Evidence-Based Hypertension Treatment in Patients With Diabetes


  • Mariana Garcia-Touza MD,

    1. From the Division of Endocrinology, Diabetes & Metabolism, Department of Internal Medicine, University of Missouri ; 1the Department of Medical Physiology and Pharmacology ; 2 and the Harry S. Truman VA Medical Center, Columbia, MO3
    Search for more papers by this author
  • and 1 James R. Sowers MD 1,2,3

    1. From the Division of Endocrinology, Diabetes & Metabolism, Department of Internal Medicine, University of Missouri ; 1the Department of Medical Physiology and Pharmacology ; 2 and the Harry S. Truman VA Medical Center, Columbia, MO3
    Search for more papers by this author

  • This research was supported by the National Institutes of Health (R01 HL73101-01A and R01 HL107910-01) and Veterans Affairs Merit System 0018 (JRS).

Mariana Garcia-Touza, MD, D109 HSC Diabetes Center, One Hospital Drive, Columbia, MO 65212


J Clin Hypertens (Greenwich). 2011;00:00–00. ©2011 Wiley Periodicals, Inc.

Both impaired glucose tolerance and diabetes are associated with substantially increased prevalence of hypertension, cardiovascular and renal disease. The goal for hypertension treatment in diabetic patients is in evolution, because of recent clinical trials. For example, the results of the recent Action to Control Cardiovascular Risk in Diabetes—BP Arm (ACCORD BP) trial failed to show an additional benefit on cardiovascular event reduction at a mean systolic BP of 119 mm Hg. A post hoc analysis of 6,400 patients with type 2 diabetes from the International Verapamil-Trandolapril Study (INVEST) also failed to show additional cardiovascular risk reduction among patients who achieved a BP <130/80 mm Hg. While the evidence fails to support a lower BP goal to reduce coronary events, there was a risk reduction in stroke events both in ACCORD and the Appropriate Blood Pressure Control in NIDDM (ABCD) trial. A number of other clinical trials also demonstrate that when systolic pressures fall to less than 130 mm Hg, a reduction in stroke but not coronary disease events occurs. Thus, the precise BP goal for diabetic patients remains unresolved. We would posit that a BP goal of 135/85 mm Hg may be a reasonable compromise when viewing the impact of BP reduction on composite stroke and coronary artery disease in extant trials.

The association between diabetes and hypertension (HTN) was first described in residents of Rancho Bernardo, California, aged 50 to 79 years, surveyed from 1972 to 1974. The association was present in both men and women at all ages, and was the strongest for persons having the best evidence for diabetes (both historical and fasting hyperglycemia). Diabetes and HTN are partially linked to overweight and obesity, which are prevalent in both conditions.1 Adjustment for overweight/obesity reduced the association considerably, but a consistent association remained. The prevalence of HTN in patients who have type 2 diabetes is up to 3 times higher than in patients without diabetes.2 This association can be explained, in part, by the presence of a maladaptive interaction of factors such as insulin resistance (IR), chronic activation of the renin-angiotensin-aldosterone system (RAAS), and the sympathetic nervous system.2,3 The interrelationship between increased adiposity and maladaptive changes in the heart and kidney in patients with IR has been called the cardiorenal metabolic syndrome (CRS).3 Guidelines around the world have been consistent in suggesting that blood pressure (BP) should be lowered to <130/80 mm Hg in the diabetic population. In this article we will review randomized controlled trials that support a lower BP target in patients with coexistent diabetes and hypertension.

Relationship Between HTN, Dyslipidemia, and IR

Insulin is an anabolic hormone that promotes liver, muscle, and fat tissue glucose uptake and its storage as glycogen in liver and muscle. Insulin also suppresses the production of glucose and very low-density lipoprotein in the liver.4 IR is a condition in which the insulin metabolic signaling response in skeletal muscle, liver, and adipose tissue is impaired. IR results from a genetic predisposition, excess weight (especially central obesity), and lack of exercise. The state of IR can, in the absence of adequate β-cell response, lead to hyperglycemia, increased advanced glycation end products, increases in free fatty acids (FFAs), and lipoprotein abnormalities. This alteration in insulin metabolic signaling leads to increased adhesion molecule expression and decreased bioavailable nitric oxide (NO) in endothelial cells, as well as increased inflammation and vascular smooth muscle migration and proliferation.5,6 High levels of FFAs are also detrimental, leading to increased oxidative stress and diminished endothelial cell NO bioavailabilty.7,8 The decreased bioavailabilty of NO reduces endothelial-mediated vasorelaxation and promotes vascular stiffness. IR is also associated with inappropriate activation of the RAAS and the sympathetic nervous system.9 Elevations in angiotensin II (Ang II) and aldosterone have, in turn, been shown to promote an impairment in systemic insulin metabolic signaling that leads to endothelial dysfunction and myocardial functional abnormalities.9,10 These two factors, decreased bioavailable NO and activation of the RAAS, promote increased sodium reabsorption and vascular remodeling contributing to the development of HTN in diabetes. Further, oxidized low-density lipoprotein accumulation in the arterial wall results in decreased arterial elasticity and increased peripheral vascular resistance (Figure 1).

Figure 1.

 Potential effects of insulin resistance on endothelial cell (EC) and skeletal muscle cell (SMC) insulin metabolic signaling. Improved EC and SMC insulin-stimulated nitric oxide (NO) bioavailability and glucose utilization ultimately improves endothelial function and systemic insulin sensitivity. Akt indicates protein kinase B; eNOS, endothelial NO synthase; FFA, free fatty acids; GLUT4, glucose transporter-4; IR, insulin receptor; IRS-1, IR substrate-1; PI3-K, phosphoinositol 3-kinase; ROS, reactive oxygen species; TCA, tricarboxylic acid. With permission from Whaley-Connell A, Sowers JR. Hypertension and insulin resistance. Hypertension. 2009;54:462–464.

Concepts Regarding Improvement of Insulin Signaling and Secretion and Effects on BP in Persons With CRS

Both pharmacologic and nonpharmacologic strategies to improve insulin secretion and metabolic signaling generally also improve endothelial function and lower BP. For example, the addition of an Ang II receptor blocker (ARB) to a diuretic has been shown to improve insulin secretion in hypertensive patients with the CRS.11,12 In support of this observation, a recent report showed that 26 weeks of treatment with an ARB, valsartan, increased glucose-stimulated insulin release and insulin sensitivity.12 In this randomized study, the effects of the ARB or placebo on β-cell function and insulin sensitivity were assessed in patients with impaired fasting glucose and/or impaired glucose tolerance (IGT), using a combined hyperinsulinemic-euglycemic and hyperglycemic clamp with subsequent arginine stimulation and a 2-hour oral glucose tolerance test. ARB treatment increased first-phase and second-phase glucose-stimulated insulin secretion compared with placebo, whereas the enhanced arginine-stimulated insulin secretion was comparable between groups. Clamp-derived insulin sensitivity was significantly increased with ARB therapy, and this therapy significantly decreased diastolic BP (DBP) and systolic BP (SBP) compared with placebo.

Prior research suggested that thiazide diuretics reduce insulin sensitivity. For example, the Study of Trandolapril/Verapamil and IR (STAR) trial13 tested the hypothesis that a fixed-dose combination of trandolapril/verapamil sustained release (T/V) was superior to a fixed-dose combination of losartan/hydrochlorothiazide (L/H) on glucose tolerance in hypertensive patients with IGT. The study showed that in patients with IGT, normal kidney function, and HTN, the fixed-dose combination of T/V reduced the risk of new-onset diabetes compared with an L/H-based therapy. These data suggested that diuretics had adverse effects on insulin secretion and/or insulin sensitivity. Further, these data collectively suggest that RAAS blockers improve insulin secretion/sensitivity and/or IR and may partly negate some of the detrimental metabolic effects of thiazide diuretics.11–13

Evidence-Based Medical Treatment in Patients With HTN and Diabetes

The current guidelines for diabetes care issued by the American Diabetes Association and other committees around the world recommend a BP goal <130/80 mm Hg in patients with diabetes. One of the first studies to provide information on optimal threshold and targets for administering antihypertensive medication to patients with diabetes was the Pretereax and Diamicron MR Controlled Evaluation (ADVANCE) trial.14 This trial was designed to evaluate the risks of major microvascular disease and macrovascular events in individuals with type 2 diabetes, which would be reduced by intensive glucose control using a sulfonylurea-based regimen targeting a hemoglobin A1c of ≤6.5%: additional BP-lowering using a single-pill combination of an angiotensin-converting enzyme (ACE) inhibitor (perindopril) and a diuretic (indapamide). A total of 11,140 patients from Europe, Canada, Asia, and Australia were randomly assigned to the BP- and glucose-lowering arms. The study ended after 4.3 and 5.5 years, respectively. Compared with placebo, additional BP-lowering of 5.6/2.2 mm Hg was associated with reductions of 9% in the primary end point, 18% in cardiovascular death, 14% in total mortality, and 21% in total renal events (P<.01). The study investigators concluded that additional BP-lowering and intensive glucose control produced independent benefits and, when combined, they significantly reduced cardiovascular mortality and improved renal outcomes.

The result of the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial (ONTARGET)15 was published in the New England Journal of Medicine in 2008. In this study, 25,588 patients were enrolled in a multicenter, double-blinded randomized trial in 40 countries. The study enrolled patients older than 55 years with coronary, peripheral, or cerebrovascular disease or diabetes with end-organ damage. They were randomized to ramipril, telmisartan, or both. The primary outcome of the study was a composite of cardiovascular death, myocardial infarction (MI), stroke, or hospitalization for heart failure. The primary composite outcome was related to baseline SBP: SBP changes from baseline to event and average in-trial SBP. The results showed that the risk of myocardial infarction did not increase with baseline SBP and was unaffected by subsequent SBP change. In contrast, stroke risk progressively increased with baseline SBP and decreased with reduction. In patients with baseline SBP <130 mm Hg, cardiovascular mortality increased with further SBP reduction. A J curve (nadir around 130) (Figure 2) occurred in the relationship between in-treatment SBP and all outcomes except stroke.16 The conclusion of this study was that in high-risk patients, the benefits from SBP-lowering <130 mm Hg are determined by a reduction of stroke, myocardial infarction is unchanged, and cardiovascular mortality is unaffected or increased.

Figure 2.

 Relative risk of the primary outcome and of the main secondary outcome. The primary composite outcome was death from cardiovascular causes, myocardial infarction, stroke, or hospitalization for heart failure. The main secondary outcome was death from cardiovascular causes, myocardial infarction, or stroke, which was used as the primary outcome in the Heart Outcomes Prevention Evaluation (HOPE) trial 5. The P value is for the comparison with the noninferiority margins. With permission from Buchner N, Banas B, Kramer BK. Telmisartan, ramipril, or both in patients at high risk of vascular events. N Engl J Med. 2008;359:426–427.

The Action to Control Cardiovascular Risk in Diabetes (ACCORD) BP trial arm17 evaluated the effect of targeting an SBP of 120 mm Hg, as compared with a goal of 140 mm Hg, among patients with type 2 diabetes at high risk for cardiovascular events. A total of 4733 participants were randomly assigned to intensive therapy (BP <120 mm Hg) or standard therapy (BP <140 mm Hg), with the mean follow-up being 4.7 years. The intensive antihypertensive therapy in the ACCORD BP trial did not considerably reduce the primary cardiovascular outcome or the rate of death from any cause. The intensive arm of BP control reduced the rate of total stroke and nonfatal stroke. In this context, the estimated number needed to treat with intensive BP therapy to prevent one stroke over 5 years was 89. There were indicators of possible harm associated with intensive BP control (systolic <120 mm Hg), including a rate of serious adverse events in the intensive arm.

The observational subgroup analysis of the International Verapamil SR-Trandolapril Study (INVEST) was published in 2010.18 Participants in this study were at least 50 years old and had diabetes and coronary artery disease. Patients received a first-line therapy of either a calcium antagonist or β-blocker followed by an ACE inhibitor, a diuretic, or both to achieve SBP <130 mm Hg and DBP <85 mm Hg. The study had 3 groups: tight control, SBP <130 mm Hg; usual control, between 140 mm Hg to 130 mm Hg; and uncontrolled, ≥140 mm Hg. Patients in the usual control group (140–130 mm Hg) had a cardiovascular event rate of 12.6% vs 19.8% for the uncontrolled group (>140 mm Hg). However, little difference existed between those with usual control and those with tight control (>130 mm Hg). The author concluded that tight control of SBP <130/80 mm Hg was not associated with improved cardiovascular outcomes compared with usual control and emphasis should be placed on maintaining SBP between 130 mm Hg and 139 mm Hg.

Guidelines currently suggest that the primary antihypertensive drug strategy in patients with diabetes should include an ARB or an ACE inhibitor.19,20 If the baseline BP is >150/90 mm Hg, a second agent should be added, preferably a thiazide diuretic, because they can add cardiovascular protection.21,22 However, recent evidence suggests that calcium channel blockers, especially amlodipine can comparatively reduce cardiovascular events.23 To test the advantages of these two types of combination therapies, the Avoiding Cardiovascular Events In Combination Therapy in Patients Living With Systolic Hypertension (ACCOMPLISH) trial24 was conducted. The trial compared treatment outcomes of ACE inhibitor plus amlodipine and ACE inhibitor plus hydrochlorothiazide combinations.25 The study was performed in very high-risk hypertensive patients, half of whom had diabetes, and found that the amlodipine-based combination was more effective than thiazide-based therapies in reducing a composite of fatal and nonfatal cardiovascular events.

Providing Renal Protection and Prevention of Microalbuminuria in Diabetes

Patients with HTN and diabetes have a 7-fold greater risk for progressing to end-stage renal disease (ESRD) and 2 to 4-fold greater risk of developing cardiovascular disease. There are data suggesting that RAAS inhibition and BP control have added renoprotective and cardioprotective effects. The Bergamo Nephrologic Diabetic Complications Trial (BENEDICT)26 found that in patients with type 2 diabetes, HTN, and normoalbuminuria, ACE inhibitor therapy with trandolapril plus verapamil or trandolapril alone delayed the onset of microalbumninuria. This effect was significant even after adjustment for baseline and follow-up DBP and SBP, proving a specific renoprotective effect of ACE inhibitor therapy. The recently published ROADMAP trial27 investigated whether treatment with an ARB would delay or prevent the occurrence of microalbuminuria in patients with type 2 diabetes, HTN, and normoalbuminuria. The BP target was <130/80 mm Hg and was achieved in nearly 80% of patients taking olmesartan and 71% of patients taking placebo. The patients were followed for a median of 3.2 years. The primary outcome was the amount of time until the first onset of microalbuminuria and the secondary end point was the onset of renal and cardiovascular events. Olmesartan was associated with delayed onset of microalbuminuria. Fewer patients in the olmesartan group than in the placebo group had nonfatal cardiovascular events (3.6% compared with 4.1%) but a greater number had fatal cardiovascular events (0.7% compared with 0.1%). It is difficult to interpret this finding and it may be related to chance. The fatal cardiovascular events were more common in patients with known preexisting coronary heart disease and who were either in the lowest quartile of BP or the highest quartile of BP reduction during follow-up. This finding was also described in the ONTARGET trial and has been called the J-curve effect. In the Nephropathy Trial (ORIENT), a higher rate of death from cardiovascular causes was also noticed. Because of these findings, the Food and Drug Administration is reviewing these data.

The trials described above have clearly shown a benefit in the prevention and delayed progression of microalbuminuria in patients with HTN and diabetes. But it remains unclear whether the prevention of nephropathy and retinopathy will be slowed by early administration of drugs that block the RAAS. The Diabetic Retinopathy Candersartan Trial (DIRECT) evaluated the effect of candesartan in the prevention of microalbuminuria in patients who were normotensive and normoalbuminuric with type 1 or type 2 diabetes. The research program comprised 3 related studies to investigate the effect of candesartan on the incidence of retinopathy in type 1 or type 2 diabetic patients who were normotensive and normoalbuminuric. Pooled results showed that the annual rate of change in albuminuria was 5.53% lower (confidence interval, 0.73%–10.14%) with candesartan than placebo. The conclusion of the study was that for 4.7 years candesartan did not prevent microalbuminuria in normotensive patients with type 1 or type 2 diabetes. A limitation of the study was the fact that it was powered for retinal and not renal end points.28 In 2009, a multicenter study was published that involved 285 normotensive patients with type 1 diabetes and normoalbuminuria who were randomly assigned to receive losartan (100 mg) daily, enalapril (20 mg) daily, or placebo and follow-up for 5 years. The primary end point of the study showed a change in the fraction of glomerular volume occupied by mesangium in kidney biopsy specimens. The retinopathy end point was a progression on the retinopathy severity scale. The change in mesangial fractional volume per glomerulus over the 5-year period did not differ significantly between the placebo group and the enalapril or losartan groups. The 5-year cumulative incidence of microalbuminuria was 6% for the placebo group; the incidence was higher in the losartan group (17%) but not with enalapril (4%). When compared with placebo, the odds of retinopathy progression by 2 or more steps was reduced by 65% with enalapril and by 70% with losartan, independent of changes in BP. The conclusion of the study showed that early blockade of the RAAS in patients with type 1 diabetes did not slow nephropathy progression but slowed the progression of retinopathy.29


All current guidelines recommend a BP goal in patients with diabetes <130/80 mm Hg to reduce cardiovascular events and the progression of diabetic nephropathy. This recommendation is based mainly in the Hypertension Optimal Treatment (HOT) and ADVANCE trials.30 Lamentably, the level of evidence used for these guidelines is level C. The ACCORD and ONTARGET trials did not find any benefit on improved cardiovascular outcome with BP <130/80 mm Hg, but did find a benefit determined by reduction of stroke. In the INVEST trial, SBP <130 mm Hg was also not associated with improved cardiovascular outcomes when compared with SBP <139 mm Hg. Post hoc analysis of these trials demonstrated that the benefit of lowering BP levels on cardiovascular risk reduction is lost when SBP levels fall below 130 mm Hg. There is an increase in cardiovascular events at SBP <120 mm Hg, which is called the J-curve effect. The J-curve effect evident in the INVEST and ONTARGET studies showed an increased risk when lowering BP <130 mm Hg in patients older than 50 years who have long-standing HTN and coronary disease.31 Contemporary evidence suggests that a BP target around 130/80 mm Hg in patients with diabetes is reasonable and possible to achieve in clinical practice. Indeed, the ACCORD study demonstrated that targeting this BP goal resulted in fewer strokes, a devastating complication that is more common in patients with diabetes.17 However, physicians should be careful in older diabetic patients with a history of coronary artery disease and/or long-standing diabetes. In this group, lowering SBP to a level around 120 mm Hg may cause increased mortality. Thus, as for the level of glycemic control, the BP goals should be individualized in patients with diabetes.


There is agreement and a good level of evidence that the first line of therapy for HTN in patients with diabetes should be an ARB or an ACE inhibitor to reduce the risk of developing microalbuminuria and proteinuria and to minimize progression of proteinuria. If the initial BP is >150/90 mm Hg or the reduction in BP is not achieved with one therapy, a second agent should be added. The ACCOMPLISH trial showed that the amlodipine-based combination was more effective than a thiazide-based combination in reducing cardiovascular events. These nondihydropyridine calcium channel blockers have the benefit of a neutral effect on glycemic control when compared with β-blockers and diuretics, which worsen insulin sensitivity.32 In this regard, the American Society of Hypertension recommends using first- and second-line antihypertensive agents that do not worsen preexisting metabolic conditions. There is no evidence to support the use of ACE inhibitors or ARBs in normotensive patients with type 1 or type 2 diabetes to prevent or delay the development of microalbuminuria or nephropathy, but this therapy has been shown to reduce the progression of retinopathy in diabetic patients.


Acknowledgments and disclosures:  The authors would like to thank Brenda Hunter for her assistance in editing this manuscript.