Hyperlipidemia in Early Adulthood Increases Long

DOI: 10.1161/CIRCULATIONAHA.114.012477
Hyperlipidemia in Early Adulthood Increases Long-Term Risk of
Coronary Heart Disease
Running title: Navar-Boggan et al.; Hyperlipidemia in Early Adulthood
Ann Marie Navar-Boggan, MD, PhD1; Eric D. Peterson, MD, MPH1;
Ralph B. D’Agostino, PhD2; Benjamin Neely, MS1; Allan D. Sniderman, MD*3;
Michael J. Pencina, PhD*1
1
Duke Clinical Research Institute, Duke University Medical Center, Durham, NC
NC; 2B
Boston
oston
t
University, Boston, MA; 3Mike Rosenbloom Laboratory for Cardiovascular Research, McGill
Centre,
Canada
University Health Cent
ntre
t e, Montreal, Ca
ana
n daa
*contributed
*c
conttrib
buteed
ed equ
equally
quual
ally
y
Correspondence:
Address forr Co
Corr
rres
rr
espo
es
pond
po
n en
nd
encce:
Ann Marie Navar-Boggan, MD, PhD and Michael J. Pencina, PhD
Duke Clinical Research Institute
Duke University Medical Center
2400 Pratt Street
Durham, NC 27705
Tel: 919-684-8111
Fax: 919-668-8830
E-mail: [email protected] or [email protected]
Journal Subject Codes: Atherosclerosis:[90] Lipid and lipoprotein metabolism, Ethics and
policy:[100] Health policy and outcome research
1
DOI: 10.1161/CIRCULATIONAHA.114.012477
Abstract
Background— Many young adults with moderate hyperlipidemia do not meet statin treatment
criteria under the new AHA/ACC cholesterol guidelines as they focus on 10-year cardiovascular
risk. We evaluated the association between years of exposure to hypercholesterolemia in early
adulthood and future coronary heart disease (CHD) risk.
Methods and Results—We examined Framingham Offspring Cohort data to identify adults
without incident cardiovascular disease to age 55 (n=1478), and explored the association
between moderate hyperlipidemia (non-high-density lipoprotein cholesterol [HDL-C] •160
mg/dL)
CHD. At median 15-year
g ) duration in early
y adulthood and subsequent
q
y follow-up,
p CHD
p,
rates
ates were significantly elevated among adults with prolonged hyperlipidemia exposure
exp
xppos
osur
uree by aage
ur
ge
55: 4.4% for those with no exposure, 8.1% for 1–10 years, and 16.5% for those with 11–20 years
exposure
after
adjustment
other
ex
xpo
posu
sure
r (p<0.001);
(p<
p<00.00
001)
1); this
th
his
i association
ass
s occia
iationn persisted
p rssisted af
pe
fteer adju
ust
stme
m nt ffor
or ot
the
herr ca
ccardiac
rd
dia
iacc risk ffactors
accto
ors
including
non-HDL-C
(HR
CII 1.0
1.05–1.85
per
ncllud
u ing non-H
HDL-C
C aatt age
gee 555
5 (H
HR 11.39,
.339, 95%
95% C
05–
5–1.
1.85
1.
855 pe
er ddecade
ecaadee off hhyperlipidemia).
yperrli
l pid
pidem
mia)
a).
Overall,
Ov
Over
erral
all,
l 85%
l,
85% of
of young
youn
yo
ungg adults
un
adul
ad
u tss with
with prolonged
pro
rolo
long
lo
nged
ng
ed hyperlipidemia
hyp
yper
erli
er
lipi
li
pide
pi
demi
de
miaa would
mi
wou
ould
ld not
not have
hav
avee been
been recommended
reccom
omme
mend
me
nded
nd
e
therapy
However,
those
for statin the
era
rapy
py at
at age
age 40,
400, under
u de
un
d r current
cu
urr
rren
entt national
en
nati
na
tion
ti
onnal guidelines.
gui
u de
deli
l nees. H
li
ow
wev
ever
er, am
er
aamong
ongg th
on
thos
o e not
os
considered statin therapy candidates at age 55, there remained a significant association between
cumulative exposure to hyperlipidemia in young adulthood and subsequent CHD risk (adjusted
HR 1.67, 95% CI 1.06–2.64).
Conclusions—Cumulative exposure to hyperlipidemia in young adulthood increases subsequent
risk of CHD in a dose-dependent fashion. Adults with prolonged exposure to even moderate
elevations in non-HDL-C have elevated risk for future CHD and may benefit from more
aggressive primary prevention.
Key words: hyperlipidemia; younger patients; coronary heart disease risk
2
DOI: 10.1161/CIRCULATIONAHA.114.012477
Hyperlipidemia is a potent risk factor for atherosclerosis and coronary heart disease (CHD) and
is present in a substantial proportion of young adults. According to data from the National Health
and Nutrition Examination Survey (NHANES), between 11.7% of adults aged 20–39 and 41.2%
of adults aged 40–64 had elevated low-density lipoprotein cholesterol (LDL-C) levels, but only
10.6% of adults aged 20–39 and 47.7% of adults age 40–64 with hyperlipidemia were on
treatment.1 The newly released American Heart Association (AHA)/American College of
Cardiology (ACC) guidelines for treatment of blood cholesterol for the prevention of
cardiovascular disease (CVD) recommend statin therapy for all adults with prevalent CVD,
LDL-C •190 mg/dl, diabetes, or 10-year risk of atherosclerotic CVD •7.5%, as assessed by the
new Pooled Cohort Equations.2 Although CHD events such as myocardial infarct
infarction
cttio
ionn pr
pres
present
esen
sen
entt
suddenly,
uddenly, the advanced extensive complex intramural lesions that lead to plaque rupture develop
over
ov
ver decades.
decad
ecad
ades
e . Since
Siinc
ncee the natural history of atherosclerosis
atheros
oscl
os
cleerosis is prolonged,
prolo
ong
n ed
d, th
thee risk of clinical events
rises
isees exponent
exponentially
tia
iall
l y la
lat
late
te iin
te
n li
life
life.
fe.. As a result,
fe
ressult, th
the
he new
he
w cholesterol
chooleest
steerool
ol guidelines
gui
uide
deli
liiness led
leed to
to a high
hig
ighh number
nuumb
mber
err of
older
ol
lde
derr adults
a ul
ad
ults
ts aged
age
gedd •60
•60 years
yeear
ears to
to be recommended
rec
ecom
om
mme
mennded
nded for
for statin
sta
taatiin therapy,
the
hera
rapy
ra
py, with
py
w th
wi
h relatively
rel
elat
ativ
at
iv
vel
e y fewer
feweer younger
fe
y ung
yo
unger
ge
adults meeting
ng sstatin
taati
tinn re
rrecommendation
c mm
co
mmen
en
nda
dati
tiion
o ttresholds.
resh
re
shol
sh
olds
ol
ds.1
ds
Studies on adults with familial hypercholesterolemia have shown CVD risk is increased
early among those with very high LDL-C levels.3 Similarly, adults with extremely low LDL-C
levels conferred by genetic polymorphisms have significantly lower than average risk of CVD.4,5
However, the association between prolonged exposure to mild to moderately elevated lipid levels
in young adulthood on an individual’s subsequent risk of CHD has not previously been well
described.6 Therefore, we used the Framingham Offspring Study to address the impact of
duration of hyperlipidemia in young adulthood (ages 35 to 55) and future risk of CHD beyond
age 55 years.
3
DOI: 10.1161/CIRCULATIONAHA.114.012477
Methods
Study Design and Sample
Our study examined data on 5124 individuals from the Offspring Cohort of the Framingham
Heart Study recruited between 1971 and 1975.7 In order to identify participants with sufficient
observation time to evaluate both the number of years of exposure to hyperlipidemia, as well as
the person’s future risk of CHD, participants were eligible for inclusion in this analysis if: 1)
they had attended Offspring Cohort examination 4 (1987-1991), 5 (1991-1995), or 6 (19951998); 2) were between the ages of 53 and 57 years; and 3) were free of CVD (defined as
myocardial infarction, angina, coronary insufficiency, transient ischemic attack, stroke, coronary
heart disease death, cardiovascular death, intermittent claudication, or heart failu
ure8) at
at tthe
he ttime
ime
im
failure
of eligibility assessment. Of exams 4, 5, and 6, the exam closest to age 55 was used as the
“baseline”
“b
bas
asel
elin
el
ine”
in
e” visit.
vis
isit. Data
Data
Da
t from prior exams were use
used
ed tto
o evaluate thee nu
nnumber
mbeer
mb
er of years of
hhyperlipidemia
hyp
yperlipidemia
pe
ia aattained
tttaine
need by
by tthe
he bbaseline
assel
elin
inee aage.
ge.. T
This
his resulted
reesullteed in
in a ssample
ampl
am
plee of 11478
4788 ad
47
adults
dul
u ts ffree
reee of
CVD
who
were
approximately
years
Participants
were
CV
VD at
a tthe
hee bbaseline
aseelin
as
ne eexamination,
xamin
aminat
a io
at
on , w
hoo w
eree ap
er
appr
prox
pr
oxiimate
ox
tely
te
ly 555
5 ye
year
arss ooff aage.
ar
g . Pa
ge
art
rtic
iccip
ipan
ntss w
erre
then
hen prospec
prospectively
ctiive
vely
ly ffollowed
o lo
ol
lowe
weed fo
fforr up to
to 20 yyears
ears
ea
rs ffor
or tthe
he development
dev
evel
elop
el
opme
op
mentt ooff CH
me
CHD
D (m
(myocardial
myo
yoca
card
ca
r ial
rd
infarction, angina, coronary insufficiency, coronary heart disease death) events. Median followup was 15 years.
Outcomes and Exposures
The primary factor of interest was the number of years of exposure to hyperlipidemia in the 20
years prior to the baseline visit at age 55 (e.g., number of years of elevated non-high-density
lipoprotein cholesterol [HDL-C] between ages 35 and 55). Consistent with the lipid measures
used in the newest Pooled Cohort Equations, hyperlipidemia in our primary analysis was defined
based on non-HDL-C, with levels •160 mg/dL considered elevated. This level is equivalent to
4
DOI: 10.1161/CIRCULATIONAHA.114.012477
the 70th percentile of the American population according to NHANES. Since Framingham
Offspring examinations occur approximately every 4 years, we interpolated hyperlipidemic
status in the years between the examinations. For individuals who developed hyperlipidemia in
the time interval between study visits, we assumed that the date of development was midway
between the two exams. For individuals with fewer than 20 years of data prior to baseline, we
conservatively assumed that the participant was free of hyperlipidemia for the time period
without data. For participants with missing data at any follow-up exam, the value from the prior
exam was carried forward.
In sensitivity analyses we also examined prior elevation of non-HDL-C as a continuous
variable. Each person’s average non-HDL-C over the preceding 20 years was cal
allcu
cula
l ted,
la
calculated,
weighted by the number of years between exams. In addition, LDL-C, rather than non-HDL-C
was evaluated
was
ev
val
alua
uate
ua
tedd using
te
ussin
ingg number of years with LDL-C
-C •
130 mg/dL. B
e au
ec
use the Friedewald equation
•130
Because
w
ass us
uused
ed to ca
alc
lcul
u at
ul
atee LD
DL-C,
L-C, aadults
du
ult
lsw
itth trig
i lyceeriidess over
ig
overr 400
400 mg/dL
mg/
g/ddL aatt any
any time
tim
me were
we e
was
calculate
LDL-C,
with
triglycerides
ex
xcl
clud
uded
ud
e ffrom
ed
rom
ro
m th
he LD
DL--C an
nal
alys
ysis
ys
is..
is
excluded
the
LDL-C
analysis.
Statistical Analysis
An
nal
alys
ysis
ys
is
First, using the number of years of hyperlipidemia by age 55, adults were stratified into three
groups: 1) those without hyperlipidemia by age 55; 2) those with 1–10 years of hyperlipidemia;
and 3) those with 11–20 years of hyperlipidemia. Kaplan-Meier survival curves were generated
to evaluate the risk of CHD over the subsequent years, and the log-rank test was used to test the
overall survival experience.
Cox proportional hazards regression models were employed to evaluate the relative risk
of increasing the number of years with hyperlipidemia on the onset of CHD events by evaluating
the association between number of years of exposure to hyperlipidemia at age 55 as a continuous
5
DOI: 10.1161/CIRCULATIONAHA.114.012477
variable between zero and twenty and future risk of CHD. To determine to what extent the
association between duration of hyperlipidemia and CHD risk could be attributed to a worse
overall health state associated with hyperlipidemia, multivariable analyses were performed
adjusting for the following standard non-lipid risk factors: age, sex, systolic blood pressure
(SBP), antihypertensive treatment, HDL cholesterol, diabetes, and smoking. Next, to determine
if the cumulative exposure to hyperlipidemia was a marker for prevalent hyperlipidemia at age
55 or if the duration of hyperlipidemia was associated with increased CHD risk independent of
lipid levels at that age, baseline (age 55) non-HDL-C level was also included in the multivariable
analysis. The final multivariable model also included adjustment for lipid-lowering therapy at
ratios
the
baseline and over the follow-up period in a time-dependent fashion. Hazard ratio
os fo
for th
he
association between duration of hyperlipidemia and future CHD risk are presented per 10-year
increase.
ncrrea
easse.
se
secondary
As sec
con
ondaaryy analyses,
anaalyse
lysees,
s, we
we repeated
repeated our
rep
ouur primary
prrim
marry analysis
anal
an
alyysis
ysiss of
of association
asssoci
as
sociat
attio
ionn between
beetw
weeen years
yearss of
year
exposure
sample
would
ex
xpo
posu
sure
su
re to
to non-HDL-C
non--HD
nonHDL
L-C •160
•1660 mg/dL
mg/d
mg
/dL
/d
L in
i oour
ur sam
ampplee off yyoung
am
oung
ou
ng aadults
dultts who
dult
who wo
oul
uldd no
nott bee
specifically
pecifically rrecommended
ecom
ec
omme
om
m nd
me
nded
ed for
for statin
sta
tati
t n therapy
ti
ther
th
erap
er
apyy under
ap
unde
un
derr the
de
th
he 2013
2013 AHA/ACC
AHA
HA/A
ACC guidelines.
gui
uide
deli
l ne
li
nes.
s.. This
Thi
hs
excluded adults with 10-year CVD risk •7.5%, diabetes with LDL-C •70 mg/dL, or LDL-C
•190 mg/dL.
Next, we assessed the proportion of adults who would have been recommended for statin
therapy at age 40 and age 50 under current guidelines based on diabetes status, LDL-C, and 10year CVD risk, stratified by years of exposure to hyperlipidemia (zero, 1–10, and 11–20 years of
hyperlipidemia). We considered both the •7.5% and the •5% risk thresholds to determine
treatment eligibility per the new guidelines. This analysis was performed to determine the extent
to which individuals with prolonged hyperlipidemia would have been identified as treatment
6
DOI: 10.1161/CIRCULATIONAHA.114.012477
candidates by the new guidelines during the period of exposure to hyperlipidemia. Since LDL-C
was estimated using Friedewald’s equation and, therefore, unavailable in adults with
triglycerides >400 mg/dL, adults with triglycerides >400 mg/dL were considered “statin eligible”
in the analysis that evaluated statin recommendations.
In sensitivity analysis, we investigated the robustness of our results to the choice of lipid
parameter and choice of threshold. First, the association between years of exposure to LDL-C
•130 mg/dL and future CHD was evaluated using multivariable Cox proportional hazards
modeling. Second, to determine whether the results depend on the 160 mg/dL non-HDL-C
threshold, the association between the weighted average non-HDL-C over the prior 20 years and
points
future risk of CHD was evaluated using restricted cubic splines. Inflection point
ts in
n tthe
he ggraph,
raph
ra
ph,,
ph
rounded
ounded to the nearest 5 mg/dL non-HDL-C were identified and then used as cut points in a
piecewise
pi
ieccew
ewiise
ise linear
line
li
nearr model
mod
o el of prior average non-HDL-C
non-HD
DL-C
C and future CHD
CHD
H rrisk.
iskk. Cox proportional
is
hazards
increase
hazard
haza
ar s modeling
modeeli
linng was
wass performed
per
erfo
form
fo
rmed
rm
ed evaluating
eval
valuatingg the
the impact
im
mpacct
ct of
of each
each 10
10 mg/dL
mg/d
mg
/ L in
/d
ncrea
crease
se iin
n ave
aaverage
verag
erag
ge
prior
elevation
risk
prio
pr
iorr el
io
elev
evat
atio
at
ionn ooff nnon-HDL-C
io
on-H
on
-H
HDL
DL-C
-C
C aand
nd ffuture
utuure
ut
ure ri
ris
sk ooff CH
sk
CHD,
D, aadjusting
djuusti
dj
ting
ti
ng ffor
or tthe
he ssame
am
me ch
ccharacteristics
araacteerisstic
ar
sticss at
at
treatment,
smoking,
baseline as inn the
the
h primary
pri
rima
mary
ma
r analysis
ry
ana
naly
ysi
siss (age,
(age
(a
gee, sex,
sex,
x SBP,
SBP,
P antihypertensive
ant
n ih
hyp
yper
erte
er
tens
te
nsiv
ns
i e tr
rea
eatm
tmen
tm
e t,, ddiabetes,
en
iabe
ia
bete
be
t s, smoking
te
g
non-HDL-C, HDL-C, and lipid therapy). Finally, to evaluate the impact of including adults on
lipid-lowering therapy at baseline, the primary analyses of the association between years of
elevation in non-HDL-C and future CHD risk were repeated excluding adults on lipid-lowering
therapy at baseline.
The analysis was approved by the Duke University Institutional Review Board. All
Framingham Offspring cohort participants gave informed consent for participation. Statistical
analysis was performed using SAS version 9.3.
7
DOI: 10.1161/CIRCULATIONAHA.114.012477
Results
Study Population
Characteristics of the study sample at baseline are presented in Table 1. A total of 124
individuals in the cohort developed CVD prior to age 55 and were not included in our sample.
The final sample included 1478 adults free of CVD at baseline with 0-20 years of
hyperlipidemia. Of these, 512 adults did not have hyperlipidemia, 389 adults had 1-10 years of
hyperlipidemia, and 577 adults had 11-20 years of hyperlipidemia exposure by baseline age.
Individuals with hyperlipidemia at baseline were more likely to be diabetic, male, and smokers,
and had higher SBP, body mass index, total cholesterol levels, and lower HDL-C levels than
those
hose without hyperlipidemia. Only 85 patients overall (5.8%) were on lipid-lowering
lipid-low
werin
in
ng treatment
trea
tr
eatm
ea
tmen
tm
ent
at the baseline visit.
Years
and Risk of CHD
Year
Ye
arss of Hyperlipidemia
ar
Hyp
yperrli
lip
pi
pidemia
During
follow-up,
155
D
urring
rin followw-up
upp, 15
55 individuals
in
ndi
divi
vidu
vi
dual
du
alss dev
ddeveloped
evel
velopeed new
w onset
onseet CHD,
CH
HD,
D, with
witth 136
1336 events
even
ev
entts
en
ts by
by median
median
med
follow-up
years.
Kaplan-Meier
CHD
baseline
fo
oll
llow
ow-u
ow
-upp of 1155 ye
year
ars.
s. Figure
Figure
Figu
ree 1 ppresents
rese
re
sen
ntss Ka
Kapl
lan
a -M
Mei
e er C
HD eevent
vent rrates
vent
attes ffrom
ro
om ba
base
selline
se
nee uup
p to
o 115
5
follow-up
stratified
hyperlipidemia
years of follo
oww up aamong
m ng
mo
n ppatients
atie
at
ient
ntss st
nt
stra
rati
tifi
ti
fied
fi
ed by tthe
he nnumber
um
mbe
berr of yyears
e rss ooff hy
ea
hype
perl
pe
rlip
ipid
ip
idem
id
emia
em
i at
ia
baseline. A dose-response pattern is seen, with progressively increasing risk of CHD as the
number of years of exposure to hyperlipidemia increases (log-rank test p<0.0001). At 15 years,
adults with 11–20 years of hyperlipidemia at baseline had an overall CHD risk of 16.5% (95%
CI 13.5–19.9%), compared with 8.1% (95% CI 5.5–11.7%) for adults with 1–10 years of
hyperlipidemia, and 4.4% (2.9–6.6%) for those without hyperlipidemia at baseline.
The unadjusted risk of CHD doubled for every ten years of exposure to hyperlipidemia
(Table 2, univariable HR 2.0, 95% CI 1.63–2.45 per decade of hyperlipidemia); this association
was attenuated, but remained statistically significant after adjusting for other standard risk
8
DOI: 10.1161/CIRCULATIONAHA.114.012477
factors including sex, age, SBP, antihypertensive therapy, smoking status, HDL-C, and diabetes
(adjusted HR 1.49, 95% CI 1.20–1.87 per decade of hyperlipidemia). In addition, the association
remained statistically significant after also adjusting for non-HDL-C at baseline (adjusted HR
1.39, 95% CI 1.05–1.85 per decade of hyperlipidemia), suggesting that previous cumulative
exposure to hyperlipidemia is associated with increased risk of CHD later in life, independent of
the cholesterol level at age 55. This association also remained significant after adjusting for
lipid-lowering therapy use at baseline and follow-up.
Statin Recommendations Under Current Guidelines
The number of participants who would be specifically targeted for statin therapy according to the
statin
Off 57
tatin benefit groups as identified by the new guidelines was calculated (Table 33).
). O
5777 adults
adul
ad
ults
with 11–20 years of hyperlipidemia at the index age of 55, 87 (15.1%) participants would have
met
criteria
met cr
rit
iter
eria
ia ffor
or sstatin
tati
ta
t n therapy at age 40, and 2011 ((34.8%)
ti
344.8%) would hhave
a e me
av
mett criteria at age 50.
These
Theese
es numbers
numberrs were
we e lower
low
wer among
amo
mong
ng
g those
tho
hose
see with
witth 1–10
1–110 years
year
arss of
ar
of hyperlipidemia
hyp
yper
erli
lipi
piide
d miaa at
at age
age 55:
55: off 389
389
adults,
(1.8%)
would
criteria
statin
would
ad
dul
u ts
ts,, 7 (1
(1.8
.8%)
.8
%) w
oulld hhave
ou
ave me
av
mett cr
crit
iter
erria ffor
or sta
tattin
ta
n ttherapy
herrapy aatt ag
agee 40 aand
nd 444
4 (1
((11.3%)
1 3%
1.
%) w
ould
uld hav
hhave
ave
ve
met criteria at
at age
age 50.
50 When
W en we
Wh
we used
u ed the
us
the
h lower
low
ower
er risk
riiskk threshold
thr
hres
esho
es
hold
ho
ld proposed
pro
ropo
pooseed by tthe
hee gguidelines
uide
ui
deli
de
l nes (10li
year CVD risk •5%) to identify adults eligible for statin therapy, 25.1% of adults with 11–20
years of hyperlipidemia at baseline would have been recommended for statin therapy at age 40,
and 51.6% would have been recommended for statin therapy at age 50.
When we restricted our analyses to those adults not recommended for statin therapy at
age 55 (i.e., 10-year CVD risk below 7.5%, LDL-C <190 mg/dL, and no diabetes with LDL-C
•70 mg/dL; n=971), the association between hyperlipidemia and risk of CHD was preserved;
adults with both 1–10 and 11–20 years of hyperlipidemia at baseline had significantly higher
rates of CHD compared with adults without hyperlipidemia (Figure 2, p<0.001). In
9
DOI: 10.1161/CIRCULATIONAHA.114.012477
multivariable models adjusting for standard risk factors and non-HDL-C at baseline, each decade
of hyperlipidemia at baseline was associated with a 67% increased risk of CHD at follow-up
(Table 2, HR 1.67, 95% CI 1.06–2.64, p=0.03).
Sensitivity Analyses
Using an LDL-C level of •130 mg/dL rather than non-HDL-C •160 mg/dL as the primary
exposure yielded similar results (appendix table 1). The results based on average prior nonHDL-C are shown in Figure 3, where they are plotted against future CHD risk approximated
using a restricted cubic spline and a piecewise linear model. The effect of weighted non-HDL-C
below 125 mg/dl on CHD was non-significant, suggesting that individuals below this cut-point
the effect of weighted
above
all have similar (lower) risk of CHD. Similarly,
a
weighted non-HDL-C
non-HDLL--C ab
bov
ovee 19
1955
this cut-point all have
mg/dl on CHD was non-significant, suggesting that individuals above
a
similar
weighted
imi
mila
larr (high)
la
(hig
(h
igh)
ig
h) risk
rissk of
o CHD. The association between
betw
wee
eenn average weig
ig
ghted
ed
d nnon-HDL-C
on-HDL-C between
mg/dL
and
wass st
statistically
Cox
modeling,
1125–195
25–195
5–
mg/d
dL an
nd CH
CHD
D wa
w
stat
atis
at
issti
t ca
callly ssignificant.
iggnifficaant.. IIn
n Co
C
oxx pproportional
ropor
orti
tionnal hhazards
ti
azards
azar
ds m
odel
od
elin
el
ing,
n ,
every
mg/dL
mg/dL
the
preceding
ev
ver
eryy 10
1 m
g/dL
g/
dL iincrease
ncreas
ncre
asee in aaverage
vera
ve
rage
ra
ge nnon-HDL-C
on-H
on-H
-HDL
DL-C
DL
-C
C bbetween
etween
etw
ween 1125–195
255–1
– 955 m
g/ddL
dL oover
verr th
ve
he pr
pre
eceedin
edin
ng 20
2
years was associated
asssooci
c atted with
wit
i h a 33%
3 % increase
33
i cr
in
c eaase in
in future
futu
fu
t re CHD
tu
CHD
D risk
ris
iskk (HR
(HR per
per 10 mg/dL
mg/
g/dL
dL increase
inc
ncre
r ase 1.33,
re
95% CI 1.23–1.45, p<0.001). After adjusting for standard risk factors including baseline nonHDL-C and lipid therapy at baseline and follow-up, this association remained statistically
significant (HR 1.20, 95% CI 1.08–1.35, p=0.001). Finally, excluding adults on lipid-lowering
therapy at baseline did not result in substantive changes to the results (appendix table 2).
Discussion
In this analysis of adults free of CVD at age 55 in the Framingham Offspring Study, we found
that those with the longest prior exposure to moderately elevated non-HDL-C had a nearly four-
10
DOI: 10.1161/CIRCULATIONAHA.114.012477
fold increased rate of CHD at follow-up. Importantly, not only does prevalent hyperlipidemia
increase future risk of CHD, but the length of exposure to hyperlipidemia in the fourth and fifth
decades of life affects future CHD risk in a dose-responsive manner as the association between
exposure to hyperlipidemia in young adulthood and future CHD remained highly significant
even after adjustment for non-HDL-C at age 55. This association was preserved in individuals
without direct recommendations for statin therapy under the current guidelines. These findings
are aligned with the biological understanding of atherosclerosis as a progressive disease due to
ongoing vessel injury over time—a substantial part of which is caused by elevated cholesterol
levels.6
In addition to the 10-year risk, which was calculated using the Pooled Coh
Cohort
ohhorrt Eq
Equa
Equations,
uati
ua
tion
ti
o s,,
on
the
he current AHA/ACC cholesterol guideline recommends when making treatment decisions to
consider
co
ons
nsid
ideer
id
er family
fam
amiily history,
hisstory,
hi
st
C-reactive protein, coro
coronary
ona
n ry artery calciu
calcium,
um,
m aankle
nklle brachial index, and
nk
lifetime
Our
suggests
sustained
moderate
lipid
levels
ifettime
ti CVD rrisk.
isk.. O
is
ur ddata
ataa su
at
sugg
gges
e ts
es
ts tthat
haat su
usttaineed mo
odera
deraatee eelevation
leevati
tion
onn ooff li
ipi
pidd le
eve
vels
ls aalso
lsso
confers
substantial
future
Given
association
duration
co
onf
n er
erss a su
ubs
b ta
tanntiial
ial rrisk
iskk ooff fu
futu
tu
ure eevents.
vent
ve
n s.
nt
s. G
iv
ven
e tthe
hee ppotent
otten
entt as
sso
s ci
c at
atio
i n betw
io
bbetween
etw
weeen du
dur
ratiionn ooff
hyperlipidemia
mid-adulthood
future
CHD
should
exposure to hy
hype
perl
pe
rlip
rl
ipid
ip
id
dem
emia
iaa bby
y mi
midd-ad
a ul
ulth
th
hoo
oodd an
andd fu
utu
ture
re C
HD rrisk,
i k, cclinicians
is
l ni
li
nici
cian
ci
a s sh
an
hou
ould
ld
d also
consider lifestyle intervention or even treatment for adults with prolonged prior exposure to
hyperlipidemia. Randomized controlled trial evidence support the clinical benefit of statin
therapy for primary prevention in adults in this age group, with a number of primary prevention
trials demonstrating that statins initiated in midlife significantly reduce future clinical events.9-11
While the new guidelines identify those patients with a very high lipid level on a single
measurement (LDL-C •190 mg/dL) as candidates for statin therapy, they do not further
differentiate risk in others based on lipoprotein levels. Under current guidelines, only one in six
adults in this cohort with prolonged duration of exposure to hyperlipidemia would have been
11
DOI: 10.1161/CIRCULATIONAHA.114.012477
directly recommended for statin therapy at age 40, and one in three at age 50. By design, our
analysis cannot answer the question whether early statin intervention in those on the
hyperlipidemic trajectory would decrease their future CHD risk. When to initiate treatment in
adults with moderately elevated non-HDL-C in early adulthood remains unknown. There are no
studies evaluating the long-term effectiveness of statin therapy in adults aged 30–50 with only
moderately elevated lipid levels and without other risk factors. Furthermore, initiating statin
therapy at younger ages would result in a much longer duration of statin use than has been
studied in randomized trials. This lack of knowledge further stresses the need for additional
research focused on the safety and efficacy of long-term statin use in early and middle adulthood
too reduce cardiovascular disease later in life. This analysis also highlights the fact
ctt tthat
h t ri
ha
risk
sk
prediction models focused on a 10-year horizon may underestimate the contribution of prolonged
exposure
ex
xpo
posu
sure
su
re to
to chronic
c ro
ch
oni
nicc disease, and the need to continue
cont
ntin
nt
inuue to evaluatee how
h w to best
ho
best incorporate 30-year
12, 13
riskk estimates
or lifetime
lifetime
if
esttim
imat
ates
ess into
into
nto current
cur
urre
rennt
re
nt prevention
preevenntiion guidelines.
guideeli
gu
eliness.12,
Limitations
Li
imi
mita
tati
ta
tion
on
ns and
and Stre
S
Strengths
treenggth
ths
Our analysiss has
has several
sev
ever
erral limitations.
lim
im
mit
itat
atio
ions
io
ns.. First,
ns
Fiirs
r t, we
we only
o ly included
on
in
ncl
clud
uded
ud
ed adults
adu
dult
l s aged
lt
aged 53–57
53–5
53
–577 who
who were
w re free off
we
CVD; therefore, the point estimates cannot be extrapolated outside of this age range.
Nevertheless, we believe that this analysis demonstrates the long-term impact of hyperlipidemia
in young adulthood. Second, our analysis defined hyperlipidemia using a non-HDL-C cutoff of
•160 mg/dL, which is consistent with how the prior Adult Treatment Panel (ATP) III guidelines
defined elevated cholesterol; given that the risk of CHD events expands with increasing levels of
non-HDL-C, the use of this or any cut-point may have falsely dichotomized a continuous
relationship. However, using a continuous approach which averaged non-HDL-C over the
preceding 20 years yielded similar results, showing that the risk associated with exposure to non-
12
DOI: 10.1161/CIRCULATIONAHA.114.012477
HDL-C increased linearly between 125–195 mg/dL. However, the number of individuals with
prior average non-HDL-C below 125 mg/dL and above 195 mg/dL is low. Therefore, we cannot
definitively determine if the relationship between prior average non-HDL-C and future CHD
continues below 125 mg/dL or above 195 mg/dL. Third, this analysis only considered the
duration of hyperlipidemia and not the duration of all other risk factors and comorbidities. While
the presence of comorbidities was accounted for at age 55, it is likely that the duration of certain
comorbidities, such as diabetes and hypertension, may also affect future CHD risk in a similar
duration-dependent manner, as demonstrated in this hyperlipidemia analysis.14 Next, our study
design excluded 124 individuals with premature CVD before age 55, representing 8% of the
initial
nitial sample. As a result, some individuals at highest risk, due to hyperlipidemia
hyperlipidemi
miaa at a young
you
oung
ng age,
age
g
may have been excluded; these exclusions would have led to an underestimation of the
association
asssooci
ciat
atio
at
ionn between
io
betw
be
wee
eenn duration of hyperlipidemia aand
ndd CHD risk.
Notably,
Notablly, our
ouur
ur study
study
dy also
als
lsoo had
haad several
sevver
se
veral strengths.
strrenggth
hs. First,
First,
Fir
rst,, oour
ur aanalysis
naly
lysi
sis iiss bbased
si
assed oon
n th
thee
Framingham
Heart
before
widespread
Fram
Fr
amin
am
ingh
in
g am H
gh
eartt Study
Stu
udy data
datta collected
coll
co
llec
ll
ectted
ed iin
n aan
n eera
r be
ra
efore
ore wi
wide
deesp
s read
read sstatin
tatiin use,
tati
use,
e aallowing
lllow
owin
ng us tto
o
evaluate the impact
imp
mpac
acct of untreated
unt
n re
reaate
t d hyperlipidemia.
h pe
hy
p rllip
ipid
idem
id
emia
em
iaa. No
Nott on
only
nly w
was
ass tthe
h ooverall
he
vera
ve
rall
ra
ll rrate
a e of llipid-lowering
at
ipid
ip
i -lowering
id
therapy in this group low, but the risk associated with increased duration of hyperlipidemia
remained significant even after adjusting for lipid-lowering therapy at baseline and follow-up
and excluding adults on lipid lowering therapy at baseline. Second, due to the length of followup, the Framingham Offspring data allows for accurate, longer-term assessment of risk factors,
as well as follow-up of hard cardiovascular endpoints. Our analysis fully utilizes the consecutive
follow-up, as well as risk factor and event ascertainment, during the course of 35 years (20 years
of potential exposure and 15 years of follow-up), which is uniquely available in Framingham.
Finally, our study design eliminates the possibility that the association seen between duration of
13
DOI: 10.1161/CIRCULATIONAHA.114.012477
hyperlipidemia and risk of coronary events is a result of residual confounding by age, as risk was
calculated for all participants starting around age 55.
Conclusions
We conclude that the exposure to hyperlipidemia in the fourth and fifth decades of life is
associated with a substantially increased risk of CHD in a dose-responsive fashion, even among
adults otherwise predicted to have low risk of CVD. Our findings suggest that adults with
longstanding moderate elevations in non-HDL-C should be added to those already identified by
the current guidelines as candidates for an informed patient-physician discussion about
appropriate lipid management strategies to reduce future risk of heart disease.
Acknowledgments:
thank
Ackn
kn
now
owle
ledg
le
dgme
dg
ment
nts:
nt
s The authors would like to th
han
ank Erin Hanley, MS,
S, ffor
o her editorial
or
contributions
Ms.. Ha
Hanley
receive
compensation
contributions,
co
onttri
ribu
b tionns to
o tthis
h s manuscript.
hi
maanuusc
s ri
ript
pt.. Ms
pt
M
Hanl
nley
ey
y ddid
id nnot
ot rec
ecei
eive
ei
ve com
om
mpe
pens
nsat
attio
i n fo
forr he
herr co
cont
ntri
ribu
ri
buuti
tion
ons,
apart
employment
where
this
study
was
conducted.
ap
apar
parrt from her emp
mploym
mp
yment at tthe
ym
h iinstitution
he
nstiitu
ns
utionn w
herre
re thi
is st
stud
udyy wa
as cond
ducteed.
Funding
work
supported
internally
Clinical
Fund
Fu
ndin
nd
ingg Sources:
in
Sour
So
urce
ur
ces:
ce
s: Th
This
is w
orkk wa
or
wass su
supp
ppor
pp
orte
or
tedd in
te
inte
tern
te
rnal
rn
ally
al
ly bby
y th
thee Du
Duke
ke C
lini
li
nica
ni
call Research
ca
Rese
Re
sear
se
arch
ar
ch Institute
Ins
nsti
titu
ti
tute
tu
te in
in
addition
Healthcare
Research
Quality.
addi
ad
diti
tion
on to
to grant
gran
gr
antt number
numb
nu
mber
er U19HS021092
U19
19HS
HS02
0210
1092
92 from
fro
rom
m the
the Agency
Agen
Ag
ency
cy ffor
or H
ealt
ea
lthc
hcar
aree Re
Rese
sear
arch
ch aand
nd Q
uali
ua
lity
ty
The content is solely the responsibility of the authors and does not necessarily represent the
official views of the Agency for Healthcare Research and Quality. The Framingham Heart Study
is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in
collaboration with Boston University (Contract No. N01-HC-25195). This manuscript has been
reviewed by Framingham Heart Study Investigators for Scientific content and consistency of
data interpretation with previous Framingham Heart Study publications.
Conflict of Interest Disclosures: Dr. Navar-Boggan has no relevant conflicts of interest to
report. Dr. Peterson has received funding for research grants from Eli Lilly and Janssen
Pharmaceuticals, and funding for serving as a consultant/participant on advisory board for
Merck, Sanofi-Aventis, Janssen Pharmaceuticals, and Boehringer Ingelheim. Dr. D’Agostino has
14
DOI: 10.1161/CIRCULATIONAHA.114.012477
no relevant conflicts of interest to report. Mr. Neely has no relevant conflicts of interest to report.
Dr. Sniderman has no relevant conflicts of interest to report. Dr. Pencina has received funding
for serving as a consultant for AbbVie.
References:
1. Pencina MJ, Navar-Boggan AM, D'Agostino RB, Sr., Williams K, Neely B, Sniderman AD,
Peterson ED. Application of new cholesterol guidelines to a population-based sample. N Engl J
Med. 2014;370:1422-1431.
2. Goff DC Jr, Lloyd-Jones DM, Bennett G, Coady S, D'Agostino RB, Gibbons R, Greenland P,
Lackland DT, Levy D, O'Donnell CJ, Robinson J, Schwartz JS, Shero ST, Smith SC, Jr., Sorlie
P, Stone NJ, Wilson PW. 2013 ACC/AHA Guideline on the Assessment of Cardiovascular Risk:
a report of the American College of Cardiology/American Heart Association Task Force on
Practice Guidelines. Circulation. 2014;129:S49-73.
3. Hopkins PN, Heiss G, Ellison RC, Province MA, Pankow JS, Eckfeldt JH, Hunt
Huunt
nt SC.
SC. Coronary
Cor
oron
o ar
on
ay
artery disease risk in familial combined hyperlipidemia and familial hypertriglyc
hypertriglyceridemia:
cerrid
idem
em
mia
ia:: a
case-control comparison from the National Heart, Lung, and Blood Institute Family Heart Study.
Circulation. 2003;108:519-523.
2003;1
; 08:519-523.
4. C
Cohen
Boerwinkle
Mosley
Hobbs
ohen JJC,
oh
C, B
oeerw
r in
inkl
klee E,
kl
E M
osle
os
leyy TH
le
TH,, Jr
Jr.,.,, H
obbs
ob
bs HH.
H.. Sequence
Seq
equeenc
ncee variations
vari
va
riat
a ions
at
ns in
in PCSK9,
P SK
PC
SK9,
9, low
low
LDL,
protection
coronary
disease.
LDL
L, and prote
tect
ctio
on aagainst
gai
ains
ai
nstt co
ns
oro
rona
nary
na
r hheart
ry
eart di
iseasse.
se. N Engl
Engl J Med.
Med
ed. 22006;354:1264-1272.
0066;3554:
00
4:12
1264
64-1
64
-127
-1
272.
27
2.
5.. Ference
Ferren
e ce BA,
BA,
A, Yoo
Yoo
oo W,
W, Alesh
Alessh I,
Al
I, Mahajan
Mah
ahaj
ajan
aj
an N,
N, Mirowska
Miro
Miro
owskka
ka KK,
KK, Mewada
Mewad
ewad
adaa A,
A, Kahn
Kahn
ahn J,, Afonso
Afonnsoo L,
Williams
Will
llia
iams KA
KA Sr,
Sr Flack
Flac
Fl
ackk JM.
JM Effect
Effe
Ef
fect
ct of long-term
long
ng-t
-ter
erm
m exposure
expos
osur
uree to lower
low
ower
er low-density
low
ow-d
- en
ensiity
y lipoprotein
lip
ipoppro
rotein
in
cholesterol be
beginning
early
life
the
coronary
egi
ginn
nnnin
ingg ea
arlyy in
n li
ife oon
n th
he ri
risk
sk ooff co
coro
ro
ona
nary
ry heart
hea
eart
rt disease:
dis
i ea
ease
se:: a Mendelian
se
M nd
Me
ndel
elia
el
iann
ia
randomization
analysis.
Coll
Cardiol.
and
ndom
omiz
izat
atio
ionn an
anal
alys
ysis
is J Am C
olll Ca
ol
Card
rdio
ioll 22012;60:2631-2639.
012;
01
2;60
60:2
:263
6311-26
2639
39
6. Sniderman AD, Furberg CD. Age as a modifiable risk factor for cardiovascular disease.
Lancet. 2008;371:1547-1549.
7. Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham
Offspring Study. Design and preliminary data. Prev Med. 1975;4:518-525.
8. D’Agostino RB, Vasan RS, Pencina MJ, Wolf PA, Cobain M, Massaro JM, Kannel WB.
General cardiovascular risk profile for use in primary care: the Framingham Heart Study.
Circulation. 2008;117:743-53.
9. Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, Beere PA, Langendorfer A, Stein
EA, Kruyer W, Gotto AM Jr. Primary prevention of acute coronary events with lovastatin in men
and women with average cholesterol levels: results of AFCAPS/TexCAPS. Air Force/Texas
Coronary Atherosclerosis Prevention Study. JAMA. 1998;279:1615-1622.
15
DOI: 10.1161/CIRCULATIONAHA.114.012477
10. Barringer TA 3rd. WOSCOPS. West of Scotland Coronary Prevention Group. Lancet.
1997;349:432-433.
11. Nakamura H, Arakawa K, Itakura H, Kitabatake A, Goto Y, Toyota T, Nakaya N, Nishimoto
S, Muranaka M, Yamamoto A, Mizuno K, Ohashi Y; MEGA Study Group. Primary prevention
of cardiovascular disease with pravastatin in Japan (MEGA Study): a prospective randomised
controlled trial. Lancet. 2006;368:1155-1163.
12. Pencina MJ, D'Agostino RB, Sr., Larson MG, Massaro JM, Vasan RS. Predicting the 30-year
risk of cardiovascular disease: the Framingham Heart Study. Circulation. 2009;119:3078-3084.
13. Berry JD, Liu K, Folsom AR, Lewis CE, Carr JJ, Polak JF, Shea S, Sidney S, O'Leary DH,
Chan C, Lloyd-Jones DM. Prevalence and progression of subclinical atherosclerosis in younger
adults with low short-term but high lifetime estimated risk for cardiovascular disease: the
coronary artery risk development in young adults study and multi-ethnic study of atherosclerosis.
Circulation. 2009;119:382-389.
14. Vasan RS, Massaro JM, Wilson PW, Seshadri S, Wolf PA, Levy D, D’Agostin
D’Agostino
inoo RB
RB..
Antecedent blood pressure and risk of cardiovascular disease: the Framingham H
Heart
Study.
eaart
r S
tudy
tu
dy..
dy
Circulation. 2002;105:48-53.
16
DOI: 10.1161/CIRCULATIONAHA.114.012477
Table 1. Characteristics of Participants Stratified by Years of Hyperlipidemia at Baseline.*
Variable
Age
Female
Smoking
Diabetes
etes
BMI
Treatment
ment for cholesterol
Treatment
ment for BP
Systolic
lic BP
Total choles
cholesterol
este
t ro
rol
HDL-C
-C
C cholesterol
chol
ch
oles
ol
e tero
rol
ol
Non-HDL-C
H LHD
LC
Total
(n=1478)
55 (54,56)
791 (53.5%)
281 (19.0%)
95 (6.4%)
27.0 (24.2,30.3)
85 (5.8%)
282 (19.1%)
127 (115,137)
209 (184,233)
48 (40,60)
1588 (132,184)
15
(132
(1
32,1
, 84
,1
84))
Duration of Hyperlipidemia at Baseline
0 Years
1-10 Years
(n=512)
(n=389)
55 (54,56)
55.0 (54,56)
336 (65.6%)
241 (62.0%)
89 (17.4%)
64 (16.5%)
19 (3.7%)
23 (5.9%)
25.4 (23.0,28.6)
26.9 (24.5,30.1)
3 (0.6%)
10 (2.6%)
59 (11.5%)
73 (18.8%)
123 (112,134)
126 (115,136)
182
18
82 (166,199)
(166
(1
6 ,199)
2155 (199,236)
( 99
(1
9 ,236)
57 (46
(46,70)
4 ,70)
499 ((39,60)
3 ,60)
39
125 (109,140)
12
(109
10 ,140
40))
165
16
65 (148,181)
(148
(1
4 ,181
81))
*
Binaryy data
dat
ata
ta presented as n (%),
(%)
% , co
ccontinuous
ontin
tin
inuo
u uss variables
var
aria
i bl
ia
bles
ess ppresented
rese
es nt
nted
ed aass me
m
median
dian
n (interquartile
(in
interqua
u rt
rtile range)
raange)
ge)
BMI indicates
ndicate
ates body mass index;
i de
in
dex;
x B
BP,
P, bblood
lood
lo
d press
pressure;
sur
u e;
e HDL-C,
HDL
DL-C
L-C
C, high
high-density
g -d
-density
ty lip
lipoprotein
i oprote
o in
n ccholesterol
hole
ho
ole
lest
ster
erol
o
17
11-20 Years
(n=577)
56.0 (54,56)
214 (37.1%)
128 (22.2%)
53 (9.2%)
(9.
9.2%
2%)
288.33 (25.4,31.0)
28.3
(25
5.44,3
, 1.
72 ((12.5%)
122.5
.5%)
5%)
150 (26.0%)
130 (119.0,140.0)
(119.0,140
230 (207,253)
(207,253
43 (37,51)
1833 (163,205)
18
( 63
(1
63,2
,205
DOI: 10.1161/CIRCULATIONAHA.114.012477
Table 2. Future Risk of CHD Per Decade of Hyperlipidemia Experienced by Age 55 in Univariable and Multivariable Analyses.
All Adults (n=1514)
Model
Univariable: duration of hyperlipidemia
Duration
tion of hyperlipidemia + standard risk factors at baseline†
Duration
tion of hyperlipidemia + standard risk factors† + non-HDL-C
at baseline
Duration
tion of hyperlipidemia + standard risk factors† + non-HDL-C
at baseline + lipid lowering therapy at baseline and follow-up
HR Per Decade of
Hyperlipidemia
(95% CI)
2.00 (1.63-2.45)
1.49 (1.20 - 1.87)
1.39 (1.05 - 1.85)
<0.0001
0.0004
0.022
1.40 (1.05 - 1.87)
0.024
*
p-value
Not Recommended for Statin
Therapy at Baseline* (n=971)
HR Per Decade of
p-value
Hyperlipidemia
(95% CI)
1.99 (1.44-2.75)
<0.0001
1.77 (1.
(1.26
.26 - 22.48)
.48)
.4
8)
0.001
0.0
1.67 ((1.06
1.06
1.
06 - 22.64)
.664)
.64)
0.026
0.0
1.60 (1.
(1.00
1 00
0 - 2.56)
2.56)
56)
00.048
0.
0
10-year
ar predicted CVD risk <7.5%, no diabetes and LDL-C •70, and LDL-C <190
Standard
ard rrisk
iskk fa
is
fact
factors:
ctor
ct
o s: ssex,
or
ex
x, age,
e,, bblood
l od
lo
o pressure, HDL-C, antihypertensive treatment,
treatme
ment
nt, sm
nt
ssmoking,
oking, diabetes
CHD indicates
cardiovascular
ndi
dica
cate
ca
tess coronary
te
corona
co
nary
ary heart
heartt disease;
dis
isea
e se
ea
s ; CI, confidence interval; CVD, cardiovasc
scul
u arr disease; HR, hazard ratio;
ul
rat
a io; LDL-C,
LDL
LD
DL-C,
L-C low-density
t lipoprotein cholesterol;
cholesterol All
other abbreviations
can
found
Table
abbr
ab
b evia
br
viati
tion
o s ca
an be fou
und iin
n Ta
T
blle 1..
†
18
DOI: 10.1161/CIRCULATIONAHA.114.012477
Table 3. Statin Recommendations at Age 40 and 50 for Adults With and Without
Hyperlipidemia at Age 55.
*
Variable
0 years (n=512)
1-10 years (n=389)
11-20 years (n=577)
Recommended for Statin Therapy,
ASCVD Risk •7.5%
Age 40
Age 50
3 (0.6%)
7 (1.8%)
87 (15.1%)
16 (3.1%)
44 (11.3%)
201 (34.8%)
Recommended for Statin
Therapy, ASCVD Risk •5%
Age 40
Age 50
6 (1.2%)
11 (2.8%)
145 (25.1%)
43 (8.4%)
77 (19.8%)
298 (51.6%)
*
Based on 2013 AHA/ACC Cholesterol Guidelines. Adults with triglycerides >400 mg/dL considered statin
recommended.
+
ACC indicates American College of Cardiology; AHA, American Heart Association; ASCVD, atherosclerotic
cardiovascular disease
Figure Legends:
Figure
Fi
igu
gure
re 11.. T
Time
ime to
im
to D
Diagnosis
iagnosis of CHD by Numbe
Number
er of
of Years of Hy
Hyperlipidemia
ype
p rlip
ip
pide
id
demia at Baseline. This
figure
fiiguure
r shows Kaplan-Meier
Kap
a lann Me
Meie
ierr curves
ie
curv
cu
rvees of
of future
futu
fu
turee risk
riisk of
of CHD
CH
HD beginning
b ginn
be
ginn
nin
ng at age
ag
ge 55 (age
(ag
ge range
r ng
ra
ngee 53–57),
53–5
53
–57),
–5
stratified
experienced
55.
Log-rank
p-value
<0.0001.
tra
rati
tiifi
f ed by
by years
year
ye
arrs of hyperlipidemia
hyp
per
erlip
ip
pid
dem
emia
ia exp
xper
xp
erie
iennced byy age
ie
age 55
5. Lo
Log
g-rrank
rank
n p-valluee <
-val
0..00
0011. *C
*CHD
HD
indicates
ndi
dicattes coro
di
coronary
rona
nary
ry
y heart
hea
eart
rt ddisease.
iseasee.
isea
Figure 2. Time to Diagnosis of CHD by Number of Years of Hyperlipidemia at Baseline Among
Adults Not Recommended for Statin Therapy at Baseline*. This figure shows Kaplan-Meier
curves of future risk of CHD stratified by years of hyperlipidemia experienced by age 55 (age
range 53–57) among adults not recommended for statin therapy at age 55. Log-rank p-value
<0.0001. *Excludes those recommended for statins: ASCVD risk •7.5%, LDL-C •190, diabetes
and LDL-C •100. +ASCVD indicates atherosclerotic cardiovascular disease; CHD, coronary
heart disease; LDL-C, low-density lipoprotein cholesterol.
19
DOI: 10.1161/CIRCULATIONAHA.114.012477
Figure 3. Prior Weighted Average Cholesterol and CHD Risk. This figure shows the shapes of
the restricted cubic spline and piecewise linear models of average non-HDL cholesterol prior to
age 55 and the centered linear predictors from the model of time to coronary heart disease. The
black curve shows the restricted cubic spline with 95% confidence intervals (dotted) and grey
curve shows the piecewise linear spline with knots at 125 mg/dL and 195 mg/dL. The x-axis is
truncated at the 5th and 95th percentiles of prior average non-HDL. *CHD indicates coronary
heart disease.
20
Incidence of Coronary Heart Disease
Years of Follow-Up from Baseline (Age 55)
Figure 1
Incidence of Coronary Heart Disease
Years of Follow-Up from Baseline (Age 55)
Figure 2
Average Non-HDL Prior to Age 55
Figure 3
Centered Linear Predictors from Model of Time to CHD