Hypertension is a major risk factor for stroke and dementia and is associated with white matter hyperintensities (WMH) and reduced brain volumes. We measured the increase in WMH volume, and rate of cerebral atrophy over two years, in hypertensive subjects participating in the Study on COgnition and Prognosis in the Elderly (SCOPE), receiving candesartan or placebo, and normotensive controls. We recruited 163 subjects who had MRI (FLAIR and volumetric T1) at 2 and 4 years after baseline assessment. From these two scans, volumetric change in WMH (n = 133) and brain atrophy rates (n = 95) were determined.Total WMH fraction increased in both normotensive and treated hypertensive groups (p < 0.01) median change: 0.05% of brain volume [range: -0.45% to 1.51%]. Deep WMH increased in hypertensive (p = 0.001) but not the normotensive group. The number of subjects with an increase of total WMH in the 5th quintile differed between the treatment groups (chi square p = 0.006), being greatest in the placebo group (32%), then candesartan (20%) then normotensive (5%). Regression analysis found significant predictors of change in WMH to be blood pressure and initial deep WMH, but not treatment group. Increased atrophy rate was predicted by baseline systolic blood pressure (p = 0.02) but was not associated with measures of WMH. Similar to WMH, there was a trend with treatment, with atrophy in normotensive < Candesartan < Placebo (Spearman's rho = 0.23, p = 0.026). Hypertension in older people is associated with increased rates of progressive whole brain atrophy and an increase in WMH. These changes are independent. Successful hypertension treatment was associated with reduced risk of WMH progression and possibly brain atrophy.

Hyperintense lesions (HL), as visualized on T2-weighted or FLAIR MRI, are a common finding in older people, but their clinical significance and influence on cognitive function remain to be clarified. We investigated the relationship between HL in deep white and gray matter structures and cognition in older subjects. We recruited 154 nondemented (Mini-Mental State Examination > 24) subjects (79 males) over the age of 70 from primary care (103 subjects with mild hypertension and 51 normotensive subjects). All subjects underwent FLAIR and proton density and T2-weighted axial 1.5-tesla MRI scans (slice thickness: 5 mm). The scans were rated for the presence and distribution of HL in the subcortical gray matter (caudate, putamen, globus pallidus, thalamus) and associated white matter tracts (internal/external capsule). Subjects (n = 149) underwent a comprehensive cognitive assessment involving tests of attention, processing speed, episodic memory, working memory, and executive function. Partial correlations (correcting for age, systolic blood pressure, and New Adult Reading Test [NART] score) were performed to investigate the relationship between cognition and white matter change. HL were found in 49% of subjects. HL in both the gray (thalamus and caudate) and white matter were significantly associated with impaired cognitive function in tasks involving processing speed and/or executive function, but showed no associations with episodic or working memory. HL in both subcortical gray matter structures and associated fiber tracts correlate with impairments in attention, executive function and processing, and memory retrieval speed in nondemented older community-dwelling subjects. Such lesions may be an important cause of age-related attentional and executive dysfunction in the elderly, as well as temporal lobe and hippocampal changes that have previously been reported to be associated with impairments to the ability to actually store and retrieve information from memory.