Keywords

Introduction

Hypertension (traditionally diagnosed after two office blood pressures ≥140/90 mm Hg) is currently the most common chronic condition for which Americans obtain medical care [1]. The age-adjusted prevalence of hypertension in the USA had been relatively stable at about 30% from 1992 to 2016 [2], but increased to 46% on 13 NOV 17, when an American College of Cardiology/American Heart Association task force reduced the threshold for this diagnosis to ≥130/80 mm Hg [3]; this definition has been rejected by both The American Association of Family Physicians and The American College of Physicians. Diabetes mellitus (diagnosed since 1997 after two fasting blood glucose measurements are >125 mg/dL, but more recently if A1c is >6.5% [4]), especially the more common type 2, also has a prevalence that is strongly influenced by age and obesity; multiple datasets suggest there has been a near doubling of the incidence of diabetes in the last 4 decades in the US. The age- and gender-dependence of the prevalence of hypertension and diabetes in the US (derived from the National Health and Nutritional Examination Surveys, NHANES, 2013–2016 [1, 5]) are shown in Fig. 11.1 and Table 11.1.

Fig. 11.1
figure 1

Age- and gender-specific prevalence of hypertension (blood pressure ≥130/80 mm Hg, or taking antihypertensive medication) in the US, according to the National Health and Nutrition Examination Surveys 2013–2016. (Data from Ref. [1]). Traditionally, men had a higher prevalence of hypertension than women until about age 60, after which the reverse has been true. Whether this can be appropriately attributed to a survivorship effect is not clear

Full size image
Table 11.1 Estimated prevalence (and 95% confidence intervals) for diagnosed diabetes, undiagnosed diabetes, and total diabetes among adults 18 years of age or older, United States, 2013–2016
Full size table

The burden of hypertension, diabetes, and their combination, is substantial. For example, in data from NHANES 2013–2016, 42.8% of American women and 49% of American men over age 20 years had hypertension; the corresponding proportions for diabetes were 12% for women and 14% for men. These figures are likely to be underestimates, because, in NHANES 2013–2016, 36.3% of those with blood pressures ≥130/80 mm Hg claimed to be unaware of the diagnosis of hypertension [1], and undiagnosed diabetes was estimated to affect 2.8% of Americans [4].

Hypertension and diabetes are just two of the all-too-commonly clustered cardiovascular risk factors in many Americans, which also include obesity and dyslipidemia (especially elevated serum levels of triglycerides and, perhaps more importantly, low-density lipoprotein cholesterol). A summary of the estimated prevalence of these inter-related risk factors, based on recent national survey data, extrapolated from the entire US adult population [1,2,3,4] (and for their overlap in a 60-year old person, the closest age of the average American with diabetes, from the Framingham Heart Study [6]), is shown in Fig. 11.2. One of the more important features of this Figure is the 67–90% overlap of diabetes with hypertension (depending on age, body-mass index, and kidney function), which provides a very strong impetus for population-based strategies to improve or prevent clinical adverse outcomes in diabetics (see below).

Fig. 11.2
figure 2

Venn diagram representing the prevalence (and strong overlap) of hypertension (33%), diabetes (11.8%), obesity (34.6%), and elevated serum low-density lipoprotein cholesterol level (>130 mg/dL, 31.1%) in the civilian, non-institutionalized US adult population (represented by the area within the square box) in recent National Surveys that included the year 2010 [1]. (The overlap proportions are taken from either national survey data (when available), or Ref. [5])

Full size image

Across the globe, both diabetes and hypertension contribute strongly to death and disability. Worldwide, raised systolic blood pressure was identified as the largest (and most important) risk factor for the Global Burden of Disease (GBD) study in 2017, accounting for 30% of deaths attributable GBD risk factors and 18% of the global disability-adjusted life years, edging out smoking (20.8% and 15%), high fasting plasma glucose (19% and 14%), and high body-mass index (13.8% and 12.2%, respectively) [7]. This was a huge change from 1990, when childhood infectious diseases were more often fatal, and reflects the increasing incidence of cardiovascular disease (and its risk factors) across the earth. An earlier analysis estimated that 26.4% (or about 972 million) of the world’s population had hypertension in 2000, with 29.2% (or 1.56 billion) projected to have the condition by 2025. Most of the growth was expected to occur in developing nations [8]. The International Diabetes Federation estimates that 463 million people had diabetes in 2019, which will increase to 700 million by 2045, because the prevalence of type 2 diabetes is increasing in every country surveyed. Perhaps because nearly 80% of people with diabetes live in low- and middle-income countries, about half of those with diabetes have not yet been diagnosed. Reflecting its importance as a risk factor for mortality (with 4.2 million deaths in 2019), the prevalence of diabetes is greatest worldwide between the ages of 40 and 59 years [9].

Pathophysiology of Hypertension in Diabetics

Although perhaps something of an oversimplification, one of the most important factors in the co-development of hypertension and diabetes is insulin resistance. This problem can be most directly studied using insulin-clamp techniques that are most appropriate in a research setting, but surrogates have been developed, including fasting and post-prandial serum insulin levels that lend themselves to large studies, including clinical trials. The results of such studies suggest that more than 50% of Americans with primary hypertension have insulin resistance. Genetics have also been implicated, because first-degree relatives of patients with hypertension also have an increased risk of insulin resistance and dyslipidemia, even if they are normotensive. Probably more important for most Americans are environmental factors, like high-fat and high-calorie diets and sedentary lifestyles that lead to central adiposity and ectopic lipid deposition. These factors probably combine with the increased risk of insulin resistance to cause inflammatory and oxidative stress, which has many negative effects. In addition to enhancing the activity of the renin-angiotensin-aldosterone and sympathetic nervous systems, and causing sodium/water retention, many maladaptive derangements occur in blood vessels. These include an increase in vascular smooth muscle cell proliferation, arterial stiffness, and vascular tone, and endothelial dysfunction, a decreased ability to vasodilate in response to appropriate stimuli (e.g., nitric oxide) [9]. Some of these effects (especially closely linked to hyperinsulinemia) appear to be mediated by an elevation in intracellular calcium concentrations within vascular smooth muscle cells; which has, in turn, been recently linked to abnormal vitamin D levels and metabolism [10]. Some believe that better understanding of these pathophysiological links between hypertension and diabetes has implications for better treatment of either condition, as antihypertensive drug classes may have differential effects on incident diabetes, and some hypoglycemic drugs may increase blood pressure (see below).

Hypertension and Type 1 Diabetes Mellitus

Type 1 diabetes (characterized by a complete lack of insulin) currently affects only about 6–8% of Americans with diabetes, with the other 92–94% having type 2 diabetes (characterized by peripheral insulin insensitivity). Most affected patients are children or adolescents, who, because of their young ages, are at low risk for hypertension. Therefore, they are also at low risk for competing causes of death (compared to their type 2 diabetic counterparts, discussed below), so a larger proportion of type 1 diabetics develop chronic kidney disease (compared to type 2 diabetics) over their lifetimes. Most type 1 diabetic patients develop “moderately increased albuminuria” (albumin:creatinine ratio 30–300 mg/gm; formerly, “microalbuminuria”), proteinuria and subsequently renal disease before hypertension (see Chaps. 3, 6, and 7 of this book for more details). However, elevated blood pressure accelerates the disease processes in these relatively younger individuals, and greatly increases their risk of both macrovascular and microvascular manifestations of diabetes. For this reason, early, intensive antihypertensive therapy is often recommended, particularly with an inhibitor of the renin-angiotensin system (for which there are copious clinical trial data using severely increased albuminuria (albumin:creatinine ratio: >300 mg/gm, formerly, “microalbuminuria”) as the endpoint) [4]. Typically, beta-blockers are best avoided as antihypertensive therapy for type 1 diabetics, as such patients are more prone to hypoglycemia, signs and symptoms of which can be diminished and even masked by beta-blockade [2, 4]. Otherwise considerations about blood pressure management in type 1 diabetics are quite similar to those for the much more common type 2 diabetics, which have been more extensively studied (see below).

Hypertension and Type 2 Diabetes

We have far more data, and thus much stronger evidence (even including some from randomized clinical trials, see below), for the role of blood pressure as a major contributor to the risk of both cardiovascular and renal disease in type 2 diabetics. Nearly all epidemiological studies, starting with the Framingham Heart Study, have consistently identified hypertension as an independent risk factor for heart disease, stroke, cardiovascular death, and end-stage renal disease, whether the subject was initially diabetic or not. Perhaps the most specific literature on the prognostic importance of hypertension in diabetics comes from a roughly 4-year follow-up study of 1145 Framingham participants after the new diagnosis of diabetes, of whom 125 died and 204 experienced a cardiovascular event [11]. After appropriate adjustments for demographic and other clinical variables, hypertension was associated with a highly significant 72% increase in the risk of mortality, and a similarly significant 57% increase in the risk of cardiovascular event in individuals with newly-diagnosed diabetes. Hypertension carried a far greater population-attributable risk than diabetes for both death (30% vs. 7%) and cardiovascular event (25% vs. 9%) in these subjects. Some would argue that conclusions drawn from observational studies are inherently weaker than observations made about primary analyses of clinical trials. Fortunately, we have abundant data from clinical trials in both diabetic subjects treated with antihypertensive drug therapy (compared to those who did not receive such therapy) [12, 13], and hypertensive subjects who were or were not diabetic (at baseline) that consistently show significant benefits of lowering blood pressure to prevent major cardiovascular and/or renal endpoints.

Cardiovascular Outcomes in Hypertensive Diabetics vs. Hypertensive Non-Diabetics

The Blood Pressure Lowering Treatment Trialists’ Collaboration published their compiled data comparing outcomes in clinical trials of antihypertensive drug therapy in diabetics vs. non-diabetics in 2005 [14]. Although their original intent was not to directly compare risks among diabetics and non-diabetics, but instead to identify the benefits of similar blood pressure-lowering treatments in these groups, their data about fatal or non-fatal myocardial infarctions (“Coronary Heart Disease”) can be rearranged as in Fig. 11.3. Random-effects meta-analysis of these data shows that diabetics consistently have a higher risk of coronary heart disease events than non-diabetics, even when treated with similar, if not identical, antihypertensive regimens. On average, diabetics in these trials (who were generally well-treated with all appropriate other therapies at the times the trials were executed) experienced an 88% (95% confidence interval, CI: 83–99%) increased risk of fatal or non-fatal myocardial infarction (P << 0.0001), compared to non-diabetics. Similar calculations indicate that the risk of fatal or non-fatal stroke in clinical trials of antihypertensive agents is highly significantly increased by 43% (95% CI: 38–52%) in diabetics, compared to non-diabetics. Similarly, cardiovascular death was significantly increased by 90% (95% CI: 84–101%) for diabetics compared to non-diabetics. These estimates are in substantial agreement with many other datasets, including those from large epidemiological studies, indicating that diabetes roughly doubles long-term cardiovascular risk, in both hypertensive and non-hypertensive individuals.

Fig. 11.3
figure 3

Meta-analysis of coronary heart disease (fatal or non-fatal myocardial infarction) in randomized clinical trials comparing antihypertensive drug regimens in hypertensive diabetics vs. hypertensive non-diabetics. (Data from Ref. [14]). The summary odds ratio for coronary heart disease across 186,620 subjects was 1.88 (95% confidence interval: 1.83–1.99) for diabetic compared to non-diabetic hypertensives

Full size image

Renal Outcomes in Hypertensive Diabetics

Although there are fewer data from randomized clinical trials for renal vs. cardiovascular endpoints, many lines of evidence strongly implicate hypertension as a major risk factor for end-stage renal disease and progressive renal disease in diabetics. Perhaps most tragic are the data collected for each patient who starts renal replacement therapy in the USA on the “Intake Form,” which are summarized annually by the United States Renal Data Systems [15]. According to the report for the year 2019, 58,372 of the 124,369 (or 46.9%) of individuals who were diagnosed with end-stage renal disease in 2017 had diabetes as the primary reason for their fate; a further 28.8% had hypertension as the primary cause of kidney failure. However, these data are likely biased, because the Intake Form allows only a single answer to a typically complex question, and the options are given alphabetically (putting “Diabetes” ahead of “Hypertension” in the list). The Intake Form was modified once, in 2001, to allow identification of more than one condition that resulted in renal replacement therapy. In that year, 15% of those reaching end-stage renal disease had diabetes alone as the cause, 33% had hypertension alone, and 39% had both hypertension and diabetes checked on the Intake Form. These data, which have not been replicated, suggest that, even (or perhaps especially) among diabetics, hypertension is a major contributor to the risk of end-stage renal disease.

In addition to these population-based epidemiological data about the risk of end-stage renal disease being significantly higher in hypertensive diabetics, many longitudinal databases also show a highly significant increase in the risk for several renal endpoints in hypertensive (compared to normotensive) diabetics. Interestingly, the Framingham Heart Study has not contributed extensively to this literature, primarily because they originally enrolled only 5209 subjects in their study, and it is far more likely that these individuals died of cardiovascular causes before they developed end-stage renal disease. However, large databases from the Multiple Risk Factor Intervention Trial [16], the Department of Veterans Affairs Medical Centers [17], and the Kaiser Permanente of Northern California Health Plan [18] have consistently shown that hypertension is a major, significant contributor to both chronic kidney disease and end-stage renal disease, even (or perhaps especially) among diabetics. Similar conclusions have been reached in long-term follow-up of populations from Finland [19], China [20], and Norway [21].

Perhaps even more compelling than epidemiological data about the importance of elevated blood pressures in diabetics in preventing kidney disease are the large number of successful clinical trials, summarized in detail below, that have shown major benefits in retarding the progression of kidney disease, and sometimes even preventing or delaying the onset of end-stage renal disease.

Hypertension Treatment Strategies in Diabetics

Lifestyle Modifications

Few would argue that intensive non-pharmacological intervention, typically starting with diet and exercise, should not be highly recommended for diabetics with elevated blood pressures [4]. Recent clinical trial data supporting these interventions in hypertensive diabetics, however, are scarce, as it is probably unethical now to randomize diabetic hypertensive patients to a strategy that does not include diet and exercise. A summary of the effects of dietary modifications (typically to lower both calories and sodium) on blood pressure can be found in an excellent review [22]. Many other lifestyle modifications have a salutary effect on blood pressure, but those highlighted in an American Heart Association Scientific Statement [23] included increased physical activity (typically aerobic exercise) and device-guided breathing. The benefits of aerobic exercise in hypertensive diabetics probably derive from both weight loss (with or without a diet plan) and improved insulin sensitivity. The best data on this point come from a large epidemiological study in Finland [24], and the Finnish Diabetes Prevention Study [25], which showed that overweight subjects with impaired glucose tolerance experienced a 58% reduction in the risk of diabetes over an average of 3.2 years after being randomized to individualized counseling about reducing weight, total and saturated fat, and increasing dietary fiber and physical activity; the benefits were directly related to successful achievement of these goals.

A beneficial lifestyle modification that should not really require much discussion is tobacco avoidance [1, 4]. Cigarettes and other forms of tobacco use increase the risk of atherosclerotic cardiovascular disease, independently of blood pressure and diabetes. Although the “evidence-base” for tobacco avoidance in diabetics, hypertensives, or the combination is lacking (primarily because it would be unethical to recommend that smokers with these problems continue using tobacco), all current guidelines recommend cessation of tobacco use, which has been shown in long-term epidemiological studies to significantly decrease the risk of most of the chronic complications of hypertension and diabetes (including cardiovascular death, myocardial infarction, stroke, and amputations).

Effects of Antihypertensive Drugs on Incident Diabetes

Although lowering elevated blood pressure is highly beneficial in preventing both cardiovascular and renal endpoints in individuals with hypertension, different classes of antihypertensive agents have disparate effects on glucose tolerance (and incident diabetes). It has been known since the late 1950s that thiazide diuretics may increase insulin requirements in diabetics, or increase the risk of incident diabetes in those who are not yet diabetic. Data on this point are confounded by the fact that hypertension itself increases these risks, presumably due to both the higher risk of overweight/obesity, and increased insulin resistance. Some beta-blockers have been noted to increase both these risks, perhaps by limiting exercise tolerance and decreasing peripheral arterial flow (and glucose uptake by large skeletal muscles). On the other hand, both angiotensin converting-enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) have been shown in randomized clinical trials to improve insulin sensitivity, and reduce the risk of incident diabetes. Network meta-analyses have been used to compare the risks of incident diabetes across all antihypertensive drug classes (including placebo/no treatment) in long-term randomized clinical trials in hypertensive individuals [26]. The most recent of these is summarized in Fig. 11.4 [27].

Fig. 11.4
figure 4

Results of network meta-analysis of incident diabetes in 34 clinical trials involving antihypertensive drugs (and placebo/no treatment). The numbers in parentheses after each class of drug are the frequency of use in clinical trials; the numbers separated by the slash below the drug class correspond to the number of incident diabetics/number of subjects at risk. (Data from Ref. [27])

Full size image

The clinical implications of these data are controversial [26, 28]. The Multiple Risk Factor Intervention Trial [29], the population-based Finnish Monitoring of trends and determinants in cardiovascular disease experience [30], the Antihypertensive and Lipid-Lowering treatment to prevent Heart Attack Trial (ALLHAT) [31], and the long-term follow-up of the Systolic Hypertension in the Elderly Trial [32] all showed no significant increase in cardiovascular risk among individuals who were not diabetic at baseline, but who developed it during afterward, compared to those who maintained euglycemia during follow-up. This conclusion can be easily faulted, however, because the duration of follow-up, particularly in clinical trials, was relatively short (e.g., in ALLHAT, the protocol called for initial testing for incident diabetes at 2 years of follow-up, and therefore limited the time for development of subsequent cardiovascular disease to 2.9 years, on average). A population-based observational study from Italy suggested (based on 11 outcome events in the new diabetic population) that new-onset diabetes was associated with a significantly higher risk of cardiovascular events (compared to non-diabetics), which did not differ from cardiovascular event rates over 13 years in those who were diabetic at baseline [33]. A population-based study from Gothenberg, Sweden concluded that it took 9 years for the cardiovascular risk of new-onset diabetes to achieve statistical significance [34]. The Framingham Heart Study also concluded that the cardiovascular risk associated with incident diabetes was time-dependent, and became significant after more than a decade for coronary heart disease, but only ~7 years for coronary heart disease death [35]. This experience was similar to that seen in the Valsartan Long-term Use Evaluation (VALUE) trial, in which the 1298 patients who developed diabetes during follow-up had a cardiac morbidity that was intermediate (hazard ratio 1.43, 95% confidence interval: 1.16–1.77) between those who were diabetic at randomization (hazard ratio 2.20, 95% confidence interval: 1.95–2.49), compared to the referent group who remained euglycemic throughout [36].

There is little doubt that, in large populations, the increased risk of incident diabetes associated with diuretics or beta-blockers (even though statistically significant), is vastly outweighed by the overwhelmingly beneficial effects of blood pressure lowering. Even if the cardiovascular and/or renal risk of incident diabetes does not increase significantly for a decade, the short-term incremental costs involved in routine medical care for diabetics will be substantial: monitoring blood glucose (and A1c twice yearly), intensifying lipid-lowering drug therapy, monitoring renal function (albuminuria and serum creatinine), and ophthalmological and podiatric screening [4]. These types of considerations have led to the common recommendation to begin antihypertensive drug therapy for most diabetics with either an ACE-inhibitor or an ARB, as they are least likely to increase plasma glucose levels or insulin requirements, and they delay the progression of albuminuria (see below) [4].

Pharmacological Treatment of Hypertension in Diabetes

Overview

Essentially all authorities agree that controlling blood pressure is beneficial for diabetics [4], but controversy exists regarding which class of antihypertensive agent should be preferred as first-line drug therapy, and what the target blood pressure should be. The American Diabetes Association still recommends either an ACE-inhibitor or an angiotensin receptor blocker for all diabetics with urinary albumin:creatinine ratio >30 mg/gm [4]. In 2005, the Blood Pressure Lowering Treatment Trialists’ Collaboration found “comparable” differences across four initial types of antihypertensive drugs (ACE-inhibitor, ARB, calcium antagonist, diuretic/beta-blocker) vs. placebo for preventing total major cardiovascular events in diabetics, and “limited evidence” that lower blood pressure goals produced larger reductions in total major cardiovascular events in diabetics [12]. This last conclusion was similar to the non-significant trend seen in the Action to Control Cardiovascular Risk in Diabetes (ACCORD)-Blood Pressure trial, regarding the primary outcome comparing systolic BPs of <120 and <140 mm Hg [37]. Some have suggested that the argument about a “preferred” initial antihypertensive drug therapy in diabetics is (or should be) moot, as nearly all diabetics (in clinical trials, as well as in general clinical practice) have required two or more blood pressure-lowering agents to achieve even the currently recommended blood pressure target of <140/90 mm Hg [4].

The numbers of diabetics with major cardiovascular events (composite of cardiovascular death, stroke, or myocardial infarction) observed (or estimated) in 28 randomized clinical trials of antihypertensive drugs involving 78,754 subjects, are summarized in Table 11.2. The results from a network meta-analysis of these data are shown in Fig. 11.5 [38]. These data support (and extend) the 2005 analyses of the Blood Pressure Lowering Treatment Trialists’ Collaboration [14], as well as those of the 2015 meta-analysis [12], both of which suggested that there were few important (or statistically significant) outcome differences across randomized initial therapies for hypertensive diabetics. These data are confounded by the fact that most placebo-treated patients in these studies received other antihypertensive agents, in addition to the randomized drug, which dilutes the protective effect of the active agent. Most authorities now hold that most differences in such analyses are more likely due to issues related to statistical power, study design, and other technical factors, rather than a clear superiority of one drug class over another for all diabetic hypertensive subjects.

Table 11.2 Major cardiovascular events observed (or estimateda) in outcome-based clinical trials of antihypertensive drugs in diabetics
Full size table
Fig. 11.5
figure 5

Results of a network meta-analysis comparing the risk of major cardiovascular events (cardiovascular death, myocardial infarction, or stroke) in 78,754 diabetic subjects across all randomized drug classes (placebo, diuretic, beta-blocker, calcium antagonist, ACE-inhibitor, angiotensin II receptor blocker) in 28 clinical trials of antihypertensive drugs (excluding the combination arms of ONTARGET, adapted from Ref. [38]). Numbers in parentheses are the number of trials using this randomized drug class. Horizontal bars indicate the 95% confidence limits; the boxes represent the odds ratios (drawn with area proportional to available statistical information). CI confidence interval, ACE-inhibitor angiotensin converting-enzyme inhibitor, ARB angiotensin receptor blocker, CCB calcium channel blocker, CV cardiovascular. Numbers separated by the slash below the drug class represent the numbers of diabetics with major cardiovascular events/numbers randomized across all trials

Full size image

ARBs

Angiotensin II receptor blockers offer many advantages for the treatment of hypertension in type 2 diabetics. They are generally effective in lowering blood pressure (particularly when combined with a diuretic or calcium antagonist), are well tolerated (even better than placebo in several comparative trials in non-diabetics), reduce both the incidence and severity of proteinuria or albuminuria, prevent major cardiovascular and renal events, and are contraindicated only in patients immediately before or during pregnancy, with known renovascular hypertension, or prior allergy to the specific agent. All ARBs (except azilsartan) are now generically available, and at least 3 (losartan, valsartan and irbesartan) are on most $9/month drug lists.

Probably the best clinical trial evidence for an angiotensin receptor blocker to prevent major cardiovascular events comes from the type 2 diabetic subgroup enrolled in the Losartan Intervention For Endpoint (LIFE) reduction trial [39]. Critics will argue that this study enrolled only patients with very strict criteria for left ventricular hypertrophy, and therefore its results may not and should not be generalizable to other patients without such abnormalities. Two years after its publication, the first author of this very report pointed out that atenolol, the initial comparator agent in LIFE, is a suboptimal once-daily antihypertensive agent [40]. Despite these objections, however, the 1195 subjects in LIFE all had hypertension, type 2 diabetes, and electrocardiographic evidence for left ventricular hypertrophy, and were randomized to initial antihypertensive therapy with either losartan or atenolol, followed by hydrochlorothiazide, and other antihypertensive drugs, as needed. Blood pressures fell from an average of 177/96 mm Hg at randomization to 146/79 mm Hg in the losartan group, compared to 148/79 mm Hg in the atenolol group. The primary composite endpoint was the first occurrence of cardiovascular death, stroke, or myocardial infarction, and was significantly reduced in the group randomized to losartan (relative risk: 0.76, 95% confidence interval: 0.58–0.98, P = 0.031), even after statistical adjustment for both baseline Framingham risk score and the degree of left ventricular hypertrophy. This unusual, post-hoc, step was pre-specified in the LIFE data analysis protocol, to reduce the probability of a Type II statistical error, which was most likely to have arisen from an unbalanced randomization process. Both all-cause and cardiovascular mortality were also significantly reduced in the losartan group (by 39% and 37%, respectively). These data were consistent with a suggestion, popular from 1995 to 2005, that how blood pressure was lowered might be an important determinant of outcomes [41]; today, most authorities agree that lowering blood pressure is more important than which agent is selected to start the process [42, 43].

Two classic, placebo-controlled, multicenter, prospective, randomized clinical trials have made type 2 diabetic nephropathy a “compelling indication” for an ARB, resulting in two FDA-approvals for this condition. Many are not aware that in both the Irbesartan Diabetic Nephropathy Trial (IDNT) [44] and the Reduction of Endpoints in Non-Insulin Dependent Diabetes Mellitus with the Angiotensin II Antagonist Losartan (RENAAL) Trial [45], potentially-eligible type 2 diabetics had their blood pressures treated with diuretics, beta-blockers, and/or other antihypertensive drugs, before randomization to an ARB, placebo (or amlodipine in IDNT). The entry criteria for the two studies were only slightly different: IDNT required type 2 diabetics between 30 and 70 years of age, a blood pressure >135/85 mm Hg, >900 mg/day of proteinuria, and a serum creatinine between 1 and 3 mg/dL in women, or 1.2–3.0 mg/dL in men. For RENAAL, type 2 diabetics had to be 30–70 years old, with urinary albumin/creatinine ratios of >300 mg/gm, and serum creatinine levels between 1.3 and 3.0 mg/dL. The results, published back-to-back in the New England Journal of Medicine, were astonishingly similar. Blood pressure was reduced in IDNT from 159/87 mm Hg at randomization, to 140/77 mm in the group randomized to irbesartan, 141/77 mm Hg in the group randomized to amlodipine, and 144/80 mm Hg in the group randomized to placebo. In RENAAL, blood pressures were reduced from 152/82 mm Hg at randomization, to 140/74 mm Hg in the losartan group, and 142/74 mm Hg in the placebo group (at the end of the study). Not only was the primary composite endpoint for both trials identical (first occurrence of doubling of serum creatinine, end-stage renal disease, or death), but also the final P-value comparing the incidence of the primary endpoint across the ARB and placebo was exactly 0.02 for each trial! The results of traditional meta-analyses summarizing these two landmark trials are shown in Fig. 11.6. It is probably not surprising that there was no overall effect on mortality, as the average age of the diabetic subjects was nearly 60 years, and nearly two-thirds had retinopathy at randomization. However, the overall highly significant delay of the primary endpoint, and the significant reduction in the number of people requiring renal replacement therapy were impressive. Pharmacoeconomic analyses of these and similar outcomes studies indicate that ARBs (even at pre-generic prices!) are cost-saving within 2 years of institution of therapy in patients with either early or late diabetic nephropathy [46].

Fig. 11.6
figure 6

Results of traditional Mantel-Haenzsel meta-analyses of comparisons of an angiotensin receptor blocker (irbesartan or losartan) vs. placebo in two landmark trials of type 2 diabetic nephropathy. (Data from Refs. [44, 45]). Horizontal bars correspond to the 95% confidence limits for each comparison; solid boxes are drawn in proportion to the number of subjects experiencing each endpoint (compared to the referent primary composite endpoint). ARB angiotensin receptor blocker, 95% CI = 95% confidence interval, IDNT Irbesartan Diabetic Nephropathy Trial, RENAAL Reduction of Endpoints in Non-Insulin Dependent Diabetes Mellitus with the Angiotensin II Antagonist Losartan. Note that these analyses ignore otherwise valid data from the amlodipine arm of IDNT

Full size image

Although not recognized by the US FDA as a valid surrogate endpoint, albuminuria and/or proteinuria have been extensively studied in both type 1 and type 2 diabetics. In nearly all trials, ARBs are quite effective in both reducing albuminuria in the short-term, and preventing its development in the longer term (typically 2–4 years). Perhaps the most famous trial of this type was the Irbesartan Microalbuminuria trial [47], published simultaneously with IDNT and RENAAL. In this prospective trial, 610 type 2 hypertensive diabetics with “microalbuminuria” (then defined as 20–200 μg/min) were randomized to placebo or low or high dose irbesartan for 2 years, and followed for the development of frank proteinuria (≥ 288 mg/d, and an increase from baseline of ≥15%). Blood pressures were barely different across the groups (145/84 mm Hg with placebo, 143/84 mm Hg with 150 mg/d, and 142/84 mm Hg with 300 mg/d of irbesartan), and there were fewer adverse effects and discontinuations in the drug-treated groups. After 2 years, only the group receiving irbesartan at 300 mg/d showed a significant (70%) reduction in the incidence of proteinuria; the lower dose had only a non-significant trend at 39%. A subsequent meta-analysis suggested that renin-angiotensin-aldosterone system inhibitors significantly reduced albuminuria in both type 1 and type 2 diabetics, with a larger effect on those with higher levels of baseline albuminuria [48].

Several of the many objections to albuminuria as a valid or useful surrogate endpoint in kidney disease in diabetics can be documented by conclusions of important studies. Especially in type 1 diabetes, the degree of albuminuria varies considerably, depending on recent blood pressure control, volume status, dietary sodium intake, and other factors. One prospective study of 75 normotensive type 1 diabetics without albuminuria at baseline suggested that patients with an increase in blood pressure during sleep predicted the development of microalbuminuria [49]. This issue can presumably be overcome by requiring two successive determinations above threshold (e.g., as in the Irbesartan Microalbuminuria trial [47]). Secondly, a 3.2-year clinical trial that enrolled 4447 type 2 diabetics without albuminuria, comparing 40 mg of olmesartan vs. placebo showed a slightly lower office blood pressure (by 3.1/1.9 mm Hg), a slowing of the rate of onset of microalbuminuria (by 23%, 95% CI: 6%–37%, P = 0.01), no difference in nonfatal cardiovascular events (P = 0.37), but an increase in cardiovascular death (15 vs. 3, P = 0.01) in the olmesartan group [50]. Although the excess death rate has been attributed to chance, an imbalance in the numbers of patients with known coronary heart disease, and other factors, many would argue that it takes far longer than 3.2 years for the disease process in hypertensive diabetics to progress from microalbuminuria to clinical cardiovascular events, suggesting that this study was underpowered to detect a significant difference in the “hard endpoints” of stroke, myocardial infarction, or cardiovascular death. The role of albuminuria as an outcomes effect modifier in chronic kidney disease is likely to remain controversial for some years to come [51].

ACE-Inhibitors

ACE-inhibitors share many of the advantages of ARBs for the treatment of diabetics with hypertension, but carry the added risk of chronic, nonproductive cough (~13%) and angioedema (0.7%). Perhaps because they were the first available agents that directly inhibited the renin-angiotensin system, they have been well tested in clinical trials that included diabetics. Perhaps the most illustrative is the Captopril Collaborative Study Group’s comparison of captopril vs. placebo in type 1 diabetics with nephropathy [52]. This trial enrolled 409 type 1 diabetics with urinary protein excretion >500 mg/d and serum creatinine <2.5 mg/dL, and used doubling of serum creatinine as the primary endpoint. After a 3-year median follow-up period, blood pressure differences between the groups were <2/4 mm Hg; significantly fewer patients in the captopril-treated group experienced doubling of serum creatinine (25 vs. 43, P = 0.007), or the secondary composite (but clinically important) endpoint of death, dialysis or transplantation (23 vs. 42, P = 0.006).

This landmark study made it difficult to justify doing similar placebo-controlled renal outcome trials in type 2 diabetics, as it was widely assumed that similar benefits should accrue. One head-to-head comparison of an ARB with an ACE-inhibitor has been done in type 2 diabetics, but it used the surrogate endpoint of decline in glomerular filtration rate (measured by iohexol clearance) as its primary outcome measure, and was successful in establishing statistical “non-inferiority” of telmisartan with enalapril in a 5-year study of 250 type 2 diabetics [53]. Many feel that this endpoint was not as robust as those used in previous renal outcome studies, and may have been unduly influenced by lack of a final measurement in 14% of the telmisartan- and 13% of the enalapril-treated subjects. Many other trials have established ACE-inhibitors as being particularly valuable for reducing proteinuria and delaying the progression of chronic kidney disease in patients without diabetes [54].

ACE-inhibitors have also been studied extensively to prevent cardiovascular disease events in diabetics. Perhaps the most optimistic effects were seen with ramipril in the 3677 diabetics randomized in the Heart Outcomes Prevention Evaluation (HOPE) [55]. Although stopped 6 months earlier than planned, the diabetics randomized to rampril enjoyed a highly significant 25% relative risk reduction for the primary composite endpoint of cardiovascular death, myocardial infarction, or stroke, as well as significant reductions in each of its components, as well as a 24% reduction in all-cause mortality and development of >300 mg/d of proteinuria. Later trials enrolling large numbers of diabetics and non-diabetics, that compared placebo with either perindopril or trandolapril were not nearly as positive, probably because of more extensive and appropriate treatment of other risk factors in both randomized groups (including antiplatelet agents, beta-blockers in subjects with a history of myocardial infarction, and lipid-lowering agents). The overwhelmingly positive results of HOPE and its diabetic substudy might be attributed to the reluctance of the Data Safety and Monitoring Board to halt the trial, which would be far more likely today (for many reasons) than in 1999.

It is important to balance the perhaps uniquely positive results of HOPE and its diabetic substudy by contrasting it with the results of the clinical trial that enrolled the largest number of type 2 diabetics ever, the Antihypertensive and Lipid-Lowering to prevent Heart Attack Trial (ALLHAT) [56], discussed further below. In their enrolled cohort of 13,101 diabetics, lisinopril was not superior to chlorthalidone in preventing any type of cardiovascular event, and may have been significantly worse in preventing stroke in black subjects (diabetic or not).

Many small studies in diabetes, heart failure, chronic kidney disease, and other conditions suggested that combining an ACE-inhibitor and an ARB might be beneficial. This seemed especially promising for reduction of albuminuria in diabetics [57], or prevention of death or rehospitalization in patients with systolic heart failure [58]. However, when the large trial (with 25,620 randomized subjects) was undertaken combining full doses of telmisartan + ramipril, there was a slightly lower blood pressure in the group given the combination, but no improvement in cardiovascular events, significantly more hyperkalemia and renal dysfunction [59], and a significantly greater risk of the composite of doubling of serum creatinine, end-stage renal disease, or death [60]. Among the 9612 enrolled diabetics, similar (non-significant) trends were observed for these important endpoints. These data suggested that there were few differences between full doses of an ACE-inhibitor and an ARB, and that the combination might be harmful to the kidney. More recently, losartan (100 mg/d) was given to 1448 type 2 diabetics with an albumin/creatinine ratio of >300 mg/gm and a baseline estimated glomerular filtration rate (eGFR) between 30 and 89.9 mL/min/1.73 m2, to which was added either placebo or lisinopril (10–40 mg/d). Although originally intended to compare regimens with regard to a “hard renal endpoint” (a composite of the first occurrence of: decline in eGFR ≥30 mL/min/1.73 m2 if baseline eGFR was ≥60 mL/min/1.73 m2, decline in eGFR of ≥50%, end-stage renal disease, or death), the trial was terminated early (despite a non-significant 12% reduction in the primary composite endpoint) because of excess hyperkalemia (6.3 vs. 2.6 events per 1000 person-years in the combination vs. monotherapy arms) and acute kidney injury (12.2 vs. 6.7 events per 1000 person-years) [61]. These data confirmed the potential harms of combining an ACE-inhibitor + ARB in type 2 diabetics, which increased the risk of shared toxicities (e.g., hyperkalemia, acute kidney injury), with no major benefit on cardiovascular or renal outcomes.

Renin Inhibitor(s)

The newest method of interfering with the renin-angiotensin system attacks the rate limiting step: hydrolysis of angiotensinogen to angiotensin I, by directly inhibiting renin. Aliskiren, the original renin inhibitor, was launched in 2007, and seemed to have many of the advantages of an ARB: dose-dependent blood pressure reductions, excellent tolerability profile, and contraindications only for pregnancy and renal artery disease. The initial trial in hypertensive type 2 diabetics with an early morning albumin/creatinine ratio between 300 and 3499 mg/gm compared losartan 100 mg/d, with or without aliskiren force-titrated from 150 mg/d for 3 months, to 300 mg/d, for another 3 months [62]. The results were quite promising: only a little (and non-significant) lowering of blood pressure, quite similar adverse effects, and a 20% overall reduction in albumin/creatinine ratio, with aliskiren + losartan, compared to losartan alone. This led to high expectations about the “hard outcomes study” that compared adding aliskiren (300 mg/day) to either an ACE-inhibitor or an ARB in 8561 diabetics with either chronic kidney disease, cardiovascular disease, or both. Although blood pressure and albuminuria were slightly lower in the group given aliskiren, the study was stopped prematurely because of significantly higher risk of hyperkalemia, hypotension, or adverse effects requiring discontinuation of drug therapy in the aliskiren group [63]. After the announcement of the trial’s early termination, other trials of aliskren in diabetics and marketing efforts for all dose forms of the aliskiren + valsartan combination were halted, and the FDA-approved product information for aliskiren was updated to include a contraindication for combining aliskiren with either an ARB or ACE-inhibitor in diabetics, and a warning against using aliskren in patients with an estimated glomerular filtration rate <60 mL/min/1.73 m2, if the patient is already taking an ACE-inhibitor or ARB.

A post-hoc analysis of the trial comparing the combination of telmisartan + ramipril to monotherapy with either in type 2 diabetics also showed a higher risk of hypotension, hyperkalemia, and the need for acute dialysis in those receiving dual inhibitors of the renin-angiotensin system [64]; excess risk was also observed in the losartan + lisinopril-treated group of the more recent trial funded by the Department of Veterans Affairs [61]. Taken together, these data indicate that monotherapy should be more advantageous than combining two drugs that interfere at different sites of the renin-angiotensin-aldosterone cascade.

Calcium Antagonists

Both dihydropyridine and non-dihydropyridine calcium antagonists have been used to lower blood pressure in many diabetic patients, based on a number of clinical trials. Early studies of non-dihydropyridine calcium antagonists showed mild-to-moderate reductions in proteinuria, which are often additive to those of renin-angiotensin system inhibitors, whereas “naked” dihydropyridine calcium antagonist tend to increase proteinuria, and were significantly inferior to an ARB in IDNT in preventing its renal endpoints [44]. As a result, most physicians now use calcium antagonists in combination with a renin-angiotensin-system inhibitor, as was commonly the case in RENAAL [45]. Calcium antagonists have no major adverse effect on glucose or cholesterol metabolism, are reasonably well tolerated, and have plentiful outcomes data from randomized clinical trials in both diabetic and non-diabetic hypertensives.

Two trials are especially illustrative of the potential benefits of calcium antagonists in diabetics: the Anglo-Scandinavian Cardiac Outcomes Trial (ASCOT) and the Avoiding Cardiovascular events through COMbination therapy in Patients LIving with Systolic Hypertension (ACCOMPLISH). The former compared amlodipine (with perindopril, as needed) and atenolol (with bendroflumethiazide, as needed), with fatal or non-fatal myocardial infarction as the primary endpoint. Concomitantly, eligible subjects were randomized to atorvastatin or placebo, which was so successful in reducing the incidence of the primary endpoint that it was stopped early, leaving the blood pressure-lowering arm of the trial with lower-than-expected statistical power. This was thought to justify a change in the primary outcome measure to total cardiovascular events and procedures for all pre-specified subgroup analyses, including that for the 5137 diabetics [65]. Although the study protocol recommended a target of <130/80 mm Hg for diabetics, their blood pressure was reduced, at 1 year, to 143/81 and 148/84 mm Hg, in the amlodipine and atenolol groups, respectively, and to 137/76 and 136/75 mm Hg at the end of the study. During follow-up, the Kaplan-Meier curves for total cardiovascular events and procedures in diabetics were super-imposable for the first 3 years, but diverged thereafter, resulting in an overall significant advantage for the amlodipine-treated group (P = 0.0261). This difference was presumably driven by putatively significant differences (P < 0.05, uncorrected for multiple comparisons) in fatal and non-fatal stroke, chronic stable angina, nonfatal stroke, peripheral arterial disease, and other revascularization procedures, all favoring amlodipine. The original primary outcome measure was not significant (P = 0.46), although the trend favored amlodipine. Overall, these results in diabetics paralleled those seen in the entire ASCOT study cohort, and have been criticized by those who believe that secondary outcomes can be properly evaluated only if the primary outcome is significant.

The ACCOMPLISH trial enrolled 11,505 high-risk hypertensive subjects (including 6946 with diabetes), and randomized them to initial therapy with benazepril and either amlodipine or hydrochlorothiazide [66]. The primary outcome measure was a composite of cardiovascular death, myocardial infarction, stroke, hospitalization for angina, resuscitated cardiac arrest, or coronary revascularization. The protocol recommended a target blood pressure of <130/80 mm Hg for all diabetics, but the average office blood pressures were 132/73 and 133/74 mm Hg in the amlodipine- and hydrochlorothiazide-treated groups during follow-up. Despite its early termination due to superiority of amlodipine over hydrochlorothiazide, diabetics randomized to the former therapy enjoyed a significant 23% reduction in the primary endpoint (P = 0.003), with significantly lower rates of coronary events (revascularization and the composite of myocardial infarction, unstable angina pectoris, or sudden cardiac death). In addition, the post-hoc renal endpoint (increase in serum creatinine by >50% and above the reference range) was significantly reduced in incidence by 47% (95% CI: 36–55%, P < 0.001) in diabetics, and even more in non-diabetics (62%). These data have caused some guideline committees to favor a calcium antagonist over hydrochlorothiazide as second-line antihypertensive therapy for diabetics, but most ALLHAT investigators believe that chlorthalidone would have produced different results, if it had been used instead of the much shorter-acting and less potent hydrochlorothiazide.

Beta-Blockers

As discussed above, most authorities currently recommend a beta-blocker for diabetics only if there is a compelling indication (e.g., post-MI, heart failure with diminished left ventricular function), because of their propensity to mask hypoglycemic signs and symptoms, potential hyperglycemia, and reduction of exercise tolerance (which may promote weight gain). Before concerns about atenolol were raised [39, 40, 65], the United Kingdom Prospective Diabetes Study 39 randomized 1158 newly diagnosed type 2 diabetics with hypertension to twice-daily captopril or once-daily atenolol, with a second randomization (discussed below) to different target office blood pressure levels. During 9 years of follow-up, significantly more subjects abandoned atenolol than captopril, but there were no significant differences across treatment arms for any of the several pre-specified endpoints (although they all favored atenolol) [67].

Diuretics

Diuretics have long been used to lower blood pressure in diabetics; for many such patients, attainment of blood pressure goals is difficult or impossible without a diuretic. These agents decrease the intravascular volume that is common in many type 2 diabetics, prevent heart failure, and counter the hyperkalemic effects of renin-angiotensin system inhibitors. Their adverse effects sometimes include erectile dysfunction, hypokalemia, and an increased risk of worsening glycemic control.

There is nonetheless a solid base of clinical trial evidence supporting the use of diuretics for hypertensive diabetics. In the Systolic Hypertension in the Elderly Program (SHEP) trial, chlorthalidone-based therapy was significantly better than placebo in reducing major cardiovascular disease events, with the same 34% relative risk reduction, but a twofold higher absolute risk reduction [68]. A meta-analysis from the Individual Data Analysis of Antihypertensive Drug Interventions project that included the Hypertension Detection and Follow-up Program, European Working Party on Hypertension in the Elderly, Swedish Trial of Older Patients with Hypertension, and SHEP showed a significant reduction in stroke (36%) and major cardiovascular events (20%) with an initial diuretic, compared to control intervention [69]. Lastly, and perhaps most importantly, as briefly mentioned above, the ALLHAT trial enrolled more diabetics than any other trial, and concluded that the diuretic they chose, chlorthalidone, was superior to all other classes of initial antihypertensive drugs for preventing one or more forms of cardiovascular disease among all hypertensives, as well as diabetics [56]. This conclusion, based largely on the inclusion of heart failure as an independent endpoint, rather than part of a composite (as originally planned), was originally quite controversial. Since then, the controversy has shifted to how large the differences are between chlorthalidone and the much more popular hydrochlorothiazide. Using very selective criteria that included data from only 9 trials, investigators from Connecticut concluded that chlorthalidone was clearly superior to hydrochlorothiazide in preventing cardiovascular events [70]; other investigators did not find a significant difference in outcomes between the two drugs in two other network meta-analyses that included data from 5 and 83 clinical trials [71, 72], although outcomes data (particularly in preventing heart failure) are far more plentiful with chlorthalidone [73].

Other Drug Classes

Most authorities agree that an alpha-1 adrenergic antagonist was inferior to low-dose chlorthalidone in preventing heart failure and combined cardiovascular disease events in ALLHAT diabetics [74]. There are many possible explanations for this disparity, including the use of seated (rather than standing) blood pressures, but it reinforces the importance of hard endpoints in clinical decision-making. Many previous studies had shown putatively beneficial effects of alpha-1 blockers on blood pressure, glucose and lipid metabolism, which were also seen in ALLHAT, but eventually found to be less important for preventing cardiovascular disease outcomes. Centrally-acting alpha-2 agonists are sometimes needed to control blood pressure, and have few adverse metabolic effects, but sedation, dry mouth, and other common adverse effects make them less popular for routine therapy of hypertension. Aldosterone antagonists are also occasionally useful, but hyperkalemia and worsened renal impairment are common adverse effects.

Blood Pressure Treatment Targets for All Diabetics?

Controversy still exists regarding the effects of a lower-than-usual blood pressure target for all diabetics. This had been a basic tenet in the diabetes and hypertension communities for many years, but was challenged by the ACCORD trial [37], rejected by JNC 8 [75], and then indirectly validated in SPRINT [76] (which included no diabetics), and eventually reinstated by the ACC/AHA 2017 US Hypertension Guideline [3].

Evidence supporting a lower-than-usual blood pressure target for diabetics came from at least 3 clinical trials: the United Kingdom Prospective Diabetes Study, the Hypertension Optimal Treatment Study, and a small multiple-intervention trial in Denmark. Back in 1985, 1148 newly-diagnosed type 2 diabetics in the United Kingdom were randomized to a “lower blood pressure target” (≤150/85 mm Hg) or “less tight control” (≤180/100 mm Hg), and followed for 8.4 years [77]. The group randomized to the lower target achieved a mean blood pressure of 144/84 mm Hg, compared to 154/87 mm Hg for the other group, and suffered significantly fewer diabetes-related endpoints (the primary outcome measure, by 24%), deaths (32%), strokes (44%), and microvascular endpoints (37%). Formal cost-effectiveness analyses, based on then-current British healthcare costs, indicated that lowering blood pressure to the lower target saved both discounted disease-free life-years and money (£1049 per endpoint-free year of life saved) [78]. Note that the incremental blood pressure reduction between the two randomized groups seen in UKPDS (10/5 mm Hg) was exactly that recommended a year earlier (for diabetics compared to non-diabetics) by the 1997 US hypertension guidelines committee, which was also supported by a pharmacoeconomic analysis showing overall cost-savings for the lower target [79]. The second trial that showed a significant benefit of a lower-than-usual blood pressure target for diabetics randomized 1501 diabetics (among the enrolled total of 18,790 subjects) to diastolic blood pressures of ≤80, ≤85, or ≤90 mm Hg [80]. Over a median of 3.8 years of follow-up, diabetics randomized to the lowest diastolic BP had a significant, 51% lower risk of major cardiovascular events, compared to those randomized to ≤90 mm Hg. The results of this trial were therefore used to support lowering the diastolic blood pressure target for diabetics to <80 mm Hg in many national and international guidelines written between 1998 and 2012. This target was seemingly supported by a small trial of 180 type 2 diabetics in Denmark, which showed a significant 55% reduction in the risk of cardiovascular complications in those who received “intensive therapy,” which included a lower-than-usual blood pressure target [81]. Extended follow-up for another 5.5 years demonstrated a significant 45% reduction in overall mortality in the “intensive therapy” group [82].

The largest and most direct test of the lower blood pressure target for diabetics was the ACCORD trial, which enrolled 10,251 subjects, and randomized them to a systolic blood pressure of <140 or <120 mm Hg [37]. Although some argue that the <120 mm Hg is too low, the average achieved systolic blood pressure in this group was 119 mm Hg, proving that it was possible to meet such a low target. However, the overall cardiovascular event rates were not significantly different (P = 0.20), although the 12% relative risk reduction favored the lower target; only the secondary endpoint of fatal or nonfatal stroke was reduced significantly (by 41%, P = 0.01). By design, the Systolic blood Pressure INtervention Trial excluded diabetic subjects (because ACCORD-BP already answered the question of whether a lower-than-usual BP goal was beneficial in this population), but the overwhelmingly positive effects of the systolic target of <120 mm Hg were robust to many subgroup analyses [76]. Rather than simply addressing the question of whether the lower-than-usual target was appropriate for diabetics, the ACC/AHA 2017 US Hypertension Guideline simply recommended a target of <130/80 mm Hg for all subjects with a blood pressure higher than that [3].

Despite the controversy, there does seem to be some support for a lower-than-usual blood pressure treatment target for diabetics with nephropathy, based on post-hoc analyses of both IDNT and RENAAL. This makes perfect sense from the precepts of preventive medicine, as the recommended treatments are nearly always more intensive (and usually more beneficial) for high-risk, compared to low-risk, groups. This principle is supported by analyses of stroke prevention with antihypertensive drugs, previously-recommended targets for LDL-cholesterol reduction across the cardiovascular risk continuum, and former post-exposure prophylaxis for needlesticks that might transmit the human immunodeficiency virus. So several very recent guidelines have recommended a lower-than-usual blood pressure for diabetics [4] (and non-diabetics [51] if the albumin/creatinine ratio is >30 mg/gm). While one might argue that this recommendation is not completely or unanimously evidence-based, it fits with the more intensive treatment of predictors of outcomes that is common in other disease states.

To summarize, the optimal treatment of hypertension in diabetics is still controversial, but probably includes lifestyle modifications whenever feasible, one inhibitor (but not two inhibitors) of the renin-angiotensin system, and sufficient other antihypertensive medications to keep the blood pressure at a level inversely proportional to the absolute risk of cardiovascular and renal disease in the individual patient, based on assessment of other all risk factors, including albuminuria [83].