Trends and expenditures in polypharmacy and hyperpolypharmacy among older adults in the United States, 2002–2017

Abstract

Background

Polypharmacy and hyperpolypharmacy are increasingly common among older adults and are associated with substantial clinical and economic burdens. Understanding trends in their prevalence and associated expenditures is essential to inform policy and targeted interventions. This study examines trends in the prevalence of polypharmacy and hyperpolypharmacy and their associated healthcare and medication expenditures among U.S. older adults from 2002 to 2017.

Methods

This retrospective cross-sectional study used data from the Medical Expenditure Panel Survey, a nationally representative survey of the noninstitutionalized U.S. population. The study included older adults aged 65 years or older. Medication burden was categorized as polypharmacy (5–9 medications) and hyperpolypharmacy (≥ 10 medications), determined annually based on prescription records. The primary outcomes were annual prevalence estimates of polypharmacy and hyperpolypharmacy. Secondary outcomes were total healthcare and prescribed medication expenditures.

Results

Our study included 61,402 adults aged 65 years or older representing a weighted population of 643 million person-years. Over 2002 − 2017, the prevalence of polypharmacy ranged from 35.1%−39.4%, while the prevalence of hyperpolypharmacy ranged from 12.5%−17.7%. Polypharmacy prevalence increased significantly until 2011 (+ 0.38%/year [0.23 ~ 0.53]; p<.001) and declined thereafter (-0.45%/year [-0.67~-0.23]; p<.001). Hyperpolypharmacy significantly increased until 2006 (+ 0.87%/year [0.14 ~ 1.59]; p=.023), with a non-significant stable period afterward. Subgroup analyses revealed a higher prevalence among individuals aged 75–84 years and among females. Asians exhibited a significant increase in polypharmacy (+ 0.83%/year [0.32 ~ 1.33]; p=.023). Total healthcare expenditures increased among individuals with polypharmacy after 2013 (+$1030/year [231 ~ 1829]; p=.016). Prescribed medication expenditures increased significantly among individuals with hyperpolypharmacy particularly after 2014 (+$939/year [319 ~ 1560]; p=.006).

Conclusion

Polypharmacy and hyperpolypharmacy remain highly prevalent among older adults in the U.S., with a significant growth in medication expenditures over time among those with hyperpolypharmacy. These findings highlight the critical need for ongoing monitoring and tailored prescribing optimization efforts targeting subgroups with the high prescribing burden to reduce the clinical and economic burdens of polypharmacy and hyperpolypharmacy in aging populations.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12877-026-07172-9.

Background

The aging population has been accompanied by a substantial increase in the use of prescription medications to manage chronic diseases leading to a corresponding rise in the number of medications taken by individuals [1]. While definitions vary across studies, polypharmacy and hyperpolypharmacy are most commonly defined as the concurrent use of ≥ 5 and ≥ 10 medications, respectively [2]. Globally, the estimated prevalence of polypharmacy among older adults is approximately 39.1% with hyperpolypharmacy affecting around 13.3% of this population [3]. Higher rates have been reported among individuals with frailty and those residing in long-term care facilities [4, 5]. In the United States (U.S.), the prevalence of polypharmacy has steadily increased over the past three decades [6–10]. It rose from 12.8% in 1988 to 24% in 1999, then to approximately 39% in the early 2010s [7, 10]. More recent population-based studies estimate the prevalence to be around 43% in the late 2010s through 2020 [6, 9]. For hyperpolypharmacy, the prevalence in the U.S. has been reported at approximately 6% to 8% [6, 8]. Additionally, significant geographic and racial disparities in polypharmacy prevalence have been documented suggesting that sociodemographic and socioeconomic factors may be important determinants of medication burden [11].

Polypharmacy has been associated with numerous adverse clinical outcomes including increased risk of falls, fractures, cognitive decline, hospitalizations, and all-cause mortality [4, 12, 13]. It is also linked to medication nonadherence and harmful drug-drug interactions [4, 12, 13]. Among older adults with frailty, hyperpolypharmacy has been particularly associated with elevated mortality risk [14, 15]. Furthermore, polypharmacy increases the likelihood of an older adult receiving potentially inappropriate medications referring to prescribing that may pose more risks than benefits [16, 17]. In economically vulnerable populations the financial burden of managing multiple prescriptions can exacerbate nonadherence and lead to unmet healthcare needs [18]. As such, it is a priority for healthcare providers to recognize and address medication overuse and inappropriate prescribing to improve health outcomes and avoid unnecessary costs among the growing population of older adults.

While the prevalence and clinical consequences of polypharmacy are well-documented, comprehensive reports on hyperpolypharmacy trends across sociodemographic subgroups utilizing nationally representative data remain relatively scarce and often lack detailed insights into associated healthcare costs [4, 6, 8]. Population subgroups defined by race, ethnicity, and other key sociodemographic factors often experience different medication access, adherence, and cost burdens, contributing to inequities in both clinical outcomes and healthcare spending that warrant solutions from both clinical and policy perspectives [19, 20]. Existing studies primarily focus on prevalence and clinical outcomes leaving a notable gap in understanding the full economic burden of poly- and hyperpolypharmacy at the population level [4, 8]. Furthermore, few studies have delineated the economic implications of polypharmacy versus hyperpolypharmacy in a longitudinal context. Therefore, a thorough understanding of these evolving trends is crucial for informing the development of cost-effective and patient-centered interventions at both clinical and policy levels.

Leveraging nationally representative data on community-dwelling older adults from 2002 to 2017, this study aims to longitudinally characterize the prevalence of polypharmacy and hyperpolypharmacy and to evaluate associated healthcare and prescription expenditures. By stratifying analyses by key demographics, this study provides insights into disparities in medication burden and costs with implications for tailored interventions and health policy planning aimed at decreasing medication burden and reducing medication-related financial impacts at the national level.

Methods

Study design and data source

This cross-sectional study was conducted utilizing the Medical Expenditure Panel Survey (MEPS) data from 2002 to 2017. MEPS is a survey collecting data on healthcare services used by individuals in the U.S., including service types, usage frequency, costs, payment methods, and insights into health insurance coverage, costs, and availability [21]. The survey is conducted annually by the U.S. Public Health Service through the Agency for Healthcare Research and Quality and the Centers for Disease Control and Prevention [22]. The MEPS sample population is derived from participants in the National Health Interview Survey, which ensures a nationally representative sample of households. The data collection process follows an overlapping panel design spanning five survey rounds conducted at intervals of approximately 5–6 months over a period of 2.5 years [23]. In addition to gathering data directly from the sample population, MEPS also incorporates a Medical Provider Component (MPC) survey. This component involves collecting information from pharmacies regarding the population sample. The MPC survey includes a computer-generated dispensing record, which encompasses details such as the date of prescription fulfillment, the National Drug Code (NDC), and the name and characteristics of the medication [24]. The study utilized the MEPS Full-Year Consolidated files, Prescribed Medicines files, and Medical Conditions files. The sample for this study was limited to older adults aged 65 years and older during the initial round of interviews. This study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies [25].

Exposures

The primary exposure was medication burden, categorized as polypharmacy (5–9 prescribed medications) and hyperpolypharmacy (≥ 10 prescribed medications), based on definitions commonly used in the literature [2]. Respondents supplied information across five in-person survey rounds conducted over a period of 2.5 years, with the recall period for each interview being 5–6 months [23, 26, 27]. The respondent provided self-reported prescription fill information, which was further verified by the MPC data [24]. Each respondent’s unique prescribed medications, identified by drug name and NDC, were utilized to calculate the medication burden. Medication counts reflect all unique prescribed medications recorded during the recall window and therefore capture cumulative rather than concurrent medication use. Based on these counts, subjects were designated as having < 5 medications, polypharmacy [5–9], or hyperpolypharmacy (≥ 10).

Study outcomes

The primary outcomes of this study were the prevalence of polypharmacy and hyperpolypharmacy. The annual prevalence of polypharmacy was calculated by dividing respondents with polypharmacy by the total population each year. The same method was used for hyperpolypharmacy. These prevalence estimates were generated for each year from 2002 to 2017. Secondary outcomes included total healthcare expenditures and prescribed medication expenditures. Total healthcare expenditures reflected the aggregate direct payments for all healthcare services including emergency visits, hospital stays, and prescription medications. Expenditure totals represent the sum of payments from all sources, including out-of-pocket spending, private insurance, public insurance, or third-party payers. Prescribed medication expenditures were calculated using the same comprehensive set of payer sources. All expenditure data were adjusted for inflation using the Consumer Price Index and standardized to 2024 U.S. dollars.

Baseline characteristics

Baseline characteristics included a range of demographic and clinical variables. Demographic characteristics comprised age, sex, race, education level, and poverty status. Age was recorded as a continuous variable and subsequently grouped into three categories: 65–74 years, 75–84 years, and 85 years or older. Sex was classified as male or female, and race was categorized as White, African American, Asian, or other. Functional status was assessed using indicators for activities of daily living (ADLs) and instrumental activities of daily living (IADLs). Comorbidities were collectively evaluated from [1] Medical Conditions Files using International Classification of Diseases, Ninth and Tenth Revision (ICD-9 and ICD-10) codes and [2] Priority Condition section of the MEPS Household Component Files [28, 29]. For each individual, an age-adjusted Charlson Comorbidity Index (CCI) score was calculated to represent comorbidity burden [29, 30]. Healthcare utilization per year including emergency room visits and hospital admissions was documented.

Statistical analysis

This study’s analysis incorporated the MEPS complex sampling design (stratification, clustering, and weighting) to ensure national representativeness using weighted record-level data. Descriptive statistics were used to summarize characteristics of the study population across 2002–2017. Continuous variables are presented as means ± standard error (SE), and categorical variables as percentages. Baseline characteristics were presented for the total population and further stratified by medication burden: less than 5, polypharmacy [5–9], and hyperpolypharmacy (≥ 10).

Trends in prevalence and expenditure were analyzed in accordance with the National Center for Health Statistics Guidelines for Analysis of Trends [31]. We first evaluated monotonic trends using linear regression. If no significant monotonic pattern was detected, polynomial regression was then used to explore nonlinear patterns. Significant polynomial trends prompted the use of Joinpoint regression to identify years of inflection. Standard errors for the annual prevalence calculation were provided for the Joinpoint regression. First-order autocorrelation estimated from the data was selected for correlated data, given that respondents in MEPS contributed data for up to 2.5 years. We selected weighted BIC as the data-driven method for model construction. Our data from 2002 to 2017 allowed up to two inflection points in the model. Upon detecting an inflection point, segmental regression computed the slopes for the trend preceding and succeeding the inflection point, which were subsequently reported as annual percentage or monetary changes. Subgroup analyses were performed by age, sex, and race. All data processing and statistical analyses were performed using R (v4.4.2) and RStudio (v1.5.57) [32]. Joinpoint regressions were conducted using the Joinpoint Regression Program (v5.4.0.0) [33, 34]. Statistical significance was defined as a p-value < 0.05 (two-sided), with a 95% confidence interval (CI). Adjustments for multiple comparisons were not made, and the results should be viewed as exploratory because of the potential for type I error.

Results

Study demographic

The aggregated data captured 61,402 individuals aged ≥ 65 years representing a weighted population of 643,127,515 over 2002–2017. Weighted baseline characteristics are summarized in Table 1. The mean age was 74 years, with 56.54% female and 85.80% identifying as White. About 55% of the population had a high school diploma or less as their highest degree, and around 60% were middle-income or higher. There were 6.53% of subjects who had limitations in at least one ADL, and 11.70% had limitations in at least one IADL. The average age-adjusted CCI was 3.82. On average, participants had 0.29 emergency department visits and 0.27 hospitalizations annually. Baseline characteristics by polypharmacy and hyperpolypharmacy are detailed in Table 1.

Table 1.

Baseline characteristics across study period (N = 643,127,515 [weighted])

	Total	Medications < 5	Polypharmacy	Hyperpolypharmacy
Unweighted size	61,402	29,756 (48.46%)	22,494 (36.63%)	9,152 (14.91%)
Weighted size	643,127,515	307,295,387 (47.78%)	238,886,435 (37.15%)	96,945,693 (15.07%)
Age
Mean ± SE	74.23 ± 0.07	73.73 ± 0.08	74.78 ± 0.09	74.46 ± 0.12
65–74	55.02%	58.72%	50.95%	53.32%
75–84	33.47%	29.98%	36.79%	36.37%
≥ 85	11.51%	11.30%	12.26%	10.31%
Sex
Male	43.46%	45.49%	42.05%	40.49%
Female	56.54%	54.51%	57.95%	59.51%
Race
White	85.80%	84.80%	86.52%	87.20%
African American	8.68%	8.78%	8.67%	8.34%
Asian	3.11%	4.03%	2.53%	1.65%
Other	2.41%	2.39%	2.28%	2.81%
Education
High school or less	54.93%	53.07%	55.52%	59.35%
Some college	21.75%	21.13%	22.61%	21.55%
College or postgraduate	22.22%	24.24%	21.17%	18.43%
Unknown	1.10%	1.56%	0.70%	0.67%
Poverty
Poor	10.04%	9.60%	10.06%	11.36%
Near poor	6.62%	5.92%	6.96%	8.00%
Low income	18.28%	17.38%	18.61%	20.34%
Middle income	29.02%	28.69%	29.10%	29.89%
High income	36.04%	38.41%	35.27%	30.41%
Functional status
Had ≥ 1 ADL limitations
Yes	6.53%	4.32%	6.80%	12.84%
No	92.53%	93.95%	92.94%	87.00%
Unknown	0.94%	1.72%	0.26%	0.16%
Had ≥ 1 IADL limitations
Yes	11.70%	7.81%	12.63%	21.74%
No	87.36%	90.47%	87.12%	78.10%
Unknown	0.94%	1.72%	0.25%	0.16%
Age-adjusted CCI	3.82 ± 0.01	3.36 ± 0.01	4.03 ± 0.02	4.78 ± 0.03
Healthcare utilization per year
Number of ER visits	0.29 ± 0.004	0.16 ± 0.003	0.33 ± 0.006	0.66 ± 0.02
Number of hospital admissions	0.27 ± 0.004	0.13 ± 0.003	0.29 ± 0.006	0.64 ± 0.014

Abbreviations: ADL Activities of Daily Living, CCI Charlson Comorbidity Index, ER Emergency Room, IADL Instrumental Activities of Daily Living, SE Standard Error

Prevalence of polypharmacy and hyperpolypharmacy

The overall prevalence of polypharmacy and hyperpolypharmacy was 37.15% and 15.07%, respectively. Trends in polypharmacy and hyperpolypharmacy prevalence are shown in Fig. 1. Within the study period, the prevalence of polypharmacy and hyperpolypharmacy ranged from 35.1% to 39.4% and 12.47%-17.72%, respectively. Joinpoint analyses identified inflection points in 2011 for polypharmacy and 2006 for hyperpolypharmacy. Prior to 2011, the prevalence of polypharmacy increased significantly (+ 0.38%/year [0.23 ~ 0.53], p<.001), followed by a decline thereafter (− 0.45%/year [-0.67~-0.23], p<.001). Prevalence of hyperpolypharmacy increased before 2006 (+ 0.87%/year [0.14 ~ 1.59], p=.02), with no significant change after 2006. The results of trend analyses, including those for the overall population and subgroup analyses, are presented in the Supplementary Tables 1–4.

Fig. 1.

Trends in the prevalence of polypharmacy and hyperpolypharmacy among US older adults, 2002-2017

Prevalence of polypharmacy and hyperpolypharmacy by age group, sex, and race

Age

Among individuals aged 65–74, polypharmacy increased from 32.9% in 2002 to 37.0% in 2011, then declined to 34.4% in 2017 (Fig. 2). Those aged 75–84 had higher prevalence with a persistent increase until 2015 (+ 0.25%/year [0.10 ~ 0.40], p=.01), followed by a notable decline (− 2.80%/year [-4.22~-1.37], p<.001). No distinct polypharmacy trend or inflection point was observed among individuals aged ≥ 85, as prevalence varied considerably from year to year. Hyperpolypharmacy among those aged 65–74 demonstrated a significant increase before 2014 (+ 0.23%/year [0.04 ~ 0.42], p=.02), while those aged 75–84 showed a significant increase prior to 2005 (+ 1.46%/year [0.09 ~ 2.83], p=.04). Trends of hyperpolypharmacy among these groups after the inflection points were both nonsignificant. Similarly, no significant trend in hyperpolypharmacy was observed among individuals aged ≥ 85.

Fig. 2.

Trends in the prevalence of polypharmacy and hyperpolypharmacy by age and sex among US older adults, 2002-2017. A Trends in prevalence of polypharmacy by age; B Trends in prevalence of hyperpolypharmacy by age; C Trends in prevalence of polypharmacy by sex; D Trends in prevalence of hyperpolypharmacy by sex

Sex

Females had a higher overall prevalence of polypharmacy though the gap narrowed over time. Among males, polypharmacy significantly increased from 32.0% in 2002 to 38.1% in 2011 (+ 0.55%/year [0.14 ~ 0.97], p=.01) (Fig. 2). In contrast, the increase before 2010 was more modest (+ 0.25%/year [0.08 ~ 0.43], p=.01) in females resulting in a narrowing gap. The declines in prevalence among males (-0.40%/year [-1.01 ~ 0.21], p=.18) and females (-0.49%/year [-0.74~-0.24], p<.001) were similar after the inflection points, even though the decline in males was not statistically significant. For hyperpolypharmacy, females did not exhibit significant trends before and after the 2007 inflection, whereas males showed a significant increase before 2014 (+ 0.40%/year [0.17 ~ 0.63], p=.003), with a non-statistically significant decline afterward.

Race

Among Whites, polypharmacy significantly rose from 35.7% to 40.4% (p=.001) before 2010, then declined (− 0.49%/year [-0.78~-0.20], p=.01) (Supplementary Figure S1). African Americans experienced a significant increase from 34.8% to 37.6% across the study period (p=.026). Asians showed the greatest overall increase, with two rises: one from 2002 to 2007 (+ 3.04%/year [1.49 ~ 4.59], p<.001) and another from 2015 to 2017 (+ 5.28%/year [-3.90 ~ 14.46], p=.23). However, only the first rise was statistically significant. Hyperpolypharmacy increased significantly among Whites (p=.03) and was not significant among African Americans (p=.06) before 2006, with overlapping trends between 2006 and 2014, followed by a subsequent decline among African Americans (− 1.46%/year [− 3.71 ~ 0.79], p=.18). However, the trends after 2006 were not significant among White and African American individuals. Asians consistently had lower hyperpolypharmacy prevalence compared to other race groups, while the trends were not significant.

Total healthcare and prescribed medication expenditures

Mean annual expenditures were $14,059 (SE ± 145) for total healthcare and $3,129 (SE ± 45) for prescribed medications (Fig. 3). Among individuals with polypharmacy, total healthcare expenditure non-significantly decreased prior to 2013 (-$118.49 USD/year [-267.86 ~ 30.87], p=.11), followed by a significant increase after 2013 (+$1030.40 USD/year [231.65 ~ 1829.20], p=.02). No inflection points were found in the < 5 medications or hyperpolypharmacy groups. Total healthcare expenditures among individuals with < 5 medications remained stable, and those with hyperpolypharmacy showed an increase (+$54.90 USD/year [-287.43 ~ 397.16], p=.74), but neither trend was significant. Increases in prescribed medication expenditures were greatest among those with hyperpolypharmacy, with non-significant increases noted before 2014 (+$98.08 USD/year [-17.93 ~ 214.09], p=.09) and statistically significant after 2014 (+$939.07 USD/year [318.65 ~ 1559.50], p=.01). Among those with polypharmacy, significant increases were seen before 2005 (+$259.21 USD/year [37.43 ~ 481.00], p=.03) and after 2012 (+$189.88 USD/year [33.05 ~ 346.70], p=.02). The results of trend analyses for expenditures are presented in the Supplementary Tables 5–6.

Fig. 3.

Trends in the Total healthcare and prescribed medication expenditure by polypharmacy category among US older adults, 2002-2017. A Total healthcare expenditure; B Prescribed medication expenditure

Discussion

This study examined the prevalence and expenditures associated with polypharmacy and hyperpolypharmacy among older U.S. adults using national data. Polypharmacy affected about 35–39% of older adults, with 12–18% experiencing hyperpolypharmacy, while the prevalence varied across groups. Total healthcare expenditures significantly increased among individuals with polypharmacy after 2013, and prescribed medication expenditures rose significantly among those with hyperpolypharmacy after 2014. These results highlight the need for targeted clinical and policy interventions to mitigate medication load and its associated economic burden in aging populations.

The prevalence of polypharmacy observed in our study (35.1–39.4%) aligns with studies using the National Health and Nutrition Examination Survey (NHANES), reporting rates between 35% and 45% during a similar period [6–10]. Consistent with prior reports, we observed a significantly increasing trend in polypharmacy through the early 2000s; however, our study uniquely identified a subsequent significant decline after 2011 [6, 9, 10]. Our estimates of hyperpolypharmacy prevalence (12.5%-17.7%) were higher than prior national reports (6%-8%), though the overall trend, marked by early significant increases followed by a non-significant plateau was consistent [6, 8]. These differences may reflect the longer medication recall period in MEPS, which likely captures a more cumulative profile of prescriptions [23, 26, 27]. Importantly, these findings indicate the need for harmonized definitions and standardized data collection methods, such as consistent recall periods, medication verification protocols, and thresholds for medication counts. This would support valid comparisons across diverse data sources and provide more accurate information for clinical care, health policy, and population-level monitoring of medication burden [2, 12].

It is generally accepted that the prevalence of polypharmacy increases with advancing age [3, 9, 12]. Our subgroup showed that individuals aged 65–74 generally had lower polypharmacy rates than those aged 75–84, with CIs that mostly did not overlap. However, it is difficult to determine differences in hyperpolypharmacy prevalence across age groups over time due to highly overlapping CIs. Additionally, due to the small sample size of adults aged 85 and older, the prevalence varied greatly with wide CIs. In addition, MEPS includes fewer community-dwelling adults aged ≥ 85 because institutionalized individuals are not captured. This reduced sampling contributes to smaller effective samples sizes and greater year to year variability further limiting the stability of prevalence estimates for this sub group. Nevertheless, comparing the overall trend lines for polypharmacy and hyperpolypharmacy across age groups, individuals aged ≥ 85 may not have the highest medication burden compared to those aged 75–84. Recent nationwide research studying medication use in individuals older than 90 years old found that medication load decreases with age in this population [35]. One explanation is the “healthy survivor” effect where frailer older adults with higher levels of comorbidities and medication burdens are less likely to reach very old age [36, 37]. Conversely, those who do survive tend to be healthier and are prescribed fewer medications. Second, preventive pharmacotherapy such as statins is frequently reevaluated in the oldest adults and may be deprescribed when the expected time to benefit exceeds projected life expectancy [38]. Additionally, clinicians may adopt less aggressive therapeutic targets and simplify medication regimens in this age group, balancing potential risks and benefits in light of age-related physiological changes and prioritizing quality of life [39–41].

The trends in the prevalence of polypharmacy by sex in our study align with studies using NHANES data, which similarly demonstrate a higher polypharmacy prevalence among females and a convergence in prevalence over time in the older population [8, 9]. While it is generally recognized that females are more susceptible to polypharmacy than males, our study suggests that this difference between older males and females in the U.S. may have diminished over the past 20 years [42]. Several factors may underline the observed convergence. One potential driver is the more rapid increase in multimorbidity among males. For example, King et al. reported that multimorbidity prevalence in men increased from 45.6% (1988–1994) to 56.1% (2013–2014), while in women it rose from 50.2% to 58.7% [43]. Furthermore, shifts in the use of sex-specific preventive therapies may contribute to the trend. The substantial decline in hormone therapy use among women following the 2002 Women’s Health Initiative trial likely reduced medication burden in this group [44]. Other factors contributing to the significant increase in male polypharmacy include improved access to preventive services, earlier detection of subclinical conditions, and more timely diagnosis and treatment of chronic diseases.

In our study, the prevalence of polypharmacy and hyperpolypharmacy was comparable between Whites and African Americans, whereas prior studies have inconsistently reported racial differences likely due to variations in study cohorts [10, 11, 45]. Overall, our study showed that Asians had a relatively lower prevalence of polypharmacy and hyperpolypharmacy, although some overlapping CIs were observed due to the limited sample size in Asians. The prevalence of polypharmacy in Asians increased significantly over the study period (+ 0.83%/year [0.32 ~ 1.33], p=.003). Nationally representative data on polypharmacy trends among Asian populations remain limited, and prior studies have often undersampled or aggregated racial minorities, potentially obscuring meaningful trends [46, 47]. Several factors may contribute to the observed increase including a growing Asian American population and improved representation in more recent datasets [46, 47]. Enhanced access to health insurance and healthcare may have been key in obtaining prescribed medications. A recent report from the U.S. Department of Health and Human Services indicated that the uninsured rate among Asian American and Native Hawaiian/Pacific Islander populations declined from 16.6% (2010) to 6.2% (2022), suggesting a broader trend of improved access to care [48]. While this report focused on non-elderly populations, insurance coverage and increased access to preventive and primary care earlier in life may influence healthcare utilization and medication use patterns in older age [49]. Nevertheless, the trends in polypharmacy and hyperpolypharmacy among Asians and their explanations warrant further exploration.

The implementation of Medicare Part D in 2006, followed by the Affordable Care Act in 2010, greatly improved medication access for older adults by expanding prescription coverage and lowering out-of-pocket costs [50]. However, this also led to higher prescription expenditures with the average total expenditure for the elderly increasing from $2,635 (2009) to $3,288 (2016) [50, 51]. Our study further analyzed trends in prescribed medication expenditure based on medication burden. Medication expenditures kept rising for older adults with hyperpolypharmacy, while those with polypharmacy saw only modest growth. The mean number of prescribed medications per older adult with hyperpolypharmacy across the study period (Supplementary Table 7) indicates that the increase may not be due to an increasing medication load per individual. Potential explanations include rising prescription costs [51], which are more influential among the hyperpolypharmacy population and warrant further investigation. Our findings on the rising prescribed medication expenditure in U.S. older adults highlight the need for systematic medication reviews and targeted deprescribing strategies to reduce unnecessary use and associated costs in this population with excessive medication burden. Clinically, implementing deprescribing interventions in multimorbid or frail populations helps mitigate adverse drug events, improve quality of life, and optimize therapeutic outcomes while controlling healthcare costs [52, 53]. Given the continued escalation in medication expenditures, addressing hyperpolypharmacy through integrated clinical and policy interventions should be a national priority in aging care.

Beyond the documented population-level prevalence, disparities, and cost in this study, poly- and hyperpolypharmacy impose substantial and often overlooked medication burdens on older adults and the family caregivers who support their medication management. The MEMORABLE realist synthesis identified five interrelated burdens that accumulate across stages of medication management and can overwhelm patients and carers when support is insufficient [54]. The study also described a multi-stage model in which medication management involves cumulative workload, emotional strain, and complex decision-making shared among older adults, caregivers, and practitioners [54]. Complementary qualitative work showed that family carers face fragmented information, unclear responsibilities, and anxiety about medication safety, demonstrating that polypharmacy-related burden extends beyond clinical risks to affect autonomy and quality of life for both parties [55]. Addressing medication burden requires integrating patient and caregiver perspectives into routine care through shared decision-making, proactive medication reviews, and structured deprescribing strategies [54, 55]. Alignment of policy initiatives and care delivery models with these patient-centered approaches will be essential to reducing the burdens of poly- and hyperpolypharmacy and supporting more manageable prescribing for the aging population.

This study has several limitations. First, the analysis was limited to community-dwelling older adults. Institutionalized populations are not included in MEPS. Given that polypharmacy and hyperpolypharmacy are typically more prevalent in institutionalized populations, caution is warranted when generalizing these findings to populations beyond the study sample [5]. Second, MEPS captures only prescribed medications and does not include over-the-counter medication and health supplements, which may result in an underestimation of total medication burden. Third, unlike surveys such as NHANES that assess prescriptions over a shorter 30-day window, MEPS uses a recall period of approximately 5 to 6 months, which reflects cumulative rather than simultaneous polypharmacy. Fourth, interchangeable drugs for the same indication due to preference, cost, or side effects were not addressed in this study, and long-term use (e.g., ≥ 2 fills within observation period), which may have inflated the number of medications taken during the recall period. Fifth, the documented medications may not accurately reflect their actual use and adherence. Last, our analyses did not stratify respondents by insurance type or Medicare–Medicaid dual eligibility, both of which represent important indicators of access and medication complexity among older adults [56, 57]. Future studies using nationally representative data should further examine longitudinal changes and differences in medication burden and associated costs across insurance coverage groups.

Conclusion

In this nationally representative study, polypharmacy and hyperpolypharmacy remain common among community-dwelling older adults in the U.S. with variation by age, sex, and race. Furthermore, prescribed medication expenditures among those with hyperpolypharmacy continue to rise significantly. These findings indicate the critical need for ongoing monitoring, tailored prescribing optimization efforts targeting subgroups with the high prescribing burden, and policies to reduce the clinical and economic burdens of polypharmacy and hyperpolypharmacy in aging populations.

Abbreviations

ADLs

Activities of Daily Living

CCI

Charlson Comorbidity Index

CI

Confidence Interval

HER

Electronic Health Record

ICD-9

International Classification of Diseases, Ninth Revision

ICD-10

International Classification of Diseases, Tenth Revision

IADLs

Instrumental Activities of Daily Living

MPC

Medical Provider Component

MEPS

Medical Expenditure Panel Survey

NDC

National Drug Code

NHANES

National Health and Nutrition Examination Survey; SE, Standard Error

STROBE

Strengthening the Reporting of Observational Studies in Epidemiology

U.S.

United States

USD

United States Dollar

Authors’ contributions

Conceptualization, A.F., C.M.C., D.M.J.; methodology, C.M.C., S.F., H.H, and D.M.J.; formal analysis, C.M.C., S.F., H.H., H.L, A.R.P. and D.M.J.; data curation, A.F., C.M.C. and S.F.; writing—original draft preparation, H.H. C.M.C, and D.M.J; writing—review and editing, all authors; supervision, D.M.J.

Funding

D.M.J. is supported by the National Institutes of Health/National Heart, Lung, and Blood Institute Loan Repayment Program (K23HL153582). Research reported in this publication was supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under award Number UM1TR005296 to the University at Buffalo. This content is those of the authors and should not be construed as the official position or policy of, nor should any endorsements be inferred by NIH, HRSA, HHS or the U.S. Government.

Data availability

The data for this study is available from the Medical Expenditure Panel Survey, a publicly available dataset that can be accessed for research purposes after a standard application and approval process.

Declarations

Ethics approval and consent to participate

This research uses Medical Expenditure Panel Survey (MEPS) data files, which are publicly available and contain only de-identified information. Pursuant to federal regulations governing human subjects research, the use of publicly available, de-identified secondary data does not constitute human subjects research and is therefore exempt from Institutional Review Board (IRB) oversight. The conduct of this research adheres to the ethical principles for research involving human data as outlined in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.