Abstract
Background:
Housing instability and HIV are both associated with early onset of aging-related health conditions, including frailty. However, little is known about the relationship between housing and frailty among people with HIV.
Methods:
We analyzed data on adults in HIV care collected during routine clinical visits at 6 sites within the US Centers for AIDS Research Network of Integrated Clinical Systems (CNICS) between 2019–2024. We measured frailty using a previously validated approach, defined as having ≥3 of 4 components (fatigue, weight loss, immobility, inactivity) vs. <3 components. Housing status was based on self-perceived living situation in the past month (“Stable”, “Unstable”, “Homeless”, or “Don’t know”). We estimated the association between most recent housing status and frailty with prevalence ratios (PR) from relative risk regression adjusted for sociobehavioral and clinical characteristics.
Results:
Among 6,961 people with HIV (84% male, 16% female) with a median age of 52 years (IQR: 40–60), 11% (n=760) were frail and 9% (n=625) were unstably housed (5% unstable, 3% homeless, 1% don’t know). Compared to individuals with stable housing, the prevalence of frailty more than doubled among those experiencing unstable housing (PR=2.41, 95% CI: 1.95, 2.97) or homelessness (PR=2.05, 95% CI: 1.56, 2.69). Stratified analyses indicated stronger associations among younger vs. older individuals and among those virally suppressed vs. unsuppressed.
Conclusions:
Housing instability and frailty were both prevalent and strongly associated among adults in HIV care, including within younger and virally suppressed subgroups. These findings highlight the importance of social determinants of health for clinical outcomes among all people with HIV.
Introduction
Amidst a nationwide housing affordability crisis and rising housing instability across the United States [1], older adults and people with HIV (PWH) face disproportionate challenges [2–4]. Housing instability includes multifaceted issues such as difficulty paying rent or mortgage payments, overcrowding, moving frequently, spending a large share of income on housing, or, at the extreme, experiencing homelessness [5,6]. Older adults comprise the fastest growing group of unstably housed people in the US, with many facing housing instability for the first time in their lives [2,7]. More than 771,000 people nationally experienced homelessness on a single night in January 2024, an all-time high and an 18% increase from 2023; people aged 55 and older accounted for one in five of those experiencing homelessness, nearly half of whom were unsheltered [8]. Among PWH, nearly one in five adults experienced homelessness or other forms of housing instability in 2022 [9]. Meanwhile, PWH are also increasingly older due to a combination of effective modern treatment as well as later-life HIV transmission [10]. However, little attention has focused on the dynamics between HIV, aging, and housing instability [11].
Living with HIV and experiencing housing instability both confer an increased risk of accelerated aging [12,13]. Frailty, a geriatric syndrome characterized by a decline in physiological function [14], manifests earlier and more often among PWH and unstably housed people than among the general population [15–17]. Both groups have a higher prevalence of risk factors for frailty such as substance use and mental health conditions, while HIV-related chronic inflammation is also thought to contribute to frailty [18–21]. Frailty results in heightened vulnerability to serious adverse health outcomes including falls, hospitalizations, and mortality [22,23]. Among PWH, frailty may also impede adherence to antiretroviral therapy (ART) and achievement of viral suppression [24], a critical target of HIV treatment and prevention efforts due to individual health benefits and diminished biological potential for onward transmission [25]. This is especially worrying in the context of housing instability, which also poses barriers to ART adherence and a higher likelihood of HIV transmission risk behaviors [26–29]. Despite the overlap in HIV, aging, and housing instability, few studies have investigated the association between housing status and aging-related conditions in PWH, including frailty [11]. Understanding the relationship between housing and healthy aging is critical for strategies to prevent and treat HIV and to improve quality of life for PWH.
As a first step, this study sought to elucidate the current combined burden of housing instability and frailty among PWH engaged in clinical care across the US. We compared the relative prevalence of frailty among PWH reporting stable housing, unstable housing, and homelessness and examined heterogeneity by sociodemographic and clinical characteristics.
Methods
Study design and participants
We conducted a cross-sectional analysis among PWH in the Centers for AIDS Research Network of Integrated Clinical Systems (CNICS), a long-running, dynamic cohort of PWH at least 18 years of age engaged in HIV primary care across the US (http://www.uab.edu/cnics) [30]. CNICS study procedures have been approved by Institutional Review Boards at each participating site and this analysis was based on deidentified data.
Of the 10 HIV clinics that contribute to CNICS, 6 sites had collected both housing status and frailty data at the time of analysis and were involved in this study. Participants included PWH who attended a visit and completed the CNICS clinical assessment of patient reported measures and outcomes (PROs) between June 2019, when housing questions were first implemented, and July 2024.
Data collection and measurement
The CNICS data repository includes extensive information across a spectrum of demographic, clinical, laboratory, and medication records for both outpatient and inpatient encounters, derived from electronic health records and other institutional data sources. The CNICS clinical assessment of PROs, which is available in multiple languages (English, Spanish, Brazilian Portuguese, Haitian Creole, Amharic), is completed by PWH on touch-screen tablet computers during routine HIV primary care visits [31,32].
Housing status assessment
The exposure of interest was self-perceived housing stability reported on the most recent PRO assessment. Housing stability was assessed with the following question: “In the past month, how would you describe your living situation?” with response options of “Stable”, “Unstable”, “Homeless”, or “Don’t know”.
To better understand the circumstances of PWH selecting each living situation response, we also tabulated responses to an additional housing-related question asking about types of locations where the respondent stayed for at least one night in the past month (e.g., their own home, someone else’s home, unsheltered, etc.). This second question was examined for descriptive purposes and was not used in the definition of housing status as the exposure variable, which focused on self-perceived housing stability.
Frailty outcomes
The primary outcome was frailty status measured at the same visit when housing status was assessed, or the closest visit within 3 years to include individuals who did not receive both assessments on the same day. Frailty, parameterized as a binary outcome, was defined as reporting ≥3 of 4 components of a modified Fried frailty phenotype [33] (mFFP: fatigue, weight loss, immobility, and inactivity). This self-reported mFFP was previously validated in CNICS and found to be highly correlated with the FFP, and has been described in detail [34]. Briefly, the 4 components of the mFFP are collected via PRO assessment: symptoms of fatigue/loss of energy and weight loss/wasting over the previous 4 weeks; current mobility, assessed in terms of self-reported ability to walk; and physical activity, assessed in terms of self-reported engagement in strenuous exercise compared to others of the same age and sex.
As a secondary analysis, we also assessed frailty status categorized into three levels based on number of reported components: robust (0), prefrail (1–2), and frail (3 or 4 components) [35]. In addition to the primary outcome of frail vs. not frail (combined robust/prefrail), we examined secondary outcomes including frail vs. robust (excluding prefrail), frail vs. prefrail (excluding robust), and prefrail vs. robust (excluding frail).
Statistical analyses
We used relative risk regression with robust standard errors [36] to estimate prevalence ratios (PR) and 95% confidence intervals (CI) for the association between housing status and each binary frailty outcome. Housing status was mainly analyzed as a 4-level variable (stable, unstable, homeless, don’t know). Primary models adjusted for sociodemographic characteristics (site, age, sex, race and ethnicity, and year), substance use (alcohol, smoking, and drug use: illicit opioids, cocaine/crack, methamphetamine/crystal), HIV clinical indicators (viral load and CD4 count), and comorbidities (diabetes, hypertension, hepatitis B, hepatitis C, and body mass index; Supplementary Table 1). Considering that many of these covariates could have bidirectional relationships with housing and frailty, we also implemented the following secondary models using a staged approach: a) only adjusted for sociodemographic characteristics; b) adjusted for sociodemographic characteristics and substance use; and c) adjusted for sociodemographic characteristics, substance use, and HIV clinical indicators, but not for comorbidities as in the primary fully adjusted model. We conducted complete case analyses because few PWH in our sample were missing covariate data (Supplementary Figure 1). All models excluded PWH missing any of the covariates in the primary adjustment set. We also conducted heterogeneity analyses stratified by sex, age (<40, 40–50, 50–60, ≥60), current HIV viral suppression status, and region. Due to smaller sample sizes in stratified models, we analyzed housing instability as a binary variable (stable vs. combined unstable/homeless, excluding “don’t know”) in heterogeneity analyses. Lastly, we conducted a sensitivity analysis limiting our sample to PWH with housing and frailty measured on the same day. All statistical analyses were conducted using Stata version 18.5 (StataCorp, College Station, TX, USA).
Results
This analysis included 6,961 PWH (84% male, n=5,845; 16% female, n=1,116; Table 1) with a median age of 52 years (IQR: 40–60). Overall, 8.9% of PWH (n=625) reported housing instability in the past month, including those who described their living situation as “unstable” (4.7%, n=330), “homeless” (2.9%, n=205) or “don’t know” (1.3%, n=90). Individuals who described their living situation as “stable” primarily reported sleeping in their own home or apartment in the past month (92%, n=5,591; Figure 1). Those with “unstable” housing most often reported sleeping in their own home or someone else’s home, as did PWH who responded “don’t know”. Those who reported experiencing homelessness most often slept in someone else’s home or in an unsheltered location (outdoors, tent, abandoned building, public space, or car). The median year of housing status assessment was 2023 (89% ≥2021, n= 6,178), and frailty was measured on the same day as housing for 80% of PWH (n=5,543).
Table 1.
Sociodemographic and clinical characteristics by housing status among people with HIV engaged in clinical care at 6 sites across the United States, 2019–2024.
| Self-perceived living situation in the past month |
|||||
|---|---|---|---|---|---|
| Overall (n=6,961) | Stable (n=6,336) | Unstable (n=330) | Homeless (n=205) | Don’t know (n=90) | |
|
| |||||
| Sex | |||||
| Female | 1,116 (16.0%) | 1,022 (16.1%) | 46 (13.9%) | 25 (12.2%) | 23 (25.6%) |
| Male | 5,845 (84.0%) | 5,314 (83.9%) | 284 (86.1%) | 180 (87.8%) | 67 (74.4%) |
| Age (years) | |||||
| <30 | 312 (4.5%) | 268 (4.2%) | 30 (9.1%) | 10 (4.9%) | 4 (4.4%) |
| 30–40 | 1,315 (18.9%) | 1,134 (17.9%) | 97 (29.4%) | 54 (26.3%) | 30 (33.3%) |
| 40–50 | 1,390 (20.0%) | 1,242 (19.6%) | 83 (25.2%) | 42 (20.5%) | 23 (25.6%) |
| 50–60 | 2,001 (28.7%) | 1,835 (29.0%) | 82 (24.8%) | 66 (32.2%) | 18 (20.0%) |
| ≥60 | 1,943 (27.9%) | 1,857 (29.3%) | 38 (11.5%) | 33 (16.1%) | 15 (16.7%) |
| Race and ethnicity | |||||
| White | 3,111 (44.7%) | 2,907 (45.9%) | 110 (33.3%) | 71 (34.6%) | 23 (25.6%) |
| Black | 2,606 (37.4%) | 2,316 (36.6%) | 161 (48.8%) | 80 (39.0%) | 49 (54.4%) |
| Hispanic | 874 (12.6%) | 772 (12.2%) | 47 (14.2%) | 42 (20.5%) | 13 (14.4%) |
| Other/Unknown | 370 (5.3%) | 341 (5.4%) | 12 (3.6%) | 12 (5.9%) | 5 (5.6%) |
| Region | |||||
| Northeast/Midwest | 1,255 (18.0%) | 1,170 (18.5%) | 55 (16.7%) | 16 (7.8%) | 14 (15.6%) |
| South | 2,896 (41.6%) | 2,637 (41.6%) | 148 (44.8%) | 68 (33.2%) | 43 (47.8%) |
| West | 2,810 (40.4%) | 2,529 (39.9%) | 127 (38.5%) | 121 (59.0%) | 33 (36.7%) |
| Alcohol use | |||||
| No drinking | 2,349 (33.7%) | 2,105 (33.2%) | 110 (33.3%) | 93 (45.4%) | 41 (45.6%) |
| Not at-risk drinking | 3,554 (51.1%) | 3,267 (51.6%) | 165 (50.0%) | 84 (41.0%) | 38 (42.2%) |
| At-risk drinking | 1,058 (15.2%) | 964 (15.2%) | 55 (16.7%) | 28 (13.7%) | 11 (12.2%) |
| Cigarette/vaping use | |||||
| Never | 2,114 (30.4%) | 2,001 (31.6%) | 56 (17.0%) | 25 (12.2%) | 32 (35.6%) |
| Past | 2,913 (41.8%) | 2,697 (42.6%) | 112 (33.9%) | 71 (34.6%) | 33 (36.7%) |
| Current (past 3 months) | 1,934 (27.8%) | 1,638 (25.9%) | 162 (49.1%) | 109 (53.2%) | 25 (27.8%) |
| Drug usea | |||||
| Never | 2,057 (29.6%) | 1,921 (30.3%) | 78 (23.6%) | 29 (14.1%) | 29 (32.2%) |
| Past | 3,418 (49.1%) | 3,148 (49.7%) | 146 (44.2%) | 81 (39.5%) | 43 (47.8%) |
| Current (past 3 months) | 1,486 (21.3%) | 1,267 (20.0%) | 106 (32.1%) | 95 (46.3%) | 18 (20.0%) |
| HIV viral load <50 copies/mL | 6,053 (87.0%) | 5,604 (88.4%) | 249 (75.5%) | 128 (62.4%) | 72 (80.0%) |
| CD4 count <350 cells/mm3 | 931 (13.4%) | 803 (12.7%) | 65 (19.7%) | 47 (22.9%) | 16 (17.8%) |
| Diabetes | 1,398 (20.1%) | 1,294 (20.4%) | 51 (15.5%) | 35 (17.1%) | 18 (20.0%) |
| Hypertension | 3,205 (46.0%) | 2,972 (46.9%) | 111 (33.6%) | 84 (41.0%) | 38 (42.2%) |
| Hepatitis B | 311 (4.5%) | 281 (4.4%) | 10 (3.0%) | 17 (8.3%) | 3 (3.3%) |
| Hepatitis C | 911 (13.1%) | 807 (12.7%) | 51 (15.5%) | 43 (21.0%) | 10 (11.1%) |
| BMI | |||||
| <18.5 | 104 (1.5%) | 89 (1.4%) | 10 (3.0%) | 4 (2.0%) | 1 (1.1%) |
| 18.5–24.9 | 1,940 (27.9%) | 1,711 (27.0%) | 123 (37.3%) | 77 (37.6%) | 29 (32.2%) |
| 25.0–29.9 | 2,551 (36.6%) | 2,355 (37.2%) | 95 (28.8%) | 73 (35.6%) | 28 (31.1%) |
| ≥30 | 2,366 (34.0%) | 2,181 (34.4%) | 102 (30.9%) | 51 (24.9%) | 32 (35.6%) |
| Frailty status | |||||
| Robust | 3,071 (44.1%) | 2,880 (45.5%) | 84 (25.5%) | 74 (36.1%) | 33 (36.7%) |
| Prefrail | 3,130 (45.0%) | 2,833 (44.7%) | 169 (51.2%) | 82 (40.0%) | 46 (51.1%) |
| Frail | 760 (10.9%) | 623 (9.8%) | 77 (23.3%) | 49 (23.9%) | 11 (12.2%) |
Data are presented as number (%). BMI: body mass index.
Illicit opioids, cocaine/crack, and methamphetamine.
Figure 1. Locations where participants stayed for at least one night during the past month by self-reported living situation, people with HIV engaged in clinical care at 6 sites across the United States, 2019–2024 (n=6,961).
Data are the percent of individuals within housing status groups that selected each response option to the question: “In the past month, please check any of the following places in which you stayed at least one night (excluding vacation)?” Individuals could select more than one response option. Responses presented as “Other” included “Residential program (e.g. drug treatment facility, halfway house)”, “Hospital/Long-term care facility”, “Jail/Prison”, and “Don’t know/Other”.
The unadjusted prevalence of frailty was 10.9% (n=760) overall, ranging from 9.8% (n=623) among PWH with stable housing, 23% (n=77) with unstable housing, 24% (n=49) experiencing homelessness, and 12% (n=11) who responded “don’t know” regarding their housing situation (Table 1). While the prevalence of frailty was similar for those reporting unstable housing and homelessness, prefrailty was more common among PWH with unstable housing (51%, n=169) than those experiencing homelessness (40%, n=82). The prevalence of frailty increased with age among PWH with stable housing, ranging from 5.0% (n=70) under age 40 to 12.7% (n=235) over age 60 (Figure 2). In comparison, PWH experiencing homelessness or unstable housing had a higher prevalence of frailty across all ages, with the highest levels among PWH aged 50 to 60 for both groups.
Figure 2. Frailty status by age and self-reported living situation during the past month, people with HIV engaged in clinical care at 6 sites across the United States, 2019–2024 (n=6,961).
Data are the percent of individuals characterized as not frail (0 frailty phenotype components), prefrail (1–2 components), and frail (3 or 4 components) stratified by age (<40, 40–50, 50–60, ≥60 years) and self-perceived housing status (stable, unstable, homeless, don’t know). The four frailty phenotype components included fatigue, weight loss, immobility, and inactivity.
In the primary, fully adjusted analysis, unstable housing was associated with a 2.41 times higher prevalence of frailty (95% CI: 1.95, 2.97) and experiencing homelessness was associated with a 2.05 times higher prevalence of frailty (95% CI: 1.56, 2.69) compared to stable housing (Table 2: Model 4). There was no association with frailty for participants who responded “don’t know” regarding their living situation (PR=1.36, 95% CI: 0.78, 2.38). In secondary analyses of discrete frailty levels, unstable housing had the strongest association with the outcome of frail versus robust (excluding prefrail; PR=2.59, 95% CI: 2.15, 3.11) while experiencing homelessness had the strongest association with frail versus prefrail (excluding robust; PR=1.93, 95% CI: 1.50, 2.48). Trends were similar but with larger point estimates in sensitivity analyses only adjusted for sociodemographic characteristics (Model 1); adjusted for sociodemographic characteristics and substance use (Model 2); and adjusted for sociodemographic characteristics, substance use, and HIV clinical indicators, but not comorbidities (Model 3).
Table 2.
Association between housing instability and frailty status among people with HIV engaged in clinical care at 6 sites across the United States, 2019–2024.
| Prevalence ratio (95% CI) |
|||||
|---|---|---|---|---|---|
| Outcome | N | Model 1: sociodemographic characteristics | Model 2: M1 factors + substance use | Model 3: M2 factors + HIV clinical indicators | Model 4 (MAIN): M3 factors + comorbidities |
|
| |||||
| Frail vs. not frail | 6,961 | ||||
| Housing status Stable (ref.) | |||||
| Unstable | 2.83 (2.30, 3.48) | 2.57 (2.08, 3.18) | 2.50 (2.02, 3.09) | 2.41 (1.95, 2.97) | |
| Homeless | 2.59 (2.01,3.34) | 2.22 (1.70, 2.90) | 2.09 (1.59, 2.75) | 2.05 (1.56, 2.69) | |
| Don’t know | 1.41 (0.80, 2.50) | 1.39 (0.78, 2.46) | 1.34 (0.76, 2.38) | 1.36 (0.78, 2.38) | |
| Frail vs. robust | 3,831 | ||||
| Housing status Stable (ref.) | |||||
| Unstable | 3.28 (2.78, 3.86) | 2.88 (2.44, 3.41) | 2.80 (2.36, 3.32) | 2.59 (2.15, 3.11) | |
| Homeless | 2.31 (1.83, 2.91) | 1.95 (1.53, 2.49) | 1.83 (1.42, 2.34) | 1.74 (1.35, 2.24) | |
| Don’t know | 1.75 (1.03, 2.97) | 1.68 (1.00, 2.84) | 1.68 (1.00, 2.81) | 1.61 (0.98, 2.64) | |
| Frail vs. prefrail | 3,890 | ||||
| Housing status Stable (ref.) | |||||
| Unstable | 1.95 (1.60, 2.38) | 1.86 (1.52, 2.28) | 1.83 (1.49, 2.24) | 1.79 (1.47, 2.19) | |
| Homeless | 2.22 (1.76, 2.81) | 2.03 (1.59, 2.60) | 1.96 (1.52, 2.52) | 1.93 (1.50, 2.48) | |
| Don’t know | 1.17 (0.67, 2.03) | 1.18 (0.68, 2.04) | 1.14 (0.66, 1.98) | 1.14 (0.66, 1.95) | |
| Prefrail vs. robust | 6,201 | ||||
| Housing status Stable (ref.) | |||||
| Unstable | 1.41 (1.29, 1.54) | 1.36 (1.24, 1.48) | 1.34 (1.23, 1.47) | 1.34 (1.23, 1.47) | |
| Homeless | 1.07 (0.92, 1.25) | 1.00 (0.86, 1.17) | 0.99 (0.85, 1.15) | 1.00 (0.86, 1.16) | |
| Don’t know | 1.18 (0.98, 1.42) | 1.16 (0.96, 1.40) | 1.15 (0.95, 1.38) | 1.16 (0.97, 1.40) | |
Data are adjusted prevalence ratios and 95% confidence intervals (CI) estimated from generalized linear models with robust standard errors. Model 1 (M1) adjusted for site, age, sex, race and ethnicity, and year of outcome assessment. Model 2 (M2) adjusted for all covariates in Model 1 in addition to alcohol use, smoking/vaping, and drug use (illicit opioids, cocaine/crack, methamphetamine/crystal). Model 3 (M3) adjusted for all covariates in Model 2 in addition to HIV viral suppression (<50 copies/ml) and low CD4 count (<350 cells/mm3). Model 4, the primary analysis, adjusted for all covariates in Model 3 in addition to diabetes, hypertension, hepatitis B and C, and body mass index.
In stratified heterogeneity analyses, we observed stronger associations between housing instability (combined unstable/homeless vs. stable, excluding “don’t know”) and frailty among PWH of younger ages compared to older adults (Figure 3 and Supplementary Table 2). In addition, we observed a stronger association among virally suppressed PWH (PR=2.47, 95% CI: 2.01, 3.03) compared to unsuppressed PWH (PR=1.81, 95% CI: 1.26, 2.60) due to a lower prevalence of frailty in stably housed, virally suppressed PWH (9.5%, n=533) than stably housed, unsuppressed PWH (12.3%, n=90), while 23–24% of unstably housed PWH were frail in both viral load strata. Results were similar in a sensitivity analysis limited to the 80% of PWH with housing and frailty measured on the same day (Supplementary Figure 2).
Figure 3. Stratified prevalence and adjusted prevalence ratios of frailty associated with housing instability compared to stable housing, people with HIV engaged in clinical care at 6 sites across the United States, 2019–2024 (n=6,871).
Left panel: Unadjusted prevalence of frailty within strata of housing status and sociodemographic/clinical characteristics. Right panel: Frailty prevalence ratio and 95% confidence interval estimated from generalized linear models comparing unstably housed people (combined “homeless” or “unstable” living situation in the past month, as per self-report) to those with “stable” housing. This analysis excluded individuals who responded “don’t know” regarding their housing situation (n=90). Models adjusted for site, age, sex, race and ethnicity, year of outcome assessment, alcohol, smoking, illegal substance use, HIV viral load, CD4 count, diabetes, hypertension, hepatitis B and C, and body mass index.
Discussion
In a large cohort of PWH engaged in care at six clinical centers across the US, we found that housing instability and frailty were highly prevalent and strongly associated conditions. More than one in five PWH experiencing homelessness or unstable housing were frail, compared to one in 10 PWH with stable housing. In adjusted analyses, experiencing homelessness or unstable housing were both associated with over twice the prevalence of frailty. These associations were even stronger among younger and virally suppressed PWH. This research provides an important contribution to the aging literature as one of the first studies to examine the associations between housing instability and aging-related health outcomes among PWH.
PWH disproportionately experience housing challenges: recent national estimates from the US Centers for Disease Control and Prevention (CDC) found that 9.3% of PWH experienced homelessness in the past 12 months, while 17.9% experienced housing instability more broadly [9]. Factors including HIV-related stigma, low socioeconomic status, and poor physical and mental health may contribute to housing instability among PWH [29]. Many PWH are also members of historically disadvantaged populations that bear the brunt of the national housing crisis: for example, 70% of PWH in the US identify as Black/African American, Hispanic/Latino, or multiracial [37], groups that have endured discriminatory housing practices [38,39] and are increasingly more likely to be housing cost-burdened [40,41].
Housing instability threatens public health efforts including the federal “Ending the HIV Epidemic in the US” (EHE) initiative, which seeks to reduce the number of new HIV infections by at least 90 percent by 2030 [42]. Housing challenges can lead to HIV risk-associated coping behaviors and can also make it more difficult for individuals to access healthcare [43,44]. Housing stress and loss of housing have been associated with increases in substance use [45] and sexual outcomes that heighten the risk of HIV transmission and acquisition [46]. In addition, many studies have shown that housing instability disrupts the HIV “treatment cascade”—including testing, engagement and retention in care, adherence to ART, and achievement of viral suppression—which prevents the onward transmission of HIV [26–28].
Aging deepens the complexity of care amidst the intersecting syndemics of HIV and housing instability. In the era of ART, over 50% of the 1.1 million people diagnosed with HIV in the US are now aged 50 years or older [37]. People over the age of 50 also accounted for 15.5% of new HIV diagnoses in 2022 [37]. At the same time, housing instability is an emerging crisis among older adults, who are the fastest growing group of unstably housed people [2]. Older adults are especially vulnerable to housing instability due to factors such as fixed incomes, declining health, and loss of social support networks [7]. Furthermore, HIV infection and housing instability are both associated with earlier onset of aging-related conditions [3,12,13]. Frailty, which is associated with increased risk of falls, cognitive decline, hospitalizations, and mortality [14], is common among both PWH and people experiencing housing instability, but its etiology is not fully understood for either group [16,17].
Our study adds key contributions to the limited literature on HIV, aging, and housing instability [11]. To our knowledge, only a single study has previously investigated the relationship between housing status and frailty among PWH, which found that temporary or unstable housing was associated with a higher frailty index in 389 older PWH receiving HIV treatment in Calgary, Canada [47]. Building on this work, we demonstrated that housing instability is associated with frailty in a large cohort of PWH receiving care at multiple sites across the US. We also examined heterogeneity by sociodemographic and clinical characteristics, which can help to identify the most affected groups and thereby target interventions. Our finding that housing instability is associated with frailty across ages, with especially elevated risk among younger PWH, points to the importance of including younger individuals in studies and screenings for frailty among PWH. We also found that PWH with suppressed HIV viral loads had a higher risk of frailty associated with housing instability, while the prevalence of frailty among virally unsuppressed PWH was elevated even among those with stable housing. This finding potentially indicates the importance of housing for health even among PWH who achieve viral suppression, and underscores the need for multipronged strategies, including housing support, for virally unsuppressed PWH. These dynamics should be further investigated in longitudinal studies, including within CNICS once more housing status data have accrued.
Our study also expands on previous work by evaluating multiple forms of housing instability, including self-perceived homelessness and unstable housing. Across the broader literature on housing and health outcomes for PWH, relatively few studies have evaluated different forms of housing instability. A recent meta-analysis found that 78% of 152 studies on housing and health among PWH used a binary indicator of housing status, typically comparing “homeless” to “not homeless” [28]. While homelessness presents as a highly visible and severe form of housing instability, other types of unstable housing may also negatively impact the health of PWH. For example, housing cost burden, frequent moving, crowding, or remaining in harmful relationships due to housing constraints could have direct and indirect health consequences, including difficulty adhering to ART [4]. Chronic exposure to stress has been increasingly linked to declining physical and mental health and accelerated aging, potentially leading to the development of frailty [48,49]. We found that people experiencing homelessness and unstable housing both had elevated levels of frailty, even after adjusting for sociodemographic and clinical characteristics. The estimated risk of frailty (and prefrailty) was actually higher for those who described their housing situation as “unstable” rather than “homeless”. These findings emphasize the importance of evaluating different forms of housing instability and could suggest the potential significance of self-perceived housing instability as a health-related stressor.
This analysis has important limitations. Our results may not be generalizable to PWH who have not yet been diagnosed with HIV or are not engaged in HIV clinical care. Additionally, our analysis was cross-sectional and thus should not be interpreted causally. This study design does not incorporate transitions between frailty states, including potential improvement [50]. It also does not address the likely possibility of bidirectional relationships between housing instability and frailty, whereby frailty could also contribute to housing instability through factors such as health and employment difficulties [51,52]. Not all individuals received housing and frailty assessments on the same day, however our sensitivity analysis showed little difference in results when restricting to the majority (n=5,543) who did. Lastly, our study period overlapped with the COVID-19 pandemic, which could have affected housing instability [53,54] and frailty [55] in complex ways that this study cannot address.
These limitations are balanced by key strengths including the large sample size, the collection of standardized PROs across sites that sets CNICS apart from many other clinical cohorts, the use of a validated frailty phenotype, and the assessment of different forms of housing instability. While the self-reported frailty phenotype we used does not objectively measure gait speed or grip strength, this measure has been validated in comparison to the FFP [34] and also shown to have a strong association with mortality that is consistent with the frailty literature among PWH [23,56]. Crucially, this simplified phenotype can be collected during routine clinical visits and has enabled frailty assessment at scale within CNICS. In addition, our housing status measure that differentiates between “unstable” and “homeless” advances research that has often relied on binary measurement of housing instability [28], though we note that these categories may still obscure important nuances. We also incorporated descriptive data on sleeping locations to better understand the circumstances of individuals in each housing category, however due to sample size constraints we could not further disaggregate housing instability in our frailty analyses.
In conclusion, this study illuminates the current combined burden of housing instability and frailty for PWH amidst a national housing crisis and population aging. While social factors such as housing instability are widely understood to be strong determinants of health among both PWH and the general population [57,58], little prior research has focused on the growing intersection of HIV, aging, and housing [11]. Quantifying the prevalence and association between housing instability and frailty among PWH is important for informing both future research and clinical care. In our large study of PWH in care across the US, housing instability and frailty were both common and strongly associated even among relatively younger and virally suppressed individuals. These findings suggest the potential importance of clinical screening for frailty as well as housing status and related social determinants of health for all PWH. Routine point-of-care screening for PROs as done in CNICS could offer avenues for intervention such as timely referrals to housing services or additional assessment and treatment of frailty. Future studies should evaluate longitudinal relationships between housing instability and frailty, including potential bidirectional associations, and consider additional measures of housing status and frailty along with other aging-related outcomes.
Supplementary Material
Acknowledgements
The authors acknowledge all CNICS study participants and personnel for their essential contributions to this work.
SOURCES OF SUPPORT:
CNICS is an NIH funded program [R24 AI067039] made possible by the National Institute of Allergy and Infectious Diseases (NIAID). The CFAR sites involved in CNICS include Univ of Alabama at Birmingham [P30 AI027767], Univ of Washington [P30 AI027757], Univ of California San Diego [P30 AI036214], Univ of California San Francisco [P30 AI027763] Case Western Reserve Univ [P30 AI036219], Johns Hopkins Univ [P30 AI094189, U01 DA036935], Fenway Health/Harvard [P30 AI060354], Univ of North Carolina Chapel Hill [P30 AI50410], Vanderbilt Univ [P30 AI110527], and Univ of Miami [P30 AI073961]. Additional support includes the National Institute of Drug Abuse (NIDA) [R01 DA047045 and R01 DA058938] and the National Institute on Aging [R33 AG067069]. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
CONFLICTS OF INTEREST: The authors report no potential conflicts of interest, including relevant financial interests, activities, relationships, and affiliations.
These findings were presented in part at: Conference on Retroviruses and Opportunistic Infections (CROI) 2025, San Francisco, CA. March 2025.
References
- 1.Rohe WM. Tackling the housing affordability crisis. Hous Policy Debate 2017; 27:490–494. [Google Scholar]
- 2.Culhane D, Treglia D, Metraux S, Kuhn R, Doran K, Johns E, et al. Emerging Crisis of Aged Homelessness. Actionable Intelligence for Social Policy (AISP), University of Pennsylvania; 2019. https://aisp.upenn.edu/aginghomelessness/ (accessed 2 Oct2024). [Google Scholar]
- 3.Kushel M Homelessness among older adults: an emerging crisis. Generations 2020. [Google Scholar]
- 4.Marcus R, de Groot A, Bachman S, Chisolm N, Quadri Y, Cabral H, et al. Longitudinal determinants of housing stability among people living with HIV/AIDS experiencing homelessness. Am J Public Health 2018; 108:S552–S560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Frederick TJ, Chwalek M, Hughes J, Karabanow J, Kidd S. How stable is stable? Defining and measuring housing stability. J Community Psychol 2014; 42:964–979. [Google Scholar]
- 6.Office of Disease Prevention and Health Promotion, U.S. Department of Health & Human Services. Housing Instability. https://odphp.health.gov/healthypeople/priority-areas/social-determinants-health/literature-summaries/housing-instability (accessed 2 Oct2024). [Google Scholar]
- 7.Henderson KA, Manian N, Rog DJ, Robison E, Jorge E, Al-Abdulmunem M. Addressing Homelessness Among Older Adults. Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services; 2023. [Google Scholar]
- 8.de Sousa T, Henry M. The 2024 Annual Homelessness Assessment Report (AHAR to Congress) Part 1: Point-In-Time Estimates of Homelessness, December 2024. U.S. Department of Housing and Urban Development; 2024. [Google Scholar]
- 9.Centers for Disease Control and Prevention. HIV Surveillance Special Report: Behavioral and Clinical Characteristics of Persons with Diagnosed HIV Infection Medical Monitoring Project, United States 2022 Cycle (June 2022–May 2023). U.S. Department of Health and Human Services; 2024. https://stacks.cdc.gov/view/cdc/159149 (accessed 25 Sep2024). [Google Scholar]
- 10.The Lancet Healthy Longevity. Ageing with HIV. Lancet Healthy Longev; 2022; 3:e119. [Google Scholar]
- 11.Chaminuka AS, Prybutok G, Prybutok VR, Senn WD. Systematic review: HIV, aging, and housing: a North American perspective, 2012–2023. Healthcare 2024; 12:992. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Wing EJ. HIV and aging. Int J Infect Dis 2016; 53:61–68. [DOI] [PubMed] [Google Scholar]
- 13.Brown RT, Hemati K, Riley ED, Lee CT, Ponath C, Tieu L, et al. Geriatric conditions in a population-based sample of older homeless adults. The Gerontologist 2017; 57:757–766. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Clegg A, Young J, Iliffe S, Rikkert MO, Rockwood K. Frailty in elderly people. The Lancet 2013; 381:752–762. [Google Scholar]
- 15.Mantell R, Hwang YIJ, Radford K, Perkovic S, Cullen P, Withall A. Accelerated aging in people experiencing homelessness: a rapid review of frailty prevalence and determinants. Front Public Health 2023; 11. doi: 10.3389/fpubh.2023.1086215 [DOI] [Google Scholar]
- 16.Kehler DS, Milic J, Guaraldi G, Fulop T, Falutz J. Frailty in older people living with HIV: current status and clinical management. BMC Geriatr 2022; 22:919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Levett TJ, Cresswell FV, Malik MA, Fisher M, Wright J. Systematic review of prevalence and predictors of frailty in individuals with human immunodeficiency virus. J Am Geriatr Soc 2016; 64:1006–1014. [DOI] [PubMed] [Google Scholar]
- 18.Cronin T, Healy D, McCarthy N, Smith SM, Travers J. Prevalence and risk factors of frailty in people experiencing homelessness: a systematic review and meta-analysis. J Frailty Aging 2025; 14:100029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Crane HM, Ruderman SA, Whitney BM, Nance RM, Drumright LN, Webel AR, et al. Associations between drug and alcohol use, smoking, and frailty among people with HIV across the United States in the current era of antiretroviral treatment. Drug Alcohol Depend 2022; 240:109649. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Eke UA, Mohanty K, Gruber-Baldini AL, Ryan AS. Frailty and aging in HIV: status post 13 years of national awareness. J Frailty Aging 2023; 12:49–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Piggott DA, Erlandson KM, Yarasheski KE. Frailty in HIV: epidemiology, biology, measurement, interventions, and research needs. Curr HIV/AIDS Rep 2016; 13:340–348. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kojima G, Iliffe S, Walters K. Frailty index as a predictor of mortality: a systematic review and meta-analysis. Age Ageing 2018; 47:193–200. [DOI] [PubMed] [Google Scholar]
- 23.Liu S, Yan Q, Jiang Y, Xiao M, Zhao J, Wang Y, et al. The impact of frailty on all-cause mortality in patients with HIV infection: a systematic review and meta-analysis. AIDS Res Hum Retroviruses 2022; 38:692–699. [DOI] [PubMed] [Google Scholar]
- 24.Olivieri-Mui B, Wilson I, Shi S, Montano M, McCarthy EP, Oh G, et al. Geriatric conditions associated with nonadherence to antiretroviral therapy among older people with HIV: the importance of frailty. AIDS Patient Care STDs 2022; 36:226–235. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cohen MS, Smith MK, Muessig KE, Hallett TB, Powers KA, Kashuba AD. Antiretroviral treatment of HIV-1 prevents transmission of HIV-1: where do we go from here? The Lancet 2013; 382:1515–1524. [Google Scholar]
- 26.Leaver CA, Bargh G, Dunn JR, Hwang SW. The effects of housing status on health-related outcomes in people living with HIV: a systematic review of the literature. AIDS Behav 2007; 11:85–100. [DOI] [PubMed] [Google Scholar]
- 27.Wolitski RJ, Kidder DP, Fenton KA. HIV, homelessness, and public health: critical issues and a call for increased action. AIDS Behav 2007; 11:167–171. [DOI] [PubMed] [Google Scholar]
- 28.Aidala AA, Wilson MG, Shubert V, Gogolishvili D, Globerman J, Rueda S, et al. Housing status, medical care, and health outcomes among people living with HIV/AIDS: a systematic review. Am J Public Health 2016; 106:e1–e23. [Google Scholar]
- 29.Marcus R, Tie Y, Dasgupta S, Beer L, Padilla M, Fagan J, et al. Characteristics of adults with diagnosed HIV who experienced housing instability: findings from the Centers for Disease Control and Prevention Medical Monitoring Project, United States, 2018. J Assoc Nurses AIDS Care 2022; 33:283–294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kitahata MM, Rodriguez B, Haubrich R, Boswell S, Mathews WC, Lederman MM, et al. Cohort profile: the Centers for AIDS Research Network of Integrated Clinical Systems. Int J Epidemiol 2008; 37:948–955. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Crane HM, Lober W, Webster E, Harrington RD, Crane PK, Davis TE, et al. Routine collection of patient-reported outcomes in an HIV clinic setting: the first 100 patients. Curr HIV Res 2007; 5:109–118. [DOI] [PubMed] [Google Scholar]
- 32.Fredericksen R, Crane PK, Tufano J, Ralston J, Schmidt S, Brown T, et al. Integrating a web-based, patient-administered assessment into primary care for HIV-infected adults. J AIDS HIV Res Online 2012; 4:47–55. [Google Scholar]
- 33.Fried LP, Tangen CM, Walston J, Newman AB, Hirsch C, Gottdiener J, et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci 2001; 56:M146–156. [DOI] [PubMed] [Google Scholar]
- 34.Ruderman SA, Webel AR, Willig AL, Drumright LN, Fitzpatrick AL, Odden MC, et al. Validity properties of a self-reported modified frailty phenotype among people with HIV in clinical care in the United States. J Assoc Nurses AIDS Care 2023; 34:158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Desquilbet L, Jacobson LP, Fried LP, Phair JP, Jamieson BD, Holloway M, et al. HIV-1 infection is associated with an earlier occurrence of a phenotype related to frailty. J Gerontol A Biol Sci Med Sci 2007; 62:1279–1286. [DOI] [PubMed] [Google Scholar]
- 36.Zou G A modified poisson regression approach to prospective studies with binary data. Am J Epidemiol 2004; 159:702–706. [DOI] [PubMed] [Google Scholar]
- 37.Centers for Disease Control and Prevention. HIV Surveillance Report: Diagnoses, Deaths, and Prevalence of HIV in the United States and 6 Territories and Freely Associated States, 2022. U.S. Department of Health and Human Services; 2024. https://stacks.cdc.gov/view/cdc/156509 (accessed 26 Nov2024). [Google Scholar]
- 38.White AG, Guikema SD, Logan TM. Urban population characteristics and their correlation with historic discriminatory housing practices. Appl Geogr 2021; 132:102445. [Google Scholar]
- 39.Bartlett R, Morse A, Stanton R, Wallace N. Consumer-lending discrimination in the FinTech Era. J Financ Econ 2022; 143:30–56. [Google Scholar]
- 40.Joint Center for Housing Studies of Harvard University. The State of the Nation’s Housing 2024. Harvard University; 2024. https://www.jchs.harvard.edu/state-nations-housing-2024 [Google Scholar]
- 41.Hess C, Colburn G, Crowder K, Allen R. Racial disparity in exposure to housing cost burden in the United States: 1980–2017. Hous Stud 2022; 37:1821–1841. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Centers for Disease Control and Prevention. Issue Brief: The Role of Housing in Ending the HIV Epidemic. U.S. Department of Health and Human Services; 2024. https://stacks.cdc.gov/view/cdc/127257 (accessed 25 Sep2024). [Google Scholar]
- 43.Kushel MB, Gupta R, Gee L, Haas JS. Housing instability and food insecurity as barriers to health care among low-income Americans. J Gen Intern Med 2006; 21:71–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Meltzer R, Schwartz A. Housing affordability and health: evidence from New York City. Hous Policy Debate 2016; 26:1508–1514. [Google Scholar]
- 45.Austin AE, Shiue KY, Naumann RB, Figgatt MC, Gest C, Shanahan ME. Associations of housing stress with later substance use outcomes: A systematic review. Addict Behav 2021; 123:107076. [DOI] [PubMed] [Google Scholar]
- 46.Groves AK, Niccolai LM, Keene DE, Rosenberg A, Schlesinger P, Blankenship KM. Housing instability and HIV risk: expanding our understanding of the impact of eviction and other landlord-related forced moves. AIDS Behav 2021; 25:1913–1922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.McMillan JM, Gill MJ, Power C, Fujiwara E, Hogan DB, Rubin LH. Comorbidities in older persons with controlled HIV infection: correlations with frailty index subtypes. AIDS Patient Care STDs 2020; 34:284–294. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Gruenewald TL, Seeman TE, Karlamangla AS, Sarkisian CA. Allostatic load and frailty in older adults. J Am Geriatr Soc 2009; 57:1525–1531. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Mooney CJ, Elliot AJ, Douthit KZ, Marquis A, Seplaki CL. Perceived control mediates effects of socioeconomic status and chronic stress on physical frailty: findings from the Health and Retirement Study. J Gerontol Ser B 2018; 73:1175–1184. [Google Scholar]
- 50.Kojima G, Taniguchi Y, Iliffe S, Jivraj S, Walters K. Transitions between frailty states among community-dwelling older people: a systematic review and meta-analysis. Ageing Res Rev 2019; 50:81–88. [DOI] [PubMed] [Google Scholar]
- 51.Palmer KT, D’Angelo S, Harris EC, Linaker C, Gale CR, Evandrou M, et al. Frailty, pre-frailty and employment outcomes in the Health and Employment after Fifty (HEAF) study. Occup Environ Med 2017; 74:476–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 52.Scheckler S, Molinsky J, Herbert C, Arenas J, Guytingco K, Soroya S. Pathways into and out of Housing Insecurity and Homelessness: Relationships Between Age, Public Program Use, and Housing Stability. Center for Financial Security, University of Wisconsin; 2023. https://rdrc.wisc.edu/publications/working-paper/wi23-14 [Google Scholar]
- 53.Alvarez T, Steffen B. Worst Case Housing Needs: 2023 Report to Congress. Office of Policy Development and Research, U.S. Department of Housing and Urban Development; 2023. https://www.huduser.gov/portal/publications/Worst-Case-Housing-Needs-2023.html [Google Scholar]
- 54.Hepburn P, Haas J, Graetz N, Louis R, Rutan DQ, Alexander AK, et al. Protecting the most vulnerable: policy response and eviction filing patterns during the COVID-19 pandemic. RSF Russell Sage Found J Soc Sci 2023; 9:186–207. [Google Scholar]
- 55.König M, Gollasch M, Komleva Y. Frailty after COVID-19: the wave after? Aging Med 2023; 6:307–316. [Google Scholar]
- 56.Ruderman SA, Fahey CA, Kyle RP. Frailty and all-cause mortality among people with HIV engaged in clinical care in the United States. 2025. [Google Scholar]
- 57.Braveman P, Gottlieb L. The Social Determinants of Health: It’s Time to Consider the Causes of the Causes. Public Health Rep 2014; 129:19–31. [Google Scholar]
- 58.Sprague C, Simon SE. Ending HIV in the USA: integrating social determinants of health. The Lancet 2021; 398:742–743. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.