Abstract
Objective
To determine the incidence and prevalence of dementia in a nationally representative cohort of US Medicare beneficiaries, stratified by important subgroups.
Design
Population based study.
Setting
Nationwide study between 2015 and 2021.
Participants
Fee-for-service Medicare beneficiaries aged 66 or older with at least one year of continuous enrollment.
Main outcome measures
Incidence and prevalence of dementia, calculated as percentage per person years or percentage of beneficiaries respectively. These metrics were also calculated in key subgroups defined by age, sex, race/ethnicity, and neighborhood socioeconomic status.
Results
A total of 5 025 039 incident cases of dementia were documented from 2015 to 2021. The overall age and sex standardized incidence decreased between 2015 and 2021 from 3.5% to 2.8%. Prevalence increased during this time from 10.5% to 11.8%. Male beneficiaries had a higher age standardized incidence than did female beneficiaries in 2015 (3.5% v 3.4%), a difference that widened by 2021 (2.9% v 2.6%; estimated difference-in-difference 0.94, 95% confidence interval (CI) 0.94 to 0.95; P<0.001). Incidence was highest in 2015 for black beneficiaries (4.2%), followed by Hispanic beneficiaries (3.7%) and white beneficiaries (3.4%), and in 2021 for black beneficiaries (3.1%) followed by white beneficiaries (2.8%) and Hispanic beneficiaries (2.6%); the difference between white and black beneficiaries narrowed from 2015 to 2021 (difference-in-difference 0.92, 95% CI 0.91 to 0.93; P<0.001) as did the difference between white and Hispanic beneficiaries (difference-in-difference 0.88, 0.87 to 0.89; P<0.001).
Conclusions
The incidence of dementia decreased from 2015 to 2021, but the prevalence increased. Disparities in these measures by race/ethnicity, sex, and neighborhood socioeconomic status should motivate future measures to promote health equity.
Introduction
The rapid aging of the US population has elevated dementia to a prominent position among medical and public health priorities. Dementia is a broad clinical syndrome encompassing several underlying pathologies that ultimately result in cognitive impairment. The burden of dementia will more than double to affect nearly 14 million people by 2060,2 with far reaching social and economic implications. A growing body of evidence suggests that the incidence and impact of dementia have important sociocultural implications, as it is disproportionately shouldered by marginalized communities including black Americans and people living in areas of socioeconomic deprivation.3 4 5
High quality evidence on recent trends in the incidence and prevalence of dementia in routine clinical practice is scarce. Previous studies have principally used survey and geographically restricted cohort designs to assess the burden of dementia.2 6 7 8 9 10 11 12 However, patients who participate in cohort and survey studies likely differ systematically from the broader population of patients with dementia, as patients residing in long term care facilities, those with more severe cognitive impairment, and those from groups historically under-represented in research are less likely to be included in other study designs. For example, a study of the National Alzheimer’s Coordinating Centers (NACC) found that, compared with the general population, participants enrolled in NACC studies were older, more educated, and less likely to be from demographic groups historically under-represented in biomedical research.13 14 15 16 17 18 Furthermore, large scale studies are needed to rigorously evaluate trends in diagnosis of dementia across demographic groups with sufficient power, especially given under-representation of non-white and Hispanic patients in existing studies and the fact that major secular changes in the US healthcare system have taken place since 2015, such as expansion of Medicaid in many states.18 19 20 Nationwide studies estimating the incidence and prevalence of dementia in routine clinical care would capture a broad population of patients with dementia and produce information directly relevant to health systems and policy makers. This approach has successfully been used for studies of stroke, cardiovascular disease, and other chronic conditions among older adults in the US.21 22 23
The objective of this study was to determine the incidence and prevalence of dementia by race, sex, and neighborhood deprivation in a large, national cohort of fee-for-service Medicare beneficiaries with sufficient sample size to allow analyses of these key subgroups, with specific attention paid to providing information on temporal changes in the total population burden of dementia that would be relevant to policy makers, which has not been previously reported in the literature. Given meaningful secular changes in the US since 2013, this study assessed the hypothesis that the incidence of dementia would decrease while the prevalence would increase in routine clinical practice in the US.
Methods
Study design and setting
This was a retrospective, nationwide study of repeated cross sectional (annual) Medicare claims. Data came from the Centers for Medicare and Medicaid Services (CMS) Virtual Research Data Center (VRDC) and included all inpatient, outpatient, and carrier fee-for-service Medicare claims for the years 2013-21. Medicare is a direct insurance program of the US government that covers almost 99% of adults over the age of 6524; the program is subdivided into two constituent programs—traditional (fee-for-service) Medicare, in which the US government directly pays for medical care, and Medicare Advantage, in which private health insurance companies receive funds from the government and provide care management services. The claims files used in this study contain computerized records of patient encounters in these settings for all participants enrolled in traditional Medicare plans. Beneficiaries enrolled in Medicare Advantage plans, which are managed by private insurance companies, are not included in these data. We evaluated trends from 2015 to 2021 to ensure at least two years of preceding data to minimize misclassification of prevalent dementia as incident dementia, on the basis of previous literature suggesting that one year look back periods may result in higher proportions of misclassification.25
Study population
For the analysis of dementia incidence, we required beneficiaries to have one preceding year of fee-for-service enrollment with no claims for dementia in any position on an inpatient, outpatient, or carrier claim (based on codes previously described, reproduced in supplementary table A).26
Variables
The variables of interest were year (to assess changes over time), race/ethnicity, sex, and neighborhood socioeconomic deprivation. For race/ethnicity, we used Medicare enrollment data to classify beneficiaries into the following mutually exclusive categories by using a previously published algorithm: black (or African American); Hispanic, other, and non-Hispanic white.27 We classified sex as male or female according to Medicare enrollment records, which are derived from Social Security Administration records that permit beneficiaries to change their recorded sex to concord with their gender identity; otherwise, sex is determined from birth certificates or other government documentation at time of social security registration. We measured neighborhood socioeconomic deprivation by using the Area Deprivation Index, which is a summary score at the census block group level. It incorporates social and economic indicators including income, employment, infrastructure, housing, and education.28 This score is reported as a national centile, with 1 representing the least deprived and 100 the most deprived neighborhoods. We used ZIP+4 codes to link beneficiaries to their corresponding census block groups.29
Endpoints
Endpoints were the age and sex standardized incidence and prevalence of dementia, which we treated as one composite entity, consistent with its classification as a broad syndrome composed of many specific underlying pathologies or etiologies. In each year, beneficiaries were included in the incident group if they had an index diagnosis of dementia before 31 December of the study year. We calculated person years as the proportion of days in each denominator year that a person was alive with fee-for-service enrollment before diagnosis of dementia. Beneficiaries were included in the prevalent dementia group for a given study year if they had a diagnosis of dementia at any point before 31 December of that year. Additionally, a patient with a diagnosis of dementia in any given year was carried forward and considered prevalent in the subsequent years until death or disenrollment. We also reported mortality in the incident cohort, which we defined as the percentage of beneficiaries with a diagnosis of dementia who died in that year, and in the prevalent cohort, which we defined as the percentage of beneficiaries who died in the subsequent calendar year.
Statistical analysis
We described baseline characteristics of the incident cohort by using means with standard deviations for continuous variables and frequencies with percentages for categorical variables. We used χ2 tests for categorical variables and Kruskal-Wallis tests for continuous variables to examine between group differences. We calculated annual incidence rates of dementia as percentage per person years and prevalence as percentage of beneficiaries. We estimated age and sex standardized prevalence and incidence rates by using a model based approach. Age was categorized into 5 year groups (65-69, 70-74, 75-79, 80-84, 85-89, ≥90). For each subgroup of interest, we calculated weights to make the age/sex distribution the same across each subgroup category by dividing the overall age/sex percentage by the age/sex percentage within each subgroup category. We then used weighted Poisson regression to estimate the standardized prevalence and incidence rates.
Participants were eligible in each study year if they were alive, aged ≥66 years, and living in the US on 1 January of each study year and were enrolled in fee-for-service Medicare for the previous 12 months. We calculated person years as the proportion of days in each denominator year that a person was alive with fee-for-service Medicare enrollment. Incidence rates were stratified by study year and by race (white, black, Hispanic, and other race/ethnicity), sex, and tenth of Area Deprivation Index. Heat maps were produced to present 2015 and 2020 prevalence and incidence rates by county. We estimated the difference-in-difference for each group (for example, male versus female) in 2015 versus 2021 to evaluate whether incidence rates changed over time for each group. Given that our data source contained only data from fee-for-service Medicare beneficiaries, we did an exploratory sensitivity analysis that extrapolates our results to the entire Medicare population by drawing from a previously published report comparing incidence and prevalence of dementia among fee-for-service Medicare beneficiaries and Medicare Advantage beneficiaries.30
Missing data
Beneficiaries missing information needed to derive the area deprivation index were excluded from analysis of neighborhood socioeconomic deprivation. Missing rates were low (228 701 beneficiaries or <5% with a diagnosis of dementia were missing Area Deprivation Index), so we used complete case analysis for those analyses.
Patient and public involvement
Patients and the public were not directly involved in the preparation of this manuscript, principally because of restrictions to accessing Medicare data. However, interactions with patients informed the motivation for this manuscript, as we observed many patients in clinical practice who would not be eligible or willing to participate in traditional cohort studies, which might lead those studies to underestimate the incidence or prevalence of dementia. We asked a member of a public advisory board to read this manuscript after submission.
Results
Table 1 shows baseline characteristics of beneficiaries with incident dementia. Overall, from the years 2015-21, 5 025 039 incident cases were recorded. Characteristics of beneficiaries by neighborhood socioeconomic status are shown in supplementary table B, which highlights high rates of medical and psychiatric comorbidities among beneficiaries with dementia.
Table 1.
Baseline characteristics of beneficiaries with incident dementia, 2015-21. Values are numbers (percentages) unless stated otherwise
| Characteristics | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 |
|---|---|---|---|---|---|---|---|
| No of incident cases | 838 824 | 796 768 | 771 082 | 724 564 | 675 130 | 589 769 | 628 902 |
| No of person years | 23 773 447 | 24 135 704 | 24 314 335 | 24 340 938 | 24 282 440 | 24 160 522 | 23 385 973 |
| Percentage (SE) age and sex standardized incidence rate* | 3.5 (0.004) | 3.3 (0.004) | 3.2 (0.004) | 3.0 (0.004) | 2.8 (0.003) | 2.5 (0.003) | 2.8 (0.003) |
| Mortality | 141 456 (16.9) | 121 333 (15.2) | 121 771 (15.8) | 113 806 (15.7) | 107 187 (15.9) | 114 379 (19.4) | 114 042 (18.1) |
| Demographics | |||||||
| Mean (SD) age, years | 81.32 (8.08) | 79.85 (8.04) | 80.00 (8.10) | 80.01 (8.11) | 80.08 (8.15) | 80.12 (8.16) | 80.05 (8.10) |
| Female sex | 498 585 (59.4) | 452 683 (56.8) | 437 842 (56.8) | 409 316 (56.5) | 381 884 (56.6) | 331 549 (56.2) | 353 803 (56.3) |
| Race/ethnicity: | |||||||
| Asian | 17 496 (2.1) | 15 629 (2.0) | 16 538 (2.1) | 16 420 (2.3) | 15 212 (2.3) | 12 951 (2.2) | 15 198 (2.4) |
| Black | 68 371 (8.2) | 60 941 (7.6) | 58 462 (7.6) | 54 572 (7.5) | 50 429 (7.5) | 43 998 (7.5) | 43 058 (6.8) |
| Hispanic | 40 593 (4.8) | 36 727 (4.6) | 36 207 (4.7) | 33 869 (4.7) | 31 172 (4.6) | 26 879 (4.6) | 28 880 (4.6) |
| Non-Hispanic white | 702 036 (83.7) | 672 150 (84.4) | 647 489 (84.0) | 607 237 (83.8) | 565 375 (83.7) | 493 844 (83.7) | 527 546 (83.9) |
| North American native | 3536 (0.4) | 3280 (0.4) | 3366 (0.4) | 3227 (0.4) | 2946 (0.4) | 2595 (0.4) | 2,745 (0.4) |
| Other | 4775 (0.6) | 4967 (0.6) | 5197 (0.7) | 4827 (0.7) | 4757 (0.7) | 4319 (0.7) | 4786 (0.8) |
| Unknown | 2017 (0.2) | 3074 (0.4) | 3823 (0.5) | 4412 (0.6) | 5239 (0.8) | 5183 (0.9) | 6689 (1.1) |
| Dual Medicaid eligibility | 144 434 (17.2) | 120 619 (15.1) | 117 576 (15.2) | 107 689 (14.9) | 97 164 (14.4) | 82 411 (14.0) | 83 565 (13.3) |
| Region | |||||||
| Northeast | 161 565 (19.3) | 149 795 (18.8) | 145 396 (18.9) | 141 006 (19.5) | 134 900 (20.0) | 115 510 (19.6) | 126 739 (20.2) |
| South | 349 931 (41.7) | 335 974 (42.2) | 319 332 (41.4) | 298 153 (41.1) | 281 467 (41.7) | 243 574 (41.3) | 256 458 (40.8) |
| Midwest | 188 263 (22.4) | 175 850 (22.1) | 173 410 (22.5) | 161 611 (22.3) | 147 950 (21.9) | 130 896 (22.2) | 136 756 (21.7) |
| West | 137 001 (16.3) | 133 332 (16.7) | 131 525 (17.1) | 122 472 (16.9) | 109 624 (16.2) | 98 918 (16.8) | 108 034 (17.2) |
| Puerto Rico | 2064 (0.2) | 1817 (0.2) | 1419 (0.2) | 1322 (0.2) | 1189 (0.2) | 871 (0.1) | 915 (0.1) |
SD=standard deviation; SE=standard error.
Incidence rate is presented as percentage per person year at risk.
Table 1 and table 2 show incidence rates and prevalence of dementia across the years 2015-21. Overall, incident cases decreased from 838 824 in 2015 to 628 902 in 2021, with a corresponding decrease in the age and sex adjusted incidence rate from 3.5% to 2.8%. During this same interval, prevalent cases increased from 2 772 188 in 2015 to 2 875 718 in 2021, reaching a maximum of 3 140 758 in 2019. This corresponded to an increase in age and sex adjusted prevalence from 10.5% in 2015 to 11.8% in 2021. Mortality increased in 2020 and 2021 both among beneficiaries with incident diagnoses in those years (19.4% and 18.1% versus 15.9% in 2019) and to a lesser extent among beneficiaries with prevalent dementia (20.0% and 17.8% versus 17.0% in 2019).
Table 2.
Characteristics of beneficiaries with prevalent dementia, 2015-21. Values are numbers (percentages) unless stated otherwise
| Variable | 2015 | 2016 | 2017 | 2018 | 2019 | 2020 | 2021 |
|---|---|---|---|---|---|---|---|
| No of prevalent cases | 2 772 188 | 2 962 259 | 3 061 866 | 3 113 375 | 3 140 758 | 2 956 102 | 2 875 718 |
| No of beneficiaries | 25 842 818 | 26 342 237 | 26 358 788 | 26 301 112 | 26 317 524 | 25 791 085 | 24 910 159 |
| Percentage (SE) age and sex standardized prevalence rate | 10.5 (0.06) | 11.1 (0.06) | 11.6 (0.07) | 11.9 (0.07) | 12.1 (0.07) | 11.7 (0.07) | 11.8 (0.07) |
| Mortality* | NA | 537 676 (19.4) | 551 315 (18.6) | 543 871 (17.8) | 529 511 (17.0) | 627 814 (20.0) | 524 780 (17.8) |
| Demographics | |||||||
| Mean (SD) age, years | 82.5 (8.1) | 82.0 (8.2) | 81.6 (8.3) | 81.4 (8.3) | 81.3 (8.2) | 81.1 (8.1) | 81.1 (8.1) |
| Female sex | 1 777 930 (64.1) | 1 865 277 (63.0) | 1 900 474 (62.1) | 1 910 789 (61.4) | 1 908 958 (60.8) | 1 788 695 (60.5) | 1 730 509 (60.2) |
| Race/ethnicity: | |||||||
| Asian | 65 353 (2.4) | 73 574 (2.5) | 79 800 (2.6) | 84 450 (2.7) | 87 171 (2.8) | 84 014 (2.8) | 84 412 (2.9) |
| Black | 260 140 (9.4) | 272 922 (9.2) | 275 796 (9.0) | 271 872 (8.7) | 267 499 (8.5) | 241 533 (8.2) | 222 105 (7.7) |
| Hispanic | 163 972 (5.9) | 178 735 (6.0) | 185 646 (6.1) | 186 914 (6.0) | 187 041 (6.0) | 173 139 (5.9) | 165 186 (5.7) |
| Non-Hispanic white | 2 247 428 (81.1) | 2 394 866 (80.8) | 2 471 740 (80.7) | 2 515 474 (80.8) | 2 538 470 (80.8) | 2 395 008 (81.0) | 2 338 300 (81.3) |
| North American native | 11 400 (0.4) | 12 183 (0.4) | 12 951 (0.4) | 13 457 (0.4) | 13 684 (0.4) | 12 694 (0.4) | 11 938 (0.4) |
| Other | 16 314 (0.6) | 18 957 (0.6) | 21 225 (0.7) | 22 736 (0.7) | 24 091 (0.8) | 23 677 (0.8) | 23 836 (0.8) |
| Unknown | 7581 (0.3) | 11 022 (0.4) | 14 708 (0.5) | 18 472 (0.6) | 22 802 (0.7) | 26 037 (0.9) | 29 941 (1.0) |
| Dual Medicaid eligibility | 712 035 (25.7) | 742 132 (25.1) | 746 349 (24.4) | 734 312 (23.6) | 714 363 (22.7) | 644 020 (21.8) | 603 165 (21.0) |
| Region | |||||||
| Northeast | 546 994 (19.7) | 572 907 (19.3) | 582 020 (19.0) | 591 446 (19.0) | 594 179 (18.9) | 560 016 (18.9) | 559 513 (19.5) |
| South | 1 161 875 (41.9) | 1 245 649 (42.1) | 1 285 163 (42.0) | 1 304 992 (41.9) | 1 318 051 (42.0) | 1 242 551 (42.0) | 1 193 078 (41.5) |
| Midwest | 606 840 (21.9) | 643 087 (21.7) | 661 890 (21.6) | 663 299 (21.3) | 670 362 (21.3) | 615 871 (20.8) | 591 812 (20.6) |
| West | 448 564 (16.2) | 492 179 (16.6) | 524 515 (17.1) | 545 512 (17.5) | 550 457 (17.5) | 530 741 (18.0) | 525 187 (18.3) |
| Puerto Rico | 7915 (0.3) | 8437 (0.3) | 8278 (0.3) | 8126 (0.3) | 7709 (0.2) | 6923 (0.2) | 6128 (0.2) |
NA=not applicable; SD=standard deviation; SE=standard error.
Beneficiaries who died in given year were prevalent in previous year. Percentage reflects proportion of deaths for each year, using previous year’s denominator.
In 2015 the age standardized incidence rate for male beneficiaries was higher than the age standardized incidence rate for female beneficiaries (3.5% (95% confidence interval (CI) 3.5% to 3.5%) versus 3.4% (3.4% to 3.4%); estimated difference 1.03 (95% CI 1.03 to 1.03); P<0.001) (supplementary table C. The difference widened in 2021, with an estimated difference in the male-to-female ratio of 1.09 (95% 1.09 to 1.10; P<0.001). The estimated difference-in-difference for male beneficiaries versus female beneficiaries in 2015 versus 2021 was 0.94 (95% CI 0.94 to 0.95; P<0.001).
In 2015 the age and sex standardized incidence rate was highest among black beneficiaries (4.2%, 95% CI 4.2% to 4.2%), followed by Hispanic beneficiaries (3.7%, 3.6% to 3.7%) and non-Hispanic white beneficiaries (3.4%, 3.4% to 3.4%), and lowest among beneficiaries of other race/ethnicity (3.1%, 3.0% to 3.1%). In 2021 black beneficiaries still had the highest incidence (3.1%, 95% CI 3.1% to 3.2%), but non-Hispanic white beneficiaries had the next highest (2.8%, 2.7% to 2.8%), followed by Hispanic beneficiaries (2.6%, 2.6% to 2.6%) and beneficiaries of other race/ethnicity (2.3%, 2.3% to 2.3%) (supplementary table D). From 2015 to 2021, differences in incidence narrowed between white beneficiaries and beneficiaries of all other groups (estimated difference-in-difference 0.92 (95% CI 0.91 to 0.93) compared with black beneficiaries, 0.88 (0.87 to 0.89) compared with Hispanic beneficiaries, and 0.93 (0.91 to 0.94) compared with beneficiaries of other race/ethnicities; all P<0.001).
Table 3 shows age and sex standardized incidence rates and prevalence of dementia for 2015 and 2021, stratified by tenth of Area Deprivation Index (1-10). For both 2015 and 2021, incidence rates and prevalence increased by tenth and are most pronounced for beneficiaries residing in the most deprived half of neighborhoods. These trends are depicted graphically in figure 1. Figure 2 shows age and sex standardized incidence and figure 3 shows age and sex standardized prevalence at the county level for 2015 and 2021.
Table 3.
Incidence and prevalence of dementia, 2015 and 2021, by tenth of Area Deprivation Index
| Year | Area Deprivation Index tenth | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1 (lowest deprivation) | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 (highest deprivation) | |
| Incidence rate (95% CI) | ||||||||||
| 2015 | 3.0 (3.0 to 3.0) | 3.2 (3.2 to 3.3) | 3.2 (3.2 to 3.3) | 3.3 (3.3 to 3.3) | 3.4 (3.4 to 3.4) | 3.5 (3.5 to 3.5) | 3.5 (3.5 to 3.6) | 3.7 (3.6 to 3.7) | 3.8 (3.8 to 3.8) | 4.1 (4.1 to 4.2) |
| 2021 | 2.4 (2.4 to 2.4) | 2.6 (2.6 to 2.6) | 2.6 (2.6 to 2.6) | 2.7 (2.6 to 2.7) | 2.7 (2.7 to 2.7) | 2.8 (2.8 to 2.8) | 2.8 (2.8 to 2.9) | 2.9 (2.9 to 2.9) | 3.0 (3.0 to 3.1) | 3.2 (3.2 to 3.3) |
| Prevalence (95% CI) | ||||||||||
| 2015 | 9.6 (9.5 to 9.6) | 10.2 (10.1 to 10.2) | 10.0 (10.0 to 10.0) | 10.0 (10.0 to 10.0) | 10.2 (10.1 to 10.2) | 10.3 (10.3 to 10.4) | 10.5 (10.5 to 10.6) | 10.8 (10.7 to 10.8) | 11.3 (11.3 to 11.4) | 12.4 (12.4 to 12.5) |
| 2021 | 10.9 (10.8 to 10.9) | 11.5 (11.4 to 11.5) | 11.3 (11.2 to 11.3) | 11.3 (11.2 to 11.3) | 11.4 (11.4 to 11.5) | 11.7 (11.7 to 11.8) | 12.1 (12.0 to 12.1) | 12.4 (12.3 to 12.4) | 13.1 (13.0 to 13.1) | 14.3 (14.2 to 14.3) |
CI=confidence interval.
Prevalence and incidence are both age and sex standardized and are reported as percentage (of beneficiaries for prevalence and of person years for incidence).
Fig 1.
Incident (top) and prevalent (bottom) rates of dementia, 2015 and 2021, per 1000 person years, by Area Deprivation Index (ADI) tenths, with 10 indicating greatest degrees of socioeconomic deprivation. Figure shows that dementia incidence and prevalence were both highest in socioeconomically deprived neighborhoods, and that this trend was generally consistent between 2015 and 2021
Fig 2.
Incident dementia rates in 2015 and 2021 by US county. Overall, incidence was generally concentrated in “stroke belt” regions of southern US in 2015, and this distribution shifted slightly northward into “rust belt” region of midwestern US in 2021, although incidence was still high in southern US in 2021.
Fig 3.
Prevalent dementia rates in 2015 and 2021 by county. Trend is similar to that seen for dementia incidence
Supplementary tables E and F report an exploratory sensitivity analysis extrapolating our results to the entire Medicare population (including Medicare advantage beneficiaries, who are not included in our current study), on the basis of a previous study that directly compared fee-for-service Medicare and Medicare Advantage beneficiaries.30 In summary, we estimate that there were likely between 370 705 and 386 645 incident cases and between 1 676 345 and 1 597 213 prevalent cases in 2021 in the Medicare Advantage population.
Discussion
This study provides contemporary information on more than five million incident cases of dementia among fee-for-service Medicare beneficiaries between 2015 and 2021. Although the overall incidence of dementia decreased over the study period, prevalence increased substantially, and our results indicate that nearly 2.9 million fee-for-service Medicare beneficiaries and nearly 4.5 million Medicare beneficiaries (including projections to the Medicare Advantage population) were likely living with a diagnosis of dementia in 2021. Our results suggest a higher prevalence than has been estimated in previous global studies, which may be attributable to greater capture of beneficiaries owing to our use of routinely collected data, misclassification due to our reliance on billing data, or differences between risk factors and diagnostic patterns in the US compared with other countries.31 32 33 34 Furthermore, the burden of dementia was unequally distributed, with the highest incidence and prevalence of dementia among black beneficiaries and those living in socioeconomically deprived neighborhoods. This study fills an important gap in the literature by describing the incidence and prevalence of dementia in a nationwide cohort of Medicare beneficiaries across six years and by sex, race, and neighborhood deprivation, characterizing the details of the incidence and prevalence of dementia in routine clinical practice at a granular level.
Strengths and weaknesses of study
Strengths of the study include its large, national sample size, its longitudinal evaluation of incidence and prevalence patterns, and its use of routinely collected data (medical claims data), which allow it to accurately reflect diagnostic patterns seen in general clinical practice. Weaknesses of the study include the following. Firstly, its reliance on Medicare claims data may reflect variations in diagnostic accuracy in routine clinical practice compared with gold standard diagnostic practices; for example, clinicians may code a dementia diagnosis for patients who do not meet the rigorous diagnostic criteria (inclusive of formal neuropsychological testing and other necessary elements) that would be applied in a research or memory clinic setting, as has been shown in previous analyses.35 36 37 38 Similarly, given our reliance on routinely collected data, patients might be misclassified as incident if they were prevalent in years preceding our data availability, although our reservation of two years (2013-14) of data outside of our primary study period decreases the likelihood that this has systematically biased our results. Secondly, Medicare claims data may not be generalizable to patients insured through other mechanisms. Importantly, beneficiaries with Medicare Advantage are not included in this study, and increased enrollment in Medicare Advantage (both owing to new enrollments and owing to beneficiary switching) occurred throughout the study period, which is likely not entirely random, as patients may elect into different plans for age or health related reasons, which could result in a changing composition of the fee-for-service Medicare population over time. Established demographic differences exist between beneficiaries enrolled in fee-for-service Medicare and Medicare Advantage.39 Thirdly, Medicare claims data are limited in their classification of race and ethnicity, although we used a validated algorithm to expand our ability to study dementia incidence by race and ethnicity. Fourthly, whether covid-19 related mortality differentially affected people with dementia (numerator) and the population at risk (denominator) owing to high mortality among older patients, especially those living in assisted facilities, is not known.40 41 Fifthly, dementia diagnoses made in routine clinical practice likely reflect a population distribution of timing of diagnosis, and this study was unable to examine important information about timing of diagnosis and subsequent longitudinal mortality rates. Sixthly, this study used the National Committee for Quality Assurance and Goodman codes for dementia ascertainment, which differ slightly from the Chronic Conditions Warehouse code list used in some other studies, which could yield different results, although most diagnosis codes overlap between the two lists. Finally, this study did not formally assess potential interactions between age, race/ethnicity, sex, neighborhood deprivation, and dementia incidence or prevalence, and future work to disentangle intersectionality in these dimensions is clearly needed.
Strengths and weaknesses in relation to other studies
This study adds to a growing body of evidence that suggests decreasing incidence of dementia and increasing prevalence in the US.9 10 42 The prevalence of dementia observed in this study of approximately 10.5% for 2015 differs from a previous three year prevalence estimate for the years 2011-13 of 14%,43 as well as a population estimate from a study that relied on clinical presentations of Alzheimer’s dementia.2 However, this is the first study to document nationwide trends in the incidence and prevalence of dementia in the US using diagnoses made in routine clinical practice. This is both a strength and a weakness of our approach; whereas carefully phenotyped cohorts provide very accurate information about dementia diagnoses at the highest standard of diagnostic rigor, this study provides information about routine practice patterns that are more broadly generalizable but likely less accurate. Our results contrast with a national interview based study (which relied on self-report of dementia and excluded participants resident in long term care facilities) that estimated a dementia prevalence of 4.0% in 2022.44 Our results also differ slightly from another large, population based study, which estimated a prevalence of dementia among fee-for-service Medicare beneficiaries of 9.0% in 2016 (compared with our estimate of 11.1%) and an incidence of 3.1% (compared with our estimate of 3.2%); these results may differ because of the study’s requirement for two separate dementia diagnoses, which is likely less sensitive but more specific than our study’s approach.30 Our study’s lack of Medicare Advantage data represents a major limitation, which we have attempted to correct through sensitivity analyses; however, changes in the demographics of the Medicare Advantage population over time could bias our results, especially those in subgroups with possible differential adoption of Medicare Advantage. However, our primary result (that is, that dementia incidence has decreased while dementia prevalence has increased) would be unlikely to be altered by our lack of inclusion of Medicare Advantage data, given that even in 2021 nearly 60% of all Medicare beneficiaries had fee-for-service Medicare and are included in our study. Additionally, although secular trends in Medicare Advantage adoption might affect the geospatial distribution of dementia incidence and prevalence observed in our study, our results are consistent with literature from chronic diseases sharing similar epidemiological features, such as cardiovascular disease, which suggest a fourfold to eightfold variation in incidence of cardiovascular disease across counties.11 45 46 47 Similarly, our results may not generalize to the small number (around 1%) of Americans who are not covered by Medicare at all.
Meaning of study: possible explanations and implications for clinicians and policy makers
The forces behind reduced incidence of dementia are likely multifactorial and may include overall improvements in the management of cardiovascular risk factors, including hypertension, hyperlipidemia, and diabetes.8 Additionally, our study suggests that increased mortality during the covid-19 pandemic could have contributed to apparent decreases in the incidence of dementia (that is, if patients who would otherwise have been given a diagnosis of dementia instead died of covid-19 or other complications). Similarly, increased mortality during the covid-19 pandemic could have masked greater increases in prevalence were the effect of the pandemic not present. Finally, the covid-19 pandemic could have decreased access to healthcare resources needed to obtain a diagnosis of dementia, decreasing apparent incidence or prevalence. The observed increase in the prevalence of dementia may be explained by several epidemiological and demographic characteristics of the disease and affected patients. Dementia is a largely irreversible, chronic pathology that typically follows a protracted course and for which older age is the most important risk factor.48 As the population ages, more people are surviving long enough to develop dementia; at the same time, those who develop the disease are also surviving longer. This may also contribute to the higher incidence and prevalence observed among women who, in the US, live an average 5.9 years longer than men.49 The study suggests that the gap in age standardized incidence between women and men widened significantly between 2015 and 2021, which contrasts with a narrowing of differences in dementia prevalence from 2000 to 2016, although data on incidence are not available.7 This greater reduction in incidence for women compared with men may be driven, in part, by reduced education inequalities and increased workforce participation among women, with unclear contributions from reductions in cardiovascular disease risk factors.50 51
Unanswered questions and future research
Future work to probe the mechanisms behind observed differences seen by sex, race/ethnicity, and neighborhood socioeconomic status could be conducted through analysis of longitudinal cohort studies with information on both cognitive outcomes and detailed life course history. Additionally, various factors, such as changes in educational attainment and risk factor burden, particularly the burden of cerebrovascular disease, which is strongly linked with dementia risk, may have implications for the epidemiology of dementia, in both the general population and specific subgroups, such as among beneficiaries from socioeconomically deprived neighborhoods, and deserve future study to delve into potential underlying mechanisms.52 53 54 Future studies evaluating healthcare utilization, particularly the use of psychotropic medications and acute medical care resources, would be highly informative to policy makers and clinicians, especially given concomitant changes in Medicaid dual eligibility owing to the Affordable Care Act. Furthermore, additional study to characterize more deeply the natural history of dementia, especially age at diagnosis and also including the onset of other psychiatric conditions, such as delirium, is needed to better understand optimal points of intervention to improve longitudinal cognitive functioning.
What is already known on this topic
More than 150 million people worldwide are predicted to develop dementia by 2050, and dementia is expected to become a leading driver of morbidity and healthcare expenditure
The burden of dementia may be higher in women, in patients from minoritized racial and ethnic groups, and among patients living in socioeconomically deprived areas
What this study adds
The incidence of dementia as diagnosed in routine clinical care in the US decreased between 2015 and 2021
Despite decreases in dementia incidence, prevalence continued to rise, with nearly 2.9 million traditional Medicare beneficiaries (around 12%) living with a dementia diagnosis in 2021
A greater burden of disease was observed in marginalized and low resource communities, highlighting the importance of policy approaches to promote equitable dementia care
Web extra.
Extra material supplied by authors
Web appendix: Supplementary tables
Contributors: BB was responsible for writing the initial draft. CBF and LZ drafted statistical analysis plans, completed the statistical analysis plan, and revised the manuscript for critical intellectual content. SS provided methodological supervision for statistical analysis and revised the manuscript for critical intellectual content. BGK, TDAR, SM, FL, BMG, YX, KGJ, and ROB revised the manuscript for critical intellectual content and provided methodological or clinical expertise to study design. ECOB and BGH conceived the study, provided supervision, procured funding, and revised the manuscript for critical intellectual content. JBL conceived the study, provided supervision, procured funding, devised the initial statistical analysis plan, and revised the manuscript for critical intellectual content. JBL is the guarantor. The corresponding author attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted
Funding: This work was internally funded by the Duke University Department of Neurology as well as by the Alzheimer’s Association, grant 24HPE-1287087, and JBL was funded by National Institutes of Health P30AG072958. The funders had no role in the collection, analysis, or interpretation of data, in the writing of the report, or in the decision to submit the manuscript for publication. The researchers had full access to the data in the study and take responsibility for the integrity of the data and data analysis.
Competing interests All authors have completed the ICMJE uniform disclosure form at www.icmje.org/disclosure-of-interest/ and declare: support from the Duke Department of Neurology, Duke University Health System, the Alzheimer’s Association, and the National Institutes of Health for the submitted work; TDAR has received fellowship funding from the Parkinson’s Foundation, the Huntington’s Disease Society of America Center of Excellence, the Tylers Hope Center for Excellence, the Randy M Schilsky Fellowship Fund, Medtronic, and Boston Scientific; SM has received grant funding from the Michael J Fox Foundation, the Parkinson’s Foundation, and the Josiah Macy Jr Foundation and honorariums, has received travel reimbursement for talks given to the Parkinson Study Group, the Parkinson’s Foundation, the American Academy of Neurology, the University of Virginia, and the University of Notre Dame, has served as site investigator for clinical trials sponsored by Takeda Pharmaceuticals, Biogen, and Inhibikase Therapeutics, and has been a consultant for Grey Matter Technologies (a wholly owned subsidiary of modality.ai) and Cerevel Therapeutics; BMG is supported by grant K23HL161426 from the National Heart, Lung, and Blood Institute, grant 23MRFSCD1077188 from the American Heart Association, and grant 2835124 from the Duke Office of Physician Scientist Development and through Bayer Pharmaceuticals as a site principal investigator; KGJ has acted as a research consultant for the University of Southern California, has received speaker fees from Eisai Inc, and has had primary investigator relationships with Eisai Inc, LEXEO Therapeutics, Athira Pharma, Annovis Bio, and the Critical Path Institute; JBL has received grant funding from Duke Bass Connections, the American Heart Association, and the National Institute on Ageing P30AG072958; no other relationships or activities that could appear to have influenced the submitted work.
Transparency: The lead author (the manuscript’s guarantor) affirms that the manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as originally planned (and, if relevant, registered) have been explained.
Dissemination to participants and related patient and public communities: We plan to disseminate this manuscript to the public through several means. Firstly, the study team will engage a community advisory board to share these results and discuss their implications for future research to improve dementia care and care giving, which is under way. The study team will work with Duke University, the University of North Carolina at Chapel Hill, and the Duke-UNC Alzheimer’s Disease Research Center to prepare press releases when the paper is published with the aim of spreading public awareness of the findings. The study team will endeavor to additionally partner with the Alzheimer’s Association, a funder of the manuscript, to maximize dissemination of the study findings. The corresponding author will make himself available to participate in interviews to help to interpret and communicate the study findings to the public. The team has prepared a lay summary that we will disseminate along with the manuscript, and JBL will prepare a video abstract summary targeted at the public to provide additional avenues to disseminate the manuscript to the public.
Provenance and peer review: Not commissioned; externally peer reviewed.
Publisher’s note: Published maps are provided without any warranty of any kind, either express or implied. BMJ remains neutral with regard to jurisdictional claims in published maps.
Ethics statements
Ethical approval
This manuscript was reported according to the RECORD statement, and a checklist was provided during submission of this manuscript. This work was approved by the Duke University Institutional Review Board, with review number Pro00106448.
Data availability statement
The conditions of the data use agreement with the Centers for Medicare and Medicaid services prohibit direct data sharing; however, investigators may apply directly to the Centers for Medicare and Medicaid services to receive the data used for the study.
References
- 1. GBD 2019 Collaborators . Global mortality from dementia: Application of a new method and results from the Global Burden of Disease Study 2019. Alzheimers Dement (N Y) 2021;7:e12200. 10.1002/trc2.12200 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population estimate of people with clinical Alzheimer’s disease and mild cognitive impairment in the United States (2020-2060). Alzheimers Dement 2021;17:1966-75. 10.1002/alz.12362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Babulal GM, Quiroz YT, Albensi BC, et al. International Society to Advance Alzheimer’s Research and Treatment, Alzheimer’s Association . Perspectives on ethnic and racial disparities in Alzheimer’s disease and related dementias: Update and areas of immediate need. Alzheimers Dement 2019;15:292-312. 10.1016/j.jalz.2018.09.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Weiss J, Beydoun MA, Beydoun HA, et al. Pathways explaining racial/ethnic disparities in incident all-cause dementia among middle-aged US adults. Alzheimers Dement 2023;19:4299-310. 10.1002/alz.12976 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Vassilaki M, Aakre JA, Castillo A, et al. Association of neighborhood socioeconomic disadvantage and cognitive impairment. Alzheimers Dement 2023;19:761-70. 10.1002/alz.12702 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Farina MP, Zhang YS, Kim JK, Hayward MD, Crimmins EM. Trends in Dementia Prevalence, Incidence, and Mortality in the United States (2000-2016). J Aging Health 2022;34:100-8. 10.1177/08982643211029716 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hudomiet P, Hurd MD, Rohwedder S. Trends in inequalities in the prevalence of dementia in the United States. Proc Natl Acad Sci U S A 2022;119:e2212205119. 10.1073/pnas.2212205119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Langa KM, Larson EB, Crimmins EM, et al. A Comparison of the Prevalence of Dementia in the United States in 2000 and 2012. JAMA Intern Med 2017;177:51-8. 10.1001/jamainternmed.2016.6807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Satizabal CL, Beiser AS, Chouraki V, Chêne G, Dufouil C, Seshadri S. Incidence of Dementia over Three Decades in the Framingham Heart Study. N Engl J Med 2016;374:523-32. 10.1056/NEJMoa1504327 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sullivan KJ, Dodge HH, Hughes TF, et al. Declining Incident Dementia Rates Across Four Population-Based Birth Cohorts. J Gerontol A Biol Sci Med Sci 2019;74:1439-45. 10.1093/gerona/gly236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Wolters FJ, Chibnik LB, Waziry R, et al. Twenty-seven-year time trends in dementia incidence in Europe and the United States: The Alzheimer Cohorts Consortium. Neurology 2020;95:e519-31. 10.1212/WNL.0000000000010022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Avan A, Hachinski V. Global, regional, and national trends of dementia incidence and risk factors, 1990-2019: A Global Burden of Disease study. Alzheimers Dement 2023;19:1281-91. 10.1002/alz.12764 [DOI] [PubMed] [Google Scholar]
- 13.Lepore M, Shuman SB, Wiener JM, Gould E. Challenges in Involving People with Dementia as Study Participants in Research on Care and Services. 2017. http://aspe.hhs.gov/reports/challenges-involving-people-dementia-study-participants-research-care-services-0.
- 14. Leggins B, Hart DM, Jackson AJ, et al. Perceptions about dementia clinical trials among underrepresented populations: a nationally representative survey of U.S. dementia caregivers. Alzheimers Res Ther 2024;16:224. 10.1186/s13195-024-01579-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Lam HR, Chow S, Taylor K, et al. Challenges of conducting research in long-term care facilities: a systematic review. BMC Geriatr 2018;18:242. 10.1186/s12877-018-0934-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Savold J, Cole M, Thorpe RJ, Jr. Barriers and solutions to Alzheimer’s disease clinical trial participation for Black Americans. Alzheimers Dement (N Y) 2023;9:e12402. 10.1002/trc2.12402 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Evangelista ND, Renteria MA, Mobley TM, et al. Selection Factors Influencing Participation in National Alzheimer’s Coordinating Center (NACC) Studies Compared to the Nationally Representative Health and Retirement Study (HRS). Alzheimers Dement 2022;18:e065949 10.1002/alz.065949 . [DOI] [Google Scholar]
- 18. Arce Rentería M, Mobley TM, Evangelista ND, et al. Representativeness of samples enrolled in Alzheimer’s disease research centers. Alzheimers Dement (Amst) 2023;15:e12450. 10.1002/dad2.12450 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Tom SE. Understanding Diversity in Life Course Social Determinants of Dementia Risk and Resilience in the Asian American Older Adult Population. JAMA Netw Open 2023;6:e231668. 10.1001/jamanetworkopen.2023.1668 [DOI] [PubMed] [Google Scholar]
- 20. Marquez DX, Perez A, Johnson JK, et al. Increasing engagement of Hispanics/Latinos in clinical trials on Alzheimer’s disease and related dementias. Alzheimers Dement (N Y) 2022;8:e12331. 10.1002/trc2.12331 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Malla G, Bodduluri S, Sthanam V, Sharma G, Bhatt SP. Access to Pulmonary Rehabilitation among Medicare Beneficiaries with Chronic Obstructive Pulmonary Disease. Ann Am Thorac Soc 2023;20:516-22. 10.1513/AnnalsATS.202204-318OC [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Fang MC, Coca Perraillon M, Ghosh K, Cutler DM, Rosen AB. Trends in stroke rates, risk, and outcomes in the United States, 1988 to 2008. Am J Med 2014;127:608-15. 10.1016/j.amjmed.2014.03.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Stewart KA, Blue L, Kranker K, et al. Hospital Use for Myocardial Infarction and Stroke Among Medicare Beneficiaries From March to December 2020. JAMA Cardiol 2021;6:1340-2. 10.1001/jamacardio.2021.2729 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lindstrom R, Keisler-Starkey K, Bunch L. Percentage of Older Adults With Both Private Health Insurance and Medicare Decreased From 2017 to 2022. 2024. https://www.census.gov/library/stories/2024/04/older-adults-health-coverage.html.
- 25. Kunicki ZJ, Bayer T, Jiang L, et al. Comparing Lookback Periods to Ascertain Alzheimer’s Disease and Related Dementias. Am J Alzheimers Dis Other Demen 2023;38:15333175231199566. 10.1177/15333175231199566 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Lusk JB, Ford C, Clark AG, et al. Racial/ethnic disparities in dementia incidence, outcomes, and health-care utilization. Alzheimers Dement 2023;19:2376-88. 10.1002/alz.12891 [DOI] [PubMed] [Google Scholar]
- 27. Eicheldinger C, Bonito A. More accurate racial and ethnic codes for Medicare administrative data. Health Care Financ Rev 2008;29:27-42. [PMC free article] [PubMed] [Google Scholar]
- 28. Singh GK. Area deprivation and widening inequalities in US mortality, 1969-1998. Am J Public Health 2003;93:1137-43. 10.2105/AJPH.93.7.1137 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Kind AJH, Buckingham WR. Making Neighborhood-Disadvantage Metrics Accessible - The Neighborhood Atlas. N Engl J Med 2018;378:2456-8. 10.1056/NEJMp1802313 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Haye S, Thunell J, Joyce G, et al. Estimates of diagnosed dementia prevalence and incidence among diverse beneficiaries in traditional Medicare and Medicare Advantage. Alzheimers Dement (Amst) 2023;15:e12472. 10.1002/dad2.12472 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Nichols E, Steinmetz JD, Vollset SE, et al. GBD 2019 Dementia Forecasting Collaborators . Estimation of the global prevalence of dementia in 2019 and forecasted prevalence in 2050: an analysis for the Global Burden of Disease Study 2019. Lancet Public Health 2022;7:e105-25. 10.1016/S2468-2667(21)00249-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Prince M, Bryce R, Albanese E, Wimo A, Ribeiro W, Ferri CP. The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement 2013;9:63-75.e2. 10.1016/j.jalz.2012.11.007 [DOI] [PubMed] [Google Scholar]
- 33. Arapakis K, Brunner E, French E, McCauley J. Dementia and disadvantage in the USA and England: population-based comparative study. BMJ Open 2021;11:e045186. 10.1136/bmjopen-2020-045186 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Liu Y, Gao X, Zhang Y, et al. Geographical variation in dementia prevalence across China: a geospatial analysis. Lancet Reg Health West Pac 2024;47:101117. 10.1016/j.lanwpc.2024.101117 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Creavin ST, Haworth J, Fish M, et al. Clinical judgment of GPs for the diagnosis of dementia: a diagnostic test accuracy study. BJGP Open 2021;5:BJGPO.2021.0058. 10.3399/BJGPO.2021.0058 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Chen Y, Tysinger B, Crimmins E, Zissimopoulos JM. 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. 10.1016/j.trci.2019.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Schliep KC, Ju S, Foster NL, et al. How good are medical and death records for identifying dementia? Alzheimers Dement 2022;18:1812-23. 10.1002/alz.12526 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Grodstein F, Chang CH, Capuano AW, et al. Identification of Dementia in Recent Medicare Claims Data, Compared With Rigorous Clinical Assessments. J Gerontol A Biol Sci Med Sci 2022;77:1272-8. 10.1093/gerona/glab377 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Murphy-Barron C, Pyenson B, Ferro C, Emery M. Comparing the Demographics of Enrollees in Medicare Advantage and Fee-For-Service Medicare. 2020. https://us.milliman.com/en/insight/comparing-the-demographics-of-enrollees-in-medicare-advantage-and-fee-for-service-medicare.
- 40.Tejada-Vera B, Kramarow E. COVID-19 Mortality in Adults Aged 65 and Over: United States, 2020. https://stacks.cdc.gov/view/cdc/121320. [PubMed]
- 41.Chidambaram P. Over 200 000 Residents and Staff in Long-Term Care Facilities Have Died From COVID-19. 2022. https://www.kff.org/policy-watch/over-200000-residents-and-staff-in-long-term-care-facilities-have-died-from-covid-19/.
- 42. 2023 Alzheimer’s disease facts and figures. Alzheimers Dement 2023;19:1598-695. 10.1002/alz.13016 [DOI] [PubMed] [Google Scholar]
- 43. Goodman RA, Lochner KA, Thambisetty M, Wingo TS, Posner SF, Ling SM. Prevalence of dementia subtypes in United States Medicare fee-for-service beneficiaries, 2011-2013. Alzheimers Dement 2017;13:28-37. 10.1016/j.jalz.2016.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Kramarow EA. Diagnosed Dementia in Adults Age 65 and Older: United States, 2022. 2024. https://stacks.cdc.gov/view/cdc/154766. [PubMed]
- 45. Spatz ES, Beckman AL, Wang Y, Desai N, Krumholz HM. Geographic Variation in Trends and Disparities in Acute Myocardial Infarction Hospitalization and Mortality by Income Levels, 1999-2013. JAMA Cardiol 2016;1:255-65. 10.1001/jamacardio.2016.0382 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Terry K, Makhlouf M, Altarabsheh SE, et al. Trends in Cardiovascular Disease Mortality by County-Level Social Vulnerability Index in the United States. J Am Heart Assoc 2023;12:e030290. 10.1161/JAHA.123.030290 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47. de Havenon A, Zhou LW, Johnston KC, et al. Twenty-Year Disparity Trends in United States Stroke Death Rate by Age, Race/Ethnicity, Geography, and Socioeconomic Status. Neurology 2023;101:e464-74. 10.1212/WNL.0000000000207446 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.National Institute on Aging. What Causes Alzheimer’s Disease? https://www.nia.nih.gov/health/alzheimers-causes-and-risk-factors/what-causes-alzheimers-disease.
- 49.Arias E, Tejada-Vera B, Kochanek KD, Ahmad FB. Provisional life expectancy estimates for 2021. 2022. https://stacks.cdc.gov/view/cdc/118999.
- 50. Tomioka K, Kurumatani N, Saeki K. Longitudinal association between lifetime workforce participation and risk of self-reported cognitive decline in community-dwelling older adults. PLoS One 2020;15:e0234392. 10.1371/journal.pone.0234392 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51. Sharp ES, Gatz M. Relationship between education and dementia: an updated systematic review. Alzheimer Dis Assoc Disord 2011;25:289-304. 10.1097/WAD.0b013e318211c83c [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52. Li XY, Kong XM, Yang CH, et al. Global, regional, and national burden of ischemic stroke, 1990-2021: an analysis of data from the global burden of disease study 2021. EClinicalMedicine 2024;75:102758. 10.1016/j.eclinm.2024.102758 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 53. Fan J, Li X, Yu X, et al. Global Burden, Risk Factor Analysis, and Prediction Study of Ischemic Stroke, 1990-2030. Neurology 2023;101:e137-50. 10.1212/WNL.0000000000207387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54. Kuźma E, Lourida I, Moore SF, Levine DA, Ukoumunne OC, Llewellyn DJ. Stroke and dementia risk: A systematic review and meta-analysis. Alzheimers Dement 2018;14:1416-26. 10.1016/j.jalz.2018.06.3061 [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
Web appendix: Supplementary tables
Data Availability Statement
The conditions of the data use agreement with the Centers for Medicare and Medicaid services prohibit direct data sharing; however, investigators may apply directly to the Centers for Medicare and Medicaid services to receive the data used for the study.