Association of modifiable risk factors with ischaemic stroke subtypes in Asian versus Caucasian populations: A systematic review and meta-analysis

Background: Ischaemic stroke (IS) is associated with various modifiable risk factors but the association of these risk factors based on TOAST classification, which characterises IS into five subtypes: large artery atherosclerosis (LAA), small vessel occlusion (SVO), cardioembolic disease (CE), other determined aetiology (ODE) and undetermined aetiology (UDE), is unknown. We aimed to summarise the published evidence for the association of modifiable risk factors with IS subtypes based on TOAST classification, specifically focussing on the Asian versus Caucasian population. Method: A comprehensive search for all the published articles was performed in electronic databases including PubMed, EMBASE, Cochrane Library, and Google Scholar from 01st January 1950 to 10th April 2022 based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Odds ratio (OR) with 95% confidence interval (CIs) along with random-effect models was used to calculate summary estimates. Results: In our meta-analysis, 32 studies with a total of 23,404 IS (14,364 in Asian vs. 9040 in Caucasian population), 7121 LAA (5219 in Asian vs. 1902 in Caucasian), 5532 SVO (3604 in Asian vs. 1928 in Caucasian), 3498 CE (1634 in Asian vs. 1864 in Caucasian), 1131 ODE (546 in Asian vs. 585 in Caucasian) and 4519 UDE (2076 in Asian vs. 2443 in Caucasian) were included. Our findings suggest a significant association between LAA and hypertension (OR = 1.07, 95% CI = 1.02-1.12), smoking (OR = 1.11, 95% CI = 1.04-1.17), dyslipidemia (OR = 1.13, 95% CI = 1.06-1.21), diabetes mellitus (OR = 1.18, 95% CI = 1.11-1.25) and atrial fibrillation (OR = 0.55, 95% CI = 0.40-0.75). Significantly strong association of hypertension, smoking, dyslipidemia, diabetes mellitus and atrial fibrillation was observed with SVO and CE stroke subtypes. Subgroup analysis based on ethnicity revealed a significant association for dyslipidemia, diabetes mellitus and atrial fibrillation in LAA for both Asians and Caucasians. Hypertension was significantly associated with SVO and ODE subtypes in both Asians and Caucasians; however, only Asian population showed significant association of hypertension in LAA and CE subtypes. The other risk factors did not show any statistical difference between the ethnic groups for the different stroke subtypes. The majority of the risk factors depicted positive association with LAA and SVO, negative with CE and neutral with ODE and UDE. Conclusion: Our findings suggest strong association of smoking, dyslipidemia and diabetes mellitus with LAA and SVO subtypes in the Caucasian population. However, only diabetes mellitus showed significant association with both LAA and SVO subtypes in Asian population as well. Thus, a majority of the traditional modifiable risk factors had a positive association in LAA and SVO, while a negative protective association was observed in CE subtype, among both the Asian and the Caucasian subgroups.