Threshold of white matter hyperintensity volume determines negative and positive associations between grey matter volume and hypertension

Abstract

AbstractBackgroundHypertension and white matter hyperintensity(WMH) burden are major risk‐factors for Alzheimer’s disease[Pantoni et al.,2010;Sweeney et al.,2018]. Prior studies show variable grey matter alterations, involving both negative and positive associations, independently related to WMH and hypertension in early stages of late‐onset AD[Vipin et al.,2018;Foster‐Dingley et al.,2015;Tuladhar et al.,2015;Den Heijer et al.,2003]. However, the effect of incremental WMH load and concomitant hypertension on brain structure in young‐onset mild cognitive impairment(MCI) remains to be elucidated. Such investigations may provide better insight into relationships between WMH, hypertension and dementia risk. We thus examine the influence of increasing WMH load on grey matter volume(GMV) as well as on the association between blood pressure(BP) and GMV in young‐onset MCI. We hypothesize that WMH burden and BP will have derogatory effects on GMV, especially under high WMH conditions.MethodT1‐MRI data from 78 MCI subjects(<65years) from the Singapore‐YouNg‐Dementia‐Cohort was studied. WMH load was divided into low(0.00‐1.49ml), medium(1.50‐4.99ml) and high(>5.00ml). Voxel‐based morphometry(VBM) assessed whole‐brain voxel‐wise changes related to increasing WMH load between hypertensive(systolic‐BP>130mmHg)[Whelton et al.,2017] and non‐hypertensive subjects. Pearson’s correlation analysis tested associations between GMV from key Executive‐Control‐Network(ECN) and Default‐Mode‐Network(DMN) regions‐of‐interest(ROI)[Shirer et al.,2012] and vascular risk‐factors.ResultWhole‐brain VBM revealed less GMV reduction in hypertensive compared to non‐hypertensive subjects in frontal and temporal regions at low‐WMH load, with more pronounced effects at medium‐WMH load (Fig.1;uncorrected‐p<0.001, cluster‐size‐threshold:p<0.05). However, at high‐WMH load, hypertensive and non‐hypertensive subjects showed no differences in GMV. Additionally, at low‐WMH load, average and ROI‐specific DMN(bilateral posterior‐cingulate‐cortex, precuneus;p<0.05) and ECN(left[p<0.05] and right[p<0.10] posterior‐parietal‐cortex) GMV related positively to WMH but showed no association at medium‐WMH load. Importantly, only high‐WMH load was associated with lower ECN(p<0.05) and DMN(p<0.10) GMV(Fig.2). These associations remained regardless of hypertension status. Additionally, high systolic‐BP related to lower GMV at high‐WMH load only(Fig.3).ConclusionOur findings indicate that in young‐onset MCI, the relationship between WMH and GMV is determined by WMH load rather than hypertension status. WMH load also influences the relationship between systolic BP and GMV in young‐onset MCI. Importantly, our results suggest that WMH and hypertension influence GMV via different mechanisms and BP recommendations for dementia prevention may be different in MCI with and without WMH.