Does Treating Hypertension Prevent Dementia?

One of the concerns of the increasing life expectancy that we are mostly enjoying is the increased chance of dementia, which increases with age and is one of the disorders that causes the most suffering not only for the patient but also for his or her family. There is currently a lot of interest in the possibility that dementia may be one of the consequences of hypertension and also that antihypertensive treatment in elderly persons may slow the progression of dementia as well as prevent strokes. A big question here is whether dementia and strokes should be considered separately for clinical decisions relating to hypertension. The widely accepted evidence that strokes are prevented by lowering blood pressure in elderly hypertensive patients means that there is really no issue in the decision to treat them. If dementia that is not obviously stroke-related is also prevented, so much the better, but it does not necessarily affect the therapeutic decision. An important practical issue is whether the prevention of dementia with antihypertensive drug treatment should be any different from the prevention of strokes. In the majority of patients, the decision to start antihypertensive medication is based on the blood pressure level and also on the presence of other risk factors, but the treatment is not normally aimed at the prevention of any specific disease. It could be, however: if an elderly patient with borderline hypertension is starting to show some mental impairment, would this affect the decision to start antihypertensive medication? At least two factors related to the prevention of dementia might influence the decision about treatment: the first is the threshold blood pressure level at which treatment is started, and the second is which type of antihypertensive therapy should be used. Both of these are discussed below. Reviewing the published literature to help answer these two clinical questions is difficult, because one of the issues concerned in the analysis of the studies is that measuring changes of dementia with time or treatment does not necessarily clarify whether the dementia is vascular dementia that results from hypertension or that related to Alzheimer’s disease, which has different brain pathology and can develop independent of blood pressure levels. For many years,1 dementia has officially been classified as two distinct disorders: vascular dementia, which is caused in part by hypertension, and that related to Alzheimer’s disease, which results from neurodegeneration characterized by specific “miliary bodies (plaques) and dense bundles of fibrils (tangles).”2 The problem is that Alzheimer’s disease is of course not routinely diagnosed by the examination of brain pathology, but by indirect tests that may overlap with the tests used to describe the neural damage caused by hypertension. Most of the diagnostic tests assess patients’ intellectual activity by interviewing and history taking, and some type of brain scanning is usually done to exclude other causes of dementia such as brain tumors, strokes, or other major changes. Another issue is that the conventional diagnosis of Alzheimer’s disease is not often made until a relatively advanced state, by which time the brain damage may be irreversible. There is currently interest in a milder form, often referred to as mild cognitive impairment (MCI), that is measured by neuropsychological testing.3 Persons with an abnormal score on this test show an increased risk of developing the full Alzheimer’s picture at a rate of 10% per year; hypertension increases the risk of MCI developing.3 There is general agreement that rates of hypertension, dementia, and Alzheimer’s disease increase with age.1 The prevalence of dementia rises rapidly with age and has been reported to be as high as 20% at the age of 80 and 40% at 90.4 Cross-sectional studies may fail to show what the causal relationships are between these variables, but longitudinal studies have shown connections. Of interest was the Maine-Syracuse Longitudinal Study of Hypertension,5 which examined the relationship between blood pressure and the Wechsler Adult Intelligence Scale over 20 years. As expected, higher blood pressure measured at baseline predicted grreater decline in intelligence in older patients (47–83 years), but more interestingly the same thing was seen in younger patients (18–46 years), in whom Alzheimer’s is not thought to be a problem. In the Framingham study, data that were collected between 1956 and 1964, before many patients were receiving antihypertensive drugs, showed that persons with higher blood pressure levels tended to show more deterioration of mental function over time than normotensive persons.6 In a population study of the elderly in Sweden,7 it was found that participants in whom dementia developed at age 79 to 85 had higher blood pressure at age 70. It has also been suggested that increased arterial stiffness, which leads to increased pulse pressure and pulse wave velocity, also accelerates dementia. One example is the Baltimore Longitudinal Study of Aging,8 which performed neuropsychological tests on aging persons (starting age, 57 years) over 11 years. Individuals with higher pulse pressure and increased pulse wave velocity showed accelerated declines in mental activity. Another is an Italian study of elderly patients whose change in mental function was evaluated over 1 year.9 It was found that a high pulse wave velocity was the best predictor of mental function decline. Despite the large number of studies investigating the effects of different types of anithypertensive drug treatment on cardiovascular events, very few data have been gathered about the effects on the progression of dementia. Only 6 randomized placebo-controlled studies have gathered any useful data. Five of these included elderly patients who had not already had a stroke. These are the Systolic Hypertension in Europe (Syst-Eur) study;10,11 the Medical Research Council’s Trial of Hypertension Treatment in the Elderly (MRC);12 the Study on Prognosis in the Elderly (SCOPE);13 the Systolic Hypertension in the Elderly Program (SHEP);14 and the Hypertension in the Very Elderly Trial Cognitive Function Assessment (HYVET-COG).15 One study—the Perindopril Protection Against Recurrent Stroke Study (PROGRESS)—enrolled patients who had already had a stroke.16 A recent Cochrane database analysis17 included 3 of the studies and concluded that there was no consistent effect of antihypertensive treatment on the progression of dementia; the problems in integration were that many patients were withdrawn from the treatment groups. As shown in the Table, these studies included elderly patients with either isolated systolic hypertension or systolic and diastolic hypertension. Most of the studies used the same measure of dementia, the Mini Mental State Exam (MMSE), which makes the results comparable. The main finding was that there was a small decline in the development of dementia with active antihypertensive treatment, but in most of the individual studies the change was not significant. One question for the difference in the drug effects is whether there were significant differences between the blood pressure changes in the active treatment or placebo groups. Thus, in the SCOPE trial,13 there was no significant difference in the dementia changes between the two groups; in both groups it was very small (the MMSE score changed from 28.5 to 28.0 in the candesartan group and 28.5 to 27.9 in the control group). The blood pressure changes in the two groups were dramatic (a decrease of 22/11 mm Hg in the candesartan group and 18/9 mm Hg in the control group) but not very different. While it is hard to interpret the significance of the dementia difference between the two groups, the fact that neither group showed a significant change over nearly 4 years of substantial blood pressure reduction is disappointing. The most recent study that examined this issue is HYVET-COG,15 which randomized hypertensive persons older than 80 years to active treatment (diuretic plus perindopril). The study was stopped after 2 years of treatment because there was a significant reduction in stroke and a marked difference in blood pressure (15/6 mm Hg) in the actively treated group. There were 263 cases of new dementia, but the rate of onset was only minimally slower in the treated group (33 per 1000 patient-years) than in the control group (38 per 1000 patient-years), which was disappointing; the difference was not significant. The authors compared the results of HYVET-COG with 3 other intervention trials that had collected similar data, all of which had shown trends for the active treatment of hypertension to produce a modest reduction in dementia (Syst-Eur, PROGRESS, and SHEP). The best known of these results was Syst-Eur, which showed a more dramatic result than other trials11(Table). However, the comparative numbers of patients in whom dementia developed during the study was initially quite small (21 in the placebo and 11 in the active treatment group). A second analysis10 with a longer follow-up (4 years rather than 2) found double the number of patients with dementia (64 vs 32), 41 of whom were classified as having Alzheimer’s disease. The benefit of antihypertensive treatment was virtually the same, a reduction of 55% in dementia. A recent prospective study (the Women’s Health and Aging Study II)18 in elderly women compared the effects of diuretics and angiotensin-converting enzyme inhibitors on the rate of impairment of cognition over 9 years, but the results are hard to interpret because the women were not randomized to the 2 drugs and findings were compared with changes in women who did not require antihypertensive treatment. The study found marked reductions in the dementia measures over the 9 years of observation in the hypertensive women treated with either drug when compared with the nontreated nonhypertensive group. Although investigations looking at the effects of different types of antihypertensive medication on the prevention of dementia have been relatively few, there is some evidence that there may be drug-specific effects. Thus, there is some evidence that neither diuretics nor candesartan, an angiotensin receptor blocker, have much protective effect, as shown in the Table. Although the Women’s Health and Aging Study II mentioned above was suggestive that both diuretics and angiotensin-converting enzyme inhibitors might be protective, it was not a randomized trial. PROGRESS also suggested the possibility of protective effects from angiotensin-converting enzyme inhibitors plus a diuretic, although the change was not significant. β-Blockers have not been shown to have any protective effect (see the MRC study in the Table). As shown in the Table, the most convincing effect was obtained in the Syst-Eur study,11 in which elderly systolic hypertensive patients were randomized to receive either the calcium channel blocker nitrendipine (a dihydropyridine similar to nifedipine) or placebo. After 2 years, there was a 50% reduction in the progression of dementia in the actively treated patients, a change that was greater than what was achieved in any of the other studies. It has been suggested that this may be a specific non–blood pressure effect, on the grounds that there may be an important role of intracellular calcium in Alzheimer’s disease and the destruction of brain cells.19 It has been accepted that the role of antihypertensive treatment in the prevention of dementia requires further study. There has never been any serious J-curve discussion suggesting that lowering blood pressure to below a certain level might actually make things worse and increase the likelihood of a stroke or dementia (in contrast to cardiac disease). However, the literature on blood pressure and dementia does not rule out the possibility. In patients with Alzheimer’s disease and vascular dementia, relatively low blood pressure and orthostatic hypotension have been reported in one small study.20 A prospective 3-year study21 found, not surprisingly, that dementia was more likely to develop in hypertensive persons, but individuals who showed a marked decrease in systolic blood pressure (20 mm Hg) showed a 2.5 times increased risk of dementia. Another prospective 10-year study7 found that persons who developed dementia were more likely to be hypertensive at the start of the observation period but showed a more marked decrease in blood pressure over the course of the study, leading to the suggestion that the development of the blood pressure decrease might be the consequence of Alzheimer’s disease–related brain damage rather than the cause. The issue of how aggressively blood pressure should be treated to reduce the progress of dementia will be addressed for the first time by the Systolic Blood Pressure Intervention Trial (SPRINT), which will be started by the National Institutes of Health in about 2 years. The trial will recruit patients aged 55 years or older with systolic blood pressure levels ≥130 mm Hg and at least one additional cardiovascular risk factor; they will be randomized to be treated to a systolic blood pressure level <140 or <120 mm Hg. While one goal of the study will be to see whether reaching the lower blood pressure value will further reduce strokes, there will also be a substudy to see whether the lower blood pressure goal will reduce the rate of age-related cognitive impairment and dementia (SPRINT-MIND substudy). Despite its growing importance as life expectancy continues to increase, the possible role of antihypertensive treatment in preventing dementia remains uncertain. It is disappointing that many of the large clinical trials that have investigated the effects of antihypertensive treatment and demonstrated the prevention of strokes have generally not evaluated changes in mental function; such changes are not that difficult to include in the studies. It could be argued that the convincing evidence that antihypertensive treatment in the elderly prevents strokes means that these patients will be treated anyway, so that more knowledge of the reduction of dementia will not have much effect on clinical recommendations, but the two questions discussed here remain unanswered. Thus, we do not know whether more aggressive treatment of hypertension in aging persons will help to prevent dementia. The superior prevention shown by a calcium channel blocker in one study raises the possibility that there may be blood pressure–independent protective effects related to intracellular calcium. This is an area of research that is in need of further investigation.