Myocardial infarction events and cardiovascular risk factor levels in Finnish- and Swedish-speaking populations of Finland

Abstract

Background. The Swedish-speaking minority of Finland is unique, because it has a higher socioeconomic status (SES) and longer life expectancy than the Finnish-speaking majority. We hypothesized that this minority may have a lower attack rate of coronary events and analysed whether this could be explained by their higher SES.

Methods. The population-based myocardial infarction (MI) registers recorded 4,845 MI events in Turku during 1988–1998. Individual-level indicators of SES were obtained from Statistics Finland. The population-based FINRISK surveys recorded cardiovascular risk factors and native languages of 10,432 people in 1987, 1997, and 2002.

Results. The age-standardized attack rate of MI was lower among the 35–64-year-old Swedish-speaking men than among Finnish-speaking men (rate ratio 0.66; 95% confidence interval 0.47–0.85) and the difference remained significant after adjustment for SES. The Swedish-speaking inhabitants had higher age-, sex-, and SES-adjusted high-density lipoprotein cholesterol, and lower triglycerides, body mass index, and diastolic blood pressure.

Conclusion. The Swedish-speaking inhabitants of Turku had lower MI morbidity and coronary mortality than the Finnish-speaking inhabitants. After controlling for SES, these differences remained significant among men, suggesting that other factors, such as differences in the risk factor profiles may also play a role.

Key words::

• Morbidity
• myocardial infarction
• socioeconomic status

Abbreviations
BMI	=	body mass index
BP	=	blood pressure
CI	=	confidence interval
HDL	=	high-density lipoprotein
MI	=	myocardial infarction
RR	=	rate ratio
SES	=	socioeconomic status

Key messages

• The Swedish-speaking population of Finland is a unique minority, because it has a higher socioeconomic status (SES) and longer life expectancy than the Finnish-speaking majority.

• The age-standardized attack rate of MI and 28-day mortality rate of coronary events was lower among the 35–64-year-old Swedish-speaking men than among Finnish-speaking men of the same age.

• The Swedish-speaking people also had higher serum high-density lipoprotein cholesterol, and lower serum triglycerides, body mass index, and diastolic blood pressure, independent of the differences in SES.

• After controlling for SES, differences in the attack rate of MI remained significant among men, suggesting that other factors, such as the discovered differences in the risk factor profiles may play a role.

Introduction

Until the early nineteenth century Finland and Sweden together composed the Kingdom of Sweden. Swedish was the only official language of the country and the language of the nobility, administration, and education. Nowadays, Finland is officially a bilingual country with a Swedish-speaking minority of about 5%. The position of this Swedish-speaking minority is in many respects very unique: they are an integral part of the Finnish culture, but in many realms of life (schools, mass media, organizations, and religion) they have their own institutions. Furthermore, it is known that this linguistic minority has lower mortality rates, a longer life expectancy (1–3), and a better socioeconomic status (SES) than the Finnish-speaking majority (4,5).

Reasons for the better health of the Swedish-speaking minority have not been fully established, but the higher SES or genetic differences have been assumed to play a role. Several studies have examined ethnic minorities in different countries (6–10), but usually these minorities have lower SES and shorter life expectancy than the majority of the population. Interestingly, the NiHonSan Study showed already in the 1970s (11) that the Japanese moving to San Francisco quickly adopt a cardiovascular risk similar to that of the Americans living there.

In our previous study we demonstrated that stroke morbidity is lower among the Swedish-speaking population in Turku, but the difference in stroke morbidity cannot totally be explained by the higher SES of the Swedish-speaking population (12). However, a better understanding of the morbidity differences and further information on the contribution of specific diseases and contributing factors could suggest possibilities for disease prevention.

Myocardial infarction (MI) remains the leading cause of death in Finland and may contribute to the differences in life expectancy between the Finnish-speaking and Swedish-speaking populations in Finland. To our knowledge, no studies have been published so far on differences in MI morbidity and coronary mortality between the language groups. We have now examined differences in coronary mortality, MI morbidity, and cardiovascular risk factor levels between the Finnish- and Swedish-speaking populations in Finland. We hypothesized that the attack rate of coronary events is lower in the Swedish-speaking minority and wanted to find out to what extent it was explained by their higher SES.

Material and methods

Turku is the fourth largest city in Finland located in the south-western part of the country, close to Sweden. It has a population of 175,000 inhabitants. About 5.2% of them are Swedish-speaking, living intermingled with the Finnish-speaking majority. During our study period the 35–99-year-old population of Turku consisted of 88,400 persons (year 1999). All symptomatic acute MI events in Finland are treated in public (municipal) hospitals, where anybody can obtain treatment for a nominal fee. All the inhabitants of Turku, regardless of native language, suffering from suspected symptoms of coronary origin, are referred for care to the Turku University Hospital.

The MI data originate from two large population-based MI registers which aimed to record all coronary events among the inhabitants of Turku. The FINMONICA MI register study was a Finnish contribution to the WHO MONICA project. It operated until 1992, and the FINAMI MI register study has continued after that. The methods of these registers have been previously described in detail (13,14).

The data of this report comprise the MI events occurring in persons aged 35–99 years with permanent residence in Turku during the 11-year period of 1988–1998. The data of the first 5 years of the present report originate from the FINMONICA MI register, and data of the last 6 years originate from the FINAMI MI register. In the FINMONICA MI register (years 1988–1992 of our data) events among persons aged < 65 years were recorded. In years 1993 and 1994 events among persons aged < 75 years were recorded. In years 1995–1997 all events among persons aged 35–99 years were recorded. In 1998 all events were recorded in persons aged 35–75 years, and in persons aged > 75 years events were recorded for those whose birthday's day of month was an odd number, and the total was calculated. The events were classified on the basis of symptoms, electrocardiogram and biomarker findings, and possible autopsy results as suggested in the consensus statement of the American Heart Association and other organizations in 2003 (15). The event rates include non-fatal definite, probable, and possible MIs, definite and probable fatal MIs, and possible fatal coronary events. Information on out-of-hospital deaths was primarily based on the review of death certificates. Deaths in which coronary heart disease or sudden death appeared as an underlying or other cause of death were evaluated for registration based on information on death certificates, medical records, and possible autopsy reports. To ensure complete coverage of out-of-hospital coronary deaths, the FINAMI register is annually linked with the National Causes-of-Death Register on the basis of the personal ID code. Potential missing cases identified by the record linkage are returned to the local register team for evaluation according to the study protocol.

The term ‘attack rate’ is used to describe the rate of all coronary events, whether incident or recurrent, and the term ‘incidence of first coronary event’ is used to describe the rate of first coronary events. The event was considered to be first if there was no evidence of a clinically recognized previous event in the patient's history.

Individual level data on native language, taxable income, profession, and education were obtained by record linkage of the MI register data with the files of Statistics Finland on the basis of the unique personal identification number. Corresponding information was obtained for the background population as well to be used in the denominators. For the analyses, income data were adjusted for inflation (year 2000) and classified into three broad categories: low, middle, and high. Data on profession were also divided into three categories: blue-collar, white-collar, and others (mainly students and unemployed persons). For retired persons, earlier profession was used. Furthermore, education was stratified into three categories: basic (≤ 9 years of full-time education), middle (9–12 years of full-time education), and higher (> 12 years of full-time education). Annual population counts in both language groups were obtained from the national Population Information System (www.vaestorekisterikeskus.fi).

Information on the prevalence of classical risk factors (hypertension, dyslipidaemia, diabetes, smoking, alcohol consumption, and obesity) among the Finnish- and Swedish-speaking populations was obtained from the data of the population-based FINRISK risk factor surveys (16). Our FINRISK data comprise people aged 25–74 years living in the Turku/Loimaa area or the Helsinki metropolitan area (including the cities of Helsinki and Vantaa) and participating in the FINRISK survey in year 1987, 1997, or 2002. Only people aged 25–64 years participated in the study in year 1987.

Information on native language for the FINRISK population sample was obtained from the national Population Information System. Socioeconomic status was measured by household income level and education level. Income was defined as total household income per year and adjusted for family size using the OECD (Organization for Economic Co-operation and Development) equivalence scale (17). Education was stratified into two categories: “basic” and “secondary or higher”. Serum levels of total cholesterol, triglycerides, and high-density lipoprotein (HDL) cholesterol were all determined in one central laboratory, and blood pressure (BP) levels, weight, and height were measured by trained nurses according to a standardized protocol. Smoking was assessed in the questionnaire, and those subjects who were currently smoking (cigarettes, cigars, or pipe) were classified as smokers. Alcohol consumption was asked about in the questionnaire as the number of alcoholic drinks consumed during the preceding week, and the results were expressed as grams of alcohol. The proportion of heavy users (average daily alcohol consumption of ≥ 3 alcohol portions in men or ≥ 2 alcohol portions in women, when one alcohol portion = 12 g of pure alcohol) was assessed. The FINRISK questionnaire included a question: ‘Has a doctor ever told you that you have diabetes, or latent diabetes?’ People, who answered ‘yes’ to one of these, were considered diabetic in the analysis.

Statistical methods

The data of the myocardial infarction registers were divided into three age-groups: 35–64, 65–74, and 75–99 years, and the attack rate of all MIs, incidence of first coronary event, and 28-day mortality rates per 100,000 inhabitants in each language group were counted separately for these age-groups. The rates were age-standardized to the European standard population (18). The 95% confidence intervals (CI) were calculated using normal approximation of the Poisson distribution for the number of events in different age-groups.

The 28-day case fatality was determined as the proportion of fatal events of all events and age- standardized according to the WHO MONICA project using the age distribution of the MONICA MI and stroke registers as the standard. The 95% CIs for the case fatality ratios were calculated using the normal approximation of binomial distribution for the number of deaths.

Age-standardized rate ratios were computed comparing the MI attack rates of the Swedish-speaking inhabitants with those of the Finnish-speaking inhabitants. The comparisons were adjusted for study year and the indicators of SES as follows: study year only (model A), study year and taxable income (model B), study year and education (model C), study year and profession (model D). The 95% CIs for the rate ratios were calculated using the normal approximation of the Poisson distribution.

The average values for total cholesterol, HDL cholesterol, triglycerides, systolic BP, diastolic BP, and body mass index (BMI), as well as the proportions of heavy alcohol users, smokers, and diabetic persons were calculated for the Finnish- and Swedish-speaking FINRISK participants. Differences between the language groups were tested using analysis of covariance for continuous variables and log-linear models for categorical variables. P values under 0.05 were considered statistically significant. The statistical analyses were carried out using SAS software, version 8 (SAS Institute Inc., Cary, NC, USA).

Results

A total of 4,845 MI events were recorded during the study period. Their distribution by language group, age, and sex is presented in Table I.

Table I. Numbers of all events, first events, and deaths among the 35–99-year-old Finnish- and Swedish-speaking inhabitants in Turku in 1988–1998.

		Women		Men
	Age	Finnish	Swedish	Finnish	Swedish
Number of events	35–64	328	9	1379	50
	65–74	427	17	788	44
	75–99	1106	74	579	44
	35–99	1861	100	2746	138
First event	35–64	253	9	944	39
	65–74	315	14	465	25
	75–99	698	37	332	25
	35–99	1266	60	1741	89
Deaths (28 days)	35–64	109	4	563	20
	65–74	197	8	419	20
	75–99	736	48	413	34
	35–99	1042	60	1395	74

In the youngest age-group (35–64 years) the attack rate of MI was higher among Finnish-speaking men and women than among Swedish-speaking men and women (Table II). The 28-day mortality rate of this age-group was also higher among Finnish-speaking men than among Swedish-speaking men. Furthermore, the incidence of first coronary event in the oldest age-group (75–99 years) was higher among Finnish-speaking women than among Swedish-speaking women. The comparisons of attack rates, incidence of first coronary events, and 28-day mortality rates among the other age-groups of women and men showed a trend towards the same direction, but the results were not statistically significant (Table II).

Table II. The age-standardized incidence of first coronary events, attack rate of all MI events, and 28-day coronary mortality rates per 100,000 inhabitants (95% CI in parentheses) among Finnish- and Swedish-speaking men and women in Turku and the rate ratios for these figures.

	Age	Swedish	Finnish	Rate ratio
Men:
Attack rate	35–64	309 (223–396)	467 (442–492)	0.66 (0.47–0.85)^a
	65–74	1877 (1318–2437)	2318 (2154–2481)	0.81 (0.56–1.06)
	75–99	4380 (3081–5680)	5582 (5117–6064)	0.78 (0.54–1.03)
Incidence of first coronary event	35–64	243 (166–319)	320 (299–340)	0.76 (0.51–1.01)
	65–74	1054 (637–1471)	1390 (1264–1517)	0.76 (0.45–1.07)
	75–99	2500 (1516–3484)	3243 (2887–3598	0.77 (0.46–1.09)
28-Day mortality rate	35–64	121 (68–174)	192 (176–208)	0.63 (0.35–0.91)^a
	65–74	860 (480–1240)	1255 (1134–1375)	0.69 (0.38–1.00)
	75–99	3365 (2229–4501)	4103 (3700–4506)	0.82 (0.53–1.11)
Women:
Attack rate	35–64	49 (17–81)	92 (82–103)	0.53 (0.18–0.89)^a
	65–74	537 (279–796)	756 (683–829)	0.71 (0.36–1.06)
	75–99	3214 (2331–3986)	3724 (3501–3947)	0.86 (0.65–1.08)
Incidence of first coronary event	35–64	49 (17–81)	72 (63–81)	0.68 (0.23–1.13)
	65–74	434 (204–663)	567 (504–630)	0.77 (0.35–1.18)
	75–99	1652 (1088–2216)	2361 (2184–2538)	0.70 (0.46–0.94)^a
28-Day mortality rate	35–64	21 (0–41)	31 (25–37)	0.68 (−0.1–1.37)
	65–74	255 (76–433)	354 (304–403)	0.72 (0.20–1.24)
	75–99	2035 (1428–2643)	2482 (2301–2663)	0.82 (0.57–1.07)

^aIndicates statistically significant difference.

The 28-day case fatalities tended to be lower among the Swedish-speaking 35–74-year-old men than among Finnish-speaking men of the same age-groups. The 95% CIs of the 28-day case fatality results were, however, overlapping in both sexes and in all age-groups (Table III).

Table III. Age-standardized 28-day case fatality (and 95% CI) of acute coronary events among the Swedish- and Finnish-speaking inhabitants aged 35–99 years in Turku during 1988–1998.

	Age	Swedish	Finnish	Ratio
Men:
First events	35–64	36.3 (22.2–50.4)	39.5 (36.3–42.6)	0.92 (0.56–1.28)
	65–74	32.5 (14.2–50.7)	55.0 (50.5–59.5)	0.59 (0.25–0.93)^a
	75–99	87.3 (76.0–98.5)	71.9 (67.2–76.6)	1.21 (1.03–1.39)
All events	35–64	38.8 (26.1–51.5)	40.5 (38.0–43.1)	0.96 (0.64–1.28)
	65–74	45.1 (30.2–59.9)	53.3 (49.8–56.8)	0.85 (0.56–1.14)
	75–99	79.2 (67.1–91.3)	71.6 (67.9–75.2)	1.11 (0.93–1.29)
Women:
First events	35–64	37.6 (13.3–61.8)	31.0 (25.2–36.8)	1.21 (0.39–2.03)
	65–74	44.5 (16.9–72.1)	42.6 (38.1–49.2)	1.02 (0.37–1.67)
	75–99	72.2 (58.4–86.1)	63.9 (60.2–67.6)	1.13 (0.90–1.36)
All events	35–64	37.6 (13.3–61.8)	31.6 (26.3–36.8)	1.19 (0.40–1.98)
	65–74	47.5 (23.3–71.7)	45.4 (40.5–50.2)	1.05 (0.51–1.59)
	75–99	65.3 (54.5–76.1)	63.9 (61.0–76.1)	1.02 (0.84–1.20)

^aIndicates statistically significant difference.

Table IV presents the age-standardized rate ratios (RR) comparing the attack rates and the 28-day mortality rates of the Swedish-speaking men and women with those of the Finnish-speaking men and women. When adjusted for the study year only, the Swedish-speaking men had almost 30% fewer myocardial infarctions than the Finnish-speaking men. After adjusting for the indicators of SES, this difference somewhat decreased but remained statistically significant. Furthermore, when adjusted for the study year only, the Swedish-speaking women had almost 20% fewer MIs than the Finnish-speaking women. This trend can also be seen when adjusted for the indicators of SES, but the results of these comparisons are not statistically significant. When adjusted for year of study only, the coronary mortality of the Swedish-speaking men was almost 30% lower than the mortality of the Finnish-speaking men. The difference, however, did not quite remain statistically significant after adjusting for the indicators of SES. No significant differences can be seen in the mortality of myocardial infarction among women.

Table IV. Rate ratios (95% CI) of acute coronary events among the 35–99-year-old Swedish- versus Finnish-speaking inhabitants in Turku during 1988–1998.

	Men		Women
	Attack rate	Mortality rate	Attack rate	Mortality rate
Model A	0.73 (0.62–0.87)^a	0.71 (0.56–0.89)^a	0.81 (0.66–0.99)^a	0.81 (0.63–1.05)
Model B	0.81 (0.68–0.96)^a	0.80 (0.63–1.01)	0.94 (0.77–1.15)	0.98 (0.75–1.27)
Model C	0.80 (0.68–0.96)^a	0.81 (0.64–1.03)	0.84 (0.69–1.03)	0.84 (0.65–1.10)
Model D	0.79 (0.66–0.94)^a	0.79 (0.62–1.00)	0.82 (0.67–1.01)	0.83 (0.64–1.09)

Model A: Adjusted for study year.

Model B: Adjusted for study year and income (lowest income tertile versus highest tertile).

Model C: Adjusted for study year and education (≤ 9 years versus > 12 years of education).

Model D: Adjusted for study year and profession (blue-collar versus white-collar).

^aStatistically significant difference between the language groups.

A total of 9,913 Finnish-speaking and 519 Swedish-speaking people participated in the FINRISK study in areas included in the present analysis. Of the Swedish-speaking people, 31.4% belonged to the lowest income tertile of the total FINRISK population, 30.4% to the middle-income tertile, and 38.2% to the highest tertile. In the Finnish-speaking population the corresponding proportions were 33.5%, 33.5%, and 33.0%. Of the Finnish-speaking people 35.4% and of the Swedish-speaking people 33.0% had basic education only.

The average values and prevalence of the classical risk factors of MI are presented in Table V. The Swedish-speaking participants had higher HDL cholesterol, lower serum triglyceride concentration, lower diastolic BP, and lower BMI than the Finnish-speaking participants. These differences remained statistically significant after adjusting for household income and education. A greater proportion of the Finnish-speaking seemed to be current smokers, but the difference did not reach statistical significance. There was no statistically significant difference in the proportion of heavy alcohol users, diabetic persons, or the level of total cholesterol or systolic blood pressure between the language groups.

Table V. Means and proportions of cardiovascular risk factors in the FINRISK studies of years 1987, 1997, and 2002 in Turku/Loimaa and the Helsinki metropolitan area. P values for the differences between the language groups.

Risk factor	Finnish-speaking (n = 9913)	Swedish-speaking (n = 519)	P^a	P^b
Cholesterol (mmol/L)	5.58	5.51	0.26	0.45
HDL (mmol/L)	1.45	1.49	0.01^c	0.02^c
Triglycerides (mmol/L)	1.45	1.33	0.02^c	0.03^c
Systolic BP (mmHg)	135.0	134.5	0.48	0.70
Diastolic BP (mmHg)	81.8	80.2	0.01^c	0.01^c
Alcohol (%)	4.0	3.9	0.80	0.75
Smoking (%)	26.7	21.8	0.19	0.41
BMI (kg/m^2)	26.5	25.7	0.001^c	0.003^c
Diabetes (%)	4.4	4.8	0.34	0.44

^aAdjusted for age and sex.

^bAdjusted for household income and education in addition to age and sex.

^cStatistically significant difference between the language groups.

HDL = high-density lipoprotein cholesterol; BP = blood pressure; BMI = body mass index.

Discussion

Our main finding was that the attack rate of MI events was lower among Swedish-speaking men than among Finnish-speaking men, and this difference remained significant after adjusting for the higher SES of the Swedish-speaking men. Among women, the attack rates showed a consistent tendency to the same direction, although the differences between language groups became marginally non-significant after adjusting for SES. Previous Finnish epidemiological health surveys have already demonstrated lower total mortality among the Swedish-speaking minority compared to the Finnish-speaking majority (1,2). Significant disparities have been demonstrated especially in cardiovascular mortality (1,19).

Several previous studies (in Finland and elsewhere) have shown higher coronary heart disease mortality and morbidity rates among people with a lower SES (20,21). Ethnic differences have, however, usually been in favour of the majority. For example, a recent report from New York City, USA, reported a considerable increase in mortality rates from several causes, including coronary heart disease, among Asians compared to white inhabitants (10). In Jerusalem, Israel, coronary event rates were much higher among Arabs than among the Jewish population (9). Against this background, the position of the Swedish-speaking minority in Finland is unique. Similar and even greater differences in cardiovascular disease risk have been observed between the Finns moving to Sweden and the Swedish population (22). In our previous study we found lower stroke morbidity and mortality among Swedish-speaking men than among Finnish-speaking men in Turku (12). In the present study, as in the previous study concerning stroke, the better SES of the Swedish-speaking minority did not fully explain their lower morbidity. The measurement of SES is, however, inevitably somewhat crude, and therefore we cannot totally exclude the possibility of residual confounding. Together these results suggest that in addition to SES there are other factors, possibly related to healthier life-style, influencing the lower MI morbidity and mortality of the Swedish-speaking population.

Our study also provided information on the classical risk factors of MI (BP, BMI, cholesterol, smoking, and alcohol consumption) among the Finnish- and Swedish-speaking inhabitants of Turku, Loimaa, and the Helsinki metropolitan area. We found that the Swedish-speaking people had higher levels of serum HDL cholesterol, lower levels of triglycerides, lower diastolic BP, and lower BMI than the Finnish-speaking people. Although statistically significant, the magnitudes of these differences were modest, and their clinical significance is uncertain. For example, the statistically significant difference in HDL cholesterol is probably clinically trivial, whereas the non-significant difference in smoking may actually have clinical relevance. This reflects the different statistical behaviour of continuous and dichotomous variables. However, all differences were to the expected direction, suggesting higher risk among Finnish-speaking persons, and thus consistent with the observed differences in the coronary mortality and morbidity. We suggest that the lower BMI may be the key feature here, and the differences in BP and lipid profile may be consequential upon the BMI difference. The Swedish-speaking people had higher education and higher household income levels, but the differences in the risk factors were independent of the differences in the socioeconomic factors. In a previous study, the Swedish-speaking people have reported lower rates of daily smoking (23). Furthermore, the Finnish-speaking people have been reported to be more prone to binge drinking (24). We were not able to confirm these findings in the present study.

Social capital has earlier been suggested to be a beneficial mediator in population health (25). The longer life expectancy and better health of the Swedish-speaking minority in Finland have also been suggested to be mediated by the larger amount of social capital of the Swedish community (26). However, another study demonstrated that the Swedish-speaking people living intermingled with the Finnish-speaking people (instead of living as a community) do not necessarily possess more social capital attributable to human connections compared to the Finnish-speaking people (27).

Significant differences in the presence and severity of coronary calcification (as a marker of coronary atherosclerosis) according to ethnicity, independent of atherosclerotic risk factors, have been demonstrated in countries with more diverse population (28). These results may suggest differences in genetic susceptibility especially between racial groups. Genetic studies have shown the Swedish-speaking population of Finland to have a Finnish admixture of about 60% in their genes (29). A more recent genome-wide structure analysis of single nucleotide polymorphisms has shown that the Swedish-speaking population of Finland has twice as large an admixture with the Swedish people in their genes as the other people of western Finland have (30). Thus, although the Finnish- and Swedish-speaking people in Finland belong to the same race, differences in genetic susceptibility may play a role in the differences in all-cause mortality and MI morbidity.

The main limitation to this study was the small number of recorded MI events due to the small size of the Swedish-speaking population of Turku. To increase the numbers, we included persons up to the age 99 years in our study. Unfortunately, however, the numbers of Swedish-speaking participants in FINRISK surveys were so small that the statistical power was insufficient for meaningful prospective cohort analyses on risk factors and MI events. Furthermore, 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 lower degree even among the oldest old (21). Our data are 10 years old, but the socioeconomic health gap in Finland has since widened, and therefore the differences between the language groups may be even more significant today (31). The strengths of this study were that the study was based on comprehensive population-based MI registers and risk factor surveys and that the unique social security numbers have enabled us to obtain accurate socioeconomic data by record linkage to the files of Statistics Finland.

In conclusion, we found that the Swedish-speaking inhabitants of Turku had lower MI morbidity and coronary mortality than the Finnish-speaking inhabitants. The difference in morbidity remained significant among men after controlling for SES, suggesting that other factors, possibly related to a healthier life-style, may also play a role. Furthermore, the Swedish-speaking people of Turku, Loimaa, and the Helsinki metropolitan area had lower levels of the classical risk factors, HDL cholesterol, triglycerides, diastolic BP, and BMI, suggesting that differences in risk factor profiles may explain some of the difference in MI morbidity.

Acknowledgements

The study has been presented as a poster presentation in the EuroPrevent 2009 congress on 7 May 2009. The authors would like to thank the FINAMI Study Group for data collection.

Declaration of interest: The study was supported by the King Gustav V and Queen Victoria's Foundation and by the Finnish Foundation for Cardiovascular Research. The authors declare that there is no conflict of interest.