Chronic kidney disease (CKD) is defined either as evidence of kidney damage or an estimated glomerular filtration rate (eGFR) below 60 ml/min/1.73 m2 persisting for at least 3 months . In the general population, Di Angelantonio et al.  pointed out that strong associations had been reported between the occurrence of nondialysis-dependent CKD and the subsequent incidence of cardiovascular events in high-risk groups, including individuals with pre-existing ischemic cardiovascular disease (CVD), whereas evidence linking renal dysfunction to the risk of CVD in individuals without known cardiovascular or renal disease was comparatively sparse and uncertain. In HIV-infected individuals, the prevalence of CKD increases over time , and some studies have shown an association between decreased kidney function and the risk of cardiovascular events [4–6]. However, these results are difficult to interpret because these studies examined various cardiovascular outcomes and included individuals with or without pre-existing CVD. Furthermore, the results were not always adjusted for traditional cardiovascular risk factors (CvRFs), and the proportion of individuals of African origin differed widely. Here, we examined whether impaired kidney function is an independent risk factor for myocardial infarction (MI) in HIV-infected individuals without prior MI, taking into account traditional CvRF, ethnic origin, HIV-related parameters and antiretroviral drug exposures.
A nested case–control study was conducted within the French Hospital Database on HIV , using data from January 2000 to December 2006 [8,9]. Cases were individuals who presented a prospectively recorded and validated first MI, as confirmed by a cardiologist. Up to five controls matched for age, sex and clinical centre were randomly selected, among HIV-infected individuals with no history of MI. When updating this nested case–control study to include individuals diagnosed with MI up to December 2009, we also collected serum creatinine levels and ethnic origin for all cases diagnosed between January 2006 and December 2009, and their matched controls. Glomerular filtration rate (GFR) was estimated using the Modification Diet in Renal Disease equation . To include all individuals in the analyses, missing values were imputed as described previously . Conditional logistic regression models were used to assess the association between eGFR below 60 ml/min/1.73 m2 and the risk of a first MI. The first step was to select, among the different ways of modelling continuous variables, the one with the lowest univariable Akaike's information criterion. After determining the crude impact of eGFR below 60 ml/min/1.73 m2 on the risk of MI, we used bivariable models, including eGFR and each individual CvRF [hypertension, smoking, family history of coronary artery disease (CAD), hypercholesterolaemia, high-density lipoprotein (HDL)-cholesterol (HDLc) level, diabetes, triglyceride level, BMI, and intravenous drug use) or each HIV-related parameter (plasma HIV-1 RNA level, current CD4+ and CD8+ T-cell counts, CD4+ T-cell nadir, CD4+/CD8+ T-cell ratio, AIDS status, cumulative exposure to protease inhibitors), to identify parameters that altered the crude eGFR odds ratio (OR) by at least 5%. The selected parameters were then included in a model, along with eGFR, and the second step was repeated to identify other parameters affecting the eGFR OR. All the parameters that were significantly associated with the risk of MI in bivariate analyses were then included in the model. Finally, to avoid over-adjustment, parameters that were not significantly associated with the risk of MI in the last model were removed.
Overall, 256 cases and 791 controls were included in the analyses. The cases were mostly men (87%), and their median age was 48 years. eGFR was below 60 ml/min/1.73 m2 in 7% of cases and 5% of controls. Only one patient in each group had a very low eGFR (<30 ml/min/1.73 m2). Compared to the controls, the cases were less frequently from sub-Saharan Africa (9 versus 15%; P < 0.0065). All traditional CvRFs, except for BMI, were more frequent in cases than in controls. BMI was below 21 kg/m2 in 22% of cases and 34% of controls (P = 0.0003). Plasma HIV-1 RNA level was above 50 copies/ml in 34% of cases and 26% of controls. The CD4+ T-cell nadir was lower in cases than in controls (P = 0.0065). More cases (32%) than controls (25%) had a CD8+ T-cell count above 1150/μl. Cases had higher cumulative exposure to protease inhibitors [5.06 years (2.32–8.40) versus 4.03 years (1.66–7.56)]. Cumulative exposure to tenofovir and to atazanavir did not differ between the two groups. In the univariable model, eGFR below 60 ml/min/1.73 m2 was associated with an increased risk of MI, with a crude OR of 1.22, without reaching statistical significance [95% confidence interval (CI) 0.90–1.66]. In multivariable models, hypertension and the log2 CD4+ T-cell nadir had the biggest impact on the change between the crude and adjusted OR (adjusted OR 1.06). When smoking and HDLc were included, the OR fell to 0.98. In the fully adjusted model, the OR was 0.99 (Fig. 1). In sensitivity analyses, the addition of diabetes, or cumulative protease inhibitor exposure, or exposure to tenofovir or to atazanavir, did not affect the results.
As in the study by Choi et al. , only a very small percentage of our population had impaired kidney function. In adjusted analyses, we found that impaired kidney function was not significantly associated with an increased risk of MI in HIV-infected individuals with no known pre-existing CAD.
The OR found here is very close to 1 and therefore differs markedly from the values obtained in previous studies. The difference cannot be explained by a power issue, because an OR of 1 would still be non-significant in the larger study. George et al.  and Campbell et al.  both included individuals with a history of CVD (26% and 44%, respectively) and used composite CVD endpoints. The proportion of individuals of African origin ranged from 47%  to 84%  in previous studies, compared to only 13% in our study.
In conclusion, we found no association between impaired kidney function and the risk of MI in HIV-infected individuals without pre-existing CAD. Hypertension, a low CD4+ T-cell nadir, smoking, and a low HDLc level were the most important confounders. Differences between studies suggest that any link between impaired kidney function and CVD might be influenced by genetic factors, as reflected by ethnic origin and by pre-existing CVD.
We thank the participants and research assistants of FHDH-ANRS CO4. The full list of investigators participating in the Clinical Epidemiology Group of the FHDH-ANRS CO4 can be found at http://http://www.ccde.fr. (Accessed 20 May 2014). FHDH-ANRS CO4 is supported by the ANRS, INSERM, and the French Ministry of Health.
Each author thoroughly contributed to the study and to the manuscript and has seen and approved the submitted version.
This work was supported by ANRS (French National Agency for Research on AIDS and Viral Hepatitis). The funding source had no involvement in the design or execution of this study.
Conflicts of interest
F.B. declares receiving travels grants from Astrazeneca. D.C. has received consultancy fees, honoraria or research grants from various pharmaceutical companies including Bristol-Myers-Squibb, Gilead Sciences, Janssen-Cilag, Merck-Sharp & Dohme-Chibret and ViiV Heatlthcare. L.C. declares receiving travels grants from Abbot, Boehringer-Ingelheim, Bristol-Myers-Squibb, Gilead Sciences, Janssen-Cilag, Merck-Sharp & Dohme-Chibret, Roche, ViiV Heatlthcare, receiving grants pending from Merck-Sharp & Dohme-Chibret, ViiV, and reports consultancy fees from Mylan. J.G. declares no conflict of interest. S.L. declares receiving travels grants from Bristol-Myers-Squibb. M.M.-K. declares no conflict of interest. M.P. declares no conflict of interest. A.S. declares receiving consultancy fees and travels grants from Gilead Sciences.
1. K/DOQJ clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification
. Am J Kidney Dis
2. Di Angelantonio E, Danesh J, Eiriksdottir G, Gudnason V. Renal function and risk of coronary heart disease in general populations: new prospective study and systematic review
. PLoS Med
3. Winston J, Deray G, Hawkins T, Szczech L, Wyatt C, Young B. Kidney disease in patients with HIV infection and AIDS
. Clin Infect Dis
4. George E, Lucas GM, Nadkarni GN, Fine DM, Moore R, Atta MG. Kidney function and the risk of cardiovascular events in HIV-1 infected patients
5. Choi AI, Li Y, Deeks SG, Grunfeld C, Volberding PA, Shlipak MG, et al. Association between kidney function and albuminuria with cardiovascular events in HIV-infected persons
6. Campbell LJ, Desai M, Hegazi A, Grunfeld C, Volberding PA, Shlipak MG. Renal impairment is associated with coronary heart disease in HIV-positive men
. HIV Clin Trials
7. Mary-Krause M, Grabar S, Lièvre L, Abgrall S, Billaud E, Boue F, et al. Cohort profile: French hospital database on HIV (FHDH-ANRS CO4)
. Int J Epidemiol
[Epub ahead of print].
8. Lang S, Mary-Krause M, Cotte L, Gilquin J, Partisani M, Simon A, et al. Impact of individual antiretroviral drugs on the risk of myocardial infarction in human immunodeficiency virus-infected patients: a case-control study nested within the FHDH ANRS Cohort CO4
. Arch Intern Med
9. Lang S, Mary-Krause M, Simon A, Partisani M, Gilquin J, Cotte L, et al. HIV replication and immune status are independent predictors of the risk of myocardial infarction in HIV-infected individuals
. Clin Infect Dis
10. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, et al. Chronic kidney disease epidemiology collaboration: using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate
. Ann Intern Med