Abstract
BACKGROUND: Heavy alcohol use, including binge drinking, is associated with high morbidity and mortality among men who have sex with men (MSM). Self-reported alcohol measures may lead to inaccurate estimates due to recall and social desirability biases. Objective alcohol biomarkers like phosphatidylethanol (PEth) can be used to corroborate self-report and could help to inform treatment approaches and research strategies for alcohol using MSM. METHODS: From 2015 to 2020, alcohol using MSM ≥18 years were enrolled in a randomized controlled trial evaluating the efficacy of naltrexone in reducing binge drinking. Using this trial's baseline data, we applied multivariable logistic regression to identify the correlates of high PEth levels (i.e., ≥87 ng/ml) and concordance between PEth levels and self-reported heavy drinking. RESULTS: Of 118 MSM, 64% had PEth levels ≥87 ng/ml and 72% had PEth levels that were concordant with self-reported heavy alcohol use. Factors significantly associated in separate models with elevated PEth levels were income ≥$60,000 (adjusted odds ratio [aOR] = 4.09; 95% CI = 1.13 to 14.82), being employed (aOR = 4.04; 95% CI = 1.45 to 11.32), episodic cannabis use (aOR = 4.63; 95% CI = 1.27 to 16.92), and any alcohol/substance use prior to or during anal intercourse (aOR = 2.52; 95% CI = 1.08 to 5.90). Living with HIV was associated with significantly lower odds of elevated PEth levels (aOR = 0.23; 95% CI = 0.09 to 0.61). Factors associated with significantly higher concordance between PEth levels and self-reported heavy alcohol use included at least weekly use of poppers (aOR = 6.41; 95% CI = 1.27 to 32.28) and polysubstance use (aOR = 2.53; 95% CI = 1.02 to 6.27). Living with HIV was associated with lower odds of concordance (aOR = 0.36; 95% CI = 0.14 to 0.97). CONCLUSIONS: PEth may enhance the detection of heavy drinking among MSM, including the identification of subpopulations that may benefit from targeted alcohol reduction interventions. However, PEth values for MSM living with HIV showed modest concordance with self-reported alcohol use and may need to be supplemented with additional biomarkers or evaluated against a different cutoff.