Objective: Discuss issues related to time-varying exposures using as an example the recently meta-analyzed literature (Baliunas et al. in Int J Public Health, 2009) on alcohol consumption and risk of HIV infection.
Methods: Cataloged sources of bias and imprecision in the context of time-varying exposures.
Results: Confounding, selection, or measurement bias may occur when standard regression approaches are used to estimate effects of time-varying exposures. The reviewed literature on alcohol consumption and HIV infection suffer from one or more of these biases.
Conclusions: Detailed prospective data and thoughtful implementation of appropriate statistical methods are needed to obtain unbiased estimates of time-varying exposures, such as alcohol consumption.