Stroke epidemiology in Southern Sweden. Trends in incidence and survival across two decades with projections into the future

Abstract: Background: Stroke is a major cause of death and disability. Investigations of the epidemiological trends for stroke might result in better understanding of stroke pathogenesis and prevention. Methods: The trends of first-ever stroke incidence and survival over 3 time periods (1983-85, 1993-95 and 2001/2002) in Lund-Orup health care district were studied. The first 2 study periods were retrospective and mainly hospital-based studies whereas the third study period used prospective and population-based design. The risk of dying after stroke was analyzed in a Cox proportional hazards analysis (with sex, age group, time period and stroke subtype in the analysis). A projection of the future number of strokes in Sweden was made. Results: The incidence rate (age standardized to the European Standard population per 100 000 person-years) was 134 in 1983-85 (95%CI 126-143; 998 patients), 158 in 1993-95 (95%CI 149-168; 1318 patients) and 144 in 2001/2002 (95%CI 130-158; 456 patients). The increase between the first 2 time periods was mainly confined to persons younger than 75 years (25% increase), and mainly observed for the stroke subtypes lacunar syndrome and posterior circulation syndrome. The survival within 28 days did not change significantly between the 3 study periods, but long-term survival improved between all 3 study periods. Between the first 2 periods, the survival improvement was seen only among patients aged ?75 years. The risk of dying for patients with stroke during 1993-95 was 0.85 and during 2001/2002 it was 0.67, compared with patients with stroke during 1983-85. During 2001/2002, 84% were hospitalized (within 14 days); we detected 2 main categories of patients who were not hospitalized: (1) elderly women with high 28-day case-fatality (61%), residing at nursing homes (86%), and (2) male patients with milder strokes and very good survival (0% 28-day and 1-year case-fatality). Based on official Swedish population projection and age and sex standardization from our incidence data, we project (presuming stable age-specific incidence rates) that the total annual number of first-ever strokes in Sweden might increase by no less than 55% up to year 2050. Conclusions: Stroke incidence and survival changed in our study area, which might contribute to an increase of the prevalence of stroke in the future. The importance of continuous surveillance of stroke epidemiology trends is highlighted by the projected increasing burden of stroke in our country.