Multiple sclerosis in Finland: incidence trends and
differences in relapsing remitting and primary
progressive disease courses
M-L Sumelahti, P J Tienari, M Hakama, J Wikström
J Neurol Neurosurg Psychiatry 2003;74:25–28
Objective: To compare the secular trends and geographical differences in the incidence of relapsing-
remitting (RRMS) and primary progressive multiple sclerosis (PPMS) in Finland, and to draw inferences
about aetiological differences between the two forms of the disease.
Methods: New multiple sclerosis cases in southern Uusimaa and the western districts Vaasa and
Seinäjoki of Finland in 1979–1993 were verified from hospital records and classified into RRMS and
PPMS. Patients met the Poser criteria for definite multiple sclerosis or otherwise satisfied the criteria for
PPMS. Disease course was categorised by the same neurologist. Crude and age adjusted incidence in
1979–1993 was estimated.
Results: During 1979–1993 the age adjusted incidence was 5.1 per 100 000 person-years in Uusi-
maa, 5.2 in Vaasa, and 11.6 in Seinäjoki. The rates in Uusimaa remained stable, while a decrease
occurred in Vaasa and an increase in Seinäjoki. Between 1979–86 and 1987–93 the incidence of
PPMS increased in Seinäjoki from 2.6 to 3.7 per 105and decreased in Vaasa from 1.9 to 0.2 per 105;
the trends were similar for RRMS.
Conclusions: There are significant differences in secular trends for multiple sclerosis incidence in Fin-
land by geographical area, but these are similar for PPMS and RRMS. The recent changes point to
locally acting environmental factors. The parallel incidence trends for RRMS and PPMS suggest similar
environmental triggers for the two clinical presentations of multiple sclerosis.
(RRMS), which may later convert to a progressive phase (sec-
ondary progressive multiple sclerosis, SPMS). Approximately
10–30% of patients have a progressive course from the onset
(primary progressive multiple sclerosis, PPMS),2–6with a
worse prognosis than for RRMS.7In addition to differences in
the natural course,3 6there is increasing support for patho-
neuroimaging8and neuropathological observations.9There is
also some evidence that the HLA class II associations may be
different.10–12Given these differences, RRMS and PPMS may
also be aetiologically distinct, and this should be reflected in
the circumstances of their occurrence.
Significant geographical and temporal differences and an
increasing occurrence are common observations in multiple
sclerosis epidemiology in high risk regions,13and have been
reported in Finland.14–19Finland belongs to a high risk region
for multiple sclerosis,13with prevalences of from 100 to 200 per
100 000 population in different areas.18The western district
Seinäjoki has for a long time stood out as a high risk focus of
the disease,15 16with a prevalence of 200/105in 1993 and an
incidence of 12 per 100 000 person-years in 1979 to 1993.17 18
This is among the highest ever reported and is twice as high as
the rates in two other areas in Finland, the western region of
Vaasa and southern Uusimaa.A familial clustering of multiple
sclerosis cases has been observed in Seinäjoki.19
The similar high multiple sclerosis incidences of 8/105and
10/105that were present in Seinäjoki and Vaasa in 1979 to
198317showed a marked change between 1984 and 1993,with
a significant increase in Seinäjoki and a decrease in Vaasa. No
change was observed in southern Uusimaa. In the present
study we compare the occurrence of multiple sclerosis broken
down by type (RRMS and PPMS), and analyse the temporal
he clinical spectrum of disease progression in multiple
sclerosis is heterogeneous.1 2In most series about 70–90%
of patients have a relapsing-remitting disease course
variations in incidence and the clinical and demographic
characteristics of the two forms of multiple sclerosis in
Finland. This allowed us to investigate whether there are geo-
graphical and secular differences in the two disease courses
which would point to aetiological differences. Such a hypoth-
esis is feasible because there are several clinical and
pathological differences between primary progressive and
relapsing remitting multiple sclerosis.3 6 8 9
Demographic data and population
Finland is situated in northern Europe between latitudes 60°
and 70°. The total population was 5.1 million in 1993. The
study districts of Uusimaa, Vaasa, and Seinäjoki are shown in
fig 1. At the end of 1993 the populations were 1 277 932,
197 042, and 179 079, respectively, in the three districts. The
annual increase in the Finnish population has been 0.4% from
1979. The population structure was affected by the migration
of young adults into Uusimaa. However, the largest age group
continues to be 40–49 years in all districts.
The western district is divided into two health care districts,
where two central hospitals provide the neurological services.
CSF and evoked potential examinations are equally available,
but MRI has only been available in Seinäjoki since 1993. In
southern Uusimaa the diagnostic facilities in the 1970s and
1980s were better; however, the situation became equal in the
three areas during the 1990s. In Finland, all multiple sclerosis
diagnoses from 1960s have been made by a neurologist in the
neurological units of university or central hospitals, following
the same diagnostic principles.
For the purpose of this study, patients with multiple sclero-
were ascertained from hospital registers by the diagnoses
multiple sclerosis or morbus demyelinans (340 and 341 in the
See end of article for
Dr M-L Sumelahti, School of
Public Health, PO Box
607, FIN-33101 Tampere,
Received 4 February 2002
In revised form
23 September 2002
Accepted 26 September
International Classification of Diseases, versions 8 and 9). Patients
were re-evaluated to meet the criteria of Poser20or otherwise
they met the criteria for PPMS.21Disease course was classified
retrospectively (by M-LS) into RRMS and PPMS groups.22In
our study, PPMS was defined as disease that showed steady
symptoms.2 5 21The RRMS group included cases with second-
ary progression after an initial relapsing-remitting course.
year from the primary
The calculations were based on definite cases of multiple
sclerosis.20Incidence was calculated from 1 January 1979 to 31
December 1993 per 100 000 person-years in age groups from
10 to 69 years23and in calendar time intervals.The significance
of the trend was assessed by classifying the total time period
into two subgroups only (1979–1986 and 1987–1993) to
reduce random variation caused by small numbers.
As the crude rates may be misleading when comparing two
calendar periods, the rates were standardised for age by an
indirect method.24The standard rates in each district were
those of 1979–1986 in each age group.The resulting standard-
ised rate (SIR) is the ratio of the observed cases in 1987–1993
to those expected if the age specific rates had been those of
1979–1986. Confidence intervals (CI) were calculated assum-
ing that the observed number of cases followed a Poisson dis-
In 1979–1993 the total number of cases was 1066, 828 (78%)
in the RRMS group and 238 (22%) in the PPMS group. The
number of cases was 736 in Uusimaa,240 in Seinäjoki,and 90
in Vaasa. The clinical and pathological characteristics of the
material are presented in table 1. CSF intrathecal IgG synthe-
sis, which was found in over 90% of the cases tested, showed
no significant differences between PPMS and RRMS groups or
between the districts. A lower female to male ratio (F/M) was
observed among PPMS cases in each district (1.2–2.0) than
among RRMS cases (1.8–2.7). The F/M ratio was especially
low for PPMS in Seinäjoki (1.2), as was the ratio for all cases
(1.6). Mean age at diagnosis was greater for PPMS (40.5
years) than for RRMS (37.5 years), and the age at diagnosis
was generally greater in the western districts of Vaasa and
Seinäjoki than in Uusimaa. At onset there was a higher
proportion of motor symptoms in Seinäjoki (30%) than in the
other districts, and the ratios of motor symptoms to other
symptoms by disease course were higher for PPMS in Seinäoki
(0.9) and lower for RRMS in Uusimaa (0.1).
The incidences of total multiple sclerosis,RRMS,and PPMS
were similar in Uusimaa and Vaasa, but twofold higher in
Seinäjoki. The difference in incidence between the RRMS and
PPMS groups was somewhat larger (4.6) in Seinäjoki and
about the same in Uusimaa and Vaasa (3.1 and 2.8) (table 2).
Trends by five year age groups were stable in Uusimaa but
showed an increase in Seinäjoki and a decrease in Vaasa. This
trend was similar for RRMS and PPMS (fig 2).Owing to small
numbers,especially in Vaasa district (n = 90),the time period
1979–1993 was divided into two periods only (1979–1986 and
1987–1993). The incidence of PPMS increased in Seinäjoki
from 2.6 to 3.7 per 105person-years and decreased in Vaasa
from 1.9 to 0.2. The trends were similar for RRMS. The
decrease in total and PPMS incidences in Vaasa was
statistically significant, but not the decrease in RRMS.
Similarly,the increase in Seinäjoki was statistically significant
for total multiple sclerosis and PPMS but not for RRMS
Over successive studies from the 1960s, Finland has been
shown to belong to a high risk region for multiple sclerosis.13
The early observation of a high risk in the western district of
Seinäjoki14became more evident in the early 1990s, when an
incidence of 13/105person-years was four times higher than
the rates in neighbouring Vaasa (3/105) and southern Uusimaa
(5/105).17The rate in Seinäjokiwas among the highestreported
worldwide.25 26In the present study we show that the secular
trend in 1979 to 1993 was increasing in Seinäjoki, decreasing
(Uusimaa in the south and Vaasa and Seinäjoki in the west).
Map of Finland showing the hospital districts under study
patients with multiple sclerosis diagnosed in
1979–1993 in Uusimaa, Vaasa, and Seinäjoki
Clinical and pathological characteristics of
CSF IgG positivity (% of tested patients)
Female to male ratio
Mean age at diagnosis (years)
Symptoms at onset (%)
Ratio of motor to other symptoms
*Cerebellar, bladder, autonomic.
PPMS, primary progressive multiple sclerosis; RRMS,
relapsing-remitting multiple sclerosis.
26 Sumelahti, Tienari, Hakama, et al
in Vaasa, and stable in Uusimaa for the disease as a whole, as
well as for the PPMS and RRMS subtypes.
The populations of western and southern districts of
Finland differ economically as well as in their migration pat-
tern, immigration being greatest in the urbanised Uusimaa
but stable in the more rural populations of Vaasa and
Seinäjoki. Owing to the longstanding isolation of Finland and
a low migration rate within the country until recently, large
areas have remained genetically isolated. The population den-
sity in the country as a whole is low and even now the growth
of the gene pool is restricted.27 28This is shown in the HLA dis-
tribution,as allelic differences in HLA can still be found across
the country.Variations in the frequencies of A9,B35,and Cw4
are common in Finland.29
Medical practice in Finland is by and large homogeneous.
Neurological facilities in the Finnish health care system are
provided in the local or central hospitals of each health care
district. Thus multiple sclerosis is diagnosed almost exclu-
sively at the neurological units of these hospitals. Populations
under study are reasonably large, representing an industrial-
ised population in Uusimaa and rural populations in Vaasa
and Seinäjoki. To improve the comparability between the dis-
tricts, the time point of definite diagnosis was classified retro-
spectively,and the diagnostic tests were recorded.The diagno-
sis was mainly clinical in over 80% of the cases.17Classification
of disease course was done by the same neurologist (M-LS)
which should reduce potential inter-rater classification
bias.4 22Because of the commonly recognised classification
problems in multiple sclerosis,30 31we decided to use only two
disease course categories, relapsing remitting and primary
progressive, the former category including patients who later
converted into the secondary progressive form.
Our population sample showing different incidence trends
in the three districts provided an opportunity to test the
hypothesis that the RRMS and PPMS subtypes could have a
different aetiology32in stable and homogeneous study popula-
tions. The fact that trends in the incidence of the two disease
subtypes moved in parallel in three districts in which there
were diverging trends suggests that the environmental
triggers for these two presentations of multiple sclerosis are
similar. Our results thus contrast with earlier findings in
western Norway,33where the incidence of PPMS remained
stable while there was an increasing incidence of RRMS and
progressive-relapsing multiple sclerosis during 1953 to 1982.
However, these data were based on a much smaller number of
patients, and the incidence trends were later shown to be
similar for both disease courses.34The fluctuating trends typi-
cal in high risk regions for multiple sclerosis were also
observed in our study and were shown to be similar for both
disease courses. Such fluctuations can most readily be
accounted for by environmental exposures.
The age adjusted incidence of PPMS in 1987–1993 was
remarkably high in Seinäjoki (3.7/105) compared with
Uusimaa (1.0/105) and Vaasa (0.2/105), indicating a long term
high risk for PPMS in Seinäjoki. To our knowledge such large
regional differences in PPMS occurrence have not been found
before within a single population sample.4 5We do not believe
this finding is hampered by diagnostic misclassification6(for
example, chronic spastic paraparesis), and this interpretation
is supported by the 94% positive CSF IgG findings among the
96% of PPMS cases tested in Seinäjoki. The distribution of
and standardised incidence ratios by calendar time, district, and disease course, 1979 to 1993
Number of cases of multiple sclerosis, incidence per 100 000 person-years in the age group 10 to 69 years,
1979 to 19861987 to 1993 1979 to 1993
DistrictNI SIRNI SIRCINI CI
0.9 to 1.2
0.8 to 1.3
0.9 to 1.1
3.8 to 4.5
0.8 to 1.2
4.8 to 5.5
0.5 to 1.1
0.02 to 0.4
0.4 to 0.8
3.0 to 4.9
0.7 to 1.7
4.1 to 6.3
0.9 to 1.4
1.02 to 1.9
1.03 to 1.5
6.9 to 9.3
2.7 to 4.3
10.1 to 13.1
CI, confidence interval; I, incidence; N, number of cases; PPMS, primary progressive multiple sclerosis; RRMS, relapsing-remitting multiple sclerosis; SIR,
standardised incidence ratio.
disease course of all multiple sclerosis cases, relapsing-remitting
multiple sclerosis (RRMS), and primary progressive multiple sclerosis
(PPMS) in Seinäjoki, Vaasa, and Uusimaa in 1979–1993.
Incidence trends per 100 000 person-years in the
Multiple sclerosis incidence trends in Finland 27
cases in our incidence data—78% in the RRMS group and 22%
in the PPMS group—follows the common pattern,5as do the
demographic features: PPMS is commonly observed among
men and is associated with a later age of onset and the likeli-
hood of initial motor symptoms.3 6 8Given the recent increase
in incidence among men in Seinäjoki, and a lower F/M ratio
compared with other districts,17it would be reasonable to sup-
pose that PPMS risk is associated with male sex in Seinäjoki.
This was not,however,the case,as the risk for PPMS was high
for both sexes (data not shown).
The causes of a locally high and increasing risk of multiple
sclerosis, an overall high male risk,17a high risk for both
disease types in Seinäjoki, and a contrasting trend in Vaasa
remain unexplained. In spite of the language difference
between the partly Swedish speaking Vaasa and Finnish
speaking Seinäjoki, the genetic background is similar.27In the
multiple sclerosis population, cases have been characterised
for HLA in the district of Seinäjoki and have been found to
have increased frequencies for B7, B12, and DR2 among both
patients and their healthy relatives.35Families were later stud-
ied for the myelin basic protein (MBP) gene on chromosome
18, a candidate gene in multiple sclerosis. Genetic linkage and
association analyses suggested that a genetic predisposition to
multiple sclerosis is closely linked to the MBP gene in this
population.36In spite of these observations, our findings of a
sustained increase in the incidence of the disease point to
environmental factors, as genetic change in populations is
slow. The environmental causes are generally suspected to be
of viral origin37but remain largely unknown at present.
The large regional differences in the incidence of multiple
sclerosis in Finland in 1989–1993 result from diverging trends
in incidence that are largely parallel for both types of disease
course. In the originally high risk area of Seinäjoki the
incidence of both types of course was still increasing from the
late 1970s to the early 1990s. In Uusimaa, the incidence of
both types has remained stable. In Vaasa, an intermediate
incidence in 1979–1983 decreased to a low levelforboth PPMS
and RRMS. This finding is of aetiological importance, as diag-
nostic practices have remained the same. The sharply diverg-
ing incidence trends in Seinäjoki and Vaasa point towards
recent changes in environmental factors,and the parallel inci-
dence trends for PPMS and RRMS in all three districts suggest
that there are similar environmental triggers for both clinical
types of multiple sclerosis.
This study was supported by the Sigrid Juselius Foundation, Helsinki
University Central Hospital,the University of Tampere,and the Minis-
try of Education in Finland through a doctoral student position in
Doctoral Programmes in Public Health (to M-LS).
M-L Sumelahti, M Hakama, School of Public Health, University of
Tampere, Tampere, Finland
J Tienari, J Wikström, Department of Neurology, University of Helsinki
Hospital, Helsinki, Finland
Competing interests: none declared
1 Charcot JM. Lectures on the disease of the nervous system. Translated by
Sigerson G. London: The New Sydenham Society, 1877:157–222.
2 McAlpine D, Compston ND, Lumsden CE. Multiple sclerosis. Edinburgh:
3 Confavreux C, Aimard G, Devic M. Course and prognosis of multiple
sclerosis assessed by the computerized data processing of 349 patients.
4 Minderhoud J, van der Hoeven J, Prange A. Course and prognosis of
chronic progressive multiple sclerosis. Results of an epidemiological
study. Acta Neurol Scand 1988;78:10–15.
5 Weinshenker BG, Bass B, Rice GP, et al. The natural history of multiple
sclerosis: a geographically based study. I. Clinical course and disability.
6 Thompson A, Polman C, Miller D, et al. Primary progressive multiple
sclerosis. Brain 1997;120:1085–96.
7 Phadke J. Clinical aspects of multiple sclerosis in north-east Scotland
with particular reference to its course and prognosis. Brain
8 Thompson A, Kermode A, MacManus D, et al. Patterns of disease
activity in multiple sclerosis: clinical and magnetic resonance study. BMJ
9 Revesz T, Kidd D, Thompson A, et al. A comparison of the pathology of
primary and secondary progressive multiple sclerosis. Brain
10 Olerup O, Hillert J, Fredrikson S, et al. Primarily chronic progressive and
relapsing/remitting multiple sclerosis: two immunogenetically distinct
disease entities. Proc Natl Acad Sci USA 1989;86:7113–17.
11 Masterman T, Ligers A, Olsson T, et al. HLA-DR15 is associated with
lower age at onset in multiple sclerosis. Ann Neurol 2000;48:211–19.
12 Weatherby S, Thomson W, Pepper L, et al. HLA-DRB1 and disease
outcome in multiple sclerosis. J Neurol 2001;248:304–10.
13 Kurtzke J. Geography in multiple sclerosis. J Neurol 1977;215:1–26.
14 Rinne U, Panelius M, Kivalo E, et al. Distribution of multiple sclerosis in
Finland with special reference to some geological factors. Acta Neurol
15 Wikström J, Palo J. Studies on the clustering of multiple sclerosis in
Finland. I. Comparison between the domiciles and places of birth in
selected subpopulations. Acta Neurol Scand 1975;51:85–98.
16 Kinnunen E, Wikström J, Porras J, et al. The epidemiology of multiple
sclerosis in Finland: Increase of prevalence and stability of foci in
high-risk areas. Acta Neurol Scand 1983;67:255–62.
17 Sumelahti M-L, Tienari P, Wikström J, et al. Regional and temporal
variation in the incidence of multiple sclerosis in Finland in 1979–1993.
18 Sumelahti M-L, Tienari P, Wikström J, et al. Increasing prevalence of
multiple sclerosis in Finland. Acta Neurol Scand 2001;103:1–6.
19 Wikström J. Studies on clustering of multiple sclerosis in Finland. II.
Microepidemiology in one high-risk county with special reference to
familial cases. Acta Neurol Scand 1975;51:173–83.
20 Poser CM, Paty DW, Scheinberg L, et al. New diagnostic criteria for
multiple sclerosis: guidelines for research protocols. Ann Neurol
21 Lublin FD, Reingold SC. Defining the clinical course of multiple sclerosis:
results of an international survey. Neurology 1996;46:907–11.
22 Kremenchutzky M, Cottrell D, Rice G, et al. The natural history of
multiple sclerosis: a geographically based study 7. Progressive–relapsing
and relapsing-progressive multiple sclerosis: a re-evaluation. Brain
23 Noseworthy J, Paty D, Wonnacot T, et al. Multiple sclerosis after age
50. Neurology 1983;33:1537–44.
24 Armitage P, Berry G. Standardization. In: Armitage P, Berry G, eds.
Statistical methods in medical research. Cambridge: Blackwell Science,
25 Robertson N, Compston DAS. Surveying multiple sclerosis in the United
Kingdom. J Neurol Neurosurg Psychiatry 1995;58:2–6.
26 Warren S, Warren K. Prevalence, incidence and characteristics of
multiple sclerosis in Westlock County, Alberta, Canada. Neurology
27 Nevanlinna HR. Genetic markers in Finland. Haematologia
28 de la Chapelle A. Disease gene mapping in isolated human
populations: the example of Finland. J Med Genet 1993;30:857–65.
29 Sirén M, Sareneva H, Lokki M, et al. Unique HLA antigen frequencies in
the Finnish population. Tissue Antigens 1996;48:703–7.
30 Khurram B, Whitaker N. Clinical and laboratory features of primary
progressive and secondary progressive MS. Neurology
31 McDonnell G, Hawkins S. Clinical study of primary progressive multiple
sclerosis in Northern Ireland, UK. J Neurol Neurosurg Psychiatry
32 Larsen J, Kvale G, Riise T, et al. Multiple sclerosis – more than one
disease. Acta Neurol Scand 1985;72:145–50.
33 Larsen JP, Aarli H, Nyland H, et al. Western Norway, a high-risk area
for multiple sclerosis: a prevalence/incidence study in the county of
Hordaland. Neurology 1984;34:1202–7.
34 Grønning M, Riise T, Kvale G, et al. Incidence of multiple sclerosis in
Hordaland, Western Norway: a fluctuating pattern. Neuroepidemiology
35 Kinnunen E, Koskimies S, Lagerstedt A, et al. Histocompatibility antigens
in familial multiple sclerosis in a high-risk area of the disease. J Neurol
Sci 1984; 65:147–55.
36 Tienari PJ, Wikström J, Sajantila A, et al. Genetic susceptibility to
multiple sclerosis linked to myelin basic protein gene. Lancet 1992;
37 Challoner P, Smith K, Parker J, et al. Plaque-associated expression of
human herpesvirus 6 in multiple sclerosis. Proc Natl Acad Sci USA 1995;
28Sumelahti, Tienari, Hakama, et al