Yusuf S, Gerstein H, Hoogwerf B, Pogue J, Bosch J, Wolffenbuttel BHR, Zinman B, for the HOPE Study Investigators . Ramipril and the Development of Diabetes. JAMA. 2001;286(15):1882–1885. doi:10.1001/jama.286.15.1882
Author Affiliations: McMaster University, Hamilton, Canada (Drs Yusuf and Gerstein and Mss Pogue and Bosch); Cleveland Clinic Foundation, Cleveland, Ohio (Dr Hoogwerf); Clinic of Maastricht, Maastricht, the Netherlands (Dr Wolffenbuttel); Mt Sinai and Toronto Hospitals, Toronto, Canada (Dr Zinman). A complete list of the HOPE investigators has been previously published.11
Context Type 2 diabetes is a growing clinical and public health problem. Preventive
efforts related to lifestyle modification are not always successful; therefore,
alternative prevention strategies need to be studied.
Objective To investigate the effectiveness of ramipril, an angiotensin-converting
enzyme inhibitor, in preventing diabetes among high-risk persons.
Design, Setting, and Participants The randomized, controlled Heart Outcomes Prevention Evaluation trial
of 5720 patients older than 55 years without known diabetes but with vascular
disease who were followed up for a mean of 4.5 years. The study included 267
hospitals in 19 countries and was conducted between 1994 and 1999.
Intervention Patients were randomly assigned to receive ramipril, up to 10 mg/d (n
= 2837), or placebo (n = 2883).
Main Outcome Measure Diagnosis of diabetes determined from self-report at follow-up visits
every 6 months, compared between the 2 groups.
Results One hundred and two individuals (3.6%) in the ramipril group developed
diabetes compared with 155 (5.4%) in the placebo group (relative risk [RR],
0.66; 95% confidence interval [CI], 0.51-0.85, P<.001).
Similar results were noted when different diagnostic criteria were used; in
the ramipril group, the RR for diagnosis of diabetes and hemoglobin A1c greater than 110% was 0.60 (95% CI, 0.43-0.85), for initiation of
glucose-lowering therapy, 0.56 (95% CI, 0.41-0.77), and for both, 0.51 (95%
CI, 0.34-0.76). These effects were also consistently seen in several subgroups
Conclusions Ramipril is associated with lower rates of new diagnosis of diabetes
in high-risk individuals. Because these results have important clinical and
public health implications, this hypothesis requires prospective confirmation.
Type 2 diabetes is an important and common risk factor for the development
of coronary artery disease, strokes, peripheral arterial disease, and renal
and eye disease. Currently, in North America, the direct and indirect costs
of diabetes and its complications exceeds $100 billion per year.1
This health and economic impact of diabetes is bound to increase, as the global
prevalence of diabetes rises from 4.2% to 5.4% by the year 2025.2
A growing amount of literature shows that the complications of diabetes
can be reduced or prevented by improving glucose control,3,4
lowering blood pressure5 and lipids,6 smoking cessation, and taking angiotensin converting
enzyme (ACE) inhibitors.7 An even more effective
approach to preventing these problems would be to prevent diabetes from developing.
Whereas recent evidence from trials suggests that lifestyle modifications
may reduce the risk of diabetes,8 the long-term
adherence to such interventions has not been high. Therefore alternative strategies
that are more easily implemented, safe and likely to prevent not only diabetes8 but also its chronic consequences, deserve to be investigated.
Recently, we demonstrated that the ACE inhibitor, ramipril, reduces
myocardial infarction, strokes, deaths, and the development of diabetic nephropathy
among high risk people both with and without a diagnosis of diabetes.7,9 We also observed that ramipril reduced
the development of diabetes in study participants without known diabetes at
randomization.9 This article describes this
finding in detail and explores possible explanations.
The design of the Heart Outcomes Prevention Evaluation (HOPE) trial
has been described in detail in previous publications. Briefly, individuals
who were 55 years or older with no evidence of left ventricular dysfunction
or heart failure and who had evidence of vascular disease or who had diabetes
and 1 other risk factor were eligible as long as they had no indication or
contraindication to receiving an ACE-inhibitor. The study was conducted in
267 hospitals in 19 countries from 1994 to 1999. All patients provided written
Of 10 576 eligible patients who participated in a run-in period
during which they received 2.5 mg ramipril once daily for 1 week followed
by matching placebo for 10 to 14 days, 1035 (9.8%) were excluded from randomization
(3.2% for side effects, 3.7% for lack of consent). Of the remaining 9541 patients,
3654 (38.3%) had a clinical diagnosis of diabetes and 5887 (61.7%) did not
at randomization. This article focuses primarily on the latter group of patients.
Of these patients, 5720 were randomized to receive up to 10 mg of ramipril
once per day or equivalent placebo. One hundred sixty-seven patients who were
randomized to receive a low dose (2.5 mg/day) of ramipril as part of the Study
to Evaluate Carotid Ultrasound changes with Ramipril and Vitamin E (SECURE).
Substudy results are not included. All patients were also randomized to receive
400 IU of vitamin E or placebo.
Follow-up visits occurred at 1 month and 6 months after randomization
and then every 6 months (mean follow-up of 4.5 years). At each visit, we documented
whether the diagnosis of diabetes had been made since the previous visit.
The primary outcome of this analysis is a new diagnosis of diabetes
recorded on the basis of self-report. This diagnosis was made blinded to treatment
allocation and, hence, is likely to be unbiased. Hemoglobin A1c
(HbA1c)levels and medications used among those diagnosed as having
diabetes were also recorded. The HbA1c levels were determined locally.
Values higher than 110% of the upper limit of normal for each laboratory were
considered to be biochemical confirmation of diabetes.
Survival curves utilizing the Kaplan Meier and log-rank procedures were
used to describe and compare the results in the 2 treatment groups. Because
of the factorial design, all analyses were stratified for randomization to
vitamin E or placebo. Subgroup analyses were conducted using tests of interaction
in the Cox regression model.
The baseline characteristics of the patients who did not have diabetes
are provided in Table 1. The proportion
of patients taking study ramipril or open label ACE-inhibitors in the active
group was 98.3% at 2 years and 89.7% at 4 years. The proportion taking open
label ACE-inhibitors in the control group was 11.6% and 27.4% respectively.
There were 102 individuals (3.6%) in the ramipril group compared with
155 (5.4%) in the placebo group (relative risk [RR], 0.66; 95% confidence
interval [CI], 0.51-0.85; P<.001) who reported
a new diagnosis of diabetes (Figure 1).
The proportion of patients diagnosed to have diabetes and a documented glycated
hemoglobin of 110% or more above the upper limit of normal (1.8% vs 3.0%;
RR, 0.60; 95% CI, 0.43-0.85; P = .003), those receiving
an oral glucose lowering agent or insulin (2.1% vs 3.6%; RR, 0.56; 95% CI,
0.41-0.77; P<.001). Those with all criteria (1.3%
vs 2.5%; RR, 0.51; 95% CI, 0.34-0.76; P<.001)
were significantly lower in the ramipril group compared with the placebo group.
Vitamin E and placebo did not differ in their effect on diabetes.
Because ramipril reduced the risk of cardiovascular events and diabetic
nephropathy, we assessed whether the higher occurrence of these clinical events
in placebo-treated patients increased the likelihood of ascertainment of diabetes
in this group. Similar stratified analyses by the occurrence of other outcomes
was also examined. As noted in Table 2
the impact of ramipril on the development of diabetes could not be explained
by any confounding factor such as preferential ascertainment in one group
vs the other or use of concomitant medications.
Figure 2 demonstrates the
results among subgroups of patients with different risk factors for developing
diabetes. The results are consistent among those with a waist-to-hip ratio
below or above the median of 0.93 and consistent among those with a body mass
index (BMI) of 27.7 kg/m2or less or higher
than 27.7 kg/m2, those with or without a history
or hypertension, those receiving or not receiving β-blockers or diuretics
at randomization. A higher proportion of individuals without diabetes who
were randomized to the placebo group than those randomized to the ramipril
group received diuretics or β-blockers (drugs that are associated with
glucose intolerance or diabetes) during the study. However, the RR for diabetes
in the subgroup of individuals who never took these drugs during the study
was consistent with the overall results (RR, 0.62; 95% CI, 0.43-0.90).
In 4074 patients weight was recorded at baseline and at study end. Weight
increased by a mean (SD) of 0.98 (6.93) kg in the active group and 0.76 (8.10)
kg in the control group.
These analyses indicate that ramipril reduces the risk of new diagnosis
of diabetes among individuals with no previous history of diabetes. The magnitude
of the benefit appears to be large and moreover, ACE-inhibitors also reduce
macrovascular and microvascular complications of diabetes.7
Although the data on new diagnoses of diabetes were collected prospectively
in the HOPE study, it was not a primary or secondary outcome of the trial.
Therefore the results should be interpreted with caution. Nevertheless, the
results are plausible given the clear statistical significance and consistency
of results across subgroups, as well as using a range of approaches to diagnosing
The Study to Evaluate Carotid Ultrasound changes in patients treated
with Ramipril and vitamin E (SECURE) supports our findings that fasting glucose
increased more with placebo (15.8 mg/dL [0.41 mmols]) than with ramipril (9.6
mg/dL [0.25 mmols]; P = .03).10
Among the patients with diabetes in the HOPE study, a significant reduction
in HbA1c levels during serial annual recordings occured during
the first 2 years (absolute difference, 0.2%)7
In the UK Prospective Diabetes Study (UKPDS)5
and in the Captopril Prevention Project,11
patients randomized to receive ACE inhibitors had lower levels of HbA1c or less development of diabetes compared with those taking β-blockers
or diuretics. It is not clear whether the differences in development of diabetes
observed in these studies are due to a protective effect of ACE inhibitors
or an adverse effect of β-blockers or diuretics.
Hypokalemia substantially impairs the insulin secretory response to
glucose,12 which may be favorably affected
by ACE inhibitors. ACE inhibitors also lower aldosterone secretion and renal
potassium wasting, which could preserve β-cell responsiveness. ACE inhibitors
may increase islet blood flow and pancreatic β-cell perfusion by reducing
angiotensin-2 mediated vasoconstriction in the pancreas.13
These effects may potentially slow or reverse the decline in β-cell function.
ACE inhibitors may reduce insulin resistance in skeletal muscles,14,15 increase insulin-mediated glucose
disposal thereby decreasing the need for pancreatic insulin secretion. The
increased insulin mediated glucose uptake by skeletal muscle in response to
an ACE inhibitor is due to increased bradykinin-mediated nitric oxide production
and not to reductions in angiotensin 2 production or action.16,17
Several observations suggest that agents that increase nitric oxide (such
as ACE-inhibitors) may also increase insulin-mediated glucose uptake. These
observations include (1) both insulin-mediated vasodilation and skeletal muscle
glucose metabolism are reduced in obese persons who do not have diabetes (ie,
individuals at risk for diabetes) and in individuals with type 2 diabetes,
(2) inhibition of nitric oxide production reproduces this effect in lean individuals,
and (3) the effect on insulin sensitivity is greater than can be accounted
for by just increased skeletal muscle blood flow.18
ACE inhibitors may also reduce insulin resistance at the liver and fat cell,
which would reduced hepatic glucose production and lower free fatty acid levels.19
Our data suggesting that ramipril, an ACE inhibitor, reduces the risk
of developing diabetes mellitus require confirmation because of the enormous
clinical and public health potential of these findings. We are therefore embarking
on a large prospective trial (Diabetes Reduction Assessment with Ramipril
and rosiglitizone Medication [DREAM]) among individuals with impaired glucose
tolerance to evaluate prospectively whether ramipril prevents diabetes.