Subcortical vascular dementia: Epidemiology and risk factors

Wide variation in definition, detection, and description of lacunar lesions on imaging

BACKGROUND AND PURPOSE

Variation in the definition of lacunar lesions on imaging and difficulties in their detection may be hampering lacunar stroke research. We assessed literature definitions of imaging lacunar lesions and the definitions and detection of lacunar lesions among small-vessel disease researchers.

METHODS

We assessed definitions of imaging lacunar lesion in 50 randomly selected articles from 3 stroke-related journals and an online survey of small-vessel disease researchers. In the literature review, we assessed clinical/imaging definitions of lacunar stroke. In the survey, we assessed lacunar lesion detection, effects of lesion appearance, background white matter lesions, and provision of relevant data.

RESULTS

Among 50 articles, imaging definitions were varied and often limited; size was stated in 21 of 43 (49%) studies of acute and in 9 of 20 (45%) studies of old lesions and site in 18 (42%) and 4 (20%), respectively. Clinical definitions also varied, and images were read mostly by nonradiologists. Among 56 survey respondents, multiple descriptions were used for recent and old, symptomatic and asymptomatic, lesions on imaging. Most agreed on definitions for site (98%) and "old lacunar infarct" (61%) size. Cavitated (vs noncavitated) lesions were usually identified as lacunar lesions; with increasing white matter lesions, however, noncavitated lesions were very unlikely to be identified, even with prior imaging available (7.8%).

CONCLUSIONS

Imaging definitions of lacunar lesions vary widely, in part due to variation in lesion detection and classification. A consensus for imaging definitions of small-vessel disease features would be helpful.

Epidemiology of vascular dementia

Although epidemiological studies are limited by diagnostic uncertainties, they suggest that stroke increases the risk of dementia. The mortality rate is higher in vascular dementia (VaD) than in Alzheimer's disease (AD). Community-based studies have provided several consistent findings: (i) age dependence with prevalence rates doubling every 5 years, (ii) a higher frequency in men and (iii) nation-to-nation differences. The prevalence of VaD ranges from 2.2% in 70- to 79-year-old women, to 16.3% in men >80 years. One sixth of acute stroke patients have preexisting dementia. The incidence of VaD has been studied much less extensively than that of AD, and substantial variations in the incidence rates have been observed: annual incidence rates (per 100,000) range from 20 to 40 between 60 and 69 years of age and from 200 to 700 over 80. The incidence rate of VaD declined over the last 2 decades, probably as a consequence of effective stroke prevention. It is generally assumed that risk factors for VaD are those of stroke, with arterial hypertension as leading factor, followed by atherosclerotic disease, low education level, alcohol abuse and heart disease. Stroke characteristics, such as lacunar infarction and left-sided hemispheric lesions, are major determinants of VaD. The cerebrovascular lesions are likely to be the only cause of dementia in strategic infarcts, in lacunar state, in hereditary cystatin C amyloid angiopathy and in CADASIL. However, white matter changes, and associated Alzheimer pathology, which are both frequent in this age category, may also contribute to the cognitive decline.