Greater ischemic stroke severity was associated with faster cognitive decline and higher dementia risk in a pooled cohort study of more than 42,000 patients.

The study included 42,342 patients aged 45 years or older without baseline stroke or dementia from 3 US prospective cohorts: the Atherosclerosis Risk in Communities study, the Framingham Offspring Study, and the Reasons for Geographic and Racial Differences in Stroke study. Cognitive follow-up lasted a median of 11.1 years, with follow-up extending up to nearly 30 years.

Researchers evaluated first-ever definite ischemic strokes and categorized stroke severity using the National Institutes of Health Stroke Scale (NIHSS). Scores of 0 to 5 were classified as minor stroke, scores of 6 to 10 as mild to moderate stroke, and scores of 11 or higher as moderate to severe stroke.

The primary outcomes were change in global cognition and incident dementia. Secondary outcomes included changes in memory and executive function. For dementia incidence, follow-up was counted beginning at baseline or age 55 years, whichever came later, because only 3 dementia cases occurred before that age.

Among 1,505 patients with first-ever definite ischemic stroke, NIHSS severity data were available for 1,055 patients. Most strokes with available severity data were minor: 879 were classified as minor, 125 as mild to moderate, and 51 as moderate to severe.

Compared with patients without stroke, those with minor stroke had nearly twice the likelihood of dementia, those with mild to moderate stroke had more than 3 times the likelihood, and those with moderate to severe stroke had about 5 times the likelihood.

A dose-response pattern was also observed for cognitive decline, particularly for global cognition and memory. Annual decline in global cognition was −0.18 points among patients without stroke, compared with −0.36 points among those with minor stroke, −0.47 points among those with mild to moderate stroke, and −0.58 points among those with moderate to severe stroke. Memory decline similarly steepened with increasing stroke severity.

The magnitude of decline may be clinically intuitive when considered in terms of cognitive aging. Based on the rate of cognitive aging observed in the study, the authors calculated that patients with mild to moderate stroke declined cognitively as if they were 1.8 years older at baseline, whereas those with moderate to severe stroke declined as if they were 2.6 years older at baseline.

Executive function showed a different pattern. It declined faster than global cognition or memory even among patients without stroke, at −0.33 points per year. Decline worsened among patients with stroke, but the severity gradient flattened at higher NIHSS levels, with mild to moderate and moderate to severe stroke showing nearly identical annual executive-function decline rates of −0.52 points per year.

The researchers noted that executive function may be particularly sensitive to vascular injury because it relies on fronto-subcortical circuits and white matter pathways. This may help explain why executive function declined relatively quickly even in patients without stroke and why additional stroke severity did not produce as clear a stepwise gradient as observed for global cognition and memory.

Exploratory analyses also suggested that blood pressure may influence some associations between stroke severity and cognitive trajectories, although findings differed by blood pressure measure, cognitive domain, and stroke severity. Higher diastolic blood pressure appeared to attenuate global cognitive decline among patients with minor and moderate to severe stroke. Higher systolic blood pressure appeared to attenuate the association between minor stroke and executive-function decline.

The researchers suggested that these differences may reflect distinct vascular mechanisms. Elevated systolic blood pressure may be linked to large artery stiffness, pulse pressure, and white matter injury, whereas diastolic blood pressure may be more closely tied to cerebral perfusion pressure. No interactions between stroke severity and vascular risk factors were observed for the dementia outcome.

Dementia ascertainment varied across the 3 cohorts. The REGARDS cohort, which contributed most of the sample, used a Six-Item Screener algorithm rather than physician-adjudicated dementia. However, sensitivity analyses excluding REGARDS produced stronger relative dementia-risk estimates than the primary analysis, with patients who had minor, mild to moderate, and moderate to severe stroke showing about 2, 5, and 8 times the likelihood of dementia, respectively, compared with patients without stroke.

Sensitivity analyses using alternative Framingham stroke-severity mappings, adjustment for apolipoprotein E epsilon 4 status among patients with available genetic data, and cohort-specific models were consistent with the primary results.

The study also excluded dementia diagnosed within 1 year following stroke to avoid capturing acute or transient cognitive changes. The researchers noted that this exclusion may have omitted common early poststroke cognitive changes and may have led to underestimation of the overall poststroke dementia burden.

Other limitations included the observational design, which precluded causal conclusions; missing NIHSS data for nearly 30% of first-ever ischemic strokes; and relatively small numbers of patients with more severe stroke. Patients with severe stroke may also have been underrepresented in follow-up because of death, disability, or fewer poststroke visits. Therefore, the true burden of cognitive decline and dementia associated with severe stroke may be higher than the study estimates suggest.

The study also lacked brain imaging data, neuroinflammatory or neurodegenerative biomarkers, stroke-location data, recurrent-stroke data, and information on peak adult cognitive ability. As a result, the findings could not determine which mechanisms or lesion characteristics contributed most to poststroke cognitive decline.

Despite these limitations, the graded associations across 3 prospective cohorts and the consistency of sensitivity analyses supported the overall finding that more severe ischemic stroke was linked to greater long-term cognitive risk. The researchers wrote that the findings highlight the importance of prevention strategies, vigilance in cognitive monitoring, and identifying mechanisms that may link stroke to cognitive decline.

Disclosures can be found in the published investigation.

Source: JAMA Network Open