P4‐127: WHITE MATTER HYPERINTENSITY VOLUME IS ASSOCIATED WITH DISPROPORTIONATE PROGRESSIVE HIPPOCAMPAL ATROPHY IN CONTROLS

Atrophy rates derived from serial MRIs are used to assess disease progression in Alzheimer's disease (AD). Understanding factors which influence atrophy rates is important in understanding pathological processes. We previously demonstrated a relationship between white matter hyperintensity (WMH) and whole-brain atrophy rates in controls. In this study we extended this work by investigating: relationships between WMH and hippocampal atrophy rates - a specific marker for AD; whether associations between WMH and hippocampal atrophy rate and WMH and whole-brain rate were independent of one another; whether any relationships were independent of cerebrospinal fluid (CSF) Aβ and tau. MRIs from Alzheimer's Disease Neuroimaging Initiative 1 visits up to 36 months inclusive were used (154 AD, 345 mild cognitive impairment, MCI and 198 controls). A multi-level mixed-effect linear regression model was used to analyse the effect of log 2 WMH (predictor) on hippocampal and brain atrophy rates (outcome measures, jointly-modelled) within diagnostic group, adjusting for intracranial volume. The association between baseline log 2 WMH and hippocampal rates adjusting for concurrent brain atrophy rate was then estimated. Similarly, the association between log 2 WMH and whole-brain rates adjusting for concurrent hippocampal rates was estimated. These models were re-run in the CSF subset, additionally adjusting for CSF Aβ and tau. In controls, greater baseline WMH was associated with higher hippocampal and whole-brain rates (Table 1). The positive association between baseline WMH and hippocampal rates remained in the CSF subset following adjustment for concurrent whole-brain rate and CSF Aβ and tau (see table 2). In MCI, greater baseline WMH was associated with greater hippocampal atrophy rates only and this relationship remained following adjustment for whole-brain rates. In the CSF subset the estimated association between WMH and hippocampal atrophy rate was of a similar magnitude, but was not statistically significant. After adjusting for Aβ and tau measures this estimated association reduced, but again was not statistically significant. In AD, there was no evidence of an association between WMH and either atrophy rate measure. These results suggest that increased WMH is associated with disproportionately higher hippocampal tissue loss in controls and MCI subjects. In controls this association remained after adjustment for CSF amyloid and tau.