Abstract
Objectives. To examine differences in the morbidity and mortality of stroke between the Finnish- and Swedish-speaking populations in Turku, taking into account the socioeconomic differences between these groups. Design. The population-based FINMONICA and FINSTROKE stroke registers recorded 5 135 stroke events among persons aged 25 – 99 years in Turku during 1988 – 1998. Events in persons aged ≥75 years were not registered in 1993 – 1995. Information on these persons’ native language and socioeconomic status (SES) (measured by taxable income, profession and years of education) were obtained by record linkage with the files of Statistics Finland. Results. Swedish-speaking men had a lower attack rate of ischaemic stroke than Finnish-speaking men (270, 95% Confidence Interval (CI) 214-326 versus 370, 95% CI 352-389, per 100 000 inhabitants per year) and the difference remained significant after adjustment for SES. Among women, the attack rates of ischaemic stroke were similar in both language groups. Conclusions. The attack rate of ischaemic stroke was lower among Swedish-speaking than among Finnish-speaking men. This difference was not totally explained by the higher SES of the Swedish-speaking population.
Finland is a bilingual country. Approximately 5% of the inhabitants belong to the Swedish-speaking minority. This minority originates mainly from the Swedish immigrants during the years 1100–1300 AD but has a rather high degree (60%) of Finnish admixture in their genes Citation1. It is known that the Swedish-speaking minority has a longer life-expectancy and better socioeconomic status (SES) than the Finnish-speaking majority Citation2, Citation3. Reasons for the life-expectancy difference and contribution of specific diseases are poorly known. Better understanding of the difference and contributing factors could provide learning opportunities for the Finnish speaking majority and suggest possibilities for disease prevention.
Stroke is a devastating disease and an important public health problem, which may contribute to the life expectancy differences between the Finnish-speaking and Swedish-speaking persons in Finland. To our knowledge, no study has analysed stroke morbidity by the language group in Finland. In the WHO MONICA Project, stroke attack rate in Turku/Loimaa area of Finland was clearly higher than in southern Sweden but of the same order of magnitude than in northern Sweden Citation4.
The town of Turku is located in southwestern Finland, close to Sweden. It has a Swedish-speaking minority of 5.2% of the population living intermingled with the Finnish-speaking majority. Turku was also a reporting unit of the WHO MONICA stroke register Citation5, and continued the stroke registration after the MONICA period in the FINSTROKE study Citation6. We have now obtained information on the native language and on indicators of SES for each individual registered in the stroke register of Turku during the period 1988–1998. The purpose of the present study was to compare the morbidity and case fatality of ischaemic and haemorrhagic stroke among Swedish- and Finnish-speaking inhabitants of Turku, taking into account the SES differences between the two population groups.
Material and methods
Turku has the population of 175 000 inhabitants. During our study period the average 25–99 year-old population living in Turku included 110 323 Finnish-speaking people (49 673 men and 60 650 women) and 6 047 Swedish-speaking people (2 778 men and 3 269 women).
Stroke data originated from the FINMONICA and FINSTROKE stroke registers Citation6. All stroke events occurring in persons aged 25 to 74 years with permanent residence in Turku were recorded during the years 1988–1998. Events occurring in persons aged=75 years were recorded in 1988–1992 and 1996–1998. To confirm the diagnosis, either a computerized tomography, magnetic resonance imaging or an autopsy was performed on 78% of the cases. The rest were based on clinical judgement and/or on cerebrospinal fluid examination. In cases where the subtype of stroke could not be determined, the FINSTROKE register used the diagnostic category ‘unclassifiable’ Citation6. In the age group 25–64 there were 1.4% of such cases and in the age group 65–74 the corresponding proportion was 5%. In present analyses we combined the unclassifiable cases with the ischemic stroke category.
Individual level data on native language, taxable income, profession and education were obtained by record linkage of the stroke register data with the files of Statistics Finland on the basis of the personal identification number, unique to every resident of Finland. For retired persons, the earlier profession was used. For the analyses, the income data were adjusted for inflation (year 2000) and classified into three categories: low, middle and high. Education was stratified into two categories: basic (=9 years of full-time education) and higher (>9 years of full-time education). Annual population counts in each language group were obtained from the National Population Information System (www.vaestorekisterikeskus.fi). 1-year and 5-year case-fatalities were identified by record linkage of the stroke register data with the National Causes of Death Register.
Statistical methods
The morbidity (per 100 000 inhabitants in each language group) and characteristics of the stroke patients (proportions of first stroke, history of myocardial infarction, diabetes, history of hypertension, current smoking, history of atrial fibrillation, the usage of acetyl salicylic acid (ASA), type of work, income tertile and education) were age-standardized to the European standard population Citation7. The 95% CIs were calculated using the normal approximation of Poisson distribution for the number of events or characteristics in the different age groups. The 28-day case fatality was defined as the proportion of fatal events of all events and age-standardized according to the practice of the WHO MONICA Project using the age distribution of myocardial infarction and stroke patients in MONICA registers as the standard. The 95% CIs for the case fatality ratio were calculated using the normal approximation of binomial distribution for the number of deaths.
Age-standardized rate ratios were computed comparing the attack rates of ischaemic and haemorrhagic strokes in Swedish-speaking inhabitants with those in Finnish-speaking inhabitants. The 95% CIs for the rate ratios were calculated using normal approximation of the Poisson distribution. The comparisons were adjusted for study year and SES as follows: study year only (model A), study year and family income (model B), study year and education (model C), study year and profession (model D).
The 1-year and 5-year prognoses after the stroke event were estimated by using Cox proportional hazards regression analyses. The Swedish-speaking population was taken as the reference category and the Finnish-speaking were compared with them. All the statistical analyses were carried out using SAS software, version 8. (SAS Institute Inc., Cary, NC, USA).
Results
A total of 5 135 stroke-events were registered in this population during the eleven year period. About 83% of these events were ischaemic strokes and about 17% were haemorrhagic strokes. Of the events, 94% were observed in Finnish-speaking inhabitants and the rest in Swedish-speaking inhabitants ().
Table I. Numbers of ischaemic and haemorrhagic stroke events, age-standardized average annual attack rates per 100 000 inhabitants, and age-standardized case fatalities in Turku by native language and gender during 1988 – 1998.
The age-standardized rate of ischaemic stroke was higher among the Finnish-speaking men than among the Swedish-speaking men (). Among women, the difference did not reach statistical significance. The rate of haemorrhagic stroke was higher among Finnish-speaking women than among Swedish-speaking women but the 95% CIs were wide. No differences were observed in the rates of haemorrhagic stroke in men or in the 28-day case fatality of any type of stroke. The 1- and 5-year case fatalities of ischaemic stroke did not differ between the language groups. The corresponding case fatalities of haemorrhagic stroke tended to be in favour of the Swedish-speaking men and women but the 95% CIs were overlapping, except for the 1-year case fatality in women (). Furthermore, the hazard ratios for the 1- and 5-year case fatalities after any kind of stroke, counted for men and women together, were also statistically insignificant: 1.10 (95% CI 0.92-1.31) and 1.10 (95% CI 0.95-1.26)
The clinical risk factors of stroke (history of diabetes, hypertension, smoking, and prior myocardial infarction) all tended to be more frequent among Finnish-speaking stroke patients than among Swedish-speaking patients but the 95% CIs were overlapping (). History of atrial fibrillation was more frequent among Finnish-speaking than among Swedish-speaking men. The registered stroke was more often first among the Swedish-speaking women than among the Finnish-speaking women. Clearest differences were, however, observed in indicators of SES. A greater proportion of Finnish-speaking than Swedish-speaking population was manual workers, belonged to the lowest income tertile and had basic education only.
Table II. Characteristics (age-standardized proportion and 95% Confidence Interval) of Finnish- and Swedish-Speaking Stroke Patients in Turku during 1988–1998.
Age-standardized rate ratios (RR) comparing the attack rates of ischaemic and haemorrhagic strokes in the Swedish-speaking inhabitants with those in the Finnish-speaking inhabitants are presented in . When adjusted for study year only, the Swedish-speaking men had almost 30% less ischaemic strokes than the Finnish-speaking men. This difference decreased after adjusting for the indicators of SES, but remained statistically significant and still the Swedish-speaking men had over 20% less ischaemic strokes than the Finnish-speaking men. No significant differences were observed among women in the rate ratios of ischaemic stroke. When adjusted for study year only, the Swedish-speaking women had 25% less haemorrhagic strokes than Finnish-speaking women. This trend was also seen after adjusting for the indicators of SES but these results were not statistically significant. Also among men the 95% CIs of the rate ratios of haemorrhagic stroke were wide.
Table III. Age-standardized rate ratios and 95% confidence intervals of the registered ischaemic and haemorrhagic stroke events in the 25–99 year old Swedish- versus Finnish-speaking inhabitants of Turku during 1988–1998.
Discussion
We found that the age-standardized attack rate of stroke was higher among the Finnish-speaking men than among the Swedish-speaking men of Turku, southwestern Finland. However, we found no significant differences between the language groups in the occurrence of haemorrhagic strokes, nor in the case fatality of any type of stroke. The attack rate differences were somewhat attenuated after taking into account the higher SES of the Swedish-speaking men, but were not totally abolished. One of four Finnish-speaking male stroke patients had a history of atrial fibrillation, while among the Swedish-speaking patients the corresponding proportion was one of ten. Also, certain other risk factors for stroke, such as hypertension, diabetes and history of myocardial infarction tended to be more frequent among Finnish-speaking stroke patients, in particular among males but the differences were not statistically significant.
Several epidemiological health surveys in Finland have demonstrated lower total mortality in the Swedish-speaking minority compared to the Finnish-speaking majority Citation2, Citation3. Significant disparities have been established especially in cardiovascular mortality Citation2, Citation8. Many studies have also found higher stroke incidence, stroke mortality and all-cause mortality among stroke-patients with lower SES than among patients with higher SES Citation9–12. Estimates of the population attributable risk have indicated that low SES may account for one third of the incidence and more than half of the mortality of ischaemic stroke Citation9. In the present study, the SES differences did not fully explain the differences in stroke attack rate between the language groups. The measurement of SES is, however, inevitably somewhat crude. Therefore, we cannot totally exclude the possibility of residual confounding.
In countries with more diverse population, studies have been done on the impact of ethnicity on stroke rate. Differences independent of conventional cardiovascular risk factors or social class have been found especially between different racial groups Citation13, which suggest differences in genetic susceptibility. The Swedish-speaking population of Finland has been shown to have a Finnish admixture of about 60% in their genes Citation1. Thus, some genetic differences exist, although both the Finnish- and Swedish-speaking are white Caucasians. Accordingly, differences in genetic susceptibility may play a role in the differences in life-expectancy and stroke morbidity.
The history of atrial fibrillation was observed in a clearly larger proportion of the Finnish-speaking than Swedish-speaking male stroke patients. However, in the background population possible differences in the prevalence of atrial fibrillation between the language groups are unknown. Atrial fibrillation at the time of stroke has been suggested to predict a higher risk of death after stroke Citation14, and it may therefore contribute to the previously observed shorter life-expectancy of the Finnish-speaking population.
In a Finnish study the observed disparities in the life-expectancy and health of the two language groups have been attributed to the larger amount of social capital of the Swedish-speaking community Citation15, as social capital has earlier been suggested to be a beneficial mediator in population health Citation16, Citation17. In another study, however, the results did not support the assumptions that the Swedish-speaking population living intermingled with the Finnish majority (instead of living as a community) would possess more social capital attributable to human relations compared with the Finnish-speaking population Citation18.
Stroke is a disease of older people and, as the majority of aged are women, the number of stroke events is higher among women. This difference should not affect our comparisons between the language groups, since the rates are age-standardized and shown separately for both genders. Socioeconomic health differences decrease with age. Differences in income level and prestige diminish after the retirement and with aging. However, the importance of SES to cardiovascular risk does not completely vanish with age but remains to a lesser degree even among the oldest old Citation19.
The main limitation of this study is the small number of Swedish-speaking people living in Turku, and therefore the small number of stroke events. The sub-categorization of education to more refined categories than the two groups (<9 years and≥9 years) was not possible, because of the small number of events. The data are ten years old, and what was true then may not be relevant today. However, the health gap between social classes in Finland has widened, and therefore the differences between the population groups may be even more important today Citation20, Citation21. We had no information on the prevalence of stroke risk factors in the Finnish- and Swedish-speaking background populations of Turku, which prevented us from including these factors in our Poisson regression models. Furthermore, our study provided no information on the level of alcohol consumption among these two language groups. It has, however, been reported earlier that Finnish-speaking persons are more prone to binge drinking Citation22, which may explain some of the Finnish-speaking men's ischaemic strokes also in our material. Our study was not designed to detect differences in traditional cardiovascular risk factors between the Swedish and the Finnish speaking population groups and was insufficiently powered to do that. We had no data on the prevalence of dyslipidemia, either. Thus, we cannot tell on the basis of these data to what extent the higher rates of ischemic stroke in Finnish speaking men than in Swedish speaking men could be due to differences in the levels of classical risk factors.
The limitations of our study are, however, balanced by its strengths: a comprehensive population-based stroke register including information of stroke risk factors, and the unique social security number of each resident of Turku enabling the record linkage to the socioeconomic data of Statistics Finland.
In conclusion, the Finnish-speaking men had a higher rate of ischaemic stroke compared to the Swedish-speaking men and a difference remained even after adjustment for the indicators of SES.
Acknowledgements
The study was financially supported by the King Gustav V and Queen Victoria Foundation and by the Finnish Foundation for Cardiovascular Research. The study also received Evo Funding from the Turku Health Centre. The authors declare that there is no conflict of interest.
References
- Virtaranta-Knowles K, Sistonen P, Nevanlinna HR. A population genetic study in Finland: Comparison of the Finnish- and Swedish-speaking populations. Hum Hered 1991; 41: 248–64
- Näyhä S. Geographic variations in cardiovascular mortality in Finland 1961-1985. Scand J Soc Med Suppl 1989; 40: 1–48
- Valkonen T. Psychosocial stress and sociodemographic differentials in mortality from ischaemic heart disease in Finland. Acta Med Scand Suppl 1982; 660: 152–64
- Stegmayr B, Asplund K, Kuulasmaa K, Rajakangas AM, Thorvaldsen P, Tuomilehto J. Stroke incidence and mortality correlated to stroke risk factors in the WHO MONICA Project. An ecological study of 18 populations. Stroke 1997; 28: 1367–74
- Tuomilehto J, Rastenyte D, Sivenius J, Sarti C, Immonen-Räihä P, Kaarsalo E, et al. Ten-year trends in stroke incidence and mortality in the FINMONICA Stroke Study. Stroke 1996; 27: 825–32
- Sivenius J, Tuomilehto J, Immonen-Räihä P, Kaarisalo M, Sarti C, Torppa J, et al. Continuous 15-year decrease in incidence and mortality of Stroke in Finland: The FINSTROKE Study. Stroke 2004; 35: 420–5
- Waterhouse J, Muir C, Correa P. Cancer incidence in five continents (IARC Scientific Publication No 15) Volume 3. International Agency for Research on Cancer, Lyon 1976
- Koskinen S. Origin of regional differences in mortality from ischaemic heart disease in Finland. Research Report 41. Helsinki, National Research and Development Centre for Welfare and Health 1994.
- Jakoljevic D, Sarti C, Sivenius J, Torppa J, Mähönen M, Immonen-Räihä P, et al. Socioeconomic status and ischaemic stroke. The FINMONICA Stroke register. Stroke 2001; 32: 1492–8
- Kunst AE, Groenhof F, Andersen O, Borgan JK, Costa G, Desplanques G, et al. Occupational class and ischemic heart disease mortality in the United States and 11 European countries. Am J Public Health 1999; 89: 47–53
- Kapral MK, Wang H, Mamdani M, Tu JV. Effect of sosio-economic status on treatment and mortality after stroke. Stroke 2002; 33: 268–73
- Avendano M, Kunst AE, Huisman M, van Lenthe F, Bopp M, Borrell C, et al. Educational level and stroke mortality: A comparison of 10 European populations during the 1990s. Stroke 2004; 35: 432–7
- Markus HS, Khan U, Birns J, Evans A, Kalra L, Rudd AG, et al. Differences in stroke subtypes between black and white patients with stroke: The South London Ethnicity and Stroke Study. Circulation 2007; 116: 2157–64
- Kaarisalo MM, Immonen-Räihä P, Marttila RJ, Salomaa V, Kaarsalo E, Salmi K, et al. Atrial fibrillation and stroke. Mortality and causes of death after the first acute ischemic stroke. Stroke 1997; 28: 311–5
- Hyyppä MT, Mäki J. Why do Swedish-speaking Finns have longer active life? An area for social capital research. Health Promot Intern 2001; 16: 55–64
- Kawachi I, Kennedy BP, Lochner K, Prothrow-Stith D. Social capital, income inequality and mortality. Am J Publ Health 1997; 87: 1491–8
- Baum F. Social capacity: Is it good for your health? Issues for a public health agenda. J Epidemiol Community Health 1999; 53: 195–6
- Volanen SM, Suominen S, Lahelma E, Koskenvuo M, Silventoinen K. Sense of coherance and its determinants: A comparative study of the Finnish-speaking majority and the Swedish-speaking minority in Finland. Scand J Public Health 2006; 34: 515–25
- Avendano M, Kunst AE, Huisman M, Lenthe FV, Bopp M, Regidor E, et al. Socioeconomic status and ischaemic heart disease mortality in 10 western European populations during 1990s. Heart 2006; 92: 461–7
- Mackenbach JP, Bos V, Andersen O, Cardano M, Costa G, Harding S, et al. Widening socioeconomic inequalities in mortality in six Western European countries. Int J Epidemiol 2003; 32: 830–7
- Rahkonen O, Talala K, Sulander T, Laaksonen M, Lahelma E, Uutela A, . Koettu terveys. Health inequalities in Finland. Trends in socioeconomic health differences 1980-2005, H Palosuo, S Koskinen, E Lahelma, R Prättälä, T Martelin, A Ostamo, et al. Ministry of Social Social Affairs and Health, Helsinki 2007; 65–73
- Hyyppä MT, Mäki J. Social participation and health in a community rich in stock of social capital. Health Educ Res. 2003; 18: 770–9