Trends in Case-Fatality of Stroke in Finland During 1983 to 1992
Background and Purpose Stroke mortality has been declining in Finland during the past 20 years. It is not known, however, whether this favorable development is attributable to the decline in the incidence or case-fatality of stroke. For this reason we examined the trends in case-fatality of stroke, including trends by subtype of stroke.
Methods The analyses were carried out using data of the community-based FINMONICA Stroke Register, which was operating in three geographic areas of Finland during 1983 to 1992. All stroke events (n=11 171) in persons aged 35 to 74 years were included in this register.
Results The 28-day case-fatality of stroke fell yearly by 3.6% (P=.01) in men and by 2.6% (P=.2) in women. At the end of the study period, the average 28-day case-fatality of all strokes was 20% in men and 21% in women. Considerable differences by subtype of stroke were observed. The 28-day case-fatalities at the end of the study period were in men—56% for subarachnoid hemorrhage, 42% for intracerebral hemorrhage, and 14% for cerebral infarction. In women, the corresponding figures were 49%, 49%, and 14%. The 28-day case-fatality of subarachnoid hemorrhage did not change during the study period, but for intracerebral hemorrhage, a significant decline was observed in men and there was a declining trend also in women. The 28-day case-fatality of cerebral infarction declined significantly in both genders.
Conclusions With the exception of subarachnoid hemorrhage, the 28-day case-fatality of stroke has fallen in Finland. It is likely that this fall has contributed to the decline in stroke mortality.
Cardiovascular disease mortality has declined in many industrialized countries since the early 1970s, although rates have increased in eastern European countries.1 2 3 4 5 6 7 Although considerable attention has been focused on the declining coronary heart disease component of cardiovascular disease mortality, less attention has been paid to the even more dramatic decline in cerebrovascular disease (stroke) mortality, which in many countries began well before that in coronary heart disease. Stroke registers can provide more accurate estimates for mortality rates than routine mortality statistics.8 Furthermore, stroke incidence and case-fatality can be estimated from register data, and thereby, reasons for the decline in stroke mortality can be examined more thoroughly. Stroke registers are an integral part of the World Health Organization’s MONICA Project,9 which was set up in the early 1980s to get more reliable information on the trends in cardiovascular disease incidence and mortality.
Results of the FINMONICA Stroke Register Study published in 1996 showed that both the incidence and the attack rate of stroke declined.10 In pooled FINMONICA data, a statistically significant decline in the incidence of stroke was observed both in men and in women. The attack rate of stroke also declined significantly in both sexes and within each of the three monitored areas. Mortality rate from stroke declined by 5.2% per year in men and by 4.7% per year in women. This decline was significant in all three areas among men but among women only in Kuopio. The incidence to mortality rate ratio increased during the study period, indicating a steeper fall in mortality than in incidence. A substantial decline in both stroke incidence and mortality was observed in all age groups. A part of the decline in stroke mortality, observed also in the official mortality statistics, can be attributed to the decline in stroke incidence during this 10-year period. In this article, we present detailed analyses of the case-fatality of stroke, including stroke subtypes, and discuss its contribution to the decline in stroke mortality during 1983 to 1992.
Subjects and Methods
The FINMONICA Stroke Register was the Finnish contribution to the World Health Organization’s Multinational MONItoring of Trends and Determinants in CArdiovascular Disease (WHO MONICA) Project, which is an international collaborative study set up in the beginning of the 1980s.11 The main aim of the study is to analyze the relation between temporal trends in incidence and mortality rates of myocardial infarction and stroke and changes in major cardiovascular risk factors over a 10-year period.
The FINMONICA Stroke Register was operating in three geographic areas of Finland: the provinces of North Karelia and Kuopio in eastern Finland and the Turku-Loimaa area in southwestern Finland.12 The stroke register was started in North Karelia and Turku-Loimaa in 1982 and in Kuopio in 1983. Data for the 10-year time period from 1983 to 1992 are given for Kuopio and Turku-Loimaa. The 10-year period for North Karelia covered the years 1982 to 1991, but for simplicity of presentation 1 year was added in this report to each study year so it is presented as a ten-year period from 1983 to 1992, along with the two other FINMONICA areas.
The methods of the FINMONICA stroke register have been described in detail previously.12 13 14 It has followed the protocol and quality control procedures of the WHO MONICA Project.9 10 11 12 13 14 15 16 17 18 The main sources of case findings were the admission diagnoses to the hospital; hospital discharge diagnoses and diagnoses from death certificates were also checked routinely. All patients, aged 25 to 74 years and residing permanently in the target areas, with symptoms and signs suggesting acute cerebrovascular disease, were evaluated for registration. At the National Public Health Institute, the FINMONICA Stroke Register data were annually cross-checked with the computerized National Death Register for completeness. Routine stroke mortality data (International Classification of Diseases codes 430-438) for the FINMONICA populations were obtained from the National Death Register maintained by Statistics Finland.19 In this analysis, data for men and women aged 35 to 74 years were used.
According to the definition of the WHO MONICA Project, the diagnosis of stroke was based on “rapidly developing clinical signs of focal (or global) disturbance of cerebral function lasting more than 24 hours (unless interrupted by surgery or death), with no apparent cause other than a vascular origin.”18 The stroke event was classified as a first-ever event if there was no evidence of a previous stroke in the patient’s history. If the patient suffered from another cerebrovascular event more than 28 days after the onset of the first attack, this other event was classified as a recurrent stroke and a new register form was filled out. The stroke event was classified as fatal if the patient died of any cause within 28 days, counted from the day of onset. If the patient died during the first or the following day from the onset of the symptoms, the case was considered fatal within two days.
Events were categorized as “definite stroke,” “not stroke,” or “unclassifiable.”9 18 Criteria for definite stroke were fulfilled when the available information permitted a clinical diagnosis of stroke. Unclassifiable was used when no diagnosis other than stroke was present to explain the event, but the available information was insufficient for determining whether symptoms and duration fully met the MONICA criteria for definite stroke. Only approximately 4% of fatal events and 0.5% of nonfatal events were categorized as unclassifiable. In the analyses of the present study, unclassifiable cases were combined with the definite stroke category.
In the analyses by subtype of stroke, strokes were divided into three categories: subarachnoid hemorrhage, intracerebral hemorrhage, and cerebral infarction. The examinations required for the classification of subcategories of stroke have been described,12 and it was documented earlier14 20 that the three subcategories mentioned above can be separated from each other reasonably reliably in the FINMONICA Stroke Register. In the beginning of the study period, the availability of computerized tomographies was limited. In cases in which computed tomographies or necropsy was not done, the subtype classification was based on other available clinical information, such as typical symptoms and signs, existence of blood in the cerebrospinal fluid, angiography, and other possible investigations. If lumbar puncture was not performed, the patient was always classified as having had an unspecified stroke. The same specialist physician (a neurologist in North Karelia, Kuopio, and Turku, and an internist in Loimaa) carried out the classification in each area during the whole study period. In 1987, the WHO MONICA Project adopted certain changes in the requirements for hemorrhagic stroke (positive findings in computed tomography or in angiography became necessary for the classification of strokes as subarachnoid hemorrhages or as intracerebral hemorrhages). In FINMONICA, however, the original classification was also retained to enable unbiased estimation of the trends in subtypes of strokes. This original classification was used in the analyses of the present report. In the beginning of the study period, 1983 to 1985, about 20% of subarachnoid hemorrhages were classified without benefit of the examinations recommended in the WHO MONICA protocol. At the end of the study period, this proportion was only 3.3%. As in the previous reports of the FINMONICA Stroke Study14 unspecified type of strokes were pooled together with thrombotic and embolic strokes, referred to as cerebral infarction.
Case-fatality is defined as the proportion of events that were fatal within the given time period, calculating the day of onset as day 0. Two categories of case-fatality are reported: 2-day and 28-day case-fatalities. The case-fatality figures are presented by dividing the 10-year monitoring period into three parts, 1983 to 1985, 1986 to 1989, and 1990 to 1992, to obtain stable estimates.
Age-standardized case-fatality rates were calculated using weights 1, 3, 7, and 12 for the 10-year age groups 35 to 44, 45 to 54, 55 to 64, and 65 to 74 years, respectively, reflecting the age distribution of all stroke events in the data of the WHO MONICA Project. Confidence intervals were calculated on the basis of binomial distribution. The trends in case-fatality of stroke were estimated from the logarithm of the annual age-standardized case-fatality rate using linear regression, and the confidence intervals of the trend were obtained in the usual manner from the standard error of regression coefficient. Regression coefficient multiplied by 100 is given as the average yearly change in percents. The Statistical Analysis System was used for computations.21 Logistic regression analyses were carried out for subcategories of stroke to find out the independent effect of age, sex, calendar year, and area on the case-fatality of stroke.
Number of Events
During 1983 to 1992, a total of 11,373 acute stroke events were registered. Patients aged 25 to 34 years (123 men and 79 women) were excluded from this analysis. Of the included events, 6307 (56%) occurred in men and 4864 (44%) in women. These events were distributed by area as follows: 5122 (46%) in Kuopio, 2890 (26%) in North Karelia, and 3159 (28%) in Turku-Loimaa. Among men, 22% (1387) were fatal within 28 days and among women 25% (1200) of cases were fatal. The lowest proportion of fatal strokes was found in Kuopio (19% among men and 21% among women), whereas in North Karelia strokes accounted for 27% and 28% and in Turku-Loimaa 23% and 27%, respectively (Table 1⇓). There were 969 subarachnoid hemorrhages, of which 500 (52%) occurred in men and 469 (48%) in women, 1128 intracerebral hemorrhages, of which 640 (57%) were in men and 488 (43%) were in women, and 9074 cerebral infarctions, of which 5167 were in men (57%) and 3907 (43%) were in women.
Of the stroke events, 636 (10%) strokes occurred out of the hospital in men and 384 (8%) in women. In Kuopio, 90 (16%) of the fatal events in men occurred out of the hospital, in North Karelia, 51 (12%), and in Turku-Loimaa, 55 (14%), and in women these were in Kuopio, 51 (18%), in North Karelia, 29 (8%), and in Turku-Loimaa, 37 (10%). In Kuopio, of the nonfatal strokes in men, 373 (16%) were not treated in the hospital, in North Karelia, 16 (1%), and in Turku-Loimaa, 51 (4%), respectively, and in women, these were in Kuopio, 234 (13%), in North Karelia, 10 (1%), and in Turku-Loimaa, 23 (2%).
Case-Fatality of Stroke
In the FINMONICA areas combined, the 28-day case-fatality of stroke declined among men aged 35 to 74 years from 24.3% in 1983 to 1985 to 19.8% in 1990 to 1992 and among women from 25.9% to 21% (Table 2⇓). It was lowest in Kuopio and highest in North Karelia. Among people aged 35 to 64 years, the 28-day case-fatality tended to be higher in women than in men, whereas in people aged 65 to 74 years the 28-day case-fatality was approximately similar in both genders (Table 2⇓). The two-day case-fatality of stroke declined from 11.3% in 1983 to 1985 to 8.6% in 1990 to 1992 in men and from 12.3% to 9.5% in women (Table 3⇓).
Twenty-eight-Day Case-Fatality of Subarachnoid Hemorrhage, Intracerebral Hemorrhage, and Cerebral Infarction
During 1983 to 1992 in FINMONICA areas combined, the 28-day case-fatality of subarachnoid hemorrhage was 56.5% in men and 46.6% in women aged 35 to 74 years, and it did not change substantially during the study period. The 28-day case-fatality of intracerebral hemorrhage declined among men aged 35 to 74 years from 57.9% in 1983 to 1985 to 41.9% in 1990 to 1992 and among women from 55.3% to 48.8% (Table 4⇓). The 28-day case-fatality of cerebral infarction among men aged 35 to 74 years declined from 17.8% in 1983 to 1985 to 13.9% in 1990 to 1992 and among women from 18.9% to 13.6% (Table 4⇓).
Trends in Case-Fatality
In the FINMONICA areas combined, the average decline in the 28-day case-fatality of stroke was 3.6% (P=.01) per year in men aged 35 to 74 years (Table 5⇓). Among the three areas, the fall was steepest in Turku-Loimaa, 5.2% per year (P=.01). Also, in women, a falling trend was observed; in Kuopio it was 4.7% per year (P=.01) and in the FINMONICA areas combined it was 2.6% per year (P=.2). No significant trends were observed in the age group 35 to 64 years, although the decline in men in Turku/Loimaa approached statistical significance (4.9% per year; P=.06). In men aged 65 to 74 years, the case-fatality of stroke fell in Turku-Loimaa by 5.7% (P=.04) per year, and in the FINMONICA areas combined it fell by 4.1% per year (P=.01; Table 5⇓). In women aged 65 to 74 years, we observed a declining trend in Kuopio by 4.2% per year (P=.06) and in Turku-Loimaa by 4.1% per year (P=.02), but no trend was observed in North Karelia. In the FINMONICA areas combined, the declining trend was not significant in women aged 65 to 74 (2.8% per year; P=.08).
The two-day case-fatality of stroke in men fell significantly in all age groups in Turku-Loimaa (Table 6⇓), and a nonsignificant declining trend was observed also in Kuopio. In combined FINMONICA areas, the two-day case-fatality of stroke fell in men aged 35 to 74 years by 3.7% per year (P=.04). In women, the decline in the 2-day case-fatality of stroke was about as large as in men, although it did not reach statistical significance because of smaller numbers.
In the FINMONICA areas combined, no improvement was found in the 28-day case-fatality of subarachnoid hemorrhage (Table 7⇓). The 28-day case-fatality of intracerebral hemorrhage fell significantly in men aged 35 to 64 years by 4.6% per year (P=.003) and in men aged 65 to 74 years by 5.5% per year (P=.007), but among women a declining trend was observed only in the age group 65 to 74 years. The 28-day case-fatality of cerebral infarction fell equally among younger women and men. In men aged 35 to 74 years, the decline was 4.6% per year (P=.02) and in women 4% per year (P=.001).
In the logistic regression analyses, calendar year and area did not influence the case-fatality of subarachnoid hemorrhage, but age increased the risk by 2.7% per year (P=.0001) and men had a 1.5 times higher risk than women (P=.008). Age and gender did not influence the case-fatality of cerebral hemorrhage, but with calender year there was a 7.6% fall each year (P=.0003) and in Turku-Loimaa the case-fatality was 1.5 times higher than in Kuopio (P=.007). Gender was not related to the case-fatality of cerebral infarction, but the risk increased with age by 5.7% per year (P=.0001). Each calendar year the risk decreased by 6.3% (P=.0001); in Turku-Loimaa the case-fatality was 1.2 times (P=.03); in North Karelia, respectively, 1.5 times higher (P=.0001) than in Kuopio.
We report a decline in 28-day case-fatality of stroke in the FINMONICA areas during the 10-year period 1983 to 1992. Of the subtypes, the decline was seen in cerebral hemorrhages and cerebral infarctions. When controlling for age and gender and area, the declines were 7.6% and 5.7% per year for these two subtypes, respectively. There are only a few reports on the trends in case-fatality of stroke in other countries. A decrease in the 30-day case-fatality rate was observed in Rochester, Minnesota, from 33 to 17% between the periods of 1945 through 1949 and 1980 through 1984.22 The Honolulu Heart Study reported a 10.4% decrease in 30-day case-fatality of thromboembolic stroke during 1985 to 1988. For hemorrhagic stroke, the change in case-fatality was 28.7% during 1985 to 1988.23 From 1984 to 1991, the case-fatality of stroke decreased among rural men in Shanghai, China, from 62.4 to 46.9% and urban women from 46.7 to 33.3%; nonsignificant trends were found among rural women and urban men.24 In Novosibirsk, Russia, a decreasing trend was reported in the case-fatality of women from 1982 to 1992, but in men the trend was increasing.25 In Auckland, New Zealand, the 28-day case-fatality of stroke tended to decline during 1981 to 1991, but the trend was not statistically significant in any age or sex group.26
It is of paramount importance that the case ascertainment is complete in case-fatality studies of stroke. This is better achieved in population-based studies, in which also out-of-hospital events are included. Finland has the national population register, national death register, and hospital discharge register, which help to secure the case findings. It is rare in Finland that suspected stroke patients are not hospitalized. FINMONICA study nurses have checked also all cases admitted to the wards of primary health centers because of cerebrovascular disease. In principle, the case-fatality trends could be biased if either case-finding procedures or classification of stroke events change during the study period. The FINMONICA Stroke Register team has made every effort to ensure that this did not happen. Particular attention was paid to the completeness of the registration of fatal events. The stroke register data were cross-checked annually with the National Death Register using computerized record linkage. Therefore, it is not likely that the coverage of fatal events would have changed during the study period. Also, the nonfatal events were identified similarly throughout the study period using hospital admission lists and hospital discharge reports of all hospitals in the monitoring areas.
The main potential source of bias is improved diagnostic accuracy because of the introduction of computerized tomographies in the FINMONICA areas during the latter half of the 1980s. These have enabled the detection of milder strokes, which could have contributed to the improvement in case-fatality. However, in the FINMONICA Stroke Register, the events were classified by a single specialist physician in each area throughout the whole study period. According to the WHO MONICA protocol, the ascertainment of strokes is based on clinical symptoms and signs, and not on the findings in CT scan or magnetic resonance imaging. These diagnostic examinations are primarily used for subtyping of stroke. It is possible that the within-area trends in case-fatality could be biased owing to diagnostic changes over time. Nevertheless, the between-area differences in the case-fatality of stroke are to some extent attributable to the differences in registration procedures. Our quality control study27 showed that suspected stroke events were classified somewhat differently by the neurologist in Kuopio than the neurologist in North Karelia. The activity of the Kuopio center in doing cerebral imaging in milder events causes results of the FINMONICA centers to be analysed also separately. As to the subtypes of stroke, we cannot exclude the possibility that the detection of small intracerebral hemorrhages improved toward the end of the study period, which may have contributed to the declining case-fatality. A bias caused by this phenomen is not, however, likely to be large.
Fatal and nonfatal strokes out-of-hospital comprise a mixture of mild strokes and very severe ones, and the attempt to determine their case-fatality is not valid. Their proportion of such stroke events was small, and it remained stable during the study period. The fatal stroke events that occurred out of the hospital were very large and caused sudden death, whereas the few nonfatal strokes not treated in the hospital were mild and the patients were dismissed home from the outpatient clinic. The study centers ascertained equal proportions of fatal events. However, the proportions of out-of-hospital nonfatal strokes were different; as in Kuopio, their proportion was somewhat larger than in North Karelia or in Turku-Loimaa.
The decline of case-fatality to the total decline in stroke mortality in Finland was observed and described in earlier studies. Observed decline in case-fatality is the result of improved medical care or may be a consequence of milder strokes caused by beneficial changes in risk factors that have occurred in Finland during that time. A declining trend in the three main risk factors, blood pressure, smoking, and cholesterol, has been observed in Finland. It was estimated that these risk factor change explain at least two-thirds of the fall in mortality from stroke in men and somewhat less in women.28 From 1987 to 1992 declining trends in blood pressure, smoking and serum cholesterol accelerated, with the exception of smoking among women.28 Antihypertensive drug treatment was started on a large scale in Finland in the 1970s. Drug treatment can explain only a part of the decrease in the mean blood pressure, because the entire blood pressure distribution has been shifted toward lower values. Salt intake was high in Finland,29 but has fallen since the late 1970s by about 15%,29 which can explain a part of the change in blood pressure. The lower levels of blood pressure in the population28 and improved control of high blood pressure in medical emergencies such as stroke may have contributed to lowering of the case-fatality of stroke.
The case-fatality rate at 28 days was on average 30% in the centers included in the stroke study of the WHO MONICA Project during 1985 to 1987 and varied threefold among populations.9 In FINMONICA, the case-fatality was among the lowest. In the present study, two-day case-fatality accounted for almost one-half of the total case-fatalities within 28 days from the onset of symptoms. The patients with a rapidly fatal stroke often have widespread atherosclerosis, and the death in acute stroke may sometimes be caused by coronary heart disease. The decrease in coronary morbidity in the Finnish population,30 31 may, thus, also have contributed to the decrease in case-fatality of stroke.
In conclusion, our findings demonstrate a declining trend in the 28-day case-fatality of stroke in FINMONICA areas during 1983 to 1992. These findings suggest that the improvement in case-fatality has contributed to the decline of stroke mortality in Finland.
We are grateful to the hospitals, primary health care centers, and health care personnel of the registration areas for supporting the local data collection. We also thank Dr Minna Kaarisalo for technical assistance in the preparation of this paper. Most of the practical work in case-finding and data-collection was done by the nurses, Elvi Surakka in North Karelia, Anu Mononen in Turku, Hannele Kastarinen, Sinikka Röynä, Eva Talasniemi, and Markku Kalinen in Kuopio, and Aino-Maija Kantee in Loimaa.
- Received May 6, 1997.
- Revision received September 18, 1997.
- Accepted September 19, 1997.
- Copyright © 1997 by American Heart Association
Central Bureau of Statistics. Deaths by Causes of Deaths, Age and Gender. Voorburg, The Netherlands; 1985. Annual Report.
Bonita R, Stewart A, Beaglehole R. International trends in stroke mortality: 1970-1985. Stroke.. 1990;21:989-992.
Modan B, Wagener DK. Some epidemiological aspects of stroke: mortality/morbidity trends, age, sex, race, socioeconomic status. Stroke.. 1992;23:1230-1236.
WHO MONICA Project, prepared by Thorvaldsen P, Asplund K, Kuulasmaa K, Rajakangas AM, Schroll M. Stroke incidence, mortality and case fatality in the WHO MONICA Project. Stroke.. 1995;26:361-367.
Tuomilehto J, Rastenyte D, Sivenius J, Sarti C, Immonen Raiha P, Kaarsalo E, Kuulasmaa K, Narva EV, Salomaa V, Salmi K, Torppa J. Ten-year trends in stroke incidence and mortality in the FINMONICA Stroke Study. Stroke.. 1996;27:825-832.
WHO MONICA Project, prepared by Tuomilehto J, Kuulasmaa K. WHO MONICA Project: assessing CHD mortality and morbidity. Int J Epidemiol. 1989;18(suppl 1):S38–S45.
Tuomilehto J, Sarti C, Narva EV, Salmi K, Sivenius J, Kaarsalo E, Salomaa V, Torppa J. The FINMONICA Stroke Register. Community-based Stroke Registration and Analysis of Stroke Incidence in Finland, 1983-1985. Am J Epidemiol.. 1992;135:1259-1270.
Sarti C, Tuomilehto J, Sivenius J, Kaarsalo E, Narva EV, Salmi K, Salomaa V, Torppa J. Stroke mortality and case-fatality rates in three geographic areas of Finland from 1983 to 1986. Stroke.. 1993;24:1140-1147.
Sarti C, Tuomilehto J, Sivenius J, Kaarsalo E, Narva EV, Salmi K, Torppa J, Salomaa V. Declining trends in incidence, case-fatality and mortality of stroke in three geographic areas of Finland during 1983-1989: results from the FINMONICA Stroke Register. J Clin Epidemiol.. 1994;47:1259-1269.
Böthig S, for the WHO MONICA Project. WHO MONICA Project: objectives and design. Int J Epidemiol. 1989;18(suppl 1):S29–S37.
WHO MONICA Project, prepared by Asplund K, Tuomilehto J, Kuulasmaa K, Torppa J. Multinational Stroke Mortality Data at the Baseline of the WHO MONICA Project: Cerebral Vascular Disease 7. Amsterdam: Elsevier Science Publishers B.V. (Biomedical Division); 1989:167-170.
World Health Organization MONICA Project. MONICA Manual. Geneva, Switzerland: World Health Organization Cardiovascular Disease Unit; 1990.
Statistics Finland. Causes of Deaths: Years 1983-1992. Helsinki, Finland: Statistics Finland; 1993. Official Statistics of Finland. ISSN No. 0787-0132.
Sarti C, Tuomilehto J, Salomaa V, Sivenius J, Kaarsalo E, Narva EV, Salmi K, Torppa J. Epidemiology of subarachnoid hemorrhage in Finland during 1983 to 1985. Stroke.. 1991;22:848-853.
SAS Institute Inc. Users guide: Statistics. Version 6, 4th ed. Vol 2. Cary, North Carolina, 1989.
Kagan A, Popper J, Reed DM, MacLean CJ, Grove JS. Trends in stroke incidence and mortality in Hawaiian Japanese men. Stroke.. 1994;25:1170-1175.
Hong Y, Bots ML, Pan X, Hofman A, Grobbee DE, Chen H. Stroke incidence and mortality in rural and urban Shanghai from 1984 through 1991. Stroke.. 1994;25:1165-1169.
Feigin VL, Wiebers DO, Whisnant JP, O’Fallon WM. Stroke incidence and 30-day case-fatality rates in Novosibirsk, Russia, 1982 through 1992. Stroke.. 1995;26:924-929.
Sarti C on Behalf of the FINMONICA Stroke Register Group. Geographic variation in the incidence of non-fatal stroke in Finland: are the observed differences real? Stroke.. 1993;24:787-791.
Vartiainen E, Sarti C, Tuomilehto J, Kuulasmaa K. Do changes in cardiovascular risk factors explain changes in mortality from stroke in Finland? BMJ.. 1995;310:901-904.
Tuomilehto J, Pietinen P, Uusitalo U, Korhonen H, Nissinen A. Changes in sodium and potassium intake in Finland during the 1980s. In: Yamori Y, Strasser T, eds. New Horizons in Preventing Cardiovascular Disease. Amsterdam: Elsevier Science Publishers B.V.; 1989:229-240.
Salomaa V, Miettinen H, Kuulasmaa K, Niemelä M, Ketonen M, Vuorenmaa T, Lehto S, Palomäki P, Mähönen M, Immonen-Räihä P, Arstila M, Kaarsalo E, Mustaniemi H, Torppa J, Tuomilehto J, Puska P, Pyörälä K. Decline of coronary heart disease mortality in Finland during 1983 to 1992: roles of incidence, recurrence and case fatality. The FINMONICA MI Register Study. Circulation.. 1996;94:3130-3137.
Immonen-Räihä P, Arstila M, Tuomilehto J, Haikio M, Mononen A, Vuorenmaa T, Torppa J, Parvinen I. 21 year trends in incidence of myocardial infarction and mortality from coronary disease in middle-age. Eur Heart J.. 1996;17:1495-1502.