Journal of Human (1999) 13, 357–358  1999 Stockton Press. All rights reserved 0950-9240/99 $12.00 http://www.stockton-press.co.uk/jhh COMMENTARY Does the treatment of isolated systolic hypertension prevent ?

CE Clarke Department of Neurology, City Hospital NHS Trust, Dudley Road, Birmingham B18 7QH, UK

Keywords: systolic hypertension; ; ; Alzheimer’s disease

Introduction years with a systolic BP of between 160 to 219 mm Hg and a diastolic BP of below 95 mm Hg. They It has been gratifying in recent years to find research were randomised to nitrendipine 10 mg to 40 mg into vascular dementia increasing. It has always daily with the possible addition of enalapril 5 mg to been the poor relation of stroke, transient ischaemic 20 mg daily, hydrochlorothiazide 12.5 mg to 25 mg attack and amaurosis fugax in terms of research, rat- daily, or both. The aim was to reduce the systolic ing alongside anterior ischaemic optic neuropathy BP by at least 20 mm Hg to reach a value of below in terms of neglect. Vascular dementia is a major 150 mm Hg. Mini-mental state examination was per- public health problem which will increase with the formed annually and if 23 or less, formal psycho- aging of the population over the next 20 years. Com- metric testing assessed the patients to decide munity-based surveys have indicated that around whether they fulfilled DSM-III-R criteria for 5% of people over the age of 65 and 15–20% of those dementia. The diagnosis of vascular dementia or over the age of 80 will suffer from severe dementia, 1 Alzheimer’s Disease was made using brain imaging with up to 40% of these being vascular in origin. and the Hachinski score. A number of community-based studies have Using the intention-to-treat analysis, after a related hypertension with cognitive decline2 and 3,4 median follow-up of 2 years, the incidence of dementia, although this has not been a universal dementia in the placebo group (n = 1180) was 7.7 finding,5–7 perhaps related to blood pressure (BP) 4 cases per 1000 patient-years, compared with the falling with the onset of dementia. Recent clinical active treatment arm (n = 1238) with an incidence trials have examined whether the treatment of iso- of 3.8 cases per 1000 patient-years. This represented lated systolic hypertension can reduce the incidence a 50% reduction in the rate of dementia on active of dementia, vascular or otherwise. The results of treatment with 95% confidence intervals for this the Systolic Hypertension in Europe Trial (Syst-Eur) reduction ranging from between 0% and 76%. The suggested that the calcium channel blocker nitrendi- level of significance is quoted as ‘P = 0.05’. The trial pine, with or without supplementary enalapril or was stopped prematurely because the second of four hydrochlorothiazide, was capable of reducing the planned interim analyses showed a significant incidence of dementia over 2 years by 50%.8 The reduction in stroke which was the primary outcome incidence of Alzheimer’s disease was also reduced measure of the study. in this study. This is in contrast with the findings of The statistical analysis and conduct of this trial the Systolic Hypertension in the Elderly Programme can be criticised on a number of counts. As the (SHEP) in which treatment with chlorthalidone with or without supplementary atenolol failed to influ- authors themselves concede in the Discussion, the ence the incidence of dementia over 4.5 years.9 incidence of dementia in the control group of 7.5 Can we rely on the findings of the Syst-Eur Trial per 1000 patient-years is low compared with the and how does this compare with previous work? findings from population-based studies with an inci- What are the implications for future research and dence of around 10 per 1000 patient-years. It is con- patient management? ceivable that, whatever the reason for this reduced incidence in the placebo group, it had a more pro- found effect on the active treatment group and there- Syst-Eur Trial methodology fore led to the apparent reduction in dementia inci- The Systolic Hypertension in Europe Trial8 dence between the groups. recruited non-demented patients over the age of 60 The actual difference of 21 cases of dementia in the placebo group compared to 11 in the active treat- ment group reached a significance value of exactly Correspondence: CE Carke 0.05. Thus, the lower confidence interval for the dif- Received 10 February 1999; accepted 11 February 1999 ference between the groups was 0%. This raises Does the treatment of ISH prevent dementia? CE Clarke 358 concerns that there was no real difference between Both of the recent trials conclusively showed that the two groups. The authors have also not been con- reducing isolated systolic hypertension significantly sistent since the a priori power calculation for the reduces the incidence of stroke and cardiac mor- dementia part of this trial used a P value of 0.01 to bidity and mortality.8,9 Therefore, it is unlikely that assess significance not 0.05. The difference between any medical ethics committee would countenance the two arms would clearly not be significant at the such a trial with a control arm comprised of no treat- 0.01 level. ment. Adding these points together with the premature What practical advice can be given to the clinician termination of the study at just 2 years, with only in this situation? Faced with a patient with isolated two post-baseline assessments of cognition, we must systolic hypertension, this should be treated to conclude that the authors’ assertion that this antihy- reduce the risk of stroke and serious cardiovascular pertensive regime can reduce the incidence of events. It may reduce the likelihood of vascular dementia is premature. dementia and even Alzheimer’s disease. The patient presenting with a dementia, of whatever type, along Comparison with previous work with systolic hypertension should also have the lat- ter reduced to within normal limits in the hope that In the Systolic Hypertension in the Elderly Pro- this will prevent or reduce further cognitive decline. gramme (SHEP), 4736 patients with systolic BP of 160 to 219 mm Hg and diastolic BP of less than 90 See Correspondence, page 419. mm Hg were randomised to chlorthalidone (12.5–25 mg daily), with supplementary atenolol (25–50 mg daily) if necessary, or matching placebo.9 After a References mean follow-up period of 4.5 years, the incidence 1 Forette F, Boller F. Hypertension and the risk of of dementia in the active treatment group was 1.6% dementia in the elderly. Am J Med 1991; 90 (Suppl which was not significantly different from the inci- 3A): 14S–19S. dence of 1.9% in the placebo group. Like the Syst- 2 Launer LJ et al. The association between midlife blood Eur Trial, SHEP demonstrated a significant pressure levels and late-life cognitive function. JAm reduction in the incidence of stroke with the treat- Med Assoc 1995; 274: 1846–1851. ment of systolic hypertension (absolute 5 year bene- 3 Hoffman A et al. Atherosclerosis, apolipoprotein E, fit of 30 events per 1000 participants) along with a and prevalence of dementia and Alzheimer’s disease in the Rotterdam study. Lancet 1997; 349: 151–154. reduction in non-fatal myocardial infarction and 4 Skoog I et al. 15-year longitudinal study of blood coronary death (absolute 5 year benefit of 55 events pressure and dementia. Lancet 1996; 347: 1141–1145. per 1000 participants). 5 Prince M, Cullen M, Mann A. Risk factors for Alzhei- The 1992 Cochrane systematic review examining mer’s disease and dementia: a case-control study based randomised control trials comparing the calcium on the MRC elderly hypertension trial. Neurology channel blocker nimodipine with placebo con- 1994; 44: 97–104. cluded that there was insufficient evidence to prove 6 Desmond DW, Tatemichi TK, Paik M, Stern Y. Risk or disprove any effect of this agent on reducing the factors for as correlates of cog- incidence of dementia.10 nitive function in a stroke-free cohort. Arch Neurol In an open label, non-controlled trial of 17 1993; 50: 162–166. 7 Guo Z, Viitanen M, Fratiglioni L, Winblad B. Low patients with multi-infarct or mixed dementia, blood pressure and dementia in elderly people: the reducing systolic hypertension to within normal Kungsholmen project. BMJ 1996; 312: 805–808. limits (135–150 mm Hg) improved cognition as mea- 8 Forette F et al. Prevention of dementia in randomised sured with the Cognitive Capacity Screening Exam- double-blind placebo-controlled Systolic Hyperten- ination.11 Interestingly, reducing BP below this level sion in Europe (Syst-Eur) trial. Lancet 1998; 352: resulted in a deterioration of cognition. 1347–1351. 9 SHEP Co-operative Research Group. Prevention of stroke by treatment in older per- Implications sons with isolated systolic hypertension. J Am Med Only two large-scale randomised controlled trials Assoc 1991; 265: 3255–3264. have examined the strategy of treating isolated sys- 10 Qizilbash N, Arrieta JL, Birks J. Nimodipine for pri- mary degenerative, mixed and vascular dementia tolic hypertension to reduce dementia (Syst-Eur8 9 (Cochrane Review). In: The Cochrane Library, Issue 1, and SHEP ). At present, no clear conclusions emerge 1999. Update Software: Oxford. from these studies. Further trials would be required 11 Meyer JS et al. Improved cognition after control of risk to resolve this question. However, it is unlikely that factors for multi-infarct dementia. J Am Med Assoc such work will be performed on ethical grounds. 1986; 256: 2203–2209.