1www.eu osu eillance.o g
Resea ch a icles
Incidence and se op e alence o ula aemia in Finland,
1995 o 2013: egional epidemics wi h cyclic pa e n
H Rossow (heidi. ossow@helsinki. i)1,2,3, J Ollg en2, J Hy önen4, H Rissanen5, O Hui u6, H Hen onen7, M Kuusi2, O Vapalah i1,3
1. Depa men o Ve e ina y Biosciences, Facul y o Ve e ina y Medicine, Uni e si y o Helsinki, Helsinki, Finland
2. In ec ious Disease Con ol Uni , Depa men o In ec ious Diseases, Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland
3. Vi ology, Uni e si y o Helsinki and Helsinki Uni e si y Hospi al, Helsinki, Finland
4. Depa men o Medical Mic obiology and Immunology, Uni e si y o Tu ku, Tu ku, Finland
5. Heal h Moni o ing Uni , Depa men o Heal h, Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland
6. Suonenjoki Uni , Na u al Resou ces Ins i u e, Suonenjoki, Finland
7. Van aa Uni , Na u al Resou ces Ins i u e, Van aa, Finland
Ci a ion s yle o his a icle:
Rossow H, Ollg en J, Hy önen J, Rissanen H, Hui u O, Hen onen H, Kuusi M, Vapalah i O. Incidence and se op e alence o ula aemia in Finland, 1995 o 2013:
egional epidemics wi h cyclic pa e n. Eu o Su eill. 2015;20(33):pii=21209. A ailable online: h p://www.eu osu eillance.o g/ViewA icle.aspx?A icleId=21209
A icle submi ed on 19 Oc obe 2014 / published on 20 Augus 2015
We s udied he incidence o epo ed ula aemia by
yea and egion and he p e alence o an ibodies
agains F ancisella ula ensis in he adul gene al
popula ion in Finland. Mo eo e , we assessed he co -
ela ion be ween ole popula ion cycles and human
ula aemia ou b eaks. The se op e alence s udy made
use o se um samples om a na ionwide popula ion-
based heal h su ey (Heal h 2000). The samples o
1,045 andomly selec ed pe sons, ep esen a i e o
he Finnish popula ion in each egion, we e sc eened
wi h an enzyme-linked immunoso ben assay (ELISA)
o he p esence o IgG an ibodies agains F. ula en-
sis, and posi i e esul s we e u he con i med by
immunoblo ing. A se ological esponse o F. ula en-
sis was ound in 2% (95% con idence in e al: 1.1–3.5)
o he popula ion. Incidence and se op e alence we e
highes in he same a eas, and ole popula ion peaks
clea ly p eceded ula aemia ou b eaks one yea la e .
In oduc ion
Tula aemia is a zoono ic disease caused by he in a-
cellula bac e ium F ancisella ula ensis [1,2]. The dis-
ease is caused p ima ily by wo o ou subspecies: he
highly i ulen ype A s ain F. ula ensis subsp. ula-
ensis which is almos comple ely es ic ed o No h
Ame ica, and he less i ulen ype B s ain F. ula en-
sis subsp. hola c ica, which occu s in many egions o
he no he n hemisphe e, including Finland [1,2].
Recu en ou b eaks wi h hund eds o cases a e
epo ed in Finland and Sweden [3,4]; in o he Eu opean
coun ies, he disease is a e, bu many coun ies epo
spo adic ou b eaks [5-7]. Geog aphically, he disease
shows a ocal dis ibu ion [8]. The pa hogen is mos
likely o pe sis in he local en i onmen bu seems
o cause epidemics only when he ecological condi-
ions a e a ou able o an ac i e in ec ious cycle [9].
Thus, app op ia e ese oi s, ampli ie s, ec o s and
sui able clima ic condi ions a e needed o an ou b eak
o human ula aemia [9]. Tula aemia is ypically associ-
a ed wi h ou doo ac i i ies, and a me s and hun e s
a e a pa icula isk o in ec ion [3-7,10]. In Finland
and Sweden, he disease is ypically mosqui o- ans-
mi ed and mos cases a e epo ed du ing Augus and
Sep embe in connec ion wi h he occu ence o la e
summe mosqui o species [3,4], whe eas in No way,
he disease occu s in au umn and win e and he mos
common sou ce is d inking wa e con amina ed by
oden s [11]. Ou b eaks in No way and Koso o* ha e
been linked o high oden densi ies [11-14].
Ai bo ne ou b eaks, mainly associa ed wi h ac i i ies
ha can gene a e ae osols, such as a ming, ga dening
o hun ing, occu occasionally in all endemic coun ies
and a e e y local [5-7,15].
Human ula aemia ypically s a s wi h non-speci ic
in luenza-like symp oms [16]. O he clinical mani es-
a ions depend mainly on he ou e o ansmission,
and he disease se e i y depends on he in ec ing sub-
species and s ain [1,16]. A e in ec ion, an ibodies
agains F. ula ensis ise slowly bu a e de ec able o
se e al yea s [17]. He e we epo a s udy conduc ed in
Finland, aiming o de e mine he incidence o ula ae-
mia and he p e alence o F. ula ensis an ibodies in
he popula ion in 2000 and 2001, o compa e he se o-
p e alence a es wi h he numbe o epo ed cases o
he pe iod 1995 o 2013, and o assess o he same
pe iod he ole o ole popula ion cycles in he empo-
al and spa ial pa e n o human ula aemia ou b eaks.
Me hods
Na ional labo a o y-based su eillance
The Finnish na ional heal hca e sys em is o ganised
in 20 geog aphically and adminis a i ely dis inc
2www.eu osu eillance.o g
heal hca e dis ic s. Labo a o y-con i med ula aemia
has been a no i iable disease by he diagnosing labo a-
o y since 1995, and clinical mic obiology labo a o ies
epo cases di ec ly o he Na ional In ec ious Disease
Regis e (NIDR) which is main ained by he Na ional
Ins i u e o Heal h and Wel a e (THL). Diagnos ic c i-
e ia o epo ing include (i) isola ion o F. ula ensis
in a clinical specimen, (ii) a mo e han ou - old ise in
se um an ibody i e o a single an ibody i e o > 160
when using an agglu ina ion assay o (iii) he p esence
o speci ic IgM and IgG an ibodies in he se um when
an enzyme-linked immunoso ben assay (ELISA) is
used. Wi h each no i ica ion, he ollowing in o ma ion
is gi en: da e and ype o specimen, da e o bi h, sex,
place o ea men and place o esidence.
Human se um samples and backg ound heal h
in o ma ion
Se um samples we e collec ed in a mul idisciplina y
epidemiological heal h su ey, he Heal h 2000 S udy,
ca ied ou in he yea s 2000–01 in Finland [18].
De ailed s udy me hods ha e been desc ibed else-
whe e [18]. Se a om adul s we e collec ed in 80 di e -
en a eas co e ing mos o he coun y. Fo he cu en
s udy, se um samples om 1,045 andomly selec ed
pe sons we e included. Fo geog aphical co e age,
all 20 heal hca e dis ic s we e ep esen ed wi h 50
samples, excep o Cen al Os obo hnia whe e only
45 samples we e a ailable and he capi al dis ic o
Helsinki and Uusimaa, which was ep esen ed wi h
100 samples, e lec ing a union o wo o me ly sepa-
a e dis ic s. An ex ensi e heal h in e iew and heal h
examina ion had been done o all pa icipan s in he
o iginal s udy. In his s udy, we especially ocused ou
analyses on common symp oms and medical condi-
ions, li ing en i onmen , occupa ion, leisu e ac i i-
ies, physical condi ion, smoking and alcohol use and
demog aphic ac o s. The Heal h 2000 Su ey was
app o ed by he E hical Commi ee o Resea ch in
Epidemiology and Public Heal h a he Hospi al Dis ic
o Helsinki and Uusimaa.
Se a om se en pa ien s who had labo a o y-con-
i med ula aemia (diagnosed a he Depa men o
Medical Mic obiology and Immunology, Uni e si y o
Tu ku) one o 16 yea s be o e se um collec ion, we e
used as posi i e con ol se a.
Se ological es ing
ELISA
All se um samples we e sc eened o F. ula ensis an i-
bodies by ELISA as p e iously desc ibed [17], wi h whole
bac e ium lysa e p epa ed om F. ula ensis li e ac-
cine s ain (LVS) as he an igen. Se um samples we e
es ed a a dilu ion o 1:100. Abso bances (OD405) we e
measu ed wi h BEP III appa a us (Siemens Heal hca e
Diagnos ics P oduc s GmbH, E langen, Ge many).
Resul s we e exp essed as a bi a y enzyme-immuno-
so ben uni s (EIU) on he scale om ze o o 100 uni s.
Con en ional ecei e ope a ing cha ac e is ic ROC)
analysis o de e mine he cu -o o he ELISA could no
be pe o med because he ue ula aemia an ibody
s a us (de ined by a gold s anda d) o ou samples o
o he ELISA es in gene al was no known. This is why
a con i ma o y Wes e n blo was employed.
Wes e n blo
Wes e n blo ing was conduc ed as ea lie desc ibed
[19]. Posi i e se a showed a ypical lipopolysaccha ide
(LPS) band pa e n a a dilu ion o 1:100. Samples we e
sc eened om he highes EIU (83.7) in ELISA down o
EIU 26.8 (including 55 se osu ey samples and se en
posi i e con ol samples). The lowes ELISA EIU alue
ha was s ill posi i e in Wes e n Blo was 28.5. The
ollowing 10 samples we e WB nega i e and sc een-
ing was s opped due o low EIU alues o he emain-
ing samples. Only samples posi i e in bo h EIA and WB
we e conside ed posi i e.
Vole da a
Da a on small mammals we e collec ed biannually
in ca 30 loca ions a ound he coun y by he Na u al
Resou ces Ins i u e, as desc ibed by p e iously [20].
The apped mammals we e p edominan ly oles,
which domina e he oden and o he small mammal
auna in Finland [20,21]. Only ole popula ions show
cyclic luc ua ions. In addi ion, h ee wild mu ine (i.e.
non- ole oden ) species inhabi Finland, bu hey a e
es ic ed o he sou he n hal o he coun y, and usu-
ally occu a low densi ies wi h seasonal, bu no cyclic
dynamics. As such, he na ional moni o ing p og amme
o ole popula ions was planned o p oduce comp e-
hensi e da a on he spa io empo al dynamics o ole
popula ion luc ua ions a ound he whole coun y. By
and la ge, ole popula ions in Finland luc ua e p o-
nouncedly in cycles o h ee o ou yea s [20]. Fo he
pu poses o his s udy, he ole popula ion da a (den-
si y indices; oles pe 100 ap nigh s) we e quali a-
i ely classi ied in o h ee dis inc cyclic phases, each
co esponding o a yea : popula ion peak yea s, popu-
la ion decline yea s and popula ion inc ease yea s.
S a is ical analyses
Se ological su ey
S a is ical analyses we e pe o med using IBM SPSS
S a is ics e sion 22 and he p og amme R map ools
package was used o cons uc he maps. Ou da ase
was a subsample selec ed om he Heal h 2000 su -
ey main s udy by s a i ied simple sampling o size 50
samples pe hospi al dis ic , and he o iginal sampling
weigh s we e eweigh ed o ma ch he popula ion size
o a gi en hospi al dis ic a yea 2000. O he e y
comp ehensi e backg ound in o ma ion, we especially
ocused ou analyses on common symp oms and medi-
cal condi ions, li ing en i onmen , occupa ion, leisu e
ac i i ies, physical condi ion, smoking and alcohol
use and demog aphic ac o s. Uni a ia e chi-squa e
es s we e compu ed aking in o accoun he sam-
pling design, i s s a a and clus e s and he sampling
weigh s o he es o independence o he ca ego ical
3www.eu osu eillance.o g
a iables. The popula ion sizes and p e alences we e
es ima ed acco dingly, aking in o accoun sampling
design. Logis ic eg ession was used o model he ela-
ionship be ween se op e alence odds and explana o y
a iables. The incidence a io o no i ied/ o al ula ae-
mia cases was calcula ed assuming ha he popula-
ion a isk and he p e alence pool we e s a iona y.
Applying a o mula o a ime s a iona y si ua ion, he
(se o)p e alence odds o ula aemia p / (1 − p) , he
mean o al incidence I and he mean du a ion o he
se op e alence D a e ela ed as p / (1 − p) = I × D, wi h
D equal o he es ima ed a e age esidual li e ime
a e he mean age o acqui ing ula aemia.
Vole and su eillance da a analysis
The associa ion be ween ula aemia ou b eaks and
phase o ole cycle was analysed wi h Poisson eg es-
sion. Poisson model calcula ions we e done by hospi-
al dis ic . The e ec o ole cyclic phase ac o was
assumed independen o he yea . Possible o e dis-
pe sion was co ec ed by Pea son chi-squa ed scale
pa ame e me hod, and possible au oco ela ions
o pe son esiduals om he model we e checked by
au oco ela ion plo s (ACF plo s).
Resul s
Na ional labo a o y-based su eillance
F om 1995 o 2013, 5,086 no i ica ions o labo a o y-
con i med ula aemia cases we e epo ed o he NIDR.
The annual numbe o no i ied cases anged om 14
o 926. The a e age annual incidence was 5.1/100,000
popula ion and he highes incidence (18/100,000 pop-
ula ion) was eco ded du ing he yea o he majo epi-
demic in 2000 (Figu e 1).
Ra es we e ypically highes in he heal h dis ic s
o No he n and Sou he n Os obo hnia and Cen al
Finland. The mean age o he cases was 45 yea s
( ange: 0–93 yea s) and 55% we e male. The annual
a ia ion in epo ed cases in he h ee heal hca e
dis ic s wi h pa icula ly high incidence is shown in
Figu e 2. Typically, epidemics occu ed in di e en dis-
ic s e e y hi d o ou h yea .

Epidemics we e s ongly seasonal, wi h he majo i y
o cases diagnosed du ing summe and ea ly au umn
(Figu e 3).
Figu e 4 shows he incidence a es by heal hca e dis-
ic in 1995–2000, i e o six yea s be o e sampling o
he se op e alence s udy.
Human se op e alence s udy
Ou s udy sampling o 1,045 pe sons comp ised 46%
men and 54% women. Thei mean age was 53 yea s
( ange: 30–92 yea s) and o geog aphical co e age,
hey ep esen ed all 20 heal hca e dis ic s. We ound
16 posi i e samples (1.5%), which a e adjus men wi h
su ey weigh s ga e an es ima ed o e all F. ula ensis
an ibody p e alence o 2.0% (95% con idence in e al:
1.1–3.5) on popula ion le el. The dis ibu ion o EIU al-
ues in ou s udy and con ol samples is shown in Figu e
5; he posi i e con ol se a showed EIU alues be ween
64.3 and 93.0.
Fi e pa icipan s had an EIU alue > 50 in addi ion o
a ypical LPS band pa e n in he Wes e n Blo (Figu e
6). One o hem epo ed being hospi alised because o
ula aemia du ing he pe iod om 1995 o 2000 (p e-
cise ime and du a ion o hospi alisa ion no a ailable).
Figu e 1
Incidence a es o labo a o y-con i med ula aemia in ec ions, Finland, 1995–2013
0
2
4
6
8
10
12
14
16
18
20
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 20132012201120102009200820072006
Yea
Incidence a e / 100,000 popula ion
4www.eu osu eillance.o g
Figu e 2
Numbe o labo a o y-con i med ula aemia cases and iming o ole popula ion peaks, by dis ic , Finland, 1995–2013
((n = 3,011)
0
50
100
150
200
250
300
350
400
450
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 201320122011201020092008
Yea
Numbe o no i ica ionsNumbe o no i ica ionsNumbe o no i ica ions
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 201320122011201020092008
Yea
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 201320122011201020092008
Yea
Vole peak phase Human cases
A. No he n Os obo hnia
0
20
40
60
80
100
120
140
160
180
B. Sou he n Os obo hnia
0
50
100
150
200
250
C. Cen al Finland
B and C: The yea 1999 was no a ypical wide-sp ead peak yea , bu showed pa chy local peaks. Fo de ails see ex .
5www.eu osu eillance.o g
Geog aphically, he se op e alence was highes in
No he n Os obo hnia (Figu e 7).
No signi ican di e ences in age o sex dis ibu ion
we e ound be ween he se oposi i e and se onega-
i e g oup. The mean age o se oposi i e pe sons in
ou s udy was 55 yea s and 50% o hem we e males.
No single isk ac o was signi ican ly associa ed wi h
se oposi i i y. F. ula ensis se oposi i i y in gene al
was highe in pe sons wi h lowe educa ional le el
(likelihood a io: 0.048), bu no single occupa ion was
ela ed o se oposi i i y. The a io o no i ied s o al
(es ima ed based on he obse ed se op e alence) F.
ula ensis in ec ions was 1/10.5.
Vole cycles and hei associa ion wi h human
ula aemia
The empo al occu ence o ole peak yea s clea ly p e-
dic ed human ula aemia ou b eaks. Ou b eaks mos ly
occu ed du ing he yea s immedia ely a e he ole
peak yea s, i.e. in decline yea s (Figu e 2). The ula ae-
mia incidence in ole decline yea s was on a e age ca
six imes highe han in oden inc ease yea s (Table).
In 10 o he 20 hospi al dis ic s, he ole decline yea s
had a signi ican ly highe ula aemia incidence (inci-
dence a e a io > 1) han he non-decline yea s. In he
emaining dis ic s, ula aemia is a e and he impac
o ole cycles could no be obse ed ( he eg ession
model could no be i ed o hose pa icula hospi al
dis ic s).
Discussion
The epidemiology o ula aemia in Finland is cha ac-
e ised by ecu en egional and seasonal ou b eaks
occu ing in sho cycles o ypically h ee o ou
yea s [3,21]. In he long e m, he incidence is high-
es in No he n Os obo hnia as shown also in he
cu en s udy. Ou b eaks in ol ing hund eds o cases
occu e e y h ee o i e yea s in some a eas such
as No he n and Sou he n Os obo hnia and Cen al
Finland [3]. In o he a eas, ou b eaks o his magni-
ude a e a e. The majo i y o cases a e epo ed in
Augus and Sep embe , du ing o igh a e he la e
summe mosqui o season. I has been shown ha he
F. ula ensis ca iage a e in mosqui oes in Sweden
inc eases wi h declining mosqui o popula ions in la e
summe and ea ly au umn [22], which may explain he
high numbe o no i ica ions in Augus and Sep embe .
Also o he impo an mosqui o- ansmi ed diseases in
Finland (such as Pogos a disease caused by Sindbis
i us) a e ansmi ed pa icula ly by he la e summe
mosqui o species [23].
On he o he hand, he e is gene ally a ime lag o se -
e al weeks be ween onse o symp oms and labo a-
o y con i ma ion o ula aemia [3]. P esumably, mos
cases no i ied in Augus and Sep embe acqui ed he
in ec ion some weeks ea lie . In ou se ological su -
ey, we ound an o e all F. ula ensis an ibody p e a-
lence o 2% on popula ion le el a e adjus men wi h
su ey weigh s (1.5% among ou s udy pa icipan s).
Se op e alence was highes in No he n Os obo hnia,
which is in line wi h he numbe o no i ica ions o he
NIDR. The obse ed se op e alence was compa able
o esul s om Ge many whe e se op e alences up o
2.3% ha e been ound [24,25]. In Ma ha’s Vineya d,
Massachuse s, whe e only landscape s we e es ed,
9.1% o he s udied popula ion was se oposi i e o
F. ula ensis [10]. In u al Aze baijan [26] and I an [27],
se op e alences signi ican ly highe han in ou s udy
we e ound. In hose s udies, oden exposu e [26] and
hun ing [27] we e shown o clea ly inc ease he isk
o ula aemia. These a e well known isk ac o s o
ula aemia.
I was su p ising ha he F. ula ensis an ibody p e -
alence was so low in Finland he Eu opean Union
Membe S a e wi h he highes epo ed ula aemia
incidence [21,24]. On he o he hand, low incidence
bu ela i ely high se op e alence in o he coun ies
p obably indica es unde diagnosing. This could be
explained by he di e en clinical pic u e: in Finland,
ula aemia is mainly mosqui o-bo ne and mani es s
as he ulce oglandula o m which is easy o diagnose
based on he ypical symp oms. In cen al and sou h-
e n Eu ope howe e , he mos common is he yphoidal
o m which is e y di icul o diagnose because simi-
la symp oms can ha e o he causes and because he
awa eness among clinicians is low. Also in Finland, a
ce ain p opo ion o in ec ions a e no no i ied, which
is no su p ising, especially when aking in o accoun
he challenging labo a o y diagnosis o ula aemia.
Gene al p ac i ione s in endemic a eas p obably o en
ea he disease based on a clinical diagnosis only and
diagnos ic labo a o y es s a e no eques ed. Some
pa ien s may also eco e a e an in luenza-like eb ile
illness wi hou seeking medical a en ion [10]. On he
o he hand, he low p e alence o an ibodies agains
F. ula ensis indica es ha ula aemia in ec ion is no
e y common and is in mos cases associa ed wi h
dis inc acu e clinical symp oms [2,16]. Based on he
Figu e 3
Cumula i e numbe o labo a o y-con i med ula aemia
cases by mon h, Finland, 1995–2013 (n = 5,086)
0
500
1,000
1,500
2,000
2,500
Jan Feb Ma Ap May Jun Jul Aug Sep Oc No Dec
Mon h
Numbe o cases

6www.eu osu eillance.o g
no i ica ions and se ocop e alence, we es ima e ha
ca one in 10 (9.5%) F. ula ensis in ec ions a e no i ied.
We ha e shown a co ela ion be ween ole popula ion
dynamics and human ula aemia ou b eaks. Human
ula aemia ou b eaks ypically occu du ing he ole
decline phase a yea a e he ole popula ion peak. In
2000, human ula aemia ou b eaks occu ed h ough-
ou Finland, including Sou he n Os obo hnia and
Cen al Finland, al hough cha ac e is ics o a wide-
sp ead ole peak in he p eceding yea we e only me in
he no he n pa o he coun y. Vole peaks occu ed in
a pa chy ashion in 1999, and pa icula ly in Sou he n
Os obo hnia, ou long- e m moni o ing si es ep e-
sen ing he la ge hospi al dis ic s may no ha e coin-
cided wi h he localised ole peak a eas [28]. We did
no include 1999 as a peak yea in ou s a is ics, bu
Figu e 4
Incidence o labo a o y-con i med F ancisella ula ensis in ec ions epo ed o he Na ional In ec ious Disease Regis e , by
heal hca e dis ic , Finland 1995–2000
Russia
Sweden
3
2
1
No way
0.0 - 2.0
A e age incidence / 100,000
>2.0 - 5.0
>5.0 - 10.0
>10.0 - 20.0
>20.0 - 40.0
Incidences pe 100,000 popula ion.
Heal hca e dis ic s: 1. No he n Os obo hnia; 2. Cen al Finland; 3. Sou he n Os obo hnia.
7www.eu osu eillance.o g
he associa ion be ween he cyclic phase and human
ula aemia was ne e heless signi ican .
We ha e p e iously shown ha oles can se e as
ampli ica ion hos s o F. ula ensis [29]. La ge ole
popula ions allow he bac e ia o eplica e in ensi ely.
Du ing a peak phase, li e oden s can shed he bac-
e ia in o he en i onmen , and in he decline phase,
in ec ed dead oden s elease la ge amoun s o F. ula-
ensis in o he en i onmen , including b eeding si es
o mosqui oes [29]. Mosqui o la ae can ake up he
bac e ia [22] and once hey become adul s, sp ead he
bac e ia o suscep ible hos s ha ac as local ampli-
ie s. This p o ides a likely explana ion o he sug-
ges ed associa ion be ween ole cycles and human
ula aemia incidence [29]. Howe e , he a ia ion in he
magni ude and he locali y o human ula aemia ou -
b eaks wa an u he analyses. High ole densi y is
p obably one, bu no he only p e equisi e o ula ae-
mia ou b eaks. The local ecological ac o s c ucial o
disease ou b eaks a e s ill no known e y well. I has
been shown ha he p esence o ce ain aqua ic amoe-
bae enhances he mul iplica ion o F. ula ensis and
ha i s in ec ion p ocess in amoebae esembles ha
in mac ophages [30]. Possibly ole ca casses con ami-
na e na u al wa e s, amoebae suppo he local pe sis-
ence o F. ula ensis in hese wa e s, mosqui o la ae
eed on hese p o ozoa and hus ge in ec ed. Wea he
condi ions in luence he amoun o mosqui oes, which
impac s on he ansmission o humans and hus he
ampli ude o he ou b eak [31].
Conclusion
In summa y, human ula aemia in Finland is ocal, and
mos o he cases occu in a ew dis ic s. This wa an s
landscape ecological analyses [8,32]. The answe may
no be simple because he endemic p o inces di e
conside ably in opog aphy. The seasonali y o ula ae-
mia occu ence s ongly indica es a majo ole o
mosqui oes in disease sp ead. The mul iannual cyclic
pa e n o he epidemics is associa ed wi h ole den-
si y cycles, wi h ole peak yea s p eceding epidemic
yea s. The in e ac ions be ween oles, mosqui oes and
F. ula ensis need s ill u he s udies, as well as he
disc epancy be ween he ela i ely low se op e alence
and conside able incidence o ula aemia.
Figu e 5
Dis ibu ion o ELISA esul s, ula aemia se op e alence
s udy, Finland, 2000–01 (n = 1,045)
ELISA: enzyme-linked immunoso ben assay; EIU:
enzyme-immunoso ben uni s.
Blue ba s: s udy samples; g een ba s: posi i e con ol samples.
300
200
100
Numbe o samples
Numbe o samples (con ol)
00
2
4
6
8
10
0 20 40 60 80 100
ELISA, EIU
Figu e 6
Wes e n blo o ep esen a i e posi i e and nega i e se a,
ula aemia se op e alence s udy, Finland, 2000–01
Lanes 1, 2 and 5: samples conside ed F ansicella ula ensis
an ibody-posi i e acco ding o a ypical ladde pa e n; lanes
3 and 4: nega i e samples om he se osu ey panel; lane 6: a
sample conside ed as an unspeci ic eac ion. M: size ma ke .
8www.eu osu eillance.o g
Table
Incidence o ula aemia in di e en hospi al dis ic s and i s ela ion o oden cycles, Finland, 1995–2013
Hospi al dis ic Vole cycle phase Cases (n) Popula ion Incidence IRR 95% Wald CI o IRR p alue
Sou he n Ka elia non decline yea s 61,417,302 0.42 3.14 0.98 10.01 0.05
decline yea s 12 902,885 1.33
Sou he n Os obo hnia non decline yea s 111 2,164,737 5.13 7.37 2.83 19.22 < 0.001
decline yea s 522 1,381,053 37.80
Sou he n Sa o non decline yea s 15 1,149,203 1.31 1.46 0.47 4.53 0.51
decline yea s 14 732,287 1.91
Helsinki and Uusimaa non decline yea s 100 15,789,947 0.63 4.29 2.15 8.55 < 0.001
decline yea s 272 10,020,284 2.71
Eas e n Sa o non decline yea s 2702,300 0.28 18.08 0.61 536.11 0.09
decline yea s 23 446,602 5.15
Kainuu non decline yea s 31,245,547 0.24 3.28 0.63 17.01 0.16
decline yea s 2253,206 0.79
Kan a-Häme non decline yea s 91,856,789 0.48 5.07 1.90 13.49 < 0.001
decline yea s 29 1,181,058 2.46
Cen al Os obo hnia non decline yea s 204 931,902 21.89 1.22 0.16 9.36 0.85
decline yea s 124 462,955 26.78
Cen al Finland non decline yea s 233 2,934,836 7.94 4.89 1.95 12.26 < 0.001
decline yea s 725 1,865,762 38.86
Kymenlaakso non decline yea s 78 1,991,154 3.92 2.58 0.86 7.69 0.09
decline yea s 128 1,267,302 10.10
Lapland non decline yea s 15 2,074,033 0.72 2.34 0.41 13.32 0.34
decline yea s 2118,189 1.69
Wes e n Os obo hnia non decline yea s 20 1,149,854 1.74 0.00 0.00 0.00 < 0.001
decline yea s 064,655 0.00
Päijänne Ta as ia non decline yea s 17 2,293,263 0.74 2.03 0.48 8.63 0.34
decline yea s 22 1,459,719 1.51
Pi kanmaa non decline yea s 104 5,098,236 2.04 3.60 0.88 14.84 0.08
decline yea s 238 3,237,195 7.35
No he n Ka elia non decline yea s 11 1,885,503 0.58 0.71 0.19 2.70 0.62
decline yea s 51,203,895 0.42
No he n Os obo hnia non decline yea s 408 4,931,836 8.27 6.75 3.31 13.75 < 0.001
decline yea s 1,060 1,898,901 55.82
No he n Sa o non decline yea s 20 2,764,993 0.72 4.95 2.04 12.00 < 0.001
decline yea s 63 1,760,755 3.58
Sa akun a non decline yea s 54 2,519,350 2.14 10.77 2.75 42.20 < 0.001
decline yea s 371 1,607,604 23.08
Vaasa non decline yea s 29 1,996,883 1.45 7.58 2.74 20.99 < 0.001
decline yea s 110 999,243 11.01
Sou h-wes Finland non decline yea s 11 5,038,620 0.22 6.87 2.63 17.95 < 0.001
decline yea s 48 3,201,242 1.50
Whole coun y non decline yea s 1,450 59,936,288 2.42 4.57 4.31 4.86 < 0.001
decline yea s 3,770 34,064,792 11.07
CI: con idence in e al; IRR: incidence a e a io.
9www.eu osu eillance.o g
*No e
This designa ion is wi hou p ejudice o posi ions on s a us,
and is in line wi h Uni ed Na ions Secu i y Council Resolu ion
1244/99 and he In e na ional Cou o Jus ice Opinion on he
Koso o decla a ion o independence.
Acknowledgemen s
This s udy was pa ially unded by he Hospi al dis ic o
Helsinki and Uusimaa (EVO THY20113 o OV) and he Paulo
Founda ion (g an o HRo in 2013). Anna Ka onen is hanked
o pe o ming he ELISA analyses. The oden moni o ing
has been pa ially suppo ed by he Eu opean p og ams
GOCE-CT-2003-010284 EDEN and FP7-261504 EDENex , and
he pape is ca alogued by he EDENex S ee ing Commi ee
as EDENex 000 (h p://www.edenex .eu), as well by Kone
Founda ion (Finland).
Con lic o in e es
None decla ed.
Figu e 7
Es ima ed se op e alence (p e alence o Wes e n blo posi i e samples), by heal hca e dis ic , ula aemia se op e alence
s udy, Finland, 2000–01 (n = 1,045)
Russia
Sweden
3
2
1
No way
0.0 - 1.0
Se o e alance %
>1.0 - 2.0
>2.0 - 4.0
>4.0 - 6.0
>6.0 - 8.0
Heal hca e dis ic s: 1. No he n Os obo hnia; 2. Cen al Finland; 3. Sou he n Os obo hnia.