Abstract
BACKGROUND: Proximal tubular dysfunction (PTD) is common in HIV-positive persons and has been associated with tenofovir disoproxil fumarate (TDF). However, few studies have assessed the natural history PTD in HIV-positive and -negative individuals, or the association of PTD with the subsequent trajectory of directly measured glomerular filtration rate (mGFR). METHODS: We followed 192 HIV-positive and 100 HIV-negative, nondiabetic participants for 3 years. We measured 3 PTD markers (normoglycemic glycosuria, fractional excretion of phosphorus, and tubular proteinuria) and mGFR (by iohexol disappearance from serum) annually. We used univariate and multivariate generalized estimating equation logistic regression to identify factors associated with PTD across all visits and linear mixed effects models to assess the association between baseline PTD and mGFR slope. RESULTS: Compared with HIV-negative participants, HIV-positive persons that were not taking antiretroviral therapy were at increased risk of PTD (adjusted odds ratio 3.33; 95% confidence interval: 1.65 to 6.71), whereas those taking a TDF-based or a TDF-sparing regimen were not at significantly increased risk of PTD. Among HIV-positive participants, uncontrolled viremia was a strong correlate of PTD. Forty-nine of 55 (89%) participants with PTD at baseline had at least 1 subsequent visit without PTD. There was no association between baseline PTD and rate of decline in mGFR over time. CONCLUSIONS: Poorly controlled HIV may be a stronger risk factor for PTD than TDF use. The individual-level variability of the PTD markers over time was high, potentially limiting their usefulness for routine screening in unselected patients. Baseline PTD was not associated with subsequent mGFR slope.