Portal de Periódicos da CAPES

Hypertension, the Metabolic Syndrome, and the Risk of Developing Diabetes: Is It Time to Change the Guidelines?

2004; Wiley; Volume: 6; Issue: 8 Linguagem: Inglês

10.1111/j.1524-6175.2004.03751.x

1751-7176

Michael A. Weber,

Diet and metabolism studies

Anyone who has attended recent hypertension meetings in the United States or Europe will have come away with the distinct impression that the metabolic syndrome has become very much a focus of interest in the discipline of hypertension. Even without going to hypertension meetings, or for that matter reading the medical literature, we have all seen accounts in the lay press about how Americans and many others are becoming progressively more obese and vulnerable to the metabolic abnormalities that lead to diabetes and its adverse cardiovascular consequences. The key intermediate step in this gloomy scenario is what is usually termed the metabolic syndrome. This condition appears to be driven in many people by obesity, and one of its cardinal diagnostic criteria is an expanded waist size. Other major features include lipid abnormalities, primarily affecting triglycerides and high-density lipoprotein cholesterol, impaired glucose tolerance (elevated fasting glucose, though still below the diabetic threshold), and increased blood pressure (though not necessarily above the hypertension diagnostic threshold). It is easy to attribute these metabolic and cardiovascular abnormalities to obesity, and in turn blame the twin forces of gluttony and physical lethargy that have currently beset us. But this may not be the only story. More than a decade ago, before the term metabolic syndrome became popular, we were already talking about what was called the hypertension syndrome.1,2 We noted two features that went beyond the contemporary definition of this term. First, apart from the metabolic findings that were presumably related to insulin resistance, these patients had early evidence for increased muscle mass and changes in diastolic function of the left ventricle of the heart as well as increases in arterial stiffness and even some early renal findings. The second interesting feature was that these findings appeared to have a strong familial pattern. Indeed, the normotensive offspring of hypertensive patients, even while in childhood or young adulthood, had many of the metabolic and other findings observed in their parents. Of course, it could be argued quite validly that children are likely to have the same lifestyles as their parents and that this is predominately an environmental rather than a genetic issue. Even so, the familial trends were observed even in the absence of obesity, suggesting that factors other than lifestyle could be critical in developing certain aspects of the hypertension or metabolic syndrome.1,2 Before we become totally preoccupied with issues of obesity, we should not forget that many hypertensive patients have acceptable body weights. It is interesting that, if anything, lean hypertensive people have a poorer cardiovascular prognosis than their obese counterparts.3 The two conditions appear to be quite distinct. Hypertension in obese patients, together with its concomitant metabolic findings, appears to be driven primarily by the obesity itself; whereas the hypertension in lean individuals appears to be driven primarily by heightened reactivity of the renin-angiotensin and sympathetic systems.4 Not surprisingly, the presence or absence of obesity in hypertensive patients even appears to affect the prognostic effects of therapy. For instance, diuretics are far more effective in reducing major events in obese than in lean hypertensive patients.5 Not only is hypertension closely related to diabetes, but it has been recognized for some time that certain drugs used to treat hypertension may accelerate the appearance of new-onset diabetes. In particular, both β blockers and diuretics have been implicated in this effect.6–10 In studies where these so-called older drugs have been compared with angiotensin-converting enzyme (ACE) inhibitors,6–11 it could be argued that perhaps the lower incidence of new-onset diabetes with the ACE inhibitors reflects a beneficial effect of these agents rather than deleterious effects of the β blockers or diuretics. However, even in comparisons with calcium channel blockers6,8,9—which are metabolically neutral—the diuretics and β blockers appear somewhat more likely to produce diabetes. The much-discussed Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT)6 was helpful in putting this issue in perspective: In high risk hypertensive patients, the diuretic, chlorthalidone, was 43% more likely than the ACE inhibitor, lisinopril, to produce diabetes, but was also 18% more likely than the calcium channel blocker, amlodipine, to produce this adverse effect. Like the ACE inhibitors, the angiotensin receptor blockers (ARBs) appear to have beneficial effects on insulin and glucose metabolism. In the Losartan Intervention For Endpoint Reduction in Hypertension (LIFE) study,11 the ARB, losartan, was associated with a 25% relative risk reduction in new-onset diabetes when compared with the β blocker, atenolol. Other ARB studies, albeit in heart failure patients, also showed a reduction in new-onset diabetes during treatment with one of these agents.12 As with some of the earlier studies, however, it wasn't clear whether the ARBs were producing beneficial effects, or alternatively whether the comparator agents—in particular, β blockers—might be causing deleterious effects. The Valsartan Antihypertensive Long-term Use Evaluation (VALUE) trial,13 which was published very recently, appears to have clarified this issue. In this trial, the ARB, valsartan, was compared with the calcium channel blocker, amlodipine. There was a relative risk reduction in new-onset diabetes of 23% in the valsartan group. Since amlodipine is known to be at least neutral in terms of its effects on insulin sensitivity and was superior to a thiazide in the ALLHAT study,6 the result in VALUE may have an important implication. This is the first time we can conclude that a drug in the ARB class has positive effects on insulin action and potentially plays a meaningful role in protecting high-risk hypertensive patients from developing diabetes. It is not clear how these agents might produce their beneficial effects. They might have such actions as enhancing the microcirculation within skeletal muscle, thereby increasing glucose clearance into muscle cells; it is also possible that by blocking the actions of angiotensin II these agents may directly enhance transport of glucose across skeletal muscle cell membranes. Alternatively, the effect of these drugs in preventing or delaying the onset of diabetes could be related to actions in the pancreas. For example, they could enhance insulin release by the β cells through increasing local blood supply and may even prevent fibrotic or other degenerative changes in the pancreas that could compromise β cell function. Whatever the mechanism might be, it now seems legitimate to claim that valsartan, and possibly other drugs in its class, has a protective effect against the emergence of diabetes. Since the differing antihypertensive drug classes have such diverse effects on the development of new-onset diabetes, bearing in mind the adverse cardiovascular and renal complications of this condition, shouldn't hypertension treatment guidelines take these issues into account when making recommendations about which drugs to prefer? And if there is concern about this matter, who are the patients most at risk? Useful information on this subject was presented very recently at the annual meetings of the American Society of Hypertension by the ALLHAT authors.14 In a follow-up to their original publication, these investigators have carried out a more in-depth analysis of the differential effects of their study drugs on incident diabetes. In particular, the ALLHAT study divided its hypertensive patients into two groups: those who at baseline had normal fasting glucose levels ( 50% of patients in the diuretic group achieved this diagnosis, but even for the other drugs the incidence of new-onset diabetes was relatively high. If we make the reasonable assumption that most of the patients in this latter group—all of whom had both hypertension and impaired glucose tolerance—almost certainly had the metabolic syndrome, this provides useful guidance as far as therapeutic approaches to hypertension are concerned. But before drawing what appears to be an inevitable conclusion regarding antihypertensive drug choices in patients with metabolic syndrome, we should briefly acknowledge the claim by the ALLHAT authors—both in their original paper,6 and as well in their more recent analysis14—that the development of new-onset diabetes in their patients was not associated with an increased risk of cardiovascular events. Although this seems to be a surprising claim, we should remember that the time of follow-up in most of these new diabetic patients could only have been 1 or 2 years, and so it might not yet have been possible for them to manifest the expected adverse consequences of this diagnosis. A recently published long-term follow-up study in such patients appears to answer this question more decisively.15 A cohort of hypertensive patients, managed by primary care physicians but followed in a systematic prospective fashion by experienced investigators, was treated for up to 15 years, with a median observation time of 6 years. They noted, first of all, that those patients most likely to develop incident diabetes were those who had relatively high baseline glucose values (as in ALLHAT) and also those who got treatment with a diuretic (again, as in ALLHAT). The second key finding in this long-term study, this time in contrast to ALLHAT, was that patients with new-onset diabetes (regardless of their treatment) had event rates for cardiovascular and renal outcomes that were almost three-fold higher than in those patients who did not develop diabetes, and in fact were similar to the rates observed in patients known to be diabetic at baseline. Right now the guidelines of the Joint National Committee on the Prevention, Detection, Evaluation, and Treatment of High Blood Pressure (JNC 7)16 do not specifically deal with the choice of appropriate drugs for patients with the metabolic syndrome. Indeed, in the absence of a compelling indication—these patients do not yet have diabetes—the guidelines, by default, would appear to recommend starting treatment with diuretics in such individuals. To be fair, neither the additional ALLHAT analysis14 nor the long-term experience,15 or for that matter the VALUE trial,13 were available at the time the guidelines were written. One thing is now abundantly clear: For patients with evidence for the metabolic syndrome, particularly those with impaired glucose tolerance, an agent such as an ARB or an ACE inhibitor is to be preferred as primary treatment to an agent such as a diuretic or a β blocker. This does not mean, incidentally, that diuretics or β blockers are to be relegated to subordinate roles. Diuretics, for example, will be used in many hypertensive patients, including those with diabetes or even those with metabolic syndrome, as low-dose adjuncts to ARBs or ACE inhibitors. When all is said and done, optimal protection of all hypertensive patients, including those with diabetes or impaired glucose tolerance, demands full control of their blood pressure. Good as they are, blockers of the renin-angiotensin system often will not achieve blood pressure goals as monotherapies and will require the addition of complementary antihypertensive drugs.