Declining incidence of ischemic stroke: What is the impact of changing risk factors? The Tromsø Study 1995 to 2012

• Published in: Stroke
• Main Authors: Vangen-Lønne, Anne M., Wilsgaard, Tom, Johnsen, Stein Harald, Løchen, Maja-Lisa, Njølstad, Inger, Mathiesen, Ellisiv B.
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Lippincott Williams & Wilkins 2017
• Subjects: epidemiology; risk factors; ischemic stroke; time trends; cohort study; Tromsø
• Online Access: https://doi.org/10.1161/STROKEAHA.116.014377

Background and Purpose: It is proposed that 20% to 40% of the decline in first-ever stroke incidence is attributed to the improvement of risk factor control. We estimated the impact of modifiable cardiovascular risk factors on the changing incidence of ischemic stroke (IS) between 1995 and 2012, using individual person data from repeated surveys in a general population. Methods: The proportion of the IS incidence decline explained by change in each risk factor over time was estimated from 1995 to 2012 by Poisson regression among 26 329 participants who attended the fourth Tromsø survey in 1994 to 1995. Hazard ratios for IS were estimated with Cox proportional hazards regression among 27 936 participants who attended at least 1 of the Tromsø surveys in 1994 to 1995, 2001, or 2007 to 2008. Age- and sex-adjusted means or prevalences of risk factors over time were estimated by generalized estimating equations. Results: There were 1226 first-ever IS during 367 636 person-years of follow-up. Changes in cardiovascular risk factors accounted for 57% of the decrease in IS incidence from 1995 to 2012. The most important contributors were decreasing mean systolic blood pressure and smoking prevalence, accounting for 26% and 17% of the observed decline, respectively. Conversely, increasing diabetes mellitus prevalence contributed negatively to the declining IS incidence. Conclusions: Changes in cardiovascular risk factors explained 57% of the decrease in IS incidence from 1995 to 2012. Reduction in systolic blood pressure and prevalence of smoking were the most important contributors.