Poor Glycemic Control Predicts Coronary Heart Disease

Events in Patients With Type 1

Diabetes Without Nephropathy
Seppo Lehto, Tapani Rönnemaa, Kalevi Pyörälä, Markku Laakso

Abstract—Patients with type 1 diabetes mellitus, especially those with nephropathy, are at increased risk for coronary heart
disease (CHD). However, information on the predictive value of cardiovascular risk factors and the degree of
hyperglycemia with respect to CHD events in patients with type 1 diabetes without nephropathy is still incomplete.
Therefore, we performed a prospective study on risk factors for CHD in patients with type 1 diabetes free of clinical
nephropathy. At baseline examination, cardiovascular risk factor levels of CHD were determined in 177 patients with
type 1 diabetes (87 men and 90 women), age 45 to 64 years at baseline and $30 years at the time of diagnosis of
diabetes. These patients were followed up to 7 years with respect to CHD events. Altogether, 20 patients with type 1
diabetes (13 men [7.3%] and 7 women [3.9%]) died of CHD and 28 patients with type 1 diabetes (17 men [9.6%] and
11 women [6.2%]) had a serious CHD event (death from CHD or nonfatal myocardial infarction). In multivariate Cox
regression analysis, a previous history of myocardial infarction (hazard ratio [HR] and its 95% confidence interval, 8.0
[3.1 to 21.0], P,0.001), high glycohemoglobin A1 (.10.4%, the highest tertile, HR 5.4 [1.4 to 20.4], P50.013), and
the duration of diabetes (.16 years, the highest tertile, HR 4.2 [1.3 to 12.9], P50.013) were the only variables
associated with CHD death even after adjustment for other cardiovascular risk factors. These variables also predicted
the incidence of all CHD events. Our results indicate that poor metabolic control is a strong predictor of CHD events
in patients with late-onset type 1 diabetes without nephropathy, independently of other cardiovascular risk factors.
(Arterioscler Thromb Vasc Biol. 1999;19:1014-1019.)
Key Words: type 1 diabetes n glucose n glycohemoglobin A1 n coronary heart disease
Downloaded from http://ahajournals.org by on May 29, 2019

S everal studies have indicated that mortality and mor-

bidity rates of coronary heart disease (CHD) are 2 to 4
times higher among patients with type 1 diabetes than in
age-matched nondiabetic subjects.1,2 In particular, patients
with clinical diabetic nephropathy and a long duration of
diabetes have an extremely high morbidity from macro-
vascular disease.2–5 Proteinuria and microalbuminuria in
patients with type 1 diabetes are known to be associated
with several adverse cardiovascular risk factors, including
hypertension and lipoprotein abnormalities, characterized
mainly by elevated serum total and LDL cholesterol and
high total and VLDL triglyceride levels, low HDL choles-
terol, and low ApoA1 levels.6 – 8 Chronic hyperglyce-
mia9 –12 and poor glycemic control13–16 have been impli-
cated in the pathogenesis of microvascular complications
in patients with type 1 diabetes, but little data are available
on the role of these factors for the development of
macrovascular complications.
Prospective studies based on representative cohorts of
patients with late-onset type 1 diabetes and without clinical
diabetic nephropathy, in whom cardiovascular risk factors
including serum lipids, lipoproteins, and indicators of glyce-
mic control had been measured at baseline, were not avail-
able. This information is, however, particularly relevant
because a high occurrence of CHD in patients with type 1
diabetes with nephropathy could be caused mainly by adverse
effects of renal disease on cardiovascular risk factors and not
by the diabetes state itself.17 Therefore, we performed a
prospective study on risk factors for CHD in representative
cohorts of Finnish patients with late-onset type 1 diabetes and
without nephropathy.
Methods
Subjects
Baseline Study
A cross-sectional study aiming to compare the prevalence of
atherosclerotic vascular disease and its risk factors in middle-
aged diabetic patients and in nondiabetic subjects was performed
from 1982 to 1984 in eastern and western Finland. A detailed
description of the procedure of obtaining representative study
populations has been reported elsewhere.18 The study population
was chosen from the national register of the Social Insurance
Institution, which includes all Finnish citizens who receive
antidiabetic medication. In brief, the inclusion criteria were the
following: (1) age from 45 to 64 years, (2) diabetes diagnosed at

Received August 5, 1998; revision accepted September 22, 1998.

From the Department of Medicine (S.L., K.P., M.L.), Kuopio University Hospital, Kuopio; and the Department of Medicine (T.R.), Turku University
Central Hospital, and the Social Insurance Institution (T.R.), Turku, Finland.
Correspondence to M. Laakso, Department of Medicine, Kuopio University Hospital, SF-70210 Kuopio, Finland. E-mail markku.laakso@uku.fi
© 1999 American Heart Association, Inc.
Arterioscler Thromb Vasc Biol. is available at http://www.atvbaha.org

1014

the age of 30 years or later, and (3) area of residence and place of
birth in the Kuopio University Hospital district (eastern Finland)
or in the Turku University Hospital district (western Finland).
Altogether, 567 diabetic patients from eastern Finland and 639
diabetic patients from western Finland participated (participation
rates were 83% and 73%, respectively). Of these diabetic patients,
120 from eastern Finland and 158 patients from western Finland
had insulin treatment at the time of examination. C-peptide
measurements after intravenous glucagon were performed in 112
of the insulin-treated patients from eastern Finland and in 151
patients from western Finland (93% and 96%, respectively). In
228 patients, type 1 diabetes was diagnosed by postglucagon
C-peptide measurement (,0.20 nmol/L) and a history of ketoac-
idosis. In all of these patients insulin treatment had been initiated
within the first year after diagnosis. The cutoff point of 0.20
nmol/L was chosen because postglucagon C-peptide values below
this limit have been shown to be associated with the occurrence of
ketoacidosis in insulin-treated diabetic subjects.19 These 228
patients (113 men and 115 women) formed the study population.
Fifty-one patients (26 men and 25 women) with urinary protein
excretion .300 mg/L and/or serum creatinine .110 mmol/L in
women and 120 mmol/L in men were excluded. Thus, the final
study population comprised 177 patients (87 men and 90 women).
The study program previously described in detail18 was performed
during 1 outpatient visit at the Clinical Research Unit of the
University of Kuopio or the Rehabilitation Research Center of the
Social Insurance Institution in Turku. The visit included an interview
regarding the history of chest pain symptoms suggestive of CHD,
smoking, alcohol intake, physical activity, and the use of drugs. All
medical records of those subjects who reported during the interview
that they had been admitted to the hospital because of chest pain
symptoms were reviewed. Review of the medical records was
performed by 2 of us (M.L. in Kuopio and T.R. in Turku) after a
careful standardization of the methods between the reviewers. The
World Health Organization criteria for verified definite or possible
Downloaded from http://ahajournals.org by on May 29, 2019
myocardial infarction (MI) based on chest pain symptoms, ECG
changes, and enzyme determinations were used in the ascertainment
of the diagnosis of previous MI.20
Smoking status was based on an interview. In all statistical
analyses, subjects were classified as nonsmokers or current
smokers.
With the subject in a sitting position after a 5-minute rest, blood
pressure was measured with a mercury sphygmomanometer and read
to the nearest 2 mm Hg. A subject was classified as having
hypertension if he or she was receiving drug treatment for hyperten-
sion or if his or her systolic blood pressure was at least 160 mm Hg
or diastolic blood pressure at least 95 mm Hg.
Biochemical Methods
All laboratory specimens were drawn after a 12-hour fast at 8 AM.
Fasting plasma glucose was determined by the glucose oxidase
method (Boehringer). Glycohemoglobin A1 (GHbA1) was deter-
mined by affinity chromatography (Isolab) (reference range, 5.5% to
8.5%). The plasma C-peptide response to glucagon was determined
according to the method of Faber and Binder.21 Serum lipids and
lipoproteins were determined from fresh serum samples drawn after
a 12-hour overnight fast. Serum total cholesterol and triglycerides
were assayed by automated enzymatic methods (Boehringer). Serum
HDL cholesterol was determined enzymatically after precipitation of
LDL and VLDL lipoproteins with dextran sulfate/MgCl2.22 LDL
cholesterol was calculated by using the Friedewald formula as
follows: LDL cholesterol5total cholesterol2HDL cholesterol2
0.453total triglycerides.
Follow-Up Study
In 1990, a postal questionnaire containing questions about hos-
pitalization because of acute chest pain was sent to every
surviving participant of the original study cohort. All medical
records of those subjects who died between the baseline exami-
nation and December 31, 1989, or who reported in the question-
naire that they had been admitted to the hospital because of chest
pain symptoms between the baseline examination and December
31, 1989, were reviewed by 1 of us (S.L.). To ensure that the data
Lehto et al April 1999 1015
collection with regard to hospital-treated MIs was complete, a
computerized hospital discharge register was used to check
hospital admissions of all participants of the baseline study, and
in case of hospitalization for an acute CHD event, medical records
were checked. The modified World Health Organization criteria
for definite or possible MI based on chest pain symptoms, ECG
changes, and enzyme determinations were used in the ascertain-
ment of the diagnosis of MI similarly as in the baseline study.20
Copies of death certificates of those patients who had died were
obtained from the files of the Central Statistical Office of Finland.
In the final classification of the causes of death, hospital records
and autopsy records were used, if available. The mortality data
included in the present article are mortality from CHD (Interna-
tional Classification of Diseases 9, Codes 410 to 414).
Statistical Methods
Data analyses were conducted with the SPSSX and SPSS/PC1
programs (SPSS Inc). The results for continuous variables are
given as mean6SEM values or percentages. The differences
between the groups were assessed by the x2 test, or Student’s
2-tailed t test for independent samples when appropriate. The
univariate and multivariate Cox regression model23 was used to
investigate the association of cardiovascular risk factors with the
incidence of CHD events.
Approval of Ethics Committee
This study was approved by the Ethics Committee of Kuopio
University Central Hospital and the Turku University Central Hos-
pital. All study subjects gave informed consent.
Results
During 7-year follow-up (mean follow-up was 7.2 years in
men and women), 20 patients (13 men [7.3%] and 7 women
[3.9%] women) with type 1 diabetes died of CHD. Alto-
gether, 28 patients with type 1 diabetes (17 men [9.6%] and
11 women [6.2%]) had a serious CHD event (death from
CHD or nonfatal MI).
Table 1 summarizes baseline characteristics of patients
with type 1 diabetes, in relation to cardiovascular mortality
and morbidity during the 7-year follow-up, by sex. Data
from eastern and western Finland were combined, because
no significant differences existed between these areas in
the levels of cardiovascular risk factors with respect to
CHD events. Men with CHD were significantly older and
had more often a history of previous MI and had higher
GHbA1 than men without CHD. Women with CHD had
more often a history of previous MI and higher levels of
GHbA1 and a longer duration of diabetes than women
without CHD.
Table 2 reports unadjusted and adjusted hazard ratios
(HRs) for cardiovascular risk factors. The highest or lowest
(for HDL cholesterol) tertile limit was used as a cutoff point
for continuous variables. In univariate analysis, a previous
history of MI and GHbA1 and the duration of diabetes were
the only variables associated with the risk of CHD death
(P,0.001) and all CHD events (P,0.01). Figure 1 demon-
strates that poor glycemic control was associated with the
incidence of CHD death similarly throughout the follow-up
period. In multivariate analysis, a previous history of MI (HR
8.0 [3.1 to 21.0], P,0.001), high GHbA1 (.10.4%, HR 5.4
[1.4 to 20.4], P50.013), and the duration of diabetes (.16
years, HR 4.2 [1.3 to 12.9], P50.013) were associated with
CHD death even after adjustment for other cardiovascular
risk factors (age, sex, area of residence, previous MI, smok-
ing, body mass index, hypertension, total cholesterol, total
 1016 Coronary Heart Disease and Type 1 Diabetes

TABLE 1. Levels of Cardiovascular Risk Factors (Mean6SEM or Percentages) in

Relation to the Incidence of CHD Events (Death From CHD or Nonfatal MI) During
7-Year Follow-Up in Patients With Type 1 Diabetes
Men Women

Without CHD With CHD Without CHD With CHD

Variable (n570) (n517) (n579) (n511)
Age (y) 53.560.5 58.661.4† 56.260.6 56.461.8
Body mass index (kg/m2) 25.160.4 24.460.8 25.560.5 26.161.4
Insulin dose (IU/d) 4562 5166 4062 4364
Previous MI (%) 4.3 35.3§ 5.1 27.3‡
Hypertension (%) 28.6 23.5 45.6 36.4
Smoking (%) 25.7 35.3 11.4 9.1
Urinary protein (mg/L) 114.260.1 132.060.2 99.060.1 161.060.3*
Total cholesterol (mmol/L) 6.2760.15 6.7660.47 6.9460.15 7.3660.22
LDL cholesterol (mmol/L) 4.2260.13 4.5460.38 4.5960.13 4.9160.20
HDL cholesterol (mmol/L) 1.4760.05 1.5660.14 1.7660.05 1.8360.14
Triglycerides (mmol/L) 1.4060.09 1.5660.22 1.4460.15 1.4860.17
Plasma glucose (mmol/L) 9.760.6 11.761.0 10.460.5 11.861.8
GHb A1 (%) 9.460.2 10.560.4* 10.160.2 11.160.4
Duration of diabetes (y) 13.861.0 15.761.6 13.060.8 22.462.0‡
MI indicates hospital-verified MI and/or Q/QS change.
*P,0.05; †P,0.01; ‡P50.001; §P,0.001.

triglycerides, and HDL cholesterol). In a similar manner, CHD, we included these variables into the Cox regression
previous MI, high GHbA1, and a long duration of diabetes models also as continuous variables. In addition, we used
were associated significantly with all CHD events in univar- different cutoff points to confirm that the use of the highest
Downloaded from http://ahajournals.org by on May 29, 2019

iate analyses. Multivariate Cox analyses demonstrated that
HRs for previous MI (3.4 [1.5 to 7.9], P50.004), high GHbA1
(2.8 [1.2 to 6.9], P50.021), and long duration of diabetes
(HR 3.9 [1.6 to 9.3], P50.002) remained almost unchanged
when adjusted for other cardiovascular risk factors. Because
lipids and lipoproteins were not associated with the risk for
tertile limit does not underestimate the significance of these
variables. Dyslipidemia was not a significant risk factor in
any of these analyses.
We did all statistical analyses also by including patients
with nephropathy in our analyses (data not shown). These
analyses also demonstrated that poor glycemic control (high

TABLE 2. Unadjusted and Adjusted HRs and Their 95% Confidence Intervals (CI) for
Cardiovascular Risk Factors to Increase the Risk for CHD Events During 7-Year Follow-Up in
Patients With Type 1 Diabetes (Cox Regression Model)
CHD Mortality (20/177) HR (95% CI) All CHD Events (28/177)

Unadjusted Multiple Adjusted* Unadjusted Multiple Adjusted*

Previous MI 6.0 (2.9–12.7)§ 8.0 (3.1–21.0)§ 3.9 (1.9–7.8)§ 3.4 (1.5–7.9)‡
Total cholesterol 1.6 (0.7–4.0) 0.8 (0.2–3.0) 1.5 (0.7–3.2) 0.9 (0.3–2.5)
(.7.2 mmol/L)
LDL cholesterol 1.0 (0.4–2.7) 0.4 (0.1–1.6) 1.3 (0.6–2.8) 0.7 (0.3–1.9)
(.4.82 mmol/L)
Total triglycerides 0.8 (0.3–2.2) 0.5 (0.2–1.6) 1.5 (0.7–3.2) 1.4 (0.6–3.4)
(.1.4 mmol/L)
HDL cholesterol 1.2 (0.6–2.5) 0.2 (0.1–1.1) 0.8 (0.3–2.1) 0.5 (0.1–1.8)
(,1.40 mmol/L)
Fasting plasma glucose 1.7 (0.7–4.2) 2.7 (1.0–7.7) 1.5 (0.7–3.3) 1.7 (0.7–3.9)
(.12.6 mmol/L)
Glycated hemoglobin A1 6.8 (2.5–19.0)§ 5.4 (1.4–20.4)‡ 3.4 (1.6–7.4)‡ 2.8 (1.2–6.9)†
(.10.4%)
Duration of diabetes 5.0 (2.0–12.4)§ 4.2 (1.3–12.9)‡ 4.5 (2.0–7.4)§ 3.9 (1.6–9.3)‡
(.16 y)
*Adjusted for age, sex, area, previous MI, smoking, body mass index, hypertension, total cholesterol, triglycerides
(log), HDL cholesterol, HbA1, and duration of diabetes. Total cholesterol was not adjusted for LDL cholesterol and vice
versa. HbA1 was not adjusted for plasma glucose and vice versa.
†P,0.05; ‡P#0.01; §P#0.001.

Figure 1. Survival curves for CHD death in patients with type 1
diabetes, by baseline GHbA1 (.10.4% versus #10.4%) during
the 7-year follow-up (univariate Cox regression model).
GHbA1) was the strongest predictor for CHD death or for all
CHD events. In a similar manner, the exclusion of patients
with a history of MI before the baseline study (9 men and 7
women) did not change the results (data not shown).
We also analyzed whether the predictive value of poor
glycemic control with respect to CHD death was modified
by the duration of diabetes (Figure 1). Median values of
GHbA1 (10.0%) and known duration of diabetes (13.0
years) were used as cutoff points. The impact of poor
glycemic control on the risk for CHD death was seen
independently of diabetes duration. Although the long
duration of diabetes increased the risk for CHD death, this
effect was much weaker than that exercised by poor
metabolic control (Figure 2).
Downloaded from http://ahajournals.org by on May 29, 2019
Discussion
Our 7-year follow-up study is the first population-based
study demonstrating the important role of glycemic control
as a predictor of CHD mortality and morbidity in patients
with late-onset type 1 diabetes without nephropathy. In
univariate analysis, the risk for CHD death was 7-fold and
the risk for all CHD events (CHD death or nonfatal MI)
3-fold among patients with type 1 diabetes with high
GHbA1 (.10.4%) compared with those with better glyce-
mic control. The association remained statistically signif-
icant even after adjustment for other cardiovascular risk
factors (Table 2). Furthermore, the impact of poor glyce-
mic control was independent of diabetes duration. Because
our study was based on only 1 single measure of glycemic
Figure 2. The 7-year incidence (%) of CHD death, by median
baseline GHbA1 and the duration of diabetes in patients with
type 1 diabetes.
Lehto et al April 1999 1017
control at baseline, it is likely that our results are even an
underestimation of the deleterious effects of hyperglyce-
mia on the risk for CHD.
Several previous studies have indicated that hypergly-
cemia predicts microvascular complications in patients
with type 1 diabetes.10 –12 Data on the risk for macrovas-
cular complications have been much more limited. Deckert
et al24 followed 259 patients with type 1 diabetes and with
slightly elevated urinary albumin excretion (age range, 19
to 51 years at baseline) for 11 years. Elevated albumin
excretion rate, but not GHbA1c, predicted the incidence of
atherosclerotic vascular disease (CHD, stroke, and periph-
eral vascular disease). GHbA1 was also not a predictor for
CHD in the Pittsburgh Epidemiology of Diabetes Compli-
cations Study.25 These 2 recent studies have significant
differences compared with our study. Both included a
much younger cohort of type 1 patients with early-onset
diabetes. Furthermore, the study by Lloyd et al25 included
patients with overt nephropathy, and the study by Deckert
et al24 combined all manifestations of atherosclerotic
vascular disease as their end point, which makes it impos-
sible to clarify risk factors for CHD in their study. That
poor metabolic control is causally associated with the risk
for macrovascular complications in type 1 diabetes has
recently gained substantial support by the Diabetes Control
and Complications Trial.26 This trial demonstrated that not
only microvascular complications, but also macrovascular
complications, were reduced (by 41%) in patients with
type 1 diabetes with good glycemic control compared with
conventionally treated diabetic patients, although the re-
duction in macrovascular events was not statistically
significant.
In patients with type 1 diabetes, total cholesterol, HDL
cholesterol, and total triglycerides are usually within
normal limits when blood glucose is controlled.27,28 Good
glycemic control in type 1 diabetes usually reduces LDL
and VLDL to normal levels29 and may raise HDL even
above the normal range.28 Abnormal lipoprotein metabo-
lism could theoretically contribute to premature athero-
sclerosis in patients with type 1 diabetes. However, in our
study, high levels of total and LDL cholesterol and total
triglycerides and low levels of HDL cholesterol failed to
predict CHD events in patients with type 1 diabetes
although our study population included a substantial num-
ber of patients with poor metabolic control (Table 2). This
could be the result of a limited number of CHD events in
our study population, but it may also indicate that in
late-onset type 1 diabetes without nephropathy, abnormal-
ities in lipids and lipoproteins do not play such a crucial
role in the development of CHD compared with the effects
of poor glycemic control. Smoking, hypertension, and
insulin dose were no different among patients with type 1
diabetes with and without CHD (Table 1).
Our study population including patients with type 1
diabetes without nephropathy provided an excellent oppor-
tunity to assess the relation between glycemic exposure
and the risk for CHD in type 1 diabetes. Based on our
findings, it is suggested that the crucial factor in the
development of CHD in patients with late-onset type 1
diabetes is hyperglycemia, because the only factors that
significantly predicted CHD events in our patients were
 1018 Coronary Heart Disease and Type 1 Diabetes
high GHbA1 and the long duration of diabetes (duration of
hyperglycemia). This implies that the risk for CHD in our
study must be explained by direct effects of hyperglycemia
itself, because clinical and biochemical characteristics
were not otherwise different between those who had a
CHD event compared with those who did not (Table 1).
However, long-lasting hyperglycemia often leads to dia-
betic kidney disease and cardiovascular risk factors be-
come abnormal when microalbuminuria, proteinuria, or
elevation of creatinine level are present. Therefore, in
addition to direct harmful effects of hyperglycemia, indi-
rect effects on cardiovascular risk factors also take place
that probably explain the massive increase in the risk for
CHD in patients with type 1 diabetes and overt
nephropathy.
Many potential biochemical and clinical mechanisms
may explain why hyperglycemia itself, independently of
changes in cardiovascular risk factors, can increase the risk
for CHD.30 Hyperglycemia is related to abnormalities in
lipoprotein particle composition, which in turn are known
to be atherogenic. Furthermore, hyperglycemia has been
reported to accelerate oxidation of lipoproteins31 and to
induce and worsen insulin resistance and hyperinsulin-
emia, both of these effects being linked with an increased
risk for atherosclerotic vascular disease.32–35 Hyperglyce-
mia can also accelerate thrombus formation among dia-
betic patients.36,37 Finally, long-lasting hyperglycemia can
cause irreversible glycation of proteins in the arterial wall,
Downloaded from http://ahajournals.org by on May 29, 2019
which may also contribute to the development of vascular
complications.37
Significant differences exist between type 1 diabetes and
type 2 diabetes regarding cardiovascular risk factors pre-
dicting CHD. In type 2 diabetes, dyslipidemia (high LDL
cholesterol, high total triglycerides, and low HDL choles-
terol) is the most important determinant of CHD events.38
According to recent studies, hyperglycemia is also associ-
ated with the risk for CHD in type 2 diabetes but it is a
weaker risk factor for CHD than is dyslipidemia.39 In
contrast, according to the present study, poor metabolic
control dominates other risk factors, including dyslipid-
emia, in patients with type 1 diabetes without diabetic
kidney disease. Whether the relation between GHbA1 and
the risk for CHD is linear or nonlinear cannot be solved by
this study because of a limited number of CHD events.
Because poor glycemic control plays an important role
in the development of macrovascular complications in type
1 diabetes, it is reasonable to assume that its treatment
reduces the risk for CHD in patients with type 1 diabetes.
The findings of the Diabetes Control and Complications
trial are also in accordance with this view.26 Therefore, the
correction of hyperglycemia seems to be rational, not only
because of the prevention of microvascular complications,
but also because it may decrease the risk for atheroscle-
rotic vascular disease in patients with type 1 diabetes.
Acknowledgments
This work was supported by grants from the Academy of Finland,
the Finnish Heart Research Foundation, and the Aarne and Aili
Turunen Foundation (to M.L. and S.L.).
References
1. Pyörälä K, Laakso M, Uusitupa M. Diabetes and atherosclerosis: an
epidemiologic view. Diabetes Metab Rev. 1987;3:463–524.
2. Krolewski AS, Kosinski EJ, Warram JH, Leland OS, Busick EJ, Asmal
AC, Rand LI, Christlieb AR, Bradley RF, Kahn CR. Magnitude and
determinants of coronary artery disease in juvenile-onset, insulin-
dependent diabetes mellitus. Am J Cardiol. 1987;59:750 –755.
3. Koivisto VA, Stevens LK, Mattock M, Ebeling P, Muggeo M,
Stephenson J, Idzior-Walus B. Cardiovascular disease and its risk factors
in IDDM in Europe. Diabetes Care. 1996;19:689 – 697.
4. Jensen T, Borch-Johnsen K, Kofoed-Enevoldsen A, Deckert T. Coronary
heart disease in young type 1 (insulin dependent) diabetic patients with
and without diabetic nephropathy: incidence and risk factors. Diabe-
tologia. 1987;30:144 –148.
5. Borch-Johnsen K, Kreiner S. Proteinuria-value as predictor of cardiovas-
cular mortality in insulin-dependent diabetes mellitus. BMJ. 1987;294:
1561–1564.
6. Howard BV, Savage BJ, Bennion LJ, Bennett PH. Lipoprotein compo-
sition in diabetes mellitus. Atherosclerosis. 1978;30:153–162.
7. Goldberg RB. Lipid disorders in diabetes. Diabetes Care. 1981;4:
561–572.
8. Briones ER, Mao SJT, Palumbo PJ, O’Fallon WM, Chenoweth W, Kottke
BA. Analysis of plasma lipids and lipoproteins in insulin-dependent and
non-insulin-dependent diabetics. Metabolism. 1984;33:42– 49.
9. Siperstein MD. Diabetic microangiopathy and the control of blood
glucose. N Engl J Med. 1983;309:1577–1579.
10. Teuscher A, Schnell H, Wilson PW. Incidence of diabetic retinopathy and
relationship to baseline plasma glucose and blood pressure. Diabetes
Care. 1988;11:246 –251.
11. Chase HP, Jackson WE, Hoops SL, Cockerham RS, Archer PG, O’Brien
D. Glucose control and the renal and retinal complications of insulin-
dependent diabetes. JAMA. 1989;261:1155–1160.
12. Klein R, Klein BE, Moss SE, Cruickshanks KJ. Relationship of hyper-
glycemia to the long-term incidence and progression of diabetic retinop-
athy. Arch Intern Med. 1994;154:2169 –2178.
13. Pirart J. Diabetes mellitus and its degenerative complications: a pro-
spective study of 4400 patients observed between 1947 and 1973.
Diabetes Care. 1978;1:168 –188.
14. Janka HU, Warram JH, Rand LI, Krolewski AS. Risk factors for pro-
gression of background retinopathy in long-standing IDDM. Diabetes.
1989;38:460 – 464.
15. D’Antonio JA, Ellis D, Doft BH, Becker DJ, Drash AL, Kuller LH,
Orchard TJ. Diabetes complications and glycemic control: the Pittsburgh
Prospective Insulin-Dependent Diabetes Cohort Study Status Report after
5 yr of IDDM. Diabetes Care. 1989;12:694 –700.
16. Klein R, Klein BE, Moss SE, Davis MD, De Mets DL. Glycosylated
hemoglobin predicts the incidence and progression of diabetic retinopa-
thy. JAMA. 1988;260:2864 –2871.
17. Borch-Johnsen K, Kremer S. Proteinuria: value as predictor of cardio-
vascular mortality in insulin-dependent diabetes mellitus. BMJ. 1987;294:
1651–1654.
18. Laakso M, Rönnemaa T, Pyörälä K, Kallio V, Puukka P, Penttilä I.
Atherosclerotic vascular disease and its risk factors in non-insulin-
dependent diabetic and nondiabetic subjects in Finland. Diabetes Care.
1988;11:449 – 463.
19. Madsbad S, Alberti KG, Binder C, Burrin JM, Faber OK, Krarup T,
Regeur L. Role of residual insulin secretion in protecting against keto-
acidosis in insulin-dependent diabetes. BMJ. 1979;2:1257–1259.
20. World Health Organization. Proposal for the Multinational Monitoring of
Trends and Determinants in Cardiovascular Disease and Protocol
(MONICA Project). WHO/MNC/82.1 Rev. 1. Geneva, Switzerland:
World Health Organization; 1983.
21. Faber OK, Binder C. C-peptide response to glucagon: a test for the
residual B-cell function in diabetes mellitus. Diabetes. 1977;26:605– 610.
22. Kostner G. Enzymatic determination of cholesterol in high density
lipoprotein fractions prepared by polyanion precipitation. Clin Chem.
1976;22:695.
23. Cox DR. Regression models and life-tables. J R Stat Soc. 1972;34:
187–201.
24. Deckert T, Yokoyama H, Mathiesen E, Ronn B, Jensen T, Feldt-
Rasmussen B, Borch-Johnsen K, Jensen JS. Cohort study of predictive
value of urinary albumin excretion for atherosclerotic vascular disease in
patients with insulin dependent diabetes. BMJ. 1996;312:871– 874.
25. Lloyd CE, Kuller LH, Ellis D, Becker DJ, Wing RR, Orchard TJ.
Coronary artery disease in IDDM: gender differences in risk factors but
not risk. Arterioscler Thromb Vasc Biol. 1996;16:720 –726.

26. The Diabetes Control and Complications Trial Research Group. The
effect of intensive treatment of diabetes on the development and pro-
gression of long-term complications in insulin-dependent diabetes
mellitus. N Engl J Med. 1993;329:977–986.
27. Nikkilä EA. Plasma lipid and lipoprotein abnormalities in diabetes. In:
Jarrett RJ, ed. Diabetes and Heart Disease. Amsterdam, The Netherlands:
Elsevier Science Publishers, B.V.; 1984:134 –167.
28. Nikkilä EA, Hormila P. Serum lipids and lipoproteins in insulin-treated
diabetics. Demonstration of increased high density lipoprotein concen-
trations. Diabetes. 1978;27:1078 –1085.
29. Rosenstock JG, Vega GL, Raskin P. Improved diabetic control decreases
LDL apoB synthesis in type I diabetes mellitus. Arteriosclerosis. 1985;
5:513A.
30. Ruderman NB, Williamson JR, Brownlee M. Glucose and diabetic
vascular disease. FASEB J. 1992;6:2905–2914.
31. Lyons TJ. Lipoprotein glycation and its metabolic consequences.
Diabetes. 1992;41(suppl 2):67–73.
32. Pyörälä K. Relationship of glucose tolerance and plasma insulin to the
incidence of coronary heart disease: results from two population studies
in Finland. Diabetes Care. 1972;2:131–141.
Downloaded from http://ahajournals.org by on May 29, 2019
Lehto et al April 1999 1019
33. Ducimetiere P, Eschwege E, Papoz L, Richard JL, Claude JR, Rosselin G.
Relationship of plasma insulin level to the incidence of myocardial
infarction and coronary heart disease mortality in middle-aged popu-
lation. Diabetologia. 1980;19:205–210.
34. Welborn TA, Wearne K. Coronary heart disease incidence and cardio-
vascular mortality in Busselton with reference to glucose and insulin
concentration. Diabetes Care. 1979;2:154 –160.
35. Rönnemaa T, Laakso M, Pyörälä K, Kallio V, Puukka P. High fasting
plasma insulin is an indicator of coronary heart disease in non-insulin-
dependent diabetic patients and nondiabetic subjects. Arterioscler
Thromb. 1991;11:80 –90.
36. Badimon JJ, Fuster V, Chesebro JH, Badimon L. Coronary atherosclero-
sis: a multifactorial disease. Circulation. 1993;87(suppl 3):II-3–II-16.
37. Colwell JA. Vascular thrombosis in type II diabetes. Diabetes. 1993;
42:8 –11. Editorial.
38. Laakso M. Epidemiology of diabetic dyslipidemia. Diabetes Rev. 1995;
3:408 – 422.
39. Laakso M. Glycemic control and the risk of coronary heart disease in
patients with non-insulin-dependent diabetes mellitus. Ann Intern Med.
1996;124:127–130.