In this study, we incorporated standardized assessment of frailty into a community cohort of HIV-infected and epidemiologically-comparable HIV-uninfected IDUs. We identified a frailty prevalence of 12.3% and found that HIV infection, particularly advanced disease stage with lower CD4 cell counts and the absence of ART or virological suppression, was strongly associated with frailty. Despite our cohort being relatively young compared to geriatric populations where frailty has been shown to presage adverse clinical outcomes, frailty was independently associated with increased mortality risk in prospective analysis even after accounting for sociodemographic variables, comorbidity, and HIV infection. Moreover, the combined effect of frailty and HIV on mortality appeared to exceed what one would expect from the additive effect of the individual exposures. In summary, these findings suggest that frailty is a useful phenotype for investigating aging among HIV-infected IDUs and could potentially identify individuals at high-risk for adverse outcomes among these highly vulnerable groups. Our data also raise the possibility that optimal HIV care with virological suppression could attenuate the development of frailty.
Aberrant inflammation and immune dysregulation have been hypothesized to characterize the natural process of aging as well as underlie the pathogenesis of chronic HIV disease 
. Based on emerging epidemiologic, clinical, and mechanistic evidence, inflammation and immune dysfunction are postulated to play a central role in frailty pathophysiology among older HIV-uninfected adults 
. We find that HIV-infected IDUs with well-controlled HIV disease are no more likely to be frail than similar individuals without HIV infection. In contrast, the likelihood of frailty was significantly higher with advanced HIV disease with inadequate virologic control. These results suggest that HIV infection without effective treatment may represent a significant, modifiable risk factor for frailty. Thus, together with data from other HIV cohorts 
, these findings suggest a putative role for HAART in arresting the progression to frailty. HAART has been shown to have a significant impact on morbidity and mortality for HIV-infected populations in general and HIV-infected IDUs specifically 
. The negative impact of late initiation of care and premature interruption of HAART on survival has also been well defined 
. Engagement in care and adherence to antiretroviral regimens continue to be a daunting challenge for the HIV-infected population 
. These effects are exacerbated among IDUs, who tend to have later diagnosis, poorer access to care, lower HAART uptake, more limited adherence, and frequent treatment interruptions with significant consequences for HIV-related morbidity and mortality 
. With less successful navigation of the HIV care continuum, HIV-infected IDUs remain at higher risk for HIV/AIDS progression, and by extension may suffer increased progression to frailty with its consequent adverse outcomes. Future investigations will need to elucidate the underlying mechanisms of frailty development in the setting of HIV and determine whether earlier ART may be effective for frailty prevention.
We found that frailty was independently associated with an increased risk of death among HIV-infected and at risk IDUs. These findings are particularly notable given that the median age of this cohort was 48 years. Frailty has been primarily linked to mortality and other adverse outcomes in significantly older cohorts (predominantly 65 years of age and older) 
. Consistent with our data, a frailty related construct has demonstrated increased mortality risk in younger populations in 2 recent studies 
. As comprehensive treatment for HIV infection evolves beyond a focus primarily on viral suppression and broadens to consider management of multiple chronic conditions to achieve healthy aging, assessment tools beyond HIV RNA and CD4 counts will be increasingly needed. Frailty may improve risk stratification and inform appropriate clinical management for aging, complex care patients living with HIV. Further, the joint impact of frailty and HIV infection on mortality suggests that efforts to reduce mortality risk through both frailty and HIV targeted interventions may translate into significant survival benefit.
Besides HIV disease markers, we found that depressive symptoms and prescription drug abuse were associated with both frailty and prefrailty. Previously, we have documented self-medication with ‘street’ drugs for symptoms in this population 
. Although directionality of these associations cannot be determined in the current analysis, it will be important to examine whether frailty interventions alleviate associated pain or other symptoms and lead to improved health-related quality of life.
Consistent with studies of frailty in HIV-uninfected older adults 
, we observed that advancing age, female gender, lower educational attainment, and increased number of comorbid conditions were associated with frailty within this IDU population. The absence of a cohabiting life partner may be a putative measure of social isolation with which frailty has also been previously associated 
. As with HIV, dysregulated inflammation has been associated with age and socioeconomic status 
. Further, hormonal influences have been postulated to play an important role in age-associated changes in inflammation among women 
. Whether hormonally mediated changes in inflammation account for sex-specific differences in frailty prevalence remains to be determined. However, inflammation may be a predominant pathway by which these factors could contribute to progression to a frail state.
Prefrailty is considered to be an intermediate state that presages progression to frailty in geriatric populations 
. Therefore, the prefrail state may provide a window of opportunity when interventions could mitigate adverse clinical outcomes. We observed consistent, although attenuated, associations of prefrailty with HIV disease and other frailty risk factors; however, prefrailty was not predictive of mortality. A larger study size or longer duration may allow identification of incremental mortality risk with prefrailty. Further follow-up also is needed to define the likelihood of transition from prefrailty to frailty among HIV-infected persons. These data will be vital to inform development of interventions to reverse or slow progression to frailty. The high prevalence of almost two-thirds of our participants with prefrailty is consistent with other studies and provides a large target population for intervention 
Our study had several limitations. Debate persists on the optimal criteria for defining frailty 
. The frailty phenotype employed in this study closely approximated the original Fried criteria, with objective measurement of weight loss, gait speed, and grip strength. Consistent with prior studies, we substituted a self-reported measure of low physical activity 
. Our weight loss parameter did not discern intentionality; however, similarly constructed frailty constructs had predictive validity roughly equivalent to the original Fried phenotype in elderly HIV-uninfected persons 
. As we used weight loss of ≥5% since last study visit (median of 6 months), our threshold for meeting this criteria required greater weight loss than the original criteria. Given the observational nature of the study and lack of temporality for HIV-frailty associations, caution is needed regarding inferences of causality. Our cohort is a predominantly African American, urban IDU cohort and as such, our findings may not be fully generalizable to other HIV-infected populations. However, significant relationships between frailty (and a frailty-related phenotype) and advanced HIV disease have been noted in several non-IDU cohorts 
. Further, we have observed a similar relationship between frailty and non-HIV-related factors in our population as reported from older HIV-uninfected populations 
. Whether similar biological mechanisms underlie the development of frailty for these different groups needs to be further investigated. Nevertheless, this population does represent those individuals particularly vulnerable to disparities in access to care and key adverse health care outcomes for whom appropriately targeted frailty interventions could have substantial clinical impact.
Improved understanding of frailty in high-risk populations may translate into clinical utility as well as strengthen our scientific understanding of the aging process. Despite improving survival and recent aging trends, HIV-infected and at risk IDUs continue to experience marked socioeconomic challenges with persistent disparities in treatment access, morbidity and mortality outcomes 
. Frailty assessment may prove useful in identifying those persons at greatest risk for premature death and allow appropriate intervention. Whether HIV infection and frailty share a single common pathway to premature death remains to be determined. However, elucidation of the mechanisms underlying frailty development may provide substantial opportunities for realizing the healthy aging of HIV-infected and drug using populations, with potential additional benefit to the general aging population.