ECG le en icula hype ophy as a isk p edic o o sudden
ca diac dea h
Kimmo Po han
a,
⁎
,1
, Tuomas Ken ä
b,1
, Teemu J. Nii anen
c,d,1
, Ma kku S. Nieminen
a,1
, Lasse Oika inen
a,1
,
Ma i Vii asalo
a,1
, Jussi He nesniemi
e, ,1
,An iM.Jula
c,1
, Veikko Salomaa
d,1
,HeikkiV.Huiku i
b,1
,
Ch is ine M. Albe
g,1
, Jani T. Tikkanen
b,g,1
a
Di ision o Ca diology, Hea and Lung Cen e , Helsinki Uni e si y Cen al Hospi al, Helsinki, Finland
b
Resea ch Uni o In e nal Medicine, Uni e si y Hospi al o Oulu, Uni e si y o Oulu, Finland
c
THL-Na ional Ins i u e o Heal h and Wel a e, Tu ku, Finland
d
THL-Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland
e
Hea Cen e , Tampe e Uni e si y Hospi al, Tampe e, Finland
Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e, Finland
g
Di ision o P e en i e Medicine, Di ision o Ca dio ascula Medicine, Depa men o Medicine, B igham and Women's Hospi al, Ha a d Medical School, Bos on, MA, USA
abs ac a icle in o
A icle his o y:
Recei ed 25 July 2018
Recei ed in e ised o m 20 Sep embe 2018
Accep ed 25 Sep embe 2018
A ailable online 27 Sep embe 2018
Backg ound: Elec oca diog aphic (ECG) le en icula hype ophy (LVH) is an es ablished isk ac o o
ca dio ascula e en s. Howe e , limi ed da a is a ailable on he p ognos ic alues o di e en ECG LVH c i e ia
specifically o sudden ca diac dea h (SCD). Ou goal was o assess ela ionships o di e en ECG LVH c i e ia o
SCD.
Me hods: Th ee adi ional and clinically use ul (Sokolow–Lyon, Co nell, R
aVL
) and a ecen ly p oposed (Pegue o–
Lo P es i) ECG LVH ol age c i e ia we e measu ed in 5730 subjec s in he Heal h 2000 Su ey, a na ional gene al
popula ion coho s udy. Rela ionships be ween LVH c i e ia, as well as hei selec ed composi es, o SCD we e an-
alyzed wi h Cox eg ession models. In addi ion, popula ion-a ibu able ac ions o LVH c i e ia we e calcula ed.
Resul s: A e a mean ollow-up o 12.5 ± 2.2 yea s, 134 SCDs had occu ed. When used as con inuous a iables, all
LVH c i e ia excep o R
aVL
we e associa ed wi h SCD in mul i a iable analyses. When single LVH c i e ia we e
used as dicho omous a iables, only Co nell was significan a e adjus men s. The dicho omous composi e o
Sokolow–Lyon and Co nell was also significan a e adjus men s (haza d a io o SCD 1.82, 95% confidence in e -
al 1.20–2.70, P= 0.006) and was he only LVH measu e ha showed s a is ically significan popula ion-
a ibu able ac ion (11.0%, 95% confidence in e al 1.9–19.2%, P=0.019).
Conclusions: Sokolow–Lyon, Co nell, and Pegue o–Lo P es i ECG, bu no R
aVL
ol age, a e associa ed wi h SCD isk
as con inuous ECG ol age LVH a iables. When SCD isk assessmen /adjus men is pe o med using a dicho o-
mous ECG LVH measu e, composi e o Sokolow–Lyon and Co nell ol ages is he p e e ed op ion.
© 2018 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://
c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Keywo ds:
Elec oca diog aphy
Epidemiology
Le en icula hype ophy
Sudden ca diac dea h
1. In oduc ion
Elec oca diog aphic (ECG) le en icula hype ophy (LVH) is an
es ablished isk ac o o ca dio ascula e en s [1–6]. ECG LVH is also
associa ed specifically wi h inc eased isk o sudden ca diac dea h
(SCD) [7,8], which is among he leading causes o dea h wo ldwide
[9]. Associa ion be ween ECG LVH and SCD emains significan also
a e adjus ing o ana omic measu es o LVH (echoca diog aphy
[echo], magne ic esonance imaging) [8,10,11], indica ing ha ad e se
elec ical emodelling pe se con eys addi ional p ognos ic alue.
ECG is widely used and a ou ine es , among o he s, in all indi id-
uals wi h hype ension. Because o i s po en ial implica ions, sea ching
o signs o LVH is one o he key s eps in he ECG assessmen . Se e al
ECG LVH c i e ia ha e been de eloped, bu hei p ognos ic alues
ha e been compa ed in only a ew s udies [6,12–14] and he e is e en
mo e limi ed da a compa ing he p ognos ic alues o di e en LVH
c i e ia specifically o SCD. The p esen s udy was pe o med o com-
pa e he ela ionships o h ee adi ional and clinically use ul
(Sokolow–Lyon, Co nell, R
aVL
) and one ecen ly p oposed (Pegue o–Lo
P es i) LVH ol age c i e ia, as well as hei selec ed composi es, o
SCD in he gene al popula ion.
In e na ional Jou nal o Ca diology 276 (2019) 125–129
⁎Co esponding au ho a : Di ision o Ca diology, Hea and Lung Cen e , Helsinki
Uni e si y Cen al Hospi al, POB 340, 00029 HUS Helsinki, Finland.
E-mail add ess: kimmo.po han@fimne .fi(K. Po han).
1
This au ho akes esponsibili y o all aspec s o he eliabili y and eedom om bias
o he da a p esen ed and hei discussed in e p e a ion.
h ps://doi.o g/10.1016/j.ijca d.2018.09.104
0167-5273/© 2018 The Au ho s. Published by Else ie B.V. This is an open access a icle unde he CC BY-NC-ND license (h p://c ea i ecommons.o g/licenses/by-nc-nd/4.0/).
Con en s lis s a ailable a ScienceDi ec
In e na ional Jou nal o Ca diology
jou nal homepage: www.else ie .com/loca e/ijca d
2. Me hods
2.1. S udy popula ion, elec oca diog aphy
The Heal h 2000 Su ey was a p ospec i e, epidemiologic su ey ha was conduc ed
in Finland be ween 2000 and 2001. The su ey popula ion (n= 8028) was a wo-s age
s a ified clus e sample d awn om he popula ion egis e and was ep esen a i e o
he en i e Finnish adul (≥30 yea s) gene al popula ion. Su ey p ocedu es consis ed o
a s uc u ed in e iew, a comp ehensi e heal h examina ion wi h ques ionnai es,
measu emen s, and a physician's clinical examina ion. Su ey was highly success ul
wi h almos 85% o he ec ui ed subjec s a ending he heal h examina ion. De ailed
Heal h 2000 Su ey me hodology epo is a ailable online [15]. Su ey was conduc ed
acco ding o he ecommenda ions o he Decla a ion o Helsinki, and was app o ed by
he Ins i u ional E hics Commi ee and by he Epidemiology E hics Commi ee o he
Helsinki and Uusimaa Hospi al Dis ic . Subjec s ga e w i en in o med consen .
Digi al 12‑lead ECGs we e eco ded a he Heal h 2000 Su ey baseline wi h Ma -
que e MAC 5000 (GE Ma que e Medical Sys ems, Milwaukee, WI). ECGs we e a ailable
om 6305 subjec s. Subjec wi h low ECG quali y, comple e le / igh bundle b anch
block, o incomple e co a ia e da a we e excluded. In addi ion, subjec s aged N80 yea s
a he su ey baseline we e excluded om adjudica ion o he cause o dea h because o
he po en ial adjudica ion imp ecision in his age g oup (due o high como bidi y a e
and low au opsy a e). A e exclusions, a o al o 5730 Heal h 2000 Su ey subjec s
we e a ailable o he p esen s udy. ECG measu emen s we e pe o med on sc een in a
blinded ashion by a single obse e wi h me hods desc ibed p e iously [16].
ECG LVH ol age ampli ude c i e ia known as Sokolow–Lyon, Co nell, R
aVL
,and
Pegue o–Lo P es i we e selec ed o he p esen s udy. Sokolow–Lyon ampli ude was cal-
cula ed as S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ); dicho omous cu o o LVH
was ≥3.5 mV [17]. Co nell ampli ude was calcula ed as S
V3
+R
aVL
; dicho omous cu o
o LVH was N2.0 mV in women and N2.8 mV in men [18]. R
aVL
ampli ude was measu ed
as R
aVL
; dicho omous cu o o LVH was N1.1 mV [19]. Pegue o–Lo P es i ampli ude was
calcula ed as he deepes S among all 12 leads + S
V4
; dicho omous cu o o LVH was
≥2.3 mV in women and ≥2.8 mV in men [20].
2.2. Follow-up and adjudica ion o he cause o dea h
Follow-up was om he Heal h 2000 Su ey baseline un il Decembe 31, 2013.
Adjudica ion o he cause o dea h was pe o med blinded o ECG da a as desc ibed
p e iously [21]. B iefly, adjudica ion was based on na ional egis e s o d ug eimbu se-
men , hospi al admission and discha ge diagnoses, and causes o dea hs. Ex ensi e
na ional egis e s a e main ained in Finland, and da a on all dea hs o Finnish ci izens
a e collec ed sys ema ically. Ou -o -hospi al dea hs and dea hs wi hin 10 days o hospi al-
iza ion we e conside ed eligible o he SCD adjudica ion. Da a om egis e s we e
analyzed independen ly by wo physicians and classified dea hs as p obable, possible,
unlikely SCD, and unknown cause o dea h. Dea hs in which a ca diac cause was he imme-
dia e o unde lying cause o dea h and he dea h was no known o be un ela ed o
a hy hmia we e defined as p obable SCDs. Dea hs in which he immedia e o unde lying
cause o dea h was nonca diac, bu ca diac disease was p esen and could easonably ha e
con ibu ed o a hy hmia based on mechanism (e.g., unexpec ed dea h due o aspi a ion
in a pa ien wi h a p io myoca dial in a c ion), o dea hs ha could ha e beena hy hmic
based on ci cums ances (e.g., dea h o a d i e in a mo o ehicle c ash, dea h while swim-
ming) we e defined as possible SCDs. Dea hs in which he e was an explained medical
cause o dea h un ela ed o ca diac disease (e.g., cance , massi e blood loss, sepsis, pulmo-
na y embolism, s oke) o a ca diac cause o dea h known o be nonsudden o un ela ed o
le hal a hy hmia (e.g., myoca dial up u e a e myoca dial in a c ion, endoca di is) we e
defined as unlikely SCDs. Dea hs wi h insu ficien da a we e defined as dea hs wi h
unknown cause. In case o disag eemen on he cause o dea h, wo addi ional indepen-
den physicians e iewed he case, and final decision was cons i u ed by consensus.
Au opsies we e pe o med in 67.2% o SCDs. In he p esen s udy, p obable and possible
SCDs we e pooled in he analyses and we e classified as SCDs.
2.3. S a is ical analysis
Analyses we e pe o med wi h R S a is ics ( e sion 3.4.0, The R Founda ion o S a is-
ical Compu ing, Vienna, Aus ia; used o Cox eg ession analyses and spline figu es),
STATA 13.0 (S a aCo p, College S a ion, TX; used o calcula e popula ion-a ibu able ac-
ions), and SPSS ( e sion 21, IBM, A monk, NY; used o all o he analyses). Values a e
gi en as mean ± SD o con inuous a iables, and pe cen ages and numbe s o ca ego ical
a iables. Bi a ia e co ela ions be ween con inuous LVH a iables we e es ed wi h
Pea son's es . Fi h's penalized maximum likelihood bias educ ion me hod o Cox e-
g ession (R package: coxph ) was used o ob ain he haza d a ios in he su i al analyses
[22,23]. Conco dance p obabili y es ima e was used o compa e he p edic i e accu acy o
he con inuous LVH a iables o SCD (R package: CPE) [24]. The alidi y o p opo ional
haza ds assump ion was e ified g aphically wi h pa ial esidual plo s (con inuous a i-
ables) and su i al p obabili y plo s (dicho omous a iables). LVH a iables we e used
bo h as con inuous and dicho omous. Selec ed composi e LVH c i e ia we e also used.
Gende , s udy baseline age, body mass index, hea a e, cu en smoking (yes/no), a e ial
hype ension (yes/no), p e ious myoca dial in a c ion (yes/no), and diabe es melli us
(yes/no) we e used as co a ia es in mul i a iable analyses. One LVH a iable was used
in he mul i a iable models a a ime. Defini ions o a e ial hype ension, p e ious myo-
ca dial in a c ion, and diabe es melli us ha e been published [16]. In e ac ion e m
(gende × LVH c i e ion) was es ed in he same model oge he wi h main e ec s
(i.e., LVH c i e ion, gende , all co a ia es). Cox models wi h penalized splines we e used
o assess and plo he ela ionship o con inuous LVH a iables o SCD isk. Popula ion-
a ibu able ac ions we e calcula ed o dicho omous ECG LVH c i e ia and selec ed
composi e c i e ia om models including all co a ia es wi h a p e iously desc ibed
me hod [25] and implemen ed as he STATA module puna cc. Popula ion-a ibu able
ac ion eflec s he p opo ion o e en s ha can be a ibu ed o a gi en isk ma ke o
he pe cen age o he cases ha would be p e en ed i a specificexposu ewe e obeelim-
ina ed om he popula ion. Two- ailed Pb0.05 was conside ed significan o all analyses.
3. Resul s
A e a mean ollow-up o 12.5 ± 2.2 yea s, 134 SCDs had occu ed.
Baseline cha ac e is ics a e shown in Table 1. Clinical a iables showed
ha , compa ed wi h subjec s wi hou SCD, subjec s wi h SCD we e
olde , mo e o en males, had highe body mass index and hea a e,
we e mo e o en smoke s, had mo e o en hype ension, p e ious myo-
ca dial in a c ion, and diabe es. ECG a iables showed ha , compa ed
wi h subjec s wi hou SCD, subjec s wi h SCD had longe QRS and QTc
du a ions, highe ol ages, and had mo e o en LVH. Depending on he
c i e ia, LVH p e alence a ied ma kedly. Pegue o–Lo P es i showed
he highes p e alence, and was ulfilled in ≈25% o all s udy subjec s.
Haza d a ios o SCD o con inuous LVH c i e ia a e shown in
Table 2. In mul i a iable adjus ed models, all c i e ia excep o R
aVL
emained significan ly associa ed wi h SCD. As an example, one milli-
me e inc ease in Co nell was associa ed wi h a 1.04- old (95% confi-
dence in e al [CI] 1.01–1.07, P= 0.008) isk o SCD. Conco dance
p obabili y es ima e was highes , 0.63 (95% CI 0.60–0.66), o Co nell,
while i was 0.56 (95% CI 0.52–0.60) o Sokolow–Lyon, 0.60 (95% CI
0.57–0.63) o Pegue o–Lo P es i, and 0.59 (95% CI 0.56–0.62) o R
aVL
.
When con inuous ECG LVH a iables we e used o es gende -LVH in-
e ac ion e ms in mul i a iable Cox models, in e ac ions we e no
ound (PN0.114 o all gende -LVH in e ac ion e ms). Adjus ed, con-
inuous associa ions be ween SCD isk and he h ee ol age c i e ia
ha we e significan ly associa ed wi h SCD in adjus ed models a e
shown in Supplemen a y Fig. I (Panels A–C). Compa ed o Sokolow–
Lyon and Pegue o–Lo P es i, inc ease in Co nell was associa ed wi h a
mo e p onounced SCD isk inc ease.
Haza d a ios o SCD o dicho omous LVH c i e ia a e shown in
Table 3. O he single c i e ia, only Co nell emained significan a e
adjus men s. O he composi e c i e ia, only composi e o Sokolow–
Lyon and Co nell emained significan a e adjus men s. O e lap o di-
cho omous Sokolow–Lyon, Co nell, and Pegue o–Lo P es i c i e ia is
shown in Fig. 1. O e lap be ween Sokolow–Lyon and Co nell was ela-
i ely small. By con as , Co nell became mos ly embedded by Pegue o–
Lo P es i. Bi a ia e co ela ion be ween con inuous Co nell and
Pegue o–Lo P es i c i e ia was high ( =0.71,Pb0.001), whe eas o he
co ela ions we e lowe (Sokolow–Lyon s. Co nell, = 0.20, Pb
0.001; Sokolow–Lyon s. Pegue o–Lo P es i, =0.21,Pb0.001).
When popula ion-a ibu able ac ions we e calcula ed o dicho o-
mous ECG LVH c i e ia and selec ed composi e c i e ia, popula ion-
a ibu able ac ion was 4.8% (95% CI –1.9–11.2%, P=0.158) o
Sokolow–Lyon,6.1%(95%CI–0.2–12.1%, P= 0.059) o Co nell,
9.5% (95% CI –2.3–20.0%, P= 0.111) o Pegue o–Lo P es i, 2.0%
(95% CI –3.1–6.8%, P= 0.442) o R
aVL
, 11.0% (95% CI 1.9–19.2%, P=
0.019) o composi e o Sokolow–Lyon and Co nell, 11.5% (95% CI –
2.1–23.3%, P= 0.095) o composi e o Sokolow–Lyon and Pegue o–
Lo P es i, and 8.3% (95% CI –3.7–18.9%, P= 0.168) o composi e o Co -
nell and Pegue o–Lo P es i. Thus, only composi e o Sokolow–Lyon and
Co nell was s a is ically significan .
4. Discussion
4.1. Main findings
Ou s udy in whi e gene al popula ion showed ha Sokolow–Lyon,
Co nell, and Pegue o–Lo P es i ECG ol age LVH c i e ia p o ided p ognos-
ic in o ma ion on SCD isk as con inuous a iables e en a e adjus ing o
126 K. Po han e al. / In e na ional Jou nal o Ca diology 276 (2019) 125–129
se e al isk ac o s, whe eas R
aVL
was no associa ed wi h SCD a e adjus -
men s. When single LVH c i e ia we e used as dicho omous a iables, only
Co nell was significan a e adjus men s. The dicho omous composi e o
Sokolow–Lyon and Co nell was also significan ly associa ed wi h SCD
a e adjus men s and was he only LVH measu e ha showed s a is ically
significan popula ion-a ibu able ac ion.
4.2. LVH as a modifie o SCD isk
Myoca dial ischemia is cen al in SCD, and co ona y disease has been
es ima ed o accoun o ≈80% all SCDs [26]. The second mos common
e iology o SCD, wi h a p opo ion o 10–15%, is (hype ophic and
dila ed) ca diomyopa hy [26]. Hype ophied myoca dium p edisposes
o malignan a hy hmias by a ious mechanisms. These include,
among o he s, educed co ona y blood flow p edisposing o ischemia;
ca diomyocy e loss and inc eased fib osis c ea ing a subs a e o elec-
ic een y and inc eased dispe sion o epola iza ion [27]; a hy hmo-
genic al e a ions in ca diomyocy e ion channel exp ession and unc ion
p edisposing o a e depola iza ions [28]. In hype ophic ca diomyop-
a hy, SCD isk has been epo ed o inc ease wi h he maximum le
en icula (LV) wall hickness [29]. Impo an ly, e en in he absence
o co ona y disease o ca diomyopa hy, inc ease in echo LV mass al-
eady wi hin no mal o mildly ele a ed ange is linea ly associa ed
wi h ad e se changes in ECG epola iza ion measu es [30]. Thus, in
he p esence o ischemia, SCD isk may be assumed o be modified, in
addi ion o o he ac o s (e.g., elec oly e dis u bances, gene ic ac o s),
by he deg ee o pa hological myoca dial hype ophy.
4.3. ECG LVH and ana omic LVH in ela ion o SCD isk
Inc ease in isk- ac o adjus ed haza d a io o SCD has been e-
po ed in he gene al popula ion bo h o ECG LVH [7,8,31] and echo
LVH [32–35]. O no e, LVH diagnosed by ECG s. echo o magne ic eso-
nance imaging con ains dis inc p ognos ic alue o SCD isk [8,10,11],
showing ha ECG LVH is a ma ke o ad e se elec ic emodelling e en
in he absence o ana omic hype ophy. Simula ion s udies indica e
ha elec ic p ope ies o he hea , especially slowed elec ic conduc-
ion eloci y (fib osis), may also inc ease ECG QRS ol ages and may
hus explain why many subjec s wi h ECG LVH do no p esen wi h an-
a omic LVH (and ice e sa), and why bo h ECG LVH and ana omic LVH
con ey p ognos ic alue independen o each o he [36]. Highligh ing
he impo ance o LVH, echo LVH has been epo ed as a leas equi a-
len o se e ely dec eased LV ejec ion ac ion as a p edic o o mo ali y
o SCD [37]. Recen ly, in an epidemiologic case-con ol s udy, he O e-
gon Sudden Unexpec ed Dea h S udy (O egon SUDS), an ECG isk
sco e which included ECG LVH (composi e o Sokolow–Lyon and Co -
nell ol ages) as a sco e componen esul ed in significan addi i e im-
p o emen abo e LV ejec ion ac ion in SCD isk es ima ion [38].
Despi e he confi med link be ween LVH and SCD, li le has been
known abou he p ognos ic alues o di e en ECG LVH c i e ia speci -
ically o SCD. The p ognos ic alues o di e en LVH c i e ia o inciden
ca dio ascula e en s a y [14], sugges ing ha some c i e ia may ou -
pe o m o he s in s a i ying specifically SCD isk. P e ious s udies
showing he ela ionship be ween ECG LVH and SCD ha e mainly e-
po ed one single LVH measu e. In one o he ea lies s udies, Minneso a
code 3–1 was used [31], whe eas in ano he epo om he 1970's he
defini ion o ECG LVH was no specified [7]. In he O egon SUDS,
Sokolow–Lyon ol age was used as a dicho omous ECG LVH a iable
and was associa ed wi h sudden ca diac a es also a e adjus ing o
o he isk ac o s, including echo LVH [8]. Mo e ecen ly in he O egon
SUDS, Romhil –Es es sco e ≥5(“defini e LVH”) was associa ed wi h sud-
den ca diac a es in he adjus ed model [39].
Ou p esen s udy may be he fi s comp ehensi e compa ison on
he pe o mance o se e al LVH c i e ia as isk p edic o s o SCD.
Th ee adi ional and clinically use ul and one ecen ly p oposed LVH
c i e ia we e analyzed bo h as con inuous and dicho omous. Da a
we e collec ed p ospec i ely. Fi s , as expec ed, significan associa ions
be ween adi ional isk ac o s and SCD was obse ed. Second, LVH
p e alence a ied ma kedly be ween c i e ia, which is no su p ising
as simila has been epo ed p e iously [6,14]. P e alence o Pegue o–
Lo P es i showed a ela i ely high p e alence (≈25% o all subjec s).
Ou s udy may be he fi s epo ing he p e alence o his new c i e ion
in he gene al popula ion. In hei wo k including mainly hype ensi e
hospi al pa ien s, Pegue o e al. epo ed ha , compa ed o Sokolow–
Lyon, Co nell, and R
aVL
,Pegue o–Lo P es i c i e ion had ma kedly highe
sensi i i y and lowe specifici y o de ec ing echo LVH [20]. Thus, ou
esul s showing highe LVH p e alence a es o Pegue o–Lo P es i
compa ed o o he c i e ia a e in line wi h he p e ious epo [20].
Table 1
Baseline cha ac e is ics o he s udy subjec s.
No SCD
(n= 5596)
SCD
(n= 134)
P alue
Clinical a iables
Age, yea s 51 ± 13 62 ± 10 b0.001
Male gende , %(n) 46(2547) 78(104) b0.001
Body mass index, kg/m
2
27 ± 5 28 ± 5 0.008
Hea a e, bea s pe minu e 63 ± 11 68 ± 14 b0.001
Cu en smoking, %(n) 22(1248) 43(58) b0.001
A e ial hype ension, %(n) 45(2503) 75(100) b0.001
P e ious myoca dial in a c ion, %(n) 2(103) 19(25) b0.001
Diabe es melli us, %(n) 5(290) 15(20) b0.001
ECG a iables
QRS du a ion, ms 93 ± 9 98 ± 14 b0.001
QTc du a ion, ms
a
409 ± 25 423 ± 29 b0.001
Sokolow−Lyon ol age, mV
b
2.5 ± 0.7 2.7 ± 0.9 0.022
Co nell ol age, mV
c
1.5 ± 0.6 1.8 ± 0.7 b0.001
Pegue o−Lo P es i ol age, mV
d
2.1 ± 0.7 2.5 ± 1.0 b0.001
R
aVL
ol age, mV
e
0.4 ± 0.3 0.6 ± 0.3 b0.001
ECG LVH by ol age c i e ia
Sokolow−Lyon, %(n)
9(482) 14(19) 0.030
Co nell, %(n)
g
7(386) 13(17) 0.016
Pegue o−Lo P es i, %(n)
h
25(1392) 34(45) 0.026
R
aVL
, %(n)
i
4(215) 8(11) 0.021
Values a e gi en as mean ± SD o con inuous a iables, and pe cen ages and numbe s o
ca ego ical a iables.
LVH indica es le en icula hype ophy; SCD, sudden ca diac dea h.
a
The maximum QT in e al o all leads wi h Baze 's o mula adjus men o hea a e.
b
S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ).
c
S
V3
+R
aVL
.
d
S
Deepes
+S
V4
.
e
R
aVL
.
S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ) ≥3.5 mV.
g
S
V3
+R
aVL
N2.0 mV (women), N2.8 mV (men).
h
S
Deepes
+S
V4
≥2.3 mV (women), ≥2.8 mV (men).
i
R
aVL
N1.1 mV.
Table 2
Haza d a ios o SCD o con inuous ECG LVH c i e ia om Cox models.
Model
a
LVH ol age c i e ia HR (95% CI)
b
P alue
Unadjus ed Sokolow–Lyon
c
1.03 (1.01–1.06) 0.003
Co nell
d
1.09 (1.06–1.12) b0.001
Pegue o–Lo P es i
e
1.05 (1.03–1.07) b0.001
R
aVL
1.12 (1.07–1.16) b0.001
Mul i a iable adjus ed Sokolow–Lyon 1.02 (1.003–1.05) 0.030
Co nell 1.04 (1.01–1.07) 0.008
Pegue o–Lo P es i 1.03 (1.01–1.05) 0.002
R
aVL
1.03 (0.98–1.09) 0.262
CI indica es confidence in e al; o he wise, abb e ia ions as in Table 1.
a
Co a ia es in he mul i a ia e analyses: gende , age a he s udy baseline, body mass
index, hea a e, cu en smoking (yes/no), a e ial hype ension (yes/no), diabe es
melli us (yes/no), p e ious myoca dial in a c ion (yes/no).
b
Haza d a io pe 100 μV (1 mm wi h he 10 mm/mV calib a ion) inc ease in he ECG
LVH measu e alue.
c
S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ).
d
S
V3
+R
aVL
.
e
S
Deepes
+S
V4.
R
aVL
.
127K. Po han e al. / In e na ional Jou nal o Ca diology 276 (2019) 125–129
Assuming ha echo LVH p e alence anges om 36% o 41% [40]inhy-
pe ension and ha echo is mo e sensi i e han ECG in de ec ing ana-
omic LVH, i seems plausible ha he cu o alue o Pegue o–Lo
P es i in he whi e adul gene al popula ion should be highe han
ha p oposed by Pegue o e al. [20] We do admi ha u u e s udies
a e needed o answe his in mo e de ail as da a on echo LVH we e
no a ailable in ou s udy.
Thi d, when LVH c i e ia we e analyzed as con inuous a iables in
mul i a iable adjus ed models, Sokolow–Lyon, Co nell, and Pegue o–
Lo P es i emained associa ed wi h SCD. By con as , R
aVL
los i s associ-
a ion wi h SCD. ECG Q, R, and S wa es a e belie ed o ep esen di e en
pa s o he depola izing a eas in he hea . Specifically, Q and R wa es
ep esen depola iza ion o he in e en icula sep um, conduc ion
sys em, and LV endomyoca dium, whe eas S wa es a e belie ed o
ep esen he depola iza ion o he en icula ee wall and myoca -
dium [20]. I has been pos ula ed ha , compa ed o S wa es, R wa es
a e less sensi i e in de ec ing mild o mode a e ana omic LVH [20]. In
con as o his, R
aVL
has also been epo ed o pe o m as good o
e en be e in echo LVH de ec ion compa ed o Sokolow–Lyon and
Co nell [5]. Risk- ac o adjus ed associa ion be ween R
aVL
and ca dio-
ascula ou comes has been epo ed [5,14], bu also conflic ing esul s
ha e been published [6]. We epo he e as a new finding ha R
aVL
as a
sole c i e ion does no seem o cap u e he malignan a hy hmogenic
isk o LVH in he gene al popula ion.
Fou h, when LVH c i e ia we e used in he p esen s udy as dicho -
omous a iables, ela ionships o Sokolow–Lyon and Pegue o–Lo P es i
o SCD we e nonsignifican in mul i a iable adjus ed models al hough
hey we e significan when used as con inuous a iables. One explana-
ion o his may be diminished s a is ical powe ela ed o dicho omi-
za ion. The composi e o Sokolow–Lyon and Co nell was he only
composi e c i e ion ha emained significan ly associa ed wi h SCD
a e adjus men s. In addi ion, composi e o Sokolow–Lyon and Co nell
was he only c i e ion ha showed s a is ically significan popula ion-
a ibu able ac ion. Use o composi e LVH c i e ia is gene ally
suppo ed because o expec ed enhanced sensi i i y in LVH de ec ion.
When c i e ia a e selec ed, a logical s ep is o combine c i e ia ha a e
bo h associa ed wi h ou come and also cap u e di e en pa ien ca ego-
ies, such as Sokolow–Lyon and Co nell. We ha e p e iously epo ed in
he Heal h 2000 Su ey ha composi e o Sokolow–Lyon and Co nell
ol ages pe o med he bes in p edic ing inciden ca dio ascula
e en s [14]. Resul s o he p esen s udy confi m ha composi e o
Sokolow–Lyon and Co nell is also alid when SCD is he endpoin .
4.4. Limi a ions
Ou s udy was pe o med in whi e gene al popula ion. Fu u e
s udies may e alua e whe he ou findings a e ep oducible also in
mul i acial popula ions as well as in popula ions wi h a mo e ele a ed
SCD isk, such as in co ona y hea disease and/o p e ious myoca dial
in a c ion.
4.5. Conclusions
Sokolow–Lyon, Co nell, and Pegue o–Lo P es i ECG, bu no R
aVL
ol age, a e associa ed wi h SCD isk as con inuous ECG ol age LVH
a iables. When SCD isk assessmen /adjus men is pe o med using a
dicho omous ECG LVH measu e, composi e o Sokolow–Lyon and
Co nell ol ages is he p e e ed op ion.
Supplemen a y da a o his a icle can be ound online a h ps://doi.
o g/10.1016/j.ijca d.2018.09.104.
Table 3
Haza d a ios o SCD o dicho omous ECG LVH c i e ia om Cox models.
Model
a
LVH ol age c i e ia HR (95% CI) PValue
Single c i e ia
Unadjus ed Sokolow–Lyon
b
1.79 (1.07–2.81) 0.027
Co nell
c
2.06 (1.20–3.31) 0.010
Pegue o–Lo P es i
d
1.53 (1.06–2.17) 0.023
R
aVLe
2.40 (1.24–4.20) 0.012
Mul i a iable adjus ed Sokolow–Lyon 1.55 (0.92–2.48) 0.096
Co nell 1.97 (1.12–3.29) 0.021
Pegue o–Lo P es i 1.40 (0.97–2.00) 0.072
R
aVL
1.36 (0.69–2.43) 0.346
Composi e c i e ia
Unadjus ed Sokolow–Lyon and/o Co nell 1.98 (1.32–2.89) 0.001
Sokolow–Lyon and/o Pegue o–Lo P es i 1.56 (1.10–2.19) 0.013
Co nell and/o Pegue o–Lo P es i 1.44 (1.003–2.05) 0.048
Mul i a iable adjus ed Sokolow–Lyon and/o Co nell 1.82 (1.20–2.70) 0.006
Sokolow–Lyon and/o Pegue o–Lo P es i 1.40 (0.98–1.99) 0.061
Co nell and/o Pegue o–Lo P es i 1.33 (0.92–1.91) 0.124
Abb e ia ions as in Tables 1 and 2.
a
Co a ia es in he mul i a ia e analyses as in Table 2.
b
S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ) ≥3.5 mV.
c
S
V3
+R
aVL
N2.0 mV (women), N2.8 mV (men).
d
S
Deepes
+S
V4
≥2.3 mV (women), ≥2.8 mV (men).
e
R
aVL
N1.1 mV.
Fig. 1. O e lap o dicho omous ECG LVH ol age c i e ia. Sokolow–Lyon: S
V1
+R
V5
o S
V1
+R
V6
(whiche e was g ea e ) ≥3.5 mV; Co nell: S
V3
+R
aVL
N2.0 mV (women), N2.8 mV
(men); Pegue o–Lo P es i: S
Deepes
+S
V4
≥2.3 mV (women), ≥2.8 mV (men).
128 K. Po han e al. / In e na ional Jou nal o Ca diology 276 (2019) 125–129
Conflic o in e es
The au ho s epo no ela ionships ha could be cons ued as a con-
flic o in e es .
Disclosu es
Veikko Salomaa has pa icipa ed in a con e ence ip sponso ed by
No o No disk. O he au ho s ha e no hing o disclose.
Acknowledgemen s
This wo k was suppo ed by g an s om he Academy o Finland
(T.K., p ojec numbe 309447), he Finnish Founda ion o Ca dio ascu-
la Resea ch (K.P., V.S.), he Finnish Medical Founda ion (K.P.), he O ion
Resea ch Founda ion (K.P., T.K.), and he Paulo Founda ion (T.K.).
Re e ences
[1] D. Le y, R.J. Ga ison, D.D. Sa age, W.B. Kannel, W.P. Cas elli, P ognos ic implica ions
o echoca diog aphically de e mined le en icula mass in he F amingham Hea
S udy, N. Engl. J. Med. 322 (1990) 1561–1566.
[2] D. Le y, M. Salomon, R.B. D'Agos ino, A.J. Belange , W.B. Kannel, P ognos ic implica-
ions o baseline elec oca diog aphic ea u es and hei se ial changes in subjec s
wi h le en icula hype ophy, Ci cula ion 90 (1994) 1786–1793.
[3] B.A. Vakili, P.M. Okin, R.B. De e eux, P ognos ic implica ions o le en icula hy-
pe ophy, Am. Hea J. 141 (2001) 334–341.
[4] M. Bombelli, R. Facche i, S. Ca ugo, F. Mado o, F. A ena e, F. Qua i-T e ano, A.
Cap a, C. Gianna asio, R. Dell'O o, G. G assi, R. Sega, G. Mancia, Le en icula hy-
pe ophy inc eases ca dio ascula isk independen ly o in-o fice and ou -o -o fice
blood p essu e alues, J. Hype ens. 27 (2009) 2458–2464.
[5] P. Gosse, E. Jan, P. Coulon, A. C eme , G. Papaioannou, S. Yeim, ECG de ec ion o le
en icula hype ophy: he simple , he be e ? J. Hype ens. 30 (2012) 990–996.
[6] C. Cuspidi, R. Facche i, M. Bombelli, C. Sala, G. G assi, G. Mancia, Accu acy and p og-
nos ic significance o elec oca diog aphic ma ke s o le en icula hype ophy in
a gene al popula ion: findings om he P essioni A e iose Moni o a e E Lo o
Associazioni popula ion, J. Hype ens. 32 (2014) 921–928.
[7] W.B. Kannel, J.T. Doyle, P.M. McNama a, P. Quicken on, T. Go don, P ecu so s o sud-
den co ona y dea h. Fac o s ela ed o he incidence o sudden dea h, Ci cula ion 51
(1975) 606–613.
[8] K. Na ayanan, K. Reinie , C. Teodo escu, A. Uy-E anado, H. Chugh, K. Gunson, J. Jui,
S.S. Chugh, Elec oca diog aphic e sus echoca diog aphic le en icula hype o-
phy and sudden ca diac a es in he communi y, Hea Rhy hm. 11 (2014)
1040–1046.
[9] W i ing G oup Membe s, D. Moza a ian, E.J. Benjamin, A.S. Go, D.K. A ne , M.J.
Blaha, M. Cushman, S.R. Das, S. de Fe an i, J.P. Desp és, H.J. Fulle on, V.J. Howa d,
M.D. Hu man, C.R. Isasi, M.C. Jiménez, S.E. Judd, B.M. Kissela, J.H. Lich man, L.D.
Lisabe h, S. Liu, R.H. Mackey, D.J. Magid, McGui e DK, E.R. Mohle 3 d, C.S. Moy, P.
Mun ne , M.E. Mussolino, K. Nasi , R.W. Neuma , G. Nichol, L. Palaniappan, D.K.
Pandey, M.J. Ree es, C.J. Rod iguez, W. Rosamond, P.D. So lie, J. S ein, A. Towfighi,
T.N. Tu an, S.S. Vi ani, D. Woo, R.W. Yeh, M.B. Tu ne , Ame ican Hea Associa ion
S a is ics Commi ee, S oke S a is ics Subcommi ee, Hea Disease and S oke S a-
is ics—2016 Upda e: a epo om he Ame ican Hea Associa ion, Ci cula ion 133
(2016) e38–360.
[10] J. Sunds öm, L. Lind, J. A nlö , B. Ze helius, B. And én, H.O. Li hell, Echoca dio-
g aphic and elec oca diog aphic diagnoses o le en icula hype ophy p edic
mo ali y independen ly o each o he in a popula ion o elde ly men, Ci cula ion
103 (2001) 2346–2351.
[11] L. Bacha o a, H. Chen, E.H. Es es, A. Ma easik, D.A. Bluemke, J.A. Lima, G.L. Bu ke, E.Z.
Soliman, De e minan s o disc epancies in de ec ion and compa ison o he p og-
nos ic significance o le en icula hype ophy by elec oca diog am and ca diac
magne ic esonance imaging, Am. J. Ca diol. 115 (2015) 515–522.
[12] P. Ve decchia, G. Schillaci, C. Bo gioni, A. Ciucci, R. Ga obigio, I. Zampi, C. Po cella i,
P ognos ic alue o a new elec oca diog aphic me hod o diagnosis o le en ic-
ula hype ophy in essen ial hype ension, J. Am. Coll. Ca diol. 31 (1998) 383–390.
[13] B.P. Hsieh, M.X. Pham, V.F. F oeliche , P ognos ic alue o elec oca diog aphic
c i e ia o le en icula hype ophy, Am. Hea J. 150 (2005) 161–167.
[14] K. Po han, T.J. Nii anen, J. Va is, I. Kan ola, H. Ka anko, M. Kähönen, M.S. Nieminen,
V. Salomaa, H.V. Huiku i, A.M. Jula, ECG le en icula hype ophy is a s onge
isk ac o o inciden ca dio ascula e en s in women han in men in he gene al
popula ion, J. Hype ens. 33 (2015) 1284–1290.
[15] S. Heis a o, Me hodology Repo , Heal h 2000 Su ey, Helsinki, Na ional Public
Heal h Ins i u e, 2008h p://u n.fi/URN:NBN:fi- e201204193320, Accessed da e:
14 June 2018.
[16] K. Po han, M. Vii asalo, A. Jula, A. Reunanen, J. Rapola, H. Väänänen, M.S. Nieminen,
L. Toi onen, V. Salomaa, L. Oika inen, P edic i e alue o elec oca diog aphic QT in-
e al and T-wa e mo phology pa ame e s o all-cause and ca dio ascula mo al-
i y in a gene al popula ion sample, Hea Rhy hm. 6 (2009) 1202–1208.
[17] M. Sokolow, T.P. Lyon, The en icula complex in le en icula hype ophy as
ob ained by unipola p eco dial and limb leads, Am. Hea J. 37 (1949) 161–186.
[18] P.N. Casale, R.B. De e eux, D.R. Alonso, E. Campo, P. Kligfield, Imp o ed sex-specific
c i e ia o le en icula hype ophy o clinical and compu e in e p e a ion o
elec oca diog ams: alida ion wi h au opsy findings, Ci cula ion 75 (1987)
565–572.
[19] J.A. Schack, R.H. Rosenman, L.N. Ka z, The aV limb leads in he diagnosis o en ic-
ula s ain, Am. Hea J. 40 (1950) 696–705.
[20] J.G. Pegue o, S. Lo P es i, J. Pe ez, O. Issa, J.C. B enes, A. Tolen ino, Elec oca dio-
g aphic c i e ia o he diagnosis o le en icula hype ophy, J. Am. Coll. Ca diol.
69 (2017) 1694–1703.
[21] P.A. Nosewo hy, A.S. Ha ulinna, K. Po han, A.M. Lah inen, A. Jula, P.J. Ka hunen, M.
Pe ola, L. Oika inen, K.K. Kon ula, V. Salomaa, C. New on-Cheh, Common gene ic
a ian s, QT in e al, and sudden ca diac dea h in a Finnish popula ion-based
s udy, Ci c. Ca dio asc. Gene . 4 (2011) 305–311.
[22] D. Fi h, Bias educ ion o maximum likelihood es ima es, Biome ika 80 (1993)
27–38.
[23] G. Heinze, M. Schempe , A solu ion o he p oblem o mono one likelihood in Cox
eg ession, Biome ics 57 (2001) 114–119.
[24] M. Gonen, G. Helle , Conco dance p obabili y and disc imina o y powe in p opo -
ional haza ds eg ession, Biome ika 92 (2005) 965–970.
[25] D. Spiegelman, E. He zma k, H.C. Wand, Poin and in e al es ima es o pa ial pop-
ula ion a ibu able isks in coho s udies: examples and so wa e, Cance Causes
Con ol 18 (2007) 571–579.
[26] H.V. Huiku i, A. Cas ellanos, R.J. Mye bu g, Sudden dea h due o ca diac a hy hmias,
N. Engl. J. Med. 345 (2001) 1473–1482.
[27] L. Choudhu y, H. Mah hold , A. Wagne , K.M. Choi, M.D. Ellio , F.J. Klocke, R.O.
Bonow, R.M. Judd, R.J. Kim, Myoca dial sca ing in asymp oma ic o mildly symp-
oma ic pa ien s wi h hype ophic ca diomyopa hy, J. Am. Coll. Ca diol. 40
(2002) 2156–2164.
[28] T. Fu ukawa, J. Ku okawa, Po assium channel emodeling in ca diac hype ophy, J.
Mol. Cell. Ca diol. 41 (2006) 753–761.
[29] P. Spi i o, P. Bellone, K.M. Ha is, P. Be nabo, P. B uzzi, B.J. Ma on, Magni ude o le
en icula hype ophy and isk o sudden dea h in hype ophic ca diomyopa hy,
N. Engl. J. Med. 342 (2000) 1778–1785.
[30] K. Po han, J. Vi olainen, T.P. Hil unen, M. Vii asalo, H. Väänänen, J. Dabek, T.
Hannila-Handelbe g, L. Toi onen, M.S. Nieminen, K. Kon ula, L. Oika inen, Rela ion-
ship o elec oca diog aphic epola iza ion measu es o echoca diog aphic le en-
icula mass in men wi h hype ension, J. Hype ens. 25 (2007) 1951–1957.
[31] B.N. Chiang, L.V. Pe lman, M. Ful on, L.D. Os ande J ., F.H. Eps ein, P edisposing ac-
o s in sudden ca diac dea h in Tecumseh, Michigan. A p ospec i e s udy, Ci cula-
ion 41 (1970) 31–37.
[32] A.W. Haide , M.G. La son, E.J. Benjamin, D. Le y, Inc eased le en icula mass and
hype ophy a e associa ed wi h inc eased isk o sudden dea h, J. Am. Coll. Ca diol.
32 (1998) 1454–1459.
[33] K. Reinie , C. De an, T. Singh, A. Uy-E anado, S. Lai, K. Gunson, J. Jui, S.S. Chugh, In-
c eased le en icula mass and dec eased le en icula sys olic unc ion ha e
independen pa hways o en icula a hy hmogenesis in co ona y a e y disease,
Hea Rhy hm. 8 (2011) 1177–1182.
[34] J.A. Laukkanen, H. Khan, S. Ku l, P. Willei , J. Ka ppi, K. Ronkainen, E. Di
Angelan onio, Le en icula mass and he isk o sudden ca diac dea h: a
popula ion-based s udy, J. Am. Hea Assoc. 3 (2014), e001285.
[35] S.H. Kone y, R.J. Koene, F.L. No by, T. Wilsdon, A. Alonso, D. Sisco ick, N.
So oodehnia, J. Go diene , E.R. Fox, L.Y. Chen, S. Adabag, A.R. Folsom, Echoca dio-
g aphic p edic o s o sudden ca diac dea h: he a he oscle osis isk in communi ies
s udy and ca dio ascula heal h s udy, Ci c. Ca dio asc. Imaging 9 (2016), e004431.
[36] L. Bacha o a, Le en icula hype ophy: disag eemen s be ween inc eased le
en icula mass and ECG-LVH c i e ia: he e ec o impai ed elec ical p ope ies
o myoca dium, J. Elec oca diol. 47 (2014) 625–629.
[37] S.M. S e ens, K. Reinie , S.S. Chugh, Inc eased le en icula mass as a p edic o o
sudden ca diac dea h: is i ime o pu i o he es ? Ci c. A hy hm. Elec ophysiol.
6(2013)212–217.
[38] A.L. A o, K. Reinie , C. Rusina u, A. Uy-E anado, N. Da ouian, D. Phan, W.J. Mack, J.
Jui, E.Z. Soliman, L.G. Te eshchenko, S.S. Chugh, Elec ical isk sco e beyond he le
en icula ejec ion ac ion: p edic ion o sudden ca diac dea h in he O egon Sud-
den Unexpec ed Dea h S udy and he A he oscle osis Risk in Communi ies S udy,
Eu . Hea J. 38 (2017) 3017–3025.
[39] N. Da ouian, A.L. A o, K. Na ayanan, A. Uy-E anado, C. Rusina u, K. Reinie , K.
Gunson, J. Jui, S.S. Chugh, The Romhil -Es es elec oca diog aphic sco e p edic s
sudden ca diac a es independen o le en icula mass and ejec ion ac ion,
Ann. Nonin asi e Elec oca diol. 22 (2017)h ps://doi.o g/10.1111/anec.12424.
[40] C. Cuspidi, C. Sala, F. Neg i, G. Mancia, A. Mo gan i, I alian Socie y o Hype ension,
P e alence o le - en icula hype ophy in hype ension: an upda ed e iew o
echoca diog aphic s udies, J. Hum. Hype ens. 26 (2012) 343–349.
129K. Po han e al. / In e na ional Jou nal o Ca diology 276 (2019) 125–129