Impact of age-related comorbidities on five-year overall mortality among elderly HIV-infected patients in the late HAART era — Role of chronic renal disease

Abstract

Objectives

To identify main prognostic factors for 5-year mortality among age-related comorbidities (ARCs) in older people living with HIV (PLHIV).

Design

A prospective, multicentre cohort study with a 5-year follow-up period in the late HAART era (from January 2008 to December 2012).

Setting

The Dat'AIDS cohort involving 12 French hospitals.

Participants

All actively followed HIV-1 infected patients aged 60 or older.

Measurements

The study endpoint was all-cause five-year mortality. The following ARCs were considered: chronic renal disease, cardiovascular diseases, cancer, chronic pulmonary disease, cirrhosis, diabetes and nutritional status. Hepatitis C (HCV), hepatitis B (HBV) co-infection and sociodemographic characteristics were also evaluated. Cox's Proportional Hazards model was used for multivariate analysis.

Results

Among 1415 PLHIV aged 60 or more patients included, mean age was 66±5.5 years; 154 died (mortality rate 2.47/100 patientyears). The most prevalent ARCs were chronic renal disease (20.1%), diabetes (14.2%) and cardiovascular diseases (12.2%). By multivariate analysis, chronic renal disease (adjusted hazard ratio (aHR)=2.25; 95% confidence interval (CI) [1.58-2.21]; p<10-4), cardiovascular diseases (aHR=2.40; 95%CI[1.64-3.52]; p<10-4), non-HIV related cancer (aHR=1.91; 95%CI[1.20-3.05]; p=0.007), cirrhosis (aHR=2.99; 95%CI[1.68-5.33]; p<10-3), HCV co-infection (aHR=2.00; 95%CI[1.18-3.38]; p=0.009), low body mass index (aHR=2.42; 95%CI[1.46-4.01]; p<10-3) and CD4 cell count < 200cells/μl (aHR=2.23; 95%CI[1.36-3.65]; p=0.002) were independently associated with 5 year mortality.

Conclusion

Due to a high prevalence, chronic renal disease and cardiovascular disease are main prognostic factors for 5-year mortality among aged PLHIV.

Introduction

Since the beginning of the highly active antiretroviral therapy (HAART) era, the HIV-infected population is ageing in high income countries (1). In France, 35% of people living with HIV (PLHIV) were 50 or older and 11% were 60 or older at last visit in 2011 (2). By 2015, half of the PLHIV in the USA will be aged 50 or older (3). Age at diagnosis is also increasing (4). Aged PLHIV have more comorbidities, and with earlier onset, than the general elderly population (5, 6, 7). The aging process in aged PLHIV may be different to that in the general aged population (7). Indeed, HIV infection and antiretroviral therapy could influence the aging process through systemic inflammation, drug toxicity, interactions with other comorbidities, and/or hepatitis co-infections (7).

This translation of age structure has an impact both on comorbidity prevalence and on overall mortality, contributing in association with therapeutic improvement, to a progressive decrease in AIDS-related deaths in favour of deaths from age-related comorbidities (ARC) (8). In the literature, this age cut-off ranges from 50 to 60 (9, 10, 11). To date, this new aged population remains poorly documented. Identifying the ARC that have the greatest impact on the patient's life course could make it possible to implement targeted preventive measures.

The aim of this study was thus to identify, among ARC, prognostic factors for 5-year mortality in PLHIV aged 60 or over followed-up in the context of a large French prospective cohort in the late HAART era (from January 2008 to December 2012). We chose the highest age threshold (>60 years old) to select older PLHIV, as this is more adapted to the context of high-income countries, and also made it possible to focus more specifically on this new and growing population of aged PLHIV.

Methods

This study involved all HIV-1 infected patients aged 60 or over on January 1st, 2008, followed-up in the context of the Dat'AIDS cohort involving 12 French hospitals. Dat'AIDS is a French multicenter prospective cohort that covers inpatients or outpatients treated in French public hospitals, including French overseas territories. It is based on a computerized medical record that is used by clinicians in real time during their consultations since 2000 (Nadis®, Fedialis Medica, Marly le Roi, France) (12). This study was performed in accordance with the principles of the Declaration of Helsinki and current French legislation related to biomedical research.

Patients were excluded if they were co-infected by HIV-2 or if they did not have at least one CD4 cell count available in the 12 months before or after January 1st, 2008. For each patient, follow-up began on January 1st, 2008 and stopped on the date of death, or date of last follow-up, or on January 1st, 2013 (whichever came first).

The study endpoint was all-cause five-year mortality. In patients lost to follow-up (defined as a date of last followup prior to January 1st, 2013 and alive at last follow-up), vital status was systematically recorded via the Centre for Epidemiology and Population Health (CESP), by linkage with the French National Institute of Statistics (INSEE), which records nearly all deaths that occur in France by centralized reception of death certificates. When available, primary cause of death was reported.

The following data were collected at baseline: age, gender, clinical centre of investigation, risk categories for HIV infection, duration of HIV infection, AIDS status, HIV related cancer, body mass index (BMI), CD4 cell count, CD4 nadir, creatinine level, HCV antibodies and HCV RNA positivity and hepatitis B surface antigen (HBsAg) positivity. The following ARC were considered at baseline, since they are known to be associated with mortality in the general population (13): history of myocardial infarction, congestive heart failure, cerebrovascular disease, cancer (non-HIV related), chronic pulmonary disease, diabetes, chronic renal disease, and cirrhosis. Although HCV co-infection and HBV co-infection are not ARCs, they were also assessed since liver-related deaths are the leading cause of death in HCV/HIV and/or HBV/HIV co-infected PLHIV (14, 15). BMI was also assessed as it is significantly associated with mortality in the aged general population, even when adjusted for comorbidities (16, 17). When appropriate, adapted International Classification of Disease, 10th Revision (ICD-10) coding algorithms for Charlson comorbidities were used (18). Other comorbidities were extracted using ICD-10 codes and diagnoses available before January 1st, 2008.

History of chronic renal disease was defined as a composite criterion comprising documented history of chronic renal disease in the patient's medical file; or an estimated glomerular filtration rate <60 ml/min/1.73m² as calculated using the Modification of Diet in Renal Disease (MDRD) formula. History of cardiovascular diseases (CVD) was defined as history of myocardial infarction, congestive heart failure or cerebrovascular disease. Chronic pulmonary disease was defined as chronic bronchitis, emphysema, chronic obstructive pulmonary disease, asthma, bronchiectasis, pneumoconiosis, hypersensitivity pneumonitis due to organic dust, chronic respiratory conditions due to chemicals, gases, fumes and vapours; chronic and other pulmonary manifestations due to radiation , chronic drug-induced interstitial lung disorders, or pulmonary heart diseases. Tobacco use was not evaluated due to a high level of missing data. HBV co-infection was defined as at least one positive HBsAg. HCV co-infection was defined as at least one positive anti-HCV antibody and/or detectable HCV-RNA viral load (at the nearest value from baseline).

Statistical Analysis

Means, standard deviations (SD), range, frequency and percentage were used to describe population characteristics, as appropriate. Associations between ARC were verified using Pearson's chi-squared test, Fisher's exact test or Student's t-test, as appropriate. Univariate analysis was performed using Cox's proportional hazards model to generate Hazard Ratios and associated 95% confidence interval. For multivariate analysis, Cox's proportional hazards model for overall mortality was used. The models were systematically adjusted for age, gender, baseline CD4 cell count, HIV viral load, duration of HIV infection and clinical centre. Any other variables with a p-value <0.20 by univariate analysis were also included in the multivariate model. A manual, step-bystep purposeful selection of covariates was used. Hazard ratio significance was determined using the global p-value of the two-sided Wald test. Significance was reached when p<0.05. Possible interactions were investigated between (1) chronic renal disease and CVD; (2) cirrhosis and HCV co-infection and HBV co-infection, if retained in the multivariate model. Model assumptions were checked. Sensitivity analyses were performed when appropriate. Model adequacy was assessed by the May and Hosmer test for adequation. We calculated Harrell's C-statistic, which is a summary measure of discrimination. In order to determine the impact in terms of public health, the etiological fraction (EF) of each of the main independent factors for mortality was calculated. Statistical analyses were performed using SAS version 9.3 (SAS Institute Inc ., Cary, North Carolina, USA).

Results

Among 18304 HIV-1 infected individuals actively followed up as of January 1st, 2008, 1583 were aged 60 or older (8.6%). Among these, 168 were excluded due to the absence of available CD4 cell count one year before or after baseline; 1415 patients (89.4%) were thus included in the study. The baseline characteristics of the study population are presented in Table 1. Most of the patients were male (77.2%) and mean age was 65.7±5.5 years. ARC prevalence was high: 766 patients (54.1%) had at least one ARC and 314 (22.2%) had at least 2. The most prevalent ARCs were chronic renal disease (20.1%), diabetes (14.2%) and CVD (12.2%). Among patients with creatinine clearance below 60 ml/min/1.73m² (18.8% of all patients), median creatinine clearance was 52 ml/min/1.73m² (interquartile range (IQR) [45-57]). Cirrhosis was present in 39 patients (2.8%), related to HCV co-infection in 28.2% of cases and to HBV co-infection in 23.1% of cases.

Table 1.

Baseline characteristics of the 1415 PLHIV aged 60 or more

	Baseline characteristics
		(N=1415)
Sex, male [n (%)]	1093	(77.2)
Age (years) [mean (±SD)]	65.7	(±5.5)
Mode of HIV infection [n (%)]
Heterosexual	630	(44.5)
Homosexual	556	(39.3)
Injecting drug user	9	(0.6)
Other	81	(5.8)
Unknown	139	(9.8)
Duration of HIV infection (years) [mean	11.9	(±6.1)
(±SD)]
ART-experienced [n (%)]	1248	(88.2)
Duration of ART treatment (years) [mean	9.6	(±4.8)
(±SD)] (n=1248)
AIDS [n (%)]	426	(30.1)
HIV-related cancer	148	(10.5)
Age-related comorbidities [n (%)]
Cancer	229	(16.2)
Non HIV-related cancer	94	(6.6)
Cardiovascular diseases	172	(12.2)
Cerebrovascular disease	77	(5.4)
Myocardial infarction	67	(4.7)
Congestive heart failure	42	(3.0)
Chronic pulmonary disease	112	(7.9)
Chronic renal disease	285	(20.1)
Diabetes	201	(14.2)
Cirrhosis	39	(2.8)
HBV co-infection	54	(3.8)
HCV co-infection	92	(6.5)
Body mass index [n (%)] (missing=29)
Obese (≥30kg/m2)	98	(7.1)
Overweight (25-29.9 kg/m2)	389	(28.1)
Normal (18.5 - 24.9 kg/m2)	823	(59.4)
Low (<18.5 kg/m2)	76	(5.5)
CD4 cell count (cells/µl) [mean (±SD)]	507	(±245)
>500 cells/µl	653	(46.2)
350 - 500 cells/µl	364	(25.7)
200 - 349 cells/µl	299	(21.1)
<200 cells/µl	99	(7.0)
CD4 nadir (cells/µl) [mean (±SD)]	210	(±174)
≥200 cells/µl	619	(43.8)
<200 cells/µl	796	(56.2)
HIV Viral Load >50 copies/ml [n (%)] (missing=3)	331	(23.4)
Estimated glomerular filtration rate <60 ml/mn/1.73m2 (MDRD) [n(%)] (missing=28)	260	(18.8)

PLHIV, people living with HIV; HIV, human immunodeficiency virus; AIDS, acquired immune deficiency syndrome;ART, antiretroviral treatment; MDRD, Modification of Diet in Renal Disease; SD, standard deviation

During the five years of follow-up (6225 patient-years), 154 (10.9%) patients died (mortality rate 2.47/100 patient-years). Mean age at death was 70.5 ±7.4 years. Causes of death are resumed in Table 2. One thousand and fifty-three surviving PLHIV were fully followed up until the end of the study period, and 208 patients were lost to follow-up (median follow up among those lost to follow-up was 2.3 years (IQR [2.0-3.4]).

Table 2.

Causes of death among the 154 HIV positive patients from the study population who died during follow-up

Causes of death	N	%
AIDS-related	14	(9.1)
Non-HIV non-viral hepatitis-related cancer	23	(14.9)
Cardiovascular diseases	25	(16.2)
Liver related	7	(4.6)
Non-AIDS-related infections	4	(2.6)
Others	16	(10.4)
Unknown	65	(42.2)

By univariate analysis, the factors associated with overall 5-year mortality were age, low BMI, AIDS status, baseline CD4 cell count (<200 cells/μl), HIV viral load (>50 copies/ml), CD4 nadir (<200 cells/μl), chronic renal disease, history of non-HIV related cancer, CVD, cirrhosis, and HCV co-infection (Table 3). No association was observed between mortality and gender, duration of HIV infection, duration of antiretroviral therapy, HIV-related cancer, chronic pulmonary disease, diabetes or HBV co-infection.

Table 3.

Factors associated with overall 5-year mortality among PLHIV aged 60 years or over by univariate and multivariate analysis (N=1415)

	Univariate analysis			Multivariate analysis (n=1386)
	HR	95% CI	p	aHR*	95%CI	p
Age (years), per additional year	1.07	[1.05 – 1.10]	<10-4	1.05	[1.02 – 1.07]	<10-3
Male sex	0.91	[0.63 – 1.31]	0.599	1.14	[0.77 – 1.70]	0.513
AIDS	1.51	[1.09 – 2.09]	0.013
Non-HIV related cancer	2.61	[1.66 – 4.09]	10-4	1.91	[1.20 – 3.05]	0.007
HIV related cancer	1.15	[0.70 – 1.88]	0.578
Cardiovascular disease	3.23	[2.28 – 4.57]	<10-4	2.40	[1.64 – 3.52]	<10-4
Myocardial infarction	2.85	[1.74 – 4.66]	<10-4
Cerebrovascular disease	2.71	[1.69 – 4.33]	<10-4
Congestive heart failure	3.27	[1.82 – 5.90]	<10-4
Chronic renal disease	2.66	[1.92 – 3.68]	<10-4	2.25	[1.58 – 2.21]	<10-4
Chronic pulmonary disease	1.54	[0.94 – 2.52]	0.083
Diabetes	1.43	[0.96 – 2.14]	0.083
Cirrhosis	4.41	[2.59 – 7.51]	<10-4	2.99	[1.68 – 5.33]	<10-3
HCV co-infection	2.18	[1.35 – 3.52]	0.002	2.00	[1.18 – 3.38]	0.009
HBV co-infection	0.80	[0.33 – 1.95]	0.622
Body mass index
Overweight	0.84	[0.56 – 1.25]	0.380	0.88	[0.58 – 1.33]	0.537
Obese	1.12	[0.60 – 2.11]	0.717	1.51	[0.80 – 2.86]	0.205
Low	3.62	[2.28 – 5.75]	<10-4	2.42	[1.46 – 4.01]	<10-3
CD4 cell count (cells/µl), ≤200	2.42	[1.52 – 3.83]	<10-3	2.23	[1.36 – 3.65]	0.002
CD4 nadir (cells/µl), ≤200	1.55	[1.11 – 2.17]	0.010
HIV viral load (copies/ml), >50	1.47	[1.04 – 2.08]	0.029
Duration of HIV infection, per one year	1.00	[0.97 – 1.02]	0.898
Duration of ART, per one year	1.01	[0.98 – 1.05]	0.440

95% CI, 95% confidence interval; HR, Hazard Ratio; aHR, Adjusted Hazard Ratio; ART, antiretroviral treatment; HIV, human immunodeficiency virus; AIDS, acquired immune deficiency syndrome; HBV, hepatitis B virus; HBC, hepatitis C virus;

^*Also adjusted for duration of HIV infection and clinical centre

Patients with CVD were more likely to have chronic renal disease (31% versus 19% in patients without CVD; p<10-4). Similarly, patients with chronic renal disease were more likely to have CVD (20% versus 10% in patients without chronic renal disease; p<10-4). Diabetes was observed in 21% of patients with chronic renal disease (versus 12% in those without; p=0.0002) and in 26% of patients with CVD (versus 13% in those without; p<10-4).

By multivariate analysis, older age, chronic renal disease, non-HIV related cancer, CVD, cirrhosis, HCV co-infection, low BMI and CD4 cell count (<200 cells/μl) were significantly associated with 5-year mortality (Table 3). In addition, there was no significant interaction between CVD and chronic renal disease or between cirrhosis and HCV co-infection. Low BMI and chronic renal disease did not satisfy the proportional hazard assumption. Therefore, we analyzed relevant HR trends over time. For low BMI, the aHR decreased over time, from an aHR of 6.36 (95%CI[3.30–12.28]; p<10-4) in the first 20 months, to an aHR of 0.79 (95%CI[0.24–2.60]; p=0.69) in the last 20 months. For history of chronic renal diseases, the aHR increased over time, from an aHR of 0.90 (95%CI[0.45-1.79]; p=0.76) in the first 20 months, to an aHR of 4.07 (95%CI[2.36–7.02]; p<10-4) in the last 20 months (figures 1a and 1b).

Figure 1a.

Adjusted hazard ratio over time for chronic renal disease

Figure 1b.

Adjusted hazard ratio over time for Low Body mass Index

Model assessment by the May and Hosmer test showed good fit. Model discrimination showed a C-statistic of 0.74, indicating good discriminatory capacity. Furthermore, defining HCV co-infection (n=92) as HCV-RNA positivity alone (n=47), and defining chronic renal disease as only an estimated glomerular filtration rate <60 ml/mn/1.73m² (n=260, 18.8%) did not significantly alter parameter estimates. Etiological fractions for chronic renal disease, CVD, cirrhosis and HCV co-infection were 20%, 15%, 5% and 6% respectively. These represent the proportion of the outcome (5-year mortality) that it would be possible to avoid if exposure were removed from the population.

Discussion

In this large multicentre prospective cohort of aged PLHIV (N=1415) followed up for five years in the late HAART era (from January 2008 to December 2012) in France, the prevalence of ARC was high, and most had a significant negative impact on survival. Low CD4 cell count (<200 cells/μl), history of chronic renal disease, CVD, non-HIV related cancer, cirrhosis, HCV co-infection and low BMI were the main prognostic factors.

The role of chronic renal disease on mortality has rarely been evaluated in PLHIV (19, 20, 21). The prevalence of chronic renal disease is significantly higher in PLHIV compared to the general population in each age stratum (5). In a large Italian cohort, prevalence of chronic renal disease was 24.6% in PLHIV aged 60 or more, versus 0.49% in the general population in the same age stratum (5). In our study, chronic renal disease was the most prevalent ARC (20.1%) and was associated with a more than two-fold increase in five-year mortality risk. Furthermore, this risk increased over time with an aHR of 4.07 (95%CI[2.36–7.02]; p<10-4) in the 20 last months period of follow-up, probably due to the progression of renal function impairment. In addition, these results did not change significantly when chronic renal disease was defined as only an estimated glomerular filtration rate <60 ml/mn/1.73m². In line with our results, two others cohorts also reported a higher risk of death in PLHIV with pre-existing chronic renal disease at the time of HAART initiation (21) and/or occurring during HAART (19). Conversely, unlike our findings, the prevalence of chronic renal disease was low (around 3%), probably due to the inclusion of younger PLHIV, and thus, the lower number of events likely limited the power of the study.

PLHIV have an increased relative risk of CVD compared with non-HIV patients (5, 22, 23, 24). In a cross-sectional national survey conducted in France in 2000, 2005 and 2010 among HIV-infected patients, the death related to CVD increased moderately over time (7, 8 and 10% respectively) and accounted for 12% in patients aged 60 years and more in 2010 (8). In our study, the prevalence and proportion of deaths related to CVD were 12.2% and 16.2%, respectively, and were similar to others studies in the same age stratum of PLHIV (5, 8). In the general population, chronic renal disease is a risk factor for CVD (25), and concurrently, CVD may promote chronic renal disease (26). In our study, patients with CVD were also more likely to have chronic renal disease, but these two ARCs were both independently associated with 5-year mortality. Furthermore, we found no significant interaction between these two diseases.

Diabetes is a common risk factor for both chronic renal disease and CVD. Several studies have addressed the risk of diabetes mellitus in the HIV-infected population but results are conflicting (27). In a Danish nationwide population-based cohort study (27), HIV-infected individuals had no increased risk of diabetes mellitus compared to the general population after the year 1998. In our study, the prevalence of diabetes was 14.2% and was comparable to the general population of similar age (28). Diabetes is associated with overall mortality in the general population (29). Diabetes tended to be associated with mortality only by univariate analysis in our study. This may be explained by the significant association between diabetes and both chronic renal disease (p<10-3) and CVD (p<10-4).

PLHIV are at increased risk of cancer, whether HIV–related (30) or not (31). We found that the increased risk mortality for cancer was due to non-HIV related cancers, and HIV-related cancers were not associated with 5-year mortality among aged PLHIV. In the HAART era, HIV-related cancers have better prognosis than before (32, 33). Nevertheless, by selecting a prevalent population with long duration of HIV infection and receiving multiple HAART regimens, instead of an “incident” HAART-naïve population, we may have selected patients who survived HIV-related cancer in the first years of their care. Similarly, a history of AIDS-related events, which included HIV-related cancers, was not retained in the final model. This means that, despite their impact at HAART initiation, these factors no longer have a major prognostic impact for the majority of aged PLHIV with long duration of HIV infection who recover satisfactory immunity.

Liver-related diseases are the leading cause of death in HIV/HCV co-infected patients (8,15). Despite a low prevalence of cirrhosis (2.8%) and HCV co-infection (6.5%) in our study, both were independently associated with mortality. This low prevalence may be explained by the low prevalence of injecting drug users in this population (0.6%). The independent adverse impact of HCV co-infection may be the consequence of the infection itself, but may also reflect a population at higher risk, since patients who had recovered from HCV co-infection also remained at higher risk of death. Nevertheless, the fact that cirrhosis was significantly related mortality, and that 28.2% of cirrhotic patients were HCV-positive, strongly encourages treatment of HCV co-infection in HIV patients, especially since highly efficient new therapies are available in this indication (34). In contrast, HBV co-infection was not associated with mortality, probably due to both a low prevalence (3.8%) resulting in a possible lack of power, and to the long term efficacy of antiretroviral drugs such as tenofovir on HBV infection (35).

In the late HAART era, more than 80% of treated PLHIV have a good immuno-virological control of their HIV infection in high income countries (2). A large proportion of our patients were also highly immuno-virologically controlled: 93% had CD4 cell count >200 cells/μl, and 77% had an undetectable HIV viral load. Only 9.1% of deaths were AIDS related. Nevertheless, CD4 cell count <200 cells/μl remained associated with mortality, indicating the importance of controlling HIVrelated immunodepression. HIV infection and antiretrovirals themselves are associated with a higher risk of chronic renal disease and metabolic complications such as insulin resistance and diabetes mellitus, dyslipidaemia, hypertension which all can increase the risk of ischaemic cardiovascular diseases (36, 37). Among the actually often prescribed antiretroviral drugs, some of them as tenofovir or atazanavir have been associated with renal function impairment, and long exposure to protease inhibitors with a higher cardiovascular risk and diabetes (38, 39). In our study, despite a long mean time of exposure (9.6 ±4.8 years), the duration of antiretroviral therapy was not associated with mortality. However, the use of specific antiretrovirals and their role on the occurrence of comorbidities as chronic renal disease and/or CVD have not been evaluated.

Low BMI was also associated with a higher risk of mortality, especially in the first 20 months after baseline and probably identified the frailest patients who are likely to die in the early period of follow-up. As in non-HIV older patients (40), obesity did not seem to negatively affect overall mortality in our study.

No study to date has evaluated the impact of multiple ARCs on overall mortality in a large representative cohort of aged 60 or more PLHIV and in a period limited to the late HAART era. The advantage of choosing a prevalent cohort is that it is closest to real-life data, based on a population seen in routine practice. Taking into account all new cases in the cohort followed between January 2008 and January 2013, or reaching the age of 60, would have overestimated the proportion of “incident” cases which, in this population, have a late presentation and thus, a different prognosis (41, 42), or may have biased age at baseline towards 60 years. Furthermore, model assessment showed good fit and good discriminatory capacity. Nonetheless, our study suffers from some limitations that deserve to be acknowledged. Cognitive disorders were not included because of the lack of reliable diagnosis in the study data. Aged PLHIV did not undergo systematic geriatric assessment in the Dat'AIDS cohort. However, it may have a role and would be a useful addition to the model in future studies, provided the assessment can be performed in the majority of patients.

It would also have been interesting to study the association between ARC and specific mortality. Unfortunately, the causes of death were not reliably recorded for all patients, precluding such an analysis. However, this did not impact on the results of the present study.

Conclusion

In conclusion, in this French representative cohort of PLHIV aged 60 or over, ARC were identified as risk factors for mortality, highlighting the necessity to implement preventive strategies to avoid their appearance and/or progression. Special attention should be paid to chronic renal disease and CVD. Indeed, apart from well-known risk factors described in the general population (43, 44), these two ARCs could be related to HIV-specific factors, such as cumulative long term HAART toxicity. Moreover, their etiological fractions were the highest among the independent prognostic factors found in this study (20% for chronic renal disease and 15% for CVD). This suggests that the implementation of preventive measures would significantly reduce mortality.

Key Points

• •Large French representative cohort of people living with HIV aged 60 or more
• •High comorbidity prevalence
• •Chronic renal disease and cardiovascular diseases were main prognostic factors and had the highest etiological fraction

Acknowledgements: Dat'AIDS scientific committee: Dellamonica P., Pugliese P., Poizot-Martin I., Cuzin L., Yazdanpanah Y., Raffi F., Cabié A., Garraffo R., Delpierre C., Allavena C., Katlama C., Valantin M.A., Duvivier C., Hoen B., Peytavin G., Jacomet C., Rey D., Delobel P., Cheret A., Chidiac C., Isnard-Bagnis C., Cotte L., Peyramond D., Bani-Sadr F., Joly V., Jovelin T., Saune K., Roger P.M., Chirouze C., May T. Dat'AIDS Study Group: P. Enel, V. Obry-Roguet, O. Faucher, S. Bregigeon , A. Ménard, I. Poizot- Martin, (Marseille) ; B. Marchou, P. Massip, E. Bonnet, M.Obadia, M. Alvarez, L. Porte, L. Cuzin, M. Chauveau, I. Lepain (Toulouse) ; P. Pugliese, L. Bentz, C. Ceppi, E. Cua, J. Cottalorda, J. Durant, S. Ferrando, JG Fuzibet, R. Garraffo, V. Mondain, A. Naqvi, I. Perbost, S. Pillet, B. Prouvost-Keller, C. Pradier, S.Wehrlen-Pugliese, P.-M. Roger, E. Rosenthal, P. Dellamonica (Nice) ; E. Billaud, C. Allavena, T.Jovelin, C. Guerbois, F. Raffi (Nantes) ; A Cheret, P. Choisy (Tourcoing) ; C. Duvivier (Pasteur Necker), M.A. Valantin, R. Agher, C. Katlama (Paris Pitié Salpétriere) ; A. Cabié, S. Abel, S. Pierre-François, B. Liautaud (Fort de France) ; D. Rey, E. Ebel, P. Fischer, M. Partisani (Strasbourg) ; C. Chirouze, Q. Gardiennet (Besançon); F. Bani-Sadr, C. Rouger, J.L. Berger, Y. N'Guyen, D. Lambert, I. Kmiec, M. Hentzien (Reims). The Nadis® EMR is developed and maintained with support from Fedialis Medica and ViiV Healthcare.

Authors contributions: All authors participated in the design of the study protocol and data collection. MH performed the data management. MH and MD performed statistical analyses. MH, MD and FBS wrote the first manuscript draft. All authors participated in interpretation of the data and writing of the final manuscript and all authors approved the final manuscript. FBS and MD were responsible for the overall supervision of the study.

Conflicts of interest: None

Sources of support: Maxime Hentzien received a one-year government grant in the context of a Master programme for Research.