Xi Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Xi Chen has a total of 303 co-authors. Here are the top co-authors by number of shared publications:
- Zhuoer Lin
Division of Health Policy and Administration, School of Public Health, University of Illinois at Chicago, Chicago, USA
12 shared publications Zhuoer Lin
Division of Health Policy and Administration, School of Public Health, University of Illinois Chicago, Chicago
9 shared publicationsThomas M. Gill
Department of Internal Medicine, Yale School of Medicine, New Haven, Connecticut
8 shared publications- Thomas M. Gill
Department of Internal Medicine, Yale School of Medicine , New Haven, Connecticut
8 shared publications - Xin Zhang
School of Statistics, Beijing Normal University, Beijing 100875, China
8 shared publications - Shanquan Chen
School of Public Health, LKS Faculty of Medicine, University of Hong Kong, Hong Kong, China
7 shared publications
Latest Publications
- Open Access
- 13 October 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Unmet care needs at the end-of-life among older adults with activity limitations
Despite the growing global demand for end-of-life care resulting from rapid social transformations, significant gaps exist in cross-national research examining unmet end-of-life care needs. Using harmonized end-of-life survey data (2010-2021) from the United States, 28 European countries, South Korea, China, and Mexico, here we show that the United States, Europe, and South Korea exhibit higher overall unmet end-of-life care needs than China and Mexico, where limited formal care provision is compensated with high levels of informal care. Death at institutions is associated with lower overall unmet end-of-life care needs in the United States and South Korea, with Mexico presenting an exception. Similarly, urban areas in Europe and South Korea exhibit lower unmet formal end-of-life care needs compared to rural areas. Among married individuals, only Mexico exhibits significantly lower unmet overall end-of-life care needs compared to unmarried individuals. Having children is consistently linked to lower unmet informal end-of-life care needs across the United States, Europe, and South Korea. Our findings highlight the complex interplay between care systems and family demographic structures in shaping end-of-life care access, and underscore the need for tailored policy approaches that strengthen both formal and informal care systems while addressing sociodemographic disparities in care access.
- 13 October 2026
- Published by Springer Nature in Nature Communications
- Vol. 17 (1)
Open AccessUnmet care needs at the end-of-life among older adults with activity limitations
Despite the growing global demand for end-of-life care resulting from rapid social transformations, significant gaps exist in cross-national research examining unmet end-of-life care needs. Using harmonized end-of-life survey data (2010-2021) from the United States, 28 European countries, South Korea, China, and Mexico, here we show that the United States, Europe, and South Korea exhibit higher overall unmet end-of-life care needs than China and Mexico, where limited formal care provision is compensated with high levels of informal care. Death at institutions is associated with lower overall unmet end-of-life care needs in the United States and South Korea, with Mexico presenting an exception. Similarly, urban areas in Europe and South Korea exhibit lower unmet formal end-of-life care needs compared to rural areas. Among married individuals, only Mexico exhibits significantly lower unmet overall end-of-life care needs compared to unmarried individuals. Having children is consistently linked to lower unmet informal end-of-life care needs across the United States, Europe, and South Korea. Our findings highlight the complex interplay between care systems and family demographic structures in shaping end-of-life care access, and underscore the need for tailored policy approaches that strengthen both formal and informal care systems while addressing sociodemographic disparities in care access.
- Open Access
- 20 August 2026
- Published by Elsevier in SSRN Electronic Journal
Fetal Pollution Exposure, Cognitive Ability, and Gender-specific Parental Investment
We examine how prenatal exposure to air pollution interacts with gender-biased parental investments to shape long-run cognitive performance. Using nationally representative survey data from China, we show that fetal PM2.5 exposure significantly reduces cognitive ability for women, particularly among those with brothers. Our evidence suggests that the family investment channel operates primarily through education rather than health, with gender-biased educational resource allocation compounding negative outcomes for females. Specifically, conditional on the same level of fetal PM2.5 exposure, females receive less homework assistance from their families and attain lower levels of education.Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.
- 20 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessFetal Pollution Exposure, Cognitive Ability, and Gender-specific Parental Investment
We examine how prenatal exposure to air pollution interacts with gender-biased parental investments to shape long-run cognitive performance. Using nationally representative survey data from China, we show that fetal PM2.5 exposure significantly reduces cognitive ability for women, particularly among those with brothers. Our evidence suggests that the family investment channel operates primarily through education rather than health, with gender-biased educational resource allocation compounding negative outcomes for females. Specifically, conditional on the same level of fetal PM2.5 exposure, females receive less homework assistance from their families and attain lower levels of education.Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.
- Open Access
- 28 June 2026
- Published by Springer Nature in Research Square
Patient-Facing AI Chatbots vs Primary Care Physicians: A Standardized Patient Field Experiment in China
Importance Primary health care (PHC) systems worldwide face persistent workforce shortages, access barriers, and quality gaps. As patients increasingly turn to low-cost easily accessible AI chatbots for health advice, real-world evidence is needed on how patient-facing AI performs compared with PHC physicians. Objective To compare patient-facing AI chatbots with PHC physicians in diagnostic accuracy, clinical safety, and patient-centered communication using standardized patients in real-world PHC settings. Design, Setting, and Participants This standardized patient field experiment was conducted from February 6 to February 27, 2025, in 62 PHC institutions in China. Standardized patients (SPs) presented the same cases of unstable angina and asthma to 248 PHC physicians and to 2 widely used AI chatbots, ChatGPT-4o and DeepSeek-R1, on the same day. The analysis included 744 SP encounters. Exposures Consultation with a PHC physician, ChatGPT-4o, or DeepSeek-R1 for the same conditions. Main Outcomes and Measures Primary outcomes included diagnostic accuracy (correct diagnosis and correct medication), clinical safety (number and rate of unnecessary tests and medication prescribing), and patient-centeredness (range 1–4, with higher scores indicating more patient-centered communications). Results Among 248 PHC physicians, 66.9% were male and the mean (SD) age was 44.7 (9.4) years. Compared with PHC physicians, ChatGPT-4o and DeepSeek-R1 had higher rates of correct diagnosis (94.0% and 99.2% vs 30.6%; P<.001), correct medication recommendation (100.0% and 100.0% vs 35.4%; P<.001), and higher patient-centeredness scores (3.28 and 3.35 vs 2.55; P<.001), particularly in understanding the whole person and finding the common ground . However, both chatbots ordered more tests and recommended more medications than physicians, resulting in higher rates of unnecessary testing (90.3% and 96.4% vs 33.1%; P<.001) and inappropriate medication recommendations (73.4% and 62.1% vs 23.4%; P<.001). Findings were robust in multivariable regression models and when local physician performance was benchmarked against a nationally representative sample of PHC providers in China. Conclusions and Relevance In this field experiment, patient-facing AI chatbots had higher diagnostic accuracy and patient-centered communication scores than PHC physicians, but were substantially more likely to recommend unnecessary tests and inappropriate medications. These findings suggest that patient-facing AI may improve access to diagnostic and communicative support in under-resourced primary care, while also creating a risk of persuasive, AI-induced low-value care. Governance of patient-facing AI should therefore focus not only on diagnostic accuracy, but also on preventing excessive recommendations that may shape patient expectations before clinical encounters.
- 28 June 2026
- Published by Springer Nature in Research Square
Open AccessPatient-Facing AI Chatbots vs Primary Care Physicians: A Standardized Patient Field Experiment in China
Importance Primary health care (PHC) systems worldwide face persistent workforce shortages, access barriers, and quality gaps. As patients increasingly turn to low-cost easily accessible AI chatbots for health advice, real-world evidence is needed on how patient-facing AI performs compared with PHC physicians. Objective To compare patient-facing AI chatbots with PHC physicians in diagnostic accuracy, clinical safety, and patient-centered communication using standardized patients in real-world PHC settings. Design, Setting, and Participants This standardized patient field experiment was conducted from February 6 to February 27, 2025, in 62 PHC institutions in China. Standardized patients (SPs) presented the same cases of unstable angina and asthma to 248 PHC physicians and to 2 widely used AI chatbots, ChatGPT-4o and DeepSeek-R1, on the same day. The analysis included 744 SP encounters. Exposures Consultation with a PHC physician, ChatGPT-4o, or DeepSeek-R1 for the same conditions. Main Outcomes and Measures Primary outcomes included diagnostic accuracy (correct diagnosis and correct medication), clinical safety (number and rate of unnecessary tests and medication prescribing), and patient-centeredness (range 1–4, with higher scores indicating more patient-centered communications). Results Among 248 PHC physicians, 66.9% were male and the mean (SD) age was 44.7 (9.4) years. Compared with PHC physicians, ChatGPT-4o and DeepSeek-R1 had higher rates of correct diagnosis (94.0% and 99.2% vs 30.6%; P<.001), correct medication recommendation (100.0% and 100.0% vs 35.4%; P<.001), and higher patient-centeredness scores (3.28 and 3.35 vs 2.55; P<.001), particularly in understanding the whole person and finding the common ground . However, both chatbots ordered more tests and recommended more medications than physicians, resulting in higher rates of unnecessary testing (90.3% and 96.4% vs 33.1%; P<.001) and inappropriate medication recommendations (73.4% and 62.1% vs 23.4%; P<.001). Findings were robust in multivariable regression models and when local physician performance was benchmarked against a nationally representative sample of PHC providers in China. Conclusions and Relevance In this field experiment, patient-facing AI chatbots had higher diagnostic accuracy and patient-centered communication scores than PHC physicians, but were substantially more likely to recommend unnecessary tests and inappropriate medications. These findings suggest that patient-facing AI may improve access to diagnostic and communicative support in under-resourced primary care, while also creating a risk of persuasive, AI-induced low-value care. Governance of patient-facing AI should therefore focus not only on diagnostic accuracy, but also on preventing excessive recommendations that may shape patient expectations before clinical encounters.
- Open Access
- 1 May 2026
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 41 (13) , 1-11
Associations of Individual- and Neighborhood-Level Factors with Racial and Ethnic Disparities in Timely Dementia Diagnosis
Background: Underdiagnosis of dementia disproportionately affects racial and ethnic minority groups. As disease-modifying treatments and advances in dementia care models emerge, it is critical to ensure equitable, timely detection of dementia among older adults. Objective: To examine racial and ethnic disparities in the timely diagnosis of dementia and determine what individual- and neighborhood-level factors explain most of these disparities. Design: Retrospective observational study using 1998–2020 Medicare-linked survey data. Participants: U.S. adults aged ≥ 65 classified as having dementia in the Health and Retirement Study (HRS), linked to Medicare claims. Main Measures: Timely diagnosis of dementia, defined as a clinical diagnosis of dementia recorded in Medicare claims within 3 years before or 1 year after the month of the HRS interview at which dementia was first observed, was compared with delayed or missed dementia diagnosis. Key Results: In this population-based study of 3, 429 U.S. older adults classified as having dementia in the HRS, non-Hispanic Black (odds ratio, 0.71 [95% CI, 0.57 to 0.89]) and Hispanic individuals (odds ratio, 0.61 [95% CI, 0.44 to 0.85]) were less likely than non-Hispanic White individuals to receive a timely diagnosis after adjusting for non-modifiable demographics and health conditions. Educational attainment explained the largest share of the disparities (48.4% for non-Hispanic Black individuals and 61.9% for Hispanic individuals). Neighborhood affluence accounted for 27.3% and 17.8%, respectively, followed by densities of health practitioner offices (15.8% and 14.6%), physician offices (9.6% and 12.4%), and dementia specialist evaluation (6.6% and 5.5%). Conclusions: A large proportion of racial and ethnic disparities in timely diagnosis among older adults with dementia were explained by educational attainment, followed by neighborhood affluence, the densities of health practitioner offices, physician offices, and dementia specialist evaluation. The findings point to key directions for targeted interventions to improve dementia detection and advance equity in dementia care.
- 1 May 2026
- Published by Springer Nature in Journal of General Internal Medicine
- Vol. 41 (13) , 1-11
Open AccessAssociations of Individual- and Neighborhood-Level Factors with Racial and Ethnic Disparities in Timely Dementia Diagnosis
Background: Underdiagnosis of dementia disproportionately affects racial and ethnic minority groups. As disease-modifying treatments and advances in dementia care models emerge, it is critical to ensure equitable, timely detection of dementia among older adults. Objective: To examine racial and ethnic disparities in the timely diagnosis of dementia and determine what individual- and neighborhood-level factors explain most of these disparities. Design: Retrospective observational study using 1998–2020 Medicare-linked survey data. Participants: U.S. adults aged ≥ 65 classified as having dementia in the Health and Retirement Study (HRS), linked to Medicare claims. Main Measures: Timely diagnosis of dementia, defined as a clinical diagnosis of dementia recorded in Medicare claims within 3 years before or 1 year after the month of the HRS interview at which dementia was first observed, was compared with delayed or missed dementia diagnosis. Key Results: In this population-based study of 3, 429 U.S. older adults classified as having dementia in the HRS, non-Hispanic Black (odds ratio, 0.71 [95% CI, 0.57 to 0.89]) and Hispanic individuals (odds ratio, 0.61 [95% CI, 0.44 to 0.85]) were less likely than non-Hispanic White individuals to receive a timely diagnosis after adjusting for non-modifiable demographics and health conditions. Educational attainment explained the largest share of the disparities (48.4% for non-Hispanic Black individuals and 61.9% for Hispanic individuals). Neighborhood affluence accounted for 27.3% and 17.8%, respectively, followed by densities of health practitioner offices (15.8% and 14.6%), physician offices (9.6% and 12.4%), and dementia specialist evaluation (6.6% and 5.5%). Conclusions: A large proportion of racial and ethnic disparities in timely diagnosis among older adults with dementia were explained by educational attainment, followed by neighborhood affluence, the densities of health practitioner offices, physician offices, and dementia specialist evaluation. The findings point to key directions for targeted interventions to improve dementia detection and advance equity in dementia care.
- 8 April 2026
- Published by Elsevier in American Journal of Preventive Medicine
- Vol. 71 (6) , 108374
Neighborhood Disorder and Risk of Dementia: The Mediating Role of Cardiometabolic Risk
- 8 April 2026
- Published by Elsevier in American Journal of Preventive Medicine
- Vol. 71 (6) , 108374
Neighborhood Disorder and Risk of Dementia: The Mediating Role of Cardiometabolic Risk
- 5 April 2026
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 74 (5) , 1395-1405
Association of Geriatric Emergency Department Care With Hospitalization and Mortality in Older Adults
Background: Since 2018, the Geriatric Emergency Department (GED) Accreditation Program has recognized Emergency Departments (EDs) that provide high‐quality care tailored to older adults. GEDs have expanded rapidly across the United States in recent years, but little is known about how GED care is associated with patient outcomes, including hospital admissions and subsequent mortality. Methods: We used the 2018–2021 Health and Retirement Study (HRS)‐Medicare linked data of adults aged ≥ 65 years. We supplemented these data with the American College of Emergency Physicians (ACEP) GED accreditation list and American Hospital Association (AHA) data. Receipt of acute care in a GED was defined as having an ED visit at a GED. Patient‐level analyses were conducted using each individual's most recent ED visit. Multivariable logistic regression models were used to estimate associations between receipt of acute care in a GED and outcomes of hospital admission and 30‐day mortality, adjusting for patient demographics, socioeconomic status, health conditions, ED visit severity, and hospital‐level characteristics. Results: Among 4563 older adults who had an ED visit, 270 (5.9%) received acute care in GEDs and 4293 (94.1%) received non‐GED care. Compared with those treated in non‐GEDs, patients treated in GEDs had significantly lower odds of hospital admission (OR, 0.61; 95% CI, 0.42–0.87; p < 0.01) and 30‐day mortality (OR, 0.62; 95% CI, 0.40–0.96; p < 0.05). Subgroup analyses showed that the association with admission was more pronounced among adults aged 65–80 years (OR, 0.43; 95% CI, 0.24–0.76; p < 0.01) and non‐Hispanic White individuals (OR, 0.51; 95% CI, 0.34–0.78). An association with lower mortality was observed among non‐Hispanic White individuals (OR, 0.51; 95% CI, 0.30–0.87; p < 0.05). Conclusions: GED care was associated with lower odds of hospital admissions and 30‐day mortality among older adults. Broader implementation may expand the reach of GED programs across diverse populations.
- 5 April 2026
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 74 (5) , 1395-1405
Association of Geriatric Emergency Department Care With Hospitalization and Mortality in Older Adults
Background: Since 2018, the Geriatric Emergency Department (GED) Accreditation Program has recognized Emergency Departments (EDs) that provide high‐quality care tailored to older adults. GEDs have expanded rapidly across the United States in recent years, but little is known about how GED care is associated with patient outcomes, including hospital admissions and subsequent mortality. Methods: We used the 2018–2021 Health and Retirement Study (HRS)‐Medicare linked data of adults aged ≥ 65 years. We supplemented these data with the American College of Emergency Physicians (ACEP) GED accreditation list and American Hospital Association (AHA) data. Receipt of acute care in a GED was defined as having an ED visit at a GED. Patient‐level analyses were conducted using each individual's most recent ED visit. Multivariable logistic regression models were used to estimate associations between receipt of acute care in a GED and outcomes of hospital admission and 30‐day mortality, adjusting for patient demographics, socioeconomic status, health conditions, ED visit severity, and hospital‐level characteristics. Results: Among 4563 older adults who had an ED visit, 270 (5.9%) received acute care in GEDs and 4293 (94.1%) received non‐GED care. Compared with those treated in non‐GEDs, patients treated in GEDs had significantly lower odds of hospital admission (OR, 0.61; 95% CI, 0.42–0.87; p < 0.01) and 30‐day mortality (OR, 0.62; 95% CI, 0.40–0.96; p < 0.05). Subgroup analyses showed that the association with admission was more pronounced among adults aged 65–80 years (OR, 0.43; 95% CI, 0.24–0.76; p < 0.01) and non‐Hispanic White individuals (OR, 0.51; 95% CI, 0.34–0.78). An association with lower mortality was observed among non‐Hispanic White individuals (OR, 0.51; 95% CI, 0.30–0.87; p < 0.05). Conclusions: GED care was associated with lower odds of hospital admissions and 30‐day mortality among older adults. Broader implementation may expand the reach of GED programs across diverse populations.
- Open Access
- 2 March 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (3) , e262427
Racial and Ethnic Reporting and Representation in US Alzheimer Clinical Trials
Importance: Alzheimer disease (AD) disproportionately affects racial and ethnic populations underrepresented in US clinical research, raising concerns about the generalizability of AD trial findings and the evaluation of treatment safety and efficacy for populations most affected by AD. Objective: To examine patterns and trends in the reporting and representation of patient race and ethnicity in US-based phase 3 AD clinical trials. Evidence Review: This systematic review examined US-based phase 3 AD drug trials identified through the Trialtrove trial database between 1997 and 2023. Trials were cross-referenced with peer-reviewed publications, ClinicalTrials.gov, pharmaceutical company reports, and conference abstracts. Completed trials were eligible for inclusion if they were designated as phase 3 drug trials targeting AD and recruited patients exclusively in the US. Primary outcomes included reporting of race and ethnicity, the number of racial and ethnic groups reported, and their representation among trial populations. Secondary outcomes included terminology used, reporting of safety or efficacy differences by race and ethnicity, and discussion of racial and ethnic representation in trial reports. Temporal trends in reporting and representation were assessed. Methodologic quality was evaluated using the Quality Rating Scheme for Studies and Other Evidence. Data collection was completed May 2024. Findings: Among 88 US-based phase 3 AD clinical trials conducted between 1997 and 2023, 71 (80.7%) had publicly available results, including 52 (59.1%) published in peer-reviewed journals. Nearly half of published trials (35 [49.3%]) did not report patient race or ethnicity. Among published trials, reporting was inconsistent and focused predominantly on White (36 [50.7%]) patients, with substantially fewer trials reporting data on Asian or Pacific Islander (11 [15.5%]), Black (20 [28.2%]), Hispanic (13 [18.3%]), or Native American (2 [2.8%]) patients. Median (IQR) enrollment of White patients was 91.3% (87.3%-93.6%), whereas enrollment of underrepresented patient populations was markedly lower, with median (IQR) enrollment of 0.9% (0.6%-1.6%) for Asian or Pacific Islander, 4.5% (3.6%-6.6%) for Black (ethnicity unspecified), 7.2% (3.7%-9.1%) for Black (non-Hispanic), 5.2% (3.1%-6.6%) for Hispanic, and 0.4% (0%-0.8%) for Native American patients. Few trials (3 of 71 [4.2%]) conducted subgroup analyses by race or ethnicity, and none reported detailed subgroup characteristics or safety or efficacy outcomes by patient race and ethnicity. Reporting practices and representation showed little improvement over time. Conclusions and Relevance: US-based phase 3 AD trials showed substantial gaps in racial and ethnic reporting and representation from 1997 to 2023, limiting the evaluation of treatment safety and efficacy across diverse populations. These findings suggest that stronger reporting standards and more inclusive trial design and recruitment strategies are needed to improve the equity and generalizability of AD trials.
- 2 March 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (3) , e262427
Open AccessRacial and Ethnic Reporting and Representation in US Alzheimer Clinical Trials
Importance: Alzheimer disease (AD) disproportionately affects racial and ethnic populations underrepresented in US clinical research, raising concerns about the generalizability of AD trial findings and the evaluation of treatment safety and efficacy for populations most affected by AD. Objective: To examine patterns and trends in the reporting and representation of patient race and ethnicity in US-based phase 3 AD clinical trials. Evidence Review: This systematic review examined US-based phase 3 AD drug trials identified through the Trialtrove trial database between 1997 and 2023. Trials were cross-referenced with peer-reviewed publications, ClinicalTrials.gov, pharmaceutical company reports, and conference abstracts. Completed trials were eligible for inclusion if they were designated as phase 3 drug trials targeting AD and recruited patients exclusively in the US. Primary outcomes included reporting of race and ethnicity, the number of racial and ethnic groups reported, and their representation among trial populations. Secondary outcomes included terminology used, reporting of safety or efficacy differences by race and ethnicity, and discussion of racial and ethnic representation in trial reports. Temporal trends in reporting and representation were assessed. Methodologic quality was evaluated using the Quality Rating Scheme for Studies and Other Evidence. Data collection was completed May 2024. Findings: Among 88 US-based phase 3 AD clinical trials conducted between 1997 and 2023, 71 (80.7%) had publicly available results, including 52 (59.1%) published in peer-reviewed journals. Nearly half of published trials (35 [49.3%]) did not report patient race or ethnicity. Among published trials, reporting was inconsistent and focused predominantly on White (36 [50.7%]) patients, with substantially fewer trials reporting data on Asian or Pacific Islander (11 [15.5%]), Black (20 [28.2%]), Hispanic (13 [18.3%]), or Native American (2 [2.8%]) patients. Median (IQR) enrollment of White patients was 91.3% (87.3%-93.6%), whereas enrollment of underrepresented patient populations was markedly lower, with median (IQR) enrollment of 0.9% (0.6%-1.6%) for Asian or Pacific Islander, 4.5% (3.6%-6.6%) for Black (ethnicity unspecified), 7.2% (3.7%-9.1%) for Black (non-Hispanic), 5.2% (3.1%-6.6%) for Hispanic, and 0.4% (0%-0.8%) for Native American patients. Few trials (3 of 71 [4.2%]) conducted subgroup analyses by race or ethnicity, and none reported detailed subgroup characteristics or safety or efficacy outcomes by patient race and ethnicity. Reporting practices and representation showed little improvement over time. Conclusions and Relevance: US-based phase 3 AD trials showed substantial gaps in racial and ethnic reporting and representation from 1997 to 2023, limiting the evaluation of treatment safety and efficacy across diverse populations. These findings suggest that stronger reporting standards and more inclusive trial design and recruitment strategies are needed to improve the equity and generalizability of AD trials.
- Open Access
- 2 March 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (3) , e261796
Risk Adjustment for Alzheimer Disease and Related Dementias in Medicare Advantage and Health Care Experiences
Importance: Failure to account for the full complexity and costs of high-need populations in the risk-adjusted capitated payment model for Medicare Advantage (MA) plans may create financial disincentives for plans to invest in comprehensive care for affected beneficiaries, potentially exacerbating health disparities. Objective: To evaluate the association of reinstatement of Alzheimer disease and related dementias (ADRD) hierarchical condition categories (HCCs) into the MA risk-adjusted payment model in 2020 with access, affordability, and quality of care for beneficiaries with ADRD. Design, Setting, and Participants: This cross-sectional study examined a nationally representative sample of MA beneficiaries from the Medicare Current Beneficiary Survey (2015-2022). Beneficiaries with ADRD and those without ADRD but with comparable neurological diseases (stroke, paralysis, or Parkinson disease) before and after 2020 were included. Data analyses were performed between January and December 2025. Exposures: Reinstatement of the ADRD HCC into the MA risk adjustment formula in 2020. Main Outcomes and Measures: Primary outcomes were accessibility of needed care, medical financial burden, satisfaction with specialist access, and satisfaction with quality of care. These outcomes were assessed using a difference-in-differences model to compare changes between the treatment and control group before and after the inclusion of ADRD HCCs in the MA risk adjustment model in 2020. Results: Among 5353 MA beneficiary observations (1239 [23.1%] aged 65-74 years; 3127 [58.4%] aged ≥75 years; 1785 male [33.3%]), 1629 (30.4%) reported a diagnosis of ADRD, and 3724 (69.6%) did not report an ADRD diagnosis. Compared with MA beneficiaries without ADRD, those with ADRD reported lower rates of difficulty accessing care (142 beneficiaries [8.7%] vs 394 beneficiaries [10.6%]) and medical financial burden (235 beneficiaries [19.3%] vs 740 beneficiaries [25.1%]), but slightly lower rates of satisfaction with specialist access (1384 beneficiaries [90.8%] vs 3267 [92.7%]) and care quality (1495 beneficiaries [92.8%] vs 3414 beneficiaries [93.0%]). Reintroducing ADRD HCCs into the MA risk-adjusted payment model was associated with a 6.62 percentage-point decrease in reporting any troubles accessing needed care (β = 0.06; 95% CI, −0.11 to −0.02; P = .005) and a 9.20 percentage-point decrease in reporting any medical financial burden (β = −0.09; 95% CI, −0.16 to −0.02; P = .009) among MA beneficiaries with ADRD. No significant association was observed for satisfaction with specialist access or with quality of care among MA beneficiaries with ADRD. Conclusions and Relevance: In this cross-sectional study of MA beneficiaries, reintroducing ADRD HCCs into the MA risk adjustment model was associated with improved care access and reduced financial burden among MA beneficiaries with ADRD. These findings suggest that risk adjustment that better reflects the costs of chronic, complex conditions may better align MA plan incentives with the needs of high-need populations and promote care equity.
- 2 March 2026
- Published by American Medical Association (AMA) in JAMA Network Open
- Vol. 9 (3) , e261796
Open AccessRisk Adjustment for Alzheimer Disease and Related Dementias in Medicare Advantage and Health Care Experiences
Importance: Failure to account for the full complexity and costs of high-need populations in the risk-adjusted capitated payment model for Medicare Advantage (MA) plans may create financial disincentives for plans to invest in comprehensive care for affected beneficiaries, potentially exacerbating health disparities. Objective: To evaluate the association of reinstatement of Alzheimer disease and related dementias (ADRD) hierarchical condition categories (HCCs) into the MA risk-adjusted payment model in 2020 with access, affordability, and quality of care for beneficiaries with ADRD. Design, Setting, and Participants: This cross-sectional study examined a nationally representative sample of MA beneficiaries from the Medicare Current Beneficiary Survey (2015-2022). Beneficiaries with ADRD and those without ADRD but with comparable neurological diseases (stroke, paralysis, or Parkinson disease) before and after 2020 were included. Data analyses were performed between January and December 2025. Exposures: Reinstatement of the ADRD HCC into the MA risk adjustment formula in 2020. Main Outcomes and Measures: Primary outcomes were accessibility of needed care, medical financial burden, satisfaction with specialist access, and satisfaction with quality of care. These outcomes were assessed using a difference-in-differences model to compare changes between the treatment and control group before and after the inclusion of ADRD HCCs in the MA risk adjustment model in 2020. Results: Among 5353 MA beneficiary observations (1239 [23.1%] aged 65-74 years; 3127 [58.4%] aged ≥75 years; 1785 male [33.3%]), 1629 (30.4%) reported a diagnosis of ADRD, and 3724 (69.6%) did not report an ADRD diagnosis. Compared with MA beneficiaries without ADRD, those with ADRD reported lower rates of difficulty accessing care (142 beneficiaries [8.7%] vs 394 beneficiaries [10.6%]) and medical financial burden (235 beneficiaries [19.3%] vs 740 beneficiaries [25.1%]), but slightly lower rates of satisfaction with specialist access (1384 beneficiaries [90.8%] vs 3267 [92.7%]) and care quality (1495 beneficiaries [92.8%] vs 3414 beneficiaries [93.0%]). Reintroducing ADRD HCCs into the MA risk-adjusted payment model was associated with a 6.62 percentage-point decrease in reporting any troubles accessing needed care (β = 0.06; 95% CI, −0.11 to −0.02; P = .005) and a 9.20 percentage-point decrease in reporting any medical financial burden (β = −0.09; 95% CI, −0.16 to −0.02; P = .009) among MA beneficiaries with ADRD. No significant association was observed for satisfaction with specialist access or with quality of care among MA beneficiaries with ADRD. Conclusions and Relevance: In this cross-sectional study of MA beneficiaries, reintroducing ADRD HCCs into the MA risk adjustment model was associated with improved care access and reduced financial burden among MA beneficiaries with ADRD. These findings suggest that risk adjustment that better reflects the costs of chronic, complex conditions may better align MA plan incentives with the needs of high-need populations and promote care equity.
- 26 November 2025
- Published by Oxford University Press (OUP) in The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences
- Vol. 81 (2)
Uncovering the biological toll of neighborhood disorder trajectories: a machine learning based weighting analysis of biomarkers in older adults
Objectives: Neighborhood physical disorder has been linked to adverse health outcomes, but less is known about longitudinal patterns of disorder trajectories and their associations with biological markers. This study examined associations between neighborhood physical disorder trajectories and metabolic and inflammatory biomarkers in U.S. older adults. Methods: We analyzed data from community-dwelling Medicare beneficiaries in the National Health and Aging Trends Study (n = 2,284, 2011–2017). Neighborhood physical disorder was assessed annually through interviewer observations over six years. Latent class analysis was used to identify exposure trajectory subgroups. Machine learning-based inverse probability weighted (IPW) regression models were conducted to estimate associations with five biomarkers, including body mass index (BMI), waist circumference, hemoglobin A1c (HbA1c), high-sensitivity C-reactive protein (hsCRP), and interleukin-6 (IL-6). Results: We identified four subgroup trajectories: stable low exposure (85%), increased exposure (3%), decreased exposure (9%), and stable high exposure (3%). Regression models revealed significant associations with HbA1c and hsCRP. Compared to the stable low exposure group, older adults with increased exposure (b = 0.04, 95% CI: 0.01–0.08), decreased exposure (b = 0.02, 95% CI: 0.01–0.05), and stable high exposure (b = 0.10, 95% CI: 0.03–0.17) exhibited higher levels of HbA1c. Only stable high exposure was associated with increased hsCRP (b = 0.25, 95% CI: 0.05–0.45). No significant associations were found for other biomarkers. Discussion: Residential environments play an important role in shaping the biological risk of aging. Incorporating routine screening for neighborhood environmental risks and implementing community-level interventions are pivotal in promoting healthy aging in place.
- 26 November 2025
- Published by Oxford University Press (OUP) in The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences
- Vol. 81 (2)
Uncovering the biological toll of neighborhood disorder trajectories: a machine learning based weighting analysis of biomarkers in older adults
Objectives: Neighborhood physical disorder has been linked to adverse health outcomes, but less is known about longitudinal patterns of disorder trajectories and their associations with biological markers. This study examined associations between neighborhood physical disorder trajectories and metabolic and inflammatory biomarkers in U.S. older adults. Methods: We analyzed data from community-dwelling Medicare beneficiaries in the National Health and Aging Trends Study (n = 2,284, 2011–2017). Neighborhood physical disorder was assessed annually through interviewer observations over six years. Latent class analysis was used to identify exposure trajectory subgroups. Machine learning-based inverse probability weighted (IPW) regression models were conducted to estimate associations with five biomarkers, including body mass index (BMI), waist circumference, hemoglobin A1c (HbA1c), high-sensitivity C-reactive protein (hsCRP), and interleukin-6 (IL-6). Results: We identified four subgroup trajectories: stable low exposure (85%), increased exposure (3%), decreased exposure (9%), and stable high exposure (3%). Regression models revealed significant associations with HbA1c and hsCRP. Compared to the stable low exposure group, older adults with increased exposure (b = 0.04, 95% CI: 0.01–0.08), decreased exposure (b = 0.02, 95% CI: 0.01–0.05), and stable high exposure (b = 0.10, 95% CI: 0.03–0.17) exhibited higher levels of HbA1c. Only stable high exposure was associated with increased hsCRP (b = 0.25, 95% CI: 0.05–0.45). No significant associations were found for other biomarkers. Discussion: Residential environments play an important role in shaping the biological risk of aging. Incorporating routine screening for neighborhood environmental risks and implementing community-level interventions are pivotal in promoting healthy aging in place.
- Open Access
- 25 September 2025
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (11) , 3464-3474
Association Between Supplemental Nutrition Assistance Program ( SNAP ) Participation and Cognitive Decline Among Older Americans: A 10‐Year Longitudinal Study
Background: This study examines whether SNAP participation slows age‐related cognitive decline among cognitively intact older adults over 10 years. Methods: We conducted a retrospective cohort study using data from SNAP‐eligible participants in the Health and Retirement Study, a nationally representative panel survey of U.S. adults aged 50 years and older. Data was collected biennially from 2010 to 2020, and linear mixed regression models were used to estimate associations between SNAP participation and cognitive decline across different population groups. Results: SNAP participation was associated with a slower cognitive decline in global cognition ( β = 0.10, 95% CI: 0.05, 0.14, p < 0.001), memory ( β = 0.07, 95% CI: 0.03, 0.11, p < 0.001), and executive function ( β = 0.03, 95% CI: 0.01, 0.04, p = 0.004). A significant three‐way interaction among SNAP participation, race/ethnicity, and time indicated a slightly faster decline in global cognition among Non‐Hispanic Black and Hispanic SNAP users. Discussion: SNAP participation may help slow age‐related cognitive decline. However, the benefits of SNAP vary across different population groups. Further research is warranted to explore the underlying mechanisms and to design interventions that can maximize the benefits of SNAP in improving cognitive decline.
- 25 September 2025
- Published by Wiley in Journal of the American Geriatrics Society
- Vol. 73 (11) , 3464-3474
Open AccessAssociation Between Supplemental Nutrition Assistance Program ( SNAP ) Participation and Cognitive Decline Among Older Americans: A 10‐Year Longitudinal Study
Background: This study examines whether SNAP participation slows age‐related cognitive decline among cognitively intact older adults over 10 years. Methods: We conducted a retrospective cohort study using data from SNAP‐eligible participants in the Health and Retirement Study, a nationally representative panel survey of U.S. adults aged 50 years and older. Data was collected biennially from 2010 to 2020, and linear mixed regression models were used to estimate associations between SNAP participation and cognitive decline across different population groups. Results: SNAP participation was associated with a slower cognitive decline in global cognition ( β = 0.10, 95% CI: 0.05, 0.14, p < 0.001), memory ( β = 0.07, 95% CI: 0.03, 0.11, p < 0.001), and executive function ( β = 0.03, 95% CI: 0.01, 0.04, p = 0.004). A significant three‐way interaction among SNAP participation, race/ethnicity, and time indicated a slightly faster decline in global cognition among Non‐Hispanic Black and Hispanic SNAP users. Discussion: SNAP participation may help slow age‐related cognitive decline. However, the benefits of SNAP vary across different population groups. Further research is warranted to explore the underlying mechanisms and to design interventions that can maximize the benefits of SNAP in improving cognitive decline.