Risk of Diabetes Mellitus with Antihypertensive Drugs

Individuals with hypertension, whether treated or untreated, are at increased risk of developing type 2 diabetes. In treated hypertensive subjects, compared with those who received no antihypertensive therapy, the risk of development of diabetes was not significantly altered with ACE inhibitors, calcium channel blockers or thiazide diuretics. Only those treated with beta-blockers were of increased risk of developing diabetes (Gress etal., 2000).

Conventional therapy BP 146/78 mmHg

Conventional therapy BP 146/78 mmHg

36 48 60

Months of follow-up

36 48 60

Months of follow-up

* Primary composite endpoint = composite of death from cardiovascular causes, non-fatal myocardial infarction, non-fatal stroke, revascularisation and amputation.

Figure 6.5 Intensive strategies to control blood pressure, lipids and glucose. Kaplin-Maier estimates for the composite endpoint of nephropathy, retinopathy, neuropathy and death from cardiovascular causes. Reproduced from Gaede P, Vedel P, Larsen N etal. (2003). Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. New England Journal of Medicine 348: 383-93. Copyright © Massachusetts Medical Society. All rights reserved.

Epidemiological studies and clinical trials support the causal link between use of beta-blockers and type 2 diabetes (Padwal and Laupacis, 2004). Beta-blockers can adversely affect glucose homeostasis, including worsening of insulin sensitivity. The diabetic potential of beta-blockers may be partly related to weight gain (Scheen, 2004).

Thiazide and thiazide-like diuretics are also cited frequently as predisposing to diabetes (Lithell, 1991; Opie and Schall, 2004). Short-term metabolic studies raised concerns about the metabolic potential of these agents (Shapiro etal., 1961). Subsequently, epidemiological studies and clinical trials suggested a causal link between use of thiazides and type 2 diabetes (Padwal and Laupacis, 2004). However, the studies that have suggested an increased risk of new-onset diabetes with thiazides have limitations: small numbers of patients; short follow-up; suboptimal definition of new-onset diabetes; lack of adequate comparison group; highly selected patients; and failure to allow for confounders (Gress etal., 2000). In the ALLHAT, there was a tendency for chlorthalidone to increase hyperglycaemia but the effect was small and not associated with increased cardiovascular events (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002).

Calcium channel blockers are generally considered to be metabolically neutral (Brown etal., 2000; Hansson etal., 2000; ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group , 2002). In the INVEST study, verapamil-based therapy reduced new-onset diabetes compared with atenolol-based therapy but the effect was modest compared with that of inhibitors of the renin-angiotensin system (Pepine etal., 2003).

Early experience with ACE inhibitors suggested no detrimental effect on fasting blood glucose after long-term administration (Neaton etal., 1993). The Swedish Trial in Old Patients with Hypertension 2 (STOP-2) study failed to show a protective effect of ACE inhibition against type 2 diabetes mellitus (Hansson etal., 1999a). The study population was elderly (average age 76 years) and the criteria for definition of diabetes were not specified. In the CAPPP study there was significant reduction in new-onset diabetes in the captopril group compared with conventional therapy (Hansson etal., 1999b). It is not clear whether this represented a protective effect of the ACE inhibitor or an adverse effect of beta-blockade or thiazide. The scientific value of the CAPPP study is reduced by a flaw in the randomisation procedure that resulted in an imbalance of baseline blood pressure and the prevalence of diabetes (Peto, 1999): this may have influenced the incidence of new-onset diabetes. In the HOPE study there was significant reduction in new-onset diabetes with ramipril compared with standard therapy (Heart Outcomes Prevention Evaluation Study Investigators, 2000). However, while the data on new-onset diabetes were prospective, new-onset diabetes was not a primary or even a secondary endpoint, and the analysis was post hoc, so the findings should be interpreted with caution.

The same investigators performed a prospective 2 x 2 study in 5269 subjects who were overweight with impaired glucose tolerance and/or impaired fasting glucose and were followed for 3 years (DREAM Trial Investigators, 2006). Approximately half of the subjects had treated hypertension. Treatment with rosiglitazone reduced the progression to diabetes by 60%, but ramipril treatment did not significantly effect the progression to diabetes. Regression to normoglycaemia, which was a secondary endpoint, was significantly increased with ramipril.

In the ALLHAT, lisinopril-based therapy was associated with 30% reduced new-onset diabetes compared with therapy based on chlorthalidone and 17% compared with amlodipine-based therapy (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002). This study provides the best evidence for an anti-diabetaginic effect of ACE inhibitors in hypertension.

In the LIFE study, losartan-based therapy was associated with a significant 25% reduction in new-onset diabetes compared with therapy based on atenolol; 70% of the time each group received concomitant thiazide therapy (Dahlof etal., 2002; Lindholm etal., 2002b). Whether this finding is due to improved insulin resistance with the angiotensin receptor blocker or a decrease in insulin sensitivity with atenolol, or both, remains uncertain. A similar reduction in new-onset diabetes has been seen with candesartan (Lithell etal., 2003; Pfeffer etal., 2003). In the Study on Cognition and Prognosis in the Elderly (SCOPE) (Lithell etal., 2003), the reduction in new-onset diabetes with candesartan was not significant but of a similar magnitude to that with losartan in the LIFE study. In the Valsartan Antihypertensive Long-term Use Evaluation (VALUE), valsartan-based therapy was associated with a 23% reduction in new-onset diabetes compared with amlodipine (Julius etal., 2004). Although the results of VALUE do not establish that angiotensin receptor blockade per se reduces or delays the onset of type 2 diabetes, the lower incidence of new-onset diabetes in valsartan-treated subjects might well reflect an anti-diabetic effect of angiotensin receptor blockers. The Antihypertensive Treatment and Lipid Profile in a North of Sweden Efficacy Evaluation (ALPINE) (Lindholm etal., 2003) confirmed the results of larger trials demonstrating a more favourable metabolic profile and lower risk of new-onset diabetes in hypertensive patients treated with an angiotensin receptor blocker compared with thiazides. Costs are greater in the short term but favourable health benefits in the long term are possible.

Recent outcome studies have reported an increased incidence of diabetes in patients treated with beta-blockers or diuretics compared with angiotensin receptor blockers, ACE inhibitors or calcium channel blockers, especially when beta-blockers and diuretics are combined. Beta-blocker/diuretic combination therapy has been associated with around 20% increase in diabetes over trial periods of approximately 5 years in comparison with therapy based on ACE inhibitors or angiotensin receptor blockers (Mason etal., 2005).

The shortcomings of the evidence must be considered. Although the major antihypertensive classes appear to exert differential effects on diabetes mellitus incidence, the data are far from conclusive (Padwal and Laupacis, 2004). Only one had diabetes mellitus as a primary endpoint, and the evidence from randomised trials is limited by sources of bias, including treatment contamination and the bias of post hoc analysis. The data from the highest quality studies suggest that diabetes is unchanged or increased by thiazides and beta-blockers, and unchanged or decreased by ACE inhibitors, calcium channel blockers or angiotensin receptor blockers.

Both diuretics and beta-blockers may exert detrimental metabolic effects leading to increased incidence of type 2 diabetes (Padwal and Laupacis, 2004). Glucose intolerance with thiazides has been attributed to potassium depletion (Helderman etal., 1983). Hypokalaemia can impair glucose tolerance by interfering with insulin release by the pancreas. In vitro studies and studies in animals and humans have suggested a possible relationship between the renin-angiotensin system and the pathogenesis of insulin resistance (Kurtz and Pravenec, 2004). Both animal and human studies have shown improvement in insulin resistance by inhibition of angiotensin II (Hovens etal., 2005). Almost half of the studies of ACE inhibitors in hypertensive non-diabetic individuals demonstrated a slight but significant increase in insulin sensitivity as assessed by insulin-stimulated glucose disposal during englycaemic hyperinsulinaemic clamp studies, while the other half failed to reveal any significant changes (Julius etal., 1991). Several clinical trials suggest that angiotensin receptor blockers have a protective effect on glucose metabolism (Dahlof etal., 2002; Lindholm etal., 2002b; Lithell etal., 2003; Pfeffer etal., 2003; Julius etal., 2004; Kurtz and Pravenec, 2004; Scheen, 2004). However, most placebo-controlled trials failed to show an anti-diabetic effect (Kurtz and Pravenec, 2004). The effects of angiotensin receptor blockers on insulin sensitivity are neutral in most studies (Scheen, 2004).

The potential mechanisms of improvement of glucose tolerance and insulin sensitivity through inhibition of the renin-angiotensin system are complex (Scheen, 2004) (Table 6.3). These may include improved blood flow and microcirculation in skeletal muscle and thereby enhancement of insulin and glucose delivery to insulinsensitive tissue (Julius etal., 1991), facilitation of insulin signally at the cellular level (Kurtz and Pravenec, 2004) and improvement of insulin secretion by the beta cells via a direct effect in blood flow to the endocrine pancreas (Carlsson etal., 1998). Blockade of the renin-angiotensin system may promote recruitment and differentiation of adipocytes, which would counteract the ectopic deposition of lipid in other tissues (liver, muscle, pancreas) thereby improving insulin sensitivity (Sharma etal., 2002).

Table 6.3 Potential antidiabetic mechanisms mediated by inhibitors of the Renin-Angiotensin system. Modified from Hovens MMC, Tamsma JK, Beishuizen ED, Huisman MV (2005). Pharmacological strategies to reduce cardiovascular risk in type 2 diabetes mellitus. An update. Drugs 65: 433-45.

Adverse effects of angiotensin II blocked by ACE inhibitors and angiotensin receptor blockers:

• insulin signalling

• tissue blood flow

• oxidative stress

• sympathetic activity

• adipogenesis

Beyond effects on the renin-angiotensin system:

• enhanced glucose metabolism by activation of bradykinin/nitric oxide pathways (ACE inhibitors)

• improved glucose and lipid metabolism by activation of PPAR — ra aPPAR — r = peroxisome proliferator-activated receptor gamma.

The availability of drugs that have anti-diabetic as well as antihypertensive properties should be of considerable clinical value. In ALLHAT, less new-onset diabetes did not translate into fewer cardiovascular events in the lisinopril group (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002). However, the complications of diabetes were reduced significantly in the ramipril group compared with conventional therapy in HOPE (Heart Outcomes Prevention Evaluation (HOPE) Study Investigators, 2000). In long-term observational studies, new-onset diabetes has the same high cardiovascular risk as that in patients with diabetes at the outset, but some years of follow-up are needed before the prognostic curves separate (Dunder etal., 2003; Eberly etal., 2003; Bartnik etal., 2004; Verdecchia etal, 2004; Kostis etal., 2005). In treated hypertensives, the risk of subsequent cardiovascular events was not dissimilar from that in previously known diabetes (Verdecchia etal., 2004); risk was almost three times that in those who did not develop diabetes, although only 63 events were recorded in this study. The average time to an event was 6 years. Since typical outcome trials in hypertension have an average follow-up of 5 years, the average duration of new-onset diabetes is only 2.5 years, leading to underestimation of the implications for cardiovascular risk.

Patients with increased fasting glucose are more likely to develop diabetes when exposed to drugs that worsen glucose tolerance (Von Eckardstein etal., 2000). The British Hypertension Society (Williams et al., 2004) recommends caution when using the combination of beta-blockers and thiazide diuretics in patients at high risk of developing diabetes: impaired glucose tolerance, strong family history of diabetes mellitus, obesity, metabolic syndrome and those of South Asian or Afro-Caribbean descent.

Delicious Diabetic Recipes

Delicious Diabetic Recipes

This brilliant guide will teach you how to cook all those delicious recipes for people who have diabetes.

Get My Free Ebook

Post a comment