Background
Living with HIV infection and experiencing housing instability are both independently associated with early onset of aging-related comorbidities including frailty. However, few studies have examined the relationship between housing status and frailty among people with HIV (PWH).
Methods
We analyzed data from PWH engaged in care at 6 sites across the US within the Centers for AIDS Research Network of Integrated Clinical Systems (CNICS) between 2019-2024. Data were collected during routine clinical visits, including self-administered surveys. We measured frailty using a previously validated approach based on having at least 3 of 4 components (fatigue, weight loss, immobility, inactivity) vs. less than 3 components. Housing status was based on self-reported living situation in the last month, comparing PWH reporting “Homeless” or “Unstable” vs. “Stable.” We estimated the association between most recent housing status and frailty with prevalence ratios (PR) from generalized linear models using complete-case data. Primary models adjusted for site, demographic characteristics, year, HIV clinical indicators (viral load, CD4 count), and substance use (alcohol, illegal substances). Sensitivity analyses removed potential mediators from the adjustment set (HIV clinical indicators and substance use).
Results
This analysis included 6,587 PWH (84% male, 16% female; 87% virally suppressed) with a median age of 52 years (IQR: 40-60). Overall, 8% of PWH (n=505) had unstable housing (3% homeless, 5% unstable), and 11% (n=700) were frail. In the primary analysis, housing instability was associated with 2.2 times higher prevalence of frailty (95% CI: 1.8, 2.6) compared to stable housing. This association was similar in models not adjusted for substance use (PR=2.4, 95% CI: 1.9, 2.8) or HIV clinical indicators (PR=2.5, 95% CI: 2.1, 3.0). In stratified analyses, we observed a stronger association among virally suppressed PWH (PR=2.5, 95% CI: 2.1, 3.1) compared to unsuppressed (PR=1.5, 95% CI: 1.1, 2.2; interaction p=0.04) due to a lower prevalence of frailty in stably housed, virally suppressed PWH (9%) than stably housed, unsuppressed PWH (13%), while 22% of unstably housed PWH were frail in both viral load strata (Table).
Conclusions
Housing instability and frailty were both prevalent and strongly associated in PWH, including among younger and virally suppressed groups. These findings highlight the importance of social determinants of health, such as housing status, for clinical outcomes among all PWH.
Table or Figure
