Essential hypertension and diabetes mellitus (DM) are both extremely common conditions and therefore it is not surprising that their co-existence is extremely prevalent. Since both are considered risk factors for coronary artery disease, cerebrovascular disease, renal failure and congestive heart failure, treatment of both conditions is essential. Whether blood pressure (BP) should be lowered to a different target in diabetic patients has been a debate for many years. Current guidelines are inconsistent regarding BP target in diabetic patients. Whereas several guidelines recommend a BP goal of <140/90 mmHg [
1,
2], some recommend a lower target for diastolic BP [
3,
4] and some recommend lower systolic BP thresholds in certain diabetic population [
5‐
7]. The class of drug treatment most appropriate for the treatment of hypertensive diabetics is also unclear and different guidelines emphasize use of different drug classes for the treatment of hypertension in diabetic patients. Recently introduced drug classes for the treatment of DM have also been found to lower BP, thus making the interaction between BP and DM even more complex. In this review, we discuss the epidemiology of diabetes and hypertension, the benefit of lowering BP in diabetic patients, the target BP and the recommended treatment to achieve the target in these patients. This review deals mainly with BP control in type 2 DM, but some of the data derived from studies that included also non diabetic patients.
Epidemiology
Hypertension is twice more common in diabetics than in non-diabetics [
8], but the definition of hypertension in diabetics is generally similar to the general populations and the threshold for treatment is persistent BP values ≥140/90 mmHg. As both hypertension and DM are highly associated with obesity, it is not surprising that their co-existence is particularly common in obese individuals [
9]. Both hypertension and DM increase significantly with increasing age and their co-existence is highest in older individuals [
10]. Patients with DM more commonly present with isolated systolic hypertension and are more resistant to treatment. In the EUROASPIRE IV survey only 54% of the diabetic patients achieved BP levels of less than 140/90 mmHg [
11]. In addition, the presence of autonomic neuropathy in diabetic patients is associated with a less nocturnal BP decrease, a higher baseline heart rate and a higher BP variability than in non-diabetics [
12‐
18].
The co-existence of DM and hypertension significantly increase the risk for coronary heart disease [
19], left ventricular hypertrophy [
20], congestive heart failure [
21] and stroke [
22] compared with either condition alone. In addition, both hypertension and DM are present in all prediction models for the occurrence of stroke in patients with atrial fibrillation [
23‐
25]. Microvascular complications are also more common in patients with co-existent hypertension and DM and both retinopathy and nephropathy are more prevalent in patients with DM and hypertension [
26,
27]. Lowering BP is particularly beneficial in diabetic patients [
28,
29], however how low should BP be is controversial.
What should be the blood pressure target in diabetes mellitus?
The BP targets in diabetic hypertensive individuals are controversial. For many years it was common practice to aim for BP targets lower than 130/80 mmHg in non-proteinuric diabetic patients. This was based on evidence from several large studies, including The Hypertension Optimal Treatment (HOT) study, the United Kingdom Prospective Diabetes Study (UKPDS) 38 and the Action in Diabetes and Vascular disease Controlled Evaluation (ADVANCE) trial [
29‐
31]. However, in most studies the achieved BP was higher than 135/85 mmHg and therefore the recommendation to lower BP to less than 130/80 mmHg was not solid [
32,
33]. Moreover, several studies reported no benefit and even harm when lower BP targets were achieved. In the Ongoing Telmisartan Alone and in Combination with Ramipril Global End point Trial (ONTARGET) study, which included 9612 diabetic patients, the composite primary outcome of death from cardiovascular (CV) causes, myocardial infarction, stroke, or hospitalization for heart failure did not differ between groups despite achievement of lower BP values in the telmisartan-ramipril arm [
34]. In the Prevention Regimen for Effectively Avoiding Second Strokes (PROFESS) trial, which included 5743 diabetics, recurrence of stroke was not less in patients receiving telmisartan despite a significant decrease in BP [
35]. The PROFESS results were different from those of the Perindopril Protection Against Recurrent Stroke Study Collaborative Group (PROGRESS) trial [
36], in which treatment with the ACE inhibitor perindopril was associated with a 38% risk reduction in the occurrence of stroke, but the PROGRESS trial included only 762 diabetic patients and they were recruited much longer following the initial stroke than in the PROFESS trial. The Telmisartan Randomised Assessment Study in ACE Intolerant Subjects with CV Disease (TRANSCEND) study [
37] was another study which 35.7% of the patients were diabetics and in which more significant BP reduction with telmisartan was not associated with CV benefit.
In the International Verapamil SR/Trandolapril (INVEST DM) study there was no difference in short term outcome in diabetic patients with coronary artery disease despite achievement of significantly lower BP (<130 mmHg vs. <140 mmHg) and in fact, there was an increased in the long term all-cause mortality in the more tightly controlled group [
38]. In the Action to Control Cardiovascular Risk in Diabetes Blood Pressure (ACCORD-BP) trial, BP reduction to <120 mmHg did not reduce mortality or overall CV outcomes, but did reduce significantly only the occurrence of stroke a pre specified secondary outcome [
39]. Moreover, intensive BP lowering was associated with an increased rate of syncope and hyperkalemia, both directly related to the intensive treatment. The results of this large prospective study, in addition to data from other studies, led most of the societies to recommend less stringent BP target in diabetic patients.
However, the results of the recent systolic blood pressure intervention trial (SPRINT) raised again the discussion what should be the target BP in diabetic patients. The SPRINT randomized 9361 persons with systolic BP > 130 mmHg and increased CV risk, but without type 2 DM, to a systolic BP target <120 mmHg (intensive treatment) or a target of <140 mmHg (standard treatment). At 1 year, the mean systolic BP was 121.4 mmHg in the intensive treatment group and 136.2 mmHg in the standard-treatment group. The study was stopped early after a median follow-up of 3.26 years owing to 25% lower rate of the primary composite outcome in the intensive-treatment group than in the standard-treatment group (P < 0.001). All-cause mortality was also lower by 27% in the intensive treatment group (P = 0.003) [
40]. The main benefit was observed in elderly subjects (>75 years) who constituted 28% of the study population [
41]. Rates of serious adverse events of hypotension, syncope, electrolyte abnormalities, and acute kidney injury or failure, but not of injurious falls, were higher in the intensive-treatment group than in the standard-treatment group [
40]. This recent study supports intensive BP lowering in non-diabetic patients with increased CV risk. The most important question came from the SPRINT is related to diabetic patients who were excluded from this study. In light of the discrepancy between the ACCORD and the SPRINT can we assume that the better results in SPRINT with intensive BP lowering does not apply to diabetic patients?
One approach is to explain why the results of the SPRINT should not be applied to diabetic patients and, unlike our previous thoughts BP targets in diabetic patients should be higher than in non-diabetics. DM has a negative influence on arteriolar function and blood flow autoregulation that shifts the pressure/flow relationship. Therefore diabetic patients are more vulnerable to compromised blood flow to vital organs when BP reaches a critical low point.
The opposite approach is that the results of the SPRINT should be applied to diabetic patients, since in most previous trials the benefits of BP reduction in diabetic patients were at least as good if not better than in non-diabetic individuals [
29,
42]. To justify this approach one should look at the effect of intensive BP lowering in diabetic patients on stroke, the long-term follow up results of the ACCORD study and the differences between the ACCORD and the SPRINT.
In the ACCORD study, despite the failure to show a decrease in primary endpoints in the intensive treatment arm the rate of stroke was significantly lower in the intensive than in the usual treatment arm [
39]. It is possible that the ACCORD trial was underpowered, with a much lower event rate than anticipated and therefore the benefit of intensive BP lowering was not observed. Recently, new results from a long-term follow-up of the ACCORD patients, dubbed the ACCORDION trial, were presented at the 2015 AHA meeting [
43]. In this extended study 3957 patients were followed for an additional 54–60 months. During this time, patients who had been in the intensive BP arm in the main trial were no longer aiming for the lower BP goals, so the difference in BP between the two groups narrowed from 14.5 mmHg at the end of the main trial to 4.2 mmHg at the end of the follow-up period. Results from the follow-up period showed a 9% non-significant reduction in the primary end point of major CV events over a median follow-up of 8.8 years from randomization. During the long-term follow-up, an interaction between BP and glycemia interventions became significant (P for interaction 0.037), with evidence of benefit for intensive BP lowering in participants randomized to standard glycemia therapy (HR = 0.79, 95% CI 0.65–0.96). These long-term results of the ACCORD trial do take on enhanced importance when viewed alongside the SPRINT results.
Several differences in the design of the studies may also explain the different results. ACCORD had lower event rates than initially predicted because of a lower CV risk profile in participants. The exclusion of participants aged >80 years led to a younger group of patients in ACCORD than in SPRINT. The mean age for ACCORD was 62 years and for SPRINT was 68 years.
Participants in the BP arm of the ACCORD were also at lower risk because patients with dyslipidemia were assigned to the lipid arm and were excluded from the BP arm.
Another significant difference in the design of the SPRINT and ACCORD studies was the use of diuretics. The treatment regimen for hypertension in the ACCORD study often used hydrochlorthiazide, and the SPRINT study primarily used chlorthalidone.
In addition, the complexity of the factorial study design in ACCORD may have made it less likely that a statistically significant difference could be demonstrated. This may suggest that if diabetic patients were included in the SPRINT they would also benefit from intensive BP lowering.
When we try to explain the reason for the difference between the SPRINT and the ACCORD it should be emphasized that the results of the SPRINT are provocative. In the recent Heart Outcomes Prevention Evaluation (HOPE)–3 trial 12,705 participants at intermediate risk who did not have CV disease were randomized to receive either candesartan at a dose of 16 mg per day plus hydrochlorothiazide at a dose of 12.5 mg per day or placebo and were followed for 5.6 years. The first co-primary outcome was the composite of death from CV causes, nonfatal myocardial infarction, or nonfatal stroke; the second co-primary outcome additionally included resuscitated cardiac arrest, heart failure, and revascularization. Therapy with candesartan plus hydrochlorothiazide was not associated with a lower rate of major CV events than placebo despite a BP decrease of 6.0/3.0 mmHg in the active treatment group. The only subgroup who benefited from BP lowering was the subgroup of participants with initial systolic BP > 143.5 mmHg [
44]. A recent study that used the extended follow-up data from the US cohort of the International Verapamil [SR]/Trandolapril Study (INVEST) showed that in hypertensive patients with coronary artery disease, achieving a systolic BP of 130–140 mmHg seems to be associated with lower all-cause mortality after approximately 11.6 years of follow-up [
45]. Similarly, the Secondary Prevention of Small Subcortical Strokes (SPS3) trial) evaluated BP goals in patients with a previous lacunar stroke testing a systolic goal of 130–149 mmHg versus <130 mmHg [
46]. This trial also did not demonstrate significant reductions in ischemic stroke or intracranial hemorrhage in the more intensive treated group. Why the results of the SPRINT showed a clear benefit of lowering systolic BP to <120 mmHg whereas other studies failed to show it?
One explanation is the technique of BP measurements. In the SPRINT, BP was measures with an automated oscillometric office BP method that eliminated the need for a human to participate in the actual measurement and therefore reduces the white coat effect. Compared with a reasonably well-done standard office-based BP, the use of an automated oscillometric office BP method will yield a systolic BP that is 7–10 mmHg lower in the same patient, measured on the same day. If this is true the systolic BP of 120 mmHg in the SPRINT is equivalent to almost 130 mmHg in clinical practice. Thus, it is reasonable to suggest in high risk patients a target systolic BP of <130 rather than <120 mmHg.
To solve the discrepancy between the various studies and to find out what should be the target systolic BP several meta- analysis were recently published (Table
1).
Table 1
Meta-analyses of anti-hypertensive treatment in diabetic patients
Effect of antihypertensive treatment at different BP levels in patients with diabetes mellitus [ 47] | 2016 | British Medical Journal | 49 | 73,738 | Only diabetic, most type 2 | 3.7 | If BP was greater than 150 mmHg, treatment reduced all-cause mortality, CV mortality, myocardial infarction, stroke and end stage renal disease. If baseline systolic BP was less than 140 mmHg, further treatment increased the risk of CV mortality with a tendency towards an increased risk of all-cause mortality |
BP lowering for prevention of CV disease and death [ 49] | 2016 | The Lancet | 123 | 613,815 | NA | NA | Every 10 mmHg reduction in systolic BP significantly reduced the risk of major CV disease events, coronary heart disease, stroke and heart failure which, in the populations studied, led to a significant 13% reduction in all-cause mortality. The effect on renal failure was not significant. Proportional risk reductions (per 10 mmHg lower systolic BP) were noted in trials with higher mean baseline systolic BP and trials with lower mean baseline systolic BP. There was no clear evidence that proportional risk reductions in major CV disease differed by baseline disease history, except for diabetes and chronic kidney disease, for which smaller, but significant, risk reductions were detected |
BP targets for hypertension in people with diabetes mellitus [ 48] | 2013 | Cochrane Database systematic reviews | 5 | 7314 | 7134 | 4.5 | Reduction in incidence of stroke in intensive BP reduction compared with standard reduction, no effect on mortality, significant increase in other serious adverse events |
BP Targets in Subjects With Type 2 Diabetes Mellitus/Impaired Fasting Glucose [ 50] | 2011 | Circulation | 13 | 37,736 | All | 4.8 ± 1.3 | A systolic BP treatment goal of 130 to 135 mmHg is acceptable. However, with more aggressive goals (<130 mmHg), the risk of stroke continues to fall, but there is no benefit regarding the risk of other macrovascular or microvascular events, and the risk of serious adverse events even increased |
Effects of intensive BP reduction on myocardial infarction and stroke in diabetes [ 51] | 2011 | Journal of Hypertension | 31 | 73,913 | 159 | NA | Tighter BP control reduced the risk of stroke by 31% compared with less tight control, whereas the reduction in the risk of MI was not significant |
A meta-analysis of 49 trials including 73,738 patients (most of them diabetic) showed that at BP values greater than 140 mmHg, BP reduction was associated with a decrease in mortality and CV morbidity. On the other hand, BP reduction in patients with initial BP values <140 mmHg resulted in increased CV mortality and a tendency towards increased overall mortality [
47]. Another meta-analysis evaluated randomized controlled trails performed only in diabetic individuals and concluded that the present evidence does not support BP targets lower than the standard targets in people with elevated BP and diabetes [
48]. A recently published meta-analysis evaluated BP lowering for prevention of CV disease and death and reported that the proportional reduction in major CV disease events by BP reduction seemed to be larger in trials done in people without diabetes or chronic kidney disease [
49]. This was attributed to different methodological characteristics in studies in diabetic patients. Another meta-analysis of 13 randomized control studies including over 37,000 diabetic hypertensive patients has shown that intensive systolic BP control to less than 130 mmHg was associated with a 10% reduction in all-cause mortality, yet no effects on microvascular or macrovascular events were noted. Regarding stroke, such an intensive BP reduction has led to a 17% risk reduction, accompanied by an additional risk reduction with further lowering systolic BP to <120 mmHg, without an increased risk for adverse effects [
50]. Another meta-analysis included 31 randomized control studies with over 73,000 diabetic hypertensive patients reported a 31% reduction in relative risk of stroke, with a 13% reduction for every 5 mmHg systolic BP or 2 mmHg diastolic BP reductions. The risk of myocardial infarction was not significantly reduced with a more intensive BP control [
51].
Thus it seems that a target of systolic BP < 130 mmHg is reasonable in most diabetic patients. In elderly diabetic patients (>80 years) but otherwise healthy, a BP target of <140–150/90 mmHg is reasonable. Lower BP levels may be adequate if tolerated by the patients. BP levels should be monitored closely in the sitting and the standing position and the treatment should be tailored to prevent excessive fall in BP [
52].
Treatment goals according to current guidelines
Although previous guidelines recommended strict BP control in diabetic patients [
53,
54], this has been challenged in recent guidelines (Table
2). The British National Institute for Health and Clinical Excellence (NICE) guidelines published in 2011 [
55] recommended commencing treatment in diabetic patients with stage 1 hypertension (Clinic BP > 140/90 mmHg and ambulatory BP monitoring (ABPM) daytime average or home BP monitoring (HBPM) average BP of >135/85 mmHg). The recently published 2016 American Diabetes Association (ADA) guidelines recommended that hypertensive diabetic patients be treated if they have a diastolic BP of >80 mmHg or a systolic BP > 140 mmHg, with a target BP value of <140/90 mmHg [
6]. These guidelines state that individuals in whom stroke risk is a concern may, as part of shared decision making, have lower systolic targets such as 130 mmHg. This is especially true if lower BP can be achieved with few drugs and without side effects of therapy. The American Heart association (AHA)/American College of Cardiology (ACC) guidelines from 2014 recommend a target BP of <140/90 mmHg, but point out that lower targets may be considered [
56]. The American Society of Hypertension (ASH)/International Society of Hypertension (ISH) guidelines from 2014 suggest a BP goal of <140/90 mmHg in diabetic patients [
2]. These values are lower than those recommended by the majority of the JNC 8 panel for non-diabetic patients aged 60–79, which was <150/90 mmHg, yet similar to those recommended for non-diabetics aged 18–60 years, and similar to the values of all non-diabetic patients by the minority view of the JNC8 [
1]. The 2013 European Society of Hypertension (ESH) and European Society of Cardiology (ESC) guidelines recommend lowering systolic BP below 140 mmHg, and diastolic BP below 85 mmHg [
3]. The Canadian Hypertension Education Program (CHEP) suggests a target BP of <130/80 mmHg [
7]. The International Diabetes Federation (IDF) suggests age-adjusted BP targets (BP target values of <130/80 mmHg for diabetic patients younger than 70 years, target values of <140/90 mmHg for patients 70–80 years old, and target values of <150/90 mmHg for patients over 80 years old) [
5].
Table 2
BP goals in diabetics according to major guidelines
Year published | 2011 | 2013 | 2014 | 2014 | 2016 | 2016 | 2012 |
Blood pressure (mmHg) | Not addressed | <140/85 | <140/90 | <140/90 | <140/90 | <130/80 | <130/80 |
Special considerations | Begin treatment if BP > 140/90 mmHg | | | | Systolic BP < 130 mmHg and diastolic BP < 80 may be appropriate for certain individuals with diabetes, such as younger patients, those with albuminuria, and/or those with hypertension and one or more additional atherosclerotic CV disease risk factors, if they can be achieved without undue treatment burden. | | <140/90 mmHg in patients 70-80 years old <150/90 mmHg in patients over 80 years old |
Recommended initial treatment | ACE inhibitor plus either a diuretic or a CCB | All classes of antihypertensive agents are recommended. RAAS blockers may be preferred, especially in the presence of proteinuria or microalbuminua | ARB or ACE inhibitor. In black patients, it is acceptable to start with a CCB or a thiazide. | Thiazide-type diuretic, CCB, ACE inhibitor or ARB | ACE inhibitor, ARB | ACE inhibitor, ARB in patients with CV or kidney disease, including microalbuminuria, or with CV risk factors | In patients without albuminuria, Thiazide-type diuretic, CCB, ACE inhibitor or ARB |