RESEARCH ARTICLE Open Access
A 10- and 15-yea pe o mance analysis o
ESC/EAS and ACC/AHA ca dio ascula isk
sco es in a Sou he n Eu opean coho
Cá ia San os-Fe ei a
1*
, Rui Bap is a
1,2
, Manuel Oli ei a-San os
1
, José Pe ei a Mou a
3,4
and Lino Gonçal es
1,2
Abs ac
Backg ound: A key s a egy o he p ima y p e en ion o ca dio ascula disease (CVD) is he use o isk p edic ion
algo i hms. We aimed o in es iga e he p edic i e abili y o SCORE (Sys ema ic CO ona y Risk Es ima ion) and PCE
(Pooled Coho Equa ions) sys ems o a he oscle o ic CVD (ASCVD) isk in Po ugal, a low CVD isk coun y, a he
10-yea landma k and a a longe , 15-yea ollow-up.
Me hods: The SCORE and PCE 10-yea isk es ima es we e calcula ed o 455 and 448 pa ien s, espec i ely.
Disc imina ion was assessed by Ha ell’s C-s a is ic. Calib a ion was analyzed by s anda dized incidence a ios (SIR).
Resul s: Du ing he 10-yea ollow-up, 7 a al ASCVD e en s ( he SCORE ou come) and 32 any ASCVD e en s ( he
PCE ou come) occu ed. The SCORE sys em showed good disc imina ion (C-s a is ic 0.83), while he PCE showed
poo disc imina ion (C-s a is ic 0.62). Calib a ion was simila o bo h sys ems, acco ding o SIR: SCORE, 0.3 (95% CI
0.1–0.7); PCE, 0.5 (95% CI 0.4–0.7). Globally, bo h 10-yea a al ASCVD isk and any ASCVD isk we e o e es ima ed in
he o e all popula ion and men. Howe e , he isk was unde es ima ed by bo h sys ems in women. Despi e an
o e es ima ion o 15-yea a al ASCVD by SCORE, he 15-yea any ASCVD obse ed incidence was 1.8 imes he 10-
yea incidence among men and 1.4 imes among women. This accele a ion o CVD isk was mo e ele an in he
lowes classes o ASCVD isk.
Conclusion: In his p ospec i e, con empo a y, Po uguese coho , he SCORE had be e disc imina o y powe and
simila calib a ion compa ed o PCE. Howe e , bo h isk sco es unde es ima ed 10-yea ASCVD isk in women.
Keywo ds: A he oscle osis, Ca dio ascula isk, Guidelines, Lipids, Myoca dial in a c ion, S oke
Backg ound
Ca dio ascula disease (CVD) emains he leading cause
o mo bidi y and mo ali y in de eloped coun ies,
despi e consis en imp o emen in ou comes [1]. Modi i-
able isk ac o s a e he majo causes o a he oscle o ic
CVD (ASCVD), accoun ing o 90% o ca dio ascula
isk [2,3]. A key s a egy in he p ima y p e en ion o
ASCVD is he use o isk p edic ion algo i hms, as he
p ophylac ic in e en ions should a ge he people who
should bene i om hem mos [1,4]. Howe e , i is
deba able which is he bes algo i hm o ASCVD isk
es ima ion.
In 2003, he Eu opean Socie y o Ca diology (ESC)
join ly wi h he Eu opean A he oscle osis Socie y (EAS)
published he SCORE (Sys ema ic CO ona y Risk Es i-
ma ion): a 10-yea isk es ima ion sys em o ACSVD
dea h [5]. The SCORE P ojec pooled da a om 12
Eu opean p ospec i e coho s udies om ele en coun-
ies, including Sou he n Eu opean coun ies such as
Spain and I aly, bu no Po ugal [5]. Ex e nal alida ion
© The Au ho (s). 2020 Open Access This a icle is licensed unde a C ea i e Commons A ibu ion 4.0 In e na ional License,
which pe mi s use, sha ing, adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as long as you gi e
app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o he C ea i e Commons licence, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his a icle a e included in he a icle's C ea i e Commons
licence, unless indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no included in he a icle's C ea i e Commons
licence and you in ended use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed use, you will need o ob ain
pe mission di ec ly om he copy igh holde . To iew a copy o his licence, isi h p://c ea i ecommons.o g/licenses/by/4.0/.
The C ea i e Commons Public Domain Dedica ion wai e (h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he
da a made a ailable in his a icle, unless o he wise s a ed in a c edi line o he da a.
* Co espondence: [email p o ec ed]
1
Ca diology Uni , Cen o Hospi ala e Uni e si á io de Coimb a, Coimb a,
Po ugal
Full lis o au ho in o ma ion is a ailable a he end o he a icle
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301
h ps://doi.o g/10.1186/s12872-020-01574-2
has been pe o med o he SCORE model in se e al
Wes e n coun ies and in he Asian popula ion [6–10].
Howe e , calib a ion o he isk cha s acco ding o ca -
dio ascula mo ali y le els o each coun y has been
p oposed o a mo e p ecise p edic ion o isk es ima es
[5]. This is o u mos impo ance in coun ies no ep e-
sen ed in he de i a ion coho s, as Po ugal.
In pa allel, he Ame ican College o Ca diology
(ACC) and Ame ican Hea Associa ion (AHA) de el-
oped he Pooled Coho Equa ions (PCE) ha es i-
ma e a composi e endpoin o 10-yea ASCVD isk,
ins ead o he SCORE-es ima ed 10-yea a al ASCVD
isk [11]. Howe e , hese equa ions we e all de i ed
om No h Ame ican coho s [11], which limi hei
applicabili y o o he popula ions. Fu he , he PCE
ha e been con o e sial because o epo s ha hey
subs an ially o e es ima e isk in bo h Ame ican and
Eu opean popula ions [12–16].
Besides he di e en p edic ion models a ailable o es-
ima e he ASCVD isk, he decision h esholds ecom-
mended o d ug he apy a e di e en o , espec i ely,
a al ASCVD (SCORE) and any ASCVD (PCE) [17].
Such con lic ing ecommenda ions may c ea e con usion
among physicians, po en ially e lec ing unce ain y
abou he ex e nal alidi y o di e en algo i hms unde
di e en se ings [18], namely o low- isk coun ies, like
hose in Sou he n Eu ope [19,20]. In Po ugal, a coun y
wi h one o he lowes ASCVD e en a es in Eu ope
[20], nei he isk sys ems ha e been alida ed and doub s
conce ning hei clinical applicabili y and p edic i e ac-
cu acy in his pa icula popula ion a e aised.
In his s udy, we aimed o de e mine he 10- and 15-
yea incidence o ASCVD e en s in a p ima y p e en ion
coho o a Sou he n Eu opean coun y. In addi ion, we
aimed o compa e he calib a ion and disc imina ion o
bo h SCORE and PCE sys ems in his coho a he p e-
de ined 10-yea landma k and a a longe , 15-yea
ollow-up, in o de o es ima e hei ex e nal alidi y o
clinical p ac ice.
Me hods
S udy design and popula ion
We conduc ed a p ospec i e single-cen e s udy, includ-
ing 663 pa ien s consecu i ely e e ed by P ima y Ca e
physicians o Cen o Hospi ala e Uni e si á io de Coim-
b a’s Lipidology Clinic om 1994 o 2007. All pa ien s
we e ollowed un il Decembe 2017. Among ec ui ed
pa icipan s, we selec ed hose aged 40–79 yea s and
wi h no p e ious his o y o hea disease and s oke a
he sc eening isi . Pa ien s wi h amilial hype choles e -
olemia o wi h missing in o ma ion a he baseline
examina ion we e excluded, esul ing in 455 indi iduals
a ailable o his s udy.
Assessmen o isk ac o s
Baseline, demog aphic and clinical a iables we e
collec ed, including age a e e al, sex, hype ension,
diabe es and smoking s a us. Hype ension was de ined
as sys olic blood p essu e ≥140 mmHg and/o dias olic
blood p essu e ≥90 mmHg, o cu en an ihype ensi e
ea men . Diabe es was de ined as as ing glucose ≥126
mg.dL
−1
o he use o hypoglycemic d ugs. Smoking s a-
us was dicho omized as cu en e sus pas /ne e
smoke s. On he i s appoin men , se e al baseline la-
bo a o y a iables and a comple e lipid p o ile on as ing
blood samples we e collec ed.
Risk classi ica ion de ini ions
The p edic o s used o es ima e isk o a i s a al o
any ASCVD e en included age, sex, smoking s a us,
o al choles e ol (TC), high-densi y lipop o ein choles-
e ol (HDL-C), sys olic blood p essu e, diabe es, and an-
ihype ensi e ea men . SCORE and PCE sys ems a y
in he age anges o which hey apply. SCORE is ecom-
mended o use in he 40- o 65- yea age ange [1]. In
con as , he age ange in which PCE is alida ed is 40–
79 yea s [11]. Thus, we compa ed he pe o mance o
bo h isk sys ems in indi iduals aged 40–79 yea s. Base-
line SCORE (using he model o low- isk coun ies) [5]
and PCE [11] p edic ed isks we e calcula ed o each
pa ien . To comply wi h eal-li e clinical use o he
SCORE, he absolu e a al ASCVD isk o hose aged
66–79 yea s was co esponding o he isk a age 65 [21].
The coho was s a i ied in ou g oups based on he
SCORE isk ca ego ies: low (SCORE < 1%), mode a e
(SCORE 1–5%), high (SCORE 5–10% o diabe es wi h-
ou o gan damage o majo CVD isk ac o ), and e y
high (SCORE > 10% o diabe es wi h o gan damage o
majo CVD isk ac o ) [1]. Majo CVD isk ac o s we e
de ined by ma kedly ele a ed single isk ac o , namely
TC > 310 mg.dL
−1
and blood p essu e ≥180/110 mmHg
[1]. Rega ding he PCE isk, he coho was also s a i-
ied in ou isk ca ego ies: low (ASCVD isk < 5%), bo -
de line (ASCVD isk 5–7.5%), in e media e (ASCVD isk
7.5–20%), and high (ASCVD isk > 20%) [4].
Ou come a iables
In SCORE, he p edic ed ou come is a al ASCVD, com-
p ising he occu ence o dea h om co ona y a e y dis-
ease, s oke, hype ension, hea ailu e, pe iphe al a e y
disease o ao ic disease [5]. The endpoin de ined by PCE
is a i s ASCVD e en , including non a al myoca dial in-
a c ion, ASCVD dea h, and s oke [11]. The e en s
among he s udy coho we e iden i ied h ough eco d
linkage wi h he na ional heal h egis y and Heal h Da a
Pla o m. We asce ained non a al myoca dial in a c ion
o s oke e en s, de ined acco ding o he Ca dio ascula
and S oke Endpoin De ini ions o Clinical T ials [22].
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 2 o 9
Cases o a al ASCVD [22] we e asce ained om he
cause o dea h lis ed on dea h ce i ica es.
S a is ical analysis
A o al o 455 pa ien s had comple e in o ma ion on he
componen s o SCORE and 448 had comple e in o ma-
ion on he componen s o PCE. The median (in e qua -
ile ange (IQR)) ollow-up ime was 15 (11–17) yea s.
All pa ien s we e ollowed up o a leas 10 yea s and
55.4% (n= 252) we e ollowed o 15 yea s ega ding a al
ASCVD and 48.9% (n= 219) ega ding any ASCVD.
Con inuous a iables we e exp essed as mean ± s and-
a d de ia ion (SD). Median and IQR we e used i he
dis ibu ion was no no mal, assessed by he use o he
Kolmogo o -Smi no es . The S uden ’s - es o no -
mal a iables and he Mann-Whi ney es o non-
no mal a iables we e used o compa isons among
g oups (male s emale). Ca ego ical a iables we e p e-
sen ed as pe cen ages and we e compa ed using chi-
squa e o Fishe ’s exac es .
The p edic i e abili y o he SCORE and PCE sys ems
o he Sou he n Eu opean popula ion was e alua ed
based on disc imina ion and calib a ion o he models.
The disc imina i e powe was compa ed wi h Ha ell’sC-
s a is ics, which akes in o accoun he iming o e en s. A
calib a ion analysis was conduc ed by using s anda dized
incidence a ios (SIR), compa ing he 10-yea p edic ed
e en a e o he obse ed a e a 10-yea ollow-up. As all
pa ien s comple ed he 10-yea ollow-up, 10-yea ob-
se ed e en a e was he ac ual a e. Rega ding he cali-
b a ion analysis a he 15-yea ollow-up, he Kaplan–
Meie me hod was u ilized o es ima e he 15-yea ob-
se ed isk o each g oup. Bo h a es we e plo ed agains
he p edic ed isk, es ima ed as mean 10-yea isk sco e in
each g oup [23]. Pa icipan s we e censo ed a he ime o
he i s occu ence o ASCVD, dea h, ime o he las
ollow-up, o a 15 yea s o ollow-up.
Conce ning head- o-head compa ison, an in e sec ion
o he wo isk sys ems speci ic subg oups was c ea ed
(n= 448). The disc imina ion was e alua ed by Ha ell’s
C-s a is ic o each isk sys em-speci ic ou come. The
pai wise di e ences among he C-s a is ics co espond-
ing o he isk sys ems we e calcula ed, and, o compen-
sa e o mul iple es ing, boo s apping was used o
ob ain 99% con idence in e als (CI) [23].
All es s we e conduc ed using STATA 13.0 (S a a-
Co p, Texas, USA).
Resul s
The mean age ± SD o he s udy sample o 455 indi id-
uals was 57.8 ± 9.6 yea s and 61.8% we e male. The o e -
all baseline cha ac e is ics we e simila be ween gende s,
excep ha women had signi ican ly highe le els o
low-densi y lipop o ein choles e ol (LDL-C) and a la ge
p opo ion o s a in he apy (Table 1).
The a e age 10-yea isk es ima es by each o he isk
sys ems we e as ollows: he mean isk o ASCVD mo -
ali y was 4.7% (5.6% in men and 3.1% in women) ac-
co ding o SCORE and he mean ASCVD isk was 13.5%
(16.2% in men and 9.0% in women) acco ding o PCE,
globally indica ing a popula ion wi h mode a e ASCVD
isk (Table 1).
Disc imina ion
We assessed he abili y o SCORE and PCE o disc imin-
a e be ween pa ien s who de eloped e en s de ined by
SCORE ( a al ASCVD) and PCE (any ASCVD) and hose
who did no , a e calcula ing he 10-yea isk o each
pa ien . Du ing he 10-yea ollow-up, 7 (1.6%) SCORE-
speci ic e en s and 32 (7.0%) PCE-speci ic e en s oc-
cu ed (Table 1). The C-s a is ic co esponding o he
model wi h isk sco e as he only co a ia e we e 0.83
and 0.62 o he SCORE-speci ic and PCE-speci ic ou -
comes, espec i ely (Table 2).
Rega ding he 15-yea ollow-up, 5 (1.9%) ASCVD
dea hs and 25 (11.3%) any ASCVD e en s we e ob-
se ed. The C-s a is ic we e 0.80 o he SCORE-speci ic
ou comes and 0.63 o PCE-speci ic ou comes (Table 2).
In he head- o-head compa ison, while he SCORE
displayed he highes C-s a is ics o a al ASCVD, he
alues we e simila conside ing any ASCVD as he ou -
come (Fig. 1). No s a is ically signi ican di e ences we e
obse ed be ween he p edic i e abili y o he wo isk
sys ems (Table 3).
Calib a ion
Sco e
While 7 a al e en s occu ed du ing he 10-yea ollow-
up pe iod, SCORE p edic ed 21.4 e en s, which means
he e we e 66% mo e expec ed e en s han obse ed
(Fig. 2).
Disco dance be ween obse ed and expec ed cases was
ound h oughou he isk con inuum, wi h a p edomin-
ance o o e es ima ion in he o e all popula ion and men
by all o he isk s a a a 10-yea and 15-yea ollow-up.
Impo an ly, SCORE unde es ima ed isk in women, pa -
icula ly in hose a high isk (5 o 10%) (Fig. 2).
PCE
Globally, he PCE p edic ed 47% mo e any ASCVD
e en s (Fig. 2). Ac ually, he PCE o e es ima ed 10-yea
isk when he p edic ed isk was 5% o highe . Bo h in
he o e all popula ion and men, when he p edic ed isk
was low (< 5%), he isk was sligh ly unde es ima ed. Re-
ga ding women, PCE unde es ima ed isk in low (< 5%)
and in e media e (7.5–20%) isk s a a. In he 219 indi-
iduals wi h 15 yea s o ollow-up, an accele a ion o he
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 3 o 9
isk was obse ed, su passing he 10-yea p edic ed isk
in low (< 5%) and bo de line (5–7.5%) isk s a a in men
and low (< 5%) and in e media e (7.5–20%) isk s a a in
women (Fig. 2).
Discussion
In ou coho o Po uguese pa ien s, he e alua ion o
PCE and SCORE isk p edic ion algo i hms e ealed ha
bo h pe o med a an accep able le el. The SCORE sys em
pe o med be e han he PCE sys em in disc imina ing
hei espec i e endpoin s. Al hough calib a ion was simi-
la , bo h isk sys ems ma kedly o e es ima ed he isk.
While bo h isk sys ems add ess he same p ac ical
ques ion, which indi iduals bene i om in e en i e
measu es, i s equa ions di e signi ican ly. The
SCOREincludesasp edic o sage,sex,smoking,
sys olic blood p essu e, and TC; i is adi ionally
applicable in indi iduals om 40 o 65 yea s, and
he p edic ed ou come is 10-yea a al ASCVD,
Table 1 Baseline cha ac e is ics and obse ed e en s in a Po uguese coho aged 40–79 yea s in p ima y p e en ion
To al (n= 455) Male (n= 281) Female (n= 174) P alue
Age –yea s 57.8 ± 9.6 57.8 ± 9.5 57.8 ± 9.7 0.44
BMI –kg.m
−2
27.8 (25.8–30.8) 28.1 (25.9–30.2) 27.5 (25.2–31.8) 0.36
Sys olic BP –mmHg 135 (125–150) 135 (125–150) 135 (125–150) 0.95
Dias olic BP –mmHg 85 (80–90) 85 (80–90) 85 (80–90) 0.98
An ihype ensi e d ugs –no. (%) 260 (57.1) 156 (55.5) 104 (59.8) 0.37
C ea inine –mg.dL
−1
0.9 (0.8–1.1) 0.9 (0.8–1.1) 0.9 (0.8–1.1) 0.15
To al choles e ol –mg.dL
−1
268.0 (226.5–309.0) 263.0 (222.0–306.0) 273.0 (231.0–316.0) 0.05
HDL-C –mg.dL
−1
46.0 (38.0–58.0) 46.0 (38.0–57.0) 47.0 (39.0–59.5) 0.21
LDL-C –mg.dL
−1
158.4 ± 57.8 151.8 ± 56.6 170.4 ± 57.9 0.02
S a in ea men - no. (%) 298 (65.5) 173 (61.6) 125 (71.8) 0.03
Diabe es –no. (%) 76 (16.7) 49 (17.4) 27 (15.5) 0.59
Cu en smoke s –no. (%) 85 (18.7) 51 (18.1) 34 (19.5) 0.71
Mean calcula ed isk
SCORE - % 4.7 ± 6.8 5.6 ± 7.9 3.1 ± 4.0 < 0.001
PCE isk - % 13.5 ± 12.5 16.2 ± 13.7 9.0 ± 8.3 < 0.001
Median calcula ed isk
SCORE - % 2.7 (0.7–6.1) 3.6 (1.5–7.5) 1.0 (0.3–5.0) < 0.001
PCE isk - % 9.5 (4.5–18.3) 12.8 (6.6–22.1) 6.1 (3.1–11.9) < 0.001
Fa al ASCVD
10-yea –no. (%) 7 (1.6) 2 (0.7) 5 (2.9) 0.11
15-yea –no. (%) 5 (1.9) 1 (0.7) 4 (3.7) 0.15
Any ASCVD
10-yea –no. (%) 32 (7.0) 15 (5.4) 17 (9.8) 0.07
15-yea –no. (%) 25 (11.3) 12 (9.8) 13 (13.5) 0.71
Con inuous a iables we e exp essed as mean ± SD. Median and IQR we e used i he dis ibu ion was no no mal, assessed by he use o he Kolmogo o -
Smi no es . The S uden ’s - es o no mal a iables and he Mann-Whi ney es o non-no mal a iables we e used o compa isons among g oups. Ca ego ical
a iables we e p esen ed as pe cen ages and we e compa ed using chi-squa e o Fishe ’s exac es . 15-yea a al and any ASCVD we e Kaplan–Meie adjus ed
ASCVD A he oscle o ic ca dio ascula disease, BMI Body mass index, BP Blood p essu e, HDL-C High densi y lipop o ein choles e ol, IQR in e qua ile ange, LDL-C
Low-densi y lipop o ein choles e ol, PCE Pooled Coho Equa ions, SCORE Sys ema ic CO ona y Risk Es ima ion, SD s anda d de ia ion
Table 2 Ha ell’s C-s a is ic SCORE and PCE in p edic ing a al
ASCVD e en s and any ASCVD e en s in a Po uguese coho
aged 40–79 yea s in p ima y p e en ion, espec i ely
OVERALL Men Women
C
SCORE
( a al ASCVD)
10-yea (n = 455) 0.83 (0.74–0.93) 0.87 (0.69–1.00) 0.87 (0.78–0.96)
15-yea (n = 252) 0.80 (0.70–0.90) 0.86 (0.67–1.00) 0.84 (0.75–0.93)
C
PCE
(any ASCVD)
10-yea (n = 448) 0.62 (0.53–0.71) 0.66 (0.52–0.79) 0.67 (0.56–0.79)
15-yea (n = 219) 0.63 (0.55–0.70) 0.64 (0.53–0.75) 0.67 (0.57–0.78)
The p edic i e abili y o he isk sco es was assessed by Ha ell’s C-s a is ic o
isk sys em-speci ic ou comes
ASCVD A he oscle o ic ca dio ascula disease, PCE Pooled Coho Equa ions,
SCORE Sys ema ic CO ona y Risk Es ima ion
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 4 o 9
dis ega ding non- a al e en s [5]. Besides he p edic-
o s included in SCORE, PCE also inco po a es e h-
nici y, HDL-C, an ihype ensi e use, and diabe es.
Fu he mo e, PCE has a la ge age ange (40–79
yea s) and p edic s he 10-yea isk o a al and
non- a al ASCVD [11].
Exis ing s udies ega ding he SCORE isk p edic ion
sys em gene ally epo ed good disc imina ion, consis -
en wi h ou indings [5,8,9,24]. Conside ing he
SCORE sys em, ou s udy epo ed a a o able C-
s a is ic o 0.83. The disc imina ion pe o mance o he
SCORE isk o p edic 10-yea ASCVD mo ali y was
e alua ed in he SCORE p ojec and he C-s a is ic
alues anged om 0.71 o 0.84 [5]. In he global popu-
la ion, we obse ed a sys ema ic isk o e es ima ion ha
has also been p e iously epo ed in low- isk popula-
ions, as in Spain [8,9], and high- isk popula ions [24].
The ex e nal alida ion s udies ha e alua e he PCE
sys em ha e gene a ed con o e sy a ound he p edic i e
accu acy o hese equa ions in con empo a y coho s
[25]. The PCE models disc imina ed isk easonably (C-
s a is ic 0.62) in ou coho , bu pe o ming wo se in
compa ison o o he s udies [11,15,16]. The o iginal
PCE C-s a is ic anged om 0.71 o 0.82 [11] and 0.74
o 0.79 in con empo a y coho s [16]. Simila o he
SCORE, we obse ed a isk o e es ima ion in ou popu-
la ion, wi h an excep ion in p edic ing ASCVD isk in
low- isk s a a (< 5%). P e ious li e a u e es ablished ha
he PCE gene ally o e es ima e isk among mode n
Eu opean and Ame ican popula ions [13,15,16,23,26].
Al hough he e a e no epo s o unde es ima ion o
ASCVD isk by he PCE, i has been widely ecognized
ha a la ge numbe o people wi h ASCVD isk p edic-
ion < 7.5% will pa adoxically expe ience ASCVD e en s
[27], ein o cing he alue o addi ional isk ma ke s o
eclassi y hose pa ien s [4,27].
Accu a e calib a ion o he isk p edic ion models is
c ucial o he success o p e en i e measu es. Al hough
bo h models o e es ima ed he isk, when obse ed isks
we e aken in o accoun , he accu acy o isk p edic ion
in women was less imp essi e. Ac ually, he isk was
unde es ima ed by he PCE isk sys em in he low- isk
(< 5%) and in e media e- isk (7.5–20%) s a a. Unde es i-
ma ion o ASCVD isk in he emale gende has only
been desc ibed once p e iously in a Ko ean coho [28].
Conce ning he SCORE sys em, he in e ac ion
be ween sex and e en s should be analyzed as an
explo a o y and hypo hesis-gene a ing inding, due o
he e y low numbe o 10-yea a al ASCVD e en s
(n= 7), especially in men (n= 2). This low ASCVD-
mo ali y a e is consis en wi h wha has been ound
in a Spanish coho o 608 non-diabe ic pa ien s on
Fig. 1 ROC cu es o he SCORE and PCE models o p edic ion o 10-yea e en s in a Po uguese coho aged 40–79 yea s in p ima y p e en ion.
aFa al ASCVD e en s and bany ASCVD e en s. The p edic i e abili y o he isk sco es was assessed by Ha ell’s C-s a is ic o isk sys em-speci ic
ou comes. ASCVD: a he oscle o ic ca dio ascula disease, SCORE: Sys ema ic CO ona y Risk E alua ion, PCE: pooled coho equa ions
Table 3 Head o head compa ison o he isk sco es acco ding
o Ha ell’s C-s a is ic in a Po uguese coho aged 40–79 yea s
in p ima y p e en ion (n= 448)
C
SCORE
C
PCE
△(C
SCORE
-C
PCE
)
(99% CI)
P alue
Fa al ASCVD
n = 7 10-YEAR 0.83 0.78 0.05 (−0.05–0.15) 0.32
n= 5 15- YEAR 0.80 0.75 0.05 (−0.04–0.15) 0.27
Any ASCVD
n= 32 10- YEAR 0.62 0.62 0.00 (−0.06–0.06) 0.99
n = 25 15- YEAR 0.63 0.62 0.00 (−0.06–0.04) 0.72
The p edic i e abili y o he isk sco es was assessed by Ha ell’s C-s a is ic o
each isk sys em-speci ic ou come. The pai wise di e ences among he C-
s a is ics co esponding o he isk sys ems we e calcula ed, and, o
compensa e o mul iple es ing, boo s apping was used o ob ain 99%
con idence in e als (CI)
ASCVD A he oscle o ic ca dio ascula disease, CI Con idence in e al, PCE
Pooled Coho Equa ions, SCORE Sys ema ic CO ona y Risk Es ima ion
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 5 o 9

p ima y p e en ion wi h a simila mode a e 10-yea
isk o a al ASCVD (2.1%), whe e only 9 ASCVD
dea hs we e epo ed (1.5%) [8]. Ne e heless,
women’s isk was unde es ima ed in high- isk women
(5–10%). An unde es ima ion o he isk by he
SCORE sys em in women has al eady been epo ed
in Aus alian women below 50 yea s [6], and in a Ma-
laysian popula ion [7]. Con e sely, in ano he Sou h-
e n Eu opean coho om I aly, he 10-yea a al
ASCVD isk was p ope ly p edic ed by SCORE [29].
Howe e , he low- isk (< 1%) women om his coho
had a 20-yea a e o any ASCVD o 3.7, and 40% o
ASCVD e en s occu ed in his class, which signals
po en ial p oblems wi h he calib a ion o his equa-
ion in women [29].
The causes o he unexpec ed indings ega ding he
unde p edic ion o a al o any ASCVD isk in women
and he e y low numbe o ASCVD dea hs in men a e
Fig. 2 Calib a ion accu acy e alua ed by compa ing obse ed and p edic ed e en s and s anda dised incidence a ios in a Po uguese coho
aged 40–79 yea s in p ima y p e en ion. A calib a ion analysis was conduc ed by using SIR, compa ing he 10-yea p edic ed e en a e o he
obse ed a e a 10- and 15-yea ollow-up. As all pa ien s comple ed he 10-yea ollow-up, 10-yea obse ed e en a e was he ac ual a e. 15-
yea obse ed e en s we e Kaplan–Meie adjus ed. SIR, s anda dised incidence a ios, SCORE: Sys ema ic CO ona y Risk E alua ion; PCE: pooled
coho equa ions
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 6 o 9
no clea . Al hough women om ou coho had highe
le els o LDL-C, hey we e also ea ed mo e commonly
wi h s a ins. In he pas , nume ous s udies ha e shown
excess mo ali y in women a e myoca dial in a c ion
[30]. In addi ion, women su e ing an acu e co ona y
synd ome a e ea ed less in ensi ely han men [31],
and, e en on p ima y p e en ion, i has been obse ed a
lowe use o lipid-lowe ing d ugs in women [32]. Inde-
penden ly o he cause o he excess isk o ASCVD
e en s in women, ou esul s sugges ha a leas in ou
popula ion, he isk s a i ica ion sys ems migh need o
be be e calib a ed among women, in o de o educe
he p opo ion o women misca ego ized and o a oid
CVD e en s by implemen ing app op ia e p e en i e
measu es [5].
Las ly, he SCORE o e es ima ed he 15-yea isk in
he o e all popula ion, p obably e lec ing he decline o
CVD mo ali y in he las decades [33]. Con e sely, e-
ga ding PCE, we obse ed an impo an inc ease in any
ASCVD e en s a 15-yea ollow-up. The obse ed 15-
yea any ASCVD incidence was 1.8 imes he 10-yea in-
cidence among men and 1.4 imes among women. This
accele a ion o ASCVD isk is e en mo e ele an in he
low- and bo de line- isk s a a in men and he low- and
in e media e- isk s a a in women, as he obse ed 15-
yea isk su passes he 10-yea p edic ed isk. A di e -
gence be ween sho - ime and long- e m isk has al eady
been desc ibed in wo I alian s udies [29,34]. Among
he ela i ely la ge g oup wi h low and in e media e
sho - e m isk, di e ences in isk ac o bu den seem o
ansla e in o ma ked di e ences in he incidence o
ASCVD e en s in he emaining li espan [35]. The added
alue o long- e m isk es ima ion becomes e en mo e
ele an in women and younge men, mo e p one o be
classi ied as being a low sho - e m isk [36].
I is essen ial o e alua e he applicabili y o isk
models o each popula ion, as isk sco es may pe o m
wo se in a di e en se ing om he one hey we e o i-
ginally de i ed [37]. I is well-known ha he incidence
o CVD e en s is declining, and Po ugal is no excep ion.
Despi e he high p e alence o hype ension and o he
ch onic diseases, Po ugal has one o he bes indica o s
o ca dio ascula mo ali y [20], namely age- and isk-
s anda dized mo ali y a es conce ning co ona y a e y
disease and ce eb o ascula disease [19]. This p oblem is
well- ecognized and he li e a u e ecommends ha
sco es should be ecalib a ed i incidence a es di e
subs an ially in he new se ings [5,13], as we ha e p e-
iously discussed o women. Addi ionally, newe s a is-
ical me hods could conside ably imp o e he accu acy
o ASCVD isk es ima es [16]. Ano he , mo e s aigh -
o wa d op ion o imp o e i s p edic i e abili y would be
he inco po a ion o addi ional isk a iables such as
co ona y a e y calcium sco e and high-sensi i i y C-
eac i e p o ein [4,27] o conside he li e ime isk o
CVD in addi ion o sho - ime (10-yea ) isk [4].
We acknowledge se e al limi a ions o ou s udy. We
ha e a limi ed sample size, ep esen a i e o he Cen e e-
gion o Po ugal, one o he egions wi h he lowes a e o
CVD e en s in ou coun y [20]. This may explain pa -
ially he o e es ima ion o he isk by bo h sys ems, and
ex apola ion o ou esul s o o he popula ions should
be done cau iously. In addi ion, only hal o he pa ien s
had a comple e 15-yea ollow-up, as many pa ien s we e
included in his coho a e 2003. Also, we canno o e -
look ha wo- hi ds o ou coho was unde s a in he -
apy; howe e , he majo i y o he pa ien s we e unde low-
and mode a e- in ensi y s a ins as hey we e ec ui ed be-
o e he e a o high-in ensi y s a ins ecommenda ion and
a e ep esen a i e o a eal-wo ld popula ion seen in
e e yday p ac ice. In e es ingly, in a sensi i i y analysis o
he Copenhagen Gene al Popula ion S udy, including s a-
in use s a baseline (7% o he popula ion) did no explain
he obse ed misma ch be ween he p edic ed and ob-
se ed e en a es [21]. Finally, we we e also no able o
adjus o p esc ip ion o o he p e en i e medica ion du -
ing ollow-up. This migh also ha e con ibu ed o some
o he obse ed misma ch. A s eng h o ou s udy is ha
he esul s o igina e om he la ges con empo a y Po u-
guese coho e alua ed so a , wi h no pa ien los o
ollow-up.
Conclusions
In his p ospec i e, con empo a y, Po uguese coho ,
SCORE had be e disc imina o y powe compa ed wi h
PCE. O e all, bo h sys ems o e es ima ed ASCVD isk
a he 10-yea landma k. Howe e , among women,
ASCVD isk was unde es ima ed wi h bo h isk sys ems.
Las ly, we also obse ed an accele a ion o he ca dio as-
cula isk a 15-yea ollow-up, pa icula ly a he lowes
isk classes. These indings may ha e he po en ial o im-
p o e isk s a i ica ion and, ul ima ely, ea men alloca-
ion in a p ima y p e en ion se ing.
Abb e ia ions
ACC: Ame ican college o ca diology; AHA: Ame ican Hea Associa ion;
ASCVD: A he oscle o ic ca dio ascula disease; CVD: Ca dio ascula disease;
EAS: Eu opean a he oscle osis socie y; ESC: Eu opean socie y o ca diology;
HDL-C: High-densi y lipop o ein choles e ol; LDL-C: Low-densi y lipop o ein
choles e ol; PCE: Pooled coho Eqs.; TC: To al choles e ol; SCORE: Sys ema ic
CO ona y isk es ima ion; SD: S anda d de ia ion; SIR: S anda dized incidence
a ios
Acknowledgemen s
We hank Ad iana Belo o assis ance wi h he s a is ical analysis.
Au ho s’con ibu ions
CSF, RB, MOS, JPM, LG designed he s udy; CSF, RB, MOS acqui ed he da a;
CSF, RB, MOS pe o med he s a is ical analysis and in e p e ed he esul s;
CSF d a ed he i s e sion o he manusc ip ; CSF, RB, MOS, JPM, LG ead
and c i ically e iewed he inal e sion o he manusc ip . All au ho s ead
and app o ed he inal manusc ip .
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 7 o 9
Funding
This wo k was suppo ed by g an POCI-01-0145-FEDER-032414. The unding
sou ce was no in ol ed in he design o he s udy, collec ion, analysis, and
in e p e a ion o da a o w i ing o he manusc ip .
A ailabili y o da a and ma e ials
The da ase s used and/o analyzed du ing he cu en s udy a e a ailable
om he co esponding au ho on easonable eques .
E hics app o al and consen o pa icipa e
W i en in o med consen was ob ained om each pa ien and he s udy
was conduc ed in acco dance wi h he Decla a ion o Helsinki and app o ed
by he local e hics commi ee (Comissão de É ica do Cen o Hospi ala e
Uni e si á io de Coimb a).
Consen o publica ion
No applicable.
Compe ing in e es s
The au ho s decla e ha hey ha e no compe ing in e es s.
Au ho de ails
1
Ca diology Uni , Cen o Hospi ala e Uni e si á io de Coimb a, Coimb a,
Po ugal.
2
Coimb a Ins i u e o Clinical and Biomedical Resea ch (iCBR),
Facul y o Medicine, Uni e si y o Coimb a, Coimb a, Po ugal.
3
In e nal
Medicine Uni , Cen o Hospi ala e Uni e si á io de Coimb a, Coimb a,
Po ugal.
4
Facul y o Medicine, Uni e si y o Coimb a, Coimb a, Po ugal.
Recei ed: 6 Ap il 2020 Accep ed: 3 June 2020
Re e ences
1. Piepoli MF, Hoes AW, Agewall S, Albus C, B o ons C, Ca apano AL, e al.
2016 Eu opean guidelines on ca dio ascula disease p e en ion in clinical
p ac ice. Eu Hea J. 2016;37(29):2315–81.
2. Yusu S, Hawken S, Ôunpuu S, Dans T, A ezum A, Lanas F, e al. E ec o
po en ially modi iable isk ac o s associa ed wi h myoca dial in a c ion in
52 coun ies ( he INTERHEART s udy): case-con ol s udy. Lance . 2004;364:
937–52.
3. Mu ay CJL, Lopez AD. Measu ing he global bu den o disease. N Engl J
Med. 2013;369(5):448–57.
4. A ne DK, Blumen hal RS, Albe MA, Bu oke AB, Goldbe ge ZD, Hahn EJ,
e al. 2019 ACC/AHA Guideline on he P ima y P e en ion o Ca dio ascula
Disease: Execu i e Summa y: A Repo o he Ame ican College o
Ca diology/Ame ican Hea Associa ion Task Fo ce on Clinical P ac ice
Guidelines. J Am Coll Ca diol. 2019;1097(19):33876–8 S0735.
5. Con oy RM, Pyö älä K, Fi zge ald AP, Sans S, Meno i A, De Backe G, e al.
Es ima ion o en-yea isk o a al ca dio ascula disease in Eu ope: he
SCORE p ojec . Eu Hea J. 2003;24(11):987–1003.
6. Goh LGH, Welbo n TA, Dhaliwal SS. Independen ex e nal alida ion o
ca dio ascula disease mo ali y in women u ilising F amingham and SCORE
isk models: a mo ali y ollow-up s udy. BMC Womens Heal h. 2014;14(1):1–11.
7. Sel a ajah S, Kau G, Hani J, Cheong KC, Hiong TG, an de G aa Y, e al.
Compa ison o he F amingham isk SCORE, SCORE and WHO/ISH
ca dio ascula isk p edic ion models in an Asian popula ion. In J Ca diol.
2014;176(1):211–8.
8. Ba oso LC, Mu o EC, He e a ND, Ochoa GF, Hue os JI, Bui ago F.
Pe o mance o he F amingham and SCORE ca dio ascula isk p edic ion
unc ions in a non-diabe ic popula ion o a Spanish heal h ca e Cen e: a
alida ion s udy. Scand J P im Heal h Ca e. 2010;28(4):242–8.
9. Bui ago F, Cañón-Ba oso L, Díaz-He e a N, C uces-Mu o E, Escoba -
Fe nández M, Se ano-A ias JM. Compa ación de las ablas REGICOR y
SCORE pa a la clasi icación del iesgo ca dio ascula y la iden i icación de
pacien es candida os a a amien o hipolipemian e o an ihipe ensi o. Re
Esp Ca diol. 2007;60:139–47.
10. Sawano M, Kohsaka S, Okamu a T, Inoha a T, Sugiyama D, Wa anabe M,
e al. Valida ion o he eu opean SCORE isk cha in he heal hy middle-
aged Japanese. A he oscle osis. 2016;252:116–21.
11. Go DC, Lloyd-Jones DM, Benne G, Coady S, D'Agos ino RB S , Gibbons R,
e al. 2013 ACC/AHA guideline on he assessmen o ca dio ascula isk: a
epo o he Ame ican college o ca diology/Ame ican hea associa ion
ask o ce on p ac ice guidelines. Ci cula ion. 2014;129(25 SUPPL. 1):49–73.
12. Ka ousi M, Leening MJG, Nanchen D, G eenland P, G aham IM, S eye be g
EW, e al. Compa ison o applica ion o he ACC/AHA guidelines, adul
ea men panel III guidelines, and Eu opean Socie y o Ca diology
Guidelines o ca dio ascula disease p e en ion in a Eu opean coho .
JAMA. 2014;311(14):1416.
13. Mo a S, Wenge NK, Cook NR, Liu J, Howa d BV, Limache MC, e al.
E alua ion o he pooled coho isk equa ions o ca dio ascula isk
p edic ion in a mul ie hnic coho om he Women’s heal h ini ia i e. JAMA
In e n Med. 2018;02215:1–10.
14. Cook NR, Ridke PM. Fu he insigh in o he ca dio ascula isk calcula o
con o e sy: he oles o s a ins, e ascula iza ions, and unde -asce ainmen
in he Women’s heal h s udy. JAMA In e n Med. 2014;174(12):1964–71.
15. De ilippis AP, Young R, Ca ubba CJ, McE oy JW, Budo MJ, Blumen hal RS,
e al. An analysis o calib a ion and disc imina ion among mul iple
ca dio ascula isk sco es in a mode n mul ie hnic coho . Ann In e n Med.
2015;162(4):266–75.
16. Yadlowsky S, Haywa d RA, Sussman JB, McClelland RL, Min YI, Basu S.
Clinical implica ions o e ised pooled coho equa ions o es ima ing
a he oscle o ic ca dio ascula disease isk. Ann In e n Med. 2018;169(1):20–9.
17. Ray KK, Kas elein JJP, Ma hijs Boekhold S, Nicholls SJ, Khaw KT,
Ballan yne CM, e al. The ACC/AHA 2013 guideline on he ea men o
blood choles e ol o educe a he oscle o ic ca dio ascula disease isk in
adul s: he good he bad and he unce ain: a compa ison wi h ESC/
EAS guidelines o he managemen o dyslipidaemias 2011. Eu Hea J.
2014;35(15):960–8.
18. Pennells L, Kap oge S, Wood A, Swee ing M, Zhao X, Whi e I, e al.
Equaliza ion o ou ca dio ascula isk algo i hms a e sys ema ic
ecalib a ion: indi idual-pa icipan me a-analysis o 86 p ospec i e s udies.
Eu Hea J. 2019;40(7):621–31.
19. GBD 2015 Heal hca e Access and Quali y Collabo a o s. Heal hca e access
and quali y index based on mo ali y om causes amenable o pe sonal
heal h ca e in 195 coun ies and e i o ies, 1990–2015: A no el analysis
om he global bu den o disease s udy 2015. Lance . 2017;390(10091):231–66.
20. Chipman A. Global hea heal h: e alua ing e o s o p omo e heal hy
hea s. In: The Economis In elligence Uni . 2018. A ailable a : h ps://
eiupe spec i es.economis .com/si es/de aul / iles/Globalhea heal h-1.pd .
Accessed 6 Ap 2020.
21. Mo ensen MB, No des gaa d BG, A zal S, Falk E. ACC/AHA guidelines
supe io o ESC/EAS guidelines o p ima y p e en ion wi h s a ins in non-
diabe ic Eu opeans: he Copenhagen gene al popula ion s udy. Eu Hea J.
2017;38(8):586–94.
22. Hicks KA, Maha ey KW, Meh an R, Nissen SE, Wi io SD, Dunn B, e al. 2017
ca dio ascula and s oke endpoin de ini ions o clinical ials. Ci cula ion.
2018;137(9):961–72.
23. Saa A, Läll K, Al e M, Ma andi T, Ainla T, Eha J, e al. Es ima ing he
pe o mance o h ee ca dio ascula disease isk sco es: he Es onian
biobank coho s udy. J Epidemiol Communi y Heal h. 2019;73(3):272-7.
24. Vikhi e a O, Pajak A, B oda G, Malyu ina S, Tamosiunas A, Kubino a R, e al.
SCORE pe o mance in cen al and eas e n Eu ope and o me So ie Union:
MONICA and HAPIEE esul s. Eu Hea J. 2014;35(9):571–7.
25. Rana JS, Tabada GH, Solomon MD, Lo JC, Ja e MG, Sung SH, e al. Accu acy
o he a he oscle o ic ca dio ascula isk equa ion in a la ge con empo a y,
Mul ie hnic Popula ion. J Am Coll Ca diol. 2016;67(18):2118–30.
26. Cook NR, Ridke PM. Calib a ion o he pooled coho equa ions o
a he oscle o ic ca dio ascula disease: an upda e. Ann In e n Med. 2016;
165(11):786–94.
27. Yeboah J, Young R, McClelland RL, Delaney JC, Polonsky TS, Dawood FZ,
e al. U ili y o non adi ional isk ma ke s in a he oscle o ic ca dio ascula
disease isk assessmen . J Am Coll Ca diol. 2016;67(2):139–47.
28. Jung KJ, Jang Y, Oh DJ, Oh BH, Lee SH, Pa k SW, e al. The ACC/AHA 2013
pooled coho equa ions compa ed o a Ko ean isk p edic ion model o
a he oscle o ic ca dio ascula disease. A he oscle osis. 2015;242(1):367–75.
29. Ve onesi G, Giampaoli S, Vanuzzo D, Gian agna F, Palmie i L, G assi G, e al.
Combined use o sho - e m and long- e m ca dio ascula isk sco es in
p ima y p e en ion. J Ca dio asc Med. 2017;18(5):318–24.
30. Lawesson SS, Al edsson J, F ed ikson M, Swahn E. Time ends in STEMI-
imp o ed ea men and ou come bu s ill a gende gap: A p ospec i e
obse a ional coho s udy om he SWEDEHEART egis e . BMJ Open. 2012;
2(2):e000726.
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 8 o 9
31. Bugia dini R, Yan AT, Yan RT, Fi che D, Lange A, Man ini O, e al. Fac o s
in luencing unde u iliza ion o e idence-based he apies in women. Eu
Hea J. 2011;32(11):1337–44.
32. Koopman C, Vaa jes I, Hein jes EM, Spie ing W, an Dis I, He ings RM, e al.
Pe sis ing gende di e ences and a enua ing age di e ences in
ca dio ascula d ug use o p e en ion and ea men o co ona y hea
disease, 1998-2010. Eu Hea J. 2013;34(41):3198–205.
33. Mensah GA, Wei GS, So lie PD, Fine LJ, Rosenbe g Y, Kau mann PG, e al.
Decline in Ca dio ascula Mo ali y. Ci c Res. 2017;120(2):366–80.
34. Di Cas elnuo o A, Cos anzo S, Pe sichillo M, Oli ie i M, de Cu is A, Zi o F,
e al. Dis ibu ion o sho and li e ime isks o ca dio ascula disease in
I alians. Eu J P e Ca diol. 2012;19(4):723–30.
35. Be y JD, Liu K, Folsom AR, Lewis CE, Ca JJ, Polak JF, e al. P e alence and
p og ession o subclinical a he oscle osis in younge adul s wi h low sho -
e m bu high li e ime es ima ed isk o ca dio ascula disease : he Ca dia
and Mesa. Ci cula ion. 2010;119(3):382–9.
36. Ma ma AK, Be y JD, Ning H, Pe sell SD, Lloyd-Jones DM. Dis ibu ion o 10-
yea and li e ime p edic ed isks o ca dio ascula disease in US adul s:
indings om he na ional heal h and nu i ion examina ion su ey 2003 o
2006. Ci c Ca dio asc Qual Ou comes. 2010;3(1):8–14.
37. Poses RM, Cebul RD, Collins M, Fage SS. The impo ance o disease
p e alence in anspo ing clinical p edic ion ules. Ann In e n Med. 1986;
105:586–91.
Publishe ’sNo e
Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in
published maps and ins i u ional a ilia ions.
San os-Fe ei a e al. BMC Ca dio ascula Diso de s (2020) 20:301 Page 9 o 9