Abstract
Background
While central nervous system (CNS)-active medication polypharmacy is prevalent among adults with early-onset dementia (EOD), its impact on nursing home (NH) admission is unclear.
Objective
To examine whether CNS-active medication polypharmacy is associated with risk for NH admission among patients with EOD.
Methods
A retrospective cohort study using 2012–2021 MarketScan Databases was conducted with adults aged 30 to 64 years diagnosed with Alzheimer's disease and related dementias (ADRD). CNS-active medication polypharmacy was assessed during a fixed 6-month period after an indexed ADRD diagnosis and defined as concurrent use of ≥3 CNS-active medications on the American Geriatrics Society Beers Criteria list that overlapped for ≥30 consecutive days. Cox hazards models were used to estimate the hazard ratio (HR) of NH admission between patients with versus without CNS-active medication polypharmacy, with adjustment for baseline covariates via inverse probability of treatment weighting.
Results
Of the 18,029 patients with EOD identified, those with CNS-active medication polypharmacy (versus those without) had an increased risk of NH admission (HR, 1.34 [95% CI, 1.16–1.56]). The risk was pronounced in patients aged >50 years, male patients, patients with a diagnosis of Alzheimer's disease or senile dementia, and patients having baseline use of anti-dementia medications.
Conclusions
In this commercially insured population of adults with EOD, receipt of CNS-active medication polypharmacy was associated with risk for NH admission. The finding, however, should be interpreted with caution, as it may reflect differences in the underlying disease severity and clinical indications treated by the studied CNS-active medications.
Keywords: Alzheimer's disease, central nervous system drugs, early-onset dementia, nursing home admission, polypharmacy
Introduction
Alzheimer's disease and related dementias (ADRD), the most common neurodegenerative disease, currently affects over 7 million Americans, with the number expected to double by 2060. 1 Although ADRD is mostly prevalent among individuals aged 65 years or older (i.e., late-onset dementia [LOD]), it can develop among adults before the age of 65 years (i.e., early-onset dementia [EOD]). EOD accounts for approximately 200,000 of the 7 million Americans with ADRD in the United States (US). 2 Globally, the estimated prevalence of early-onset was 119 per 100,000 person-years. 3
Use of central nervous system (CNS)-active medications is prevalent among patients with EOD. In a multi-year, cross-sectional study using US commercially insured data, we found that one in four patients with EOD simultaneously used three or more CNS-active medications (termed CNS-active medication polypharmacy) listed on the American Geriatrics Society (AGS) Beers Criteria for longer than 30 days between 2013 and 2021. 4 This high prevalence of CNS-active medication polypharmacy among patients with EOD was nearly 1.8-fold as high as the prevalence (13.9% in 2018) of CNS-active medication polypharmacy observed in patients with LOD. 5 In addition, the annual crude prevalence of CNS polypharmacy among adults with EOD (21.2%–25.0%) was higher than for chronic pain (5.1%–5.9%), depression (14.8%–21.7%), or epilepsy (20.0%–22.3%) from 2013 to 2021. 4 Together, these findings indicate that patients with EOD may be at high risk for receiving CNS-active medication polypharmacy, raising safety concerns associated with this potentially inappropriate prescribing practice.
To our knowledge, limited population-based research has studied the impact of CNS-active medication polypharmacy—a potential risk factor of nursing home (NH) admission among patients with EOD. Prior studies have largely focused on patients with LOD,6,7 but those findings cannot be generalized to patients with EOD due to differences in dementia symptoms, 8 dementia progression, 9 and dementia-related neuropsychiatric symptoms. 10 Compared with patients with LOD, those with EOD tend to present changes in personality and behaviors (versus memory change), 8 have faster rate of cognitive decline,9,10 and are more likely to develop seizures 11 and neuropsychiatric symptoms, such as depression (41% versus 26%), and anxiety (28% versus 14%), 10 factors that independently or collectively contribute to the risk of NH admission. Prior studies are also limited in studying a single CNS drug class (e.g., antipsychotics or antidepressants) as a predictor of NH admission,6,7 with few data describing the impact of using multiple CNS-active medications concurrently for patients with dementia. The lack of such data may hinder doctors’ and caregivers’ ability to provide early intervention to delay or prevent costly NH admission for patients with EOD.
While our previous cross-sectional study demonstrated a high prevalence of CNS-active medication polypharmacy among US commercially insured patients with EOD between 2013 and 2021, 4 a question remains understudied regarding the risk of NH admission in relation to CNS-active medication polypharmacy. Thus, we conducted a retrospective cohort study to examine the association between use of CNS-active medication polypharmacy and subsequent risk for NH admission among US commercially insured patients diagnosed with ADRD between 30 and 64 years.
Methods
Study design and data source
We conducted a retrospective cohort study using the Merative MarketScan Commercial Claims and Encounters database from January 1, 2012, to December 31, 2021. MarketScan contains healthcare administrative claims data of individuals and their dependents who receive employer-sponsored coverage from medium to large firms in the US. 12 Data describing individual demographics (e.g., age and sex), medical encounters in inpatient and outpatient settings, and prescription medications dispensed are available over time, allowing for the conduct of longitudinal studies. We used prescription medications to measure CNS-active medication polypharmacy (key exposure). MarketScan also contains information on dates and places of services from over 90 healthcare settings, including skilled NHs and long-term care facilities. Such information allows for detecting when patients were admitted to and received services from an NH facility (key outcome measure). The Ohio State University institutional review and privacy boards exempted this study from review and waived the requirement for obtaining patient-informed consent because the data are de-identified. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Study sample
The study sample included individuals 30 to 64 years of age who were diagnosed with ADRD, defined as having at least two encounters from inpatient or outpatient claims with an International Classification of Diseases, Ninth or Tenth Revision Clinical Modification (ICD-9-CM/ICD-10-CM) code in any diagnostic position (Supplemental Table 1). These diagnostic codes were used in a prior study identifying patients with LOD 13 and are commonly used in various algorithms identifying patients with ADRD. 14 The requirement of at least two medical encounters with ADRD diagnostic codes was used to minimize the rule-out diagnosis of pseudodementia that is secondary to other conditions (e.g., psychiatric symptoms) or that mimics dementia symptoms but are potentially reversible (e.g., brain bleeding from traumatic brain injury), 15 which are common in early-onset dementia, 16 thereby increasing the possibility of identifying true cases. 17 Given that the typical diagnostic process for ADRD begins with a first clinical visit for physical examination followed by MRI or CT and neuropsychological assessment and then reaches a definitive diagnosis based on the clinical evaluations at the second visit, 18 we allowed patients to enter the cohort at the date of their second medical encounter with an ADRD diagnosis, at which the condition is verified (i.e., index date). Patients were further required to have continuous enrollment in medical and drug insurance for 12 months before the index date (baseline) to measure covariates (detailed in “Covariate Measures” section) and 6 months after the index date to measure CNS-active medication polypharmacy. The 12-month baseline was used to improve identification of true ADRD cases with at least 2 medical encounters for ADRD. Supplemental Figure 1 shows a schematic diagram of the study design.
To ensure complete capture of CNS-active medication polypharmacy over 6 months following the index date (exposure assessment), we excluded patients hospitalized during this period, as prescription medication information is unavailable during hospital stays. Patients were followed up from the end of the 6-month exposure assessment period until the first NH admission, medical insurance disenrollment, study end (December 31, 2021), or the end of a 1-year follow-up. To assess the risk of NH admission, we excluded patients who had an NH stay during the 12-month baseline and 6-month exposure assessment period (18 months). Figure 1 shows the study sample selection flowchart.
Figure 1.
Flowchart of the retrospective cohort study sample of individuals with early-onset dementia. ADRD: Alzheimer's disease and related dementias.
CNS-active medication polypharmacy
We used MarketScan pharmacy files to capture prescription CNS-active medications, including antipsychotics, antidepressants, benzodiazepines, anticonvulsants, opioids, and Z-drugs, all of which are listed in the AGS Beers Criteria as drugs considered potentially inappropriate when used concurrently (Supplemental Table 1). 19 For each CNS therapeutic class, we excluded injectable drugs primarily used in inpatient settings, for which prescription dispensing claims are unavailable. The primary key exposure, CNS-active medication polypharmacy, was defined as concurrent use of three or more CNS medications of interest, regardless of therapeutic classes, that overlapped for ≥30 consecutive days measured during a fixed 6-month exposure assessment period after the indexed ADRD diagnosis. For example, patients who concurrently received prescriptions from the same therapeutic category (e.g., SSRI and SNRI from antidepressants), along with oxycodone from opioids, for ≥30 days were classified as having CNS-active medication polypharmacy. A cutoff of ≥30 days to define CNS polypharmacy has been used in prior research.4,5 Defining the drug exposure as concurrent use of CNS-active medications for 30 or more days allows for minimizing the potential overestimation of prescribed CNS-active medications that were intended for use as needed. Similar to prior studies in assessing CNS-active polypharmacy,4,5 we identified daily exposure of each CNS-active medication according to the fill date and days’ supply of the drug. For each CNS medication, we allowed a gap of 7 days or fewer between prescription fills to account for potential delays in refills.
Outcome measures
The outcome of interest was NH admission, defined as an admission to a nursing facility or skilled nursing facility. A nursing facility assists with daily activities (e.g., bathing, dressing, and eating) mainly for a long-term period, while a skilled nursing home facility provides care focusing on rehabilitation and medically necessary services for a short-term period. 20 We selected this outcome because NH admission indicates that individual patients need professional assisted care in an NH facility due to clinical reasons, such as a decline in daily living skills or complex health conditions (e.g., ADRD) requiring ongoing care. 21 Consistent with a prior study, 22 we used the variable of place of service from the Merative MarketScan to determine whether individuals were admitted to an NH facility. 12
Covariate measures
Covariates were measured 12 months before the indexed ADRD diagnosis (baseline). Demographic characteristics included age (categorized into <44, 45–49, 50–54, 55–59, and 60–64 years), sex (male versus female), US geographical region (Northeast, Midwest, West, and South), and rurality (rural versus metropolitan residential area). Selected clinical indications for the CNS-active medications of interest included behavioral symptoms, anxiety disorders, depression, psychotic disorders, chronic pain, and seizure disorder. Other clinical conditions included cancer, neurodegenerative disorders, substance use disorder, diabetes, hypertension, cardiovascular diseases, pulmonary disorders, kidney disorders, liver disorders, gastrointestinal tract disorders, and fall injuries, all of which are risk factors associated with the use of the CNS medications of interest or NH institutionalization.13,23,24 We also measured healthcare utilization (use of any hospitalization or emergency department visit) and use of dementia medications (donepezil, galantamine, memantine, rivastigmine, and tacrine) at baseline. To assess the overall comorbidity, we calculated the total number of comorbid conditions (excluding the aforementioned clinical outcomes) classified using the Clinical Classification Software developed by the Healthcare Cost and Utilization Project. 24 Finally, we examined subtypes of the indexed ADRD diagnosis, including frontotemporal dementia, vascular dementia, Alzheimer's disease, senile dementia, and dementia unspecified,4,5 and included the year of the indexed ADRD diagnosis as a covariate to account for changes in diagnosis and management of ADRD over the 10-year study period. 25 Supplemental Table 1 shows the diagnosis codes for the disease, condition, or medication studied as covariates.
Statistical analysis
We reported baseline characteristics among patients with EOD, overall, and by patients with versus without CNS-active medication polypharmacy. Differences in baseline covariates between the exposure groups were assessed, with standardized mean differences (SMDs) > 0.100 indicating covariate imbalance.26,27 We adjusted potential differences in baseline variables between groups via inverse probability of treatment weighting (IPTW), calculated as the inverse of the propensity score for patients with CNS-active medication polypharmacy and the inverse of 1 minus the propensity score for patients without CNS-active medication polypharmacy. 28 The propensity score was calculated using a binary logistic regression model that included the receipt of CNS-active medication polypharmacy as the dependent variable and baseline covariates as independent variables. Each IPTW value was truncated at the first and 99th percentiles to reduce the influence of outliers on estimates. In the final analysis, we used a Cox proportional hazards model weighted by IPTW to estimate the adjusted hazard ratio (HR) and 95% confidence interval (CI) for the risk of NH admission between patients with versus without CNS-active medication polypharmacy. We examined the proportional hazards assumption using the Schoenfeld residuals test 29 and found that the assumption was satisfied.
Sensitivity and subgroup analyses
We conducted several sensitivity analyses to evaluate the robustness of our findings. First, we trimmed the propensity score at the 0.5th and 99.5th percentiles and at the 2nd and 98th percentiles.30,31 Second, we ran a Fine-Gray subdistribution hazard model to assess the robustness of our findings by accounting for insurance disenrollment as a competing event, because risk of NH admission is not observable when patients disenroll from health insurance. 32 Third, we examined the association between CNS-active medication polypharmacy and risk of NH admission in a sample that additionally included patients who had a hospital stay of less than 30 days (n = 3431) during the 6-month exposure assessment period. Fourth, we stratified the main analyses according to age groups (>50 versus ≤50 years), sex (male versus female), and key risk factors of NH admissions in adults with ADRD, 23 including subtypes of indexed ADRD diagnosis and baseline use of anti-dementia medications. Finally, we computed E-values to estimate the influence of unmeasured confounders on estimated associations. 33
None of the reported baseline covariates, key exposure, and outcome variables had missing values. All analyses were performed using SAS, version 9.4 (SAS Institute Inc., Cary, NC), and all tests were two-sided with the statistical significance set at p < 0.05.
Results
Characteristics of the identified sample
Of 27,727 enrollees aged < 65 years who had at least 2 medical claim encounters coded for ADRD between January 2012 and December 2021, we identified 18,029 eligible patients with EOD after applying eligibility criteria (Figure 1). Patients in the study sample had a mean age of 57.4 (standard deviation [SD] ± 6.6), with 13.6% aged 50 years or younger and 54.7% female (Table 1). During the follow-up, 52.6% of participants were censored due to medical insurance disenrollment, and there was no statistical difference in the proportions of patients lost to follow-up due to medical insurance disenrollment between those with versus without CNS-active polypharmacy (51.8% versus 52.9%, p = 0.11) (Supplemental Table 2).
Table 1.
Baseline characteristics of commercially insured adults with early-onset dementia, overall and by exposure to CNS-active medication polypharmacy.
| Characteristic | Overall sample, No. (%) | Receipt of CNS-active medication polypharmacy | SMD | ||
|---|---|---|---|---|---|
| Yes | No | Before IPTW | After IPTW | ||
| No. (%) | No. (%) | ||||
| Total, No. | 18,029 (100) | 3535 (100) | 14,494 (100) | ||
| Age, years | |||||
| Mean ± (SD) | 57.4 ± (6.6) | 56.8 ± (6.8) | 57.6 ± (6.6) | 0.112 | 0.002 |
| ≤44 | 1171 (6.5) | 259 (7.3) | 912 (6.3) | 0.116 | 0.047 |
| 45–49 | 1277 (7.1) | 304 (8.6) | 973 (6.7) | ||
| 50–54 | 2263 (12.6) | 487 (13.8) | 1776 (12.3) | ||
| 55–59 | 4335 (24.0) | 876 (24.8) | 3459 (23.9) | ||
| 60–64 | 8983 (49.8) | 1609 (45.5) | 7374 (50.9) | ||
| Sex | |||||
| Male | 8160 (45.3) | 1365 (38.6) | 6795 (46.9) | 0.071 | 0.003 |
| Female | 9869 (54.7) | 2170 (61.4) | 7699 (53.1) | ||
| Rurality | |||||
| Metropolitan | 13,189 (73.2) | 2575 (72.8) | 10,614 (73.2) | 0.008 | 0.001 |
| Rural | 4840 (26.8) | 960 (27.2) | 3880 (26.8) | ||
| Region of United States | |||||
| South | 8594 (47.7) | 1859 (52.6) | 6735 (46.5) | 0.153 | <0.001 |
| Northeast | 3163 (17.5) | 500 (14.1) | 2663 (18.4) | ||
| Midwest | 3632 (20.1) | 681 (19.3) | 2951 (20.4) | ||
| West | 2640 (14.6) | 495 (14.0) | 2145 (14.8) | ||
| Indications for CNS-active medications | |||||
| Behavioral symptoms | 8427 (46.7) | 1901 (53.8) | 6526 (45.0) | 0.175 | 0.087 |
| Anxiety disorders | 5964 (33.1) | 2005 (56.7) | 3959 (27.3) | 0.624 | 0.037 |
| Depression | 7254 (40.2) | 2431 (68.8) | 4823 (33.3) | 0.759 | 0.037 |
| Psychotic disorders | 1217 (6.8) | 423 (12.0) | 794 (5.5) | 0.231 | 0.029 |
| Chronic pain | 7988 (44.3) | 2221 (62.8) | 5767 (39.8) | 0.474 | 0.004 |
| Epilepsy | 2225 (12.3) | 718 (20.3) | 1507 (10.4) | 0.277 | 0.021 |
| Substance use disorders | 1902 (10.5) | 693 (19.6) | 1209 (8.3) | 0.329 | 0.004 |
| Subtype of Indexed ADRD diagnosis | |||||
| Frontotemporal dementia | 918 (5.1) | 137 (3.9) | 781 (5.4) | 0.180 | 0.030 |
| Vascular dementia | 1334 (7.4) | 294 (8.3) | 1040 (7.2) | ||
| Alzheimer's disease | 1401 (7.8) | 256 (7.2) | 1145 (7.9) | ||
| Senile dementia | 3545 (19.7) | 437 (12.4) | 3108 (21.4) | ||
| Dementia unspecified | 10,831 (60.1) | 2411 (68.2) | 8420 (58.1) | ||
| Comorbid condition | |||||
| Cancer | 2189 (12.1) | 451 (12.8) | 1738 (12.0) | 0.023 | 0.010 |
| Neurodegenerative disorders | 4326 (24.0) | 1013 (28.7) | 3313 (22.9) | 0.133 | 0.008 |
| Diabetes | 5031 (27.9) | 1121 (31.7) | 3910 (27.0) | 0.104 | 0.021 |
| Hypertension | 8849 (49.1) | 1963 (55.5) | 6886 (47.5) | 0.161 | <0.001 |
| Cardiovascular diseases | 6547 (36.3) | 1477 (41.8) | 5070 (35.0) | 0.140 | 0.023 |
| Pulmonary disorders | 9316 (51.7) | 2227 (63.0) | 7089 (48.9) | 0.287 | <0.001 |
| Kidney disorders | 1881 (10.4) | 498 (14.1) | 1383 (9.5) | 0.141 | 0.013 |
| Liver disorders | 1491 (8.3) | 379 (10.7) | 1112 (7.7) | 0.106 | 0.014 |
| Gastrointestinal tract disorders | 4260 (23.6) | 1187 (33.6) | 3073 (21.2) | 0.280 | 0.017 |
| Fall injuries | 1580 (8.8) | 514 (14.5) | 1066 (7.4) | 0.232 | 0.006 |
| Total comorbidities, mean (± SD) | 10.1 ± 6.4 | 12.9 ± 7.1 | 9.5 ± 6.0 | 0.521 | 0.009 |
| Use of health care services | |||||
| Hospitalization | 3184 (17.7) | 988 (27.9) | 2196 (15.2) | 0.315 | 0.020 |
| ED visit | 6532 (36.2) | 1757 (49.7) | 4775 (32.9) | 0.345 | 0.014 |
| Medication use | |||||
| Antidementia drugs | 6179 (34.3) | 1072 (30.3) | 5107 (35.2) | 0.105 | 0.018 |
| Year of Indexed date | |||||
| 2013 | 2795 (15.5) | 549 (15.5) | 2246 (15.5) | 0.125 | 0.039 |
| 2014 | 2027 (11.2) | 383 (10.8) | 1644 (11.3) | ||
| 2015 | 2096 (11.6) | 436 (12.3) | 1660 (11.5) | ||
| 2016 | 2662 (14.8) | 590 (16.7) | 2072 (14.3) | ||
| 2017 | 2013 (11.2) | 399 (11.3) | 1614 (11.1) | ||
| 2018 | 1949 (10.8) | 354 (10.0) | 1595 (11.0) | ||
| 2019 | 1912 (10.6) | 320 (9.1) | 1592 (11.0) | ||
| 2020 | 2575 (14.3) | 504 (14.3) | 2071 (14.3) | ||
ADRD: Alzheimer's disease and related dementias; CNS: central nervous system; ED: emergency department; SD: standard deviation; SMD: standardized mean difference; IPTW: inverse probability of treatment weighting.
Baseline characteristics between patients with versus without CNS-active medication polypharmacy
Of 18,029 patients with EOD, 3535 (19.6%) had CNS-active medication polypharmacy (Table 1). Before IPTW adjustment, patients with (versus without) CNS-active medication polypharmacy were more likely to be younger (15.9% versus 13.0% age ≤50 years, SMD = 0.116) and residing in the US South (52.6% versus 46.5%, SMD = 0.153). As expected, patients with (versus without) CNS-active medication polypharmacy were more likely to receive a diagnosis of clinical conditions indicated for the CNS-active medications of interest, including behavioral symptoms (53.8% versus 45.0%, SMD = 0.175), anxiety disorders (56.7% versus 27.3%, SMD = 0.624), depression (68.8% versus 33.3%, SMD = 0.759), psychotic disorders (12.0% versus 5.5%, SMD = 0.231), chronic pain (62.8% versus 39.8%, SMD = 0.474), epilepsy (20.3% versus 10.4%, SMD = 0.277), and substance use disorders (19.6% versus. 8.3%, SMD = 0.329). Patients with (versus without) CNS-active medication polypharmacy were more likely to have received diagnoses of clinical conditions, including pulmonary disorders (63.0% versus 48.9%, SMD = 0.287), hypertension (55.5% versus 47.5%, SMD = 0.161), cardiovascular diseases (41.8% versus 35.0%, SMD = 0.140), and fall injuries (14.5% versus. 7.4%, SMD = 0.232). In addition, a higher proportion of patients with (versus without) CNS-active medication polypharmacy experienced all-cause hospitalization (27.9% versus. 15.2%, SMD = 0.315) and an emergency department visit (49.7% versus. 32.9%, SMD = 0.345) during the 12-month baseline period. After IPTW, the distribution of all baseline characteristics was balanced between groups with versus without CNS-active polypharmacy, with SMDs for characteristics<0.100 (Table 1).
Combinations of concurrent use of CNS-active medications
Among 3535 patients with EOD who had CNS-active medication polypharmacy during the 6 months following the indexed ADRD diagnosis, the mean (SD) duration of receiving ≥3 CNS-active medication classes was 123 (54) days. Use of antidepressants accounted for 94.7% of patients with CNS-active medication polypharmacy, followed by benzodiazepines (63.9%), opioids (56.6%), antipsychotics (51.4%), and anticonvulsants (48.0%) (Table 2). Antidepressants were the most commonly prescribed class of CNS-active medications studied. Co-use of an antidepressant with another antidepressant was 66.0%, with a benzodiazepine was 63.4%, with an opioid was 55.8%, with an antipsychotic was 51.1%, and with an anticonvulsant was 46.1%. The percentages of patients with co-use of benzodiazepines with other CNS-active medication classes (except for antidepressants) were high, ranging from 19.0% to 50.1%.
Table 2.
Combinations of CNS-active medications among adults with young-onset dementia and CNS-active medication polypharmacy.
| Medication class | Percentage of 3535 patients with CNS-active medication polypharmacy | ||||||
|---|---|---|---|---|---|---|---|
| Overall** | Antidepressant* | Benzodiazepine* | Opioid* | Antipsychotic* | Anticonvulsant* | Z-drug* | |
| Antidepressant | 94.7 | 66.0 | 63.4 | 55.8 | 51.1 | 46.1 | 19.8 |
| Benzodiazepine | 63.9 | 19.0 | 53.1 | 50.1 | 42.8 | 20.3 | |
| Opioid | 56.6 | 38.4 | 40.2 | 45.8 | 22.0 | ||
| Antipsychotic | 51.4 | 21.8 | 42.0 | 15.5 | |||
| Anticonvulsant | 48.0 | 27.5 | 17.2 | ||||
| Z-drug | 20.3 | 4.6 | |||||
CNS: central nervous system.
*The frequency indicates percentage of 3535 patients with CNS-active medication polypharmacy prescribed as a specific combination of CNS-active medications within or between medication classes. Patients may have more than one drug combination.
**Overall percentage of 3535 patients with CNS-active medication polypharmacy by the specific medication class of CNS medications.
Main analysis
The crude incidence rate for NH admission in the overall sample was 25.04 per 1000 patient-years, with a higher rate being observed for patients with versus without CNS-active medication polypharmacy (29.6 versus 23.9 per 1000 patient-years) during 1-year follow-up (Table 3). After adjustment for covariates via IPTW, patients with (versus without) CNS-active medication polypharmacy had a higher risk for NH admission (HR, 1.34 [95% CI, 1.16–1.56]). The 1-year risk of NH admission in patients with (versus without) CNS-active medication polypharmacy was 3.29% (95% CI, 2.44%–4.42%) versus 2.37% (95% CI, 2.08%–2.70%) (Table 3). The HR estimate remained unchanged after trimming extreme weights of IPTW at various percentiles. A similar result (HR, 1.32 [95% CI, 1.13–1.53]) was observed when conducting sensitivity analysis with the Fine-Gray model, with disenrollment from insurance treated as a competing event (Supplemental Table 3). The E-value of the association with risk of NH admission was 2.01 for patients with versus without CNS medication polypharmacy, indicating strong resistance to unmeasured confounding factors. Sensitivity analysis of adding patients with a hospital stay of <30 days during the exposure assessment period in the study sample yielded findings consistent with the main analysis (Supplemental Table 4).
Table 3.
Associations between CNS-active medication polypharmacy and risk for nursing home admission among commercially insured adults with early-onset dementia.
| n | No. of events | Follow-up (days) | PY | Ratea | Risk for nursing home admission | 1-Year risk | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean ± SD | Median (IQR) | Crude HR (95% CI) | p | aHRb (95% CI) | p | % (95% CI) | |||||
| Overall sample | 18,029 | 314 | 254 ± 128 | 321 (136–365) | 12,539 | 25.04 | - | - | - | - | 2.82 (2.38–3.32) |
| By Chronic Concurrent Use of CNS-active medications | |||||||||||
| For <30 days | 14,494 | 241 | 253 ± 128 | 320 (136–365) | 10,072 | 23.92 | Reference | 2.37 (2.08–2.70) | |||
| For ≥30 days | 3535 | 73 | 254 ± 128 | 327 (134365) | 2466 | 29.60 | 1.24 (0.95–1.61) | 0.102 | 1.34 (1.16–1.56) | <0.001 | 3.29 (2.44–4.42) |
SD: standard deviation; IQR: interquartile range; PY: patient-years; IR: incidence rate; aHR: adjusted hazard ratio; CI: confidence interval; CNS: central nervous system.
Rate = IR/1000 PY.
Adjusted for variables listed in Table 1 using inverse probability of treatment weights in the Cox proportional hazards model.
Subgroup analysis
In the stratification analysis by age, the association between CNS-active medication polypharmacy and risk of NH admission was significant among patients with EOD who were older than 50 years (HR, 1.39 [95% CI, 1.20–1.61]), but not among those ≤50 years of age (HR, 0.57 [95% CI, 0.27–1.19]) (Table 4). When stratified by sex, there was an association among male patients (HR, 1.91 [95% CI, 1.54–2.36]) but not among female patients (HR, 0.94 [95% CI, 0.75–1.16]). An association between CNS-active medication polypharmacy and risk of NH admission varied by type of ADRD diagnosed at the index date. An association was observed among patients with a diagnosis of Alzheimer's disease (HR, 2.86 [95% CI, 1.72–4.76]) or senile dementia (HR, 1.47 [95% CI, 1.09–1.98]), but not among patients with a diagnosis of other ADRD types. We also observed an increased risk of NH admission associated with CNS-active medication polypharmacy in patients with anti-dementia medication use (HR, 1.84 [95% CI, 1.52–2.25]), but not among patients without the medication use at baseline.
Table 4.
Associations of CNS-active medication polypharmacy with risk of nursing home admission, stratified by selected baseline characteristics.
| By selected characteristic (n = sample size) | Patients with versus without CNS-active polypharmacy | ||||
|---|---|---|---|---|---|
| Crude HR (95% CI) | p | Adjusted HR (95% CI) | p | p for Interaction | |
| Age | |||||
| ≤50 years (n = 2795) | 1.47 (0.52– 4.17) | 0.47 | 0.57 (0.27–1.19) | 0.16 | 0.06 |
| >50 years (n = 15,234) | 1.25 (0.96–1.64) | 0.10 | 1.39 (1.19–1.61) | <0.001 | |
| Sex | |||||
| Female (n = 9869) | 1.11 (0.77–1.59) | 0.58 | 0.94 (0.75–1.16) | 0.55 | <0.001 |
| Male (n = 8160) | 1.46 (1.00 –2.14) | 0.052 | 1.91 (1.54–2.36) | <0.001 | |
| Dementia subtype a | |||||
| Frontotemporal dementia (n = 918) | 1.95 (0.77– 4.91) | 0.16 | 1.42 (0.82–2.47) | 0.26 | 0.67 |
| Vascular dementia (n = 1334) | 1.44 (0.66–3.13) | 0.36 | 0.90 (0.52–1.57) | 0.72 | |
| Alzheimer's disease (n = 3545) | 2.12 (0.80–5.57) | 0.051 | 2.86 (1.72–4.76) | <0.001 | |
| Senile dementia (n = 1441) | 1.38 (0.76–2.51) | 0.31 | 1.47 (1.09–1.98) | 0.02 | |
| Dementia unspecified (n = 10,831) | 1.13 (0.79–1.63) | 0.50 | 1.09 (0.88–1.35) | 0.43 | |
| Anti-dementia medication use | |||||
| Yes (n = 6179) | 1.29 (0.89–1.87) | 0.19 | 1.85 (1.52–2.25) | <0.001 | 0.01 |
| No (n = 11,850) | 1.82 (0.89–1.85) | 0.20 | 1.05 (0.84–1.32) | 0.64 | |
HR: hazard ratio; CNS: central nervous system; CI: confidence interval.
Discussion
In this population-based cohort study of privately insured adults diagnosed with ADRD before the age of 65 years, patients with (versus without) concurrent use of three or more CNS-active medications included in the AGS Beers Criteria had a 34% higher risk of NH admission during the 1-year follow-up period. The risk associated with CNS-active medication polypharmacy was particularly pronounced among patients with EOD who were older than 50 years, were male, received a diagnosis of certain dementia types, or received antidementia drugs at baseline. Our findings suggest that receipt of CNS-active medication polypharmacy was independently associated with subsequent risk of NH admission among patients with EOD, especially among older male patients with specific types of ADRD diagnosis.
Limited population-based studies have examined CNS-active medication polypharmacy and its associated adverse effects, such as the risk of NH admission among adults with EOD. The existing literature has primarily focused on older populations (≥65 years) and has studied the association of single or multiple use of the studied CNS-active medications with risk of sedation, 34 falls or fractures,35,36 cognitive impairment,37,38 and risk of NH admission.6,7,39 Consistent with the findings observed in older populations,6,7,39 our study of patients with EOD observed an increased risk of NH admission associated with the use of multiple CNS-active medications simultaneously. Thus, our results provide supporting evidence that similar to older adults with LOD, patients with EOD are susceptible to risks associated with the use of multiple CNS-active medications that have been listed to be avoided for older adults by AGS Beers Criteria.
The reasons for the observed heightened risk of NH admission among patients with CNS-active medication polypharmacy are multifaceted. The increased risk may be due to the exacerbation of underlying conditions (e.g., depression, anxiety, behavioral symptoms, sleep disorders, and pain) for which the CNS-active medications were prescribed. These neuropsychiatric conditions are prevalent and often co-occur in patients with ADRD, particularly at the advanced stage. 40 Thus, the use of three or more of the CNS-active medications studied may be an indicator, signaling worsening neuropsychiatric symptoms of ADRD among patients with EOD. On the other hand, a regimen of multiple CNS-active drugs used to treat several diseases in patients with EOD may be due to a prescribing cascade, that is, medications being added to treat or prevent adverse effects of other medications already prescribed. 41 Evidence has shown that patients with ADRD have a high risk of prescribing cascades involving CNS-active medications.42–44 For example, antidepressants are used to manage behavioral symptoms commonly experienced by patients with dementia, and subsequent headaches and difficulty in sleeping, potential side effects of antidepressants, may trigger the prescribing of opioids and Z-drugs, leading to the concurrent use of three CNS-active medications. The potential prescribing cascades involving CNS-active medications have led to initiatives to de-prescribe these potentially high-risk medications for older adults, particularly those with ADRD.42–44
Of the CNS-active medications studied, antipsychotics and benzodiazepines are two therapeutic drug categories that have drawn great clinical attention due to their potential side effects, including risk of stroke, death, and falls, in patients with dementia. 45 Co-use of antipsychotics and benzodiazepines was prevalent, present in 50.1% of patients with EOD who had CNS-active medication polypharmacy in the present study. Current clinical guidelines recommend cautious use of these two CNS-active medications.19,46 In general, the guidelines advised against using antipsychotics or benzodiazepines for neuropsychiatric symptoms of dementia (e.g., delirium, psychosis, agitation, and insomnia) unless the symptoms are severe and other non-pharmacological treatment options have failed. When considered clinically needed, antipsychotics can be used for a short duration (e.g., < 1 week) at a low dosage (e.g., between 0.5 and 1.0 mg/day of risperidone).19,46,47
In our subgroup analysis, the increased risk of NH admission with receipt of CNS-active medication polypharmacy was observed among patients who were male and aged > 50 years, but a null association was noted among patients who were female and aged ≤50 years. Reasons for the different associations observed between different sex-stratified subgroups are unclear but could be explained by sex differences in neuropsychiatric symptoms of dementia. Studies indicated that female patients tended to present depressive and anxiety symptoms, while male patients tended to exhibit disruptive (e.g., aggression) behaviors.48,49 Disruptive behaviors (versus depressive symptoms) were associated with higher caregiver burden 50 and burnout 51 in dementia and are often the major reason for NH replacement for patients with dementia.52,53 It is possible that aggressive behaviors were one of the underlying conditions for which multiple CNS-active medications were prescribed, and the symptoms, if insufficiently controlled, increased the risk of NH admission among male patients with EOD in the present study.
The present study results provide referential data for clinicians, patients, and their caregivers regarding potential risks associated with the receipt of CNS-active medication polypharmacy among patients who developed dementia between 30 and 64 years of age. Similar to patients with LOD, we observed that those with EOD faced an excessive risk of NH admission when using multiple CNS-active medications simultaneously. EOD often represents a distinct patient group, with a greater contribution of genetic factors, higher likelihood of presenting changes in personality, behaviors, or language (versus memory loss), and different contributing non-genetic factors (e.g., having traumatic brain injury versus cerebrovascular conditions), compared to LOD.54,55 These differences may drive faster disease progression,9,10 contribute to unique behavioral symptom profiles (high prevalence of depression and anxiety) 10 that require CNS-active medications, and lead to increased risk of NH institutionalization in patients with EOD.54,55 Clinical attention is needed in reviewing the use of CNS-active medications among patients with characteristics associated with an increased risk, as identified in the present study. Also, communication among prescribers and between prescribers and caregivers is important in reducing the prescribing of CNS-active medication polypharmacy in patients with dementia, who often receive multiple CNS-active medications from several prescribers. 56
Several limitations in our study should be noted. First, we measured the exposure by dispensed prescription claims and did not know whether patients consumed the medications. Second, the studied CNS-active medications were selected based on the AGS Beers Criteria updated in 2019, rather than 2023, 57 in which skeletal muscle relaxants were added as a therapeutic category of CNS-active medication. Third, administrative claims data lack clinical detail regarding the specific indication for CNS-active medication prescriptions. Consequently, we could not assess the clinical appropriateness of the observed polypharmacy. Fourth, we cannot eliminate the possibility of unmeasured confounders. While accounting for various confounders, our results were subject to residual confounding due to unmeasured confounders, including severity of underlying conditions for which CNS-active medications are prescribed, cognitive and physical function status, and sociodemographic variables (e.g., race and ethnicity), which are often unavailable in administrative claims data. Yet, only unmeasured confounders that were associated with high risk (HR ≥2.01) for both the CNS-active medication polypharmacy and the outcome could explain away our observed effect estimates, according to our E-values calculation. 33 Fifth, over half of the study patients with EOD were censored due to medical insurance disenrollment, likely from loss of employment, 58 during the 1-year follow-up. Such censoring prevented us from observing any NH admissions, leading to an underestimated incidence rate of the outcome. Nevertheless, our sensitivity analysis treating insurance disenrollment as a competing event yielded results consistent with the main analysis, suggesting that the loss to follow-up due to insurance disenrollment was likely non-informative and did not introduce substantial overestimation bias into our primary estimates. Sixth, the validity of diagnostic codes for ADRD and its subtypes remains suboptimal, with the positive predictive values ranging between 50% to 70% and sensitivity ranging between 30% to 50% when compared to clinical cognitive assessments.59,60 Our study sample likely missed patients who already had symptoms of early-onset dementia but were underdiagnosed. Finally, our study findings are generalizable only to commercially insured adults with EOD in the US.
Conclusion
In this cohort study of commercially insured patients with EOD, concurrent use of three or more CNS-active medications for 30 days or longer was associated with an increased risk of NH admission. The 1-year risk was pronounced among patients with EOD who were male, older than 50 years, had a diagnosis of dementia undetermined, or had baseline use of anti-dementia medication. These findings suggest that regular assessment of medication regimens that involve the use of CNS-active medications is needed and may reduce the risk of NH admission among patients with EOD.
Supplemental Material
Supplemental material, sj-docx-1-alr-10.1177_25424823261472831 for Central nervous system–active medication polypharmacy and risk for nursing home admission among adults with early-onset dementia: A retrospective cohort study by Nistha Shrestha, ChienWei Chiang and Yu-Jung Jenny Wei in Journal of Alzheimer's Disease Reports
Acknowledgements
The authors have no acknowledgments to report.
Footnotes
ORCID iDs: Nistha Shrestha https://orcid.org/0000-0002-8062-0247
Yu-Jung Jenny Wei https://orcid.org/0000-0001-8861-6907
Ethical considerations: The Ohio State University's Institutional Review Board approved and waived patient informed consent and HIPAA authorization for this study because of minimal risk and lack of feasibility to contact the subjects in Marketscan data.
Consent to participate: Not applicable
Consent for publication: Not applicable
Author contribution(s): Nistha Shrestha: Data curation; Formal analysis; Investigation; Methodology; Validation; Visualization; Writing – original draft; Writing – review & editing.
ChienWei Chiang: Data curation; Formal analysis; Investigation; Project administration; Resources; Software; Validation; Writing – review & editing.
Yu-Jung Jenny Wei: Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Resources; Supervision; Validation; Writing – review & editing.
Funding: The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute on Aging, (grant number R01AG073442).
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Supplemental material: Supplemental material for this article is available online.
Data availability statement
Data source used for this paper is licensed to The Ohio State University through a data user agreement with Merative MarketScan. As such, restrictions apply to the availability of this dataset, and it is not publicly available.
References
- 1.Alzheimer’s Association. 2025 Alzheimer’s disease facts and figures. Alzheimers Dement 2025; 2025: e70235. [Google Scholar]
- 2.Griffin P, Apostolova L, Dickerson BC, et al. Developments in understanding early onset Alzheimer’s disease. Alzheimers Dement 2023; 19: S126–S131. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Hendriks S, Peetoom K, Bakker C, et al. Global prevalence of young-onset dementia: a systematic review and meta-analysis. JAMA Neurol 2021; 78: 1080–1090. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Wei YJ, Shrestha N, Chiang C, et al. Prevalence and trend of central nervous system-active medication polypharmacy among US commercially insured adults with vs without early-onset dementia: a multi-year cross-sectional study. Alzheimers Res Ther 2024; 16: 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Maust DT, Strominger J, Kim HM, et al. Prevalence of central nervous system-active polypharmacy among older adults with dementia in the US. JAMA 2021; 325: 952–961. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Brodaty H, Connors MH, Xu J, et al. Predictors of institutionalization in dementia: a three year longitudinal study. J Alzheimers Dis 2014; 40: 221–226. [DOI] [PubMed] [Google Scholar]
- 7.Dorenlot P, Harboun M, Bige V, et al. Major depression as a risk factor for early institutionalization of dementia patients living in the community. Int J Geriatr Psychiatry 2005; 20: 471–478. [DOI] [PubMed] [Google Scholar]
- 8.Picard C, Pasquier F, Martinaud O, et al. Early onset dementia: characteristics in a large cohort from academic memory clinics. Alzheimer Dis Assoc Disord 2011; 25: 203–205. [DOI] [PubMed] [Google Scholar]
- 9.Koedam EL, Pijnenburg YA, Deeg DJ, et al. Early-onset dementia is associated with higher mortality. Dement Geriatr Cogn Disord 2008; 26: 147–152. [DOI] [PubMed] [Google Scholar]
- 10.Gumus M, Multani N, Mack ML, et al. Progression of neuropsychiatric symptoms in young-onset versus late-onset Alzheimer’s disease. Geroscience 2021; 43: 213–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Amatniek JC, Hauser WA, DelCastillo-Castaneda C, et al. Incidence and predictors of seizures in patients with Alzheimer’s disease. Epilepsia 2006; 47: 867–872. [DOI] [PubMed] [Google Scholar]
- 12.IBM MarketScan Research Databases for life sciences researchers. 2018.
- 13.Maust DT, Gerlach LB, Gibson A, et al. Trends in central nervous system-active polypharmacy among older adults seen in outpatient care in the United States. JAMA Intern Med 2017; 177: 583–585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Walling AM, Pevnick J, Bennett AV, et al. Dementia and electronic health record phenotypes: a scoping review of available phenotypes and opportunities for future research. J Am Med Inform Assoc 2023; 30: 1333–1348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen Y, Tysinger B, Crimmins E, et al. Analysis of dementia in the US population using medicare claims: insights from linked survey and administrative claims data. Alzheimers Dement (N Y) 2019; 5: 197–207. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Mouta S, Fonseca Vaz I, Pires M, et al. What do we know about pseudodementia? Gen Psychiatr 2023; 36: e100939. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Festa N, Price M, Moura LM, et al. Evaluation of claims-based ascertainment of Alzheimer disease and related dementias across health care settings. JAMA Health Forum 2022; 3: e220653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Singh Solorzano C, Orini S, Festari C, et al. Clinical pathways for diagnosing neurocognitive disorders: insights from process mining a memory clinic cohort. J Am Med Dir Assoc 2025; 26: 105774. [DOI] [PubMed] [Google Scholar]
- 19.By the 2019 American Geriatrics Society Beers Criteria® Update Expert Panel. American Geriatrics Society 2019 Updated AGS Beers Criteria® for potentially inappropriate medication use in older adults. J Am Geriatr Soc 2019; 67: 674–694. [DOI] [PubMed] [Google Scholar]
- 20.Goodwin JS, Li S, Middleton A, et al. Differences between skilled nursing facilities in risk of subsequent long-term care placement. J Am Geriatr Soc 2018; 66: 1880–1886. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Toth M, Palmer L, Bercaw L, et al. Trends in the use of residential settings among older adults. J Gerontol B Psychol Sci Soc Sci 2022; 77: 424–428. [DOI] [PubMed] [Google Scholar]
- 22.Yang J, Boytsov N, Carlson JJ, et al. Health care resource utilization and costs among patients with multiple myeloma with exposure to double-class or triple-class multiple myeloma treatments: a retrospective US claims database analysis. J Manag Care Spec Pharm 2023; 29: 917–926. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Gaugler JE, Duval S, Anderson KA, et al. Predicting nursing home admission in the U.S: a meta-analysis. BMC Geriatr 2007; 7: 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Vetrano DL, Damiano C, Tazzeo C, et al. Multimorbidity patterns and 5-year mortality in institutionalized older adults. J Am Med Dir Assoc 2022; 23: 1389–1395.e4. [DOI] [PubMed] [Google Scholar]
- 25.Porsteinsson AP, Isaacson RS, Knox S, et al. Diagnosis of early Alzheimer’s disease: clinical practice in 2021. J Prev Alzheimers Dis 2021; 8: 371–386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput 2009; 38: 1228–1234. [Google Scholar]
- 27.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009; 28: 3083–3107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects. Biometrika 1983; 70: 41–55. [Google Scholar]
- 29.Xue X, Xie X, Gunter M, et al. Testing the proportional hazards assumption in case-cohort analysis. BMC Med Res Methodol 2013; 13: 88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Crump RK, Hotz VJ, Imbens GW, et al. Dealing with limited overlap in estimation of average treatment effects. Biometrika 2009; 96: 187–199. [Google Scholar]
- 31.Stürmer T, Rothman KJ, Avorn J, et al. Treatment effects in the presence of unmeasured confounding: dealing with observations in the tails of the propensity score distribution—a simulation study. Am J Epidemiol 2010; 172: 843–854. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509. [Google Scholar]
- 33.VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the E-value. Ann Intern Med 2017; 167: 268–274. [DOI] [PubMed] [Google Scholar]
- 34.Richardson K, Loke YK, Fox C, et al. Adverse effects of Z-drugs for sleep disturbance in people living with dementia: a population-based cohort study. BMC Med 2020; 18: 51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Hart LA, Marcum ZA, Gray SL, et al. The association between central nervous system-active medication use and fall-related injury in community-dwelling older adults with dementia. Pharmacotherapy 2019; 39: 530–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Guan Q, Men S, Juurlink DN, et al. Opioid initiation and the hazard of falls or fractures among older adults with varying levels of central nervous system depressant burden. Drugs Aging 2022; 39: 729–738. [DOI] [PubMed] [Google Scholar]
- 37.Wright RM, Roumani YF, Boudreau R, et al. Effect of central nervous system medication use on decline in cognition in community-dwelling older adults: findings from the health, aging and body composition study. J Am Geriatr Soc 2009; 57: 243–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wattmo C, Wallin AK, Londos E, et al. Risk factors for nursing home placement in Alzheimer’s disease: a longitudinal study of cognition, ADL, service utilization, and cholinesterase inhibitor treatment. Gerontologist 2011; 51: 17–27. [DOI] [PubMed] [Google Scholar]
- 39.Zuckerman IH, Langenberg P, Baumgarten M, et al. Inappropriate drug use and risk of transition to nursing homes among community-dwelling older adults. Med Care 2006; 44: 722–730. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Lyketsos CG, Lopez O, Jones B, et al. Prevalence of neuropsychiatric symptoms in dementia and mild cognitive impairment: results from the cardiovascular health study. JAMA 2002; 288: 1475–1483. [DOI] [PubMed] [Google Scholar]
- 41.O'Mahony D, Rochon PA. Prescribing cascades: we see only what we look for, we look for only what we know. Age Ageing 2022; 51: afac138. [DOI] [PubMed] [Google Scholar]
- 42.Phelan EA, Balderson BH, Fujii MM, et al. Deprescribing central nervous system-active medications among community-dwelling older adults with dementia in primary care: a feasibility study. Int J Environ Res Public Health 2025; 22: 1611. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Hung A, Growdon ME. Deprescribing considerations for central nervous system-active polypharmacy in patients with dementia. J Am Geriatr Soc 2025; 73: 343–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Green AR, Boyd CM, Quiles R, et al. Deprescribing for People with dementia: a roadmap. Drugs Aging 2025; 42: 795–806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Korkatti-Puoskari N, Tiihonen M, Caballero-Mora MA, et al. Therapeutic dilemma’s: antipsychotics use for neuropsychiatric symptoms of dementia, delirium and insomnia and risk of falling in older adults, a clinical review. Eur Geriatr Med 2023; 14: 709–720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.APA Work Group on Alzheimer’s Disease and other Dementias, Rabins PV, Blacker D, et al. American Psychiatric Association practice guideline for the treatment of patients with Alzheimer’s disease and other dementias. Second edition. Am J Psychiatry 2007; 164: 5–56. [PubMed] [Google Scholar]
- 47.Neufeld KJ, Yue J, Robinson TN, et al. Antipsychotic medication for prevention and treatment of delirium in hospitalized adults: a systematic review and meta-analysis. J Am Geriatr Soc 2016; 64: 705–714. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Resnick B, Galik E, McPherson R, et al. Gender differences in disease, function, and behavioral symptoms in residents with dementia. West J Nurs Res 2022; 44: 812–821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Lovheim H, Sandman PO, Karlsson S, et al. Sex differences in the prevalence of behavioral and psychological symptoms of dementia. Int Psychogeriatr 2009; 21: 469–475. [DOI] [PubMed] [Google Scholar]
- 50.Miyamoto Y, Tachimori H, Ito H. Formal caregiver burden in dementia: impact of behavioral and psychological symptoms of dementia and activities of daily living. Geriatr Nurs 2010; 31: 246–253. [DOI] [PubMed] [Google Scholar]
- 51.Hiyoshi-Taniguchi K, Becker CB, Kinoshita A. What behavioral and psychological symptoms of dementia affect caregiver burnout? Clin Gerontol 2018; 41: 249–254. [DOI] [PubMed] [Google Scholar]
- 52.Chan DC, Kasper JD, Black BS, et al. Presence of behavioral and psychological symptoms predicts nursing home placement in community-dwelling elders with cognitive impairment in univariate but not multivariate analysis. J Gerontol A Biol Sci Med Sci 2003; 58: 548–554. [DOI] [PubMed] [Google Scholar]
- 53.Toot S, Swinson T, Devine M, et al. Causes of nursing home placement for older people with dementia: a systematic review and meta-analysis. Int Psychogeriatr 2017; 29: 195–208. [DOI] [PubMed] [Google Scholar]
- 54.Seath P, Macedo-Orrego LE, Velayudhan L. Clinical characteristics of early-onset versus late-onset Alzheimer’s disease: a systematic review and meta-analysis. Int Psychogeriatr 2024; 36: 1093–1109. [DOI] [PubMed] [Google Scholar]
- 55.Mendez MF. Early-onset Alzheimer disease and its variants. Continuum (Minneap Minn 2019; 25: 34–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.Vordenberg SE, Davis RC, Strominger J, et al. Clinician contributions to central nervous system-active polypharmacy among older adults with dementia in the United States. J Am Geriatr Soc 2025; 73: 422–430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 57.By the American Geriatrics Society Beers Criteria Update Expert P. . American Geriatrics Society 2023 updated AGS Beers Criteria(R) for potentially inappropriate medication use in older adults. J Am Geriatr Soc 2023; 71: 2052–2081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 58.Bagnasco G, Bakx P, Licher S, et al. Earnings losses in young-onset dementia: Population-based study with admin data. Alzheimers Dement 2025; 21: e14588. doi: 10.1002/alz.14588 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 59.Wang T, Pate V, Kim DH, et al. Developing a novel algorithm to identify incident and prevalent dementia in medicare claims-the ARIC study. Am J Epidemiol 2025; 194: 3537–3548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 60.McCarthy EP, Chang CH, Tilton N, et al. Validation of claims algorithms to identify Alzheimer’s disease and related dementias. J Gerontol A Biol Sci Med Sci 2022; 77: 1261–1271. [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
Supplemental material, sj-docx-1-alr-10.1177_25424823261472831 for Central nervous system–active medication polypharmacy and risk for nursing home admission among adults with early-onset dementia: A retrospective cohort study by Nistha Shrestha, ChienWei Chiang and Yu-Jung Jenny Wei in Journal of Alzheimer's Disease Reports
Data Availability Statement
Data source used for this paper is licensed to The Ohio State University through a data user agreement with Merative MarketScan. As such, restrictions apply to the availability of this dataset, and it is not publicly available.