Abstract
We evaluated the relationship between neighborhood disadvantage (measured by the Area Deprivation Index [ADI]) and frailty. We performed a secondary analysis, pooling cross-sectional data collected from 209 people with HIV (PWH) aged ≥50 years enrolled in studies in Colorado (CO) and Missouri (MO). MO participants (N = 137) had a higher ADI (µ= 70, ơ2 = 25) compared to CO (µ= 32, ơ2 = 15; p < .001). No significant differences in ADI were observed between frailty categories when cohorts were examined either separately or combined; however, when comparing individual frailty criteria, the most apparent differences by neighborhood disadvantage were seen among those with limited physical activity (μ = 67, ơ2 = 28) compared to those without (μ = 55, ơ2 = 29, p = .03). Neighborhood disadvantage was associated with low physical activity but not with overall frailty status. Future research should examine how access to physical activity spaces varies based on ADI, as this could be crucial in preventing frailty.
• We investigated the relationship between neighborhood disadvantage and frailty among two demographically and geographically distinct cohorts of older PWH. • Neighborhood disadvantage was associated with low physical activity among older PWH but not with overall frailty status. • This study provides some insights to generate hypotheses regarding the association between SDOH and frailty among older PWH.
• ADI may be an important factor when assessing physical health and physical activity. • Future studies considering differences in access to spaces for physical activity may be a key factor in frailty prevention as it correlates with ADI.What this paper adds
Applications of study findings
The proportion of older adults living with HIV has drastically shifted over the past three decades of the AIDS epidemic. In 2018, 51% of the people with HIV (PWH) in the US were aged 50 and older (Centers for Disease Control and Prevention [CDC], 2022a). Within the next decade, this percentage will increase to 73% (Smit et al., 2015). While the development of effective antiretroviral therapy (ART) has led to markedly improved life expectancy, older adults may experience unique challenges in HIV and general care (Iriarte et al., 2021).
Individuals with HIV appear to have an accentuated aging process, with ongoing immune activation and chronic inflammation that may contribute to higher rates of multimorbidity and frailty (Erlandson, 2020; Erlandson et al., 2017; Falutz et al., 2021; Strain et al., 2022). Frailty is a syndrome characterized by decreased physiological reserve and increased vulnerability to stressors, predisposing an individual to adverse health outcomes (Fried et al., 2001; Piggott et al., 2016). Within the fields of geriatrics and HIV, one of the most commonly used ways to define frailty is with the Fried Frailty Phenotype (FFP), which includes five specific parameters (weight loss, weakness, slowness, exhaustion, and low physical activity) (Fried et al., 2001; Kehler et al., 2022). Frailty is common and can present earlier in PWH (Falutz et al., 2021; Iriarte et al., 2021). Among this group, the pooled prevalence of frailty is 10.9% for those aged 50 and older, and its onset is nearly ten years earlier compared to individuals without HIV (Desquilbet et al., 2007; Yamada et al., 2022). More importantly, frailty in PWH is associated with an increased risk of physical and cognitive decline, falls, increased use of health care services and hospitalizations, poor quality of life, and increased mortality (Falutz et al., 2021; Iriarte et al., 2021).
Both frailty and HIV disproportionally impact individuals of lower socioeconomic status (Erlandson & Piggott, 2021; Iriarte et al., 2022). Divergent experiences on social determinants of health (SDOH) interfere with health equity, that is, just access to attaining the highest level of health (CDC, 2022b). SDOH are conditions in the environments where people are born, live, learn, work, play, worship, and age that affect a wide range of health, functioning, and quality-of-life outcomes and risks (Healthy People 2030, 2023). These SDOH significantly impact people’s health and include conditions such as safe housing, transportation, neighborhoods, education, job opportunities, and income (Healthy People 2030, 2023). The National Health and Aging Trends Study, a nationally representative sample of adults 65 years and older in the US, showed that frailty prevalence is significantly higher among Latino/Hispanic and African American/Black populations compared to non-Hispanic White (Bandeen-Roche et al., 2015). This frailty disparity tends to persist across income quartiles independent of the number of comorbidities (Usher et al., 2021). Additionally, frailty also has been associated with greater socioeconomic challenges, including low income, unemployment, occupation, and low educational level in both populations with and without HIV (Bandeen-Roche et al., 2015; Dugravot et al., 2020; Iriarte et al., 2021, 2023; Piggott et al., 2016; Tan et al., 2022). A longitudinal study conducted with PWH showed that lower socioeconomic status (SES) is associated with greater likelihood of frailty and a lower likelihood of frailty recovery (Piggott et al., 2020), resulting in a double hit of lower physiologic resilience stress among those with potentially simultaneous heightened stress exposure (Erlandson & Piggott, 2021).
One measure that captures area-level SDOH is the area deprivation index (ADI). The ADI is a validated composite measure of neighborhood SES disadvantage that includes four domains: income, housing, employment, and education (Center for Health Disparities Research, University of Wisconsin, 2023; Kind & Buckingham, 2018). Previous studies conducted with older populations have revealed disparities in health outcomes such as chronic disease care, COVID-19 prevalence/mortality, rehospitalizations, and Alzheimer’s disease as a function of ADI (Hu et al., 2021; Kind et al., 2014; Kitchen et al., 2021; Powell et al., 2020). However, evidence of the association between area-level deprivation on PWH is limited. One study conducted by Edmonds et al. (2021) reported that ADI was associated with the cumulative incidence of death over four years among women with HIV in the US. Another study indicated the association between ADI and brain aging in PWH alongside traditional HIV metrics such as viral load and CD4 cell count (Petersen et al., 2023). Conversely, another study found that living in socioeconomically disadvantaged neighborhoods (higher ADI) was associated with a lower risk of HIV/AIDS among US Veterans (Oluyomi et al., 2023).
Examining geographic neighborhood-level factors provides a critical framework for implementing HIV care. While SDOH are relevant for HIV care and frailty prevention (Erlandson & Piggott, 2021), the specific impact of neighborhood disadvantage (i.e., low SES) on frailty remains poorly understood. To address these critical knowledge gaps, we examined the relationship between neighborhood disadvantages, as measured by ADI, and frailty among two demographically and geographically distinct cohorts of older PWH. It was hypothesized that frail older PWH reports would be associated with a higher ADI (i.e., greater disadvantage).
Methods
Study Design and Sample
This is a cross-sectional study among older PWH enrolled in two cohorts from the Denver, Colorado (CO) and St. Louis, Missouri (MO) metropolitan areas. CO and MO were selected due to their differing demographic compositions and geographic settings. In CO, out of a total of 13,442 PWH, the HIV prevalence is reported as 87% male and 13% female. The racial distribution among PWH in CO is as follows: 54.9% White, 25.2% Latino, and 15.2% Black/African American. Regarding healthcare metrics, 84.5% of PWH are linked to care, 65.9% receive care, and 60.6% achieve viral suppression (AIDSVu, 2024a). In MO, out of a total of 13,103 PWH, the prevalence of HIV is 81.6% male and 18.4% female. The racial distribution among PWH in MO is: 45.1% White, 6.9% Latino, and 42.9% Black/African American. In terms of healthcare metrics, 77.8% of PWH are linked to care, 77.1% receive care, and 66.8% are virally suppressed (AIDSVu, 2024b).
The CO cohort included PWH enrolled in the Exercise for Healthy Aging Study (Erlandson et al., 2018) and from the High-Intensity Exercise Study to Attenuate Limitations and Train Habits in Older Adults With HIV (Oliveira et al., 2022). The MO cohort included PWH enrolled in the Frailty and Brain Integrity in Older HIV-infected Individuals (Paul et al., 2018) and from the Exercise Training to Improve Brain Health in Older HIV + Individuals (Cooley et al., 2023) studies. All participants in both cohorts were aged ≥50 years, sedentary, on antiretroviral therapy, and had a primary address in CO or MO. Participants signed a written informed consent before enrollment, and the studies were approved by the local institutional review board at each site.
Additional exclusion criteria for MO participants included a current or past history of confounding neurological disorders (e.g., stroke), untreated severe depression, head injury with loss of consciousness >30 min, < 8 years of education, and a positive result on a urine drug screen for a substance other than alcohol, tobacco, or marijuana. Exclusion criteria for CO participants have been previously published but included major contraindications to enrollment in an exercise study, such as severe mobility limitation, unstable angina, supplemental oxygen requirement, and uncontrolled hypertension, among others (Erlandson et al., 2018; Oliveira et al., 2022).
Study Variables
Frailty
We used the FFP to assess frailty. We employed a previously established population-independent cutoff to determine frailty status, including non-frail (scored as 0), pre-frail (scored 1–2), and frail (scored 3–5). Frail status was determined when three or more of the following Fried frailty criteria were present: weakness, slowness, weight loss, low activity, and exhaustion. Weakness was assessed by the average of 3 dominant hand grip strength measurements and defined by applying previously defined sex and body mass index (BMI) cutoffs (Fried et al., 2001). Slowness was defined by the average of 2 readings on a 4-m walk: men ≤173 cm and women ≤159 cm in height who required ≥6.22 seconds or men >173 cm and women >159 cm in height who required ≥5.33 seconds to complete the walk met the criterion for slowness; gait speed was dichotomized at ≤1 or >1 m/sec. Weight loss was defined by a self-report of an unintentional weight loss of ≥10 pounds during the past year. Low activity was measured as being “limited a lot” in response to the Short Form (SF)–36 question, “Does your health limit you in vigorous activities such as running, lifting heavy objects, or participating in strenuous sports?” Exhaustion was defined as experiencing at least 3–4 times per week the feeling that “everything I do is an effort” or “sometimes I just cannot get going.” The summary frailty score is a count of items for which the respondent scored “1” (range 0–5) (Fried et al., 2001). For the current analyses, pre-frail and frail groups were combined into one group.
Neighborhood Disadvantage
To evaluate neighborhood disadvantage, we utilized the ADI built from 17 socioeconomic indicators (income, education, employment, and housing quality) available in the US Census 2015 (Center for Health Disparities Research, University of Wisconsin, 2023; Kind & Buckingham, 2018). The ADI allows for rankings of neighborhoods by socioeconomic disadvantage in a region of interest (e.g., at the state or national level) and is generated for block groups corresponding to participants’ home addresses. Each neighborhood is defined as a Census block group. In this study, ADI is presented in national percentile rankings from 1 (least disadvantaged) to 100 (most disadvantaged).
Data Analysis
Demographics, clinical characteristics (e.g., duration of HIV, current CD4 t-cell count, detectable viral load), health characteristics (e.g., Montreal Cognitive Assessment [MoCA] (MoCA cognition, 2023)), proportion of frail individuals, and national ADI were first compared between CO and MO participants using chi-square or Mann–Whitney U tests. General linear models were used to assess the differences in national ADI by frailty group (non-frail vs. pre-frail/frail). Due to potential cohort differences, this analysis was first conducted separately within each cohort and subsequently within both cohorts combined. Race/ethnicity and sex were included as covariates in individual cohort analyses, while race/ethnicity, sex, and cohort were included in the combined analysis. National ADI was also compared between individuals who did or did not meet each individual frailty criteria (weakness, slowness, weight loss, low physical activity, and exhaustion) using general linear models with race/ethnicity, sex, and cohort included as covariates.
Results
Demographic and Clinical Characteristics of the CO and MO Cohorts (n = 209).
Note. AA = African American; ADI = area deprivation index; MoCA = Montreal Cognitive Assessment.
aADI scores are provided in national percentile rankings at the block group level from 1 to 100 and in deciles from 1 to 10 for each state. A block group with a ranking of 1 indicates the lowest level of “disadvantage” within the nation or state, and an ADI with a ranking of 100 or 10 indicates the highest level of “disadvantage” for the nation or state, respectively.
National ADI by Frailty Status
When examining cohorts separately, national ADI did not differ by frailty status. In the CO cohort, pre-frail/frail PWH reported a slightly higher national ADI compared to non-frail, while the opposite pattern was observed within the MO cohort. However, neither reached the threshold for significance (p-values >.05, CO η2 = 0.01, MO η2 = 0.004). Similarly, there was no significant difference between frailty categories when the two cohorts were combined (p = .82, η2 = 0.00) (Figure 1). Comparison of the Differences in National ADI Scores by Frailty Status. Note. Frailty was assessed using the Fried criteria and classified as non-frail (0 criteria met) or pre-frail/frail (1–5 criteria). Black diamond represents mean ADI National Rank for each location. Neither reached the threshold for significance. CO cohort: F = 1.2, p = .27. MO cohort: F = 0.88, p = .35. Race/ethnicity and sex were included as covariates in individual cohort analyses. ADI = area deprivation index.
National ADI by Frailty Criteria
Comparison of the Differences in National ADI Scores by Frailty Components in the Combined Group.
Note. Number of participants meeting each criteria: Weight loss n = 18; Weakness n = 53; Slowness n = 43; Exhaustion n = 79; Low activity n = 51. ADI = area deprivation index.
Discussion
To our knowledge, this is among the first studies to examine the relationship between neighborhood disadvantages (ADI) and frailty among two demographically and geographically distinct cohorts of older PWH. The main contribution of this study is to provide insights regarding the association between SDOH and frailty among older PWH. There is limited research on this topic, especially in the context of older PWH. Findings from previous studies on the relationship between ADI and frailty suggest that this is an emerging area of inquiry (Brothers & Rockwood, 2019; Fritz et al., 2020).
Although we did not find a significant relationship between overall frailty status and ADI, socio-ecological approaches from HIV risk and prevention research may be informative in trying to understand frailty in people aging with HIV (Brothers & Rockwood, 2019). As an indicator of area-based socioeconomic disadvantage, the ADI is rooted in four factors: income, housing, employment, and education (Center for Health Disparities Research, University of Wisconsin, 2023; Kind & Buckingham, 2018). These modifiable factors may interact to increase or decrease the chances of harmful outcomes (Rhodes, 2002). For instance, neighborhood disadvantages such as walkability and safety may determine physical activity levels and social engagement and, therefore, mediate the pathway between frailty and other adverse health outcomes such as falls and death (Brothers & Rockwood, 2019; Rhodes, 2002). This example is consistent with the Edmonds et al. study (2021), wherein ADI was associated with mortality among women with HIV in the U.S. These findings suggest that place-based characteristics should be considered along with individual-level factors, which can enrich the scientific understanding of how neighborhood characteristics relate to frailty status among older PWH.
Our finding of greater ADI among PWH who reported low physical activity due to health limitations is consistent with recent research within the general aging population demonstrating that low SES is associated with lower self-reported physical activity, and both measures strongly predicted increased longitudinal frailty risk (Kheifets et al., 2022). Although further research is needed to fully examine the specific aspects of disadvantaged neighborhoods that most directly contribute to low physical activity (e.g., personal safety and lack of open space), factors such as lower income, reduced access to healthy or plentiful food, or access to and the quality of medical care may impact physical health (French et al., 2019; Kurani et al., 2021). Older PWH who live in a more disadvantaged neighborhood may represent a cohort of individuals who could benefit from additional monitoring of physical activity and health.
Considering the differences between CO and MO populations, several factors could potentially contribute to the observed variations in physical activity levels and frailty among older PWH. Consistent with the HIV prevalence reported in these states (AIDSVu, 2024a; 2024b), we found the racial distribution differs significantly between the two states, with a higher percentage of Black/African American individuals in MO compared to CO. Racial disparities in healthcare access, socioeconomic status, and neighborhood environments can impact health outcomes such as physical activity levels and frailty among PWH (Bandeen-Roche et al., 2015; Dugravot et al., 2020; Iriarte et al., 2021, 2023; Piggott et al., 2016; Tan et al., 2022). Likewise, the diverse urban and rural areas in CO and MO offer varying environmental factors and resources, affecting access to parks, recreation, and walkable neighborhoods, which may influence physical activity levels in our study groups. Future research should explore how these factors interact to shape health outcomes, leading to targeted interventions and policies for this vulnerable population’s well-being.
Gait speed is a simple and effective indicator of age-related disease and functionality (Stover et al., 2023). In this study, we found that those participants meeting the slowness criteria were less disadvantaged (lower ADI) in the combined cohorts. These results are inconsistent with previous research that showed that slow gait speed may be associated with socioeconomic disparities (e.g., occupational class, low SES, low education level, and social participation limitations) among older adults (Plouvier et al., 2016; Sialino et al., 2021; Warren et al., 2016). This finding may be explained by the type of instrument used to assess the slowness criteria. Although the 4-m walk test has been widely used among older adults and PWH, it is possible that in our sample, the instrument was not sensitive enough to detect variations in daily life walking performance (Van Ancum et al., 2019). Additionally, the age of the CO and MO cohorts was relatively young and predominantly male, which is not typical of the populations in which the gait speed criteria were developed. This, plus the small sample size, is possibly driving our results. Moreover, the effect size of this relationship was small; therefore, further research with larger populations is needed to re-assess this association.
Limitations
With cross-sectional data, we cannot infer causality. The use of a convenience sample from two existing cohorts in CO and MO limits the generalization of the results. We accounted for demographic variations between the CO and MO cohorts by incorporating adjustments for cohort, race/ethnicity, and sex. Nonetheless, this study underscores that the demographic and SDOH profiles of PWH can vary significantly across diverse locations, even within the U.S. Not adjusting for multiple comparisons is also another limitation of this study. It is important to note that our non-significant results do not rule out a relationship between frailty and ADI, and further research is needed to expand the work by conducting studies with bigger sample sizes and longitudinal designs.
Conclusions
Our objective was to examine the relationship between neighborhood disadvantage and frailty. Neighborhood disadvantage was associated with lower physical activity and weakness and less slowness in older PWH but not with overall frailty status. Thus, ADI may be an important factor when assessing physical health and physical activity. Future studies should consider differences in access to spaces for physical activity as it correlates to ADI, which may be a key factor in frailty prevention.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: On the Colorado site, this work was supported by the Gilead Sciences Research Scholars Program in HIV (to KME), the National Institute of Aging of the National Institutes of Health [K23AG050260, K24AG082527, and R01AG066562] to KME, and the NCATS Colorado CTSA UM1TR004399. On the Missouri site, this work was supported by grants from the National Institutes of Health R01NR015738 (to BA) and R01NR014449-04 (to BA). The funding sources had no role in data collection, analysis, interpretation, trial design, or patient recruitment. No payments were made in the writing of this manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
