 |
|
|
| |
Dementia incidence and prevalence in older adults with HIV: a 23-year retrospective cohort study
|
| |
| |
Download the PDF here
Download the PDF here
Discussion
Incident dementia diagnoses among people with HIV have declined and are converging with rates observed in demographically matched peers without HIV. Encouragingly, this downward trend was observed across all race/ethnicity and sex subgroups of people with HIV. However, dementia prevalence remains elevated - 38.31 cases per 1000 persons in the recent period (2020-2023) - with consistently elevated rates across all demographic subgroups, likely reflecting higher cumulative risk from earlier years. Therefore, dementia screening and care will need to remain a continued focus in HIV primary care for years to come.
While declining trends in dementia incidence were generally consistent across demographic subgroups, a modest increase among Hispanic individuals with HIV during 2020-2023 warrants further investigation, especially given Centers for Disease Control (CDC) projections that Hispanic populations will face the largest increases in dementia incidence in coming years [16]. In terms of dementia prevalence, people with HIV had a significantly higher burden of dementia across all racial/ethnic groups, highlighting the need to integrate HIV-specific considerations into broader aging care strategies. Given that people with HIV often experience social isolation and have fewer informal caregiving options, adapting existing HIV and dementia care infrastructure to meet increasing demands could support more inclusive care and positively impact the wellbeing of people aging with HIV [17,18]. Ensuring equitable access to dementia and HIV-related services will also be vital, particularly for racial and ethnic minority populations who often experience poorer dementia and HIV health outcomes [19-22].
Although our study population was predominantly male (86%), dementia prevalence was higher among women, regardless of HIV status, consistent with general population studies [31,32]. As the HIV population continues to age, factors that contribute to dementia risk among women, including menopausal transitions, hormone replacement therapy [33-37], and life-course adversity affecting cognitive reserve [32,38,39], may widen differences in dementia prevalence by sex. Understanding how these factors interact with HIV is an important area for future research.
Lam, Jennifer O.a,b; Fan, Dongjiea; Pothamsetty, Navyaa; Samiezade-Yazd, Zahraa; Hu, Haihongc; Lopez, Errold; Lee, Catherinee; Lea, Alexandra N.a; Hou, Craig E.f; Towner, William J.d,g; Horberg, Michael A.b,c; Silverberg, Michael J.a,b,e
AIDS April 4 2026
Objective:
To compare dementia incidence and prevalence by HIV status, race/ethnicity, and sex.
Design:
A retrospective cohort, 2000-2023.
Methods:
Adults with HIV aged at least 50 years and 1 : 20 matched individuals without HIV from Kaiser Permanente, a U.S. healthcare system, were included. Dementia diagnoses were identified via electronic health records. We estimated rates of incident dementia diagnoses and prevalence, overall and by time period (2000-2004, 2005-2009…2020-2023) using Poisson regression, and assessed trends using Joinpoint regression. Covariate-adjusted rate ratios compared dementia by HIV status, with sub-analyses stratified by race/ethnicity and sex.
Results:
Among 24 762 people with HIV and 494 963 people without HIV (86.9% men, 45.5% White, 23.1% Black, 20.3% Hispanic), incident dementia diagnoses declined from 2000 to 2023 in both people with and without HIV (-7.68 and -2.70% per period, respectively).
Overall, the incidence of dementia diagnosis was higher in people with HIV (adjusted incidence rate ratio [aIRR]=1.72, 95% CI=1.59-1.85).
In the most recent period (2020-2023), this difference was not statistically significant (aIRR=1.16, 95% CI=0.99-1.35), partly due to increases in diagnoses among people without HIV during this period.
Dementia prevalence remained higher in people with HIV, overall (adjusted prevalence ratio [aPR]=1.71, 95% CI=1.61-1.82) and in 2020-2023 (aPR=1.59, 95% CI=1.46-1.73), with similar patterns by race/ethnicity and sex.
Conclusion:
Incident dementia diagnoses have declined in people with HIV and are approaching those of people without HIV, with consistent trends across demographic subgroups. However, prevalence remains elevated, likely reflecting excess risk from earlier years. These findings highlight the need for sustained attention to cognitive health and the integration of dementia-related services in HIV care.
Introduction
In the U.S. and globally, age-related cognitive decline is of increasing concern in the aging HIV population. People with HIV are known to experience a greater burden of cognitive impairments compared to the general population [1-4]. Recent studies also suggest an elevated risk of age-associated dementia among people with HIV, even among those on effective antiretroviral therapy (ART) [5-8], which could cause substantial health burden for both individuals and healthcare systems. Contributing factors include comorbidities that affect vascular health, behavioral factors such as substance use, and persistent HIV reservoirs-all of which may adversely impact neurocognition and accentuate brain aging [9-12].
Despite growing recognition of the potential impacts of age-associated dementia in this population, limited evidence exists on how dementia burden among people with HIV compares to that of the general population. Further, no prior study has characterized the epidemiology of age-associated dementia among people with HIV by race and ethnicity. In the general population, the overall frequency of dementia is projected to triple within the next 3 decades due to population aging, becoming a leading driver of morbidity and healthcare expenditures [13]. Given the disproportionately higher dementia risk among people with HIV, understanding dementia epidemiology in this population will be critical to prepare for future healthcare demands.
To contribute knowledge on the burden of dementia in the aging HIV population, we assembled a diverse cohort of adults with HIV and demographically matched comparators without HIV from three integrated healthcare systems in the United States. In this study, we describe our initial analyses of incident dementia diagnoses and prevalence, overall and stratified by race/ethnicity and sex.
-------------------
In trend analyses stratified by race/ethnicity, declining dementia incidence was observed across all racial/ethnic groups, regardless of HIV status (Fig. 1; Supplemental Table 2, https://links.lww.com/QAD/D749). Among Hispanic people with HIV, however, the sIR increased modestly in the final time period-from an sIR of 2.08 (1.31, 3.15) in 2015-2019 to an sIR of 2.37 (1.60, 3.38) in 2020-2023, though this was not statistically significant and case numbers were small (23 dementia cases in 2015-2019 and 33 dementia cases in 2020-2023). This trend was not observed in any other racial/ethnic group. In trend analyses stratified by sex, dementia incidence declined consistently across all time periods for both men and women.
Discussion
Incident dementia diagnoses among people with HIV have declined and are converging with rates observed in demographically matched peers without HIV. Encouragingly, this downward trend was observed across all race/ethnicity and sex subgroups of people with HIV. However, dementia prevalence remains elevated - 38.31 cases per 1000 persons in the recent period (2020-2023) - with consistently elevated rates across all demographic subgroups, likely reflecting higher cumulative risk from earlier years. Therefore, dementia screening and care will need to remain a continued focus in HIV primary care for years to come.
This study helps fill a critical knowledge gap by describing the burden of dementia among people with HIV by race/ethnicity and sex, an important step in understanding cognitive health in an aging and increasingly diverse HIV population. While declining trends in dementia incidence were generally consistent across demographic subgroups, a modest increase among Hispanic individuals with HIV during 2020-2023 warrants further investigation, especially given Centers for Disease Control (CDC) projections that Hispanic populations will face the largest increases in dementia incidence in coming years [16]. In terms of dementia prevalence, people with HIV had a significantly higher burden of dementia across all racial/ethnic groups, highlighting the need to integrate HIV-specific considerations into broader aging care strategies. Given that people with HIV often experience social isolation and have fewer informal caregiving options, adapting existing HIV and dementia care infrastructure to meet increasing demands could support more inclusive care and positively impact the wellbeing of people aging with HIV [17,18]. Ensuring equitable access to dementia and HIV-related services will also be vital, particularly for racial and ethnic minority populations who often experience poorer dementia and HIV health outcomes [19-22].
Declining dementia incidence may be due to improved ART and better management of vascular risk factors common in people with HIV, such as cardiovascular disease, hypertension, and smoking [23-25]. However, maintaining consistent HIV viral suppression can be challenging, particularly in older populations where cognitive decline may impair ART adherence. In our cohort, 95.4% of individuals with detectable viral load at baseline had at least one subsequent episode of detectable viral load, and 23.2% of those with undetectable baseline viral load later experienced viral rebound, which could reflect the bidirectional relationship between HIV control and cognition.
Since our study period overlapped with the COVID-19 pandemic, changes in healthcare utilization may have influenced dementia recognition and diagnoses, impacting dementia estimates in the final years of our study. For instance, emerging knowledge about the neurologic effects of COVID-19 infection could have increased rates of cognitive screening [26-28]. Additionally, differential COVID-19-related mortality among people with dementia, or in the overall population of older adults may have affected observed incidence, though whether this varied by HIV status is unknown [29,30]. Continued monitoring of postpandemic trends will be needed to assess whether the observed convergence of dementia incidence persists and whether it varies by HIV status or demographic subgroup.
Although our study population was predominantly male (86%), dementia prevalence was higher among women, regardless of HIV status, consistent with general population studies [31,32]. As the HIV population continues to age, factors that contribute to dementia risk among women, including menopausal transitions, hormone replacement therapy [33-37], and life-course adversity affecting cognitive reserve [32,38,39], may widen differences in dementia prevalence by sex. Understanding how these factors interact with HIV is an important area for future research.
This study had several limitations. First, dementia diagnoses were identified using ICD codes from routine clinical data, which could be influenced by changes in diagnostic practice and coding over time. Matching of people with and without HIV by year of cohort entry, along with mapping ICD codes to consistent dementia categories across the study period, would likely have minimized the impact of these cohort effects. Also, these data would capture patterns of care generally seen in clinical practice. Second, small numbers of dementia cases among Asian participants limited our ability to conduct stratified analyses or examine temporal trends in this group, an issue also present in other studies [32,40,41]. Further, underdiagnosis of dementia has been reported across racial and ethnic groups, including Asian, Black, and Hispanic populations, due to a combination of cultural, linguistic, and healthcare-related factors [40,42-46]. In our study, low case counts among Asian participants may reflect these broader patterns as well as cultural norms around aging specific to this population [41,43,47,48]. Third, analyses may have been affected by residual confounding from dementia risk factors not routinely recorded in the EHR, such as physical activity levels and environmental exposures. Fourth, models were adjusted for neighborhood-level education as a proxy for individual-level educational attainment, which may not fully capture the influence of education on dementia risk. Finally, although we report dementia types descriptively, models comparing incidence and prevalence by HIV status were not stratified by type. However, though dementia types may differ in etiology and trajectory, clinical care needs would largely overlap; therefore, our conclusions regarding the importance of high-quality dementia care would still be applicable.
This study also had notable strengths. KP serves a large population of older adults with HIV in an integrated healthcare setting and is broadly representative of the racial and ethnic diversity of the underlying community [49]. By harmonizing 23 years of KP EHR and HIV registry data across three U.S. regions, we assembled a demographically diverse study population suitable for both race-stratified and sex-stratified analyses. People with HIV had stable access to ART, and ART uptake was high (82% at baseline), reducing potential confounding effects of suboptimally treated HIV. In addition, our inclusion of a matched comparator group of people without HIV from the same healthcare setting enabled robust comparisons of dementia burden and helped account for potential differences in dementia diagnosis due to differential access to care. The ability to adjust for individual-level comorbidities and healthcare utilization further strengthened our estimates. Finally, because this study was conducted within integrated health systems our findings may generalize to other insured populations with similar access to care.
Conclusion
This study offers one of the most comprehensive longitudinal assessments of dementia among people with HIV in the modern ART era. The decline in incident dementia diagnoses is an encouraging trend that may reflect improved HIV treatment and care. However, the persistently higher prevalence underscores the need for sustained attention to cognitive health and integration of dementia-related services into HIV care.
--
|
| |
|
|
|
|
|
