Impact of Diabetic Retinopathy on Cardiac
Outcome After Coronary Artery Bypass Graft
CARDIOVASCULAR
Surgery: Prospective Observational Study
Takayuki Ono, MD, Takeki Ohashi, MD, Teiji Asakura, MD, Nagara Ono, MD,
Minoru Ono, MD, Noboru Motomura, MD, and Shinichi Takamoto, MD
Departments of Cardiothoracic Surgery and Anesthesiology, The University of Tokyo Hospital, Tokyo, and Department of
Cardiovascular Surgery, Heart Center, Nagoya Tokushukai General Hospital, Aichi, Japan
Background. Diabetic retinopathy is a manifestation of
more severe diabetes. We sought to assess the impact of
diabetic retinopathy on cardiac outcome of coronary
artery bypass graft surgery (CABG).
Methods. We prospectively assessed the status of the
retina of 74 consecutive diabetics who were referred for
first-time CABG, and compared cardiac outcome of
CABG in diabetics with retinopathy with that in those
without retinopathy. Cardiac events included recurrent
angina or congestive heart failure that needed admission
to hospital, myocardial infarction, repeat revasculariza-
tion, and cardiac death.
Results. Twenty-six diabetics had retinopathy and 48
diabetics did not have retinopathy. Diabetics with reti-
nopathy were likely to have higher hemoglobin A1c level
(p 0.048), and receive insulin treatment (p 0.0065). In
D iabetic retinopathy is a frequent and early sign of
microvascular complication. The risk of retinopa-
thy is directly related to the degree and duration of
hyperglycemia. Ophthalmologists all know that diabetics
with advanced retinopathy are commonly in poor gen-
eral health, and have a poor life expectancy. Epidemio-
logic evidence [19] demonstrated that diabetics with
retinopathy have high mortality rates of heart disease.
The Wisconsin Epidemiologic Study of Diabetic Retinop-
athy (WESDR), which consists of a total of 2,366 diabetic
patients who were followed up for 16 years, demon-
strated that after controlling for age and sex, diabetics
with poorer visual acuity and more severe retinopathy
had a significant increased risk for death from coronary
artery disease [8, 9].
Recently, my colleagues and I have shown [10] that the
presence of retinopathy is a strong independent risk
factor of all-cause mortality after coronary artery bypass
graft surgery (CABG) in diabetics. The 12-year overall
survival rate was 40% for diabetics with retinopathy,
compared with 88% for those without retinopathy. Poor
prognosis of diabetics with retinopathy might be contrib-
Accepted for publication July 18, 2005.
Address correspondence to Dr Takayuki Ono, Department of Cardiovas-
cular Surgery, The University of Tokyo Hospital, 7-3-1, Hongo, Bunkyo-
Ku, Tokyo, 113-8655, Japan; e-mail: takohno@hotmail.com.
2006 by The Society of Thoracic Surgeons
Published by Elsevier Inc
the 12 months of follow-up, 13 cardiac events occurred in
diabetics with retinopathy and 7 in those without reti-
nopathy (p 0.0021). Among diabetics with retinopathy,
heart failure or death due to heart failure accounted for
54% (7 of 13) of these cardiac events. Kaplan-Meier
analysis showed significant difference in cardiac event-
free survival between the two groups (p < 0.001). After
adjustment for differences in patients characteristics,
diabetic retinopathy remained a predictor of cardiac
event (adjusted relative risk 4.2, 95% confidence inter-
val, 1.5% to 11.9%; p 0.0067).
Conclusions. After CABG, diabetics with retinopathy
have a substantially increased risk of cardiac events,
especially of congestive heart failure.
(Ann Thorac Surg 2006;81:608 12)
2006 by The Society of Thoracic Surgeons
uted by the high prevalence of cardiac events after
CABG. However, little is known about cardiac outcome
of CABG in diabetics with retinopathy. In this study,
therefore, we prospectively assessed 12-month cardiac
outcome after CABG in diabetics with retinopathy, as
compared with those without retinopathy.
Patients and Methods
Patients referred to the Heart Center at Nagoya To-
kushukai General Hospital, Japan for CABG between
November 2001 and April 2003 were eligible for the study
provided they had diabetes mellitus treated with hypo-
glycemic agents or insulin injection, and provided that
they had stable angina or unstable angina. Patients were
excluded if they required concomitant cardiac proce-
dures, or they had previously undergone CABG, but
there were no eligibility restrictions for ejection fraction,
age, or urgency of surgery. All patients gave written
informed consent (approved by our Institutional Ethics
Committee), and completed ophthalmic examination for
detection of diabetic retinopathy within one week prior
to CABG by an ophthalmologist. According to a modifi-
cation of the Diabetic Retinopathy Study and the Early
Treatment Diabetic Retinopathy Study grading scale [10],
the severity in the worst affected eye was used and the
0003-4975/06/$32.00
doi:10.1016/j.athoracsur.2005.07.057
Ann Thorac Surg ONO ET AL 609
2006;81:608 12 RETINOPATHY AND CARDIAC EVENT AFTER CABG

Table 1. Baseline Characteristicsa

Diabetics With Diabetics Without
Retinopathy Retinopathy
Characteristics (n 26) (n 48) p Value

CARDIOVASCULAR
Demographic profile
Mean age - year 56.8 8.0 56.4 7.3 0.85
Female sex - no. 19 32 0.75
Status of diabetic retinopathy
Mild-to-moderate nonproliferative - no. 11
Severe nonproliferative - no. 7
Proliferative - no. 8
Medical history
Prior myocardial infarction - no. 13 21 0.63
Unstable angina - no. 10 23 0.59
Hypertension - no. 16 23 0.33
Current smoking - no. 13 31 0.32
Treatment with insulin - no. 10 5 0.0065
Clinical profile
Mean blood fasting glucose - mg/dL 111.8 22.2 107.1 19.0 0.098
Hemoglobin A1c - % 6.3 0.34 6.0 0.61 0.048
Mean serum total cholesterol - mg/dL 200 40.4 197.7 43.0 0.81
Mean serum triglycerides - mg/dL 163 59.4 146 70.3 0.32
Mean serum creatinine - mg/dL 1.2 0.3 1.1 0.3 0.14
Preoperative angiographic profile
Triple-vessel disease - no. 22 41 0.99
Mean ejection fraction - % 53.1 10.3 55.5 12.9 0.41
Coronary-artery-bypass grafting
ITA grafting - no. 23 45 0.99
Mean number of grafts 2.9 1.0 3.0 0.8 0.73
Complete revascularization - no. 24 45 0.99
a
Plus-minus values are mean SD.
ITA internal-thoracic-artery.

patients with retinopathy were grouped into three cate-
gories of retinopathy: those with mild-to-moderate non-
proliferative retinopathy, those with severe stage of non-
proliferative retinopathy, and those with proliferative
retinopathy. Each patient underwent off-pump CABG
performed by two surgeons experienced in the off-pump
procedure, and the goal of surgery was to obtain com-
plete revascularization [11, 12]. We have been performing
off-pump CABG since February 1998. Before the study
began, the off-pump procedure had been the routine
procedure for all patients who need CABG. After CABG,
antiplatelets were prescribed for all patients. Other rou-
tine medications included oral diuretics, cholesterol-
lowering agents, nitrates, calcium-channel blocker, beta-
blockers, and angiotensin-converting enzyme inhibitors
as appropriate. After discharge from the hospital, medi-
cations were given as deemed appropriate by the respon-
sible physicians.
Outcomes
All patients were followed for 12 months after the oper-
ation. The primary outcome variables were recurrent
angina or congestive heart failure that needed admission
to hospital, myocardial infarction, repeat revasculariza-
tion (surgery or angioplasty), and death due to cardiac
causes. Patients who died were classified according to
their cause of death. The remaining patients who had
had one or more nonfatal events were classified on the
basis of their first event. Congestive heart failure was
defined as paroxysmal nocturnal dyspnea or orthopnea,
and evidence of radiographic pulmonary edema. Myo-
cardial infarction was diagnosed based on typical symp-
toms, electrocardiographic changes, and cardiac enzyme
elevations.
Statistical Analysis
Data were analyzed using SPSS for Windows (version
10.0, SPSS, Chicago, IL). Dichotomous data were com-
pared by the 2 test or Fishers exact test for categorical
variables. Descriptive data for continuous variables are
presented as mean SD. The Students t test was used
for continuous variables. Cardiac event-free survival
rates were estimated according to Kaplan-Meier methods
and were compared using the log-rank test. We used a
Cox proportional-hazards model to adjust for differences
between patients characteristics of the two groups. Two-
sided p values of less than 0.05 were considered to
indicate a statistically significant difference.
 610 ONO ET AL Ann Thorac Surg
RETINOPATHY AND CARDIAC EVENT AFTER CABG 2006;81:608 12

Table 2. Cardiac Events After Coronary Artery Bypass Grafting

Diabetics With Diabetics Without
Retinopathy Retinopathy
(n 26) (n 48) p Value
CARDIOVASCULAR

Event in hospital
Death from cardiac causes 1 (3.8%) 1 (2.1%) 0.99
Fatal myocardial infarction 0 1 (2.1%) 0.99
Nonfatal myocardial infarction 0 0 -
Fatal congestive heart failure 1 (3.8%) 0 0.35
Nonfatal congestive heart failure 0 0 -
Repeated revascularization 1 (3.8%) 1 (2.1%) 0.99
Angioplasty 1 (3.8%) 0 0.35
Bypass surgery 0 1 (2.1%) 0.99
Recurrent angina 0 0 -
Any event 2 (7.7%) 2 (4.2%) 0.61
Event during follow-up
Death from cardiac causes 2 (7.7%) 0 0.12
Fatal myocardial infarction 0 0 -
Nonfatal myocardial infarction 2 (7.7%) 1 (2.1%) 0.28
Fatal congestive heart failure 2 (7.7%) 0 0.12
Nonfatal congestive heart failure 4 (15.4%) 1 (2.1%) 0.04
Repeated revascularization 1 (3.8%) 1 (2.1%) 0.99
Angioplasty 1 (3.8%) 1 (2.1%) 0.99
Bypass surgery 0 0 -
Recurrent angina 2 (7.7%) 2 (4.2%) 0.61
Any event 11 (42.3%) 5 (10.4%) 0.0026
Event at 1 year
Congestive heart failure 7 (26.9%) 1 (2.1%) 0.0022
Myocardial infarction angina repeat 6 (23.1%) 6 (12.5%) 0.32
revascularization
Any event at 1 year 13 (50.0%) 7 (14.6%) 0.0021

Results Cardiac events after CABG are summarized in Table 2.

Seventy-four diabetics were studied. Twenty-six (35.1%)
diabetics have retinopathy and 48 (64.9%) did not have
retinopathy. Baseline characteristics of the two groups
are shown in Table 1. Of 26 diabetics with retinopathy, 11
patients were grouped as having mild-to-moderate non-
proliferative retinopathy, seven as having severe nonpro-
liferative retinopathy, and eight as having proliferative
retinopathy. As compared with diabetics without reti-
nopathy, diabetics with retinopathy were more likely to
have higher hemoglobin A1c level (p 0.048), and
receive insulin treatment (p 0.0065). Other patients
characteristics, including ejection fraction and serum
creatinine level, did not differ significantly in the two
groups. The mean number of grafts per patient was 2.9
among diabetics with retinopathy and 3.0 among diabet-
ics without retinopathy. Complete revascularization was
achieved in 92.3% of diabetics with retinopathy and
93.8% of diabetics without retinopathy. Follow-up an-
giography was performed in 21 diabetics with retinopa-
thy (62 grafts) and 30 diabetics without retinopathy (94
grafts). The overall patency rates were 93.5% and 94.7%,
respectively (p 0.74). Twenty-three patients (31.1%) did
not undergo follow-up angiography because they had no
symptoms.
In the 12 months of follow-up, there were four cardiac
deaths and 16 nonfatal cardiac events (total 20 patients
[27% of study patients]). Most (80%) cardiac events oc-
curred after discharge from the hospital. At 12 months, 13
of 26 (50%) diabetics with retinopathy had cardiac events,
compared with 7 of 48 (15%) diabetics without retinopa-
thy (p 0.0021).
Cardiac events were categorized to two groups accord-
ing to their cause. Twelve patients had coronary athero-
sclerotic events (recurrent angina, myocardial infarction,
and repeat revascularization), and eight congestive heart
failures occurred. In diabetics without retinopathy, most
(6 out of 7) cardiac events were due to coronary athero-
sclerosis. In diabetics with retinopathy, congestive heart
failure or death due to congestive heart failure accounted
for 54% (7 out of 13) of cardiac events. It was diabetics
with advanced retinopathy (4 of those with severe non-
proliferative retinopathy, and 3 of those with proliferative
retinopathy) who had congestive heart failure. A much
larger proportion of diabetics with retinopathy had con-
gestive heart failure during follow-up than did diabetics
without retinopathy (p 0.0022).
A 12-month cardiac event-free survival was 50.0% (95%
confidence interval [CI], 41.2% to 59.8%) for diabetics
Ann Thorac Surg
2006;81:608 12
Fig 1. Kaplan-Meier estimates of survival free from angina, conges-
tive heart failure, myocardial infarction, repeat revascularization,
and death due to cardiac causes, according to the presence or ab-
sence of diabetic retinopathy (DR). (CABG coronary artery by-
pass graft surgery.)
with retinopathy, whereas 85.4% (95% CI, 80.3% to 90.5%)
for diabetics without retinopathy. As shown in Figure 1,
cardiac event-free survival curves differed significantly
according to the presence or absence of retinopathy (p
0.001). After adjustment for differences in patients char-
acteristics (hemoglobin A1c level and insulin treatment),
the presence of retinopathy remained a predictor of
cardiac event (adjusted relative risk, 4.2; 95% CI, 1.15 to
11.9; p 0.0067).
Comment
Diabetic retinopathy is a manifestation of more severe
diabetes, which is known to be associated with a higher
incidence of adverse cardiac events in any given time
period. In this study, we have shown that diabetics with
retinopathy had an unexpected increased risk of cardiac
events even after CABG, as compared with diabetics
without retinopathy. Especially, diabetics with advanced
retinopathy are at a high risk of developing congestive
heart failure.
Several epidemiologic studies demonstrated that the
population with diabetic retinopathy have a high inci-
dence of cardiac events in any given time period [1316].
The Framingham Heart Study and the Framingham Eye
Study examined the association between diabetic reti-
nopathy and the occurrence of cardiovascular events
including myocardial infarction, coronary insufficiency,
angina, congestive heart failure, stroke, and intermittent
claudication [13]. Among the 206 type 2 diabetics, the
odds rates for diabetic retinopathy and cardiovascular
events were 14.3, 2.0, and 0.3 for ages 52 to 64, 65 to 74,
and 75 to 85 years, respectively. The Milan Study on
Atherosclerosis and Diabetes [16] prospectively studied
735 patients with type 2 diabetes mellitus who were
screened for unknown asymptomatic coronary artery
disease. During 5-year follow-up a total of 42 ischemic
ONO ET AL 611
RETINOPATHY AND CARDIAC EVENT AFTER CABG
heart events occurred: 3 fatal myocardial infarction, 20
nonfatal myocardial infarction, and 19 angina pectoris.
Multivariate analysis demonstrated that diabetic retinop-
athy was an independent predictor of incidence of car-
diac events (adjusted hazard rate 2.37; p 0.036;
CARDIOVASCULAR
1.06 5.31). In addition, Wong and colleagues [17] clearly
demonstrated that the diabetics with retinopathy had an
increased risk of developing congestive heart failure
independent of coronary artery disease. However, there
has been little study examining cardiovascular events of
diabetics with retinopathy who undergo coronary revas-
cularization. Kim and colleagues [18] studied 115 diabet-
ics with retinopathy and 205 diabetics without retinopa-
thy who underwent percutaneous coronary intervention
(PCI), and assessed the association between the presence
of retinopathy and the occurrence of major cardiovascu-
lar events including death, myocardial infarction, cere-
brovascular event, and repeat revascularization. The
2-year survival rate was 96.3% for diabetics with retinop-
athy, as compared with 99.6% for diabetics without reti-
nopathy (p 0.02). However, the 2-year cardiovascular
event-free survival rate was not statistically different
between the two groups (67.7% and 73.8% in patients
with retinopathy and without retinopathy, respectively,
p 0.17). A recent report [19] demonstrated that the
severity of diabetic retinopathy correlates with the inci-
dence of in-stent restenosis after PCI, especially for
proliferative retinopathy.
It is generally accepted that hemoglobin A1c and
treatment with insulin injection are markers of the sever-
ity of diabetes. However, hemoglobin A1c is a measure of
short-term (3 or 4 months) glucose control. In diabetics
who strictly treated with insulin therapy, glucose control
would be good and the status of diabetes could be less
severe. In contrast, among ophthalmologists diabetic
retinopathy is used as a measure of how well glucose was
controlled over time and therefore a precise measure of
the severity of diabetes. Our study suggested that cardiac
outcome after CABG is closely related to the status of
diabetes mellitus independent of short-term glucose
control.
This study carries the clinical implications for diabetics
who undergo CABG. In the present study, in the 12
months of follow-up after CABG, adverse cardiac events
occurred in as many as 50% of diabetics with retinopathy.
This high prevalence of cardiac events might contribute
to poor prognosis in diabetics with retinopathy undergo-
ing CABG. Nowhere is this more visible than in the
eye, where the retinal circulation represents the cerebro-
vascular microcirculation [20], and the status of the retina
is a good predictor of cardiac events in diabetics. As a
result, pre-CABG retinal examination in diabetics has
become part of our practice guidelines. Diabetics with
retinopathy would be a target for more aggressive man-
agement after CABG to improve cardiac outcome of the
diabetic population. A better knowledge of the relation-
ship between diabetic retinopathy and cardiac outcome
after CABG would be useful for developing a future
therapeutic strategy for diabetics requiring CABG.
 612 ONO ET AL
RETINOPATHY AND CARDIAC EVENT AFTER CABG
Study Limitations
Limitations of this study include, first, that the effects of
postoperative medication on cardiac events were not exam-
ined because medications were given as deemed appropri-
CARDIOVASCULAR
ate by the responsible physicians after hospital discharge.
Second, diabetic retinopathy was assessed only within one
week prior to CABG, and we do not have information about
progression of retinopathy during the year of follow-up.
Third, our patient sample was small. A larger sample and
longer follow-up would be needed to assess influences of
the severity of retinopathy on cardiac outcome.
References
1. Davis MD, Hiller R, Magli YLM, et al. Prognosis for life in
patients with diabetes: relation to severity of retinopathy.
Trans Am Ophthalmol Soc 1979;77:144 70.
2. Sullivan P, Caldwell G, Alexander N, et al. Long-term
outcome after photocoagulation for proliferative diabetic
retinopathy. Diabet Med 1990;7:788 94.
3. Neil A, Hawkins M, Potok M, et al. A retrospective popula-
tion-based study of microalbuminuria as a predictor of
mortality in NIDDM. Diabetes Care 1993;16:996 1003.
4. Sasaki A, Uehara M, Horiuchi N, et al. A 15-year follow-up
study of patients with non-insulin-dependent diabetes mel-
litus (NIDDM) in Osaka, Japan. Factors predictive of the
prognosis of diabetic patients. Diabetes Res Clin Pract 1997;
36:417.
5. Helbig H, Kellner U, Bornfeld N, et al. Life expectancy of
diabetic patients undergoing vitreous surgery. Br J Ophthal-
mol 1996;80:640 3.
6. Rajala U, Pajunpaa H, Koskela P, et al. High cardiovascular
disease mortality in subjects with visual impairment caused
by diabetic retinopathy. Diabetes Care 2000;23:957961.
7. Henricsson M, Nilsson A, Heiji A, et al. Mortality in diabetic
patients participating in an ophthalmological control and
screening programme. Diabet Med 1997;14:576 83.
8. Klein R, Moss SE, DeMets DL. Relation of ocular and
systemic factors to survival in diabetes. Arch Intern Med
1989;149:266 72.
Ann Thorac Surg
2006;81:608 12
9. Klein R, Klein BE, Moss SE, et al. Association of ocular
disease and mortality in a diabetic population. Arch Oph-
thalmol 1999;117:148795.
10. Ono T, Kobayashi J, Sasako Y, et al. The impact of diabetic
retinopathy on long-term outcome following coronary artery
bypass graft surgery. J Am Coll Cardiol 2002;40:428 36.
11. Ono T, Asakura T, Ohashi T, et al. A simple method of
triggering balloon counterpulsation accurately during off-
pump coronary artery bypass surgery. Ann Thorac Surg
2005;79:7235.
12. Ono N, Ono T, Asakura T, et al. Usefulness of unipolar
epicardial ventricular electrogram for triggering of IABP
during off-pump coronary artery bypass surgery in patients
with hemodynamic instability complicating acute coronary
syndrome. Anesth Analg 2005;100:937 41.
13. Hiller R, Sperduto RD, Podgor MJ, et al. Diabetic retinopathy
and cardiovascular disease in type II diabetics. The Framing-
ham Heart Study and the Framingham Eye Study. Am J
Epidemiol 1988;128:4029.
14. Mittinen H, Haffner SM, Lehto S, et al. Retinopathy predicts
coronary heart disease events in NIDDM patients. Diabetes
Care 1996;19:1445 8.
15. Diglas J, Willinger C, Neu U, et al. Morbidity and mortality
in type 1 and 2 diabetes mellitus after the diagnosis of
diabetic retinopathy. Dtsch Med Wochenschr 1992;117:
1703 8.
16. Fagila E, Favales F, Paleari F, et al. Cardiac events in 735 type
2 diabetic patients who underwent screening for unknown
asymptomatic coronary heart disease: 5-year follow-up re-
port from the Milan study on atherosclerosis and diabetes
(MiSAD). Diabetes Care 2002;25:2032 6.
17. Wong TY, Rosamond W, Chang PP, et al. Retinopathy and
risk of congestive heart failure. JAMA 2005;293:639.
18. Kim YH, Hong MK, Song JM, et al. Diabetic retinopathy as a
predictor of late clinical events following percutaneous cor-
onary intervention. J Invasive Cardiol 2002;14:599 602.
19. Kim SH, Han JG, Oh BH, et al. Grade of diabetic retinopathy
has a powerful predictive relevance of in-stent restenosis
after percutaneous coronary intervention. Circulation 2004;
110[suppl 3]:III 482.
20. Ono T, Okita Y, Ando M, et al. Retrograde cerebral perfusion
in human brains. Lancet 2000;356:1323.