Objectives: Some evidence suggests that HIV infection is associated with premature frailty—a syndrome typically viewed as being related to ageing. We determined the prevalence and predictors of frailty in a population of HIV-infected individuals in South Africa.
Design: Case-control study of 504 adults more than the age of 30 years, composed of 248 HIV-infected adults and 256 age- and gender-matched, frequency-matched HIV-seronegative individuals.
Methods: Frailty was defined by standardized assessment comprised of ≥3 of weight loss, low physical activity, exhaustion, weak grip strength, and slow walking time. Independent predictors of frailty were evaluated using multivariable logistic regression.
Results: The mean ages of the HIV-infected and HIV-seronegative groups were 41.1 ± 7.9 years and 42.6 ± 9.6 years, respectively. Of the HIV-infected adults, 87.1% were receiving antiretroviral treatment (median duration, 58 months), their median CD4 count was 468 cells/μL (interquartile range = 325–607 cells/μL) and 84.3% had undetectable plasma viral load. HIV-infected adults were more likely to be frail than HIV-seronegative individuals (19.4% vs. 13.3%; P = 0.07), and this association persisted after adjustment for confounding variables [adjusted OR = 2.14; 95% confidence interval (95% CI): 1.16–3.92, P = 0.01]. Among HIV-infected individuals, older age was a strong predictor of frailty, especially among women (women: OR = 2.55 per 10-year age increase; men: OR = 1.29 per 10-year age increase, P-interaction = 0.01). Lower current CD4 count (<500 cells/μL) was also independently associated with frailty (OR = 2.84; 95% CI: 1.02 –7.92, P = 0.04).
Conclusions: HIV infection is associated with premature development of frailty, especially in women. Since higher CD4 counts were associated with lower risk of frailty, earlier initiation of antiretroviral treatment may be protective.
*International Centre for Eye Health, London School of Hygiene & Tropical Medicine, London, United Kingdom
†Department of Clinical Research, Faculty of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, London, United Kingdom
‡MRC Tropical Epidemiology Group, Faculty of Epidemiology and Population Health, London School of Hygiene & Tropical Medicine, London, United Kingdom
§Department of Ophthalmology, Faculty of Health Sciences, University of Cape Town
‖Groote Schuur Hospital, Cape Town, South Africa
¶Desmond Tutu HIV Centre, Institute of Infectious Diseases and Molecular Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa
#Department of Clinical Research, Faculty of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, London, United Kingdom.
Correspondence to: Sophia Pathai, MSc, MRCOphth, International Centre for Eye Health, Department of Clinical Research, Faculty of Infectious and Tropical Diseases, London School of Hygiene & Tropical Medicine, Keppel St, London, WC1E 7HT, United Kingdom (e-mail: firstname.lastname@example.org).
S. Pathai was supported by the Wellcome Trust Grant #090354/Z/09/Z. R. Wood was supported by International Epidemiologic Database to Evaluate Aids (IEDEAA) 5U01AI069924-02 and Cost-Effectiveness of Preventing AIDS Complications (CEPAC) 5 R01AI058736-02. S. Lawn was supported by the Wellcome Trust Grant #088590.
The authors have no other funding or conflicts of interest to disclose.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.jaids.com).
Received June 22, 2012
Accepted September 10, 2012
Major reductions in HIV-associated mortality have occurred as a result of the global scale-up of highly active antiretroviral therapy (ART). This is largely due to prevention of AIDS-related events but is also due to a decrease in non-AIDS–related events and deaths.1–3 Despite these benefits, evidence is emerging that patients receiving ART are at an increased risk of age-related non-AIDS morbidity and mortality compared with HIV-seronegative individuals.4–6 Several of these conditions are classically associated with the normal ageing process but seem to occur at an earlier age in HIV-infected persons compared with age-matched HIV-seronegative individuals. It is possible that not only are HIV cohorts ageing chronologically, but they may also be undergoing accelerated physiologic and immunologic senescence.
Frailty is a clinical syndrome initially described in geriatric populations. It reflects a concept of decreased physiologic and functional reserve and a subsequent decrease in adaptation to external or intrinsic stressors. Frailty is characterized by multiple pathologies, low physical activity, and slow motor performance,7,8 and leads to cognitive and physical decline manifest as an increased risk of mortality, falls, and hospitalization. HIV infection has been associated with premature development of frailty, and it has been speculated that this may emerge as an important clinical syndrome in HIV-infected individuals. The prevalence of premature presentation of frailty in HIV-infected populations is reported to range between 5% and 20%, depending on the study population.9–12 However, limitations of these reports include studying single sex cohorts and differences in frailty criteria. There is also substantial potential for confounding because of differential exposure to potential risk factors between the study population and the reference uninfected population (if used), thereby altering their risk of age-related outcomes.
The epidemiology of HIV and AIDS in sub-Saharan Africa is changing; extensive ART scale-up has led to reduced mortality rates and a rapidly expanding cohort of HIV-infected African patients who are living much longer than they were previously.13,14 However, it is unclear whether these individuals will also be subject to premature ageing in the longer term. Estimates from previous studies from Europe and North America where the epidemiology of HIV differs substantially from that in sub-Saharan Africa, may not be generalizable to African HIV cohorts. In this study, we determined the prevalence and predictors of frailty in an HIV-infected population in sub-Saharan Africa.
Between March and December 2011, unselected HIV-infected individuals older than 30 years were enrolled from a community-based HIV treatment center in Nyanga district of Cape Town, which has been previously described.15,16 All participants had a confirmed serological diagnosis of HIV and were either about to commence ART (ART-naive) or were already on first-line ART. Participants who had active opportunistic infections (OIs) were not recruited; however, participants who had active tuberculosis (TB) (ie, non-symptomatic but still receiving treatment for TB) were enrolled.
A control group of HIV-uninfected participants was recruited using frequency matching by gender and 5-year age categories. HIV-seronegative individuals were enrolled from participants confirmed to be HIV-negative attending an HIV prevention trials site (Emavundleni Centre), located within the same district as the HIV treatment center. These 2 centers were chosen as individuals attending them were drawn from the same community and were therefore likely to have similar sociodemographic characteristics.
The study was approved by the London School of Hygiene and Tropical Medicine Ethics Committee and the University of Cape Town Faculty of Health Sciences Research Ethics Committee and adhered to the tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants.
Assessment of Frailty and Data Collection
Physical frailty was defined by the presence of 3 or more of the 5 criteria: (1) unintentional weight loss (self-report of weight loss was verified in 100 (20%) participants by weighing of the participant at the time of data collection followed by referral to clinic records. Self-report and clinical records were congruent in 80 participants), (2) self-reported low physical activity, (3) self-reported exhaustion, (4) weak grip strength, and (5) slow walking time (see Appendix, Supplemental Digital Content, http://links.lww.com/QAI/A358). All of these 5 components described in the original phenotype by Fried et al17 were used to determine the presence of frailty. However, we used the proxy described by Onen et al10 for the physical activity measure (for description, see Appendix, Supplemental Digital Content, http://links.lww.com/QAI/A358). Grip strength of the dominant hand was measured 3 times using a grip dynamometer (Jamar Plus+ Digital Hand Dynamometer, Jamar, Sammons Preston, Rolyan, Bolingbrook, IL). The average of 3 weight measurements was recorded in kilograms (kg) to one decimal point. Walking time was assessed using the method of Cesari et al.18 The average of 2 trials (in meters per second) was used for analysis. Participants were excluded from the determination of grip strength if they had pain or arthritis of the dominant hand and excluded from the walking test if they had paralysis of an extremity or side of the body or needed to use a walking aid.
Sociodemographic information and medical history were obtained via a questionnaire administered in the participant's first language (Xhosa or English). Sociodemographic and behavioral variables of interest included education, alcohol consumption, smoking history, and income (salaried income and/or social welfare grant). Clinical information was obtained from medical case notes where required. Comorbidity was defined as the concurrent presence of one or more chronic diseases or conditions, including cardiovascular disease, chronic renal failure, airways disease, and malignancy (both AIDS and non-AIDS defining). Cardiovascular diseases included myocardial infarction and cerebrovascular disease. Blood pressure (BP) was measured using a digital sphygmomanometer. Hypertension was defined as a systolic BP of 140 mm Hg or higher, diastolic BP of 90 mm Hg or higher, or the combination of self-reported high BP diagnosis and the use of antihypertensive medications. Height was measured in meters, weight in kilograms, and body mass index (BMI) was calculated as weight per height.2 HIV-related conditions were classified according to the WHO staging system and were based on historical assessment done at the time of enrolment into the ART service. ART was defined as the use of 3 or more antiretroviral drugs, and treatment duration was recorded in months. Nadir and current CD4 count and HIV RNA plasma viral load (VL) were available from medical records, current values being measured within an 8-week period before frailty assessment. VL suppression was defined as HIV RNA <50 copies/mL.
Analyses were conducted on participants with criteria available for determination of frailty phenotype (3 or more of the criteria present). Participants were categorized as “frail” and “non-frail,” using the criteria given above. Comparisons between categorical groups were performed using χ2 tests. Continuous variables were compared using Student t test if normally distributed and the Mann–Whitney U test if nonnormally distributed. Variables were log transformed where appropriate. All P values were 2-tailed and considered significant if P < 0.05. Univariable logistic regression was performed to estimate odds ratio (OR) and 95% confidence intervals (CIs) of factors associated with frailty. Multivariable logistic regression was then used to evaluate independent factors associated with frailty overall and within the case and control groups, respectively. HIV-infected individuals were also analyzed based on their ART status (naive or on treatment). All analyses were performed with Stata 11 (Stata Corp, College Station, TX).
Of the 504 participants, 248 had HIV-infection and 256 were HIV-seronegative individuals. The 2 groups of participants were similar in terms of age and gender (Table 1), but HIV-infected individuals tended to have higher levels of education (87.9% vs. 81.6%, P = 0.05) and reported lower levels of alcohol use and smoking consumption than HIV-seronegative individuals (P < 0.05). HIV-infected individuals reported a greater income than HIV-seronegative individuals (41.6% vs. 32.0% received more than ZAR1000 a month (approximately USD125), which was related to a greater proportion of the HIV-infected group receiving social welfare grants compared with HIV-seronegative individuals (data not shown). Although TB in the form of current or past disease was more common in the HIV group, other comorbidity was slightly higher in the control group. BMI was lower in HIV-infected individuals compared with HIV-seronegative individuals (27.7 ± 6.5 vs. 31.3 ± 8.8 kg/m2, P < 0.0001). Among females, HIV-infected individuals reported fewer pregnancies than HIV-seronegative individuals (P = 0.006).
Clinical characteristics of the HIV-infected individuals are also reported in Table 1. Overall, 72.9% had a history of WHO stage 3 or stage 4 defining illness. The current CD4 count among participants receiving ART was 468 cells/μL [interquartile range (IQR) = 325–607 cells/μL] and 84.3% had undetectable VL. Median treatment duration on ART was 58 months (IQR, 34–75 months). About 12.9% of the HIV-infected group were ART naive and had correspondingly lower CD4 counts and higher VL (Table 1).
Prevalence of Frailty and Association With HIV
Assessment of frailty was possible in all participants (n = 504, with 3 participants excluded from the walking test but still contributing frailty data). Frailty outcomes within the study population are reported in Table 2. The prevalence of frailty was greater among HIV-infected individuals than HIV-seronegative individuals (19.4%, 95% confidence interval [CI]: 14.4% to 24.3% vs. 13.3%, 95% CI: 9.1% to 17.5%; P = 0.07). Levels of “prefrailty” (ie, scoring 1 or 2 of the frailty criteria) were similar between the 2 groups. Of the frailty indicators, weight loss and slow walking time were more common in the HIV-infected group, however, these differences did not reach statistical significance.
Multivariable analysis adjusting for baseline characteristics and a priori confounders showed that HIV infection was independently associated with frailty ( adjusted OR = 2.14; 95% CI: 1.16 to 3.92; Table 3). Age was also a strong predictor, and this association was stronger in females (P for interaction = 0.03). A higher level of education reduced the odds of frailty (OR = 0.44; 95% CI: 0.21 to 0.90). Low BMI was also strongly associated with frailty. TB was not included in the model because of its high correlation with HIV status. When TB was analyzed as a potential binary outcome of interest (no history vs. current/past) the association with frailty was not as strong as that for HIV (OR = 1.73; 95% CI: 0.94 to 3.08) and did not reach statistical significance.
Predictors of Frailty in the HIV-Infected Group
Among the 216 HIV-infected individuals on ART, the prevalence of frailty was 18.0% (95% CI: 13.2% to 23.8%). In univariable analyses, older age and lower socioeconomic status (including education level) and alcohol consumption were associated with higher odds of frailty.
In the multivariable model, there was evidence of an interaction between gender and age with a strong association between increasing age and frailty within females but not males (P for interaction = 0.001). Low BMI was also a strong predictor of frailty (P trend = 0.01). Among the HIV-related covariates, a current CD4 count less than 500 cells/μL was associated with increased risk of frailty (OR = 2.84; 95% CI: 1.02 to 7.92). No associations were found with HIV RNA levels (current or nadir), duration of treatment, or ART regimen. Socioeconomic variables were not associated with frailty in this model. In a multivariable analysis restricted to females on ART, a similar association was seen with current CD4 count. Age continued to act as a strong predictor and low BMI. A history of 3 or more pregnancies was also associated with frailty in this group [OR = 3.42 (1.03–11.36), P = 0.04] (data not shown).
In the ART naive group (n = 32), 28.1% (95% CI: 13.7% to 46.7%; n = 9) were frail. The only factor associated with frailty in a multivariable analysis was WHO clinical stage 3 or 4. Age and gender and other HIV-related covariates were not associated.
To assess the possibility of current diagnoses of TB (the most common OI in this setting)15 causing false misclassification of frailty, we constructed a separate multivariable model excluding such patients (n = 10). In this model, HIV still remained strongly predictive of frailty (OR = 2.08; 95% CI: 1.12 to 3.86, P = 0.02). In a further model of HIV-infected participants that similarly excluded those with a current diagnosis of TB, CD4 count remained a strong predictor of frailty (OR = 2.89; 95% CI: 1.02 to 7.92, P = 0.04) (Table 4).
Predictors of Frailty in the Control Group
Among HIV-seronegative individuals, frailty was independently associated with older age, female gender, and smoking (Table 5). There was also some evidence that TB was associated with frailty. In view of the strong association of female gender with frailty, we performed subanalyses restricted to females. In adjusted analyses, increasing age was still a strong predictor of frailty.
This study provides clear evidence that HIV infection is strongly associated with a 2-fold risk of premature frailty in this African population. Premature frailty was most prevalent in women who comprise the majority of the HIV-infected population in sub-Saharan Africa. These findings have potentially important implications for long-term morbidity among the millions of patients receiving ART long-term in Africa and may have an important bearing on the optimum timing of ART initiation.
Our HIV-related frailty prevalence of 19.4% is higher than estimates from other regions. A prevalence of 9% was reported from a clinic population in the United States (mean age of 42 years).10 The Women's Interagency HIV Study (a prospective cohort in 5 US cities) found a prevalence of 12% in HIV-infected women with clinical AIDS (median age of 41 years).11 In the Multicenter AIDS Cohort Study (a longitudinal study of men who have sex with men), a frailty prevalence of 5%–14% depending on age and duration of HIV infection was reported from 1994–2005 data9,12 and 8% in 2009–2010 among men aged 40–49 years.19 In the SUN study (a US observational cohort of HIV-infected adults, median age of 47 years), 5% of participants were frail.20 The variation in estimates is likely to be attributable to differences in study design and clinical demographics of patients recruited (eg, ART status and degree of immunodeficiency).
In agreement with previous studies of participants on ART, current CD4 count was a strong independent predictor of frailty.9,11,12,19 However, we did not find an association with duration of ART or nadir CD4 cell count. These findings are entirely consistent with previous analyses of factors associated with the risk of incident TB and of mortality in this cohort.21,22 Thus, although these 3 variables are interrelated,23 it seems that current CD4 count best captures current “well-being” in this ART cohort. Thus, the best way to prevent frailty may be to maintain high CD4 counts through early initiation of ART. Potential reductions in non-AIDS–related morbidity and mortality from earlier initiation of ART are currently under investigation in the Strategic Timing of AntiRetroviral Treatment study24 and prevention of frailty may be another important benefit of such a strategy.
It is possible that frailty status could be misclassified in some HIV-infected participants because of recent or current OIs. However, we think this unlikely for several reasons. Participants with acute symptomatic OIs were not eligible for enrolment in the study and a large majority of patients were clinically stable on long-term ART (median 58 months). They had a median CD4 count of 468 cells/ μL and so were at low risk of current comorbidity, and previous WHO stage 3 and 4 defining illnesses were remote. A small number of participants had current diagnoses of TB for which continuation phase treatment was being received. Estimates of HIV as a predictor of frailty did not greatly change when these participants were excluded from analyses, suggesting that misclassification was minimal. Thus, we believe that the observed association of frailty with the HIV-infected group is likely to be related to HIV infection itself rather than as a consequence of symptoms related to current OIs.
Female gender was an important predictor of frailty in both study populations. In the general population, frailty is more common among females.25,26 Men may be protected by greater muscle mass and higher testosterone levels27,29 reflecting the greater biological capital they achieve before age-related decline. Two US-based studies did not find an association of female gender with frailty,10,20 possibly due to the gender composition (predominantly male) of the study populations or reduced statistical power to detect an interaction. The interaction of gender and age in the HIV-infected group in our study is a novel finding. This interaction may be related to effects of decreased circulating estrogens with increasing age and subsequent inflammation. In physiological ageing, low levels of estrogens may be associated with increased levels of proinflammatory cytokines that have been linked to sarcopenia.28 This effect may be exacerbated or modulated by HIV infection or ART. In resource-constrained settings, it is possible that women are nutritionally deficient compared with men, exacerbated by multiple pregnancies.
In sub-Saharan Africa, the proportion of elderly people infected with HIV is increasing.13 In 2007, approximately 3 million people aged 50 years or older were living with HIV in sub-Saharan Africa, comprising 14% of the adult HIV population.30 With 5.1 million people in sub-Saharan Africa having started ART by the end of 2010, the number of people aged 50 years or older living with HIV will inevitably continue to increase.31 These evolving demographics may necessitate a shift from treatment of primarily OIs toward management of non-AIDS–related conditions. Furthermore, if premature ageing is scientifically validated, HIV-infected patients >50 years of age or older may come to be considered “old.”32 Chronic age-related disease within African HIV-infected populations is likely to place a significant burden on health care budgets and human resources. The high prevalence of HIV-related frailty also has additional economic implications. Those who are frail are less likely to be economically productive and more likely to need assistance from families and welfare grants. Interventions to improve frailty may be simple such as progressive resistance exercises. This intervention improved the strength of HIV-infected adults in Brazil.33 However, given the multifactorial aetiology of frailty, other systemic mechanisms may also require intervention.
A key strength of this study is the inclusion of an age/gender-matched control group with a similar sociodemographic profile to the HIV-infected individuals. The hypothesis of premature ageing in HIV has received criticism primarily due to limitations in characterization of participants, in particular the possibility of differential exposure to potential risk factors between HIV-infected and HIV-uninfected populations.32,34,35 For example, in the US Veterans Aging Cohort, HIV-infected veterans were more likely to have a history of substance misuse compared with age- and sex-matched uninfected veterans.36 The differential exposures to risk factors between HIV-infected and HIV-uninfected populations and residual confounding could result in an apparent increased risk of age-related outcomes. By recruiting from the same community, we aimed to reduce the likelihood of differential risk exposure. Compared with HIV-seronegative individuals, the HIV-infected group in this study tended to be more affluent (likely reflecting receipt of welfare grants) and also more educated as has been observed elsewhere in South Africa.37 HIV remained strongly associated with frailty even after adjustment for socioeconomic factors.
This study has some limitations. We studied a frailty-like phenotype rather than the previously defined and validated frailty phenotype.17 However, this modified phenotype is comparable to phenotypes used in other studies of HIV-related frailty.9,10,12 The frailty phenotype within the context of HIV has not been fully established or validated to date. The main limitation of the frailty criteria relates to their subjective nature. Grip strength and walking time measured at the time of data collection may not necessarily represent the participant's overall ability, and a longitudinal evaluation of these parameters would be optimal. Similarly, reports of exhaustion relied on self-report. The original phenotype used a weighted score of kilocalories expended to assess low physical activity. This may be difficult to ascertain in resource-limited settings, and the approximate adopted by Onen et al10 may be more useful in these environments. Although we relied on self-report of weight loss, validation of a proportion of the reports was favorable and no differential misclassification between groups was observed.
The study design means that a causal relationship between HIV and frailty cannot be concluded, nor can a temporal relationship be established. We are unable to infer whether HIV infection or ART is primarily responsible for the strong association with frailty, partly due to a low proportion of ART-naive participants (12.9%). Often such patients are more acutely ill and may be less willing or able to take part in research studies. In the Multicenter AIDS Cohort Study cohort, the presence of frailty before ART initiation was an independent predictor of the development of AIDS or death despite ART.38 Thus, frailty status may assist in establishing risk of morbidity and mortality. In the present study, we were unable to assess whether a diagnosis of frailty was related to outcomes typically associated with the syndrome in older adults, such as falls, hospitalization, and death. To fully assess if individuals in sub-Saharan Africa meeting the definition of frailty are at increased risk of these outcomes will require longitudinal studies.
In conclusion, HIV is an important predictor of frailty in this African population. Sub-Saharan Africa is undergoing significant HIV-related demographic changes, leading to an ageing HIV-infected population. Chronic age-related conditions will impact this population and HIV-related premature ageing will likely compound this disease burden. Early initiation of ART at higher CD4 counts may maintain CD4 counts at higher levels and protect against development of the frailty phenotype. As access to ART expands, and patients continue to age and live with HIV infection, longitudinal studies are needed to assess the evolution of frailty within HIV-infected populations and its impact on morbidity and mortality.
1. Mocroft A, Brettle R, Kirk O, et al.. Changes in the cause of death among HIV positive subjects across Europe: results from the EuroSIDA study. AIDS. 2002;16:1663–1671.
2. The Antiretroviral Therapy Cohort Collaboration. Causes of death in HIV-1-infected patients treated with antiretroviral therapy, 1996-2006: collaborative analysis of 13 HIV cohort studies. Clin Infect Dis. 2010;50:1387–1396.
3. El-Sadr WM, Lundgren JD, Neaton JD, et al.. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med. 2006;355:2283–2296.
4. Deeks SG. Immune dysfunction, inflammation, and accelerated aging in patients on antiretroviral therapy. Top HIV Med. 2009;17:118–123.
5. Lohse N, Hansen AB, Pedersen G, et al.. Survival of persons with and without HIV infection in Denmark, 1995-2005. Ann Intern Med. 2007;146:87–95.
6. Guaraldi G, Orlando G, Zona S, et al.. Premature age-related comorbidities among HIV-infected persons compared with the general population. Clin Infect Dis. 2011;53:1120–1126.
7. Fulop T, Larbi A, Witkowski JM, et al.. Aging, frailty and age-related diseases. Biogerontology. 2010;11:547–563.
8. Hubbard RE, Woodhouse KW. Frailty, inflammation and the elderly. Biogerontology. 2010;11:635–641.
9. Desquilbet L, Margolick JB, Fried LP, et al.. Relationship between a frailty-related phenotype and progressive deterioration of the immune system in HIV-infected men. J Acquir Immune Defic Syndr. 2009;50:299–306.
10. Onen NF, Agbebi A, Shacham E, et al.. Frailty among HIV-infected persons in an urban outpatient care setting. J Infect. 2009;59:346–352.
11. Terzian AS, Holman S, Nathwani N, et al.. Factors associated with preclinical disability and frailty among HIV-infected and HIV-uninfected women in the era of cART. J Womens Health (Larchmt). 2009;18:1965–1974.
12. Desquilbet L, Jacobson LP, Fried LP, et al.. HIV-1 infection is associated with an earlier occurrence of a phenotype related to frailty. J Gerontol A Biol Sci Med Sci. 2007;62:1279–1286.
13. Mills EJ, Rammohan A, Awofeso N. Ageing faster with AIDS in Africa. Lancet. 2011;377:1131–1133.
14. Mills EJ, Bärnighausen T, Negin J. HIV and aging—preparing for the challenges ahead. N Engl J Med. 2012;366:1270–1273.
15. Lawn SD, Myer L, Bekker LG, et al.. Burden of tuberculosis in an antiretroviral treatment programme in sub-Saharan Africa: impact on treatment outcomes and implications for tuberculosis control. AIDS. 2006;20:1605–1612.
16. Lawn SD, Myer L, Orrell C, et al.. Early mortality among adults accessing a community-based antiretroviral service in South Africa: implications for programme design. AIDS. 2005;19:2141–2148.
17. Fried LP, Tangen CM, Walston J, et al.. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci. 2001;56:M146–M156.
18. Cesari M, Kritchevsky SB, Penninx BW, et al.. Prognostic value of usual gait speed in well-functioning older people—results from the Health, Aging and Body Composition Study. J Am Geriatr Soc. 2005;53:1675–1680.
19. Margolick JB, Li X, Detels R, et al.. Earlier occurrence of the frailty phenotype in HIV+ men than in HIV-men. Paper presented at: 18th Conference on Retroviruses and Opportunistic Infections. Boston, MA; February 27 – March 2, 2011.
20. Onen NF, Patel P, Baker J, et al.. Frailty and pre-frailty in a contemporary cohort of HIV+ adults. 19th Conference on Retroviruses and Opportunistic Infections. Seattle, WA; March 5–8, 2012.
21. Lawn SD, Little F, Bekker LG, et al.. Changing mortality risk associated with CD4 cell response to antiretroviral therapy in South Africa. AIDS. 2009;23:335–342.
22. Lawn SD, Myer L, Edwards D, et al.. Short-term and long-term risk of tuberculosis associated with CD4 cell recovery during antiretroviral therapy in South Africa. AIDS. 2009;23:1717–1725.
23. Battegay M, Nuesch R, Hirschel B, et al.. Immunological recovery and antiretroviral therapy in HIV-1 infection. Lancet Infect Dis. 2006;6:280–287.
24. Babiker AG, Emery S, Fätkenheuer G, et al.. Considerations in the rationale, design and methods of the Strategic Timing of AntiRetroviral Treatment (START) study. Clin Trials. E-pub ahead of print: 2012.
25. Puts MT, Lips P, Deeg DJ. Sex differences in the risk of frailty for mortality independent of disability and chronic diseases. J Am Geriatr Soc. 2005;53:40–47.
26. Avila-Funes JA, Helmer C, Amieva H, et al.. Frailty among community-dwelling elderly people in France: the three-city study. J Gerontol A Biol Sci Med Sci. 2008;63:1089–1096.
27. Walston J, Fried LP. Frailty and the older man. Med Clin North Am. 1999;83:1173–1194.
28. Joseph C, Kenny AM, Taxel P, et al.. Role of endocrine-immune dysregulation in osteoporosis, sarcopenia, frailty and fracture risk. Mol Aspects Med. 2005;26:181–201.
29. Gasparotto AS, Sprinz E, Lazzaretti RK, et al.. Genetic polymorphisms in estrogen receptors and sexual dimorphism in fat redistribution in HIV-infected patients on HAART. AIDS. 2012;26:19–26.
30. Negin J, Cumming RG. HIV infection in older adults in sub-Saharan Africa: extrapolating prevalence from existing data. Bull World Health Organ. 2010;88:847–853.
31. World Health Organisation. GLOBAL HIV/AIDS RESPONSE: Epidemic Update and Health Sector Progress Towards Universal Access, Progress Report 2011. Geneva, Switzerland: WHO, UNICEF, UNAIDS; 2011:97.
32. Capeau J. Premature aging and premature age-related comorbidities in HIV-infected patients: facts and hypotheses. Clin Infect Dis. 2011;53:1127–1129.
33. Souza PMLD, Jacob-Filho W, Santarém JM, et al.. Effect of progressive resistance exercise on strength evolution of elderly patients living with HIV compared to healthy controls. Clinics (Sao Paulo). 2011;66:261–266.
34. Fisher M, Cooper V. HIV and ageing: premature ageing or premature conclusions? Curr Opin Infect Dis. 2012;25:1–3.
35. Martin J, Volberding P. HIV and premature aging: a field still in its infancy. Ann Intern Med. 2010;153:477–479.
36. Goulet JL, Fultz SL, Rimland D, et al.. Aging and infectious diseases: do patterns of comorbidity vary by HIV status, age, and HIV severity? Clin Infect Dis. 2007;45:1593–1601.
37. Johnson L, Budlender D. HIV Risk Factors: A Review of the Demographic, Socio-Economic, Biomedical and Behavioural Determinants of HIV Prevalence in South Africa; Centre for Acurial Research, University of Cape Town; 2002. Cape Town, South Africa: CARE Monograph No 8.
38. Desquilbet L, Jacobson LP, Fried LP, et al.. A frailty-related phenotype before HAART initiation as an independent risk factor for AIDS or death after HAART among HIV-infected men. J Gerontol A Biol Sci Med Sci. 2011;66:1030–1038.