ABSTRACT Alcohol use is common among people living with HIV (PWH) and is a consistent predictor of poor antiretroviral therapy (ART) adherence. It has also been associated with HIV virologic failure and mortality, but these results have been somewhat mixed and the relationships between level of alcohol use and these outcomes are not well defined. Interventions to reduce alcohol use among PWH have shown modest or mixed results. Thus, the level of alcohol use needed to cause harm for PWH and the efficacy of interventions to reduce the harm are uncertain. This uncertainty stems in part from the near ubiquitous reliance on self-report to measure alcohol use, which may be inaccurate due to recall bias or social desirability bias and lead to spurious or obscured results, and from inconsistent alcohol measurement. Objective biomarkers can be leveraged to supplement self-report or as alternatives to self-report. The leading alcohol biomarker is phosphatidylethanol (PEth), which can be found in whole blood or dried blood spots, detects prior 2-4 weeks alcohol use, and is correlated with total alcohol consumed. Several research studies of PWH have conducted or plan to conduct PEth testing, making possible an unprecedented opportunity to pool a large number of observations with PEth and HIV data to provide definitive answers to these questions. We propose the Biomarkers for Alcohol/HIV Research (BAHR) Study to gather and pool these data, which will include more than 8,000 PWH with 15,000 observations, and use PEth to resolve past alcohol/HIV research uncertainties, to guide future interventions, and to provide measurement guidance for future research. We will determine the relationship between PEth- measured alcohol use and HIV virologic failure and mortality risk among PWH who are on ART using data from six studies (Aim 1). We will conduct individual participant data meta-analyses of alcohol/HIV intervention studies (15 have agreed to participate) to examine evidence of the efficacy of the interventions to reduce PEth- measured alcohol use, and their further impact on virologic failure (Aim 2). For both these aims, we will compare the results using PEth alone to those obtained using self-report alone, and self-report combined with PEth, to guide future alcohol measurement in research. Lastly, because PEth is expensive and inaccessible in low-resource and non-research settings, we will examine the predictive value of a combination of common laboratory tests as a low-cost alternative to PEth testing, leveraging the extensive testing being conducted in a 6000-person study (Aim 3). These analyses will provide tangible advancements for the alcohol/HIV field, namely definitive answers on the relationship of alcohol use to HIV virologic failure and mortality risk; the efficacy of alcohol interventions studies to reduce alcohol use and decrease virologic failure; information on the comparability of results using biomarkers versus self-report to measure alcohol use; and evidence on the predictive ability of a low-cost alcohol risk score for further testing and potential increased availability in low-resource and non-research settings.

NARRATIVE We propose to leverage several research studies of persons with HIV (PWH) that measured phosphatidylethanol (PEth), a biomarker of how much alcohol someone has consumed over the past month that is found in the blood, to investigate key questions about how alcohol use affects the health of persons with HIV and how to reduce alcohol use. We will also use these data to explore other ways to measure alcohol use via common blood tests.