THE RELATIONSHIP OF HYPERTENSION, DIABETES AND OTHER VASCULAR RISK FACTORS WITH POSTSTROKE COGNITIVE FUNCTION: THE STROKOG (STROKE AND COGNITION) CONSORTIUM

Abstract

While hypertension, diabetes, atrial fibrillation (AF), smoking and past stroke are associated with increased risk of dementia, it is not known whether they contribute to cognitive impairment in patients with recent stroke. This study examines the relationship between vascular risk factors and cognitive function in stroke patients from a combined sample of international cohorts. Eleven studies from the STROKOG consortium provided individual participant data (IPD). Demographic data were harmonised and for each participant, using appropriate normative data from individuals without stroke, z-scores for 5 domains (attention, memory, language, visuospatial and executive function) were calculated from neuropsychological tests completed 3–6 months after index stroke/TIA. Linear mixed effects models in a one-stage IPD meta-analysis were used to examine the association between each cognitive domain z-score and risk factor, adjusting for sex, age and education. Among 2,981 stroke/TIA participants from 9 countries (61% male, mean age 66 years, 94% ischaemic stroke, 4% TIA and 2% haemorrhagic stroke), the mean z-scores in all domains were nearly 1 SD below normal. 70% had a history of hypertension, 30% diabetes, 14% AF, 14% past stroke*and 47% were ex- or current smokers. Diabetes and a history of past stroke were strongly related to poorer cognition in all domains, with global cognition scores (the average of 5 domains) being 0.44 SD less (95%CI: -0.56, -0.31; p-value <0.001) in participants with diabetes and 0.37 SD less (-0.53, -0.22; <0.001) in those with past stroke. Hypertension was associated with poorer cognition in all domains except for memory, with a 0.19 SD ([-0.32, -0.07]; 0.002) lower global cognition score. Smoking contributed negatively to memory and visuospatial domains (-0.14 SD [-0.25,-0.03]; 0.016), (-0.22 SD [-0.35, -0.09]; 0.001), but not the other domains. The relationships between AF and domain scores were not significant (Table 1). Diabetes and history of past stroke are both associated with poorer cognitive performance in stroke patients. Hypertension and smoking are also associated with poorer cognitive function although the effects are smaller and not in all domains. *History of past stroke indicates having one or more strokes prior to the index event.