Effects of aspirin on dementia and cognitive function in diabetic patients: the ASCEND trial

Abstract

Aims: Aspirin is widely used in cardiovascular disease prevention but is also associated with an increased risk of bleeding. The net effect of aspirin on dementia and cognitive impairment is uncertain.

Methods and results: In the ASCEND trial, 15 480 people from the UK with diabetes and no history of cardiovascular disease were randomized to aspirin 100 mg daily or matching placebo for a mean of 7.4 years. The 15 427 ASCEND participants with no recorded dementia prior to baseline were included in this cognitive study with a primary pre-specified outcome of 'broad dementia', comprising dementia, cognitive impairment, or confusion. This was ascertained through participant, carer, or general practitioner report or hospital admission diagnosis, by 31 March 2019 (∼2 years beyond the scheduled treatment period). The broad dementia outcome occurred in a similar percentage of participants in the aspirin group and placebo group: 548 participants (7.1%) vs. 598 (7.8%), rate ratio 0.91 [95% confidence interval (CI), 0.81-1.02]. Thus, the CI excluded proportional hazards of >2% and proportional benefits of >19%.

Conclusion: Aspirin does not have a large proportional effect on the risk of dementia. Trials or meta-analyses with larger total numbers of incident dementia cases to increase statistical power are needed to assess whether any modest proportional 10-15% benefits of 5-7 years of aspirin use on dementia exist.

Clinical trial registration: Current Controlled Trials number, ISRCTN60635500; ClinicalTrials.gov number: NCT00135226.

Keywords: Aspirin; Cardiovascular disease; Dementia.

© The Author(s) 2022. Published by Oxford University Press on behalf of European Society of Cardiology.

Figures

Structured Graphical Abstract

The effect of aspirin on dementia in the ASCEND trial: Study design and summary results.

Figure 1

Study profile. aPer cent of those not known to have died.

Figure 2

Effect of randomization to aspirin on dementia and cognitive function. Box sizes for the rate ratios are proportional to the amount of statistical information. aFifty-three individuals (26 on aspirin, 27 on placebo) are censored due to the broad dementia outcome before randomization. bCognitive z-score difference is adjusted for age at test and sex. The axis reversed to maintain the direction of ‘favours aspirin’ to the left.

Figure 3

Association of the incidence of vascular and bleed events with dementia. NEdem, number of participants with the non-fatal incident event and, subsequently, the broad dementia outcome; NE, total number of patients with the non-fatal incident event. aAdjusted for age, sex, randomized allocation, number of hospitalizations, incidence of the other non-fatal incident events, prior disease, and baseline predictors of dementia selected by backwards selection with psel = 0.2. bNon-disabling events (analyses are censored at disabling stroke or intracranial bleed). Intracranial bleed includes haemorrhagic stroke. cRate ratios associated with having the combined events tend to be higher than those with individual types because, for people with more than one type of event, they encompass the risks associated with each of the types of event.

Figure 4

Effect of randomization to aspirin on dementia in a meta-analysis of three trials. Box sizes for the rate ratios are proportional to the amount of statistical information. aASPREE trial: a randomized placebo-controlled trial of 100 mg daily aspirin in 19 114 healthy elderly individuals with a median follow-up of 4.7 years. The secondary endpoint of dementia is used. bJPAD trial: a randomized open-label trial of 80–100 mg daily aspirin in 2536 patients with diabetes with a median follow-up of 11.4 years. cThe ASCEND broad dementia outcome is used. dFixed-effects meta-analysis of ASPREE, JPAD, and ASCEND.