Effects of Medicare Eligibility at Age 65 Among Individuals With and Without Functional Disability

Abstract

Background

Medicare coverage at age 65 improves access to and use of care and alleviates financial hardship for the general population. However, less is known whether the effects differ between individuals without and with functional disability.

Objectives

To examine the effects of Medicare eligibility at age 65 on health insurance coverage, financial burden of care, and access to care among individuals without and with functional disability.

Design

We used a regression discontinuity design, which exploits the discontinuity in eligibility for Medicare at age 65 and compares individuals just before and after age 65.

Participants

Our analysis included 19,876 individuals (aged 59–71) without functional disability and 8376 individuals with functional disability from the 2014–2021 Medical Expenditure Panel Survey.

Main Measures

We assessed health insurance coverage, financial burden of care, and access to care.

Results

Medicare eligibility led to increases in any and Medicare coverage for both groups, but those with functional disability had a decrease in Medicaid coverage by − 2.6 percentage points. Medicare eligibility resulted in lower financial burden of care for both groups, but the effects were greater among those with functional disability (− $578 vs. − $344 for out-of-pocket spending, − 3.7 vs. − 4.9 percentage points for cost-sharing, and − 2.5 vs. − 0.8 percentage points for paying medical bills over time). Although Medicare eligibility led to a decrease in delayed medical care among those without functional disability (− 2.1 percentage points), no change was observed among those with functional disability. Notably, access to care remained limited among those with functional disability after obtaining Medicare eligibility (8.6% and 3.9% for being unable to get medical care and experiencing delay in getting medical care).

Conclusion

Medicare coverage can reduce financial hardship, especially for individuals with functional disability. However, there is a need to develop policies that ensure equitable access to care for those with functional disability.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11606-024-09060-7.

Functional disability, defined as acquired difficulties in performing daily tasks or maintaining independent living, becomes increasingly prevalent with advancing age in the USA.¹ In 2019, 66.1% of US older adults reported experiencing functional limitations, with 46.8% indicating significant difficulties and 19.3% reporting some level of impairment.² This trend often results from a complex interplay of age-related physical and cognitive changes, chronic health conditions, and social factors.^3,4 Adults with functional disability experience a high degree of co-occurring medical, behavioral health, and social complexity, incurring greater health care expenditures.^5–12

The additional costs associated with functional disability can impose a disproportionately high financial burden, but Medicare coverage has the potential to reduce financial hardship for individuals with functional disability, possibly improving access to care. Medicare is one of the largest public health insurance programs in the USA. Most Americans become eligible for Medicare at age 65. Near-universal access to Medicare coverage at age 65 decreases the financial burden of care and improves access to care,^13–18 potentially improving health outcomes. This phenomenon is more pronounced among marginalized populations, including racial/ethnic minorities, immigrants, individuals with chronic conditions, and individuals with food insecurity.^14,19–22

However, less is known about the impacts of Medicare coverage among individuals with functional disability. On the one hand, Medicare coverage may decrease financial hardship for individuals with functional disability. Evidence suggests that individuals with functional disability had substantially high unmet need for care.^23–26 This may be attributable to insufficient health insurance coverage. While some people with functional disability qualify for Medicare coverage before age 65 through Social Security Disability Insurance, many do not because they do not meet Social Security’s definition of disability as the inability to work any significant job due to a medically diagnosable physical or mental condition that either will end in death or has lasted/will last at least 12 months. On the other hand, Medicare coverage may not necessarily translate in improvements in access to care for individuals with functional disability as Medicare coverage for disabling conditions is limited.²⁷ Medicare covers acute and post-acute care services, including inpatient hospital stays, skilled nursing care, hospice, and some home health care, but has limited coverage of disabled conditions, including long-term care, home health care, and durable medical equipment.²⁸ Consequently, Medicare beneficiaries with functional disabilities experience challenges in accessing care and often experience unmet need for care compared to Medicare beneficiaries without functional disabilities.^23,29–31

In this study, we examined the effects of Medicare eligibility at age 65 on health insurance coverage, financial burden of care, and access to care among individuals without and with functional disability. We hypothesized that Medicare eligibility would alleviate the financial burden of care for both groups, but the effects would be larger among those with functional disability than those without functional disability. We also hypothesized that Medicare eligibility would lead to a decrease in delayed medical care among those without functional disability, but no change would be observed among those with functional disability, potentially due to the limited Medicare coverage for services that support disabling conditions.

METHODS

Data

We used data from the 2014–2021 Medical Expenditure Panel Survey (MEPS), which is a nationally representative survey of the US civilian non-institutionalized population. The survey collects information on participants’ demographic, socioeconomic, and health-related characteristics. MEPS collects data from two main sources: the Household Component (HC) and the Medical Provider Component (MPC). In our study, we used the full-year consolidated data files from the HC and the medical conditions files from the MPC. The COVID-19 pandemic caused significant disruptions to field operations of the survey. However, a recent examination shows that the quality of the collected data has been minimally affected.³²

Study Sample

We first identified adults aged 18 and older from the MEPS (n = 180,814). We then applied a data-driven method to determine an optimal age range (bandwidth) centered around age 65.³³ This approach is crucial because individuals significantly younger or older than 65 may have different characteristics that could confound the results. By focusing on a more homogeneous age range, we could reduce potential errors and improve the precision of causal inference. Consistent with prior research,^13–18 a range of 6 years was selected. Thus, our study sample includes individuals aged 59–71 (n = 35,484). We excluded individuals with missing data from the analysis (n = 1673). Furthermore, individuals exactly at age 65 were not included as this population might not have been eligible for Medicare the first day of the month they turn 65, given that eligibility begins on the first day of the month in which they turn 65 (n = 2808).

Next, we categorized individuals into two groups: individuals without and with functional disability. Following prior research,²⁴ we identified functional disability based on six questions assessing difficulties in seeing, hearing, memory/concentration, walking, self-care, and performing errands due to a physical, mental, or emotional condition. The initial version of these questions was developed and tested in the 1990s with a focus on identifying disabling conditions. Currently, these six questions are used in over ten national surveys conducted by various US federal departments and agencies, such as the Census Bureau, and the Departments of Education, Labor, Housing and Urban Development, and Justice.³⁴ Moreover, the six-item set adheres to the established data standard for survey questions on disability codified by the US Department of Health and Human Services, serving as a fundamental measure.³⁵ These following questions from the MEPS included “Does anyone in the family have any difficulty seeing?”, “Does anyone in the family have any difficulty hearing?”, “Do any of the adults in the family experience confusion or memory loss such that it interferes with daily activities?”, “Does anyone in the family have any difficulty walking, climbing stairs, grasping objects, reaching overhead, lifting, bending or stooping, or standing for long periods?”, “Does anyone in the family receive help or supervision with personal care such as bathing, dressing, or getting around the house?”, and “Because of a physical, mental, or emotional condition, do you have difficulty doing errands alone such as visiting a doctor’s office or shopping?” If the response was “yes”, a follow-up question was asked to determine which household member(s) had difficulty. We assigned a value of 1 to indicate the presence of a difficulty and 0 to indicate none. The scores from the six questions were then aggregated. Functional disability was assessed using a binary measure, categorized as either absent (no difficulties in any of the six questions) or present (at least one difficulty in any of the six questions).

Outcomes

We included three types of outcomes. First, we included measures of health insurance coverage: any coverage (defined as being insured), Medicare coverage (including Medicare only, Medicare with private coverage, and Medicare with Medicaid), Medicaid coverage, no Medicare but any forms of private or public coverage. These categories are not mutually exclusive. Second, we included measures of the financial burden of care: annual out-of-pocket spending per person (defined as the total amount of health care expenses covered directly by individuals or families), cost-sharing (calculated as the percentage of annual out-of-pocket spending divided by total health care spending), problems paying medical bills, and paying medical bills over time. Out-of-pocket spending was adjusted to 2021 dollars using the Personal Consumption Expenditures Price Index for health care. Finally, we included measures of access to care: being unable to get medical care and experiencing delay in getting medical care. We treated out-of-pocket spending and cost-sharing as continuous variables while all others were binary. Detailed questionnaires for each outcome are provided in the ESM Appendix.

Covariates

To control for characteristics that may confound the effect of Medicare eligibility, we included race/ethnicity, sex, having a usual source of care, immigrant status, employment status, marital status, education, family income, census region of residence, number of chronic conditions a total of 20 chronic conditions from the medical condition files, and health-related quality of life from the 12-item Short Form Survey. The SF-12 score is a standardized health measure scored on a scale of 0 to 100, with higher scores indicating better health.

Statistical Analysis

We used a regression discontinuity design to estimate the effects of Medicare eligibility at age 65 on individuals without and with functional disability. Regression discontinuity addresses selection bias in comparing individuals with and without an intervention by exploiting an arbitrary cutoff in program eligibility, which offers transparent visual evidence of changes in the outcome at the cutoff.^36,37 The approach relies on the assumption that individuals within a narrow bandwidth around the cutoff have similar observed and unobserved characteristics other than the likelihood of obtaining Medicare coverage.

We first examined discontinuities in outcomes at the Medicare eligibility cutoff. Then, we examined the effects of Medicare eligibility at age 65 on the outcomes of interest. Specifically, we conducted a parametric regression discontinuity model with a quadratic age trend while adjusting for the individual-level covariates described above. We separately conducted the analysis for individuals without and with functional disability. To determine whether the effects of Medicare eligibility were more pronounced among individuals with functional disability than individuals without functional disability, we analyzed an interaction term between functional disability and Medicare eligibility at age 65. Moreover, we conducted sensitivity analyses by changing the bandwidth to 5 or 7 years, excluding individuals from the 2020–2021 MEPS, excluding individuals with Medicaid coverage, excluding individuals with Medicare before age 65, and using alternative models. Furthermore, prior research suggests that Medicare eligibility at age 65 led to a positive impact on self-reported health,¹⁸ potentially leading to shifts in self-reported functional disability. As this may result in a change in sample composition at age 65, we examined whether the prevalence of functional disability changed significantly at age 65. For all analyses, we included year-fixed effects. We also used survey weights to generate nationally representative estimates and accounted for the complex survey design in standard error estimation. While some individuals participated across multiple years, we treated the data for each year as an independent annual cross-sectional survey. Thus, we adjusted the SEs for clustering within individual.

RESULTS

Sample Characteristics

Our final sample included 19,876 individuals without functional disability and 8376 individuals with functional disability. Overall, demographic, socioeconomic, and health characteristics did not differ substantially between individuals below and above age 65 years in both groups, except for employment status (Table 1).

Table 1.

Characteristics of Individuals Without and With Functional Disability Before and After Medicare Eligibility at Age 65 Years, 2014–2021

	Without functional disability		With functional disability
Characteristics	Age 59–64 y (N = 10,797; weighted N = 124,849,044)	Age 66–71 y (N = 9079; weighted N = 94,177,311)	Age 59–64 y (N = 4080; weighted N = 40,825,688)	Age 66–71 y (N = 4296; weighted N = 41,441,417)
Race/ethnicity, %
White	58.7	63.7	56.5	62.5
Black	17.4	13.0	15.7	14.6
Hispanic/Latino	16.2	14.9	22.0	16.7
Asian	5.7	6.4	2.1	2.9
Other or multiple	1.9	2.0	3.7	3.3
Sex, %
Men	46.8	45.4	42.2	45
Women	53.2	54.6	57.8	55.0
Having a usual source of care, %	83.1	90.3	89.1	92.8
Immigrant, %	77.8	81.4	87.0	86.5
Employed, %	66.5	33.1	29.9	17.3
Married, %	62.9	61.4	41.8	47.8
Education, %
Less than high school	12.4	11.1	19.8	17.9
High school	40.0	38.4	46.1	43.5
College	40.9	45.2	27.7	32.2
Family income, %
< 199% of FPL	24.8	26.2	53.6	43.4
200–399% of FPL	27.9	27.7	24.3	27.6
> 399% of FPL	47.3	46.1	22.1	29.0
Region, %
Northeast	17.3	15.8	15.4	13.7
Midwest	20.9	20.4	22.5	22.2
South	36.4	37.5	42.7	41.8
West	25.4	26.3	19.4	22.3
Number of chronic conditions, %
0	59.4	48.9	26.9	24.7
1–2	35.6	43.3	49.2	51.0
3–5	4.9	7.6	21.9	22.5
6 +	0.1	0.2	2.0	1.7
Health-related quality of life, median (25th, 75th percentile)
SF-12 physical component score	53 (47, 56)	52 (44, 56)	35 (26, 46)	36 (27, 46)
SF-12 mental component score	55 (50, 58)	57 (52, 59)	48 (39, 56)	52 (42, 58)

Abbreviation: FPL, federal poverty level

Health Insurance Coverage

Medicare eligibility at age 65 led to increases in any and Medicare coverage for both groups, but individuals with functional disability had a decrease in Medicaid coverage (Table 2). Medicare eligibility led to significant increases in Medicare coverage among those without and with functional disability, but the increase was relatively smaller among those with functional disability than those without functional disability (61.5 percentage points [95% CI, 57.2–65.8] and 86.4 percentage points [95% CI, 85.4–87.4]). In both groups, the increase was primarily driven by those with Medicare only or those with Medicare combined with private coverage. Although the proportion of individuals with both Medicare and Medicaid increased, the magnitude of this increase was relatively modest. Any insurance coverage also increased at age 65, but the increase was relatively low among both groups (3.8 percentage points [95% CI, 2.7–4.9] and 4.5 percentage points [95% CI, 4.0–5.1] for those without and with functional disability). Notably, those with functional disability had a significant decrease in Medicaid coverage by 2.6 percentage points (95% CI, − 4.0, − 1.2). Health insurance coverage among individuals without and with functional disability by age is presented in ESM Appendix Figures A-B.

Table 2.

Health Insurance, Financial Burden of Care, and Access to Care Among Individuals Without and With Functional Disability Before and After Medicare Eligibility at Age 65 Years, 2014–2021

	Without functional disability				With functional disability
	Unadjusted values, weighted % or median (25th, 75th percentile)		Regression discontinuity at age 65 years		Unadjusted values, weighted % or median (25th, 75th percentile)		Regression discontinuity at age 65 years
Outcomes	Age 59–64 y (N = 10,797; weighted N = 124,849,044)	Age 66–71 y (N = 9079; weighted N = 94,177,311)	Adjusted absolute change, estimates (95% CI)	Adjusted relative change, %	Age 59–64 y (N = 4080; weighted N = 40,825,688)	Age 66–71 y (N = 4296; weighted N = 41,441,417)	Adjusted absolute change, estimates (95% CI)	Adjusted relative change, %
Health insurance
Any coverage, %	93.5	99.6	4.5 (4 to 5.1)	4.8	93.9	99.8	3.8 (2.7 to 4.9)	4.0
Medicare coverage, %	4.8	96.3	86.4 (85.4 to 87.4)	1800.0	32.2	97.1	61.5 (57.2 to 65.8)	191.0
Medicare only, %	2.6	43.7	33.4 (32.2 to 34.6)	1284.6	14.7	44.5	23.2 (15.8 to 30.7)	157.8
Medicare and private, %	0.9	47.8	49.3 (47.2 to 51.3)	5477.8	6.5	38.7	30.9 (19.5 to 42.3)	475.4
Medicare and Medicaid, %	1.2	4.4	3.2 (2.6 to 3.8)	266.7	10.8	13.1	5.9 (2.8 to 9.1)	54.6
Medicaid coverage, %	6.6	5.2	− 1.4 (− 2.6 to 0.1)	− 21.2	24.6	14.4	− 2.6 (− 4 to − 1.2)	− 10.6
No Medicare but any public or private, %	84.3	2.9	− 77.9 (− 79.1 to − 76.6)	− 92.4	56.8	2.2	− 51 (− 56.3 to − 45.8)	− 89.8
Financial burden of care
Out-of-pocket spending, $	489 (107, 1354)	556 (171, 1313)	− 344 (− 439 to − 249)	− 30.1	572 (141, 1700)	679 (214, 1819)	− 578 (− 1012 to − 144)	− 37.9
Cost-sharing, %	18.8 (0.1, 40.9)	14.9 (5.5, 31.8)	− 4.9 (− 7 to − 2.7)	− 17.6	9.5 (1.8, 25.4)	9.2 (2.7, 22.0)	− 3.7 (− 6.9 to − 0.5)	− 19.3
Paying medical bills over time, %	15.6	11.8	− 0.8 (− 3.3 to 1.7)	− 5.1	26.8	18.7	− 2.5 (− 4.8 to − 0.2)	− 9.3
Problems paying medical bills, %	7.2	5.0	− 0.4 (− 2.0 to 1.1)	− 5.6	20.6	13.2	− 1.8 (− 5.6 to 2)	− 8.7
Access to care
Unable to get medical care, %	1.7	0.9	− 1.1 (− 2.4 to 0.3)	− 64.7	8.5	3.9	− 0.7 (− 4.9 to 3.5)	− 8.2
Delay in getting medical care, %	5.4	3.2	− 2.1 (− 3.4 to − 0.9)	− 38.9	14.5	8.6	2.2 (− 0.4 to 4.7)	15.2

Functional disability was determined by binary responses to six questions assessing difficulties in seeing, hearing, memory/concentration, walking, self-care, and performing errands due to a physical, mental, or emotional condition and categorized into two levels: without functional disability (no difficulties on all six questions) and with functional disability (≥ 1 difficulties). The effects of Medicare eligibility at age 65 years was estimated using a parametric regression model with a quadratic age trend. All adjusted regressions controlled for (self-reported) race/ethnicity, sex, employment status, marital status, education, family income, US census region of residence, number of chronic conditions, health-related quality of life from the 12-item Short Form Survey, and year-fixed effects. Survey weights were used to adjust sample characteristics to be representative of the US population. We excluded individuals aged 65 years as they might not be eligible for Medicare for a full year. All characteristics were reported as nationally representative weighted percentages

Financial Burden of and Access to Care

The effects of Medicare eligibility at age 65 on the financial burden of care and access to care differed between individuals without and with functional disability (Figs. 1 and 2 and Tables 1 and 2). Medicare eligibility resulted in lower out-of-pocket spending and cost-sharing for both groups, but the effects were more pronounced among those with functional disability (− 37.9% and − $578 [95% CI, − 1012, − 144] vs. − 30.1% and − $344 [95% CI, − 439, − 249] for out-of-pocket spending and − 19.3% and − 3.7 percentage points [95% CI, − 6.9, − 0.5] vs. − 17.6% and − 4.9 percentage points [95% CI, − 7.0, − 2.7] for cost-sharing). Furthermore, Medicare eligibility led to a significant decrease in paying medical bills over time among those with functional disability by 9.3% (95% CI, − 2.5 percentage points [95% CI, − 4.8, − 0.2]). However, no change was observed among those without functional disability. Finally, Medicare eligibility led to a significant decrease in delayed medical care among those without functional disability by 38.9% (− 2.1 percentage points [95% CI, − 3.4, − 0.9]). However, there was no significant change among those with functional disability. Notably, access to care remained limited among those with functional disability after obtaining Medicare eligibility (8.6% and 3.9% for being unable to get medical care and experiencing delay in getting medical care).

Figure 1.

Financial burden of care and access to care among individuals without functional disability by age.

Figure 2.

Financial burden of care and access to care among individuals with functional disability by age.

We also found that the interaction term between functional disability and Medicare eligibility was statistically significant for all outcomes, except for insurance coverage and out-of-pocket spending (ESM Appendix Table A).

Although functional disability becomes more prevalent with age, there was no clear evidence that the prevalence of functional disability increased significantly at age 65 (ESM Appendix Figure C). Our sensitivity analyses were largely consistent with the main findings, showing that Medicare eligibility at age 65 was associated with larger decreases in financial burden of care among those with functional disability than those without functional disability, but did not lead to a decrease in delayed medical care among those with functional disability (ESM Appendix Tables B-C).

DISCUSSION

We found differential effects of near-universal access to Medicare coverage at age 65 between individuals without and with functional disability. First, Medicare eligibility led to increases in any insurance and Medicare coverage for both groups, but those with functional disability had a decrease in Medicaid coverage. Second, Medicare eligibility resulted in lower financial burden of care for both groups, but the effects were greater among those with functional disability. Finally, Medicare eligibility led to a decrease in delayed medical care among those without functional disability, but no change was observed among those with functional disability.

Our study suggests that Medicare coverage plays a critical role in improving care for individuals with functional disability. This reaffirms prior evidence that Medicare coverage is associated with gains in access to care and financial burden of care.^13–21 Notably, larger improvements in these outcomes were detected among individuals with functional disability. This is likely because this population has greater needs for medical care and Medicare offers extensive coverage for many basic health services, which can lower financial burden of care. Furthermore, prices for Medicare-covered services are often lower than prices in commercial insurance.³⁸ Consequently, individuals with functional disability may have experienced lower out-of-pocket costs than those with private coverage or those without any insurance coverage. Another notable finding is that Medicare eligibility resulted in a decrease in Medicaid coverage among individuals with functional disability. This finding suggests a substitution effect, where individuals may seek alternative options due to limited Medicaid coverage.

Our study offers evidence of gaps in Medicare coverage for individuals with functional disability consistent with prior research showing that Medicare beneficiaries with disabilities have high need for care, but unmet need for care remain high.^3,11 However, our finding necessitates a cautious interpretation because we did not find evidence that Medicare eligibility lowered access to care among individuals with functional disability. As access to care remained limited among individuals with functional disability even after obtaining Medicare coverage, this contextually implies that the role of addressing the multifaceted needs of individuals with functional disabilities might be limited. Thus, multiple policy interventions may be needed. For example, Medicare coverage for disabling conditions is limited for home health care and durable medical equipment.²⁸ Thus, expanding Medicare coverage could improve access to adequate services and potentially reduce unmet need for care. Also, developing models that integrate the financing and delivery of services covered in Medicare and Medicaid could address social needs both inside and outside of health care settings.^31,32 Individuals with functional disability not only suffer from medical risk factors but also are likely to suffer from social risk factors, underscoring the unique needs of those with disabilities.¹²

Limitations

This study has several limitations. First, the validity of the regression discontinuity counts on several assumptions. One is that the outcomes of interest would evolve smoothly with age in the absence of Medicare eligibility at age 65. However, life changes at the age of retirement may impact the outcomes differently. Another is that individuals cannot manipulate Medicare eligibility at age 65, but may anticipate coverage and delay care until age 65. However, evidence suggests that this is unlikely to happen.³⁹ Second, we could not adjust for all potential comorbidities and confounders. However, the regression discontinuity design helps to alleviate such bias by comparing individuals with similar characteristics above and below the age cutoff. Third, we employed a widely recognized index to assess functional disabilities, but there are some concerns about the validity of the measure. The measure failed to identify about 20% of adults aged 18–64 who reported having disabilities.⁴⁰ Fourth, the impacts of Medicare eligibility are more likely to be strongest for the uninsured, but we could not directly test the impacts due to limited availability of longitudinal data. Fifth, we examined a limited set of outcomes, and thus our findings may not be applicable to other types of outcomes. Relatedly, our outcome measures were self-reported and may be subject to reporting bias. Finally, we treated the data for each year as an independent annual cross-sectional survey, but this assumption may not fully apply because some individuals might participate across multiple years.

CONCLUSIONS

Our analysis showed that Medicare eligibility at age 65 led to greater decreases in financial hardship among individuals with functional disability than individuals without functional disability. Despite persistent high levels of unmet need for care among individuals with functional disabilities, no improvement was observed. These findings suggest that near-universal access to Medicare coverage has the potential to reduce financial hardship for individuals with functional disability, yet limited Medicare coverage for disabling conditions may not alleviate unmet need for care. Alternative policies are needed to ensure that individuals with functional disability have improved access to care.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding:

This work was partly supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT).

Declarations:

Conflict of interest:

The authors declare that they do not have a conflict of interest.