Background: Oropharyngeal candidiasis (OPC) is the most frequently observed oral infection in HIV-infected individuals. Historically, lower CD4 counts have been associated with an increased prevalence of OPC in HIV-infected patients, but HIV viral load has also recently been recognized as a possible predictive factor.
Objective: We examined the impact of viral load and blood CD4 cell count on the occurrence of OPC using modern exploratory statistical analyses.
Methods: The exploratory and inferential methods of classification and regression trees (CARTs) and logistic regression were used to compare the impact of viral load and CD4 cell counts on OPC status in 161 HIV-infected individuals from an outpatient clinic population in New Orleans.
Results: The use of stepwise logistic regression and CART to classify individual OPC status both identified viral load as the most important covariate, followed by CD4 cells counts. Age, sex, and highly active antiretroviral therapy use were also found to be associated with OPC status.
Conclusions: These data strongly suggest that low viral load distinguishes those not at risk for OPC with high viral load, which also includes a heterogeneous set of predictors for OPC status, has the highest impact on OPC classification.