Fatty acid ethyl esters in hair as alcohol markers: estimating a reliable cut-off point by evaluation of 1,057 autopsy cases.

Alcohol abuse is a widespread problem, especially in Western countries. Therefore, it is important to have markers of alcohol consumption with validated cut-off points. For many years research has focused on analysis of hair for alcohol markers, but data on the performance and reliability of cut-off values are still lacking. Evaluating 1,057 cases from 2005 to 2011, included a large sample group for the estimation of an applicable cut-off value when compared to earlier studies on fatty acid ethyl esters (FAEEs) in hair. The FAEEs concentrations in hair, police investigation reports, medical history, and the macroscopic and microscopic alcohol-typical results from autopsy, such as liver, pancreas, and cardiac findings, were taken into account in this study. In 80.2 % of all 1,057 cases pathologic findings that may be related to alcohol abuse were reported. The cases were divided into social drinkers (n = 168), alcohol abusers (n = 502), and cases without information on alcohol use. The median FAEEs concentration in the group of social drinkers was 0.302 ng/mg (range 0.008-14.3 ng/mg). In the group of alcohol abusers a median of 1.346 ng/mg (range 0.010-83.7 ng/mg) was found. Before June 2009 the hair FAEEs test was routinely applied to a proximal hair segment of 0-6 cm, changing to a routinely investigated hair length of 3 cm after 2009, as proposed by the Society of Hair Testing (SoHT). The method showed significant differences between the groups of social drinkers and alcoholics, leading to an improvement in the postmortem detection of alcohol abuse. Nevertheless, the performance of the method was rather poor, with an area under the curve calculated from receiver operating characteristic (ROC curve AUC) of 0.745. The optimum cut-off value for differentiation between social and chronic excessive drinking calculated for hair FAEEs was 1.08 ng/mg, with a sensitivity of 56 % and a specificity of 80 %. In relation to the "Consensus on Alcohol Markers 2012" by the SoHT, an increase in the cut-off value for FAEEs in the proximal hair segment 0-3 cm from 0.5 to 1 ng/mg may be advisable to avoid excessive numbers of false positive results.