“…Ordinary least squares regression was used to determine whether the inclusion of race or surrogates of race, APOL1 variants and genetic African ancestry, improved ability to predict log-transformed serum creatinine. Additional variables considered for model inclusion included age at study entry, birth sex, BMI [32], hypertension [32], diabetes [33], hepatitis C infection [34], hepatitis B infection [34], history of intravenous drug use [35], CD4 + cell count, log-transformed HIV viral load, tenofovir use, protease inhibitor use, indinavir use, atazanavir use, lopinavir use, integrase inhibitor use, and overall antiretroviral therapy use. Model residuals were assessed for functional form of continuous variables, and all models were tested for collinearity of variables using the variance inflation factor.…”