Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Nov 10.
Published before final editing as: J Am Med Dir Assoc. 2026 Nov 7;28(1):106540. doi: 10.1016/j.jamda.2026.106540

Trends in Antipsychotic Initiation and Use Among Adults with Dementia, 2016–2025: A Pharmacoepidemiologic Study Using the TriNetX Research Network

Wanhong Zheng 1, Chris Feghali 1, Krista Ulisse 1, Khalid Sharif 1, Matthew Zell 1
PMCID: PMC13747660  NIHMSID: NIHMS2214395  PMID: 42998354

Abstract

Objective

To examine prescribing trends in prevalence and initiation of individual antipsychotics among a predominantly US-based cohort of adults with dementia from 2016 to 2025, with demographic stratification and formal trend analysis.

Design

Retrospective cohort study.

Setting and Participants

This study used TriNetX Research Network data from 111 healthcare organizations. Adults aged ≥65 years with dementia between January 1, 2016, and December 31, 2025, were included; individuals with prior psychotic disorders were excluded.

Methods

Eighteen FDA-approved antipsychotics, including 14 second-generation and 4 first-generation agents, were identified from electronic health record (EHR) prescription data. Annual prevalence and initiation (proportion and rates) were calculated overall and by age, sex, and race. Joinpoint regression was used to estimate annual percent change (APC) and average annual percent change (AAPC) in initiation.

Results

Among 950,573 dementia patients (mean age, 82 years; 56.4% female), prevalence of antipsychotic use increased from 2016 to 2025. Quetiapine and haloperidol were the most frequently used agents, increasing from 3,605 to 18,432 and from 4,745 to 15,829 per 100,000, respectively. Olanzapine prevalence increased from 1,205 to 9,264 per 100,000 patients. Initiation rates rose substantially, driven by select antipsychotics; quetiapine initiations increased nearly fivefold, while olanzapine showed steep increase (AAPC, 30.51%; P<.001). Use was highest among individuals aged 75–84 years and varied by sex and race. Joinpoint analyses demonstrated late-period acceleration for several agents, including quetiapine, olanzapine, risperidone, and brexpiprazole.

Conclusions and Implications

Antipsychotic use in dementia increased substantially over the past decade, driven primarily by rising initiation of selected agents. These findings highlight marked agent-specific and demographic heterogeneity in prescribing patterns and underscore the need for continued monitoring of antipsychotic use in this high-risk population.

Keywords: Antipsychotic, dementia, trends, pharmacoepidemiology

Summary:

In this retrospective cohort study of 950,573 adults with dementia, antipsychotic use increased substantially from 2016 to 2025, driven primarily by rising initiation of selected agents despite safety concerns.

Introduction

The global life expectancy is increasing1. As the population ages, dementia has become a major global health challenge. It is also a significant cause of mortality in the elderly population2. A recent study showed that the lifetime risk of dementia in US adults over the age of 55 years is 42%3. Approximately 75% of individuals with dementia experience behavioral symptoms, which are associated with nursing home placement, adverse effects on interpersonal functioning, and shorter survival4,5. Although disease-modifying therapies such as lecanemab and donanemab have recently been approved for early Alzheimer’s disease, pharmacological treatment of dementia remains primarily focused on symptom control6. While non-pharmacological treatment of behavioral symptoms is traditionally recommended first line4, antipsychotic use among dementia patients has been extensively documented across various care settings.

Antipsychotic prescribing patterns among patients with dementia vary across healthcare settings and regions. In the United States, with regulatory initiatives of reducing antipsychotic use, antipsychotic prescribing rate in nursing homes decreased from 40% to 14–15%7. In contrast, its use among community-dwelling dementia patients nearly doubled from 9% to 16% between 2010 and 2018, despite Food and Drug Administration (FDA) black-box warnings and Choosing Wisely recommendations8. More broadly, antipsychotic use per 100 persons aged ≥65 rose from 2.67 to 4.05 between 2015 and 2024, with long-term use increasing approximately 65–71%9. International studies have likewise demonstrated persistent but variable antipsychotic use in dementia populations across Europe, Australia, and Japan, highlighting the influence of regional regulatory policies, healthcare systems, and prescribing practices10–12.

Second-generation agents (SGAs) dominate geriatric antipsychotic prescribing, as first-generation agents (FGAs) use fell from 22.2% to 14.3%9. Among individual agents, quetiapine has become the most widely prescribed, accounting for approximately 61% of regular antipsychotic prescriptions in some cohorts. This pattern emerged partly following 2003 cerebrovascular warnings for risperidone and olanzapine that did not initially apply to quetiapine13,14. Nonetheless, prescribing patterns vary across countries and care settings. Risperidone remains clinically important because it is the only antipsychotic licensed for short-term treatment of behavioral and psychological symptoms of dementia in several jurisdictions, including in the EU, UK, and Australia15–17. Haloperidol remains the most prescribed FGA16. In May 2023, brexpiprazole became the first drug with FDA approval specifically for agitation associated with Alzheimer’s dementia18.

Demographic correlates of antipsychotic use have also been identified: male sex is independently associated with higher likelihood of being prescribed antipsychotic agents despite female predominance of dementia16. Racial disparities exist in different settings, with Black patients more likely to receive antipsychotics in community settings while Hispanic and Asian patients are more likely to receive antipsychotics in nursing facilities8,19. In the UK, Black and Asian patients received antipsychotics for significantly longer durations than White patients, suggesting differential exposure to potentially inappropriate long-term use15.

Previous research often used limited, regional, or cross-sectional data focused on antipsychotic classes rather than specific agents. To address this gap, we analyzed temporal trends in the prevalence and incidence of 18 antipsychotic medications among adults with dementia from 2016 to 2025 using a large multinational electronic health record network comprising a predominantly US population, with results stratified by demographic characteristics.

Methods

Data Source and Study Design

This retrospective cohort study used TriNetX Research Network data, a real-time database of de-identified electronic health records (EHR) from healthcare organizations (HCOs) located primarily in the United States, with additional participation from non-US institutions. It contains longitudinal patient-level data on demographics, diagnoses, and medications and is widely used for epidemiologic and real-world studies. All data are HIPAA-compliant and de-identified. As no direct human subject interaction occurred, the study was classified as non-human subjects research and exempt from West Virginia University IRB review.

Study Population

Patients with dementia were identified using International Classification of Diseases, Tenth Revision (ICD-10) diagnosis codes, including Alzheimer’s disease (G30), vascular dementia (F01), dementia in other diseases classified elsewhere (F02), unspecified dementia (F03), and frontotemporal dementia (G31.0). To focus on late-onset dementia and enhance clinical validity, the study population was restricted to individuals aged 65 years or older at the time of cohort entry, and early-onset dementia cases were excluded.

To minimize confounding by primary psychotic disorders, patients with any lifetime diagnosis of a psychotic disorder were excluded, including schizophrenia (F20), schizoaffective disorder (F25), delusional disorder (F22), bipolar disorder with psychotic features (F31.2, F31.5), and major depressive disorder with psychotic features (F32.3, F33.3).

Antipsychotics Exposure Definition

Antipsychotic exposure was defined based on prescription records documented in the EHR. A total of 18 U.S. Food and Drug Administration (FDA)-approved antipsychotics were included in the analysis. These consisted of 14 second-generation (atypical) antipsychotics: clozapine, olanzapine, aripiprazole, risperidone, paliperidone, quetiapine, ziprasidone, lurasidone, asenapine, iloperidone, brexpiprazole, cariprazine, lumateperone, and pimavanserin; and 4 first-generation (typical) antipsychotics: haloperidol, chlorpromazine, fluphenazine, and perphenazine. Because some antipsychotic medications are available in both oral and injectable formulations, exposure was defined irrespective of route of administration. Oral and short-acting injectable formulations were included, while long-acting injectable (LAI) formulations were excluded.

Outcome Measures and Statistical analysis

Outcome measurements include incident cases, incidence proportion, incidence rate, prevalent cases, and prevalence.

Annual incidence and prevalence of each antipsychotic medication were estimated using the TriNetX analytic platform for each calendar year from 2016 through 2025. The lookback period was defined as the entirety of a patient’s available medical record within the TriNetX system, consistent with the platform’s default settings.

Incident antipsychotic use was defined as the first recorded prescription of a given antipsychotic within a calendar year among patients who had no prior record of that medication during the lookback period but had documented activity during the same year. Incidence proportions were calculated as the number of new cases divided by the number of eligible patients at risk for that medication in the specified year. Incidence rates were calculated by the platform using person-time denominated in person-days and were subsequently standardized and expressed as cases per 100,000 person-years to improve interpretability given the relatively low absolute event counts. Each individual could contribute incident cases to multiple antipsychotics but only once per medication.

Annual prevalence was defined as the proportion of patients with at least one recorded prescription for a given antipsychotic during each calendar year, regardless of prior use. All outcome measures were stratified by age group, sex, and race.

Basic descriptive statistics were used to summarize annual incidence and prevalence estimates overall and strata. Continuous variables were summarized using counts and proportions. To evaluate temporal trends in antipsychotic initiation, joinpoint regression analysis was performed on annual incidence rates. Joinpoint models were used to identify statistically significant changes in trends over time and to estimate annual percent change (APC) for each identified segment. The average annual percent change (AAPC) was calculated to summarize overall temporal trends across the study period. Analyses were conducted using weighted regression, with weights derived from the standard errors of the incidence rates, thereby accounting for variability in rate estimates across years.

Software and visualization

Initial analyses were conducted using the TriNetX analytics environment. Original TriNetX data were exported to Microsoft® Excel® (Version 2602) for additional data processing, descriptive analyses, and figure generation. Temporal trend analyses were performed using Joinpoint Regression Software (Version 6.0.1; National Cancer Institute)20.

Results

Baseline characteristics

The study cohort included 950,573 patients with dementia (Table 1). The mean age was 82 years (SD 6; range 65–90). Females comprised most of the cohort (56.43%). Most patients were non-Hispanic or Latino (69.62%), although ethnicity was undocumented for 26.01% of the cohort. By race, White patients accounted for 65.62% of the population, followed by Black or African American (12.29%) and Asian patients (5.68%). Geographically, 89% of the population are from USA, with majority from South (27%) and Middle Atlantic (20%) areas.

Table 1.

Baseline demographic characteristics of the study cohort (N = 950,573)

Characteristic Value
Age, years
 Mean (SD) 82 (6)
 Minimum–Maximum 65–90
Sex, n (%)
 Male 413,689 (43.52%)
 Female 536,408 (56.43%)
 Unknown 475 (0.05%)
Ethnicity, n (%)
 Not Hispanic or Latino 661,789 (69.62%)
 Hispanic or Latino 41,540 (4.37%)
 Unknown 247,244 (26.01%)
Race, n (%)
 White 623,766 (65.62%)
 Black or African American 116,825 (12.29%)
 Asian 53,993 (5.68%)
 American Indian or Alaska Native 3,422 (0.36%)
 Native Hawaiian or Other Pacific Islander 3,992 (0.42%)
 Other race 24,430 (2.57%)
 Unknown 124,145 (13.06%)
Geographic distribution, n (%)
US Regions
 New England 84,167 (9%)
 Middle Atlantic 188,638 (20%)
 East North Central 73,080 (8%)
 West North Central 78,615 (8%)
 South Atlantic 259,155 (27%)
 East South Central 20,480 (2%)
 West South Central 65,577 (7%)
 Mountain 28,858 (3%)
 Pacific 58,703 (6%)
Ex-US Regions
 Ex-US 106,926 (11%)

Annual prevalence of antipsychotics

Table 2 shows that across 2016–2025, the annual prevalence of all 18 antipsychotics increased, with the highest prevalence consistently observed for quetiapine and haloperidol. Within 10 years, quetiapine use rose 5.1-fold (3,605 per 100,000 to 18,432 per 100,000), while olanzapine exhibited 7.7-fold increase (1,205 to 9,264 per 100,000). Mid-range prevalence agents such as aripiprazole increased from 580 to 2,423 per 100,000, whereas newer or less commonly used agents like brexpiprazole remained comparatively low but increased significantly over time (10.7-fold). Lumateperone was marketed in 2019 and remained 0 through 2020 but increased modestly thereafter (2 in 2021 to 16 in 2025). For relative increase, cariprazine had the highest change (66-fold) and the second was pimavanserin (19-fold).

Table 2.

Annual prevalence’s of 18 antipsychotics in dementia patients (per 100,000), 2016–2025

2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
Quetiapine 3,605 4,455 5,344 6,337 7,710 9,135 11,048 13,055 15,414 18,432
Haloperidol 4,745 5,536 6,420 7,319 8,286 9,308 10,682 11,994 13,802 15,829
Olanzapine 1,205 1,475 1,813 2,249 2,826 3,542 4,482 5,639 7,246 9,264
Risperidone 1,326 1,536 1,765 2,014 2,293 2,603 2,974 3,351 3,880 4,513
Aripiprazole 580 669 793 919 1,060 1,203 1,407 1,653 1,998 2,423
Ziprasidone 262 318 380 446 540 630 763 909 1,130 1,322
Brexipiprazole 107 121 146 172 190 203 228 324 655 1,147
Chlorpromazine 199 232 262 293 320 356 398 434 467 520
Pimavanserin 21 76 111 148 189 217 245 275 342 403
Lurasidone 52 66 80 93 103 114 131 153 185 221
Clozapine 56 58 69 74 83 91 98 108 112 132
Cariprazine 2 5 10 18 25 36 50 71 97 131
Fluphenazine 19 21 22 25 26 27 31 47 76 112
Perphenazine 51 59 62 62 62 64 64 66 69 72
Paliperidone 14 15 17 19 22 24 28 30 34 41
Asenapine 13 15 17 15 15 16 16 17 17 21
Lumeteperone 0 0 0 0 0 2 2 5 11 16
Iloperidone 7 8 8 7 7 7 5 6 5 6

Stratified distributions

Figure 1 shows stratified distributions of antipsychotic use by age, sex, and race in 2025. Panel 1A illustrates that the largest shares of antipsychotics were generally observed in the 75–84 age groups. For the most commonly prescribed medications (quetiapine, haloperidol, and olanzapine), the 80–84 age group accounted for the biggest share of users, followed by the ≥85 group. In contrast, several second-generation antipsychotics showed relatively greater concentration in younger-old age strata, e.g. aripiprazole use was highest among patients aged 70–79, with a smaller proportion among those aged ≥85 (13.23%). Similar patterns were observed for pimavanserin and clozapine, for which more than half of users were aged 70–79 years. Overall, patients aged 65–69 years contributed the smallest proportion of use across most antipsychotics (mostly < 20%) in the dementia cohort.

Figure 1. Stratified distributions of antipsychotic use among adults with dementia in 2025.

Figure 1.

Figure 1.

Figure 1.

Panel A shows the age distribution of prevalent users for each of the 18 antipsychotics, stratified into 65–69, 70–74, 75–79, 80–84, and ≥85 years. Panel B shows the sex distribution of prevalent users (female, male, and unknown sex). Panel C shows the racial distribution of prevalent users categorized as White, Black or African American, Asian, American Indian or Alaska Native, Native Hawaiian or Other Pacific Islander, Other Race, and Unknown Race. Values are presented as proportions of prescribed users for each antipsychotic.

Gender distributions (Panel 1B) varied by medication. The female proportion was high for perphenazine (73.59% female), lurasidone (65.6% female) and aripiprazole (63.55% female). Quetiapine (55.51% female) and haloperidol (53.18% female) also had a modest female predominance. In contrast, male predominance was observed for chlorpromazine (73.38%), pimavanserin (58.96%), iloperidone (54.55%), ziprasidone (51.75%), and clozapine (51.63%).

For race distribution (Panel 1C), White patients constituted the largest share (47.59% to 81.02%), mostly > 60%. Black patients accounted for a substantial minority for several agents, including 14.53% for haloperidol and 13.35% for olanzapine, but were comparatively higher for fluphenazine (24.78%) and iloperidone (23.81%). Notably, the proportion with unknown race was sizable for some agents (e.g., 23.57% for chlorpromazine) and was particularly high for clozapine (37.43%). Paliperidone seems to be disproportionally used among Native Hawaiian and Pacific Islanders (5.56%).

Antipsychotic initiation and temporal trends

Initiation rates of antipsychotics increased substantially across most agents (Table 3). Throughout the study period, quetiapine initiations rose nearly 5-fold and haloperidol more than tripled. Olanzapine and risperidone also had steady and pronounced growth, paralleling their increasing prevalence over time. Brexpiprazole demonstrated a marked acceleration in uptake after 2022. Others also showed gradual increases in initiation, although at lower absolute rates. In contrast, initiation of perphenazine has decreased over time.

Table 3.

Annual initiation proportions of 18 antipsychotics (per 100,000), 2016–2025

2016 2017 2018 2019 2020 2021 2022 2023 2024 2025
Quetiapine 1,303 1,589 1,773 2,061 2,709 3,116 3,876 4,379 5,332 6,289
Haloperidol 1,465 1,638 1,832 2,011 2,299 2,632 3,186 3,493 4,215 4,853
Olanzapine 468 541 657 834 1,082 1,415 1,808 2,226 2,990 3,865
Risperidone 429 485 531 601 707 818 910 994 1,177 1,394
Brexipiprazole 20 25 36 39 33 30 39 111 347 539
Aripiprazole 147 162 191 213 242 274 325 360 459 531
Ziprasidone 89 107 121 138 178 211 262 298 389 449
Pimavanserin 21 57 49 56 70 72 72 74 107 113
Chlorpromazine 56 61 62 63 67 76 86 91 100 109
Fluphenazine 6 5 4 6 4 4 6 20 36 47
Clozapine 13 12 19 17 21 23 22 25 24 37
Lurasidone 21 19 18 20 18 19 22 22 31 34
Cariprazine 2 4 6 9 9 12 15 21 26 30
Paliperidone 3 3 4 4 5 5 7 4 8 9
Perphenazine 13 15 11 12 7 8 6 6 7 5
Lumeteperone 0 0 0 0 0 2 2 4 6 5
Asenapine 3 5 4 2 2 2 2 2 2 3
Iloperidone 2 2 2 2 2 2 2 2 2 3

With corresponding data presented in Tables 2 and 3, Figure 2 highlights the prescribing changes of five antipsychotics with the highest prevalence or initiation in 2025.

Figure 2. Temporal trends in prevalence and initiation of the 5 most commonly used antipsychotics among adults with dementia, 2016–2025.

Figure 2.

Panel A shows annual prevalence (per 100,000) of quetiapine, haloperidol, olanzapine, risperidone, and aripiprazole. Panel B shows annual initiation proportions (per 100,000) of quetiapine, haloperidol, olanzapine, risperidone, and brexpiprazole.

Joinpoint regression analyses revealed distinct trends in antipsychotic initiation in dementia 2016–2025 (eTable 1 in Supplement 1, Figure 3). Among high-incidence antipsychotics (Panel A), quetiapine, olanzapine, haloperidol, and risperidone demonstrated consistently rising initiation rates with significant acceleration after 2023. Olanzapine exhibited the steepest overall increase (AAPC 30.51%, p < 0.001). In the intermediate-incidence group (Panel B), aripiprazole and ziprasidone showed significant and progressive increases with higher APCs in the later years, whereas chlorpromazine and pimavanserin displayed mild and continuous increasing trend. Brexpiprazole demonstrated a pronounced inflection point after 2022, with a dramatic surge in initiation (APC 145.43%, p < 0.001), visually evident as a sharp upward trajectory. Low-incidence antipsychotics (Panel C) exhibited heterogeneous patterns: fluphenazine and lurasidone transitioned from mild decreasing to significant growth after 2021, cariprazine showed rapid early growth followed by sustained increases, clozapine showed steady increases without detectable joinpoints. In contrast, very low-incidence antipsychotics (Panel D) remained infrequently initiated overall; paliperidone and lumateperone (shorter observation time 2021–2025) demonstrated consistent growth, whereas perphenazine showed a significant long-term decline (AAPC −9.86%, p = 0.01).

Figure 3. Temporal trends in initiation rates of 18 antipsychotics among adults with dementia, 2016–2025.

Figure 3.

Antipsychotics are grouped according to initiation rate level (per 100,000 person-years) in 2025. Panel A shows high-incidence antipsychotics (quetiapine, haloperidol, olanzapine, and risperidone). Panel B shows intermediate-incidence antipsychotics (aripiprazole, brexpiprazole, chlorpromazine, pimavanserin, and ziprasidone). Panel C shows low-incidence antipsychotics (cariprazine, clozapine, fluphenazine, lurasidone, and lumateperone). Panel D shows very low-incidence antipsychotics (asenapine, iloperidone, paliperidone, and perphenazine). Solid lines represent fitted joinpoint regression models. Points represent observed annual incidence rates.

Sensitivity Analysis

As a sensitivity analysis, we repeated all primary analyses using data restricted to the network US Collaborative Network, with 66 US HCO(s) queried and 66 HCO(s) responded. The final cohort included 844,408 patients who matched the same query criteria. Patterns of antipsychotic initiation and prevalence in the U.S. cohort were highly consistent with those observed in the global analysis, with quetiapine and haloperidol remaining the most frequently prescribed agents and SGAs driving the overall increase in use. The magnitude of increase was comparable across datasets. Temporal trends similarly demonstrated late-period acceleration for several high-utilization medications, with preservation of joinpoint patterns and overall trajectory. Minor differences in absolute rates were observed, likely reflecting regional variation in prescribing practices. These findings support the robustness of the primary results and suggest that observed trends are not driven by geographic heterogeneity within the TriNetX network. Detailed results are presented in eTable 2–4 in Supplement 1.

Discussion

In this large EHR-based cohort from a predominantly US population of older adults with dementia, we observed pronounced and accelerating growth in antipsychotic use from 2016 to 2025, driven largely by increased initiation rather than prolonged use. Prescribing patterns varied by age, sex, and race, and were dominated by select agents - mainly quetiapine, olanzapine, risperidone, and haloperidol. Joinpoint analyses revealed distinct inflection points and late-period accelerations for several medications, including newer agents and agents which received FDA approval for the treatment of agitation associated with dementia during this period, indicating a rapidly evolving prescribing landscape. Overall, these findings raise important questions regarding clinical decision-making, safety, and adherence to guideline-recommended care in this high-risk population.

The reasons for the increasing use of antipsychotics in dementia are likely multifactorial. Previous studies suggest antipsychotics are commonly used to manage behavioral and psychological symptoms of dementia (BPSD), with the prevalence ranging from 6–17% in community settings to 12–59% in nursing homes16,21,22. With growing recognition of these symptoms, which affect a large proportion of dementia patients and are closely associated with caregiver burden, institutionalization, and increased healthcare utilization, use of antipsychotics has increased over time, particularly for long-term exposure, driven largely by SGAs and declining use of FGAs9. On the other hand, the prevalence of dementia continues to increase with population aging, the absolute number of patients presenting with severe neuropsychiatric symptoms may also be rising, creating greater demand for symptom management. Previous studies show quetiapine is the most frequently prescribed agent, while risperidone remains widely used and haloperidol is the most common typical antipsychotic13,14. Consistent with historical findings, our study shows that quetiapine and haloperidol remain the most frequently initiated antipsychotics in patients with dementia. The rate of prescription of these two agents accelerated from 2023–2025, reflecting an increasing need for sedation and agitation control. Olanzapine prescriptions demonstrated a greater relative increase (AAPC 30.51, p < 0.001), with the steepest incidence increase occurring between 2024–2025. Practical considerations such as a relatively fast onset, scheduled and as needed dosing regimen, parenteral formulations, and sedating properties may contribute to its utilization, despite limited regulatory approval for dementia-related symptoms. Its relatively higher metabolic burden over other agents is an area of concern, however. Finally, increasing clinical complexity of patients with dementia, including coexisting psychiatric disorders, may have further contributed to increased antipsychotic prescribing. However, because our study did not evaluate prescribing indications, the relative contribution of these factors cannot be determined.

Our findings should also be interpreted in the context of care setting differences. A Texas Medicare study found declining antipsychotic use in nursing facilities16, but TriNetX does not reliably capture care setting. The increase observed in our study may therefore reflect rising use in outpatient or community settings, which could help explain differences from prior reports.

Demographically, a Canadian study found that although women account for most treated patients, male sex is associated with higher likelihood of use23. Our study showed trends toward sex-specific predominance for a few agents in males: chlorpromazine (73.38%), pimavanserin (58.96%), iloperidone (54.55%), ziprasidone (51.75%), clozapine (51.63%); but the rest 13 agents are more used in females. This is a novel finding of this study. Several factors could have contributed to these sex differences. Men more often exhibit agitation, aggression, and disinhibited behaviors24,25, which may contribute to greater use of the more sedating agents chlorpromazine, ziprasidone, and clozapine. Also, Dementia with Lewy bodies (DLB) and Parkinson’s disease dementia (PDD) are more common in men, and clozapine and pimavanserin are often favored in these conditions because of their lower risk of extrapyramidal adverse effects26. Finally, women have a higher risk of QTc prolongation and torsades de pointes than men27. As ziprasidone and iloperidone are strongly associated with QTc prolongation28, they may be prescribed less often to older women.

Older age is generally associated with increased antipsychotic prescribing9. This was reflected in our findings: use in the dementia cohort had peak proportions in patients aged 75–84 years for commonly prescribed agents such as quetiapine, haloperidol, and olanzapine. Previous studies showed that quetiapine is often preferred in the oldest patients due to its sedating properties, whereas risperidone and olanzapine may be used more cautiously in the very elderly because of their higher cerebrovascular risk21,29. Our results demonstrated a notable quetiapine concentration in the 80–84 (27.72%) and ≥85 (24.22%) age groups; olanzapine and risperidone showed similar patterns. In contrast, several SGAs (e.g., aripiprazole, lurasidone, cariprazine) were relatively more concentrated in younger 70–79 age group. Across most agents, patients aged 65–69 years consistently accounted for the smallest share of use, underscoring a pronounced age gradient in prescribing patterns. Several factors may contribute to this pattern. First, BPSD, especially agitation, aggression, psychosis, and sleep disturbances, tend to become more frequent and severe as cognitive impairment progresses30, increasing the need of pharmacologic intervention in the older patient group. Second, patients in the older age groups are more likely to reside in long-term care or nursing home settings, where neuropsychiatric symptoms are common and antipsychotics are often used when nonpharmacologic approaches are insufficient. Third, greater caregiver burden, safety concerns, and risks of injury related to agitation or psychosis may prompt treatment in advanced age despite known safety concerns31. However, because our study lacked information on symptom severity, care setting, and prescribing indications, these explanations remain speculative.

In terms of racial and ethnic disparities, consistent with previous reports, we observed that White patients comprised the majority of users across most agents, while Black patients represented a substantial minority and were disproportionately represented for certain medications (e.g., fluphenazine 24.78%, iloperidone 23.81%, haloperidol 14.53%). Several factors may contribute to these differences. Prior studies have noted higher antipsychotic use among Black individuals in community settings and greater use among Hispanic and Asian individuals in nursing facilities16,32; and Black patients are prescribed haloperidol and fluphenazine, and FGAs generally, at higher rates than White patients33,34. Additionally, prior disparities in mental health diagnosis and treatment have raised concerns about racial differences in antipsychotic prescribing. Our data cannot determine the reasons for the observed differences.

FGAs use remained high among dementia patients, particularly haloperidol and chlorpromazine. Despite a decline in perphenazine (AAPC −9.86%, p = 0.01), haloperidol remained the second most prescribed agent, with initiation rates increasing over 3-fold (1,465 to 4,853 per 100,000, 2016–2025). Chlorpromazine use nearly tripled, and fluphenazine showed marked post-2021 acceleration (APC 96.48%). These trends are concerning given known risks in dementia and may reflect reliance on these agents for acute agitation, underscoring the need for safer, guideline-concordant prescribing.

In May 2023, the FDA approved brexpiprazole as the first drug specifically to treat agitation associated with Alzheimer’s dementia18. We reported an APC of 145.43 for 2022–2025, indicating that the initiation of brexpiprazole prescriptions surged dramatically, after the FDA approval for this indication. Likewise, lumateperone was marketed in 2019 then had uptake in 2021, a real-world adoption example among new and low-incidence agents.

Our study has both strengths and limitations. Instead of class level summary, we presented drug-by-drug trends across 18 antipsychotics over 10 years duration and demographic heterogeneity. Concurrent prevalence and initiation trajectories are both included, allowing a clearer inference about what is driving increasing antipsychotic utilization. Trend analysis identified inflection points and highlighted that the increases in antipsychotic prescribing have not been linear during this time. Finally, clinically interpretable demographic stratification at the drug level offered readers the opportunity for actionable insight into “who gets what prescription” beyond overall prescription rates. Limitations included retrospective observational study nature that prohibits causal relationship conclusion. Information on care setting, medication dosing, treatment duration, or adherence were not included. Although the TriNetX Research Network includes HCOs from multiple countries, the study population was predominantly derived from US HCOs (89%). Furthermore, country-level identifiers and detailed regional information for non-US participants were not available within the analytic dataset. Consequently, the findings primarily reflect prescribing patterns within the US healthcare environment and should not be interpreted as representative of individual non-US countries or global prescribing trends.

Conclusions and Implications

In this large-scale multinational real-world retrospective study, we provide a comprehensive, agent-specific characterization of prescribing and initiation trends of 18 antipsychotics among patients with dementia from 2016 to 2025. Our findings support that antipsychotic use in dementia increased substantially over the past decade, driven primarily by rising initiation of select antipsychotics, particularly quetiapine, haloperidol, olanzapine, and risperidone, with emerging contributions from newer agents such as brexpiprazole and lumateperone. These increases were not uniform across medications but displayed pronounced heterogeneity in growth magnitude, temporal inflection points, and demographic distributions by age, sex, and race.

Supplementary Material

Supplement 1

Funding:

Research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under Award Number U54 GM104942. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflict of interest: None

Conflict of Interest Statement:

All authors affirm that there are no relevant financial or non-financial conflicts to report.

Informed consent and patient details

The authors have NOT obtained informed consent from participants or their legal representatives. Only deidentified data were used.

Studies in Human

This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place.

Ethics committee approval was not required under relevant laws and institutional guidelines. As no direct human subject interaction occurred, the study was classified as non-human subjects research and exempt from West Virginia University IRB review.

References

  • 1.Garmany A, Terzic A. Global healthspan-lifespan gaps among 183 World Health Organization member states. JAMA Netw Open. 2024;7(12):e2450241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Du L, Yu L, Wang T, Boyle PA, Barnes LL, Marquez DX, et al. Lifetime risk of incident dementia and incident mild cognitive impairment in older adults. Alzheimers Dement. 2026;22(2):e71173. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Fang M, Hu J, Weiss J, Knopman DS, Albert M, Windham BG, et al. Lifetime risk and projected burden of dementia. Nat Med. 2025;31(3):772–776. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Watt J, Porter J, Tavilsup P, Chowdhury M, Hatch S, Ismail Z, et al. Guideline recommendations on behavioral and psychological symptoms of dementia: a systematic review. J Am Med Dir Assoc. 2024;25(5):837–846.e21. [DOI] [PubMed] [Google Scholar]
  • 5.Toot S, Swinson T, Devine M, Challis D, Orrell M. Causes of nursing home placement for older people with dementia: a systematic review and meta-analysis. Int Psychogeriatr. 2017;29(2):195–208. [DOI] [PubMed] [Google Scholar]
  • 6.Ballard C, Doherty P, Ismail Z, Corbett A, Cummings JL. Symptomatic treatment for Alzheimer disease: current evidence and future directions. Nat Rev Neurol. Published online September 1, 2026. doi: 10.1038/s41582-026-01260-5 [DOI] [PubMed] [Google Scholar]
  • 7.Centers for Medicare & Medicaid Services. Antipsychotic medication use data report: 2021 Q4. Published July 29, 2022. Accessed June 5, 2026. https://www.cms.gov/files/document/antipsychotic-medication-use-data-report-2021q4-updated-07292022.pdf
  • 8.Yang AW, Leng M, Arbanas JC, Tseng CH, Fendrick AM, Sarkisian C, et al. Trends in antipsychotic prescribing among community-dwelling older adults with dementia, 2010–2018. Health Aff Scholar. 2025;3(2):qxaf021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Olfson M, Xie F, Bushnell G, Hua J, Miles J, Crystal S. Antipsychotic medication use by older adults. JAMA Psychiatry. 2026;83(2):214–216. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Bezabhe WM, Radford J, Salahudeen MS, et al. Ten-Year Trends in Psychotropic Prescribing and Polypharmacy in Australian General Practice Patients with and without Dementia. J Clin Med. 2023;12(10):3389. Published 2023 May 10. doi: 10.3390/jcm12103389 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Sánchez-Soblechero A, Gil M, Rojo JM, Muñiz R, Hoyos-Alonso MC, Olazarán J. Trends of antipsychotic prescribing for people with dementia in Spain. J Alzheimers Dis. 2025;107(1):104–111. doi: 10.1177/13872877251359682 [DOI] [PubMed] [Google Scholar]
  • 12.Kuroda N, Hamada S, Sakata N, et al. Antipsychotic use and related factors among people with dementia aged 75 years or older in Japan: A comprehensive population-based estimation using medical and long-term care data. Int J Geriatr Psychiatry. 2019;34(3):472–479. doi: 10.1002/gps.5041 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Schneider C, O’Mahony D, Byrne S, Dalleur O, Spinewine A, Huibers CJA, et al. Prescribing and deprescribing of atypical antipsychotics in older multimorbid patients. BMC Geriatr. 2025;25(1):962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Olfson M, King M, Schoenbaum M. Antipsychotic treatment of adults in the United States. J Clin Psychiatry. 2015;76(10):1346–1353. [DOI] [PubMed] [Google Scholar]
  • 15.Smith H, Petersen I, Hayes JF, Jordan KP, Boman M, Banerjee S, et al. Antipsychotic prescriptions in people with dementia in primary care: a cohort study investigating adherence of dose and duration to UK clinical guidelines. Lancet Psychiatry. 2025;12(10):758–767. [DOI] [PubMed] [Google Scholar]
  • 16.Kim Y, Krause T, Samper-Ternent R, Teixeira A. Antipsychotic use in older adults with dementia: community and nursing facility trends in Texas, 2015–2020. J Am Med Dir Assoc. 2025;26(3):105463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Bony H, Lloyd RA, Hotham ED, et al. Differences in the prescribing of potentially inappropriate medicines in older Australians: comparison of community dwelling and residential aged care residents. Sci Rep. 2020;10(1):10170. Published 2020 Jun 23. doi: 10.1038/s41598-020-66991-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Grossberg GT, Sanford A, Montano CB, Porsteinsson AP, Scanland S, Worz C, et al. A US-based practitioner’s guide to diagnosis, evaluation, and evidence-based treatment of agitation in Alzheimer’s dementia: expert consensus recommendations. Postgrad Med. 2025;137(6):469–485. [DOI] [PubMed] [Google Scholar]
  • 19.Komossa K, Rummel-Kluge C, Hunger H, Schmid F, Schwarz S, Duggan L, et al. Olanzapine versus other atypical antipsychotics for schizophrenia. Cochrane Database Syst Rev. 2010;(3):CD006654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.National Cancer Institute. Joinpoint Regression Program. Surveillance Research Program. 2026. [Google Scholar]
  • 21.Tampi R, Jeste D. Dementia is more than memory loss: neuropsychiatric symptoms of dementia and their nonpharmacological and pharmacological management. Am J Psychiatry. 2022;179(8):528–543. [DOI] [PubMed] [Google Scholar]
  • 22.Mühlbauer V, Möhler R, Dichter M, Zuidema S, Köpke S, Luijendijk H. Antipsychotics for agitation and psychosis in people with Alzheimer’s disease and vascular dementia. Cochrane Database Syst Rev. 2021;12(12):CD013304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Maclagan LC, Maxwell CJ, Harris DA, Campitelli MA, Diong C, Lapane KL, et al. Sex differences in antipsychotic and benzodiazepine prescribing patterns: a cohort study of newly admitted nursing home residents with dementia in Ontario, Canada. Drugs Aging. 2020;37(11):817–827. [DOI] [PubMed] [Google Scholar]
  • 24.Ferretti MT, Iulita MF, Cavedo E, Chiesa PA, Schumacher Dimech A, Santuccione Chadha A, Baracchi F, Girouard H, Misoch S, Giacobini E, Depypere H, Hampel H, & Women’s Brain Project and the Alzheimer Precision Medicine Initiative (2018). Sex differences in Alzheimer disease - the gateway to precision medicine. Nature reviews. Neurology, 14(8), 457–469. 10.1038/s41582-018-0032-9 [DOI] [PubMed] [Google Scholar]
  • 25.Silvestri C, Almici V, Libri I, Mattioli I, Cosseddu M, Turrone R, Rivolta J, Grassini C, Caratozzolo S, Alberici A, Marengoni A, Pilotto A, Borroni B, Padovani A, & Benussi A. (2024). Sex Differences in the Severity and Progression of Neuropsychiatric Symptoms Across Different Dementia Types. Neurology. Clinical practice, 14(3), e200299. 10.1212/CPJ.0000000000200299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Kaila LV, Ikeda M, Sultana J, Chouliaras L, O’Brien JT, & Taylor JP. (2025). The evolving therapeutic landscape of dementia with Lewy bodies. The Lancet. Neurology, 24(12), 1038–1052. 10.1016/S1474-4422(25)00323-0 [DOI] [PubMed] [Google Scholar]
  • 27.Tisdale JE, Chung MK, Campbell KB, Hammadah M, Joglar JA, Leclerc J, Rajagopalan B, & American Heart Association Clinical Pharmacology Committee of the Council on Clinical Cardiology and Council on Cardiovascular and Stroke Nursing (2020). Drug-Induced Arrhythmias: A Scientific Statement From the American Heart Association. Circulation, 142(15), e214–e233. 10.1161/CIR.0000000000000905 [DOI] [PubMed] [Google Scholar]
  • 28.Melo L, Pillai A, Kompella R, Patail H, & Aronow WS. (2024). An Updated Safety Review of the Relationship Between Atypical Antipsychotic Drugs, the QTc Interval and Torsades de Pointe As: Implications for Clinical Use. Expert opinion on drug safety, 23(9), 1127–1134. 10.1080/14740338.2024.2392002 [DOI] [PubMed] [Google Scholar]
  • 29.Yunusa I, Alsumali A, Garba AE, Regestein QR, Eguale T. Assessment of reported comparative effectiveness and safety of atypical antipsychotics in the treatment of behavioral and psychological symptoms of dementia: a network meta-analysis. JAMA Netw Open. 2019;2(3):e190828. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wiels WA, Wittens MMJ, Zeeuws D, Baeken C, Engelborghs S. Neuropsychiatric Symptoms in Mild Cognitive Impairment and Dementia Due to AD: Relation With Disease Stage and Cognitive Deficits. Front Psychiatry. 2021;12:707580. Published 2021 Aug 17. doi: 10.3389/fpsyt.2021.707580 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.American Psychiatric Association Council on Geriatric Psychiatry. Resource Document on the Use of Antipsychotic Medications to Treat Behavioral Disturbances in Persons with Dementia. Washington, DC: American Psychiatric Association; 2014. Accessed August 26, 2026. https://www.psychiatry.org/getattachment/8a9eee0c-877e-4b51-9248-535ab6f4b565/Resource-2014-Antipsychotic-Medications-Dementia.pdf [Google Scholar]
  • 32.Shireman TI, Fashaw-Walters S, Zhang T, Zullo AR, Gerlach LB, Coe AB, et al. Federal nursing home policies on antipsychotics had similar impacts by race and ethnicity for residents with dementia. J Am Med Dir Assoc. 2023;24(9):1283–1289.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Cook TB, Reeves GM, Teufel J, Postolache TT. Persistence of racial disparities in prescription of first-generation antipsychotics in the USA. Pharmacoepidemiol Drug Saf. 2015;24(11):1197–1206. doi: 10.1002/pds.3819 [DOI] [PubMed] [Google Scholar]
  • 34.Medina C, Akinkunmi A, Bland N, et al. Differences in schizophrenia treatments by race and ethnicity-analysis of electronic health records. Schizophrenia (Heidelb). 2024;10(1):48. Published 2024 Apr 26. doi: 10.1038/s41537-024-00470-4 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1

RESOURCES