[Two of the authors respond:]
We acknowledge that our interpretation of the Syst-Eur trial was overly optimistic. Many criticisms have been levelled at this trial as reported by Forette and colleagues.[1–3] In the intention-to-treat analysis the incidence of all dementia was apparently reduced by 50%. Although this reduction, from 7.7 to 3.8 cases per 1000 patient-years, was statistically significant (p = 0.05), the 95% confidence interval (CI) was 0%-76%. In the per-protocol analysis, reduction was 60% (p = 0.03) and the 95% CI was 2%-83%.4 It has been pointed out that there appeared to be a large number of subjects not completing the protocol and that assignment of as few as 2 of these subjects could change the conclusions.2 However, Forette and colleagues reported that when subjects who could not be traced before publication were subsequently included in the analysis, the original conclusions were validated.5
The intention of Forette and colleagues' study was to determine whether antihypertensive treatment would reduce the incidence of vascular dementia.4 The finding of only 2 incident cases of vascular dementia among 32 cases of all dementia is not surprising. Although vascular dementia is considered to be much less common now than in the past, its distinction from Alzheimer's disease is less clear than once thought. Not only do they share common risk factors,6 but there is mounting evidence that the occurrence of strokes in people with Alzheimer's neuropathological changes may precipitate or exaggerate the clinical manifestations of dementia.7
These observations provide a plausible rationale for the findings of Forette and colleagues. However, the SHEP study failed to demonstrate that antihypertensive treatment reduced the incidence of dementia.8 These apparently conflicting results might be explained by differences in the 2 study populations; for example, the mean pretreatment diastolic blood pressure was 10 mm Hg lower in the SHEP study than in the Syst-Eur trial. It has been suggested that the use of a calcium-channel blocker in the Syst-Eur study could explain the apparent difference in findings.[4, 9] However, data from the Canadian Study of Health & Aging indicate that people with hypertension who were taking calcium-channel blockers were more likely to experience a decline in cognitive performance than those taking other antihypertensive medications.10
We thank Mark Clarfield for his observations and hope that this debate will be settled by a definitive trial using all causes of dementia as an end point. Indeed, such a study is already underway, the Study on Cognition & Prognosis in the Elderly (SCOPE).11 In the meantime we hope that systolic hypertension will be appropriately recognized and treated in older people, as this will reduce the incidence of stroke. Whether such treatment will also decrease the incidence of dementia remains to be seen.