Cardiovascular Medications and Dementia Risk in Older Adults: A Literature Review and Disproportionality Analysis Using OpenVigil FAERS Data

Abstract

Background

Cardiovascular medications are commonly prescribed to older adults; however, their potential association with cognitive decline remains poorly understood.

Objective

This study aimed to systematically evaluate the relationship between cardiovascular drugs and the risk of dementia.

Methods

(1) A retrospective disproportionality analysis of the FDA Adverse Event Reporting System data, accessed via OpenVigil 2.1, which examined 97 cardiovascular drugs across 14 therapeutic categories in patients aged ≥60 years, and (2) a literature review of case-reports of drug-induced cognitive impairment.

Results

Of the 97 drugs analyzed, disproportionate reporting signals (indicating more frequent reporting than expected by chance) were identified for 38 (39.2%) across four types of dementia: dementia (13.4%), Alzheimer’s disease (16.5%), vascular dementia (18.6%), and dementia with Lewy-bodies (6.2%). ACE inhibitors exhibited the highest signal rate (75.0%). Thirteen case-reports were identified, primarily involving statins (53.8%). Discontinuation of the drug resulted in cognitive improvement in 12/13 cases.

Conclusions

This study identifies disproportionate dementia-related adverse event reporting for nearly 40% of cardiovascular drugs examined, with ACE inhibitors and ARBs showing the highest signal rates. However, these findings are preliminary and require validation through future pharmacoepidemiological studies.

Keywords

cardiovascular medications dementia pharmacovigilance FAERS cognitive impairment drug safety

Introduction

Dementia is a progressive neurodegenerative disorder characterized by a decline in cognitive function, memory, and the ability to perform activities of daily living. It represents a major public health challenge, particularly among the older adult population, with Alzheimer’s disease, vascular dementia, and dementia with Lewy bodies being the most prevalent subtypes.^1,2 As the global population ages, the burden of dementia is expected to rise sharply, underscoring the need for a deeper understanding of its risk factors and potential contributors.³

Globally, approximately one new case of dementia occurs every 3.2 seconds. In 2020, over 55 million people worldwide were affected by dementia, and this number is anticipated to almost double every 20 years, reaching 78 million by 2030 and 139 million by 2050.⁴ In 2015, the global cost was US$ 818 billion, rising to US$ 1.3 trillion by 2019, with projections of US$ 2.8 trillion by 2030.⁵

Studies have indicated a link between dementia, Alzheimer’s disease, and vascular dementia with cardiovascular conditions and their associated risk factors, such as hypertension, obesity, diabetes, and smoking.^6,7 According to the Lancet Commission, hypertension and its midlife risk factors, including obesity and lack of physical activity, contribute to approximately 2-9.6% of dementia cases worldwide.⁸

Consequently, cardiovascular drugs are widely prescribed for older adults, often for prolonged periods of time. While these medications play a crucial role in managing cardiovascular morbidity and mortality, emerging evidence suggests that cardiovascular drugs may either protect against or contribute to cognitive decline, independent of their cardiovascular effects. Reports of drug-induced cognitive impairment have raised concerns regarding the potential association between cardiovascular pharmacotherapy and the onset or exacerbation of dementia.^2,8-10

Despite these concerns, the relationship between cardiovascular drugs and dementia remains unclear. Most available data are limited to isolated case reports or small observational studies, making it difficult to draw definitive conclusions from them. Randomized clinical trials often exclude patients with pre-existing cognitive impairment and are underpowered to detect rare neuropsychiatric adverse events (AEs). Observational cohort studies are subject to confounding factors. In this context, pharmacovigilance methods, such as disproportionality analysis, are valuable tools for detecting safety signals in real-world populations, complementing traditional epidemiological approaches.¹¹

This review aimed to systematically evaluate the association between cardiovascular drugs and dementia risk using a two-phase approach: (1) retrospective disproportionality analysis of the FDA Adverse Event Reporting System (FAERS) data to identify signals of dementia associated with cardiovascular medications and (2) literature review of published case reports to characterize the clinical presentation, causality, and recovery patterns of suspected drug-induced dementia/cognitive impairment. We acknowledge the important methodological considerations for interpreting these data sources. FAERS reports of dementia may include both true adverse drug reactions and coincidental diagnoses in patients who developed dementia while taking these medications, as cardiovascular disease and dementia share common risk factors, and dementia has a multifactorial, insidious onset. However, disproportionate reporting patterns warrant investigation as potential safety signals. Case reports demonstrating clear temporal relationships between drug exposure and cognitive symptoms with improvement upon discontinuation can provide complementary evidence of potential causal associations. By integrating large-scale pharmacovigilance data with clinical observations, this study aimed to provide a comprehensive assessment of the potential neurocognitive risks associated with cardiovascular therapies.

Methods

Study Design

Phase 1: Disproportionality Analysis

Data Source and Database

Data were extracted from the FDA Adverse Event Reporting System (FAERS) accessed through OpenVigil 2.1 (https://openvigil.pharmacovigilance.org/), a web-based pharmacovigilance tool that provides pre-processed FAERS data.¹² The database included adverse event reports from Q4/2003 to Q4/2024. OpenVigil 2.1 has several advantages, including automated drug name mapping, duplicate detection algorithms, and standardized Medical Dictionary for Regulatory Activities (MedDRA) terminology.

Study Population

This study focused on older adult patients aged ≥60 years. No restrictions were placed on sex, allowing for a comprehensive assessment across both male and female patients.

Drug Selection and Classification

Cardiovascular medications were selected based on the Anatomical Therapeutic Chemical (ATC) Classification System and grouped into 14 major drug classes. A total of 97 individual drugs across these classes were included in this analysis (Table 1).

Table 1.

Cardiovascular Drug Classes and ATC Codes.

Drug class	ATC codes	Drugs
ACE inhibitors	C09AA	Benazepril, captopril, enalapril, fosinopril, lisinopril, perindopril, quinapril, ramipril
Angiotensin II receptor blockers	C09CA	Candesartan, eprosartan, irbesartan, losartan, olmesartan, telmisartan, valsartan
Antiarrhythmics	C01AA, C01BA, C01BB, C01BC, C01BD	Flecainide, propafenone, quinidine, disopyramide, procainamide, lidocaine, mexiletine, amiodarone, dofetilide, dronedarone, ibutilide, digoxin
Anticoagulants	B01AA, B01AB, B01AE, B01AF, B01AX	Apixaban, dabigatran, heparin, rivaroxaban, warfarin, fondaparinux, edoxaban, enoxaparin, dalteparin
Antiplatelet agents	B01AC	Aspirin, clopidogrel, dipyridamole, prasugrel, ticagrelor, ticlopidine
Beta blockers	C07AB, C07AA	Acebutolol, atenolol, bisoprolol, metoprolol, nadolol, propranolol, sotalol, nebivolol, timolol
Calcium channel blockers	C08CA, C08DA, C08DB	Amlodipine, diltiazem, felodipine, nifedipine, verapamil, isradipine, nicardipine, nisoldipine, clevidipine
Carbonic anhydrase inhibitors	S01EC01	Acetazolamide
Cholesterol-lowering medications	C10AA, C10AB, C10AC, C10AD, C10AX	Atorvastatin, fluvastatin, lovastatin, pitavastatin, pravastatin, rosuvastatin, simvastatin, niacin, ezetimibe, cholestyramine, colesevelam, colestipol, alirocumab, evolocumab, fenofibrate, gemfibrozil
Combined alpha and beta-blockers	C07AG	Carvedilol, labetalol
Loop diuretics	C03CA	Furosemide, torsemide, bumetanide
Potassium-sparing diuretics	C03DA, C03DB	Amiloride, spironolactone, eplerenone, triamterene
Thiazide and thiazide-like diuretics	C03AA, C03BA	Chlorothiazide, chlorthalidone, hydrochlorothiazide, indapamide, metolazone
Vasodilators	C01DA, C02DB, C02DC	Isosorbide dinitrate, isosorbide mononitrate, hydralazine, nitroglycerin, minoxidil, nitroprusside

Outcome Definition

Dementia-related adverse events were identified using the MedDRA Preferred Terms (PTs).

PT-Dementia

General dementia reports (PT: 10012259).

Dementia Alzheimer’s Type

Alzheimer’s disease-specific reports (PT: 10012218).

Vascular Dementia

Vascular etiology-specific reports (PT: 10046643).

Dementia With Lewy Bodies

Lewy body pathology-specific reports (PT: 10012242).

It is important to note that these MedDRA categories are not mutually exclusive; a single adverse event report may be coded under multiple dementia-related terms.

Statistical Measures

Disproportionality analysis serves to identify drug-event combinations that are reported with greater frequency than would be anticipated by chance. It is crucial to recognize that a signal does not establish causality; instead, it highlights statistical associations that necessitate further evaluation through rigorous pharmacoepidemiological studies. For each drug-dementia pair, the following disproportionality measures were calculated.

2 × 2 Contingency Table

	Dementia	Other AEs	Total
Target Drug	a	b	a + b
Other Drugs	c	d	c + d
Total	a + c	b + d	a + b + c + d (N)

A 2*2 contingency table captures the incidences of dementia (targeted event) and all other adverse events for the target drug – cardiovascular drugs, and all other drugs in the database to calculate PRR.

Proportional Reporting Ratio (PRR)

The PRR compares the proportion of dementia reports for the target drug to the proportion of dementia reports for all other drugs. A PRR >1 indicates dementia is reported more frequently with the target drug.¹³

PRR = [a / (a + b)] / [c / (c + d)]

95 % CI = \exp [\ln (PRR) \pm 1.96 \times (\sqrt{1 / a} + 1 / c - 1 / (a + b) - 1 / (c + d))]

Reporting Odds Ratio (ROR)

The ROR estimates the odds of reporting dementia versus other adverse events for the target drug compared with all other drugs. This is analogous to the odds ratio used in epidemiological studies.¹⁴

ROR = (a \times d) / (b \times c)

95 % CI = \exp [\ln (ROR) \pm 1.96 \times (\sqrt{1 / a} + 1 / b + 1 / c + 1 / d)]

Chi-Square With Yates’ Correction

The chi-square test evaluates whether the association between the drug and dementia is statistically significant, with Yates’ correction applied to reduce bias in small sample sizes.

χ^{2} = N \times {(| ad - bc | - N / 2)}^{2} / [(a + b) (c + d) (a + c) (b + d)]

Signal Detection Criteria

A drug-dementia combination was considered a positive signal when ALL three criteria were met: Total Reports ≥2; Chi-square (Yates correction) ≥ 4 and Proportional Reporting Ratio (PRR) ≥ 2.0.¹³

Statistical Analysis

Additional statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA). Statistical significance was set at P < 0.05 for all analyses.

Between-Class Analysis

The chi-square test for independence was used to evaluate the associations between drug classes and signal detection. Odds ratios with 95% confidence intervals were used to quantify the signal detection likelihood by drug class.

Between-Dementia-Type Analysis

Kruskal-Wallis test was used to compare PRR distributions across the three dementia types. Spearman’s rank correlation to assess correlation of signal strength across dementia types.

Phase 2: Literature Review of Case Reports

Search Strategy

A comprehensive literature search was conducted using PubMed/MEDLINE to identify relevant studies published between January 2000 and December 2024. We selected this timeframe to reflect the utilization of recent cardiovascular medications and to align with the availability of the FAERS database. Our emphasis was on case reports and case series rather than larger epidemiological studies, as these provide detailed insights into the timing of drug exposure and onset of symptoms, as well as information on symptom reversibility following drug discontinuation. These data are essential for identifying the acute cognitive effects induced by drugs, which might be overlooked or diluted in extensive observational studies evaluating long-term dementia risk. The search was restricted to English-language publications to ensure a comprehensive understanding and accurate data extraction. The search strategy employed a combination of Medical Subject Headings (MeSH) terms and free-text keywords related to cardiovascular medications and cognitive adverse events, including various terms for dementia, cognitive impairment, and memory disorders. The complete search strategy with specific terms, Boolean operators, and search filters is provided in the supplemental materials.

Study Selection

The retrieved citations were systematically screened based on the title and abstract content, followed by a full-text review of potentially relevant articles to determine their final inclusion in the review.

Data Extraction

Data extraction was performed using a standardized approach to ensure the consistent capture of relevant information across all included studies. The study characteristics recorded included the first author, publication year, country of origin, and study design (case report vs case series). Patient-level data encompassed demographic information, such as age and sex, along with relevant clinical background, including cardiovascular comorbidities, diabetes status, and baseline cognitive function when reported.

Drug exposure details were comprehensively documented, including the specific medication name and therapeutic class, prescribed daily dosage, duration of treatment prior to the emergence of cognitive symptoms, and the clinical indication for which the medication was prescribed. The nature and characteristics of the reported cognitive events were systematically captured, with particular attention to the type and severity of cognitive impairment described.

Causality assessment data were extracted when available, including evaluation of temporal relationships between drug initiation and cognitive symptoms, documentation of drug discontinuation (dechallenge) and associated outcomes, any rechallenge events and their results, and the authors’ overall causality determination. Clinical outcome information focused on the reversibility of cognitive symptoms, time course of recovery where applicable, availability of long-term follow-up data, and the final cognitive status reported in each case.

Ethical Considerations

This study used publicly available de-identified data from the FAERS database and published literature. No patient-specific information was accessed or analyzed in this study. The disproportionality analysis used aggregate data from OpenVigil, and the systematic review included only published case reports. As a retrospective analysis of anonymized secondary data, this study was exempt from institutional review board approval.

Results

Overview of Signal Detection

A total of 97 cardiovascular drugs from 14 drug classes were analyzed in this study. Signals were detected for 13 drugs (13.4%) for dementia, 16 drugs (16.5%) for Alzheimer’s type dementia, 18 drugs (18.6%) for vascular dementia, and six drugs (6.2%) for Dementia with Lewy bodies (Table 2).

Table 2.

Summary of Signal Detection by Dementia Type.

Dementia type	Drugs with signals (%)	Drug classes with ≥1 signal
PT-dementia	13/97 (13.4)	8/14
Alzheimer’s type	16/97 (16.5)	7/14
Vascular dementia	18/97 (18.6)	8/14
Dementia with lewy bodies	6/97 (6.2)	5/14

Detailed Signal Detection Results

Signal detection analysis identified potential associations between cardiovascular medications and dementia-related adverse events in multiple dementia subtypes. Detailed results, including drug names, total reports, chi-square statistics, and proportional reporting ratios (PRR) with 95% confidence intervals, are presented in the supplementary data (Supplemental Tables S1-S4). Notable signals were detected for anticoagulants (apixaban, dabigatran), angiotensin II receptor blockers (valsartan, telmisartan), and cholesterol-lowering medications (atorvastatin, rosuvastatin) across different dementia phenotypes, with PRR values ranging from 2.02 to 9.81, suggesting that these drug classes warrant further pharmacoepidemiological investigation.

Statistical Analysis of Signals

Drug Class Association With Signal Detection

Chi-square analysis was used to evaluate the association between drug classes and signal detection across all dementia types (Table 3). Among the 97 drugs analyzed, 38 (39.2%) were associated with signal detection in at least one type of dementia diagnosis. The highest signal rates were observed among ACE inhibitors (75.0%, n = 6/8) and angiotensin II receptor blockers (71.4%, n = 5/7), whereas several drug classes demonstrated no signals.

Table 3.

Association Between Drug Class and Signal Detection (Between-Class Analysis).

Drug class	Signal present/Total	Signal rate (%)	Standardized residual^a
ACE inhibitors	6/8	75.0	1.6
ARBs	5/7	71.4	1.4
Antiplatelet agents	3/6	50.0	0.4
Vasodilators	3/6	50.0	0.4
Anticoagulants	4/9	44.4	0.3
Beta blockers	4/9	44.4	0.3
Cholesterol-lowering medications	6/16	37.5	−0.1
Loop diuretics	1/3	33.3	−0.2
Antiarrhythmics	3/12	25.0	−0.8
Calcium channel blockers	2/9	22.2	−0.8
Thiazide/Thiazide-like diuretics	15	20.0	−0.7
Total	38/97	39.2	—

χ² = 15.60, df = 13, P = 0.27.

^aStandardized residuals are shown for Signal Present category. Bold values indicate residuals >1.5.

The chi-square test revealed no statistically significant association between the drug class and overall signal detection (χ² = 15.60, df = 13, P = 0.27). Examination of standardized residuals identified ACE inhibitors (residual = 1.6) and ARBs (residual = 1.4) as having higher-than-expected signal rates, although neither exceeded the threshold for statistical significance (±1.96).

Between-Dementia-Type Analysis: Signal Patterns Across Dementia Categories

An analysis of 388 drug-dementia combinations revealed an overall signal rate of 13.7% (53/388) across all dementia types (Table 4). ARBs had the highest signal rate (25.0%), followed by antiplatelet agents and vasodilators (both 20.8%). Chi-square analysis indicated no significant association between drug class and signal detection when considering all dementia types (χ² = 12.5, df = 13, P = 0.49).

Table 4.

Signal Detection Rates by Drug Class Across Dementia Types.

Drug class	Signal present/Total	Signal rate (%)
ARBs	7/28	25.0
Antiplatelet agents	5/24	20.8
Vasodilators	5/24	20.8
ACE inhibitors	6/32	18.8
Beta blockers	6/36	16.7
Anticoagulants	5/36	13.9
Cholesterol-lowering medications	8/64	12.5
Calcium channel blockers	4/36	11.1
Thiazide/Thiazide-like diuretics	2/20	10.0
Antiarrhythmics	4/48	8.3
Loop diuretics	1/12	8.3
Total	53 (13.7)	13.7

χ² = 12.5, df = 13, P = 0.49.

PRR Distribution Comparison Across Dementia Types

Kruskal-Wallis testing revealed highly significant differences in the proportional reporting ratio distributions across the four dementia types (H = 39.37, df = 3, P < 0.001) (Table 5). General dementia and vascular dementia showed the highest signal detection patterns (mean ranks = 218 and 216, respectively), followed by Alzheimer’s disease (mean rank = 210).

Table 5.

Comparison of Signal Detection Patterns Across Dementia Types.

Dementia type	N	Mean rank^a	Signal detection pattern
General dementia	97	218	Highest signal detection
Vascular dementia	97	216	High signal detection
Alzheimer’s disease	97	210	Moderate signal detection
Dementia with lewy bodies	97	134	Significantly lower signal detection

Kruskal-Wallis H = 39.37, df = 3, P < 0.001.

^aHigher mean rank indicates greater signal detection.

Cross-Dementia Signal Correlation Analysis

Spearman’s rank correlation analysis was used to assess the similarity of signal strength patterns between the dementia types (Table 6). The strongest correlation was observed between general dementia and Alzheimer’s disease (ρ = 0.48, P < 0.001), indicating moderate similarity in signal detection patterns. This correlation may be attributable to two possible factors: (1) Alzheimer’s disease is the most prevalent dementia subtype and (2) the MedDRA coding categories are not mutually exclusive, and reports may be coded with both general “dementia” and specific “Alzheimer’s disease” terms. Moderate correlations were also found between vascular dementia and dementia with Lewy bodies (ρ = 0.40, P < 0.001).

Table 6.

Cross-Dementia Signal Correlation Matrix (Spearman’s ρ).

	General dementia	Alzheimer’s	Vascular	Lewy bodies
General dementia	1.00	0.48***	0.15	0.33**
Alzheimer’s disease	0.48***	1.00	0.24*	0.25*
Vascular dementia	0.15	0.24*	1.00	0.40***
Dementia with lewy bodies	0.33**	0.25*	0.40***	1.00

*P < 0.05, **P < 0.01, ***P < 0.001.

Notably, the correlation between general dementia and vascular dementia was weak and non-significant (ρ = 0.15, P = 0.16), while other correlations were significant but modest (range: 0.24-0.40), suggesting that while some commonalities exist in drug safety profiles across dementia types, each dementia category maintains distinct pharmacovigilance characteristics.

Multivariable Analysis: Predictors of Signal Detection

Multivariable logistic regression analysis was used to examine the independent effects of drug class and dementia type on the signal detection (Table 7). The model included all 14 drug classes and four dementia types as categorical predictors, with vasodilators and dementia with Lewy bodies as the reference categories.

Table 7.

Multivariable Logistic Regression Analysis: Predictors of Signal Detection.

Variable	Odds ratio	95% CI	P-value
Drug class			0.90
ACE inhibitors	0.87	0.23-3.35	0.84
ARBs	1.28	0.34-4.80	0.72
Antiarrhythmics	0.34	0.08-1.42	0.14
Anticoagulants	0.61	0.15-2.41	0.48
Antiplatelet agents	1.00	0.24-4.11	1.00
Beta blockers	0.75	0.20-2.87	0.68
Calcium channel blockers	0.47	0.11-1.99	0.30
Cholesterol-lowering medications	0.54	0.15-1.87	0.33
Loop diuretics	0.34	0.03-3.34	0.35
Thiazide/Thiazide-like diuretics	0.41	0.07-2.46	0.33
Other drug classes^a	<0.001^b	—	—
Dementia type			0.07
General dementia	2.40	0.86-6.66	0.09
Alzheimer’s disease	3.09	1.14-8.36	0.03*
Vascular dementia	3.58	1.34-9.58	0.01*

*P < 0.05.

Reference categories: Vasodilators (drug class), Dementia with Lewy bodies (dementia type).

Total categories: 14 drug classes, 4 dementia types.

Model χ² = 14.01, df = 16, P = 0.60.

^aIncludes Carbonic Anhydrase Inhibitors, Combined alpha/beta-blockers, and Potassium-Sparing Diuretics.

^bPerfect separation - no signals observed.

Drug class was not a significant predictor of signal detection when controlling for dementia type (overall Wald χ² = 7.08, df = 13, P = 0.90). All drug classes demonstrated wide confidence intervals spanning the null value, indicating substantial uncertainty in effect estimates. The three drug classes exhibited perfect separation with zero observed signals.

Dementia type emerged as a marginally significant predictor of signal detection (overall Wald χ² = 6.93, df = 3, P = 0.07). Patients with Alzheimer’s disease demonstrated significantly higher odds of signal detection than those with dementia with Lewy bodies (OR = 3.09, 95% CI: 1.14-8.36, P = 0.03). Similarly, patients with vascular dementia showed significantly elevated odds relative to the dementia with Lewy bodies reference group (OR = 3.58, 95% CI: 1.34-9.58, P = 0.01).

Literature Review of Case Reports

Study Selection and Characteristics

The systematic literature search identified 213 potentially relevant studies in PubMed. After title and abstract screening, 45 studies were subjected to a full-text review. Following the application of the inclusion and exclusion criteria, eight studies reporting individual case reports of drug-induced cognitive impairment met our criteria for inclusion (Table 8). These eight studies included 13 individual case reports published between 1990 and 2021.

Table 8.

Characteristics of Included Studies.

First author	Year	Country	Cases (n)
Rogers¹⁵	1990	United Kingdom	3
Fisher¹⁶	1992	United States	1
Orsi¹⁷	2001	United States	1
King¹⁸	2003	United States	2
Galatti¹⁹	2006	Italy	1
Padala²⁰	2006	United States	1
Suraweera²¹	2016	Sri Lanka	2
Porter²²	2021	United Kingdom	2

Patient Demographics and Case Details

The 13 individual case reports included patients aged 32-81 years (mean 67.2 ± 15.8 years), with female predominance (8/13). Most patients had normal baseline cognitive function (11/13), while only two had mild pre-existing cognitive impairment. Cardiovascular comorbidities were common and were present in over half of the patients.

Three major drug classes have been implicated in the development of cognitive impairment. Statins accounted for the largest proportion of dementia cases (7/13), with simvastatin being the most frequently reported drug (30.8% of all cases). Beta-Blockers accounted for 4/13 of the cases and were equally distributed between propranolol and atenolol. Direct oral anticoagulants (DOACs) comprised 2/13 of the cases and included apixaban and rivaroxaban.

Cognitive Manifestations and Temporal Patterns

Short-term memory impairment was the predominant clinical feature, affecting 10/13 patients. This manifested as difficulty in retaining new information, forgetting recent conversations, misplacing items, and an inability to recall recent events. Other significant cognitive domains affected included attention/concentration difficulties (5/13) and executive dysfunction (4/13), which presented as impaired decision-making, planning difficulties, and reduced problem-solving abilities. Daily functioning was compromised in over half of the patients (8/13), indicating a clinically meaningful cognitive decline.

The temporal relationship between drug exposure and symptom onset varied considerably. Early onset cases (≤1 month) comprised 6/13 of the reports, while others experienced delayed symptom development, with some cases not manifesting until after three months of therapy (Table 9).

Table 9.

Patient Demographics and Baseline Characteristics.

Characteristic	n or mean ± SD
Age (years)
Mean ± SD	67.2 ± 15.8
Range	32-81
Sex
Female	8/13
Male	5/13
Baseline cognitive status
Normal cognition	11/13
Pre-existing mild impairment	2/13
Comorbidities
Cardiovascular disease	7/13
Hypertension	6/13
Diabetes mellitus	2/13
Psychiatric conditions	3/13
Beta-blockers
Propranolol	2/13
Atenolol	2/13
Statins
Simvastatin	4/13
Atorvastatin	2/13
Rosuvastatin	1/13
Direct oral anticoagulants
Apixaban	1/13
Rivaroxaban	1/13
Cognitive domain affected
Memory (short-term/episodic)^a	10/13
Memory (long-term)	3/13
Executive function	4/13
Attention/concentration	5/13
Language/word-finding	4/13
Functional impact
Behavioural changes	3/13
Daily functioning impairment	8/13
Time period
≤1 week	2/13
1 week - 1 month	4/13
1–3 months	4/13
>3 months	3/13

^aShort-term memory impairment was the predominant manifestation.

The detailed drug exposure characteristics are presented in Supplemental Table S5. Daily dosages varied considerably within the drug classes, with statin doses ranging from 10-40 mg daily. Notably, two cases demonstrated dose-dependent effects, with cognitive symptoms appearing after dose escalation rather than after the initial drug exposure. The duration of drug exposure before cognitive symptom onset showed a bimodal distribution, with early onset cases (≤1 month) comprising 6/13 of the reports.

Causality Assessment

Temporal relationships between drug initiation and cognitive symptom onset were established in all 13 patients (100%) (Supplemental Table S6). Most cases demonstrated probable or definite causality, with only two cases classified as having a possible causality.

Dechallenge was performed in all cases, with drug discontinuation leading to cognitive improvement in 12/13 patients. Rechallenge was attempted in only two cases, both involving statin therapy. In both instances, rechallenge resulted in the recurrence of cognitive symptoms, providing strong evidence for drug causality.

Clinical Outcomes and Recovery Patterns

The reversibility of cognitive symptoms following drug discontinuation was consistent across the drug classes (Supplemental Table S7). Complete cognitive recovery occurred in 8/13 patients, while partial recovery was documented in 4/13 patients. Only one patient showed no improvement after discontinuation of the drug.

The time course of recovery varied according to the drug class and the individual patient factors. Recovery within one week occurred in 2/13 cases, both involving DOACs. Most patients (6/13) demonstrated recovery within 1-4 weeks, while 3/13 required 1-3 months for full recovery.

An exploratory analysis was performed to examine whether the length of treatment prior to the onset of symptoms affected recovery outcomes. The cases were divided into groups based on the duration of exposure: short (≤1 month, n = 5), medium (1–3 months, n = 3), and long (>3 months, n = 5). The rates of complete recovery were similar across these categories. Notably, patients with prolonged exposure, including one who received treatment for 3 years and another for many years, experienced recovery periods comparable to those with shorter exposure durations (days to weeks). This indicates that the potential for recovery is primarily determined by the type of drug and individual characteristics rather than the overall duration of exposure.

Drug Class-Specific Analysis

Table 10 presents a comparative analysis of the cognitive impairment patterns across the three drug classes. Each drug class demonstrated distinct characteristics in terms of onset timing, severity, and recovery patterns.

Table 10.

Drug Class-Specific Cognitive Impairment Patterns.

Characteristic	Beta-blockers (n = 4)	Statins (n = 7)	DOACs (n = 2)
Onset pattern
Early (≤1 month)	2 (50)	3 (42.9)	2 (100)
Delayed (>3 months)	1 (25)	2 (28.6)	0 (0)
Severity
Mild-moderate	4 (100)	7 (100)	0 (0)
Severe (hospitalization)	0 (0)	0 (0)	2 (100)
Recovery rate	4 (100)	6 (85.7)	2 (100)
Time to recovery
≤4 weeks	4 (100)	4 (66.7)	2 (100)
>4 weeks	0 (0)	2 (33.3)	0 (0)

Statin-Associated Cognitive Impairment

Seven patients with statin-associated cognitive impairment exhibited characteristic patterns. The onset of symptoms showed a bimodal distribution, with some patients developing symptoms within days to weeks (n = 3), while others experienced delayed onset after months of therapy (n = 4). Two cases clearly demonstrated dose-dependent effects, with cognitive symptoms appearing after dose escalation rather than after the initial drug exposure.

Beta-Blocker-Associated Cognitive Impairment

Four cases of beta-blocker-associated cognitive impairment revealed patterns related to the lipophilicity of these drugs. Cases involving propranolol, a highly lipophilic beta-blocker, demonstrated more severe cognitive effects than those involving atenolol, a hydrophilic beta-blocker. Recovery occurred in all beta-blocker cases (100%) within 1-6 weeks of drug discontinuation.

DOAC-Associated Cognitive Impairment

Two cases involving direct oral anticoagulants (apixaban and rivaroxaban) presented with particularly severe cognitive and behavioural manifestations that required psychiatric hospitalization. Recovery was rapid and complete in both cases, occurring within 2-3 weeks of drug discontinuation.

Discussion

This review provides a comprehensive evaluation of the potential association between cardiovascular medications and dementia-related adverse effects in older adults. Our findings underscore the efficacy of pharmacovigilance techniques in identifying potential neurocognitive side effects of these drugs. Disproportionality analysis is valuable for detecting rare or delayed safety signals that may not be readily apparent in randomized trials. The study suggests that certain cardiovascular drugs, particularly ACE inhibitors, angiotensin II receptor blockers (ARBs), statins, and direct oral anticoagulants (DOACs), may be disproportionately associated with cognitive decline and dementia. By analyzing FAERS data in conjunction with published case reports, we obtained novel insights into the complex relationship between cardiovascular drugs and dementia. We identified dementia-related signals in nearly 40% of the cardiovascular drugs evaluated, with ACE inhibitors, ARBs, and antiplatelets exhibiting the highest proportional reporting ratios, whereas calcium channel blockers, diuretics, and antiarrhythmics were less frequently associated with dementia. Supporting these findings, the case review indicated that statins, beta-blockers, and direct oral anticoagulants were most frequently linked to cognitive side effects, with short-term memory loss being the most prevalent. Notably, 92% of patients experienced improvement after discontinuing the medication, and some exhibited a reproducible relapse upon rechallenge, providing strong evidence of a causal relationship. Notably, dementia with Lewy bodies demonstrated a significantly lower signal detection pattern (mean rank = 134) compared to other dementia subtypes. While DLB is one of the most common causes of dementia after Alzheimer’s disease (AD) and vascular dementia, it is rarely reported in pharmacovigilance databases. This discrepancy may reflect the known underdiagnosis of DLB in clinical practice, where its complex symptomatology (cognitive fluctuations, visual hallucinations, and parkinsonism) is often attributed to other conditions. The lower signal rate for DLB in FAERS likely reflects reporting patterns and diagnostic challenges rather than a true difference in drug-induced risk, as adverse event reporters may be less likely to recognize or code drug-related cognitive impairment as being DLB-specific.

These results both support and differ from the increasing epidemiological evidence indicating that cardiovascular medications may reduce the risk of dementia. A recent case–control study based on a Swedish registry revealed that prolonged use (5-9 and ≥10 years) of diuretics, beta-blockers, calcium channel blockers, RAS-acting agents, lipid-lowering drugs, and oral anticoagulants significantly lowered the incidence of dementia, with odds ratios between 0.75 and 0.96 compared to those who did not use these drugs. Conversely, short-term use of all cardiovascular drug classes (1-4 years) was linked to a heightened risk (ORs, 1.13-1.30), and antiplatelet therapy, regardless of duration, was associated with an increased risk compared to non-users (ORs, 1.13-1.25).²³ Similarly, an umbrella review of meta-analyses found consistent links between cardiovascular risk factors, such as coronary heart disease, heart failure, atrial fibrillation, and midlife hypertension and dementia.⁷ Overall, these findings underscore that cardiovascular pharmacotherapy can act as both protective and potentially harmful factors in cognitive health.

Statins are an example of this duality. Large-scale meta-analyses involving over seven million participants have indicated that statin use significantly reduces the risk of dementia compared to non-users (hazard ratio [HR] 0.86; 95% confidence interval [CI]: 0.82 to 0.91; P < 0.001). Statin use was also associated with a reduced risk of Alzheimer’s disease (HR 0.82; 95% CI: 0.74-0.90; P < 0.001) and vascular dementia (HR 0.89; 95% CI: 0.77-1.02; P = 0.093). Furthermore, individuals exposed to statin therapy for more than three years demonstrated a statistically significant 63% reduction in dementia risk (HR 0.3685; 95% CI: 0.2979-0.4558; P < 0.001).²⁴ However, cognitive impairment associated with statin use has been documented in case studies, observational research, and randomized controlled trials (RCTs).^25-27 In the UK Biobank cohort, overall statin use was linked to an increased risk of Alzheimer’s disease and dementia; nevertheless, subgroup analyses suggested a potential protective effect in APOE ε4 carriers, particularly among males, underscoring the complex and interaction-dependent nature of the effects of statins on cognitive outcomes.²⁸ Therefore, the effects of statins on both cognitive impairment and AD onset remain inconsistent and warrant further investigation.

These adverse effects, recognized by regulatory authorities, are rare yet clinically significant, with factors such as lipophilicity, high dosage, and individual susceptibility contributing to their occurrence.²⁹ Beta-Blockers demonstrate a similar pattern; although they are advantageous for cardiovascular conditions, observational studies suggest that they may double the risk of vascular dementia in the general older adult population.³⁰ In relation to DOACs, our literature review identified sudden behavioural and cognitive decline, consistent with previous findings that associate these agents with neuropsychiatric symptoms, potentially attributable to mitochondrial toxicity.²²

Meta-analyses investigating various categories of antihypertensive medications have found no correlation between the use of diuretics and the incidence of dementia. Conversely, the findings for beta-blockers and ACE inhibitors were inconsistent. There is some evidence suggesting a potential protective effect of calcium channel blockers and angiotensin receptor blockers.⁸ Ding et al³¹ did not identify any specific subtype of antihypertensive drugs as superior in relation to dementia prevention.

The apparent contradiction between ensuring safety at the population level and causing harm at the individual level can be elucidated by methodological differences. While registry and cohort studies typically reflect average outcomes, often emphasizing reductions in vascular risk, pharmacovigilance databases and case reports focus on rare but serious idiosyncratic incidents. Although spontaneous reports are constrained by reporting biases, the absence of denominators, and imprecise diagnostic coding, consistent signals across various drug classes and their biological plausibility bolster our confidence in these findings. The bimodal temporal pattern of symptom onset observed in our case studies (early: ≤1 month vs delayed: >3 months post-initiation) suggests that different underlying mechanisms may be involved in drug-induced cognitive impairment. Mechanistic evidence supports pathways that are both beneficial and harmful.^23,24,29

The biological processes that lead to acute cognitive impairment are not fully understood and are likely to differ depending on the drug class involved. Possible mechanisms include effects on cerebral blood flow, anticholinergic properties, and lipophilic drug ability to cross the blood-brain barrier. However, detection bias is a significant factor to consider, as patients who are newly prescribed cardiovascular drugs often receive more frequent clinical evaluations, which may reveal pre-existing cognitive issues. The timing of medication initiation and identification of cognitive symptoms do not necessarily imply a causal relationship, a distinction that cannot be conclusively determined using spontaneous reporting data.

Although our findings are preliminary and require further validation through prospective studies, they have significant implications for clinical practice. The identification of dementia-related reported and documented cases of reversible cognitive impairment highlights the necessity for clinical awareness of medication-induced cognitive decline, particularly in older adults who take multiple medications. Vigilant monitoring of cognitive changes following the initiation of a new drug and reduction of medication when cognitive symptoms manifest can substantially enhance the quality of life. These findings advocate for the integration of regular cognitive assessments into cardiovascular care and underscore the importance of personalized risk-benefit evaluations, especially for patients with pre-existing cognitive issues or those on multiple cardiovascular medications. Future research should explore the effects specific to various dementia subtypes and identify the patient populations at risk. Prospective pharmacoepidemiological studies, pragmatic deprescribing trials, and mechanistic research incorporating biomarkers of vascular damage and amyloid–tau pathology are essential for elucidating causal pathways. Importantly, conducting research in low- and middle-income countries, where the prevalence of cardiovascular risk factors and dementia is increasing rapidly, is crucial for maintaining global relevance.⁶

The present study highlights the dual role of cardiovascular pharmacotherapy in the management of dementia. On the one hand, prolonged use of antihypertensives, statins, and anticoagulants may mitigate the risk of dementia onset. Conversely, certain medications, particularly lipophilic statins, beta-blockers, and DOACs, may precipitate temporary cognitive decline in susceptible patients. The challenge for clinicians and researchers is to balance these beneficial and adverse effects, underscoring the necessity for continuous vigilance, judicious prescribing, and targeted research.

Limitations

This study had several notable limitations that must be considered when interpreting its results. First, the FAERS database is susceptible to substantial reporting bias, as health care providers and patients are more likely to report severe or unexpected incidents, potentially exaggerating the association between certain drugs and dementia incidence. The absence of a denominator renders it impossible to determine the absolute risk of cognitive impairment associated with specific drugs. Confounding by indication represents a major limitation, as patients on cardiovascular medications frequently have underlying cardiovascular conditions that independently increase their risk of dementia. This complicates the differentiation between the cognitive effects attributable to drugs and those arising from underlying disease processes. Our analysis did not adequately account for potential drug-drug interactions that could contribute to cognitive impairment and other confounding factors. While FAERS reports identify suspect drugs, the reporter’s attribution of causality to a specific medication cannot be independently validated, particularly in patients taking multiple concurrent medications. The disproportionality analysis method cannot establish causality or control for confounding exposures. The case report literature review is constrained by publication bias, as cases with positive outcomes or clear causality are more likely to be published than those with uncertain results. The limited number of case reports (n = 13) restricts the generalizability of our findings and precludes a meaningful subgroup analysis. Taken together, these limitations mean that the associations observed should be interpreted cautiously and regarded as hypothesis generating.

Conclusion

This comprehensive pharmacovigilance study elucidates the potential neurocognitive risks associated with cardiovascular medications in the elderly population. Through a systematic analysis of FAERS data and clinical case reports, we determined that approximately 40% of the examined cardiovascular drugs exhibited a disproportionate number of dementia-related adverse event reports. ACE inhibitors, angiotensin receptor blockers, and antiplatelet agents had the highest signal detection rates. The analysis of case reports revealed that statins, beta-blockers, and direct oral anticoagulants were the most frequently associated with reversible cognitive impairment. While cardiovascular drugs remain essential for reducing morbidity and mortality, clinicians must remain vigilant regarding potential cognitive side effects. These hypothesis-generating findings suggest that increased awareness of the potential cognitive effects may be appropriate when initiating cardiovascular therapies in older adults. While these findings do not support changes in prescribing practices, they indicate the potential value of remaining alert to cognitive changes after drug initiation and promptly evaluating symptoms as they arise. However, these associations require confirmation through prospective pharmacoepidemiological studies before any modifications to clinical guidelines can be recommended. Patient-centered prescribing, careful monitoring, and deprescribing when appropriate are crucial for safeguarding both cardiovascular and cognitive health. These findings should be viewed as preliminary signals requiring confirmation in prospective, well-designed clinical studies before influencing the clinical guidelines.

Supplemental Material

Supplemental material - Cardiovascular Medications and Dementia Risk in Older Adults: A Literature Review and Disproportionality Analysis Using OpenVigil FAERS Data

Supplemental material for Cardiovascular Medications and Dementia Risk in Older Adults: A Literature Review and Disproportionality Analysis Using OpenVigil FAERS Data by Jehath Syed, PhD, Amruta Potdar, M. Pharm and Sri Harsha Chalasani, PhD in Journal of Geriatric Psychiatry and Neurology

Supplemental Material

Supplemental material - Cardiovascular Medications and Dementia Risk in Older Adults: A Literature Review and Disproportionality Analysis Using OpenVigil FAERS Data

Footnotes

Author Contributions

Jehath Syed: Conceptualization, Data curation, Investigation, Methodology, Writing- original draft; Amruta Potdar: Data curation, Investigation, Writing- original draft; Sri Harsha Chalasani: Conceptualization, Supervision and Validation, Writing-review & editing

Declaration of Conflicting Interests

Ms. Amruta Potdar is employed by Eli Lilly and Company. Her contribution to this manuscript was independent and not related to her employment.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jehath Syed

Supplemental Material

Supplemental material for this article is available online.

References

Liu

Deng

Gong

Chen

. Global burden of alzheimer's disease and other dementias in adults aged 65 years and over, and health inequality related to SDI, 1990-2021: analysis of data from GBD 2021. BMC Public Health. 2025;25(1):1256. doi:10.1186/s12889-025-22378-z. Published 2025 Apr 3.

Cermakova

Fereshtehnejad

Johnell

Winblad

Eriksdotter

Religa

. Cardiovascular medication burden in dementia disorders: a nationwide study of 19,743 dementia patients in the Swedish dementia registry. Alzheimers Res Ther. 2014;6(3):34. doi:10.1186/alzrt264. Published 2014 Jun 16.

GBD 2019 . Dementia forecasting collaborators. Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the global burden of disease study 2019. Lancet Public Health. 2022;7(2):e105-e125. doi:10.1016/S2468-2667(21)00249-8

Alzheimer's Disease International . Dementia statistics. Alzheimer's disease International. Published 2020. Accessed August 24, 2025.https://www.alzint.org/about/dementia-facts-figures/dementia-statistics/

Alzheimer's Disease International . World alzheimer report 2015: the global impact of dementia: an analysis of prevalence, incidence, cost and trends. Alzheimer's Disease International. Published August 2015. Accessed August 24, 2025.https://www.alzint.org/u/WorldAlzheimerReport2015.pdf

Liu

Geng

. A systematic analysis for disease burden, risk factors, and trend projection of alzheimer's disease and other dementias in China and globally. PLoS One. 2025;20(5):e0322574. doi:10.1371/journal.pone.0322574. Published 2025 May 7.

Brain

Greene

Tang

EYH

, et al. Cardiovascular disease, associated risk factors, and risk of dementia: an umbrella review of meta-analyses. Front Epidemiol. 2023;3:1095236. doi:10.3389/fepid.2023.1095236. Published 2023 Feb 9.

Livingston

Huntley

Sommerlad

, et al. Dementia prevention, intervention, and care: 2020 report of the lancet commission. Lancet. 2020;396(10248):413-446. doi:10.1016/S0140-6736(20)30367-6-

Marvanova

. Drug-induced cognitive impairment: effect of cardiovascular agents. Ment Health Clin. 2016;6(4):201-206. doi:10.9740/mhc.2016.07.201. Published 2016 Jun 29.

10.

Reimers

Odin

Ljung

. Drug-induced cognitive impairment. Drug Saf. 2025;48(4):339-361. doi:10.1007/s40264-024-01506-5

11.

Sakaeda

Tamon

Kadoyama

Okuno

. Data mining of the public version of the FDA adverse event reporting system. Int J Med Sci. 2013;10(7):796-803. doi:10.7150/ijms.6048. Published 2013 Apr 25.

12.

OpenVigil . Open tools for data-mining and analysis of pharmacovigilance data. Accessed April 10, 2025.https://openvigil.pharmacovigilance.org/

13.

Evans

Waller

Davis

. Use of proportional reporting ratios (PRRs) for signal generation from spontaneous adverse drug reaction reports. Pharmacoepidemiol Drug Saf. 2001;10(6):483-486. doi:10.1002/pds.677

14.

van Puijenbroek

Bate

Leufkens

Lindquist

Orre

Egberts

. A comparison of measures of disproportionality for signal detection in spontaneous reporting systems for adverse drug reactions. Pharmacoepidemiol Drug Saf. 2002;11(1):3-10. doi:10.1002/pds.668

15.

Rogers

Bowman

. Cognitive impairment associated with beta-blockade in the elderly. Postgrad Med J. 1990;66(782):1050-1052. doi:10.1136/pgmj.66.782.1050

16.

Fisher

. Amnestic syndrome associated with propranolol toxicity: a case report. Clin Neuropharmacol. 1992;15(5):397-403. doi:10.1097/00002826-199210000-00008

17.

Orsi

Sherman

Woldeselassie

. Simvastatin-associated memory loss. Pharmacotherapy. 2001;21(6):767-769. doi:10.1592/phco.21.7.767.34577

18.

King

Wilburn

Wofford

Harrell

Lindley

Jones

. Cognitive impairment associated with atorvastatin and simvastatin. Pharmacotherapy. 2003;23(12):1663-1667. doi:10.1592/phco.23.15.1663.31953

19.

Galatti

Polimeni

Salvo

Romani

Sessa

Spina

. Short-term memory loss associated with rosuvastatin. Pharmacotherapy. 2006;26(8):1190-1192. doi:10.1592/phco.26.8.1190

20.

Padala

Potter

. Simvastatin-induced decline in cognition. Ann Pharmacother. 2006;40(10):1880-1883. doi:10.1345/aph.1H014

21.

Suraweera

de Silva

Hanwella

. Simvastatin-induced cognitive dysfunction: two case reports. J Med Case Rep. 2016;10:83. doi:10.1186/s13256-016-0877-8. Published 2016 Apr 5.

22.

Porter

KMF

Hargreaves

De Souza

Goddard

. Treatment with the direct oral anticoagulants (DOACs) apixaban and rivaroxaban associated with significant worsening of behavioural and psychological symptoms of dementia (BPSD). BMJ Case Rep. 2021;14(3):e240059. doi:10.1136/bcr-2020-240059. Published 2021 Mar 8.

23.

Ding

Wennberg

Engström

Modig

. Use of common cardiovascular disease drugs and risk of dementia: a case-control study in Swedish national register data. Alzheimer's Dement. 2025;21(1):e14389. doi:10.1002/alz.14389

24.

Westphal Filho

Moss Lopes

Menegaz de Almeida

, et al. Statin use and dementia risk: a systematic review and updated meta-analysis. Alzheimer's Dement. 2025;11(1):e70039. doi:10.1002/trc2.70039. Published 2025 Jan 16.

25.

Bettermann

Arnold

Williamson

, et al. Statins, risk of dementia, and cognitive function: secondary analysis of the ginkgo evaluation of memory study. J Stroke Cerebrovasc Dis. 2012;21(6):436-444. doi:10.1016/j.jstrokecerebrovasdis.2010.11.002

26.

Evans

Golomb

. Statin-associated adverse cognitive effects: survey results from 171 patients. Pharmacotherapy. 2009;29(7):800-811. doi:10.1592/phco.29.7.800

27.

Wagstaff

Mitton

Arvik

Doraiswamy

. Statin-associated memory loss: analysis of 60 case reports and review of the literature. Pharmacotherapy. 2003;23(7):871-880. doi:10.1592/phco.23.7.871.32720

28.

Dagliati

Peek

Brinton

Geifman

. Sex and APOE genotype differences related to statin use in the aging population. Alzheimer's Dement. 2021;7(1):e12156. doi:10.1002/trc2.12156. Published 2021 May 2.

29.

Belessiotis-Richards

Hayes

Feng Yap

, et al. Systemic medications and dementia risk: a systematic umbrella review. Mol Psychiatr. 2025;30:5578-5599. doi:10.1038/s41380-025-03129-3. Published online 24 July.

30.

Holm

Ricci

Di Martino

, et al. Beta-blocker therapy and risk of vascular dementia: a population-based prospective study. Vasc Pharmacol. 2020;125-126:106649. doi:10.1016/j.vph.2020.106649

31.

Ding

Davis-Plourde

Sedaghat

, et al. Antihypertensive medications and risk for incident dementia and alzheimer's disease: a meta-analysis of individual participant data from prospective cohort studies. Lancet Neurol. 2020;19(1):61-70. doi:10.1016/S1474-4422(19)30393-X

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB

0.40 MB