Allen John P, Litten Raye Z
National Institute on Alcohol Abuse and Alcoholism, Vienna, Virginia, USA.
Alcohol Clin Exp Res. 2003 Oct;27(10):1667-70. doi: 10.1097/01.ALC.0000091224.78880.47.
Biochemical markers of heavy drinking are playing increasingly prominent roles in alcohol treatment efficacy studies, especially in those designed to evaluate medications. Among these roles are serving as inclusion or exclusion criteria for research participants, corroboration of self-report of drinking status, assessment of the safety of the agent being evaluated, and determination of treatment outcome.
Recent alcohol medication development trials that included biomarker information were reviewed and critiqued from the perspectives of how biomarker measures were used and how findings on them were reported.
Although generally the application of biomarkers as inclusion criteria is not recommended, they may aid in exclusion of potential subjects (e.g., elevated liver function measures in trials of agents that could result in liver damage). Biomarkers are most commonly used as indicators of outcome, usually serving as secondary outcome variables. The relationship of outcome findings on biomarker and self-report measures is positive, but only moderate. As used to date, biomarkers of drinking tend to be less sensitive than well-standardized and properly administered self-report measures. Nevertheless, they do provide a useful, unique source of information on drinking status.
The contribution of biomarkers to alcoholism clinical research would be enhanced if certain design strategies were incorporated into their application and if critical information were included in the research publication. This article offers a series of recommendations to improve on their use in a research context.
