To the editor
Regular aspirin use is routinely recommended for protection against cerebrovascular and cardiovascular events. Although cerebrovascular pathology contributes importantly to risk of dementia, it is not well-established whether aspirin use attenuates this risk. Our previous work has demonstrated that transient ischemic attack (TIA) or report of stroke-like symptoms are associated with increased risk of cognitive impairment in cross-sectional analyses1 and with incident cognitive decline during longitudinal follow-up2 in the Reasons for Geographical and Racial Differences in Stroke (REGARDS) study3.
Our objective was to determine whether regular aspirin use protects against incident cognitive decline in REGARDS. The REGARDS study has enrolled 30,239 participants for whom extensive demographic and health data were collected. Several cognitive assessments have occurred during longitudinal follow-up of the cohort. A measure of global cognitive status, the Six-item Screener (SIS) has been administered annually since 2003. Additional cognitive measures were subsequently added to the protocol for consistency with the 5-minute neuropsychological battery recommended by the National Institute of Neurological Disorders and Stroke-Canadian Stroke Network (NINDS-CSN) Harmonization Standards.4 These tasks, administered every 2 years, evaluate memory (Word List Learning—WLL and Word List Recall—WLR) and executive function (Animal Fluency Test—AFT and Letter Fluency—LF)‥
Longitudinal SIS data were available for 23,915 participants who were cognitively normal at baseline (SIS > 4): 38% black, 43% female, mean age 64 years (SD=9.2) at enrollment, with an average follow-up of 5.9 years. Analyses of the SIS data indicate that those who did not use aspirin regularly have a higher likelihood of incident impairment on the SIS (SIS< 5 on most recent assessment) in univariate models (OR: 1.11, 95% CI: 1.09, 1.13). However, after adjustment for demographic factors, the association between regular aspirin use and incident impairment on the SIS is no longer significant (OR 0.99, 95% CI 0.89, 1.09). Additional inclusion of Framingham Stroke Risk total scores or individual factors did not change this finding.
For separate analyses of cognitive change on the WLL, WLR, AFT and LF measures, difference scores were calculated based on each participant’s first and last assessments on each measure, adjusted for initial score values. There were 12,231 participants with longitudinal WLL data: 35% black, 56% female, mean age 64 (SD=8.4), with an average follow-up of 3.6 years. In ANCOVA models the association between aspirin use and change in WLL score was significant before (Model 1), but not after adjustment for demographics (Model 2), or risk factors (Models 3 & 4). Analysis of the WLR data produced similar findings. To understand the reduction of association between aspirin use and change in WLL and WLR scores due to demographic factors, we studied the impact of adding each demographic variable individually and found that age is the dominant factor responsible for the effect attenuation. Similar results were obtained for AFT and LF (see Table).
Table.
Multivariable results for cognitive change scores by aspirin use.
| Model 1 | Model 2 | Model 3 | Model 4 | |
|---|---|---|---|---|
| WLL (n = 12,231) | ||||
| Aspirin users | −0.099 | −0.44 | −1.12 | −0.32 |
| Non-aspirin users | 0.418 | −0.35 | −1.21 | −0.32 |
| p-value | <0.0001 | 0.23 | 0.34 | 0.97 |
| WLR (n= 11,722) | ||||
| Aspirin users | 0.0096 | −0.13 | −0.39 | −0.087 |
| Non-aspirin users | 0.27 | −0.064 | −0.41 | −0.050 |
| p-value | <0.0001 | 0.043 | 0.62 | 0.27 |
| Letter F Fluency (n = 5,126) | ||||
| Aspirin users | −0.4962 | −0.8868 | −1.092 | −0.8766 |
| Non-aspirin users | −0.3092 | −0.7691 | −1.067 | −0.7400 |
| p-value | 0.053 | 0.22 | 0.84 | 0.18 |
| Animal Fluency (n = 8,366) | ||||
| Aspirin users | −1.01 | −1.37 | −1.47 | −1.36 |
| Non-aspirin users | −0.65 | −1.22 | −1.33 | −1.22 |
| p-value | <0.0001 | 0.086 | 0.21 | 0.13 |
Model 1: adjusted for baseline score
Model 2: adjusted for first assessment, demographics (age, sex, race, education, income, region—stroke belt vs non-stroke belt)
Model 3: adjusted for first assessment, demographics, and Framingham Stroke Risk Profile factors considered individually (age, systolic blood pressure, antihypertensive medication, diabetes, current smoker, history of cardiovascular disease, a trial fibrillation, left ventricular hypertrophy)
Model 4: adjusted for first assessment, demographics, Framingham Stroke Risk Profile total score
The protective effects of aspirin for heart disease and as a secondary preventive treatment for stroke are well-documented. Many epidemiologic studies have identified stroke and heart disease as independent risk factors for cognitive decline or dementia. It would stand to reason that daily aspirin use could reduce the risk of cognitive decline through reduction of cardiac and cerebrovascular disease, and that this effect should be larger in people having higher risk of stroke. Previous small studies5 have reported a protective effect, and this was seemingly confirmed in data from the Canadian Health Study6 and the Baltimore Longitudinal Study of Aging7.
Other longitudinal cohorts have not replicated these findings. The Women’s Health Study found no association between aspirin use and cognitive performance over 9.6 years of follow-up.8 Analysis of data from 3,229 participants in the Cardiovascular Health Study over age 65 reported no protective effect for aspirin.9 A 5-year RCT of low-dose aspirin, the Aspirin for Asymptomatic Atherosclerosis trial, reported no association between cognitive decline and aspirin use.10
We examined cognitive change in REGARDS along two dimensions: memory and executive function. We also were able to capture clinically relevant incident impairment in global cognitive status. Given the large sample, REGARDS is sufficiently powered to detect even small associations, should they exist. However, in this large national sample of black and white adults, we saw no relationship between daily aspirin use and cognitive change over 2 to 6 years of follow-up after controlling for the impact of age.
ACKNOWLEDGMENTS
This research project is supported by a cooperative agreement U01 NS041588 from the National Institute of Neurological Disorders and Stroke, National Institutes of Health, Department of Health and Human Service. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Neurological Disorders and Stroke or the National Institutes of Health. Representatives of the funding agency have been involved in the review of the manuscript but not directly involved in the collection, management, analysis or interpretation of the data. The authors thank the other investigators, the staff, and the participants of the REGARDS study for their valuable contributions. A full list of participating REGARDS investigators and institutions can be found at http://www.regardsstudy.org
Conflict of Interest: Dr. Kelley has served as a consultant to Lilly and has received research support from Novartis, Merck, Lundbeck, Lilly, and from NIH.
Dr. McClure receives research support from Genzyme Corporation, NIH, and NASA.
Dr. Wadley receives grant funding from Genzyme Corporation, receives research support from NIH,, served as an invited speaker for the Alzheimer’s Association Research Roundtable, and serves on the editorial boards of Current Gerontology and Geriatrics Research and Journal of Aging Science.
Dr. Unverzagt has served as a consultant to Piramal Life Sciences, Eli Lilly and Company, and UCB Biosciences; serves as an editorial board member for Neuropsychology, and Journal of the International Neuropsychological Society; receives research support from NIH and Posit Science Inc; holds stock in Eli Lilly, Inc; and provides medico-legal reviews.
Dr. Kissela is a consultant for Allergan, has been paid to adjudicate clinical trial events for AbbVie and Reata, and receives research support from NIH.
Dr. Kleindorfer serves as a consultant for Genentech, and receives research support from the NIH and the CDC.
Dr. G. Howard receives research support from NIH, and is on study advisory boards for Cerevast and PhotoThera.
Footnotes
Author Contributions: All authors meet the criteria for authorship stated in the Uniform Requirements for Manuscripts Submitted to Biomedical Journals. The corresponding author (Brendan Kelley) affirms that he has listed everyone who contributed significantly to the work and has obtained written consent from all contributors.
Brendan J. Kelley, Leslie A. McClure, Frederick W. Unverzagt, Brett Kissela,; Dawn Kleindorfer, George Howard, and Virginia G. Wadley contributed to conception and design, acquisition of data, analysis and interpretation of data; drafting the article or revising it critically for important intellectual content; and final approval of the version to be published.
REFERENCES
- 1.Wadley VG, McClure LA, Howard VJ, et al. Cognitive status, stroke symptom reports, and modifiable risk factors among individuals with no diagnosis of stroke or transient ischemic attack in the REasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Stroke. 2007;38:1143–1147. doi: 10.1161/01.STR.0000259676.75552.38. [DOI] [PubMed] [Google Scholar]
- 2.Kelley BJ, McClure LA, Letter AJ, et al. Report of stroke-like symptoms predicts incident cognitive impairment in a stroke-free cohort. Neurology. 2013;81:113–118. doi: 10.1212/WNL.0b013e31829a352e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Howard VJ, Cushman M, Pulley L, et al. The reasons for geographic and racial differences in stroke study: objectives and design. Neuroepidemiology. 2005;25:135–143. doi: 10.1159/000086678. [DOI] [PubMed] [Google Scholar]
- 4.Hachinski V, Iadecola C, Petersen RC, et al. National Institute of Neurological Disorders and Stroke-Canadian Stroke Network vascular cognitive impairment harmonization standards. Stroke. 2006;37:2220–2241. doi: 10.1161/01.STR.0000237236.88823.47. [DOI] [PubMed] [Google Scholar]
- 5.Jonker C, Comijs HC, Smit JH. Does aspirin or other NSAIDs reduce the risk of cognitive decline in elderly persons? Results from a population-based study. Neurobiol Aging. 2003;24:583–588. doi: 10.1016/s0197-4580(02)00188-4. [DOI] [PubMed] [Google Scholar]
- 6.Lindsay J, Laurin D, Verreault R, et al. Risk factors for Alzheimer's disease: a prospective analysis from the Canadian Study of Health and Aging. Am J Epidemiol. 2002;156:445–453. doi: 10.1093/aje/kwf074. [DOI] [PubMed] [Google Scholar]
- 7.Stewart WF, Kawas C, Corrada M, et al. Risk of Alzheimer's disease and duration of NSAID use. Neurology. 1997;48:626–632. doi: 10.1212/wnl.48.3.626. [DOI] [PubMed] [Google Scholar]
- 8.Kang JH, Cook N, Manson J, et al. Low dose aspirin and cognitive function in the women's health study cognitive cohort. BMJ. 2007;334:987. doi: 10.1136/bmj.39166.597836.BE. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Szekely CA, Breitner JC, Fitzpatrick AL, et al. NSAID use and dementia risk in the Cardiovascular Health Study: role of APOE and NSAID type. Neurology. 2008;70:17–24. doi: 10.1212/01.wnl.0000284596.95156.48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Price JF, Stewart MC, Deary IJ, et al. Low dose aspirin and cognitive function in middle aged to elderly adults: Randomised controlled trial. BMJ. 2008;337:a1198. doi: 10.1136/bmj.a1198. [DOI] [PMC free article] [PubMed] [Google Scholar]