Introduction: Use of angiotensin II (ATII)-stimulating antihypertensive medication (AHM), including angiotensin receptor blockers (ARBs) and dihydropyridine calcium channel blockers (CCBs), has been associated with lower dementia risk. Previous studies had relatively short followup periods. The aim of this study is to investigate if these effects are sustained over longer periods.Methods: This post hoc observational analysis was based on data from a dementia prevention trial (preDIVA and its observational extension), among Dutch community-dwelling older adults without prior diagnosis of dementia. Differential associations between AHM classes and incident dementia were studied after 7.0 and 10.4 years, based on the median follow-up durations of dementia cases and all participants.Results: After 7 years, use of ATII-stimulating antihypertensives [hazard ratio ¼ 0.68, 95% confidence interval (CI) ¼ 0.47-1.00], ARBs (hazard ratio ¼ 0.54, 95% CI ¼ 0.31-0.94) and dihydropyridine CCBs (hazard ratio ¼ 0.52, 95% CI ¼ 0.30-0.91) was associated with lower dementia risk. After 10.4 years, associations for ATIIstimulating antihypertensives, ARBs and dihydropyridine CCBs attenuated (hazard ratio ¼ 0.80, 95% CI ¼ 0.61-1.04; hazard ratio ¼ 0.75, 95% CI ¼ 0.53-1.07; hazard ratio ¼ 0.73, 95% CI ¼ 0.51-1.04 respectively), but still suggested lower dementia risk when compared with use of other AHM classes. Results could not be explained by competing risk of mortality. Conclusion:Our results suggest that use of ARBs, dihydropyridine CCBs and ATII-stimulating antihypertensives is associated with lower dementia risk over a decade, although associations attenuate over time. Apart from methodological aspects, differential effects of antihypertensive medication classes on incident dementia may in part be temporary, or decrease with ageing.

Background: Midlife hypertension is a risk factor for all-cause dementia. Treatment of hypertension with antihypertensive medication (AHM) may therefore be a promising strategy to delay or prevent dementia. Recent reports have suggested that AHM are associated with differential effects on dementia risk, due to class-specific mechanisms.In this study, we assess the association between different AHM-classes and incident dementia, using data from the preDIVA observational extension (POE) study. Method:All participants who reported use of AHM at baseline were included in the analyses. We distinguished angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor blockers (ARBs), beta-blockers, calcium channel blockers (CCBs), and diuretics. Data on incident dementia was collected during the randomised phase at 2, 4, 6-8 and, after discontinuation of the intervention, 10-12 years after baseline.Cox proportional hazards regression was used to study the association between use of different AHM-classes, compared with use of any other AHM class, and incident dementia. A post-hoc analysis was performed, dividing follow-up time in two periods (preDIVA median follow-up acting as cut-off point) to account for the effect of time between exposure and dementia.Result: 1953 participants who reported AHM use at baseline were included, with a mean age of 74.5 (±2.5) and a mean systolic blood pressure of 156.4 mmHg (±21.5).After a median follow-up of 10.4 years (IQR 6.7-11.0), 228 (11.7%) participants had developed dementia. Adjusted hazard ratios (HR) for dementia were 1.07 (95%CI 0.80-1.42) for ACE inhibitors, 0.75 (95%CI 0.53-1.06) for ARBs, 1.00 (95%CI 0.77-1.31) for beta-blockers, 0.92 (95%CI 0.68-1.24) for CCBs, and 1.02 (95%CI 0.78-1.33) for diuretics. ARBs and CCBs were inversely associated with dementia during the first time period (HR 0.46, 95%CI 0.25-0.84 and HR 0.62, 95%CI 0.38-0.99 respectively), which was not apparent anymore during the second one. Conclusion:Despite short-term beneficial associations of CCBs and ARBs and a promising HR for ARBs also after longer follow-up, associations between any AHMclass and incident dementia were not convincing after complete follow-up. These findings are possibly best explained by the long follow-up in older participants with high mortality-and dementia rates, possibly attenuating AHM-class effects on dementia incidence.