Table 1

Summary of key observational studies investigating the association between aspirin use and mental illness.

Author, year; study


Study design

Agent and dosage


Psychiatric measure

Presentation of results

Key finding

Sturmer, 1996; EBSHP

ASA use affected decline of cognitive function

Cohort study

ASA: <1, 1 to 2 and, >2 tablets/day ASA effect hypothesized to be more dependent on frequency than on dose, mean daily dose was not used

3,631 NII >65 years old; 2,773 at 3-year follow-up, and 2,023 at 6-year follow-up



No significant effect seen. Modest benefit of ASA, especially with intermittent use, on decline of cognitive function

Henderson, 1997

ASA prevented dementia or cognitive impairment

Community survey. Two-wave: cross-sectional at 2 wave; longitudinal from 1to 2 wave

ASA yes/no (unknown dose)

1,045 participants 70 years old at baseline; 588 people with cognitive assessment at both waves


Mean and SD

Cross-sectional: no significant difference in cognitive tests. Longitudinal: no differences in cognitive decline or incidence of dementia

Stewart, 1997; BLSA

Reduced risk of AD in ASA users


ASA yes/no; NSAIDs yes/no; acetaminophen yes/no

1,686 participants in baseline study

BIMCT, MMES, Immediate and Delayed Cued Recall, BNT, CVF, TMT A and B, Clock Drawing and other constructions, CESDS, PFAQ, NART


Non-significant AD risk ratio for ASA users. Duration of ASA use and the risk of AD were not significantly associated

Peacock, 1999; ARIC study

Association of regular use of NSAIDs or ASA with cognitive function

Cross-sectional cohort study

NSAIDs yes/no; ASA yes/no

13,153 participants, 48 to 67 years old

Delayed word recall test, WAIS-R digit symbol subtest, WFT


Weak negative association (<0.1) between current ASA and word fluency & recall. No association of lifetime ASA and cognitive index scores. ASA treatment for 8 years weakly associated with better word fluency

Landi, 2003

Relationship between NSAIDs and AD


ASA yes/no; NSAIDs yes/no

2,708 participants admitted to home care programs


Mean and SD

NSAID users had a nearly 50% lower risk of cognitive impairment. For subjects using ASA, the risk estimated was similar; 67% decreased risk of cognitive impairment associated with non-ASA NSAID use

Nilsson, 2003; OCTO-Twin

ASA protective for AD

Cross-sectional and longitudinal

Low-dose ASA (75 mg/day + 3500 mg/week) 250, 500 mg

702 participants >80 years old, 91 with dementia at baseline; 88 developed dementia during observational period; 315 people at follow-up

DSM-III-R criteria for dementia, NINCDS/ARDRA criteria for AD, NINDS/AIREN criteria for vascular dementia, MMSE


At 9-year follow-up, significant association between ASA and lower frequency of AD and all dementia; significantly more likely that intact cognitive function was maintained in those taking ASA. Use of low-dose ASA alone did not affect the risk ratio

Cornelius, 2004; Kungsholmen Project

Association between ASA and NSAIDs with AD and overall dementia, and influence of apoE ε4

Longitudinal cohort study

ASA 75 to 500 mg/day yes/no; NSAIDs yes/no. Differences in dosages not considered

1,301 subjects

75 years old dementia-free at baseline; 987 at 1*follow-up (314 died); 650 at 2*follow-up (281 died)

DSM III-R criteria for dementia, Hachinski scale for differential diagnoses AD versus VaD versus Mixed dementia, MMSE, neurological and psychiatric examinations, neuropsychological assessment


6-year increased AD risk in the ASA/apoE ε4 negative group

Jonker, 2004; LASA

Protective effect of ASA on cognitive decline in older people

Community-based study; ?case-control

NSAIDs yes/no; ASA <100 mg yes/no

612 participants, 62 to 85 years old

MMSE, AVLT, coding task


3-year follow-up of decline in episodic memory (immediate recall) for ASA users was reduced by more than three times. Effect of ASA was significant only in >75-year-olds

Shepherd, 2004; SOP Study

ASA protective against AD

ASA yes/no

151 NII,

75 years old, dementia-free

MMSE, Logical Memory and Similarities subtest from WAIS-R, BNT, visuo-perceptual abilities from the JLOT, COWAT

Mean and SD

ASA use was a significant positive predictor of performance on the Logical Memory test

Arvanitakis, 2008; ROS

Relation between NSAIDs/AD, change in cognition, and AD pathology


ASA yes/no; non-ASA NSAIDs yes/no

1,019 Catholic clergy, mean age 75 years old, dementia-free

As reported previously [12]


At 1 year-no apparent relation of ASA to incident AD, change in cognition, or AD pathology

Szekely, 2008; CHCS

Association between NSAIDs, ASA, and acetaminophen with dementia and AD


ASA yes/no; NSAIDs yes/no; acetaminophen yes/no. No dosage reported

3229 participants >65 years old, dementia-free 1228 ASA users

NINCDS/ARDRA criteria for AD, ADDTC criteria for VaD, Mixed dementia diagnosis, 3MSE, MRI


At 10 years, risk of AD, VaD, and all-cause dementia was not associated with use of ASA

Almeida, 2010

ASA decreased prevalence of depression and cognitive impairment


Not reported

5,556 men 69 to 87 years old



ASA not associated with lower OR of depression or cognitive impairment in >75-year-old men. Discontinuation of ASA between the two assessments related to greater OR of depression than non-users

Pasco, 2010; GOS

ASA reduced the risk for depression; ASA + statin reduced risk of de novo depression

Case-control study, retrospective cohort analysis

ASA yes/no

386 women >50 - years old. 1* MDD >50 years old versus no MDD. No prior MDD, followed up from baseline or time of exposure to ASA, until 1* MDD or 10-year follow-up


OR and HR

OR for MDD in the ASA group was 0.18, P = 0.1 The prevalence of exposure to statins + ASA was lower for women with MDD; OR for MDD was 0.15 (95% CI 0.03 to 0.65, P = 0.01). Exposure to statins + ASA was associated with a reduced risk for MDD

Waldstein, 2010; BLSA

Relation between ASA and NSAIDs and age-related change in cognitive functions

Cross-sectional and longitudinal

ASA yes/no; NSAIDs yes/no

2300 dementia-free

Digits Forward and Backward portions of WAIS-R, CVLT, BVRT, TMT, Letter & Category Fluency, BNT, MMSE, BIMCT


Cross-sectional: use of ASA related to better average performance across testing sessions on measures of verbal and visual learning, and memory and global mental status. Longitudinal: ASA related to greater prospective decline on BIMCT and the BVRT. Significant effects of ASA use were noted for the BVRT, the CVLT learning slope and short free recall, and the MMSE, and indicated better average levels of function for ASA users

Clinical population

Ketterer, 1996; Coronary angiography

Regular ASA prophylactic therapy for IHD associated with emotional distress

ASA 80 to 325 mg

174 men

CMS, Framingham Type A Scale, KSSFC

ASA associated with less depression and anxiety or worry on the KSSFC

Stanford, 1999; HF or previouS OHT or CB

Usual schedule of drug therapy maintained

135 participants

Profile of Mood States

Mean and SEM

More positive mood in ASA groups due to less fatigue. Tension and TMD in ASA patients just failed to reach criterion for statistical significance

Broe; 2000; SOP Study dementia


80% on ASA 175 mg; no high dose

163 NII

75 years old with different dementias categories, and 373 control subjects

NINCDS-ADRDA criteria for AD

Inverse association between ASA and AD, but not other dementia, not dosage-related

Mendlewicz, 2006; treatment-resistant DP

Accelerating effect of ASA in combination with fluoxetine

Open-label, uncontrolled

Treated openly during 4 weeks with ASA 160 mg/day in addition to their current antidepressant treatment

21 participants underwent

4 weeks SSRI


Mean and SD

SSRI + ASA showed a global response rate of 52%. Remission was achieved in 43% of the total sample and 82% of the responder sample. In the responder group, a significant improvement was seen within week 1, which was sustained until day 28

Stolk, 2010; bipolar disorder

NSAIDs and glucocorticoids ameliorate bipolar symptoms

ASA 30 or 80 mg/day or ASA >80 mg/kg or non-selective NSAIDs or COX-2i or GCs + lithium

5145 participants receiving lithium

Incidence density of medication events (change in medication or increase of >30% of the current dose)

Incidence density ratio

Subjects receiving ASA 30 or 80 mg/day were 17% less likely to have a medication event; with >80 mg/kg ASA, non-selective NSAIDs, COX-2i, and GCs were not significant, but non- selective NSAIDs and GCs significantly increased medication events

3MSE, Modified Mini-Mental State Examination; AD, Alzheimer's disease; ADDTC, Alzheimer Disease Diagnostic and Treatment Centers; apoE, apolipoprotein E; ASA, acetylsalicylic acid (aspirin); AVLT, Auditory-Verbal Learning Test; BIMCT, Blessed Information-Memory-Concentration Test; BNT, Boston Naming Test; BVRT, Benton Visual Retention Test; CESDS, Center for Epidemiologic Studies Depression Scale; CIE, Canberra Interview for the Elderly; CMS, Cook-Medley Scale; COWAT, Controlled Oral Word Association Test; COX-2i, cyclooxygenase-2 inhibitor; CVF, Competing Values Framework; CVLT, California Verbal Learning Test; DSM-III-R, Diagnostic and Statistical Manual of Mental Disorders, 3rd revision; EBMT, East Boston Memory Test; F-U, follow-up; GC, Glucocorticoids; GDS, Geriatric Depression Scale; HDRS, Hamilton Depression Rating Scale; HR, hazard ratio; IHD, ischemic heart disease; JLOT, Judgment of Line Orientation; KSSFC, Ketterer Stress Symptom Frequency Checklist; MDD, Major Depressive Disorder; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging; NART, National Adult Reading Test; NII, Non-institutionalized individuals; NINCDS-ADRDA, National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer's Disease and Related Disorders Association; NINDS/AIREN, National Institute of Neurological Disorders and Stroke and Association Internationale pour la Recherché et l'Enseignement en Neurosciences; NSAID non-steroidal anti-inflammatory drug; OR, odds ratio; PFAQ, Pfeffer Activities Questionnaire; PMS, Profile of Mood States; SCID I RV-NP, Structured Clinical Interview for DSM-IV Axis I Disorders, Research Version, Non-patient; SD, standard deviation; SE, standard error; SEM, standard error of the mean; SLMT, Symbol Letter Modalities Test; SPMSQ, Short Portable Mental Status Questionnaire; SSRI, selective serotonin reuptake inhibitor; TMD, Total Mood Disturbance Score; TMT, Trail Making Test; VaD, vascular dementia; vs, versus; WAIS-R, (Wechsler Adult Intelligence Scale - Revised; WFT, Word Fluency Test.

Berk et al. BMC Medicine 2013 11:74   doi:10.1186/1741-7015-11-74

Open Data