Hair Analysis Panel Discussion:
Section: Appendix C, Michael Greenberg
Hair Analysis: Exploring the State of the Science
The laboratory analytical methods available are capable of defining the qualitative existence of a variety of pharmaceuticals, drugs of abuse, and occupational/environmental toxicants. The operative word here is "qualitative". Quantitation of specific levels are not, in my opinion and experience, either generally reliably reproducible and/or clinically useful. Specific analyte levels are essentially of little or no value in the determination of so-called cut-off levels (e.g., PELs, TLVs, "safe levels"), "normal levels," or other designators which rely on reference levels. In addition, the analytical techniques currently in use are capable of providing "segmental analysis" of hair, which in turn can provide a historical picture of various qualitative (not quantitative) exposures over time. In addition, hair analysis may help to derive an essential time frame which may indicate, based on the average rate of hair growth, the time of inception for various exposures.
The amount of hair needed for analysis may be dependent on the specific analyte sought as well as the temporal relationship between exposure and hair harvest.
One of the most important shortcomings for hair analysis, as it currently exists, is the fact that reference ranges may often be unreliable. Laboratories frequently base their reference ranges for specific analytes on limited case reports in the medical literature or exclusively on data derived from animals, which has limited applicability to humans. These facts contribute to substantial limitations with regard to interpretation of results. Variability undoubtedly exists from one laboratory to another. Certainly these facts limit the clinician's ability to interpret and utilize hair-derived values beyond the potential qualitative information that might come from hair testing of any individual.
Thus, at this time, it may be prudent to recommend that hair testing for all substances (drugs of abuse, occupational toxicants, environmental toxicants) be limited to qualitative determinations as opposed to quantitative determinations. The goal of quantitation for any laboratory analyte is to derive clinical algorithms that translate into levels that indicate disease, dysfunction, or specific risks for disease or dysfunction. With regard to hair testing in its current state, there is little evidence that there is sufficient reliability to use quantitation for these purposes.
Laboratory washing procedures prior to digestion may significantly alter the hair content of various analytes. For example, when hair is tested for THC, if it is washed with methanol, THC concentrations may be reduced by as much as 85% (Forensic Drug Abuse Advisor, 1996) by virtue of this process. It is reasonable to expect that similar degradations in analyte concentration occur when other analytes are involved.
Hair pigmentation is a critical factor in the interpretation of the concentration of certain compounds and their metabolites incorporated into hair. Melanin is responsible for the pigmentation. The color and the melanin content of human hair samples differs over a wide range. Once deposited into hair, chemicals may remain detectable for a period of months to years. However, if disposition into hair is influenced by those properties attributed to hair color, then certain persons may test positive more frequently than other persons. Removal of the melanin from hair digests prior to hair analysis may reduce the effect of melanin on the total chemical concentration by excluding the drug bound to the pigment. In one study (Hold KM, et al), the effect of melanin removal by centrifugation of hair digests on cocaine concentrations was investigated. Two sets of hair samples from five cocaine users were analyzed for cocaine and metabolites. A solution consisting of 10 mL of 0.5M Tris buffer (pH 6.4) to which is added 60 mg D,L-dithiothreitol, 200 mg SDS, and 200 U Proteinase K, was used to digest the hair. Two milliliters of this solution was added to 20 mg of hair and incubated at 37 degrees in a shaking water bath (90 oscillations/min) overnight. The samples were removed from the water bath and mixed. One set was centrifuged at 2000 rpm and divided into supernatant and melanin pellet. The other set was not centrifuged. Internal standards were added to all tubes. The samples were further extracted, derivatized, and analyzed by gas chromatography-mass spectrometry. A mean of 8.8% (standard deviation [SD] 7.0%) of the total cocaine concentration (supernatant and pellet) was left behind in the pellet. The same experiment was repeated—except that the melanin pellet was redigested with 0.1 N HCl. After redigestion of the melanin pellet, the mean cocaine concentration in the pellet was 3.8% +/- 4.0% (mean +/- SD) of the total cocaine concentration in hair. These investigators felt that their data demonstrate that removal of melanin from hair digests by centrifugation does not eliminate hair color bias when interpreting cocaine concentrations.
TOPIC #2: FACTORS INFLUENCING THE INTERPRETATION OF ANALYTICAL RESULTS
Exposure of hair sample to the external environment could be
an important factor
in confounding results on both a quantitative and qualitative basis. By way of example,
many over the counter hair coloring preparations contain lead acetate (e.g., Grecian formula). This may persist for long periods on hair shafts and thus confound hair testing results for lead. It is also unclear if the use of coloring agents containing lead acetate alters or enhances the hairs ability to bind other analytes or potential toxicants.
Based on the medical literature that describes the use of hair testing for substances of abuse there are differences in hair uptake of various substances based on ethnicity. For example, negroid hair has been suggested to bind cocaine residues with greater affinity than caucasoid hair.
There are also reports in the literature that the ability to bind various chemicals and drugs may depend on endogenous hair color as well as if hair has undergone bleaching. For example, bleached hair radically lowers the drugs [of abuse] content of hair. This may explain the observation that many competitors on the professional biking circuit sport bleach blond hair (Kintz). Blond hair has been shown to not bind cocaine or its metabolites as well as pigmented hair (Hubbard). In addition, there was no evidence of a dose-related incorporation of these drugs and metabolites into non-pigmented hair. The concern is that similar circumstances may occur with regard to specific occupational or environmental toxicants and chemicals.
Based on a study presented by Reid et al. indicating that gray hair takes up less cocaine than non-gray hair, it is possible that gray hair may also alter the utility of hair analysis in other settings. The Reid study evaluated cocaine levels in the same individuals by comparing the levels in gray and non gray hairs from the same person. In a similar study (Rothe et al.), hair samples from 15 patients receiving medical treatment with amitriptyline, carbamazepine, chlorprothixene, diclofenac, doxepine, indomethacine, maprotiline, or metoclopramide, or with a chronic heroin and cocaine abuse, were separated into white and pigmented fibers and both fractions were independently investigated by GC-MS. The drugs were found in pigmented fibers as well as in white fibers, but the concentrations in the white fibers were smaller than in the pigmented ones for most of the samples investigated. The concentration ratio of the drugs or their metabolites in both hair fractions (white/pigmented) was found to be between 0.09 and 1.57 (mean 0.70, 30 concentration pairs). There are large differences in this ratio between different subjects with the same drug; whereas for different drugs in the same subject—in many cases—similar ratios were measured. As reasons, a different grade of pigmentation of the hair and the influence of the drug structure are discussed. From these results it follows that the natural hair color is an important parameter in the evaluation of drug concentration in hair. Again, similar effects may be seen when dealing with occupational and environmental toxicants.
The rate of hair growth may be an important factor in the ability to identify the presence of various materials based on time of exposure. Sources usually indicate that head hair grows at the rate of 1-2 cm per month. That in itself represents a range encompassing up to a 100% difference in hair growth rate. Obviously, comparisons of individuals whose hair growth rates differ by a factor of 100% is problematic.
TOPIC #3: TOXICOLOGIC CONSIDERATIONS
Relatively little is known about the biological uptake of specific substances with regard to the concentration delivered to and incorporated into hair. There is essentially no data that reliably establishes the relationship between chemical concentrations in the hair and blood or other target organs for most chemcials. More specifically, and more importantly, no dose-response data currently exists with regard to chemical concentrations in the hair and blood or other target organs. In addition, no disease predictive value exists for any quantatative data that has been derived to date with regard to the hair concentration of drugs or chemicals.
Rollins et al have suggested that the ionization state of any given chemical is what determines whether or not it will bind with hair melanin. These investigators reported that catiomnic drugs are more likely to bind with melanin when compared with anionic drugs. This study may provide some guidance with regard to the binding ability of other toxicants of concern.
TOPIC # 4: DATA GAPS AND RESEARCH NEEDS
The data gaps that most significantly limit the use of hair testing in public health evaluations are 1) the lack of accurate and reliable reference range data and 2) the lack of specific information about dose response relationships with regard to the relationship between chemical concentrations in the hair and blood or other target organs. In my estimation, these two items constitute the most pressing research needs with regard to hair testing.
Future studies must address these basic data gaps in order to
even begin to decide if hair testing has clinical screening or other
TOPIC # 5: IDENTIFYING SCENARIOS FOR WHICH HAIR ANALYSIS MAY BE APPROPRIATE
Hair testing for acute exposures is clearly not the best alternative for determination of either dose or exposure with regard to any potential toxicant. If acute exposure is defined as the pre-distribution time frame, then blood or urine testing would be far superior to hair testing in any scenario. However, in the event of a single exposure (as opposed to an ongoing exposure) the use of hair testing after the completion of the pre-distribution phase of kinetics may be helpful in qualitatively identifying the fact that exposure has indeed occurred and/or generally timing that exposure. The use of hair analysis in this setting may have forensic as well as public health value.
In the setting of chronic exposures, hair analysis may have value in identifying and documenting a given exposure. This, again, may have forensic, civil-legal, and risk assessment value for individuals as well as communities and populations. Obviously, the length of any given individual's hair may limit the use of hair analysis, as well as how frequently the hair is cut.
In any scenario, however, the state of the art is such that specific and measurable health effects will generally not be uncovered by hair analysis. In addition, public health and/or individual risk assessment determinations will be limited by whatever conclusions may be drawn by what is essentially a qualitative and not quantitative toxicological evaluation.
One interesting study (Al-Delaimy, et al) used hair analysis to measure the relation between workplace smoking policies and exposures to environmental tobacco smoke (ETS) of workers in bars and restaurants. In this study, 114 workers were questioned about sources of exposure to ETS and smoking habits, and details of the smoke-free policy in their work place were recorded. A hair sample was collected from each participant and tested for nicotine. Among non-smoking workers, hair nicotine levels varied strongly according to the smoke-free policy at their place of work. Those working in 100% smoke-free restaurants had much lower levels than staff working in bars with no restrictions on smoking, and levels were intermediate for staff working in places with a partial smoking ban. Hair nicotine levels among nonsmokers working in places with no restriction on smoking were similar to hair nicotine levels of active smokers. The findings from this study highlight the substantial levels of exposure of bar and restaurant staff from patrons' smoking.
The potential sources for confounding variables in the hair testing
arena are truly legion. This fact is demonstrated in one instance
by a paper from Japan wherein investigators sought to draw a relationship
between head hair mercury and health. However, in the end, these
investigators discovered that "some subjects who showed a high total
mercury level made habitual use of toilet soap containing much mercury."
Thus, the confounding effect of an unusual source for a heavy metal
can interfere with effective hair analysis.
- Kintz P et al. Abstract presented at American Academy of
Forensic Sciences meeting, Reno,
Nevada, February 2000.
- Hubbard D et al. Society of Forensic Toxicologists meeting,
Snowbird, Utah, 2000.
- Reid R. et al. Society of Forensic Toxicologists meeting,
Snowbird, Utah, 2000.
- Rollins D et al. Society of Forensic Toxicologists meeting,
Snowbird, Utah, 2000.
- Rothe M et al. Effect of pigmentation on the drug deposition
in hair of grey-haired subjects. Forensic Sci Int 1997 Jan 17;84(1-3):53-60.
- Harada M et al. Monitoring of mercury pollution in Tanzania:
relation between head hair mercury and health. Sci Total Environ
1999 Mar 9;227(2-3):249-56.
- Holde et al. Quantitation of cocaine in human hair: the
effect of centrifugation of hair digests. J Anal Toxicol 1998
- Al-Delaimy W et al. Nicotine in hair of bar and restaurant
workers. N Z Med J 2001 Mar 9;114(1127):80-3.