Introduction
The most reliable and widely used methods of forensic hair comparison involve microscopic analysis and, more recently, DNA analysis. However, a wide variety of other methods have been proposed and find use in special applications, usually as an augmentation of conventional methods. These techniques can be divided into four broad and somewhat overlapping categories: microchemical methods, instrumental chemical methods, physical methods and biochemical methods.
Microchemical methods
The reverse of detecting treatments on hairs is to deliberately treat a hair with various dyes or reagents and then microscopically observe and compare the changes produced. In one example of such a micro-chemical testing technique, changes produced on hairs by treatment with mercaptoacetic acid were observed by scanning electron microscopy. Another example used Lucifer Yellow CH dye. These examples suggest that further research into this technique would be useful.
Instrumental Chemical Methods
Many instrumental chemical methods can play a role in forensic hair comparison. The most promising of these are the methods used to examine cosmetic treatments on hairs. Other instrumental chemical methods are targeted at naturally occurring or environmentally deposited components of the hair.
The most widely reported of this latter group of methods is pyrolysis gas chromatography (Py-GC). Although early work with packed column Py-GC failed to provide useful results, it was felt that the enhanced resolving power of capillary column Py-GC would provide a potential method of individualizing human hair. It was found, however, that the relative amounts of the pyrolysis products of hair proteins were the same from person to person, and that the same sort of variability occurred within a set of samples from the same individual as within a set of samples from different individuals. A subsequent study, found three major components (benzene, toluene, and styrene) whose capillary column Py-GC pyrograms differed significantly between individuals and remained constant over time. In a blind trial hairs could be correctly sorted into groups and it was concluded that although the method was not suitable for individualizing hairs on its own, it could prove useful in combination with other methods.
A promising application of Py-GC is the analysis of nicotine in hair. In a study of 48 subjects an increasing nicotine concentration gradient was found from root to tip, indicating adsorption of nicotine from the outside. A successful distinction was made between smokers and nonsmokers and a good correlation was shown between nicotine concentration and self-reported exposure. This method has potential to provide useful information to aid investigations (smoking status) as well as additional comparison characteristics.
Many different instrumental methods are currently used to analyze trace elements in human hair. Although this is a viable method for environmental and medical studies, there are many problems associated with its use in forensic hair comparison. Trace elements in human hair can arise from metabolism, cosmetic preparations or the external environment. Trace element content of hair is affected by personal factors (age, sex, diet, etc.), external factors (shampoos, air-borne contaminants, etc.), frequency of hair washing, the structure of the hair itself, and by sample treatment prior to analysis. Furthermore, hair trace element content varies with location on the body, from hair to hair on the same body, from root to tip along the hair shaft, across the diameter of the hair, and according to the stage in the hair growth cycle. Because of such variation and sample treatment problems, trace element analysis of hair is not considered to be a reliable method of forensic hair comparison.
A major use of chemical instrumental methods has been drug profiling of human hair. Reports of opiates, barbiturates, methaqualone, codeine, cocaine, morphine and phencyclidine being detected in hair have appeared in the literature in recent years. Radio-immunoassay has been the most common method of analysis. All work to date has concentrated on the toxicological use of hair analysis for drugs. The large number of licit and illicit drugs available and the correlation between crimes of violence and drug abuse make this a promising area for further research as a method for comparing known and questioned hair samples.
The use of GC-MS to detect oxidative dyes and other hair products was discussed elsewhere. Fourier transform infrared spectroscopy (FTIR) also has potential in the analysis of hair sprays and other cosmetic treatments. High pressure liquid chromatography (HPLC) may also prove useful for the analysis of hair dyes and rinses.
Much current work in instrumental analysisisrelated to applications of combined instruments such as gas chromatography/mass spectrometry (GC-MS) and HPLC-FTIR. It ispossible that the future will bring applications of these combined instrumental methodsto forensic hair comparison.
Other instrumental chemical methods have been reported in single studies which have not been followed-up or generally adopted.
Physical Methods
The use of physical properties of hair in forensic hair comparison has been limited. The potential of spec-trofluorometry in the comparison of photolumines-cent propertieswaspointed out in 1975 and the significance of some elastic constants in hair comparison has been investigated. Many years ago, density was proposed as a hair comparison characteristic. Those physical properties thus far investigated have been found to have large within-hair and within-individual variationswhich have tended to discourage their use. The sophisticated computers and statistical packagesof today and the future might be able to overcome thisproblem, provided that these variations are significantly less than the variations that exist between individuals.
Biochemical Methods
Biochemical methods provide some promising areas for future research in forensic hair comparison. For several years, ABO grouping of hairs has been successfully used routinely in Japan and other East Asian countries. However, when applied to non-Mongoloid hairs, ABO grouping has been found to give results that are somewhat erratic. Japanese workers have shown that blood group substances appear to be localized in the medulla. Thisknowledge could lead to future research which might increase the reliability of ABO grouping with Caucasian and Negroid hairs.
For several years, enzyme typing of hair root sheaths was a method routinely used in many forensic laboratories. Phosphoglucomutase, esterase D and glyoxylase were the systems most commonly used. Other workers developed a procedure for simulta-neoustyping of erythrocyte acid phosphatase, adeny-late kinase and adenosine deaminase in human hair root sheaths. However, since root sheaths occur in only a small percentage of questioned hair, the application of thismethod issomewhat limited. Aswith blood and semen analysis, in recent years DNA-based methodshave superseded enzyme typing methodsfor hair.
Isoelectric focusing in polyacrylamide gels for non-carboxymethylated keratins has proven useful in spe-ciesdetermination and could be used to detect individual variation in a given population. It has been suggested that sodium dodecyl sulfate-polyacry-lamide gel electrophoresis (SDS-PAGE) could be applied to forensic hair comparison. It was noted, however, that further research was needed. Following a good discussion of hair comparison by electro-phoresis another worker concluded that: ‘The results of a study of a limited number of human hair samples by two-dimensional electrophoresis are sufficiently encouraging to recommend that a trial be started in which hair samples from a much larger number of individualsbe examined . . . Further research is required to assess the potential of the technique but present indications are that it may become a useful supplement to microscopic description of hair.’ (Marshall, 1984)
Amino acids analysis is another biochemical method with some future possibilities, although the variability in amino acid composition of hairs may cause problems.
Concluding Remarks
Let us now look at the effect some of the previously discussed other analytical methods might have on evidential value. Making as many sets of independent comparisons as possible will greatly reduce the probability of coincidental matches. By doing so, provided they do not also greatly increase the probability of examiner errors, these other analytical methodologies should lead to a large increase in the evidential value of forensic hair comparison.
Two notes of caution should, however, be introduced when considering any future research on hair comparison methodology. First, the probabilities of type I errors(incorrect elimination) and type II errors (incorrect association) vary inversely. As the probability of type II errors is decreased by new methods, the probability of type I errors will often increase. Accordingly, in evaluating new methods we need to ask whether or not the method significantly decreases the probability of type II errors and, if so, if it also increases the probability of type I errors to an unreasonable level.
A second reservation or caution must be expressed with regard to research on hair characteristics which are sensitive to the environment or consumer products. If, for example, a suspect is not apprehended shortly after a crime, his or her hair could have been subject to cosmetic treatment between commission of the crime and the subsequent submission of the known hair sample to a forensic laboratory. Thus the cosmetic treatment characteristics of the known hair sample could be either deliberately or accidentally altered. When compared with a questioned hair found at the crime scene or in the victim’s clothing, a type I error would likely result if too much emphasis were placed on cosmetic treatment characteristics in such circumstances.
Collaborative studies provide an excellent means of assessing the costs and benefits of any new hair comparison methodology.
