Hair analysis for drugs of abuse. Hair color and race differentials or systematic differences in drug preferences?

There is currently a debate in the literature on chemical drug analysis con cerning the contribution of biophysical attributes associated with specimen s and specimen donors to assay outcome. In recent years this debate has foc used on hair analysis, but has in the past also been raised in urinalysis i nterpretation. In this article we examine several aspects of that controver sy. First, we present data regarding the effects of hair color on the distr ibution of positive hair testing results for three drug classes. We compare these results to negative hair samples from comparable donors. This data i s derived from head hair from preemployment donors that was classified acco rding to seven visual color categories. We determined the distribution of c olors for hair samples devoid of any of three assayed drugs (amphetamines, cocaine, and cannabinoids). Subsequently, this distribution was compared wi th the distributions for hairs that had rested positive for amphetamines, c ocaine or cannabinoids. We examined a total of 2000 randomly selected sampl es; 500 negative hair samples and 500 positive samples for each of three dr ugs: cannabinoids, cocaine, and amphetamine. We also evaluated ethnic/racia l factors in relation to positive urinalyses for various ethnic/racial grou ps. We examined approximately 4000 urine specimens from two different group s, each constituting around 2000 specimens. In addition to ethnicity/race a nd urinalysis outcome, we also examined the relationship between the hair c olor distributions of urine donors and the corresponding urinalysis results for the three drug classes. We also compared them to drug-negative samples . Our summary impression is that the observed outcome patterns were largely consistent with differences in drug preferences among the various societal groups. There was little evidence of a pattern attributable to hair color bias alone or selective binding of drugs to hair of a particular color. Lik ewise, there was no discernible pattern associated with race or ethnicity t hat would lend support to a "race effect" in drug analysis. (C) 2000 Elsevi er Science Ireland Ltd. All rights reserved.