FormalPara Key Points

There is a persistent, increased risk of major adverse cardiovascular/cerebrovascular event onset with long-term use of antipsychotic or cognitive-enhancing medications in those with behavioral and psychological symptoms of dementia.

Both antipsychotics and cognitive enhancers are associated with decreased time to negative cardiovascular event onset, and antipsychotic medications had a greater negative impact compared with cognitive enhancers.

Higher doses of medications decreased time to a MACCE onset, with the greatest risk for those taking high doses of antipsychotic medications compared with cognitive enhancers.

1 Introduction

The DSM-5 defines dementia as a neurocognitive disorder with diverse causes and clinical features. Dementia is characterized by significant cognitive decline from a previous level of performance in one or more cognitive domains, such as complex attention, executive function, learning and memory, language, perceptual-motor, or social cognition [1]. This decline not only affects cognitive abilities but also significantly interferes with everyday activities, distinguishing dementia from normal aging and from mild cognitive impairment (MCI).

A vast majority (97%) of patients with dementia also experience clinically significant noncognitive neuropsychiatric symptoms, collectively referred to as behavioral and psychological symptoms of dementia (BPSD) [2]. These symptoms disrupt perception, thought content, mood, and behavior [3] and are often categorized into five clusters: apathy (lack of initiative and eating disturbances), psychosis (sleep disturbances, hallucinations, and delusions), aggression/agitation (verbal/physical), hyperactivity (disinhibition, psychomotor agitation, and restlessness), and affective (dysphoria, depression, and anxiety) [4]. It is common for patients to exhibit multiple BPSD simultaneously, with approximately 50% experiencing at least four neuropsychiatric symptoms [5]. Among these symptoms, depression, apathy, and anxiety are the most prevalent, while hallucinations, disinhibition, and euphoria are less common but clinically significant [5].

BPSD are strongly associated with distress in both patients and caregivers, leading to negative outcomes such as increased morbidity and mortality, greater healthcare utilization, functional impairment, emotional distress, and earlier nursing home placement [6]. These symptoms exacerbate cognitive decline and physical dysfunction, correlating with the degree of functional and cognitive impairment. Anxiety and depression have been identified as precursors to future cognitive decline, increasing the risk of dementia, particularly Alzheimer's disease (AD) [7]. Additionally, higher numbers of neuropsychiatric symptoms are linked to lower 3-year survival rates [8], with psychosis specifically associated with increased mortality and accelerated cognitive decline [9]. BPSD pose a significant challenge in medical management and are a major cause of institutionalization [10], highlighting the importance of targeting these symptoms alongside cognitive decline in treatment strategies [11].

Antipsychotic drugs (APs) target specific symptoms of BPSD including agitation, aggression, psychosis, and inappropriate behaviors [12] and have been reported to produce better improvements in treating BPSD than placebos [13]. Numerous APs have dopamine D2 blocking properties in cortico-limbic regions that contribute to the alleviation of psychosis and behavioral excitation [12]. Although, many cause extrapyramidal side effects (movement dysfunction), APs are still used to treat BPSD with a prescription rate of about 20–50% [13] and are often used for sustained periods lasting more than 6 months [14].

APs are generally classified into two groups based upon their receptor antagonism potency: typical and atypical [15]. Typical APs, such as phenothiazines (e.g., chlorpromazine, fluphenazine), butyrophenones (e.g., haloperidol), and benzamides (e.g., sulpiride), primarily target dopamine receptors and often induce severe extrapyramidal effects due to their high potency [12]. On the other hand, atypical APs such as risperidone and olanzapine act on both dopamine and serotonin receptors, with lower potency and a reduced risk of extrapyramidal effects [16]. Despite their recognized adverse effects, APs are commonly prescribed off-label to manage neuropsychiatric symptoms in older individuals with dementia, especially AD. Orsel et al. [17] indicated that approximately 14–50% of people with dementia are treated with two or more psychotropic drugs; however, for those diagnosed with AD, the use of two or more psychotropic drugs increased from 5.9% 5 years before to 18.3% 4 years after diagnosis. Despite being frequently prescribed for the purpose of alleviating BPSD [18], the clinical efficacy of atypical APs is limited, and they pose many safety concerns, particularly regarding cardiovascular incidents and mortality risk [19]. While studies suggest improvement in symptoms like agitation and psychosis with atypical APs, they do not significantly enhance overall functioning or quality of life in populations with AD [20].

Dysfunction in the acetylcholine (ACh) neurotransmitter system is observed in dementia [21], potentially contributing to cognitive decline [22] and the behavioral and emotional responses observed in BPSD [23]. Additionally, overactivation of glutamate receptors, particularly NMDA receptors, is implicated in synaptic dysfunction underlying memory deficits [24]. Cognitive enhancers, including cholinesterase inhibitors (ChEIs) such as donepezil, galantamine, and rivastigmine and the NMDA receptor antagonist memantine, are commonly used in dementia treatment to slow cognitive decline and control disruptive behaviors [25]. Studies show positive effects of ChEIs in improving or stabilizing preexisting BPSD and postponing the appearance of new symptoms. Donepezil, galantamine, and rivastigmine have demonstrated efficacy in treating mild to moderate BPSD symptoms, with specific benefits observed in symptoms such as delusion, paranoia, sleep disturbance, and anxiety [26,27,28]. Memantine has also shown effectiveness in reducing agitation, aggression, and delusions [29].

Despite their effectiveness in reducing psychosis and behavioral disturbances in individuals with dementia [13], both APs and cognitive enhancers have drawn warnings from the US Food and Drug Administration (FDA) due to their association with various adverse events. These events include metabolic symptoms such as weight gain, hyperlipidemia, and diabetes, as well as cerebro/cardiovascular events, gait disturbance, cognitive decline, and even sudden death [30]. Evidence suggests that the use of these medications is linked to an increased risk of major adverse cardiac and cerebrovascular events (MACCE). Recent meta-analyses have indicated an elevated all-cause mortality associated with AP use in patients with dementia [31], with specific drugs such as olanzapine and risperidone being linked to stroke in older adults [32, 33]. Large-scale studies have shown that AP use is correlated with a 35% higher likelihood of mortality compared with non-users [34], with this heightened risk becoming evident within 3 months of treatment initiation and persisting with prolonged usage, as demonstrated in analyses of the Medicare database [35].

Despite the acknowledged risks, pharmacological interventions remain a common choice for managing BPSD [36]. Therefore, this study aimed to examine whether individuals prescribed APs and/or cognitive enhancers experience a faster onset of MACCE compared with those not receiving such treatment. We hypothesized that individuals prescribed APs or cognitive enhancers would exhibit a shorter survival time, indicating a quicker onset of MACCE, than those who were not prescribed these medications. Additionally, we anticipated that higher doses of either medication type would correlate with a shorter survival time compared with lower doses.

2 Methods

2.1 Ethics

The Loma Linda University Institutional Review Board reviewed the study proposal (IRB # 5230414) and provided a determination of approval without full human subject review on 12 September 2023.

2.2 Study Design

A study comparing entire survival experiences between treatment groups (cognitive enhancers only, APs only, both cognitive enhancers and Aps, and no drug treatment) was conducted using a sample of deidentified patients from Loma Linda University Medical Center database. Data were extracted for patient sociodemographics (age when diagnosed with dementia, sex, and race), clinical features (dementia diagnosis, number of comorbidities, MACCE, date of diagnoses, and number of MACCE before dementia diagnosis), and drug-related characteristics (drug treatments, drug dosage and frequency, and number of medications). Participants were organized into one of four groups based on the medication used to treat their BPSD symptoms: cognitive enhancers, AP drugs only, both drugs, and no drug treatment. For our secondary analysis, participants in the cognitive enhancers only and APs only groups were further organized into three groups based off the dosage used to treat their symptoms: low dose, medium dose, and high dose.

2.3 Participants

The study population (N = 13,562) consisted of individuals aged ≥ 50 or older who were diagnosed with a form of dementia and BPSD symptoms. Inclusion criteria included having data on (a) dementia diagnosis, (b) list of all diagnoses including psychiatric illness, (c) medication history, and (d) sociodemographic information including age, sex, smoking status, and race. Exclusion criteria included (a) missing information from their records such as age or date when diagnosed with dementia and date of MACCE diagnosis, (b) current diagnosis of any psychiatric illness not related to their dementia status, or (c) previous history of treatment with an AP or cognitive enhancer prior to dementia diagnosis.

2.4 Variables

2.4.1 Medication Treatment

APs and cognitive enhancers were identified using the WHO Anatomical Therapeutic Chemical (ATC) classification, with a total of five different APs (olanzapine, haloperidol, quetiapine, aripiprazole, and risperidone) and four cognitive-enhancing medications (memantine, donepezil, rivastigmine, and galamantine) included in the data.

AP dosages and maximum doses are dependent on the reason the drug was prescribed. Since many different types of symptoms can be considered as part of BPSD, we used the maximum recommended dose for BPSD when determining categorization of a drug into our low-dose, medium-dose, and high-dose groups. According to the British Columbia (BC) BPSD Algorithm (2014) [37], the maximum dose for treating BPSD using olanzapine is 10 mg. As such, 3 mg or less was considered a low dose, 4–6 mg a medium dose, and 7 mg or greater a high dose. According to Cloak and Khalili [38], the maximum recommended dose for treating BPSD using haloperidol is 5 mg. As such, doses of 2 mg or less were considered a low dose, doses between 2 and 3 mg were a medium dose, and doses 3 mg or greater were a high dose. The maximum recommended dose for treating BPSD using quetiapine is 200 mg [37]. As such, doses 75 mg or less were considered a low dose, doses between 75 and 150 mg were a medium dose, and doses 150 mg or greater were a high dose. The maximum recommended dose for treating BPSD using aripiprazole is 10 mg [37]. As such, doses 3 mg or less were considered a low dose, doses between 3 and 6 mg were a medium dose, and doses 6 mg or greater were a high dose. The maximum recommended dose for monotherapy maintenance doses for treating BPSD using risperidone is 2 mg [37]. As such, doses 0.5 mg or less were considered a low dose, doses between 0.5 and 1 mg were a medium dose, and doses 1 mg or greater were a high dose.

According to Da Re et al. [39], the maximum recommended dose for treating BPSD using memantine is 20 mg. As such, doses of 10 mg or less were considered a low dose, doses between 10 and 15 mg were a medium dose, and doses 15 mg or greater were a high dose. The maximum dose for treating BPSD using donepezil is 10 mg [40]. As such, doses 5 mg or less were considered a low dose, doses between 5 and 7.5 mg were a medium dose, and doses 7.5 mg or greater were a high dose. According to Patel and Gupta [41], the maximum recommended dose for treating BPSD using oral capsule rivastigmine is 12 mg. As such, doses 4 mg or less were considered a low dose, doses between 4 and 9 mg were a medium dose, and doses 9 mg or greater were a high dose. The maximum recommended dose for treating BPSD using galamantine is 24 mg [40]. As such, doses 12 mg or less were considered a low dose, doses between 12 and 18 mg were a medium dose, and doses 18 mg or greater were a high dose.

2.4.2 MACCE

The primary outcome assessed for inclusion in the prediction analysis was major adverse cardiac and cerebrovascular events (MACCE). MACCE was defined as a composite of myocardial infarction (MI), ischemic or hemorrhagic stroke, transient ischemic attack (TIA), aneurysm, cardiogenic shock, complete heart block, cardiac arrest, and congestive heart failure.

2.4.3 Covariates

Covariates included demographics (age when diagnosed with dementia, sex, and race), and comorbidities (smoking status and number of MACCE occurring before dementia diagnosis). Smoking status was divided into four categories: never, former, sometimes, and every day.

2.5 Data Sources

Deidentified patient data from Loma Linda University Medical Center database were collected from August 2008 until August 2023 and were used to create our sample. The use of patients’ medical and pharmacotherapeutic records was authorized by the clinical directors of all participating departments.

2.6 Statistical Analysis

An initial Cox proportional hazards regression analysis was performed on data from individuals diagnosed with dementia to evaluate the effects of medication treatment (1 = cognitive-enhancers only, 2 = AP only, 3 = cognitive enhancers and APs only, and 4 = no medications) on duration with dementia until MACCE, after controlling for the effects of participant race (1 = white/Caucasian, 2 = Asian, 3 = Hispanic/Latino(a), 4 = Black/African American, 5 = Native Hawaiian/Pacific Islander), smoking status (1 = unknown, 2 = never, 3 = former, 4 = sometimes, 5 = every day), age when diagnosed with dementia, sex (1 = male, 2 = female), and number of MACCE before dementia diagnosis.

A second Cox proportional hazards regression analysis was performed on data from individuals diagnosed with dementia to evaluate the effects of AP medication dosage (1 = low dose, 2 = medium dose, 3 = high dose) on duration with dementia until MACCE, after controlling for the effects of participant race (1 = white/Caucasian, 2 = Asian, 3 = Hispanic/Latino(a), 4 = Black/African American, 5 = Native Hawaiian/Pacific Islander), smoking status (1 = unknown, 2 = never, 3 = former, 4 = sometimes, 5 = every day), age when diagnosed with dementia, sex (1 = male, 2 = female), and number of MACCE before dementia diagnosis. A third Cox proportional hazards regression analysis was performed on data from 236 individuals diagnosed with dementia to evaluate the effects of AP medication dosage (1 = low dose, 2 = medium dose, 3 = high dose) on duration with dementia until MACCE, after controlling for the effects of participant race (1 = white/Caucasian, 2 = Asian, 3 = Hispanic/Latino(a), 4 = Black/African American, 5 = Native Hawaiian/Pacific Islander), smoking status (1 = unknown, 2 = never, 3 = former, 4 = sometimes, 5 = every day), age when diagnosed with dementia, sex (1 = male, 2 = female), and number of MACCE before dementia diagnosis. All Cox regression analyses were performed using SPSS (version 28). Prior to analysis, the data were examined for outliers and for violations of the assumptions of survival analysis; none were found. No cases were censored.

Given a hazard ratio of 0.67, an alpha level of 0.05, and a desired power of 0.90, a total sample size of 267 events (across all groups) is required to detect a statistically significant difference in survival between groups with 90% power. With an event rate in the control group of 11% over a follow-up period of 7.5 years, a total sample size of 527 subjects (across all groups) would be needed to observe the required number of events. Our final sample fulfills these criteria.

3 Results

3.1 Patient Characteristics

From the initial 13,562 patients, 12,436 individuals were excluded from the study sample owing to missing data. The final sample consisted of 1126 patients, most of whom were female (N = 597; 52.9%) with a mean age of 78.19 ± 6.71 years when diagnosed with dementia. Patients with a diagnosis of “dementia” rather than a more specific diagnosis (N = 495) accounted for 43.9% of the sample, 82.4% had no previous history of a MACCE (N = 928), and 58% presented with no history of smoking tobacco (N = 655). Many participants self-identified their race as white/Caucasian (N = 981; 86.9%). In the initial analysis, participants were assigned into one of four treatment groups: cognitive enhancers only (N = 228, 20.2%), AP drugs only (N = 478, 42.3%), both drugs (N = 259, 22.9%), and no drug treatment (N = 164, 14.5%). In the secondary analysis, those prescribed cognitive enhancers only were assigned into groups based off their dosage levels: low dose (N = 96, 24.5%), medium dose (N = 100, 25.5%), and high dose (N = 32, 8.2%). Additionally, those prescribed APs only were assigned into groups based off their dosage levels: low dose (N = 187, 78.2%), medium dose (N = 39, 16.3%), and high dose (N = 13, 5%). Further information on participant demographics can be found in Table 1.

Table 1 Participant demographics (N = 1126)
Full size table

3.2 The Effect of Drug Treatment on MACCE in Patients with Dementia

The results of this survival analysis indicated that there was a significant effect of medication factors on duration until MACCE (χ2(3) = 166.091, p < 0.001). Specifically, medication treatment significantly influenced duration with dementia until MACCE onset after controlling for all other covariates, such that the odds of experiencing a MACCE were 96.3% higher for individuals who were treated with both AP drugs and cognitive-enhancing drugs than those on no medications [hazard ratio (HR) 1.963, 95% CI [1.601, 2.406], p < 0.001]. The mean time until MACCE for those treated with both APs and cognitive enhancers was 3.871 years, while those not treated with any medication experienced a mean time of 6.351 years until MACCE (N = 164, 14.5%). Additionally, those treated with only AP drugs had 78.7% greater odds of experiencing a MACCE sooner than those not treated with APs (HR 1.787, 95% CI [1.527, 2.092], p < 0.001). The mean time until MACCE for those treated with APs was 4.220 years. Those treated with only cognitive-enhancing drugs had 31.9% greater odds of experiencing a MACCE than those not prescribed cognitive enhancers (HR 1.319, 95% CI [1.101, 1.581], p = 0.003). The mean time until MACCE for those treated with cognitive enhancers was 5.931 years.

Race, sex, and number of MACCE before dementia diagnosis also predicted duration of dementia until MACCE. When compared with white participants, the odds of Black/African American individuals experiencing a MACCE sooner were 71.7% greater (HR 1.717, 95% CI [1.269, 2.323], p < 0.001), after controlling for all other covariates. Additionally, the odds of Asian individuals experiencing a MACCE sooner after dementia diagnosis were 43.2% greater (HR 1.432, 95% CI [1.033, 1.985], p = 0.031) and 56.8% greater for Hispanic/Latino(a) individuals (HR 1.568, 95% CI [1.009, 2.437], p = 0.045). The odds of a MACCE occurring after dementia diagnosis were 13.8% greater for males (HR 1.138, 95% CI [1.007, 1.285], p = 0.038) and 59.2% greater for every additional MACCE that occurred before dementia diagnosis (HR 1.592, 95% CI [1.381, 1.835], p < 0.001). Further information on the results of the Cox regression examining treatment type on survival can be found in Table 2. A plot of the survival functions for medication treatments at the means of all other covariates is shown in Fig. 1. Year 6 provides the largest separation between treatment type and time until MACCE. The graph indicates that, for those treated with both APs and cognitive enhancers, about 82% of patients with dementia had experienced a MACCE.

Table 2 Cox regression analyses of factors on with MACCE occurrence
Full size table
Fig. 1
figure 1

MACCE survival according to group by cox proportional hazards regression. BPSD behavioral and psychological symptoms of dementia, MACCE major adverse cardiovascular/cerebrovascular events. Survival curve plotting the probability that a MACCE has not occurred by (and including) each time point (2.5-year intervals). Solid red line, no drugs; solid blue line, cognitive enhancers; solid green line, antipsychotic drugs only; solid yellow line, both cognitive enhancers and antipsychotics

Full size image

3.3 The Effect of Drug Dosage on MACCE in Patients with Dementia

A secondary Cox regression survival analysis indicated that there was a significant effect of cognitive-enhancing medication dosage on duration until a MACCE (χ2(12) = 46.884, p < 0.001). When compared with low dosages of cognitive-enhancing drugs, high dosages significantly influenced duration with dementia until MACCE occurrence after controlling for all other covariates, such that the odds of experiencing a MACCE were 76% higher for individuals on a high dose of cognitive-enhancing drugs (HR 1.760, 95% CI [1.147, 2.702], p = 0.010). The mean time until MACCE for those treated with high-dose cognitive enhancers was 3.781 years. Individuals taking a medium dose of cognitive enhancers experienced 38% greater odds (HR 1.380, 95% CI [1.029, 1.850], p = 0.032). The mean time until MACCE for those treated with medium dose cognitive enhancers was 4.480 years. A plot of the survival functions for AP and cognitive enhancer dosages is shown in Fig. 2A. Year 4 provides the largest separation between cognitive enhancer dosage and time until MACCE, suggesting that, after 4 years, approximately 62% of patients with dementia had experienced a MACCE.

Fig. 2
figure 2

MACCE survival curves by drug dosage using Cox proportional hazards regression. BPSD behavioral and psychological symptoms of dementia, MACCE major adverse cardiovascular/cerebrovascular events. A A survival curve plotting the probability that a MACCE has not occurred by (and including) each time point (2.5-year intervals) for cognitive enhancer dosing. Solid blue line, low doses of cognitive enhancers; solid green line, medium doses of cognitive enhancers; solid red line, high doses of cognitive enhancers. B A survival curve plotting the probability that a MACCE has not occurred by (and including) each time point (2.5-year intervals) for antipsychotic dosing. Solid blue line, low doses of antipsychotics; solid green line, medium doses of antipsychotics; solid red line, high doses of antipsychotics

Full size image

There was also a significant effect of AP dosage on duration until MACCE, χ2(12) = 29.963, p = 0.003. When compared with low dosages of APs, high dosages significantly influenced duration with dementia until MACCE occurrence after controlling for all other covariates, such that the odds of experiencing a MACCE were 238.0% higher (HR 3.380, 95% CI [1.880, 6.077], p < 0.001). The mean time until MACCE for those treated with high dose APs was 2.692 years. Individuals taking a medium dose of APs had 48.0% greater odds of experiencing a MACCE (HR 1.480, 95% CI [1.035, 2.116], p = 0.032). The mean time until MACCE for those treated with medium dose APs was 4.308 years. A plot of the survival functions for AP and cognitive enhancer dosages is shown in Fig. 2B. Year 6 provides the largest separation between dosage and time until MACCE in graph B, suggesting that, for those prescribed high-dose APs, about 87% of patients with dementia had experienced a MACCE. Further information on both Cox regression analyses examining the effect on drug dosage on survival can be found in Table 2.

4 Discussion

In this study, we found an increased risk for sooner MACCE onset for individuals treated with both APs and cognitive enhancers, as well as for those prescribed higher drug dosages. Our findings also indicate that the following patient characteristics may also explain the occurrence of sooner MACCE onset: history of MACCE before dementia diagnosis, race (African American/Black, Asian, or Hispanic), and sex (male).

The association of AP and cognitive enhancer drug use with MACCE occurrence has been previously described in the literature. Previous research has demonstrated that the risk of stroke is elevated shortly after the initiation of AP treatment in individuals with AD [42]. Commencing APs is linked to a higher age-standardized incidence rate and an increased likelihood of experiencing a stroke within the first 2 months compared with those who do not initiate AP treatment. Additionally, a meta-analysis of three randomized controlled trials on ziprasidone and quetiapine revealed a rise in the prolongation of the QTc interval, increasing the risk of ventricular tachyarrhythmias [43]. Moreover, two retrospective cohort studies have linked the use of APs to a significantly increased risk of sudden cardiac mortality in patients with other psychiatric illnesses. Straus et al. [44] found that the current use of APs in the general population is associated with an increased risk of sudden cardiac death, even at a low dose and in persons who use APs for indications other than schizophrenia. These instances indicate that APs, frequently administered to individuals with diverse psychiatric disorders, including dementia, are linked to sudden cardiac mortality, raising significant concerns for all populations, particularly those with dementia who might receive such prescriptions. As for cognitive enhancers, several possible cholinergic cardiac side effects have been reported including hypotension, bradycardia, heart block, and QT/QTc prolongation [43]. While adverse cardiac events with cognitive enhancers are less common [44], older adults treated with these drugs may be more susceptible to adverse events due to the high prevalence of cardiac comorbidities. Thus, prescribers are becoming more concerned about the possibility of severe side effects from AChEIs, particularly in older adult patients with AD [45].

Our findings have shown that there is a significant decrease in the time to MACCE onset for those prescribed both APs and cognitive enhancers. Of the 259 patients in our sample who were diagnosed with BPSD and treated with both APs and cognitive-enhancing drugs, approximately 212 patients (82%) were diagnosed with MACCE within 6 years of beginning BPSD treatment. In comparison, approximately 80% of those solely taking APs and 70% of those only taking cognitive enhancers to reduce BPSD symptomology had experienced a MACCE within 6 years. This may be explained by the fact that certain APs tend to cause weight gain in addition to cardiovascular and metabolic abnormalities, such as increased risk of obesity, metabolic syndrome, type 2 diabetes mellitus, and related cardiovascular morbidity [46, 47]. Studies on APs, particularly second-generation APs (SGAs) such as olanzapine and risperidone, have been associated with weight gain, dyslipidemia, and insulin resistance in older adults [48]. Comparatively, the use of APs in dementia has been consistently linked to a 1.5–1.7 times increased risk of mortality, and a 2-3-fold increased risk of cerebrovascular events (CVAE) [46]. Another study found that, among hospitalized adults, typical APs were linked to elevated mortality or cardiopulmonary arrest rates, whereas atypical APs showed increased risk solely among adults aged 65 years and older, with a total of 691 outcome events (0.5% of total hospitalizations), comprising 515 deaths and 176 cardiopulmonary arrests [49]. A less noticeable but comparable pattern was observed in nursing home residents. Other literature suggests that atypical or conventional AP users are more likely to experience avoidable adverse events such as infections, falls, and worsening cognitive function [50]. Our study is consistent with previous findings and provides an estimated timeframe for the onset of MACCE. However, to our knowledge, ours is the first study to examine the relationship between AP medications, cognitive enhancers, and MACCE within dementia populations, specifically focusing on the context of treating BPSD and examining the impact of medication dosage on the time to MACCE onset. Therefore, we must interpret these findings in the context of similar but limited previous research.

There are nonpharmacological interventions for the treatment of BPSD including indirect techniques aimed at decreasing symptoms of BPSD by working with caregivers, adapting to the environment, or directly targeting problematic symptoms. Indirect interventions (caregiver training, multidisciplinary team approaches, individualized treatment plans, and modifying environmental factors) primarily focus on educating care staff, physicians, and pharmacists. However, according to a systematic review, these techniques are effective in the short term and do not address the complexity of BPSD or the factors that influence symptoms long term [50]. Direct interventions include (1) cognitive/emotion-oriented interventions (validation therapy, simulate presence therapy), (2) sensory stimulation interventions (light therapy, music therapy, transcutaneous electrical nerve stimulation), and (3) behavioral management techniques [51]. Cognitive training involves the targeted practice and training of specific cognitive domains, while cognitive rehabilitation takes a more tailored therapeutic approach and aims to improve functioning in daily life rather than cognitive performance [52]. In addition, psychosocial therapies are intended to enhance self-esteem, well-being, and social/communication skills [53]. Regardless of one’s cognitive capacity or functional ability, these strategies attempt to introduce a range of tailored meaningful activities for patients with dementia, in hopes of ensuring that individuals are fully able to participate and benefit. Studies on the effects of physical activity have shown improvements in patient mood, quality of life, falls, cardiovascular health, and rates of disability [54, 55]. Frequent exercise can also increase cardiovascular fitness, muscle mass, arterial compliance, energy metabolism, and overall functional capacity [56]. Therefore, the current literature suggests there are alternatives to medication for treatment of BPSD, with varying levels of effectiveness.

This study has some limitations worth acknowledging. First, data were obtained through a Loma Linda University Medical Center database, which included deidentified data from patients in a blue zone. In this region, there is a particularly high number of centenarians, as people tend to live longer than average and maintain good health. Many people utilizing Loma Linda Health Care are affiliated with the Seventh Day Adventist religion, who adhere to a vegan or vegetarian diet, prioritize exercise, and refrain from smoking and drinking. However, no data on religion/spirituality or individual lifestyle were obtained in this study. Thus, the sample used for our study may not be representative of true survival rates or MACCE occurrences in the general population. Second, in our study, smoking status did not consider exposure to second-hand smoke, which is known to be a risk factor for heart disease [57]. Third, the impact of drug use on health outcomes, particularly regarding anticholinergic burden (ACB) (the cumulative effect of taking multiple drugs with anticholinergic activities), may be influenced by both cumulative and time-related effects. The duration of exposure to medications containing anticholinergic properties, coupled with the quantity of these medications taken, can amplify their potential health effects. This cumulative effect signifies the collective strain placed on the body from the continuous use of various drugs with anticholinergic activity over an extended period. Within the scope of this study, the individual anticholinergic burden of each participant remains undetermined. Fourth, comorbidities and geriatric syndromes such as malnutrition, sarcopenia, and frailty can significantly increase the risk of developing MACCE [58]. While no data on these comorbidities were obtained in this study, these conditions do tend to increase with age, and age and years since onset of dementia were included as covariates in our analyses. Fifth, while it is important to consider the adherence to AP or cognitive enhancers prescriptions to evaluate the relationship with MACCE, no information on this was available to us considering the data were derived from a university healthcare system database. Sixth, this study did not account for the severity of dementia as there were no available data in the Loma Linda Health Care database. Thus, we could not account for the influence these factors may have had on calculated survival rates. Seventh, the cognitive enhancer variable included ChEIs and memantine, which is a NMDA receptor antagonist, into the same category. Thus, it is unclear whether ChEIs and NMDA receptor antagonists have different influences on time until MACCE onset. Eighth, given the limited scope of this study, we did not examine other treatments. Our analysis focused solely on the use of cognitive enhancers, APs, a combination of both, and no medication treatment. Thus, we did not investigate the prescription of other pharmacological treatments, such as antidepressants, nor did we have data on the use of nonpharmacological treatments. However, SSRIs [59] and tricyclic antidepressants [60] have been associated with a higher risk of hospitalizations for cardiovascular events in older adults. Therefore, further studies should include other classes of medications in their analysis. Ninth, the current study does not compare the risk profile for different AP and cognitive-enhancing medications. It remains for future studies to determine whether some AP or cognitive-enhancing medications have a different risk pattern than the one reported for the aggregate data in this study.

We also speculate on several measures and future directions based on these findings. First, there may be a need for more precise risk assessment tools to identify individuals with BPSD who are at the highest risk of MACCE associated with medication use. Additionally, further research could explore alternative treatment modalities, such as nonpharmacological interventions or lower-risk medications, for managing BPSD symptoms. Longitudinal studies tracking medication use and cardiovascular outcomes in larger and more diverse populations could provide additional insights into the long-term risks and benefits of pharmacological treatments for BPSD. Further, due to the association of SSRIs and antidepressants with MACCE among older adults, future research should analyze the impact of other types of pharmacological treatments on MACCE onset among older adults with BPSD. Lastly, enhancing healthcare provider education and awareness regarding the potential cardiovascular risks associated with AP and cognitive enhancer use in individuals with dementia could help optimize treatment decisions and improve patient outcomes.

The strengths of this study lie in its innovative approach and unique contributions to the existing literature on the intersection of AP medications, cognitive enhancers, and MACCE within dementia populations, specifically focusing on the context of treating BPSD. Firstly, while previous research has explored the individual impacts of MACCE on dementia and the use of AP in dementia separately, this study stands out for its comprehensive examination of the relationship between these factors. By investigating how both APs and cognitive enhancers contribute to the risk of MACCE within dementia populations, the study provides a more holistic understanding of the potential cardiovascular and cerebrovascular risks associated with these commonly prescribed medications. This integrative approach is vital for clinicians and researchers seeking to optimize treatment strategies and mitigate potential adverse effects in patients with dementia experiencing BPSD. Secondly, the study’s focus on medication dosage represents a significant methodological advancement. While previous studies have explored the overall association between medication use and MACCE risk, this study goes a step further by specifically examining the impact of dosage on the time to MACCE onset. By quantifying the relationship between medication dosage and cardiovascular/cerebrovascular health outcomes, the study offers valuable insights into the dose–response relationship and helps elucidate the potential dose-dependent risks associated with AP and cognitive enhancer use in dementia populations. This nuanced understanding of dosage effects can inform more precise prescribing practices and facilitate personalized treatment approaches tailored to individual patient needs and risk profiles. Overall, the current study’s innovative approach, combined with its novel contributions to understanding the relationship between medication use, dosage, and MACCE risk in dementia populations, strengthens its significance within the field of geriatric psychiatry and has important implications for clinical practice, research, and policymaking aimed at improving the safety and efficacy of pharmacological interventions for BPSD.

The results of this study suggest a persistent, increasing risk of MACCE onset with long-term use of AP or cognitive-enhancing medications in those with BPSD. However, our findings indicate that higher doses of APs are particularly risky and present many adverse effects on cardiovascular/cerebrovascular health, some of which are medically serious. Predicting an estimate of the expected survival after beginning drug treatment for BPSD is a useful measure for individuals and their families, as well as for clinicians providing treatment for these symptoms. These research findings underscore the importance of carefully balancing the benefits and risks of pharmacological interventions for BPSD. Clinicians need to weigh the urgency of treating severe symptoms against the potential risks of MACCE associated with medication use. Often, there is a sense of urgency when treating more severe symptoms of BPSD; however, clinicians should consider the risk of MACCE for these individuals and which factors influence the length of survival until MACCE onset. Moreover, considering factors that influence the length of survival until MACCE onset should be an integral part of treatment decision-making for individuals with BPSD.

4.1 Conclusions

There is a persistent, increased risk of major adverse cardiovascular/cerebrovascular event onset with long-term use of AP or cognitive-enhancing medications in those with behavioral and psychological symptoms of dementia. Both APs and cognitive enhancers are associated with decreased time to negative cardiovascular event onset, and APs had a greater negative impact compared with cognitive enhancers. In addition, higher doses of medications decreased time to a MACCE onset, with the greatest risk for those taking high doses of AP medications compared with cognitive enhancers. These novel findings highlight the need for cardiovascular/cerebrovascular monitoring in patients with dementia and BPSD.