Pendlebury ST, Chen PJ, Welch SJV, Cuthbertson FC, Wharton RM, Mehta Z, and Peter M. Rothwell PM. Methodological Factors in Determining Risk of Dementia After Transient Ischemic Attack and Stroke: (II) Effect of Attrition on Follow-Up. Stroke. 2015

In a recently published article in Stroke, Sarah Pendlebury and her colleagues examine the methodological problems associated with ascertainment of post-stroke dementia.

This article is the third in a series of three articles on this topic. In their first two papers, they demonstrated that baseline selection criteria and selective attrition from face-to-face follow-up result in under-representation of those patients with the greatest likelihood of having dementia. In this paper, they specifically aim to understand the extent to which the feasibility of cognitive testing is limited. They also assessed factors that are associated with untestability with short cognitive tests (MMSE, MoCA, Telephone Interview for Cognitive Status and abbreviated mental test score). 

Patients were prospectively recruited from 2002 to 2007 for the Oxford Vascular Study (OXVASC) after TIA or stroke. Cognitive testing was attempted at ascertainment and then on a pre-determined schedule.

Of the 1,236 patients recruited (mean age 75.2 years), 403 had TIA and the remainder were predominantly ischemic stroke patients (5% primary intracerebral hemorrhage). They dichotomized stroke as minor or major, with major defined as NIHSS>3; 370 (30%) had major stroke.

The percentage testable was 76% at cohort entry, 82% at 1 month, 88% at 6 months, 87% at 1 year and 83% at 5 years. Only 46% of patients with major stroke were testable. Notably, being testable was associated with younger age, less pre-morbid disability and less pre-morbid dementia. Reasons for untestability included stroke-related deficits such as dysphasia, hemiparesis, reduced level of consciousness and acute confusion. The proportion of testable stroke patients increased with time after stroke (perhaps due to recovery and also death of sicker patients). Untestability was less of a problem for TIA patients.

A number of papers have recently demonstrated the “mixed pathology” seen in brains of patients with dementia. Understanding the contribution of vascular disease to dementia is crucial for the intelligent design of dementia therapeutics and prevention strategies. This series of papers delineates the methodological issues that result in the possible underestimation of the contribution of vascular disease to dementia. Future studies of stroke-related dementia should anticipate and account for the limitations described by these authors.