Abstract
Background:
The extension of Medicare telehealth waivers following the COVID-19 public health emergency presents an unprecedented opportunity to examine the longitudinal relationship between telehealth usage and cognitive functioning in later life. We analyzed transitions in telehealth use and investigated whether associations between telehealth use and cognitive functioning varies across different levels of cognitive performance.
Methods:
Using a longitudinal cohort design with data from the National Health and Aging Trends Study (NHATS; 2021–2023), we examined patterns of telehealth access and use among Medicare beneficiaries aged 70+ (N=3,116). Cognitive functioning was assessed using a comprehensive battery measuring memory, orientation, and executive function. We categorized telehealth access/use into four categories: no access to internet-enabled devices, basic use (email/texting only), broader internet use without telehealth, and telehealth use. Using unconditional quantile regression with fixed effects, we examined associations between telehealth access/use and cognitive functioning across the cognitive performance distribution.
Results:
While 64% of telehealth users maintained their usage over consecutive years, 44% of those without device access remained without access. Compared to basic internet use, lack of access to internet-enabled devices was associated with cognitive decline (b=-0.64, p<.05), whereas telehealth use was associated with better cognitive functioning (b=0.57, p<.01). The association between telehealth use and cognitive functioning was only significant at the lower end of the cognitive distribution (10th percentile: b=1.43, p<.01), with no significant associations observed at higher levels.
Conclusions:
The stable nature of digital access and usage over time—with most older adults either maintaining their initial level or showing limited upward mobility in digital engagement—underscore the importance of early intervention to promote internet access and skills. Telehealth use was linked to cognitive benefits only at the lower end of the distribution, suggesting the importance of targeted interventions to promote telehealth access among the older population at risk of cognitive impairment.
Keywords: Telehealth, National Health and Aging Trends Study, Cognition, Digital Divide
Introduction
The rapid acceleration of telehealth adoption during the COVID-19 pandemic has transformed healthcare delivery for older adults, with telehealth remaining integral to the healthcare landscape beyond the public health emergency. The extension of Medicare telehealth flexibilities through 2024 has removed existing restrictions on originating sites and expanded the scope of providers who can deliver telehealth services, including qualified occupational therapists, physical therapists, and mental health counselors1. While overall telehealth use among adults decreased since the peak of the pandemic2, healthcare organizations demonstrated sustained commitment to digital transformation, with the percentage of hospitals offering at least one form of telemedicine services increasing from 46% in 2017 to 72% in 20213. Despite this widespread institutional adoption of telehealth, significant questions remain about how older adults with varying levels of cognitive functioning navigate and benefit from digital health services. Understanding this relationship has important public health implications as we seek to optimize telehealth delivery while ensuring accessible and effective healthcare for older adults across the cognitive spectrum.
Several critical knowledge gaps exist regarding telehealth use and cognitive functioning in older adults. First, because Medicare reimbursement for telehealth services was previously restricted to clinics, hospitals, and other specific health facilities in rural regions4, our understanding of how the general population of older adults maintain or modify their telehealth engagement over time remains limited. Second, while research has documented cognitive benefits associated with general internet use in later life5,6, we know little about how healthcare-related digital engagement influences cognitive functioning7,8. Third, to our knowledge, no studies have examined how cognitive benefits associated with telehealth usage may vary across different levels of cognitive functioning —a crucial consideration for developing targeted telehealth-based interventions that account for varying cognitive capabilities.
This study addresses these knowledge gaps using longitudinal data from the National Health and Aging Trends Study (2021–2023). First, we examine patterns of telehealth access and usage transitions among Medicare beneficiaries aged 70 and older in the years following the COVID-19 pandemic. We then investigate the associations between different patterns of digital engagement—ranging from no access to active telehealth use—and cognitive functioning, using quantile regression to examine whether these associations vary across the cognitive functioning distribution.
Methods
Data.
This study employed data from the National Health and Aging Trends Study (NHATS), a nationally representative longitudinal study of Medicare beneficiaries. Conducted annually since 2011, NHATS collects information on sociodemographic characteristics, cognition, functional status, and health behaviors among Medicare beneficiaries through in-home interviews. Our sample included 3,116 respondents aged 70 and older in 2021, after excluding proxies (n=78) and those with missing data (n=272).
Dependent Variable: Cognitive Function.
Cognitive function was assessed at baseline and during each annual follow-up. NHATS employs a comprehensive cognitive battery assessing memory (immediate and delayed 10-word recall; range=0–20), orientation (date, month, year, day of week, naming president and vice president; range=0–8), and executive function (clock drawing test; range=0–5). We used a composite score (total range 0–33), with higher scores indicating better function. Additional details are available (Freedman et al. 2024).
Independent Variable: Telehealth Access and Use.
We assessed telehealth access and use through a hierarchical classification. First, access to internet-enabled device was asked with four questions: (a) Do you have a working cell phone, (b) Do you have a working computer in your home: (c) In the last month, have you used a computer anywhere else; and (d) Do you have a tablet computer like iPad that works by touching the screen. If respondents answered “no” to all four questions, they were classified as having “no access” to internet-enabled devices (n=166).
Next, if respondents had access to internet-enabled device, they were asked about the extent of their use. Respondents who reported sending messages by email or texting (i.e., In the last month, have you ever sent messages by email or texting?) but reported no other use (In the last month, have you ever gone on the internet or online for any other reason?) were classified as demonstrating “basic internet use” (n=1,220).
Respondents who reported using the internet beyond email/texting were further classified based on self-reported telehealth use in the past year (e.g., “In the last year, have you gone on the internet or online to have a visit with any of your medical providers? This is sometimes called a telehealth visit”). “Telehealth non-use” (n=1,034) was assigned to those who reported internet use beyond email/texting but no telehealth visits, while “telehealth use” (n=696) was assigned to those who reported having online medical provider visits.
Covariates
Our models included time-varying covariates to account for potential confounders for the association between telehealth access/usage and cognitive functioning. Sociodemographic characteristics included living alone, residence in a metropolitan area, and Medicaid enrollment. Health conditions included comorbidities diagnosed by a physician, functional impairment status (activities of daily living limitations), self-rated health, and depressive symptoms (Patient Health Questionnaire-4).
Analytic Strategy
We began by analyzing the pattern of change in telehealth access and usage and estimated transition probabilities across consecutive observations during the study period. We then employed unconditional quantile regression (UQR) models with fixed effects (FE) to investigate the links between telehealth and cognitive functioning across the population cognitive functioning distribution.9 While general linear models, including FE models, estimate the effect of a predictor on the outcome at the mean (i.e., the average effect) conditional on other covariates, UQR allows for examining potential heterogeneous effects across the unadjusted outcome distribution. To this end cognitive functioning, we conducted analyses at 10th, 25th, 50th, 75th, and 90th percentiles, capturing effects at lower (10th and 25th percentiles), median (50th percentile), and higher (75th and 90th percentiles) levels of cognitive functioning. We also estimated the average effect using a FE model to provide a reference point.
Results
Sample characteristics varied across telehealth access/use classifications at baseline (Table 1). Telehealth users were younger (mean age=79.49), more likely to be married (58.9%), college-educated (52.6%), and living in metropolitan areas (89.1%) compared to those with no access/use. Overall, respondents on average were 82 years old, with the majority being female (58%), Non-Hispanic White (73%), metropolitan residents (81%), and classified as having no dementia (83%; Supplementary Table 1). The mean cognitive functioning score during the study period was 18.9, with a slightly left-skewed distribution (Supplementary Figure 1). The 10th, 25th, 50th, 75th, and 90th percentiles of cognitive functioning were 11.9, 16.3, 20.2, 23.3, and 25.7, respectively.
Table 1.
Characteristics of NHATS Respondents in 2021
| No Accessa (N=166) |
Basic Useb (N=1,220) |
Telehealth Non-Usec (N=1,034) |
Telehealth Used (N=696) |
|
|---|---|---|---|---|
| Variables | N (%) / Mean (SD) | N (%) / Mean (SD) | N (%) / Mean (SD) | N (%) / Mean (SD) |
| Female | 116 (69.9%) | 739 (60.6%) | 573 (55.4%) | 373 (53.6%) |
| Agee | 87.08 (6.82) | 83.09 (6.37) | 80.14 (5.63) | 79.49 (5.53) |
| Married | 30 (18.1%) | 426 (34.9%) | 549 (53.1%) | 410 (58.9%) |
| Live Alone | 97 (58.4%) | 502 (41.1%) | 338 (32.7%) | 197 (28.3%) |
| Race/Ethnicity | ||||
| Non-Hispanic White | 95 (57.2%) | 731 (59.9%) | 881 (85.2%) | 568 (81.6%) |
| Non-Hispanic Black | 54 (32.5%) | 360 (29.5%) | 107 (10.3%) | 99 (14.2%) |
| Non-Hispanic other | 4 (2.4%) | 32 (2.6%) | 25 (2.4%) | 12 (1.7%) |
| Hispanic | 13 (7.8%) | 97 (8.0%) | 21 (2.0%) | 18 (2.6%) |
| Education | ||||
| Less than high school | 65 (39.2%) | 333 (27.3%) | 221 (21.4%) | 107 (15.4%) |
| High school degree | 50 (30.1%) | 419 (34.3%) | 221 (21.4%) | 107 (15.4%) |
| Some college | 30 (18.1%) | 282 (23.1%) | 335 (32.4%) | 205 (29.5%) |
| College degree or more | 21 (12.7%) | 186 (15.2%) | 417 (40.3%) | 366 (52.6%) |
| Income (median, $)f | 20,981(58,838) | 27,703 (43,403) | 55,000(66,495) | 66,631 (89,529) |
| Medicaid enrollment | 51 (30.7%) | 288 (23.6%) | 61 (5.9%) | 38 (5.5%) |
| Metropolitan area | 132 (79.5%) | 957 (78.4%) | 822 (79.5%) | 620 (89.1%) |
| ADL difficultyg | 1.21 (1.50) | 0.823 (1.267) | 0.44 (0.90) | 0.570(1.06) |
| Health conditionsh | 3.11 (1.35) | 3.01 (1.34) | 2.57 (1.31) | 2.78 (1.28) |
| Dementiag | ||||
| No Dementia | 83 (50.0%) | 900 (73.8%) | 958 (92.6%) | 651 (93.5%) |
| Possible dementia | 23 (13.9%) | 156 (12.8%) | 46 (4.4%) | 21 (3.0%) |
| Probable dementia | 60 (36.1%) | 164 (13.4%) | 30 (2.9%) | 24 (3.4%) |
Note. Respondent N=3,116; ADL = Activities of Daily Living
No internet-enabled device or if any, does not know how to use.
Used internet-enabled devices to only messages by email or texting.
Used internet-enabled device for other reasons besides email or texting.
Used internet-enabled device to have a visit with a medical provider.
Ranges from 71 to 103.
Imputed household income in dollars before taxes in 2020.
Number of difficulties with activities of daily living (e.g., eating, bathing, toileting, dressing, and getting out of bed).
Number of physician-diagnosed health conditions (heart attack, heart disease, hypertension, arthritis, osteoporosis, diabetes, lung disease, stroke, dementia, or cancer).
NHATS protocol divides the level of cognitive impairment classification into three levels using clinical diagnosis of Alzheimer’s disease and related dementia; the AD8 score; and performance-based scoring of memory, orientation, and executive functioning.
Figure 1 illustrates the probabilities of transitioning in telehealth access and usage over consecutive observations during the study period. Among respondents with no access to internet-enabled devices, 69% continued without access, 29% transitioned to basic use, 2% moved to telehealth non-use, but none transitioned to telehealth use. Of those with basic use, 81% remained in the same category, 6% lost access, 11% transitioned to a broader internet use, and 2% moved to telehealth use. Among individuals in the telehealth non-use group, 53% continued internet use without telehealth, <1% lost access, 13% reduced usage to basic use, 33% and transitioned to telehealth use. Finally, among telehealth users, 68% continued telehealth use, <1% lost access, 5% reduced usage to basic use, and 26% transitioned to non-telehealth internet use.
Figure 1. Telehealth Access and Usage Transitions Among Older Adults (2021–2023).
This figure shows the probabilities of transitions in telehealth access and usage among older adults over two years. Each arrow represents the observed percentage of individuals transitioning from an originating state to a destination. Respondent N=3,116.
The results from models examining the associations between telehealth access/use and cognitive functioning are presented in Table 2. In the FE model, having no access was associated with a lower cognitive function (b=-0.64, p<.05) compared to basic internet use, while both telehealth non-use (b=0.38, p<.05) and telehealth use (b=0.57, p<.01) were linked to better cognitive functioning relative to basic use. However, these average effects were not consistently observed across the cognitive functioning distribution, as the estimates from the UQR models indicated the associations were significant only at the 10th percentile of cognitive functioning (telehealth non-use: b=1.11, p<.05; telehealth use: b=1.43, p<.01). Telehealth access/use was unrelated to cognitive functioning at higher percentiles.
Table 2.
Estimates From Fixed Effects (FE) and Unconditional Quantile Regression With Fixed Effects (UQR-FE) Models Predicting Cognitive Function
| FE | UQR-FE | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Variable | Mean | 10th | 25th | 50th | 75th | 90th | ||||||
| Round 12 | −0.75*** | (0.07) | −1.09*** | (0.21) | −0.67*** | (0.15) | −0.59*** | (0.12) | −0.49*** | (0.13) | −0.79*** | (0.15) |
| Round 13 | −0.80*** | (0.08) | −1.71*** | (0.22) | −1.02*** | (0.16) | −0.54*** | (0.13) | −0.25 | (0.14) | −0.63*** | (0.17) |
| Telehealth use/accessa | ||||||||||||
| No Accessb | −0.64* | (0.31) | −2.21 | (1.63) | 0.09 | (0.60) | 0.12 | (0.33) | −0.25 | (0.27) | 0.01 | (0.30) |
| Telehealth non-usec | 0.38* | (0.16) | 1.11* | (0.50) | 0.54 | (0.39) | 0.13 | (0.28) | 0.42 | (0.27) | 0.03 | (0.28) |
| Telehealth used | 0.57** | (0.18) | 1.43** | (0.51) | 0.70 | (0.43) | 0.21 | (0.32) | 0.48 | (0.32) | 0.48 | (0.35) |
| Live alone | −0.16 | (0.22) | 0.28 | (0.63) | −0.15 | (0.44) | −0.29 | (0.35) | −0.32 | (0.37) | −0.21 | (0.44) |
| Medicaid | −0.15 | (0.21) | −1.63 | (0.99) | 0.05 | (0.53) | −0.04 | (0.31) | 0.15 | (0.22) | 0.14 | (0.24) |
| Metropolitan | 0.43 | (0.63) | 0.03 | (2.12) | −0.28 | (1.36) | 0.80 | (0.96) | 1.21 | (0.68) | 0.59 | (0.87) |
| Health conditionse | −0.19 | (0.10) | −0.03 | (0.31) | −0.47* | (0.22) | −0.17 | (0.16) | −0.23 | (0.18) | 0.09 | (0.21) |
| Any ADL difficulty | −0.20 | (0.11) | −0.33 | (0.36) | −0.25 | (0.24) | 0.09 | (0.19) | −0.16 | (0.19) | −0.01 | (0.22) |
| Any ADL help | −0.90*** | (0.18) | −3.32*** | (0.69) | −0.47 | (0.41) | −0.15 | (0.27) | −0.36 | (0.27) | −0.40 | (0.32) |
| Self-rated healthf | −0.03 | (0.06) | 0.18 | (0.24) | 0.04 | (0.15) | 0.02 | (0.10) | −0.02 | (0.11) | −0.14 | (0.12) |
| Depressive symptomsg | −0.09* | (0.04) | −0.08 | (0.14) | −0.16 | (0.08) | 0.00 | (0.05) | −0.09 | (0.05) | −0.01 | (0.06) |
Note. Based on 8,218 observations from 3,116 respondents. Robust standard errors presented in parentheses. Cognitive function was assessed using a comprehensive batter measuring memory, orientation, and executive function; higher scores indicate better function.
Reference category: Basic use: Used internet-enabled devices to only messages by email or texting.
No internet-enabled device or if any, does not know how to use.
Used internet-enabled device for other reasons besides email or texting.
Used internet-enabled device to have a visit with a medical provider.
Number of physician-diagnosed health conditions (heart attack, heart disease, hypertension, arthritis, osteoporosis, diabetes, lung disease, stroke, dementia, or cancer).
Ranges from 1(poor) to 5(excellent).
Patient Health Questionnaire for Depression (PHQ-4), higher score indicating more depressive symptoms.
p < .05.
p < .01.
p < .001.
Discussion
The integration of telehealth into routine care has become one of the lasting impacts of the COVID-19 public health emergency. While telemedicine use has declined from its unprecedented peak of 42% of Medicare outpatient visits in April-May 2020 to approximately 5% in 2024, both patients and physicians continue to value telehealth availability as a care option10. Notably, other forms of digital health technologies have shown sustained or increased usage, such as remote patient monitoring, wearables, and patient portal messaging system10,11. For older adults, this digital transformation offers opportunities to overcome mobility-related barriers to healthcare access, but it also poses challenges for those with limited digital literacy or cognitive impairment. As healthcare delivery becomes increasingly digitized, understanding how these technological advances affect older adults across the cognitive functioning spectrum has become increasingly critical.
In this context, this study uncovered several key findings regarding telehealth use and cognitive functioning among older adults in three years following the global pandemic. First, the relatively high stability in telehealth use (68% continuing use) among existing users indicates that once older adults adopt telehealth services, they tend to maintain their usage. At the same time, we observed notable “stickiness” in digital access and usage patterns, characterized by ordered transitions between levels of engagement. Most transitions between categories followed a stepwise pattern, with few individuals skipping intermediate stages of adoption (e.g., only 2% moved directly from basic use to telehealth use). This pattern aligns with the conceptualization of the digital divide as a multi-layered phenomenon, where disparities in access (the first layer), use (the second layer), and outcome (the third layer) compound over time12, underscoring the challenges of rapid digital adoption in this population. In this light, healthcare systems should consider implementing stepped-care approaches to telehealth adoption, with enhanced support and resources for those with low digital literacy or access.
Second, the fixed-effects model estimating the average effect indicated that broader internet engagement including telehealth use was associated with better cognitive functioning compared to basic internet use, while lack of access was associated with cognitive decline. This finding extends prior research documenting cognitive benefits of general internet use among older adults5,13, by demonstrating that more complex digital engagement beyond texting and emailing may offer additional cognitive benefits. However, quantile regression models provided a more nuanced understanding of the average effect, as the findings demonstrated that such cognitive benefits were experienced only by the older adults at the lower end of the cognitive functioning distribution. This concentrated effect aligns with emerging evidence that providing cognitive stimulation via structured digital activities may be both feasible and acceptable for older adults with mild cognitive impairment14,15. Moreover, the enhanced healthcare access via telehealth may be particularly valuable for those with lower cognitive functioning, who tend to encounter significant barriers to timely medical care16,17. With proper design and oversight, telehealth could provide meaningful benefits for older adults with lower cognitive functioning, particularly in maintaining cognitive performance and quality of life8, thereby helping to close the health disparity in the population.
We recognize important limitations of the study. First, our data lacks granular information about telehealth engagement, including service types, frequency of use, and caregiver assistance. The items on ‘access’ assess device availability rather than ownership, and some reported ‘access’ may primarily reflect technology access and use by others. As such, our findings cannot truly distinguish between individuals and support systems that enable technology access/use. Second, while our fixed-effects models account for time-invariant confounders, unobserved time-varying factors might influence both telehealth use and cognitive functioning. This is particularly relevant given rapid post-pandemic changes in healthcare delivery and telehealth policies that may have affected digital engagement and cognitive assessment opportunities. As such, we are not able to describe the study findings as causal estimates. Future research should incorporate claims-level or geospatial confounders such as healthcare infrastructure changes, broadband access variations, and insurance coverage shifts.
In summary, our findings reveal both challenges and opportunities in digital healthcare for older adults. We observed stability in telehealth access and usage over consecutive years, as well as ordered transitions. The study also demonstrated cognitive benefits of telehealth, which were concentrated among those with lower cognitive functioning. These findings underscore the need for proactive strategies to establish digital access and literacy, along with carefully designed telehealth interventions that can adapt to varying levels of cognitive functioning in later life.
Supplementary Material
Funding
This research was supported by the National Institute on Aging (R01AG062477) and by the MedStar Health Research Institute (Early Investigator Award).
References
- 1.Anonymous. MLN fact sheet. Telehealth services. Centers for Medicare & Medicaid Services; 2024. [Google Scholar]
- 2.Lucas J, Wang X. Declines in telemedicine use among adults: United States, 2021 and 2022. National Health Statistics Report. U.S. Department of Health and Human Services; 2024. [PubMed] [Google Scholar]
- 3.Jiang JX, Ross JS, Bai G. Unveiling the adoption and barriers of telemedicine in US hospitals: a comprehensive analysis (2017–2022). J Gen Intern Med 2024;39(13):2438–2445; doi: 10.1007/s11606-024-08853-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shaver J. The state of telehealth before and after the COVID-19 pandemic. Prim Care 2022;49(4):517–530; doi: 10.1016/j.pop.2022.04.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kim YK, Han SH. Internet use and cognitive functioning in later life: focus on asymmetric effects and contextual factors. Gerontologist 2021;62(3):425–435; doi: 10.1093/geront/gnab149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Firth J, Torous J, Stubbs B, et al. The “online brain”: how the internet may be changing our cognition. World Psychiatry 2019;18(2):119–129; doi: 10.1002/wps.20617. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Chumbler NR, Mann WC, Wu S, et al. The association of home-telehealth use and care coordination with improvement of functional and cognitive functioning in frail elderly men. Telemed J E Health 2004;10(2):129–137; doi: 10.1089/tmj.2004.10.129. [DOI] [PubMed] [Google Scholar]
- 8.Lai FH, Yan EW, Yu KK, et al. The protective impact of telemedicine on persons with dementia and their caregivers during the COVID-19 pandemic. Am J Geriatr Psychiatry 2020;28(11):1175–1184; doi: 10.1016/j.jagp.2020.07.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Rios-Avila F, Maroto ML. Moving beyond linear regression: implementing and interpreting quantile regression models with fixed effects. Sociol Methods Res 2024;53(2):639–682; doi: 10.1177/00491241211036165. [DOI] [Google Scholar]
- 10.Mehrotra A. Next steps in payment and regulatory policy for telehealth. Statement. Committee on Ways & Means: United States House of Representatives; 2024. [Google Scholar]
- 11.Ndayishimiye C, Lopes H, Middleton J. A systematic scoping review of digital health technologies during COVID-19: a new normal in primary health care delivery. Health Technol (Berl) 2023;13(2):273–284; doi: 10.1007/s12553-023-00725-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Van Deursen AJAM, Helsper E, Eynon R, et al. The compoundness and sequentiality of digital inequality. Int J Commun 2017;11:452–473. [Google Scholar]
- 13.Kamin ST, Lang FR. Internet use and cognitive functioning in late adulthood: longitudinal findings from the Survey of Health, Ageing and Retirement in Europe (SHARE). J Gerontol B Psychol Sci Soc Sci 2020;75(3):534–539; doi: 10.1093/geronb/gby123. [DOI] [PubMed] [Google Scholar]
- 14.Schepens Niemiec SL, Lee E, Saunders R, et al. Technology for activity participation in older people with mild cognitive impairment or dementia: expert perspectives and a scoping review. Disabil Rehabil Assist Technol 2023;18(8):1555–1576; doi: 10.1080/17483107.2022.2116114. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Doi T, Verghese J, Makizako H, et al. Effects of cognitive leisure activity on cognition in mild cognitive impairment: results of a randomized controlled trial. J Am Med Dir Assoc 2017;18(8):686–691; doi: 10.1016/j.jamda.2017.02.013. [DOI] [PubMed] [Google Scholar]
- 16.Mullins MA, Bynum JPW, Judd SE, et al. Access to primary care and cognitive impairment: results from a national community study of aging Americans. BMC Geriatr 2021;21(1):580; doi: 10.1186/s12877-021-02545-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lovett RM, Benavente JY, Opsasnick LA, et al. Associations between cognitive impairment severity and barriers to healthcare engagement among older adults. J Appl Gerontol 2023;42(7):1387–1396; doi: 10.1177/07334648231166289. [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.