Resea ch A icle
Sc eening Sleep Diso de ed B ea hing in S oke Uni
Ki si Väy ynen,1,2 Ka i Ko elainen,2Heikki Numminen,2Ka ja Mie inen,3Anna Keso,3
Mi ja Tenhunen,1Heini Huh ala,4and Sa i-Leena Himanen1,3
1Depa men o Clinical Neu ophysiology, Medical Imaging Cen e and Hospi al Pha macy, Tampe e Uni e si y Hospi al,
33521 Tampe e, Finland
2Depa men o Neu oscience and Rehabili a ion, Tampe e Uni e si y Hospi al, 33521 Tampe e, Finland
3School o Medicine, Uni e si y o Tampe e, 33520 Tampe e, Finland
4School o Heal h Sciences, Uni e si y o Tampe e, 33520 Tampe e, Finland
Co espondence should be add essed o Sa i-Leena Himanen; sa i-leena.himanen@u a. i
Recei ed 25 Ma ch 2014; Re ised 12 May 2014; Accep ed 14 May 2014; Published 27 May 2014
Academic Edi o : Ma co Zucconi
Copy igh © 2014 Ki si V¨
ay ynen e al. This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion License,
which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
In acu e s oke, OSA has been ound o impai ehabili a ion and inc ease mo ali y bu he e ec o cen al apnea is mo e unclea .
The aim o he p esen s udy was o e alua e he easibili y o using limi ed ambula o y eco ding sys em (sleep ma ess o e alua e
noc u nal b ea hing and EOG-elec odes o sleep s aging) in sleep diso de ed b ea hing (SDB) diagnos ics in mild acu e ce eb al
ischemia pa ien s and o disco e he p e alence o a ious SDB-pa e ns among hese pa ien s. 42 pa ien s wi h mild ischemic
s oke o ansien ischemic a ack we e s udied. OSA was ound in 22 pa ien s (52.4%). Cen al apnea was ound in wo pa ien s
(4.8%) and sus ained pa ial obs uc ion in only one pa ien (2.4%). Sleep s aging wi h EOG-elec odes only yielded a simila
ou come as sco ing wi h s anda d ules. OSA was ound o be common e en a e mild s oke. I s ea ly diagnosis and ea men
would be a ou able in o de o imp o e eco e y and educe mo ali y. Ou esul s sugges ha OSA can be assessed by a limi ed
eco ding se ing wi h EOG-elec odes, sleep ma ess, and pulse oxime y.
1. In oduc ion
Sleep diso de ed b ea hing (SDB) is common in ce eb o as-
cula disease [1]. In pa icula , obs uc i e sleep apnea (OSA)
cons i u es a isk ac o o ce eb o ascula e en s, and i is
equen ly seen in pa ien s wi h bo h s oke and ansien
ischemic a ack (TIA) [1,2]. OSA is known o inc ease s oke
mo ali y and i wo sens he ou comes o ehabili a ion [3–
5]. Con inuous posi i e ai way p essu e (CPAP) ea men o
OSAins okepa ien shasbeen ound odec easemo ali y,
imp o e unc ional eco e y, inc ease subjec i e well-being
and mood, and inhibi ecu en s okes [6–10]. The e ec o
cen alapneasonmo ali yand unc ional eco e yins oke
pa ien s is less s udied, bu in one s udy cen al apnea did no
inc ease mo ali y a e s oke as OSA did [11].
E en hough ou ine sc eening o OSA in he s oke uni s
is becoming mo e common, i is no usually pe o med in
he s oke uni s in Finland. One eason o his migh be ha
pa ien s wi h bo h s oke and OSA do no p esen he ypical
clinical pic u e o sleep apnea; hey a e o en nei he obese no
sleepy [4,12,13]. In addi ion, i migh be ha sleep s udies a e
conside ed cumbe some in acu e hospi al wa ds. Pa ien s in
s oke uni s may ha e many di e en moni o ing de ices, as
well as nasal supplemen a y oxygen he apy, and pe o ming
noc u nal polyg aphy wi h addi ional senso s is pe haps no
a o ed.
Sleep ma esses p o ide a nonin asi e means o diagnose
SDB. Bo h he SCSB (s a ic cha ge sensi i e bed) and he
Em i mo emen senso ha e been p o en o be sui able
in de ec ing sleep apnea [14,15]. SDB diagnos ics wi h he
mo emen senso is pa ien - iendly, and he manual sco ing
p ocedu e, pe o med in 3 min epochs, is easy and apid.
Howe e , in limi ed polyg aphy se ings wi hou EEG, he
apnea-hypopnea index (AHI) ends o be lowe han in
ull polysomnog aphy (PSG) eco dings [16]. In acu e wa ds
pa ien s o en sleep poo ly due o he nu sing and ea men
Hindawi Publishing Co po a ion
Sleep Diso de s
Volume 2014, A icle ID 317615, 7 pages
h p://dx.doi.o g/10.1155/2014/317615
2Sleep Diso de s
ime schedules. The e o e some es ima ion o o al sleep ime
would be bene icial.
Wi h his s udy we aim o disco e he p e alence o
obs uc i e and cen al sleep apnea as well as p olonged
pa ial obs uc ion, among mild s oke and TIA pa ien s
in Finland. P olonged pa ial obs uc ion is associa ed wi h
con inuous inc ease in espi a o y esis ance wi hou apneas
o hypopneas [17–19]. I is easily de ec ed by he SCSB and
heEm i senso s,andi sp e alenceins okepa ien shas
no been p e iously s udied.
In addi ion we aim o de e mina e whe he limi ed polyg-
aphy consis ing o mo emen senso and pulse oxime y
in o de o e alua e noc u nal b ea hing combined wi h
elec ooculog aphy-signal (EOG) in o de o e alua e sleep
ime will p o ide a diagnos ic means o adequa e quali y in
de ec ing SDB.
2. Ma e ials and Me hods
Pa ien s we e ec ui ed om he S oke Uni o Tampe e Uni-
e si y Hospi al be ween May and Oc obe , 2010. All pa ien s
we e hospi alized due o ischemic ce eb al s oke o ansien
ischemic a ack (TIA). The ini ial eligibili y c i e ia o he
s udy we e as ollows: (a) TIA o mild ischemic s oke, he
se e i y de ined as Na ional Ins i u es o Heal h S oke Scale
(NIHSS) sco e <12, and (b) no p e ious CPAP ea men o
OSA. The esea che e alua ed he NIHSS sco e a he wa d
and he ollowing demog aphic da a we e ga he ed: age, sex,
BMI, neck ci cum e ence, p e ious diseases, smoking, use o
alcohol, subjec i e sleep p oblems (sno ing and apnea) in he
weeks be o e he s oke, and inc eased day ime sleepiness.
Fo sleepiness he ollowing ques ion was p esen ed: “we e
yousleepie hanusualacoupleo daysbe o e hes oke?”
The s udy was app o ed by he E hical Commi ee o he
Pi kanmaa Hospi al Dis ic and all he subjec s ga e hei
w i en in o med consen . I OSA was ound, he subjec was
e e ed o he Sleep Cen e o e alua e he need o CPAP
ea men . The pulmonologis hen began CPAP ea men
acco ding o he s anda ds o ou hospi al.
2.1. Reco ding Mon age. The polyg aphic eco dings we e
pe o med du ing he i s nigh in he hospi al, whene e
possible. The eco dings we e made in he s oke uni o a
he neu ologic wa d. The limi ed polyg aphy se ing consis ed
o wo EOG channels (EOG P8-A2, EOG P18-A2), elec o-
ca diog aphy, pulse oxime y, and he Em i senso (32 cm ×
62 ×0.4 cm, placed unde he ho acic a ea o he pa ien ).
In 14 andomly selec ed pa ien s noc u nal sleep quali y and
b ea hing we e assessed wi h ull polysomnog aphy, which
consis ed o he same senso s as he limi ed polyg aphy and
addi ionally six EEG channels (F3-A2, C3-A2, O1-A2, F4-
A1, C4-A1, and O2-A1), submen al and an e io ibialis mus-
cle elec omyog aphy, ho acic and abdominal espi a o y
mo emen s by induc i e bel s, and posi ion. Ai low was
measu ed wi h a he mis o and also wi h a nasal p essu e
ansduce i he pa ien did no ha e nasal supplemen a y
oxygen he apy. A sampling a e o 2 Hz was used o pulse
oxime y, 10 Hz o espi a o y mo emen s, 500 Hz o ECG,
and 200 Hz o he Em i senso and he o he signals.
2.2. Visual Analysis. The ull polysomnog aphies (PSG) we e
classi ied in o he sleep s ages acco ding o he s anda d
c i e ia. The apnea-hypopnea index (AHI) was calcula ed as
he numbe o obs uc i e apneas and hypopneas (hypopnea
ule 4b in [20]) pe hou o sleep. A ousals we e sco ed
acco ding o he c i e ia o he ASDA [21]. To e alua e he
easibili y o he in ended EOG sleep s aging we classi ied he
ull polysomnog aphies in o he sleep s ages by means o he
EOG-signals, wi hou EEG o EMG, oo [22]. The amoun s o
sleep and sleep s ages comp ised by he wo di e en me hods
we e hen compa ed.
The limi ed polyg aphs (POL) we e classi ied in o he
sleep s ages u ilizing only he EOG signals. The espi a o y
analysis o he limi ed polyg aphs was pe o med by he Em i
signal. Em i signal was sco ed in o one o en ca ego ies in
3 min epochs as in ou p e ious wo k [15,19]. The ma ess
signal ca ego ies used we e as ollows: no mal b ea hing
(NB), pe iodic b ea hing ype 1 (P1), obs uc i e pe iodic
b ea hing ypes 1–3 (OP1–3), cen al pe iodic b ea hing
(CP), p olonged pa ial obs uc ion (inc eased espi a o y
esis ance, IRR), la ge mo emen a i ac s las ing >40 s (M),
epochs wi h a leas ou sho pe iodic mo emen s in all
channels wi hou espi a o y a ia ion (pe iodic mo emen s,
PM), and wake epochs (W) wi h EEG-de ined wake ulness
o mo e han 50% o ime. The wake ulness epochs we e no
included in he u he analyses. The espi a o y esul s o he
ma ess b ea hing analyses a e p esen ed in pe cen ages o
o al sleep ime de i ed om he EOG sco ing.
The EEG sco ings we e pe o med by wo clinical neu-
ophysiologis s. The median sco ing ag eemen was 84.1%
(76.4–86.7%)wi h hemedianKappa aluebeing0.71( ange
0.67–0.81). The consensus sco ings we e used in he inal
analyses. The u he analyses o he eco dings we e pe -
o med om he i s sleep onse a e 8 PM o he inal
awakening in he mo ning. Obs uc i e AHI ≥15/h was
selec ed o designa e ma ked obs uc i e sleep apnea. Based
on ou p e ious s udy an obs uc i e AHI o 15/h co esponds
well wi h obs uc i e pe iodic b ea hing ime (%OP1 + %OP2
+%OP3)o 21%inEm i sco ing[15]. Thus he cu -o
alues o obs uc i e AHI ≥15/h and obs uc i e pe iodic
ime pe cen age o ≥21 we e selec ed o de ine ma ked OSA
and such pa ien s we e classi ied in o he OSA g oup. The
same cu -o alues we e used o de ine he cen al apnea
g oup(cen alAI≥15/h o cen al pe iodic b ea hing o
≥21%). The hi d g oup was comp ised o he pa ien s ha ing
a p olonged pa ial obs uc ion (IRR) pe cen age ≥21%, wi h
bo h obs uc i e AHI and cen al AI less han 15/h (IRR
g oup). The emaining pa ien s composed he non-SDB
g oup.
2.3. S a is ics. In he s a is ical compa isons, nonpa ame -
ic es s we e used, since all he a iables we e no no -
mallydis ibu ed.Thecompa isonbe ween he wosleep
s aging p ocedu es (s anda d sco ing e sus EOG sco ing)
was pe o med by Spea man’s co ela ion coe icien and
Sleep Diso de s 3
Cohen’s Kappa-analysis [23]. Pai -wised compa isons o
ela ed pa ame e s we e conduc ed by using he Wilcoxon
es . The pa ien g oups we e compa ed wi h he Mann-
Whi ney es . The demog aphic da a (sex, age, BMI, p e ious
dep ession, p e ious s oke, p e ious hea ailu e, hype en-
sion, hype choles e olemia, diabe es melli us, hypo hy eosis,
as hma, COPD, p e ious o he lung diseases, use o obacco,
use o alcohol, NIHSS-sco e, sno ing, wi nessed apneas, and
inc eased sleepiness) we e e alua ed wi h he bina y logis ic
eg ession analysis. In s a is ical analyses 𝑃 alues <0.05 we e
conside ed s a is ically signi ican .
3. Resul s
Al oge he 46 pa ien s olun ee ed o he s udy. Fou een
pa ien s we e ec ui ed o ull polysomnog aphy bu only
wel e eco dings we e comple ed (86%). One pa ien wan ed
he senso s o be de ached soon a e s a ing he eco ding,
and one eco ding e mina ed unexpec edly a 11 p.m. Thi y-
wo pa ien s we e examined by using he limi ed polyg aphy.
Two o hese eco dings (6%) we e excluded om he
analyses; one pa ien wan ed he eco ding o be s opped jus
be o e alling asleep and in one eco ding bo h he Em i
and pulse oxime y signals we e missed. In o al, 42 pa ien s
we e accep ed in o he analyses. The polysomnog aphy g oup
(PSG) consis ed o one emale pa ien and ele en male
pa ien s, whe eas in he limi ed polyg aphy (POL) g oup, 17
pa ien s we e emale and 13 we e male. The analysis s a
imes a ied be ween 20 : 06 and 02 : 45, wi h median o 21 : 19.
In o de o ensu e he easibili y o he sleep s age sco ing
basedon heEOGelec odes,wecompa ed hesleeps age
esul s de i ed by using he wo sco ing me hods in he PSG
g oup. The sleep s aging ag eemen be ween he s anda d
sco ingme hodand heEOGsco ingwas82.2%(median,
ange 56.3–88.2%). The median Kappa alue was 0.72 (0.41–
0.80), indica ing subs an ial ag eemen [23]. The o al sleep
ime (TST) es ima ed by he EOG sco ing co ela ed well
wi h he TST measu ed by s anda d sco ing (Spea man’s
co ela ion coe icien 0.964). The sleep s age pa ame e s a e
p esen ed in Table 1. The o al amoun o sleep was highe
by heEOGsco ing hanby hes anda dsco ing(353.5min
e sus 321 min, esp., Table 1), bu he di e ence emained
a he edge o s a is ical signi icance. The pe cen ages in he
amoun s o a ious sleep s ages be ween he wo sco ing
p ocedu es did no yield o s a is ically signi ican di e ences.
Figu e 1 depic u es he hypnog ams wi h he lowes sleep
s aging ag eemen be ween he s anda d sco ing and he
EOG sco ing (56.3%, Kappa 0.41 = mode a e conco dance).
The sleep s age pa ame e s o all 42 pa ien s a e p esen ed
in Table 2. The median sleep e iciency, as well as he amoun s
o deep sleep and REM-sleep, was in gene al qui e low e en
i mos o he subjec s slep subjec i ely well o ai ly well
du ing he s udy nigh (45% and 36%, esp.). Se en een
pe cen o he pa ien s slep poo ly and one pa ien (2%) el
ha hehadno slep a all(his o alsleep imewas3h).
Eigh ou o he wel e subjec s in he PSG g oup had
AHI ≥15/h (one emale, 7 men). In he POL g oup 14 pa ien s
had obs uc i e pe iodic b ea hing ≥21% o TST (8 emale, 6
male). The espi a o y pa ame e s o he PSG g oup and he
POL g oup a e p esen ed in Table 3.Wealsocalcula ed he
AHI o sleep pe iod ime in he PSG g oup (sleep pe iod
ime, ime omsleeponse o he inalawakeningwi h
wake ulness ime inside he sleep pe iod included). The AHI
o he sleep pe iod ime was signi ican ly lowe han he AHI
de i ed om TST (13.3/h, ange 0.5–48.5, e sus 18.1/h ange
1.2–67.9, 𝑃 alue0.002).Whileeigh pa ien sou o wel ehad
AHI >15/h when TST was applied, only h ee pa ien s had
AHI >15/h i sleep pe iod ime was used.
Al oge he 17 subjec s (40.5%, 9 males, 8 emales) did no
ha e sleep diso de ed b ea hing (non-SDB g oup, Table 4)
when hePSGand hePOLg oupswe epooled.The
emaining pa ien s had ei he obs uc i e o cen al b ea h-
ing diso de ; no one had bo h abno mal obs uc i e and
abno mal cen al indices. OSA was ound in 22 pa ien s
(52.4%, OSA g oup). Nine o hese pa ien s we e emale
(41%). Cen al apnea was ound in wo male pa ien s (4.8%,
cen al g oup). IRR was ound only in one emale pa ien
(2.4%, IRR g oup). In gene al he s udy subjec s had e y
mild s oke as assessed by he NIHSS and hey we e no
e y o e weigh . S a is ical compa isons we e made be ween
henon-SDBand heOSAg oupsonlyas hecen alg oup
and he IRR g oup emained so small. The e we e only ew
s a is ically signi ican di e ences be ween he non-SDB and
OSA g oups; he subjec s in he non-SDB g oup had mo e
slow wa e sleep (N3), and hei oxygen desa u a ion indices
(ODI4) we e lowe .
Mo e demog aphic pa ame e s a e p esen ed in Table 5.
Hype ension, sno ing, and inc eased sleepiness we e e-
quen in all subjec g oups. The demog aphic a iables
we e e alua ed wi h he bina y logis ic eg ession analysis.
Only inc eased sleepiness was iden i ied as an independen
p edic o o hes okein henon-SDBg oupwi hOR=5.33
(CI 1.32–21.53).
4. Discussion
In he p esen s udy he p e alence o SDB among mild s oke
and TIA pa ien s was 59.5%. Ou esul o he p e alence
o OSA (52.3%) in pa ien s wi h TIA o mild s oke is in
conco dance wi h hose o a p e ious s udy [24]. Only wo
subjec swe e ound oha ema kedcen alapnea(4.8%),
bu his co esponds wi h he indings in a p e ious me a-
analysis [25]. The p e alence o IRR in he p esen s udy was
lowe (2.4%) han he p e alence among ou sleep labo a o y
pa ien s (10.2%) [15]. The cu -o limi we p e iously used was
lowe (≥15%)bu e eni he aluehadbeenused, henumbe
o IRR-pa ien s in he p esen s udy would no ha e changed.
OSA inc eases sympa he ic one [26], inducing hype -
ension which may be a p edisposing ac o o s oke.
Howe e , he e a e many physiologic mechanisms behind he
inc ease in blood p essu e in OSA. Fo example, he hypoxia
associa ed wi h pe iodic apnea augmen s sympa he ic ac i i y
and blood p essu e wi h an addi ional inc ease caused by
epe i i e a ousals, es o ing he uppe ai way pa ency [27–
29]. IRR ep esen s a di e en SDB en i y since he pe iods
wi h pa ial obs uc ion may las se e al minu es wi hou
4Sleep Diso de s
Table 1: Sleep s age pa ame e s o he PSG g oup wi h he wo di e en sco ing sys ems.
S anda d sco ing EOG sco ing 𝑃 alue
Median (min–max) Median (min–max)
TST (min) 321 (48.5–506.5) 353.5 (49.5–498.5) 0.051
SEI (%) 72.2 (42–90.9) 69.1 (42.9–89.4) 0.48
% N1 12.8 (6.5–41.2) 15.4 (6.8–48.5) 0.068
% N2 69.5 (36.1–83.2) 66.2 (37.4–82.4) 0.33
% N3 7.0 (0–23.6) 5.9 (0–17.7) 0.17
% REM 11.2 (0–22.3) 12.9 (0–25.4) 0.61
S anda d sco ing: sleep s age sco ing acco ding o he s anda d ules; EOG sco ing: sleep s age sco ing based on he wo EOG (elec ooculog aphy) channels
only; TST: o al sleep ime; SEI: sleep e iciency index e e ed o analysis ime; % N1–N3: pe cen age o N1–N3 sleep e e ed o TST; % REM: pe cen age o
REM sleep e e ed o TST. S a is ical compa isons wi h he Wilcoxon es .
Wake
REM
N1
N2
N3
(a)
Wake
REM
N1
N2
N3
22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00
(b)
Figu e 1: Example o he hypnog ams o he same pa ien de i ed by he wo di e en sleep s aging p ocedu es. The uppe hypnog am
ep esen s s anda d sleep s aging and he lowe hypnog am ep esen s EOG sco ing. The hypnog ams esemble each o he bu he e a e
di e ences in bo h sleep e iciency (69% by s anda d sleep s aging, 59% by EOG-sco ing) and he a angemen o he sleep s ages. The pa ien
had mode a e OSA (AHI 18.8/h).
Table 2: The sleep quali y pa ame e s o he 42 pa ien s.
Median Min Max
TST (min) 368.5 48.5 510.5
SEI (%) 70.8 27.5 97.6
% N1 8.8 1.8 41.2
% N2 46.9 3.3 83.2
% N3 6.5 0 26.2
%REM 7.1 0 31.2
TST: o al sleep ime; SEI: Sleep E iciency Index e e ed o analysis ime; %
N1–N3: pe cen age o N1–N3 sleep e e ed o TST; % REM: pe cen age o
REM sleep e e ed o TST.
a ousals. In ha way he inc ease in sympa he ic ac i i y
induced by a ousals does no eme ge. In addi ion he p e-
ailing ai low agains esis ance du ing IRR is capable o
ac i a ing he pulmona y s e ch ecep o s, which leads o
he inhibi ion o sympa he ic ac i a ion [30]. Thus, i migh
be ha IRR does no cons i u e a ma ked isk ac o o
hype ension. Indeed, i has been p esen ed ha women
wi h IRR ha e less hype ension han women in he gene al
popula ion [31]. I is possible ha he isk o ca dio ascula
e en s o IRR pa ien s migh be educed as compa ed o OSA
pa ien s. Thus he p esen s udy inc eases ou knowledge o
IRR, bu i s p o ound signi icance s ill emains somewha
obscu e and equi es u he e alua ion.
In he p esen s udy, he neck ci cum e ence measu e-
men did no di e be ween he OSA g oup and he non-
SDB g oup. This migh be due o he ac ha he BMI o
ou pa ien s was no high in gene al. Simila esul s ha e
been ound ea lie [13,24]. In addi ion, ou OSA pa ien s
did no p esen inc eased day ime sleepiness be o e s oke,
unlike hose who did no ha e OSA. The absence o excessi e
day ime sleepiness has been no ed be o e [13]. I may be ha
pa ien s wi h OSA do no necessa ily ecognize inc eased
Sleep Diso de s 5
Table3:The espi a o ypa ame e so hePSGg oupand hePOLg oup.
PSG g oup POL g oup
Median Min Max Median Min Max
AHI obs uc i e (𝑛/h) 18.1 1.2 67.9
AHI cen al (𝑛/h) 0 0 1.2
ODI4 4.5 0 32.5 6.0 0 64.8
% OP1–OP3 27.2 2.6 69.3 14.1 0 45.8
% CP 0 0 4.7 0 0 51.6
% IRR 0.8 0 9.6 0.5 0 25.4
SaO2min (%) 87.5 79.0 93.0 85.0 75.0 91.0
SaO2mean (%) 93.7 90.7 96.5 93.2 89.5 97.1
Pulse (bpm) 55 46 73 58 46 86
AHI obs uc i e: numbe o obs uc i e apneas and hypopneas pe hou o sleep; AHI cen al: numbe o cen al apneas pe hou o sleep; ODI4: numbe o
desa u a ions pe hou o sleep; % OP1–OP3: pe cen age o ime wi h obs uc i e pe iodic b ea hing as de ec ed by Em i senso ; % CP: pe cen age o imewi h
cen al pe iodic b ea hing as de ec ed by Em i senso ; % IRR: pe cen age o ime wi h p olonged pa ial obs uc ion as de ec ed by Em i senso ; SaO2min:
minimum oxygen sa u a ion pe cen age; SaO2mean:a e ageoxygensa u a ionpe cen age;Pulse:meanpulsepe minu e.
Table 4: Some demog aphic and sleep pa ame e s o he di e en subjec g oups.
Non-SDB, 𝑛=17 OSA, 𝑛=22 Cen al, 𝑛=2 IRR, 𝑛=1 Non-SDB e sus OSA
Median (min–max) Median (min–max) Median (min–max) Value 𝑃 alue
BMI 26.0 (19.9–32.7) 26.6 (21.9–37.4) 25.3 (25.2–25.3) 27.5 0.705
AGE 62 (45–79) 69 (37–82) 77 (75–78) 68 0.131
TST 426 (49–511) 321 (105–490) 300 (282–318) 416 0.116
SEI 72.8 (42.0–97.6) 70.3 (27.5–95.7) 65.2 (61.3–69.1) 67.2 0.236
% N1 8.9 (2.6–41.2) 9.0 (1.8–23.0) 12.5 (11.9–13.1) 8.2 0.962
% N2 45.9 (33.1–73.9) 57.8 (3.3–83.2) 44.7 (42.7–46.7) 53.6 0.168
% N3 12.4 (3.3–26.2) 4.3 (0–23.6) 1.5 (1.1–1.8) 2.4 0.005
% REM 6.9 (0–22.3) 10.8 (1.0–31.2) 6.9 (5.2–8.5) 3.1 0.539
ODI4 1.8 (0–8.2) 16.4 (0–64.8) 9.7 (4.2–15.1) 0.7 0.000
SaO2min (%) 87 (80–91) 85 (75–93) 87 (84–89) 85 0.053
SaO2mean (%) 93.6 (90.7–97.1) 93.3 (89.5–96.8) 94.5 (93.3–95.7) 91.7 0.975
Pulse (bpm) 61 (50–86) 58 (46–78) 51 (50–52) 51 0.133
NIHSS 2 (0–6) 1 (0–4) 3 (2–4) 1 0.138
Neck (cm) 40 (34–44) 41 (38–53) 41 (40–42) 39 0.231
NIHSS: Na ional Ins i u es o Heal h S oke Scale; Neck: neck ci cum e ence in cen ime e s.
The es o he abb e ia ions a e as in Tables 1and 3. S a is ical compa isons be ween he Non-SDB g oup and he OSA g oup wi h he Mann-Whi ney es .
sleepiness due o he ch onic sleepiness induced by OSA. The
p oblem wi h ou s udy is ha sleepiness was no e alua ed
s uc u ally, so i is no possible o d aw s ong conclusions.
In a p e ious me a-analysis he p e alence o emale
pa ien s wi h OSA and s oke was ound o be 35.4% [25].
In ou s udy, he pe cen age o emales wi h OSA was 41%.
In ha way, he po ion o emale pa ien s is clea ly la ge
han he p e alence o OSA in emales in gene al [32]. I
hasbeenshown ha emalesmayp esen a ypicalSDB-
ela ed symp oms, o example, insomnia and dep ession
[33],and he ema kablepo iono emalepa ien samong he
OSA + s oke pa ien s may, in pa , accoun o he a ypical
subjec i e complain s among hese pa ien s. I is possible ha
ou ine anamnesis as i con en ionally ela es o OSA does
no necessa ily e eal SDB in s oke pa ien s, which s esses
he need o OSA sc eening in s oke uni s.
Pa ien s end o sleep poo ly in he hospi al, and also
he objec i e sleep quali y o ou pa ien s was poo , as
assessed by sleep e iciency. In mos o he sleep eco d-
ings we u ilized only he wo EOG channels in he sleep
s age sco ing (EOG sco ing) o achie e an es ima e o TST.
As a esul , he o al sleep ime ob ained by he EOG
sco ing co ela ed well wi h he s anda d sco ing esul s.
Acco ding o p e ious s udies sleep s aging by EOG sig-
nal seems ela i ely eliable and p omising [22,34]and
because heEOGsignalscanbe eco dedbydisposable
elec odes, he a achmen o he elec odes is easy. Ou
esul s sugges ha sleep s aging wi h EOG esul s in
mo e alid diagnos ic SDB ou come han con en ional
espi a o y polyg aphy wi hou sleep s aging e en i he
EOG sco ing p ocedu e may o e es ima e he amoun o
sleep.

6Sleep Diso de s
Table 5: Demog aphic pa ame e s o he 42 pa ien s.
Non-SDB OSA Cen al IRR
Dep ession (%) 5.9 4.5 0 0
P e ious s oke (%) 5.9 13.6 0 0
Hea ailu e (%) 23.5 27.3 0 0
Hype ension (%) 47.1 63.6 50 100
Hype choles e olemia (%) 11.8 13.6 0 0
Diabe es melli us (%) 5.9 4.5 0 0
Hypo hy eosis (%) 5.9 4.5 0 100
As hma (%) 5.9 0 0 0
COPD (%) 17.6 4.5 0 0
Tobacco use (%) 47 18.1 50 0
Alcohol use (%) 29.4 22.7 0 0
Sno ing (%) 52.9 54.5 50 100
Wi nessed apneas (%) 5.9 22.7 50 0
Inc eased sleepiness 76.5 40.9 0 100
COPD: ch onic obs uc i e pulmonal disease.
Ou o he aim was o e alua e he easibili y o limi ed
polyg aphy wi h a sleep ma ess in sc eening SDB. The
clinical e alua ion o ma ess signal is easy and as and,
depending on he eco ding de ice used, adding di e en
addi ional signals is o en simple. In he p esen s udy, he
pe cen age o success ul eco dings was highe in he POL
g oup han in he PSG g oup, which suppo s he use o he
limi ed eco ding mon age in acu e wa ds. I he numbe
o eco ded pa ame e s is kep low, he sys em does no
dis u b he pa ien , enabling eco dings o se e al nigh s
when needed.
5. Conclusion
In hep esen s udyweconcen a edonmilds okeand
TIA pa ien s only, and he eco dings we e made soon a e
he s oke. In gene al, pa ien s wi h TIA and mild s oke
eco e well, bu , acco ding o he cu en knowledge, hey
ha e an inc eased isk o s oke ecu ence [35]. As ecu en
s oke migh be inhibi ed by he p ope ea men o SDB
[6], sc eening o SDB should be a ou ed in he s oke uni s.
Ou esul s sugges ha limi ed and less dis u bing eco ding
se ing is su icien o e eal OSA in s oke uni s.
Con lic o In e es s
The au ho s decla e ha he e is no con lic o in e es s
ega ding he publica ion o his pape .
Acknowledgmen s
The s udy was inancially suppo ed by Tekes, he Na ional
Technology Agency o Finland, and by he Compe i i e
Resea ch Financing o he Expe Responsibili y a ea o
Tampe e Uni e si y Hospi al G an nos. 9M014 and 9P013.
Re e ences
[1] C.Basse i,M.S.Ald ich,R.D.Che in,andD.Quin ,“Sleep
apnea in pa ien s wi h ansien ischemic a ack and s oke: a
p ospec i e s udy o 59 pa ien s,” Neu ology, ol.47,no.5,pp.
1167–1173, 1996.
[2]O.Pa a,A.A boix,S.Bechiche al.,“Timecou seo sleep-
ela ed b ea hing diso de s in i s -e e s oke o ansien
ischemic a ack,” Ame ican Jou nal o Respi a o y and C i ical
Ca e Medicine, ol.161,no.2I,pp.375–380,2000.
[3] D.C.Good,J.Q.Henkle,D.Gelbe ,J.Welsh,andS.Ve huls ,
“Sleep-diso de ed b ea hing and poo unc ional ou come a e
s oke,” S oke, ol.27,no.2,pp.252–259,1996.
[4] Y. Kan e ko, V. E. Haj e k , V. Z i a n o i c , J. R a b o u d, an d T. D.
B adley, “Rela ionship o sleep apnea o unc ional capaci y and
leng h o hospi aliza ion ollowing s oke,” Sleep, ol.26,no.3,
pp. 293–297, 2003.
[5]O.Pa a,A.A boix,J.M.Mon se a ,L.Quin
´
o, S. Bechich,
and L. Ga c´
ıa-E oles, “Sleep- ela ed b ea hing diso de s: Impac
on mo ali y o ce eb o ascula disease,” Eu opean Respi a o y
Jou nal, ol.24,no.2,pp.267–272,2004.
[6] M. ´
A. Ma ´
ınez-Ga c´
ıa, J. J. Sole -Ca alu˜
na, L. Eja que-
Ma ´
ınez e al., “Con inuous posi i e ai way p essu e ea men
educes mo ali y in pa ien s wi h ischemic s oke and obs uc-
i e sleep apnea: a 5-yea ollow-up s udy,” Ame ican Jou nal o
Respi a o y and C i ical Ca e Medicine, ol.180,no.1,pp.36–41,
2009.
[7] M. ´
A. Ma ´
ınez-Ga c´
ıa, R. Galiano-Blanca , P. Rom´
an-
S´
anchez, J.-J. Sole -Ca alu˜
na, L. Cabe o-Sal , and E. Salcedo-
Maiques, “Con inuous posi i e ai way p essu e ea men in
sleepapneap e en snew ascula e en sa e ischemics oke,”
Ches , ol.128,no.4,pp.2123–2129,2005.
[8] O. Sandbe g, K. A. F anklin, G. Buch , S. E iksson, and Y.
Gus a son, “Nasal con inuous posi i e ai way p essu e in s oke
pa ien s wi h sleep apnoea: a andomized ea men s udy,”
Eu opean Respi a o y Jou nal, ol.18,no.4,pp.630–634,2001.
[9] C.M.Ryan,M.Bayley,R.G een,B.J.Mu ay,andT.D.B adley,
“In luence o con inuous posi i e ai way p essu e on ou comes
Sleep Diso de s 7
o ehabili a ion in s oke pa ien s wi h obs uc i e sleep apnea,”
S oke, ol.42,no.4,pp.1062–1067,2011.
[10]T.E.Wessendo ,Y.-M.Wang,A.F.Thilmann,U.So gen ei,
N. Konie zko, and H. Teschle , “T ea men o obs uc i e sleep
apnoea wi h nasal con inuous posi i e ai way p essu e in
s oke,” Eu opean Respi a o y Jou nal, ol.18,no.4,pp.623–629,
2001.
[11] C. Sahlin, O. Sandbe g, Y. Gus a son e al., “Obs uc i e sleep
apnea is a isk ac o o dea h in pa ien s wi h s oke: a 10-yea
ollow-up,” A chi es o In e nal Medicine, ol.168,no.3,pp.297–
301, 2008.
[12] C. L. Basse i, M. Milano a, and M. Gugge , “Sleep-diso de ed
b ea hing and acu e ischemic s oke: diagnosis, isk ac o s,
ea men , e olu ion, and long- e m clinical ou come,” S oke,
ol. 37, no. 4, pp. 967–972, 2006.
[13] M. A z , T. Young, P. E. Peppa d e al., “Dissocia ion o
obs uc i e sleep apnea om hype somnolence and obesi y in
pa ien s wi h s oke,” S oke, ol.41,no.3,pp.e129–e134,2010.
[14] O. Polo, “Pa ial uppe ai way obs uc ion du ing sleep. S udies
wi h he s a ic cha ge-sensi i e bed (SCSB),” Ac a Physiologica
Scandina ica, Supplemen , ol.145,no.606,pp.1–118,1992.
[15] M. Tenhunen, E. Elomaa, H. Sis onen, E. Rauhala, and S.-L.
Himanen, “Em i mo emen senso in e alua ing noc u nal
b ea hing,” Respi a o y Physiology and Neu obiology, ol.187,no.
2, pp. 183–189, 2013.
[16] J.Hedne ,L.G o e,M.Bonsigno ee al.,“TheEu opeansleep
apnoea da abase (ESADA): epo om 22 Eu opean sleep
labo a o ies,” Eu opean Respi a o y Jou nal, ol.38,no.3,pp.
635–642, 2011.
[17] G. Bao and C. Guilleminaul , “Uppe ai way esis ance
synd ome-one decade la e ,” Cu en Opinion in Pulmona y
Medicine, ol.10,no.6,pp.461–467,2004.
[18] L. He n´
andez, E. Balles e , R. Fa ´
ee al.,“Pe o manceo nasal
p ongs in sleep s udies: spec um o low- ela ed e en s,” Ches ,
ol. 119, no. 2, pp. 442–450, 2001.
[19] M.Tenhunen,E.Rauhala,J.Vi kkala,O.Polo,A.Saas amoinen,
and S.-L. Himanen, “Inc eased espi a o y e o du ing sleep
is non-in asi ely de ec ed wi h mo emen senso ,” Sleep and
B ea hing, ol.15,no.4,pp.737–746,2011.
[20]C.Ibe ,S.Ancoli-Is ael,A.L.ChessonJ .,andS.F.Quan,
“The new sleep sco ing manual— he e idence behind he ules,”
Jou nal o Clinical Sleep Medicine, ol.3,no.2,p.107,2007.
[21] M. Bonne , D. Ca ley, M. Ca skadon e al., “EEG a ousals:
sco ing ules and examples. A p elimina y epo om he Sleep
Diso de s A las Task Fo ce o he Ame ican Sleep Diso de
Associa ion,” Sleep, ol.15,no.2,pp.173–184,1992.
[22] A.-M. Lapinlampi and S.-L. Himanen, “Sleep s aging wi h
on opola EEG de i a ion,” Sleep and Hypnosis, ol.6,no.2,
pp.59–65,2004.
[23]J.R.LandisandG.G.Koch,“Themeasu emen o obse e
ag eemen o ca ego ical da a,” Biome ics, ol.33,no.1,pp.
159–174, 1977.
[24]W.Chan,S.B.Cou s,andP.Hanly,“Sleepapneainpa ien s
wi h ansien ischemic a ack and mino s oke: oppo uni y
o isk educ ion o ecu en s oke?” S oke, ol.41,no.12,
pp. 2973–2975, 2010.
[25] K. G. Johnson and D. C. Johnson, “F equency o sleep apnea
in s oke and TIA pa ien s: a me a-analysis,” Jou nal o Clinical
Sleep Medicine, ol.6,no.2,pp.131–137,2010.
[26] V.K.Some s,A.L.Ma k,D.C.Za ala,andF.M.Abboud,“Con-
as ing e ec s o hypoxia and hype capnia on en ila ion and
sympa he ic ac i i y in humans,” Jou nalo AppliedPhysiology,
ol.67,no.5,pp.2101–2106,1989.
[27] R. L. Ho ne , D. B ooks, L. F. Koza , S. Tse, and E. A. Phillipson,
“Immedia e e ec s o a ousal om sleep on ca diac au onomic
ou low in he absence o b ea hing in dogs,” Jou nal o Applied
Physiology, ol.79,no.1,pp.151–162,1995.
[28] C.P.O’Donnell,T.Ayuse,E.D.King,A.R.Schwa z,P.L.Smi h,
and J. L. Robo ham, “Ai way obs uc ion du ing sleep inc eases
blood p essu e wi hou a ousal,” Jou nal o Applied Physiology,
ol.80,no.3,pp.773–781,1996.
[29] J. Ringle , R. C. Basne , R. Shannon e al., “Hypoxemia alone
does no explain blood p essu e ele a ions a e obs uc i e
apneas,” Jou nalo AppliedPhysiology, ol.69,no.6,pp.2143–
2148, 1990.
[30] D. R. Seals, N. O. Suwa no, M. J. Joyne , C. Ibe , J. G. Copeland,
and J. A. Dempsey, “Respi a o y modula ion o muscle sym-
pa he ic ne e ac i i y in in ac and lung dene a ed humans,”
Ci cula ion Resea ch, ol.72,no.2,pp.440–454,1993.
[31] U. An alainen, O. Polo, T. Vahlbe g, and T. Saa es an a,
“Reimbu sed d ugs in pa ien s wi h sleep-diso de ed b ea hing:
a s a ic-cha ge-sensi i e bed s udy,” Sleep Medicine, ol. 11, no. 1,
pp.49–55,2010.
[32] T. Telaki i, M. Pa inen, M. Kosken uo, T. Salmi, and J. Kap io,
“Pe iodic b ea hing and hypoxia in sno e s and con ols: ali-
da ion o sno ing his o y and associa ion wi h blood p essu e
and obesi y,” Ac a Neu ologica Scandina ica, ol.76,no.1,pp.
69–75, 1987.
[33] A.Valipou ,H.Lo halle ,H.Rausche ,H.Zwick,O.C.Bu ghu-
be , and P. La ie, “Gende - ela ed di e ences in symp oms o
pa ien s wi h suspec ed b ea hing diso de s in sleep: a clinical
popula ion s udy using he Sleep Diso de s Ques ionnai e,”
Sleep, ol.30,no.3,pp.312–319,2007.
[34]J.Vi kkala,J.Hasan,A.V
¨
a i, S.-L. Himanen, and K. M¨
ulle ,
“Au oma ic sleep s age classi ica ion using wo-channel elec o-
oculog aphy,” Jou nal o Neu oscience Me hods, ol.166,no.1,
pp. 109–115, 2007.
[35] M. F. Giles and P. M. Ro hwell, “Risk o s oke ea ly a e
ansien ischaemic a ack: a sys ema ic e iew and me a-
analysis,” Lance Neu ology, ol.6,no.12,pp.1063–1072,2007.