Epidemiology of stroke in an urban population - aspects of time, place and person

The present thesis explored the epidemiology of stroke in Malmö, Sweden, an urban population of approximately 250000 inhabitants. Incidence of stroke in Malmö has been monitored since 1989. The National census investigation in 1990 was used to get information about the background population in the city. Previous studies of the seasonal and weekday variations in incidence of stroke have shown inconsistent results. In this study, the temporal variation in stroke incidence was explored among 7129 patients with a first stroke between 1989 and 1999. No relationship could be found between season or weekday and incidence of stroke. However, the case-fatality rate (within 28 days) was significantly higher among patients with stroke in winter. Several studies have shown that divorced or widowed men and women have higher mortality rates than those who are married. Whether incidence of stroke similarly differ between these groups is less clear. In a study of stroke incidence in Malmö between 1990 and 2000, we found that incidence of stroke was increased in divorced men and women and in widows/widowers. Never married men had not any increased risk. Among men and women who initially were married, divorce or death of spouse was followed by an increased incidence of stroke. Previous studies have shown that incidence of stroke varies between countries and ethnic groups. We explored whether incidence of stroke was related to country of birth. Inhabitants from former Yugoslavia and Hungary had an increased risk of ischemic stroke. People from former Soviet Union and China or Vietnam had an increased risk of intracerebral haemorrhage, and people born in Finland had an increased risk of subarachnoid haemorrhage. People born in Rumania had lower risk than those born in Sweden. Previous studies from the city have shown that incidence of stroke show substantial differences between the residential areas in the city. Areas with high incidence are characterised by less favourable socioeconomic circumstances. We explored whether the survival among the stroke cases similarly was related to the residential area. The 28-day mortality, 1-year mortality and long-term mortality (until 2001) was higher in patients from areas with low socioeconomic level. These relationships reached significance mainly for the longer follow-up periods and for patients below 75 years of age. It is concluded that incidence of stroke is related to marital status and country of birth. The residential area and season of the stroke event are associated with the prognosis after the stroke. Key words: Stroke, case fatality, temporal trends, marital status, life events, stroke outcome, survival, ethnic groups, incidence, epidemiology, geography, Sweden