Original research  1
Homocysteine is an independent predictor of long-term
cardiac mortality in patients with stable coronary artery
disease in the era of statins
Loukianos S. Rallidisa, Nikolaos Kosmasa, Taxiarchoula Rallidib,
Georgios Pavlakisc, Estela Kiouria and Maria Zolindakid
Background  Homocysteine (Hcy) is considered a risk
factor for cardiovascular disease.
Objective  To explore the long-term prognostic value of
Hcy in patients with stable coronary artery disease (CAD)
in the era of statins.
Methods  A total of 876 consecutive patients with
stable CAD were recruited and followed up for a median
of 6.1 years. Lipids and Hcy levels were measured at
baseline. Primary endpoints were cardiac death and
secondary endpoints were hospitalizations for acute
coronary syndrome, myocardial revascularization,
arrhythmic event or ischemic stroke.
Results  Follow-up data were obtained from 842
patients of whom 70 had a cardiac death (8.3%), while
258 (30.6%) met the secondary endpoints. Seven hundred
four patients (83.6%) were on statins. In univariate Cox
regression analysis Hcy predicted the occurrence of
cardiac death [hazard ratio: 1.030; 95% confidence interval
(CI): 1.018–1.042, P < 0.001] but not the occurrence of
secondary endpoints (hazard ratio: 1.010; 95% CI: 0.999–
1.020, P = 0.081). Hcy remained an independent predictor
Introduction
The prognosis of patients with coronary artery disease
(CAD) has been improved significantly over the past
two decades. This is mainly due to progresses in revas-
cularization techniques, introduction of more effective
antiplatelet treatments and the use of statins. Despite
this progress, a significant number of patients with CAD
develop cardiovascular events. This persistent recur-
rence risk can be attributed to residual risk caused by
lipids [therapeutic gap in achieving low-density lipo-
protein-cholesterol (LDL-C) targets, lack of treatment
of high lipoprotein(a), etc.], inflammation, thrombosis,
etc. [1,2].
Based on prospective observational studies and clinical
data, homocysteine (Hcy) has long been considered a
risk factor for cardiovascular disease (CVD) [3]. Multiple
potential atherogenic properties have been proposed
such as endothelial dysfunction by decreasing nitric oxide
bioavailability, induction of oxidative stress and enhance-
ment of proinflammatory actions [4]. Additionally, Hcy
0954-6928 Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2019 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
of cardiac death after adjustment for conventional risk
factors, ejection fraction and statin use (hazard ratio:
1.030; 95% CI: 1.017–1.044, P < 0.001). Patients in the
highest tertile of Hcy levels (>14.1 μmol/L) had three
times higher risk of cardiac death compared with patients
in the lowest tertile (<10.3 μmol/L) (hazard ratio = 3.036,
CI: 1.983–4.649, P < 0.001).
Conclusion  Hcy is an independent predictor of cardiac
death in patients with stable CAD in the era of statins.
Coron Artery Dis XXX: 000–000 Copyright © 2019 Wolters
Kluwer Health, Inc. All rights reserved.
Coronary Artery Disease 2019, XXX:000–000
Keywords: homocysteine, risk factor for cardiac death,
stable coronary artery disease, statins
a
Second Department of Cardiology, University General Hospital,
‘Attikon’,  bSchool of Dentistry, National and Kapodistrian University of
Athens,  cKAT General Hospital of Attica, Athens and  dBiochemistry Laboratory,
General Hospital of Nikea, Piraeus, Greece
Correspondence to Loukianos S. Rallidis, MD, FESC, Second Department of
Cardiology, University General Hospital “Attikon”, 1 Rimini, 12462 Chaidari,
Athens, Greece
Tel: +30 210 992 9106; fax: +30 210 5832351; e-mail: lrallidis@gmail.com
Received 28 March 2019 Accepted 25 August 2019
increases thrombogenic potential via several mechanisms
such as increased platelet reactivity, attenuated anticoag-
ulant processes and increased tissue factor expression [5].
Several studies have shown an association of raised Hcy
levels with prognosis in the setting of acute coronary
syndrome (ACS) [6–8]. However, there are only a few
studies regarding the prognostic role of hyperhomocyst-
einemia in patients with stable CAD [9,10]. In our study,
we explored the long-term prognostic value of Hcy in
patients with stable CAD in the era of statins which is
the major contributor to the improved outcome of these
patients the past two decades.
Methods
Study population
A total 876 patients with stable CAD were initially
recruited from three major general hospitals in Athens
area, Greece. All patients were participants of the
Lipoprotein-Associated phospholipasE A2 in stable cor-
onary aRTEry diSease (LAERTES) study, as previously
DOI: 10.1097/MCA.0000000000000800
2  Coronary Artery Disease  2019, Vol XXX No XXX
described [11] which is an ongoing prospective, hospi-
tal-based registry investigating risk factors and progno-
sis in patients with stable CAD. Subjects were included
if they had been previously hospitalized for ACS, had a
myocardial revascularization intervention or had under-
gone coronary angiography for chest pain and there
was a documentation of CAD. Coronary artery stenosis
was defined as at least 50% reduction in lumen diame-
ter of any of the three coronary arteries or their primary
branches.
Exclusion criteria were ACS or coronary artery bypass
grafting (CABG) within the previous 6 months, clinical
evidence of heart failure, chronic kidney disease (creati-
nine levels > 1.5 mg/dL), age >75 years or coexistent neo-
plasm or inflammatory disease.
Patients underwent clinical examination and special
attention was paid to reporting the risk factors and their
medication. The following definitions were used: hyper-
tension, blood pressure ≥140/90 mmHg or antihyperten-
sive treatment; hypercholesterolemia, total cholesterol
>200  mg/dL (5.2  mmol/L); diabetes mellitus, fasting
plasma glucose >125 mg/dL (6.94 mmol/L) or glucose
lowering treatment. Smokers were defined as those who
smoked at least one cigarette per day. Body mass index
was calculated as weight (in kilograms) divided by stand-
ing height (in square meters). Waist circumference was
also measured with the patient standing, after a regular
expiration, to the nearest centimeter, midway between
the lowest rib and the iliac crest.
Finally, all patients underwent an echocardiographic
study and the ejection fraction of left ventricle was
measured using the biplane method of discs (modified
Simpson’s rule).
Biochemical measurements
Venous fasting blood samples were collected between
08:00 and 09:00 for the same day measurements of lipids
and Hcy, while the remaining serum was stored and frozen
at −80oC until additional biomarkers assayed. Serum lev-
els of total cholesterol, high-density lipoprotein-choles-
terol and triglycerides were measured enzymatically on
an automatic analyzer (Dimension RXL; Dade Behring,
Marburg, Germany). LDL-C was calculated accord-
ing to the Friedewald formula. Plasma concentration of
Hcy was determined using chemiluminescent micropar-
ticle immunoassay technology, on the ARCHITECT
(Abbott). The intraassay and interassay coefficient of var-
iation of all measured biochemical parameters was <7%.
Follow-up
After the initial appointment, patients were followed up
at 6-month intervals by telephone interview by trained
cardiologists. In addition, between September and
December 2018, all patients were contacted to assess
their status. When a patient was not found, data were
Copyright © 2019 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
obtained by telephone interview with family members or
the treating physician. The primary endpoint was cardiac
mortality and the secondary endpoint was readmission
for ACS, arrhythmic event (life-threatening ventricular
arrhythmia) or ischemic stroke. All myocardial revas-
cularizations (percutaneous coronary intervention or
CABG) performed due to clinical deterioration were not
reported separately but were included in the ACS group.
Noncardiac deaths were not included in the analysis. In
the case of death, death certificates were obtained and
when this was not possible the cause of death was verified
by verbal or written contact with the treating physician.
The study was approved by the ethics committee of our
institution and all subjects gave signed informed consent.
Statistical analysis
Continuous variables are presented as means  ± SD,
while nonnormally distributed variables are presented
as medians and interquartile ranges. We used the t test
for independent samples to compare means for nor-
mally distributed variables or the Mann–Whitney test
for skewed variables. The Shapiro–Wilk test for normal-
ity was used in order to evaluate the t test assumption.
Categorical variables are presented as absolute and rel-
ative frequencies, and associations between categorical
variables were tested using the chi-square test. Hazard
ratios and corresponding 95% confidence intervals (CIs)
were calculated using Cox proportional hazards models.
Unadjusted Kaplan–Meier survival curves illustrated the
association between Hcy levels and the incidence of car-
diac deaths over time, and the log-rank test evaluated
statistical significance. A P value <0.05 was considered
significant. The SPSS version 25 (SPSS, Inc., Chicago,
Illinois, USA).
Results
The median follow up period was 6.1 years (3.7–7.1).
Of the initially recruited 876 patients, follow-up data
were not obtained from 25 patients due to several rea-
sons (denied to participate, change of address, etc.). In
addition, nine patients whose death was considered non-
cardiovascular were excluded. Therefore, analysis was
confined to 842 CAD patients. Baseline characteristics
according to occurrence of cardiac death are shown in
Table 1.
Of the 842 patients during follow-up 70 patients died
(cardiac death) (8.3%), while 258 (30.6%) met the sec-
ondary endpoints. Seven hundred four patients (83.6%)
were on statins. In univariate Cox regression analysis,
Hcy predicted the occurrence of cardiac death (hazard
ratio: 1.030; 95% CI: 1.018–1.042; P < 0.001) but not the
occurrence of secondary endpoints (hazard ratio: 1.010;
95% CI: 0.999–1.020, P = 0.081). Hcy remained an inde-
pendent predictor of cardiac death after adjustment for
conventional risk factors, ejection fraction and statin use
 Homocysteine and long-term prognosis in stable coronary artery disease Rallidis et al.  3

Table 1  Baseline clinical characteristics and drug treatment of the study population according to the occurrence of cardiac death

Characteristics Cardiac death (n = 70) No cardiac death (n = 772) P value

Age (years) 61.2 ± 13 59.1 ± 14 0.835

Males, n (%) 62 (88.6%) 667 (86.4%) 0.743
Hypertension, n (%) 40 (57.1%) 368 (47.7%) 0.162
Diabetes mellitus, n (%) 24 (34.3%) 177 (23.3%) 0.060
Family history for CAD, n (%) 19 (27.2%) 322 (41.6%) 0.063
Ejection fraction of left ventricle (%) 44.1 ± 10.9 50.5 ± 10.1 <0.001
BMI (kg/m2) 28.3 ± 4.6 28.7 ± 4.1 0.517
Waist circumference (cm) 102.7 ± 14 102.4 ± 11.8 0.897
Homocysteine (μmol/L) 19.1 ± 13.6 13.9 ± 8.6 <0.001
Total cholesterol (mg/dL/mmol/L) 169.8 ± 60/4.39 ± 1.55 173.7 ± 46/4.49 ± 1.19 0.504
Triglycerides (mg/dL) 122.0 (89.5–177.0) 129.5 (93.0–180.0) 0.369
Triglycerides (mmol/L) 1.38 (1.01–2.0) 1.46 (1.05–2.03)
LDL cholesterol (mg/dL/mmol/L) 110.3 ± 54/2.85 ± 1.4 107.2 ± 39/2.77 ± 1.0 0.550
HDL cholesterol (mg/dL/mmol/L) 39.8 ± 9.9/1.03 ± 0.26 41.3 ± 10.7/1.07 ± 0.28 0.216
Ejection fraction (%) 43.2 ± 13.2 50.8 ± 10.2 <0.001
Medication
  Antiplatelet treatment 62 (88.6%) 674 (87.3%) 0.906
  Lipid lowering medication 56 (81.4%) 648 (83.9%) 0.716
 Beta-blockers 47 (67.1%) 547 (70.9%) 0.709
  ACE inhibitors or AT1 blockers 53 (75.7%) 563 (72.9%) 0.713

ACE, angiotensin converting enzyme; BMI, body mass index; CAD, coronary artery disease; HDL, high-density lipoprotein; LDL, low-density lipoprotein.

Fig. 1

Event-free Kaplan–Meier survival curves in patients with stable coronary artery disease according to tertile levels of Hcy.

(hazard ratio: 1.030; 95% CI: 1.017–1.044, P < 0.001). In subgroup of patients with LDL-C levels below 70 mg/dL
addition, adjustment with other medical treatment (data (n = 155) after adjustment for conventional risk factors,
not shown) did not change the correlations. Finally, Hcy ejection fraction and statin use (hazard ratio: 1.082; 95%
was an independent predictor of cardiac death in the CI: 1.029–1.137; P = 0.002).

Copyright © 2019 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
4  Coronary Artery Disease  2019, Vol XXX No XXX
Patients in the highest tertile of Hcy (>14.1 μmol/L) had
three times higher risk of cardiac death compared with
patients in the lowest tertile (<10.3 μmol/L) (hazard
ratio = 3.036, 95% CI: 1.983–4.649, P < 0.001) [Fig. 1].
Discussion
In our population which comprised patients with stable
CAD, we found that serum Hcy levels predicted cardiac
mortality in a quantitative manner. The prognostic abil-
ity remained even after adjustment for conventional risk
factors and particularly statin use. No significant associa-
tions were found between Hcy levels and the occurrence
of secondary cardiovascular end-points. This is the first
study to affirm the prognostic role of Hcy in patients with
stable CAD in the era of statins, the main driver of mor-
tality reduction of the past decades. Interestingly, Hcy
preserved its prognostic ability for cardiac mortality even
in patients with optimal LDL-C levels, that is, LDL-C
levels below 70 mg/dL.
Hcy has been studied extensively in respect to its rela-
tionship with CVD. It was found repeatedly to be a
strong predictor of CAD occurrence among the gen-
eral population [12]. Veeranna et al. [13] found that the
addition of Hcy levels to the Framingham Risk Score,
improved significantly its ability to predict cardiovascu-
lar events in intermediate cardiovascular risk individuals.
The debate as to whether Hcy is a causal factor for CAD
or an ‘innocent’ bystander of the process remains active
[14]. It is partly driven by the generally negative results
of intervention trials that managed to reduce Hcy levels
effectively by B vitamins supplementation but did not
reduce the risk of cardiovascular events [15]. Maron and
Loscalzo [16], in their review, concluded that the results
of intervention trials were disappointing, but also high-
lighted some of their shortcomings, including that none
of them targeted at the treatment of very high levels
of Hcy, where a potential benefit might have appeared.
Mendelian randomization studies also failed to demon-
strate a causal role of Hcy for CAD development [17].
In our study, Hcy levels showed a long-term statistically
significant correlation with cardiac death in patients with
stable CAD independently of all traditional risk factors
and statin treatment. Numerous studies have correlated
high Hcy levels and increased short- or long-term cardiac
mortality in patients with ACS [6–8], but only two stud-
ies explored the prognostic value of patients with stable
CAD. In particular, one study [9] performed 20 years ago
reported that Hcy predicted adequately cardiovascu-
lar mortality in 587 patients with angiographically con-
firmed CAD after a median follow-up of 4.6 years. In
another study [10] of 247 patients with a previous his-
tory of premature myocardial infarction (men < 55 years,
women < 60 years) Hcy was an independent predictor of
total and cardiac mortality. However, none of these stud-
ies can directly be compared with ours because only 6 and
Copyright © 2019 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
6.5% of their population, respectively, were on hypolipi-
demics, while in our cohort, the vast majority were taking
statins.
Another observational study which recruited 1412
patients with stable CAD during the period 1994–1997
showed that increased Hcy levels were an independent
predictor of total mortality [18]. However, information on
concomitant medical treatment is missing.
We found no statistically significant relationship between
Hcy levels and the complex endpoint of major adverse
cardiovascular events. Other studies also demonstrate
mixed results. In accordance with our study, Retterstol et
al. [10] reported that Hcy had no effect on the composite
end point of major cardiac events. In another study with
a mixed population of both stable and unstable CAD,
there was no significant association between Hcy levels
and the complex endpoint of death, myocardial infarction
and rehospitalization for rest unstable angina in 483 CAD
patients [19].
Numerous possible mechanisms have been proposed, as
to how hyperhomocysteinemia might promote the pro-
gression of atherothrombosis [20–22]. Hcy is involved in
the promotion of intracellular production of free oxygen
species, inhibition of nitric oxide synthesis, activation
of thrombosis, endothelial dysfunction, hyperuricemia,
monocyte activation and production of inflammatory
mediators such as interleukin-8 and heme carrier pro-
tein. Another mechanism that has been suggested is
endothelial cell DNA hypomethylation associated with
accumulation of S-adenosylhomocysteine, which is a
precursor molecule of Hcy [23]. Hcy seems also to medi-
ate inhibition of cholesterol efflux from macrophages in
vitro and, thus, promote cholesterol accumulation inside
these cells [24].
Conclusion
Hcy is an independent long-term predictor of cardiac
death in patients with stable CAD, in the era of statins.
Acknowledgements
Conflicts of interest
There are no conflicts of interest.
References
1 Ridker PM. Residual inflammatory risk: addressing the obverse side of the
atherosclerosis prevention coin. Eur Heart J 2016; 37:1720–1722.
2 Cai A, Li L, Zhang Y, Mo Y, Mai W, Zhou Y. Lipoprotein(a): a promising
marker for residual cardiovascular risk assessment. Dis Markers 2013;
35:551–559.
3 Cybulska B, Kłosiewicz-Latoszek L. Homocysteine–is it still an important
risk factor for cardiovascular disease? Kardiol Pol 2015; 73:1092–1096.
4 Cheng Z, Yang X, Wang H. Hyperhomocysteinemia and endothelial dys-
function. Curr Hypertens Rev 2009; 5:158–165.
5 Undas A, Brozek J, Szczeklik A. Homocysteine and thrombosis: from basic
science to clinical evidence. Thromb Haemost 2005; 94:907–915.
6 Fu Z, Qian G, Xue H, Guo J, Chen L, Yang X, et al. Hyperhomocysteinemia
is an independent predictor of long-term clinical outcomes in Chinese
 Homocysteine and long-term prognosis in stable coronary artery disease Rallidis et al.  5
octogenarians with acute coronary syndrome. Clin Interv Aging 2015;
10:1467–1474.
7 Ma Y, Li L, Geng XB, Hong Y, Shang XM, Tan Z, et al. Correlation between
hyperhomocysteinemia and outcomes of patients with acute myocardial
infarction. Am J Ther 2016; 23:e1464–e1468.
8 Omland T, Samuelsson A, Hartford M, Herlitz J, Karlsson T, Christensen
B, Caidahl K. Serum homocysteine concentration as an indicator of
survival in patients with acute coronary syndromes. Arch Intern Med 2000;
160:1834–1840.
9 Nygård O, Nordrehaug JE, Refsum H, Ueland PM, Farstad M, Vollset SE.
Plasma homocysteine levels and mortality in patients with coronary artery
disease. N Engl J Med 1997; 337:230–236.
10 Retterstol L, Paus B, Bohn M, Bakken A, Erikssen J, Malinow MR, Berg K.
Plasma total homocysteine levels and prognosis in patients with previous
premature myocardial infarction: a 10-year follow-up study. J Intern Med
2003; 253:284–292.
11 Rallidis LS, Tellis CC, Lekakis J, Rizos I, Varounis C, Charalampopoulos A,
et al. Lipoprotein-associated phospholipase A(2) bound on high-density
lipoprotein is associated with lower risk for cardiac death in stable coro-
nary artery disease patients: a 3-year follow-up. J Am Coll Cardiol 2012;
60:2053–2060.
12 Peng HY, Man CF, Xu J, Fan Y. Elevated homocysteine levels and risk of car-
diovascular and all-cause mortality: a meta-analysis of prospective studies. J
Zhejiang Univ Sci B 2015; 16:78–86.
13 Veeranna V, Zalawadiya SK, Niraj A, Pradhan J, Ference B, Burack RC, et al.
Homocysteine and reclassification of cardiovascular disease risk. J Am Coll
Cardiol 2011; 58:1025–1033.
14 Ganguly P, Alam SF. Role of homocysteine in the development of cardiovas-
cular disease. Nutr J 2015; 14:6.
15 Martí-Carvajal AJ, Solà I, Lathyris D, Dayer M. Homocysteine-lowering inter-
ventions for preventing cardiovascular events. Cochrane Database Syst Rev
2017; 8:CD006612.
Copyright © 2019 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
16 Maron BA, Loscalzo J. The treatment of hyperhomocysteinemia. Annu Rev
Med 2009; 60:39–54.
17 Smulders YM, Blom HJ. The homocysteine controversy. J Inherit Metab Dis
2011; 34:93–99.
18 Anderson JL, Muhlestein JB, Horne BD, Carlquist JF, Bair TL,
Madsen TE, Pearson RR. Plasma homocysteine predicts mortality inde-
pendently of traditional risk factors and C-reactive protein in patients with
angiographically defined coronary artery disease. Circulation 2000; 102:
1227–1232.
19 Zairis MN, Ambrose JA, Manousakis SJ, Stefanidis AS, Papadaki OA,
Bilianou HI, et al.; Global Evaluation of New Events and Restenosis After
Stent Implantation Study Group. The impact of plasma levels of C-reactive
protein, lipoprotein (a) and homocysteine on the long-term prognosis
after successful coronary stenting: the Global Evaluation of New Events
and Restenosis After Stent Implantation Study. J Am Coll Cardiol 2002;
40:1375–1382.
20 Djuric D, Jakovljevic V, Zivkovic V, Srejovic I. Homocysteine and homo-
cysteine-related compounds: an overview of the roles in the pathology of
the cardiovascular and nervous systems. Can J Physiol Pharmacol 2018;
96:991–1003.
21 Esse R, Barroso M, Tavares de Almeida I, Castro R. The contribution of
homocysteine metabolism disruption to endothelial dysfunction: state-of-the-
art. Int J Mol Sci 2019; 20:867.
22 Li T, Yu B, Liu Z, Li J, Ma M, Wang Y, et al. Homocysteine directly interacts
and activates the angiotensin II type I receptor to aggravate vascular injury.
Nat Commun 2018; 9:11.
23 Kim J, Kim H, Roh H, Kwon Y. Causes of hyperhomocysteinemia and its
pathological significance. Arch Pharm Res 2018; 41:372–383.
24 Jin P, Bian Y, Wang K, Cong G, Yan R, Sha Y, et al. Homocysteine acceler-
ates atherosclerosis via inhibiting lxrα-mediated ABCA1/ABCG1-dependent
cholesterol efflux from macrophages. Life Sci 2018; 214:41–50.