Kidney Dysfunction and Markers of Inflammation in the Multicenter AIDS Cohort Study

Abstract

Background: Human immunodeficiency virus (HIV)-infected individuals are at higher risk for chronic kidney disease than HIV-uninfected individuals. We investigated whether the inflammation present in treated HIV infection contributes to kidney dysfunction among HIV-infected men receiving highly active antiretroviral therapy.

Methods: The glomerular filtration rate (GFR) was directly measured (using iohexol) along with 12 markers of inflammation in Multicenter AIDS Cohort Study participants. Exploratory factor analysis was used to identify inflammatory processes related to kidney dysfunction. The estimated levels of these inflammatory processes were used in adjusted logistic regression analyses evaluating cross-sectional associations with kidney function outcomes.

Results: There were 434 HIV-infected men receiving highly active antiretroviral therapy and 200 HIV-uninfected men. HIV-infected men were younger (median age, 51 vs 53 years) and had higher urine protein-creatinine ratios (median, 98 vs 66 mg/g) but comparable GFRs (median, 109 vs 106 mL/min|1.73 m(2)). We found an inflammatory process dominated by markers: soluble tumor necrosis factor receptor 2, soluble interleukin 2 receptor α, soluble gp130, soluble CD27, and soluble CD14. An increase of 1 standard deviation in that inflammatory process was associated with significantly greater odds of GFR ≤90 mL/min/1.73 m(2) (odds ratio, 2.0) and urine protein >200 mg/g (odds ratio, 2.3).

Conclusions: Higher circulating levels of immune activation markers among treated HIV-infected men may partially explain their higher burden of kidney dysfunction compared with uninfected men.

Keywords: HIV infection; chronic kidney disease; glomerular filtration rate; immune activation; inflammatory markers.

Figure 1.

Percentile box plots displaying the distribution of glomerular filtration rate (GFR) and urine protein-creatinine ratio (uPCr) among the sample of human immunodeficiency virus (HIV)–infected (HIV⁺) and HIV-uninfected (HIV⁻) men from the Multicenter AIDS Cohort Study. The shaded areas represent proportion with GFR values ≥140 mL/min/1.73 m² or ≤90 mL/min/1.73 m² (right) and proportion with uPCr >200 mg/g (left), comparing HIV-infected with HIV-uninfected samples.

Figure 2.

Standardized biomarker profiles showing the average standardized levels of the 12 inflammatory biomarkers stratified by human immunodeficiency virus (HIV) serostatus and by glomerular filtration rate (GFR; ≤90 vs >90 mL/min/1.73 m²). Upper panel shows biomarker values standardized to an overall mean and standard deviation for each biomarker, ignoring HIV serostatus. Lower panel shows biomarker values standardized to a serostatus-specific mean for each biomarker. Abbreviations: CCL2, monocyte chemoattractant protein 1; CRP, C-reactive protein; CXCL8, interleukin 8; CXCL10, interferon γ–induced protein 10; IL-6, interleukin 6; IL-10, interleukin 10; sCD14, soluble CD14; sCD27, soluble CD27; sgp130, soluble gp130; sIL2rα, soluble interleukin 2 receptor α; sTNFr2, soluble tumor necrosis factor (TNF) receptor 2.

Figure 3.

Distribution of inflammatory process levels for each of the 3 identified inflammatory processes. Upper panel, distribution stratified by kidney function: low filtration (glomerular filtration rate [GFR] ≤90 mL/min/1.73 m²) hyperfiltration (GFR ≥140 mL/min/1.73 m² for men aged ≤40 years, subtracting 1 mL/min/1.73 m² for each year of age >40), or normal filtration (GFR >90 to <140 mL/min/1.73 m²). Lower panel, distribution stratified by normal versus high urine protein levels (urine protein-creatinine ratio, ≤200 vs >200 mg/g).

Figure 4.

Results from adjusted logistic regressions showing the estimated odds ratios (ORs) for the associations between the levels of the 3 inflammatory processes and 6 outcomes: low kidney filtration (glomerular filtration rate [GFR] ≤90 mL/min/1.73 m²), hyperfiltration (GFR ≥140 mL/min/1.73 m² for men aged ≤40 years, subtracting 1 mL/min/1.73 m² for each year of age >40), high urine protein-creatinine ratio (>200 mg/g), diabetes, hypertension, and hepatitis C virus (HCV) infection. Models were adjusted for all the predictors noted. Effect estimates associated with human immunodeficiency virus (HIV) serostatus are also presented, without adjustment for inflammatory process levels.