peers, the YO had significantly thinner cortices of the temporo-parietal area; whereas the OO had significantly greater loss in the anterior cortex (anterior cingulate, lateral orbitofrontal, pars triangularis and insula) and in temporal lobe. Volume of WMHs was distinctive of dementia status in YO but not in OO. The emphasis on decline in medial temporal and posterior brain regions in the non-demented OO is different from the documented “fronto-temporal” pattern in young elders and the temporo-parietal pattern in typical Alzheimer's disease. Absence of dementia in the OO correlates with relatively preserved anterior cortex and the temporal lobe, which could inform mechanisms of cognitive maintenance in late life.