Odds of Frailty Twice Higher in Older Adults With HIV

HIV infection independently predicted frailty in a unique comparison of older adults with versus without HIV. Body composition measures were the variables most strongly related to frailty in this AGEhIV cohort analysis.

Frailty is a clinically identifiable syndrome, which the AGEhIV investigators define as "enhanced vulnerability to stressors and increased risk of adverse outcomes." The frailty phenotype defined by Fried and colleagues has been linked to increased morbidity and mortality in the general population. Previous research found a higher frailty phenotype prevalence in HIV populations than in HIV-negative comparison groups. The AGEhIV team conducted this study to compare the prevalence of frailty and prefrailty in HIV-positive people 45 and older and a matched HIV-negative group.

The AGEhIV cohort comprised 597 HIV-positive people at least 45 years old who were recruited at the Academic Medical Center in Amsterdam. AGEhIV researchers also recruited 551 HIV-negative people at an Amsterdam sexual health clinic and from the Amsterdam Cohort Studies on HIV/AIDS. The investigators selected HIV-negative controls to create a group matched to the HIV cohort in age, gender and ethnicity, with a similar geographical, socio-demographic and behavioral background. HIV-positive and negative cohort members were screened twice yearly for age-associated morbidities.

Participants had a frailty phenotype if they had at least three of five conditions: (1) weight loss, (2) low physical activity, (3) exhaustion, (4) low grip strength and (5) slow walking speed. Participants had prefrailty if they met two of the five criteria. For each participant, researchers counted the number of age-associated noncommunicable comorbidities diagnosed during follow-up. They also determined waist and hip circumference and body mass index (BMI) for each participant and measured markers of inflammation and immune activation.

The analysis included 521 people with HIV and 513 without HIV. Median age stood at 52.8 years in the HIV group and 52.1 years in the control group. Most participants in the HIV group (88.7%) and the HIV-negative group (85.8%) were men. The HIV group included a higher proportion of blacks (12.7% versus 5.5%, P < .001) and a higher proportion of smokers (32.4% versus 24.5%, P = .02). Higher proportions of people with HIV had two age-associated comorbidities (24.0% versus 18.7%) or three or more comorbidities (12.3% versus 5.1%) (P < .001). Waist-to-hip ratio was significantly higher (worse) in people with than without HIV (median 0.97 versus 0.92, P < .001), and people with HIV had a slightly but significantly lower body mass index (BMI) (24.2 versus 24.5 kg/m2, P = .02). The HIV group had significantly higher levels of some markers of inflammation and immune activation. Most people with HIV (94.4%) were taking antiretroviral therapy at enrollment, and 93.3% of treated people had an undetectable viral load.

Significantly higher proportions of people with HIV were frail (10.6% versus 2.7%) or prefrail (50.7% versus 36.3%) (P < .001). Greater frailty and prefrailty prevalence with HIV held true for every age group: 45 to 50, 50 to 55, 55 to 60, 60 to 65 and over 65. Logistic regression analysis adjusted for age, sex, black race, smoking, and hepatitis C coinfection determined that people with HIV had more than twice higher odds of frailty (odds ratio [OR] 2.39, 95% confidence interval [CI] 1.85 to 3.09, P < .0001). Further adjustment for waist-to-hip ratio slightly attenuated the association between HIV and frailty (OR 1.93, 95% CI 1.46 to 2.55, P < .001). Adding BMI to the model further attenuated the impact of HIV (OR 1.74, 95% CI 1.31 to 2.32, P < .001). BMI below 20 kg/m2 independently raised the odds of being in a higher frailty category in people with HIV (OR 6.14, 95% CI 3.10 to 12.18, P < .001) but not in HIV-negative controls.

A multivariable model focused solely on people with HIV determined that each year with a CD4+ count below 200 cells/mm3 raised the odds of being in a higher frailty category 14% (OR 1.14/year, 95% CI 1.00 to 1.30, P = .04). Each year of protease inhibitor therapy raised the odds of being in a higher frailty category 5% (OR 1.05/year, 95% CI 1.01 to 1.10, P = .01). Depressive symptoms were independently associated with a higher frailty category, but markers of inflammation and immune activation were not linked to worse frailty.

The authors note that the impact of higher waist-to-hip ratio in mediating the impact of HIV on frailty suggests that both abdominal obesity and lean hips (due to lipoatrophy or sarcopenia) "may contribute to the development of frailty in the context of treated HIV infection." At the same time, BMI below 20 kg/m2 was independently associated with a higher frailty category in people with HIV (but not in HIV-negative controls). The AGEhIV team suggests that low BMI "may be a persistent consequence of advanced HIV-disease and related weight loss." The investigators observe that depression may be either a cause or a consequence of frailty. They propose that frailty results from advanced HIV disease (marked by a low CD4+ count and low BMI), which might be avoided by starting antiretroviral therapy earlier.