The matter of white and gray matter in cognitive impairment

Abstract: Cognitive impairment spans from minor subjective cognitive impairment to disabling dementia. Many biomarkers have been developed to monitor different aspects of cognitive impairment. Magnetic resonance imaging is the most used neuroimaging biomarker in research and can measure gray matter (GM) and white matter (WM) changes. Although there is a consensus that atrophy in GM is a marker for neuronal loss, there is little evidence assessing the role of WM changes. The aim of this thesis is to first develop a tool to reliably measure the changes in WM in the form of white matter hyperintensities (WMH) and second to evaluate the role of WM and GM changes in the early stages of cognitive decline. In Study I and Study II, a fully automated method for segmentation of WMH has been developed and validated. Validation results indicated that the WMH segmentation was performed with high similarity to manual delineation and with superb reproducibility. In Study III, coronary heart disease (CHD) and hypertension, which are known to contribute to WM damage, were examined and their effect on GM and WM changes was investigated on a group of 69 individuals with 30-year follow-up. We showed that CHD and hypertension indeed affect the GM volume and thickness and the effect of CHD is partially independent of hypertension. However, the results indicate no significant effect on WMH, which we believe is due to the fact that WMH were measured as a crude total volume. In Study IV, a pipeline was developed to isolate the WM tract connecting each GM region to the rest of the brain and to measure the burden of WMH on each tract, hereinafter tractbased WMH. We used a cohort of 257 cognitively normal (CTL), 87 subjective cognitive impairment (SCI) and 124 mild cognitive impairment (MCI) subjects and examined their GM volume, tract-based WMH and cognitive performance. Our results indicated that the fraction of variance in GM volume that can be explained by tract-based WMH in SCI subjects is significantly higher than in both CTL and MCI subjects. The results also showed that in subjects with high and low cognitive performance, tract-based WMH can barely explain any GM volume change. However, in subjects with slight cognitive impairment tract-based WMH can explain the changes in GM volume. In summary, we investigated different ways of measuring the damage of WMH and showed that the role of WMH is more pronounced when measuring them in relation to the WM tract they affect. The effect of WMH on GM has been shown to be mainly in the earlier stages of cognitive impairment.