The high risk of frailty in middle-aged people living with HIV
Frailty, the condition of depleted energy and physiological decline that is often seen in the elderly and that makes them less capable of recovering from illness, accidents and the stresses of life, has been reported to be significantly more common in people living with HIV — even in middle-aged people on antiretroviral treatment (ART) —than in people without HIV according to a growing number of studies including a recent analysis of the Multicenter AIDS Cohort Study (MACS) (described in a summary on aidsmap), the Women’s Interagency HIV study, and the AGEhIV Cohort Study from the Netherlands). These studies have primarily been cross-sectional — in other words, based on data that was collected only once from a population, like a snapshot at a fixed point-in-time.
Now, a new analysis of the AGE h IV study that followed the cohort over time has found that people living with HIV on ART had twice the risk of progressing from a robust state to frailty than a matched group of people without HIV. This was true both in the univariate analysis and in an analysis that was adjusted for age, risk group (e.g., men who have sex with men, people who inject drugs, etc.), ethnicity and education — though people in the cohort who were not HIV-positive but who had as high a number of comorbidities, waist-to-hip ratios and depressive symptoms as the people with HIV were also more likely to become frail.
In addition, the analysis found that regardless of whether it was found in people with or without HIV in the cohort, “frailty was a strong predictor of both mortality and comorbidity development,” said Dr. Eveline Verheij of the Academic Medical Center of the GGD (Amsterdam’s Health Service), who presented the study findings at AIDS 2018 on July 26.
So, while it may seem like something of a buzzkill to talk about premature ageing and frailty in the era of U=U (undetectable equals untransmittable), it may be time that more of us begin to take more notice.
Background [and a rather lengthy digression]
Anyone who has watched elderly loved ones dwindle away into shadows of their former selves has seen frailty — but not everyone may be aware that frailty is considered to be syndrome in geriatric medicine, though the operational definitions in the medical literature vary substantially.
Features of frailty have long been associated with HIV and AIDS — ever since it was first called ‘slim disease’ in east Africa. AIDS-related wasting syndrome was attributed both to opportunistic infections in the GI track and, eventually, to inflammation and metabolic changes in the gut brought on by HIV itself. There were other characteristics of accelerated aging as well, which led some researchers to postulate that a better understanding of the pathogenesis of AIDS would provide insights into how frailty develops in the elderly.
With the advent of ART, much of this research was moved to the back burner — but it continued to simmer there. Frank wasting syndrome became much less common, but it was quickly supplanted by lipodystrophy and other body composition changes. Soon other comorbidities such as heart disease began to occur more often at younger ages than usual in people living with HIV — and warnings went out that premature ageing might be something we would have to worry about.
Initially, this was attributed to the toxicity of the first antiretroviral drugs (which indeed deserved some of the blame). Eventually, and particularly as treatments improved, the pendulum swung back towards the long-term effects of HIV infection and chronic inflammation. This was perhaps mediated by ongoing low-level HIV replication or other physiological consequences of HIV infection such as the metabolic effects of damage in the gastrointestinal track (leaky gut syndrome) and acquired deficits in the immune system that are never fully recovered in some individuals, especially if there was a delay in initiating treatment (the norm until quite recently). Other fingers pointed at risk factors and substance use (such as smoking — which is particularly ageing) or conditions such as depression that are more common among people living with HIV.
In the twenty years after the first reports of lipoatrophy, there have been many further reports of increased age-related comorbidities, including osteoporosis, chronic kidney disease, certain cancers and sarcopenia (where lean muscle mass is lost and replaced by fat, compromising muscle quality). However, it is important to add that the majority of people with HIV have been doing quite well as long as they have been in care and taking ART. They’ve been doing so well, in fact, that some researchers have recently questioned whether the reports of premature ageing were somewhat overblown.
Some have pointed to data from a recently published large multicohort study in the US and Europe (the Antiretroviral Therapy Cohort Collaboration) that concluded that people living with HIV who started ART within the last ten years and had reached a CD4 cell count of around 350 by the end of the first year on treatment, would have a life expectancy similar to that of an HIV-negative person’s [though this was only over the second and third year of the study follow-up] —a marked improvement over the survival of people starting ART during earlier periods. This could be because the antiretroviral drugs used today are markedly safer, and probably also because treatment is started earlier — now as soon as possible after testing positive and before too much physiological and immunologic damage has been done. In addition, some of the improved outcomes seem to be because clinicians have heeded the warnings and ramped up the clinical management of heart disease and other comorbidities in their HIV-positive patients — with improved screening, prevention and treatment.
That’s great news, though achieving these results requires a consistently high standard of care across HIV-treating facilities, which actually does not exist in many settings. In fact, one can safely infer that such results are dependent upon a lifelong retention in high quality care across a number of medical disciplines — a precarious situation for people who live where there are increasing barriers to accessing quality public services —including in the United States under the current administration. How quickly might someone falter if they fall through the cracks and there are gaps in their health care coverage? What will happen to those of us who lose Medicaid or health insurance? What if your antiretrovirals are free, but you cannot get referrals to specialist care for your comorbidities?
Moreover, it is little consolation if you are one of those individuals whose CD4 cell count doesn’t quite reach 350 after a year on treatment — which is far more likely for those who were only diagnosed with HIV after the age of 50 – or if you are one of the ragtag lot of plucky survivors who started treatment more than ten years ago, with the old drugs. This lengthy digression has been for those who are more at risk. That population is growing —more than half of the HIV-positive population in the US is now over the age of 50.
But back to Dr. Verheij’s presentation.
The AGEhIV study started in 2010, enrolling 596 people with HIV who were above the age of 45 at the HIV clinic of the Academic Medical Center. This cohort was compared to 550 similarly aged people without HIV who were attending the public health service of Amsterdam. The researchers assessed frailty using a system proposed by Fried et al that looks for the presence of five factors — slow gait, weakness (often assessed by grip strength), low levels of physical activity (sedentary behavior), exhaustion (low endurance), and weight loss (generally involuntary) in order to tally up a ‘Frailty Score’ that classifies individuals as either robust, pre-frail or frail. If none of the five factors are present, the person is scored as being robust; when three or more of these factors are found in a person, they are considered frail while those with one to two factors are categorized as being ‘pre-frail.’ This is much simpler (though some would argue, less nuanced) than other commonly used approaches (more on this below).
Having previously shown that people with HIV had a much greater risk of being frail than the group without, Dr. Verheij and colleagues used data from the cohort gathered at two visits during two later periods (2012-2014, and 2014 to 2016) to assess impact of frailty on mortality, comorbidity development, and to see whether people living with HIV were at higher risk of becoming frail.
The cohorts were relatively well matched for age (both a median of ~52) and HIV risk group (70.2% were men who have sex with men in the HIV negative group, and 75.8% of the group with HIV), but a number of other variables were significantly different. For instance, the HIV negative group were more likely to be Dutch and better educated and had slightly lower waist to hip ratios and rates of depression. In addition, 38.9% of the HIV-negative cohort vs. 51.7% of the HIV-positive had one or more comorbidities at baseline.
In the HIV-positive cohort, the median time at study entry since HIV diagnosis was 12.2 years (range 6.6 to 17.1 years). Close to 96% were on ART and virally suppressed, with a median cumulative exposure of just over 10 (range 4.4-14.5) years. Although they had a decent CD4 cell count at study entry (median 565, range 433-740), the nadir was considerably lower: median 170 with a range from 70-260 — almost a third of them had been diagnosed with AIDS. The low nadir CD4 cell and earlier AIDS diagnosis are important because both have been associated with a higher incidence of a variety of comorbidities in individuals with suppressed viral loads and high CD4 cell counts.
The analysis of data collected at the first study visit had found that the prevalence of being frail was three times higher in the HIV-positive group than in the HIV-negative group (10% vs 3%). On the opposite end of the spectrum, a quick look at the graph below shows that the proportion of people with HIV who were robust at the ages of 45 to 49 was much lower than in the HIV-negative cohort, and in fact, quite similar to the proportion in that cohort who still remained robust above 65 years of age (only around 40%). Note, this is consistent with an earlier report from the MACS that found that the prevalence of frailty in 55-year old men living with HIV was similar to that in HIV-negative men over the age of 65.,
New results in more detail:
The mortality analysis was based upon the entire cohort, using Kaplan Meier and Cox regression for multivariate analysis. Overall, there were 17 deaths, five among the HIV-negative participants and 12 among the HIV-positive. Though the mortality rate was low, the frail participants had (to no surprise) a higher risk of dying (log-rank p-value < 0.001). Perhaps because the number of deaths has so far been low, the risk was not attenuated by HIV status, age or comorbidities.
However, the risk of developing a comorbidity was also similar in both groups. This analysis looked only at those for whom data from subsequent visits were available, 479 people in the HIV-negative group, and 497 people in the HIV-positive group. Over 3727 person-years of follow-up, 276 participants developed in total 329 new comorbidities. Three quarters of the new diagnoses were of hypertension, chronic obstructive pulmonary disorder (COPD), renal disease and osteoporosis. Diabetes, cancers, cardiovascular disease and heart failure also occurred.
Although the finding of similar rates of new comorbidities seems counter-intuitive, it should be remembered that the HIV positive cohort already had a significantly higher burden of comorbidities at the start of the study — and many had had multiple comorbidities. It might also be a testament to the effectiveness of the antiretrovirals and quality of care these individuals were receiving (consistent with the Antiretroviral Therapy Cohort Collaboration). It may also support indications that most of the damage to the long-term health of people living with HIV is already in place by the time they go onto ART.
And yet, despite the similar incidence of new comorbidities, the higher number of those progressing to frailty (notably, from a robust state) seems to show that something different is happening in the HIV-positive group. The frailty analysis included only those who were robust at the first study visit (284 in the HIV-negative group and 204 of the HIV-positive group). HIV infection was associated with twice the risk of progression to frailty — and, in an adjusted analysis, this difference was not affected by age, risk group, ethnicity or education.
The difference in risk was somewhat attenuated by waist-to-hip ratio, though not significantly so. This finding is notable because, in a key study from 2012 that looked at frailty and body composition (using DEXA and MRI), both central obesity and fat redistribution (particularly fat infiltration in the muscle) were found to be important predictors of frailty in people living with HIV,  though a recent MACS analysis found that higher waist circumference (abdominal obesity) and sarcopenia were associated with frailty whether men were HIV-positive or not. Regardless, frailty in the context of obesity can be insidious, because the fat may conceal the visible loss of muscle from both the individual at risk and from their care providers.
In the multivariate analysis further adjusted to combine a high waist-to-hip ratio and number of comorbidities, frailty was still more common in the HIV-positive group, but the finding was no longer statistically significant. When the analysis was adjusted further to account for depression, the association of HIV status with frailty weakened even more. This may suggest that in this mostly gay middle-aged male cohort, those who were overweight, seriously ill and suffering depression were more likely to become frail regardless of HIV status. But the fact remains that people living with HIV are, on the whole, more likely to have fat redistribution (and likely to have suffered muscle loss), comorbidities and be depressed and thus, more likely to be frail.
Despite the overly rosy projections some have made about normal life expectancies on ART, given the data showing that frailty had a strong impact on mortality, over time, it will likely have a significant impact on life expectancy that will be disproportionately higher in middle aged and elderly people living with HIV. The community could be in serious jeopardy of losing more of our precious survivors, unless something is done.
So what can be done?
Dr. Verheij said that the first step would be to recognize those at risk for developing frailty and associated adverse health outcomes because it is easier to maintain robustness in those who are currently robust, while those who are pre-frail or frail are more likely to stay that way.
A range of interventions were mentioned by Dr. Verheij and audience members, including smoking cessation (a no-brainer) and strengthening connections within social networks. Mobility and exercise are also key to countering frailty, but as one recent systemic literature review in the elderly frail found, some types of exercise interventions are effective, while others are not. Others stressed mental health services to counter depression and ‘toxic stress’.
“This is something that we should put in our clinics, screening our patients so that we could identify comorbidities before they [become] advanced,” said Dr. Cissy Kityo Mutuluuza of Uganda who was one of the session’s moderators.
But this raises another complicated issue: How exactly can care providers screen for frailty in a regular HIV clinic setting?
This was not explained during Dr. Verheij’s presentation, but according to the earlier paper from the study, the screen involved a short questionnaire about unintentional weight loss in the previous year, physical activity and exhaustion based upon questions drawn from other screening tools (a review of the literature finds that some other studies used more exacting measures of energy expenditure), while assessing gait required a timed walk test (of 15 feet). Grip strength was measured using something called a Jamar handheld dynamometer.
In other words, it would require someone trained and with the proper tools and space to perform the assessment. It would be interesting to hear if anyone is routinely doing such assessments in their middle aged and elderly HIV positive patients.
There are many other subjective and objective frailty assessment methods (at least 25 according to one recent paper). One such approach that has been used in several studies of people living with HIV is the Frailty Index proposed by Rockwood and Mitnitski and., This involves tallying up the total number of accumulated ‘deficits’ out of a long checklist (usually at least 30 health variables) of symptoms, signs, laboratory abnormalities, diseases, and disabilities that cover all the body’s systems and that are strongly associated with advanced age. It would be interesting to see whether this approach would better identify those at risk of poor health outcomes if used in the same cohort, though they might be testing somewhat different things.
A better consideration socio-demographic factors that may be associated with frailty may also be warranted. A number of studies highlight the importance of food security and nutrition. The Women’s Interagency HIV study found higher rates of frailty among women living with HIV with low incomes.
Another audience member asked Dr. Verheij whether they had investigated inflammatory serum biomarkers or of microbial translocation that might be predictive of progression to frailty. She said that she had, but she could not find any association between a range of biomarkers and progression to frailty or of an increased risk of comorbidity development.
In other words, there’s no simple lab test.
The danger in all of this is that a slide towards frailty may be modifiable at first but if, bit by bit, cumulative changes go unnoticed and unaddressed, one may reach a point where it may be too late and will become extremely difficult to reverse.
This suggests, to this writer at least, that even though everyone may want a moment to celebrate the data on U=U and be cheerleaders for early treatment, we should not be dismissive of the needs of those who are not doing so well, those who came before and who continue to be at risk of poor outcomes. They must not be swept under the carpet.
Until we can agree on how to best identify those at risk, a state of hyper vigilance may be necessary among middle aged people living with HIV — and we need to focus on prevention as there appears to be no fully effective restorative treatment.
This piece is quite long, but it is only scratching the surface. Time permitting, I hope to write further on this subject sometime in the near future.
Note: Here’s the video for the entire NCD session.
 Hawkins KL et al. Abdominal obesity, sarcopenia, and osteoporosis are strongly associated with frailty in MACS. AIDS, online edition, 2018.
 Gustafson, DR et al. Frailty and constellations of factors in aging HIV-infected and uninfected women – the Women’s Interagency HIV study. J Frailty Aging. 2016;5(1):43-8. doi: 10.14283/jfa.2016.79.
 Kooij KW et al. HIV infection is independently associated with frailty in middle-aged HIV type 1-infected individuals compared with similar but uninfected controls. AIDS 30: 241-250, 2016.
 Verheij E et al. Increased risk of both mortality and incident comorbidity among frail HIV-positive and HIV-negative participants in the AGEhIV Cohort Study, and increased risk of frailty progression in those with HIV. Abstract THAB0105.
 Pathophysiology of the AIDS wasting syndrome.
 Margolick JB, Chopra RK. Relationship between the immune system and frailty: pathogenesis of immune deficiency in HIV infection and aging. Aging (Milano) 1992; 4:255–7.
 Charlotte Y et al. Progressive Proximal-to-Distal Reduction in Expression of the Tight Junction Complex in Colonic Epithelium of Virally-Suppressed HIV Individuals. PLoS Pathogens, 2014; 10 (6): e1004198 DOI: 10.1371/journal.ppat.1004198
 Serrano-Villar et al. The CD4:CD8 ratio is associated with markers of age-associated disease in virally suppressed HIV-infected patients with immunological recovery. HIV Med. 2014 Jan;15(1):40-9. doi: 10.1111/hiv.12081. Epub 2013 Sep 6.
 Engel T et al. HIV and Aging – Perhaps Not as Dramatic as We Feared? Gerontology. 2018 Jun 15:1-11. doi: 10.1159/000489172. [Epub ahead of print]
 The Antiretroviral Therapy Cohort Collaboration. Survival of HIV-positive patients starting antiretroviral therapy between 1996 and 2013; a collaborative analysis of cohort studies. The Lancet HIV, online publication 10 May 2017. See http://dx.doi.org/10.1016/S2352-3018(17)30066-8.
 Fried LP et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001; 56(3):M146–156.
 Desquilbet L et al. HIV-1 infection is associated with an earlier occurrence of a phenotype related to frailty. Journals of Gerontology Series A—Biological Sciences & Medical Sciences. 2007; 62(11):1279–86.
 Shah K et al. A new frailty syndrome: Central obesity and frailty in older adults with the human immunodeficiency virus. J Am Geriatr Soc 2012; 60:545-9; PMID:22315957
Wasserman P, Segal-Maurer S, Rubin DS. High prevalence of low skeletal muscle mass associated with male gender in midlife and older HIV-infected persons despite CD4 cell reconstitution and viral suppression. Journal of the International Association of Providers of AIDS Care. 2014; 13(2):145–152.
 Hawkins, Ibid.
 Apóstolo J et al. Effectiveness of interventions to prevent pre-frailty and frailty progression in older adults: a systematic review. JBI Database System Rev Implement Rep. 2018 Jan; 16(1): 140–232. Published online 2018 Jan 12.
 Pritchard JM et al. Measuring frailty in clinical practice: a comparison of physical frailty assessment methods in a geriatric out-patient clinic. BMC Geriatr. 2017; 17: 264. Published online 2017 Nov 13. doi: 10.1186/s12877-017-0623-0.
 Rockwood K, Mitnitski A. Frailty defined by deficit accumulation and geriatric medicine defined by frailty. Clinics in geriatric medicine. Feb; 2011 27(1):17–26.
 Brothers TD et al. Predictors of transitions in frailty severity and mortality among people aging with HIV. PLoS ONE 12(10):e0185352. https://doi.org/10.1371/journal.pone.0185352.
 Smit et al. Ibid.
 Gustafson et al. Ibid.