Validity of Head Hair Analysis and Methods of Assessing Multiple Chemical Sensitivity

Roger D. Masters

Research Professor of Government, Dartmouth College

            As a method of preliminary evidence of the presence of toxic substances harmful to health and behavior, analysis of head hair samples has many advantages.     Among these, the most important is the ability to screen for a large number of toxic and biologically necessary elements (which can be of vital importance because excessive levels of biologically necessary elements such as zinc and manganese can have negative effects, especially in combination with other toxins such as lead, cadmium, or mercury).    Other advantages include the non-invasive character of the test, which avoids problems of taking samples, as well as relatively low cost compared to a full array of blood, urine, and head hair analyses.   Hence, head hair analysis is now increasingly being used as a preliminary test to see whether individuals have absorbed poisons linked to behavioral or health problems. 

I. THE ISSUE: IS HEAD-HAIR ANALYSIS A VALID  METHODOLOGY?

A recent challenge to the validity of head hair analysis was published in the Journal of the American Medical Association[1].   Specialists have noted that this study was marked by flawed design and a questionable conclusion.  There is no doubt that some laboratories engaging in head hair analysis provide unreliable results.  Those established in the field were thus not surprised by confirmation of this view.   The problem of research design concerns the very small number of laboratories studied and the failure to use any method that would discriminate between reliable and unreliable laboratories (e.g., blind tests of aliquots from a single subject at each lab).  While it is proper to note that some laboratories do not provide scientifically accurate data, it does not logically follow that all use of the head hair analysis technique is totally unscientific and completely unreliable.

There are three general reasons to contest the blanket rejection of head hair analysis on the basis of evidence which, at most, shows that some laboratories are unreliable.  First, the problem of quality control and potential confounding factors characterizes many laboratory procedures, and is especially serious in tests of lead levels in capillary blood (see Section II below).  Second, properly conducted head hair analyses have the benefits of measuring long lasting excretion of a wide variety of toxins in a cost-effective manner – and hence have a valuable place in conjunction with other environmental and diagnostic testing.  And finally, but most important, these advantages have been amply demonstrated by the excellent diagnostic and treatment results of the Pfeiffer Treatment Center, where sophisticated research has shown the role of chemical imbalances in the diagnosis and treatment of such diverse conditions as autism[2] and assaultive young males.[3]     This last point is especially telling, since an editorial endorsing the JAMA  article critical of head hair testing concluded that a role in effective therapy would be a principal reason to accept the technique.[4]

After making points like those noted above, one commentator on the issue added: “The World Health Organization, the International Atomic Energy Agency, and the U.S. Environmental Protection Agency have all recommended hair analysis for determination of heavy metals in certain instances.  The EPA stated in a report that ‘…if hair samples are properly collected and cleaned, and analyzed by the best analytic methods, using standards and blanks as required, in a clean and reliable laboratory by experience personnel, the data are reliable.’ ”[5]

II. THE EXAMPLE OF BLOOD LEAD TESTING

The example of blood lead testing illustrates the difference between the specific tissue being analyzed and reliability of results with.[6]  Although widely used and now conventional, blood lead measurements based on a capillary (or “finger-prick”) sample are more widely in use than venous blood lead tests (which require the insertion of a needle in a vein).   Capillary tests are less expensive and, for wide scale population screening, often preferred due to the greater ease of administration and lower cost. Unfortunately, capillary tests are also more likely to be confounded by toxins on the surface of the skin or “lead dilution occurring with serum leakage when the puncture is squeezed.” [7] Even though venous blood lead is considered more accurate, the greater difficulty in large-scale testing has resulted in many surveys of children’s blood lead using capillary tests.   Moreover, venous testing can also be subject to confounding factors: “Even in cases of severe lead poisoning, lead appears in whole blood at the trace or subtrace level.   Extreme care is therefore required to minimize contamination and sample lead loss.  Collection precautions include the use of lead-free blood tubes, needles, anticoagulant, and a thoroughly cleaned puncture site.” [8]

There is thus no perfect solution. In some settings, venous blood lead testing will inevitably reduce substantially the proportion of children in a target population that are actually sampled.  On the other hand, using capillary testing may increase coverage but reduce accuracy.   Other test media, such as urine, bone or teeth, each have advantages and drawbacks.    Typically, the solution to this dilemma is to use capillary blood lead tests for a first assessment of a population, with venous blood lead testing as a follow-up for cases that require further analysis.

III. HEAD HAIR ANALYSIS ON SMALL SAMPLES AS A PRELIMINARY SCREENING METHODOLOGY

In much the same way as capillary blood tests in many American schools, head hair analysis can be used as a preliminary screening device.  In the study of the Oujè Bougoumou Cree, a small number of head hair samples was analyzed to provide a first test of the hypothesis that heavy metals have been absorbed by individuals in a population that suffers health or behavioral deficits and  is exposed to toxins.  To avoid unreliability due to poor laboratory procedures, samples were tested by Doctor’s Data, Inc. – a laboratory nationally known not only for its high standards but for its commitment to scientific research using data from its head hair analyses.[9] 

Even using a well-reputed laboratory, a preliminary examination like the one conducted for our Report is not the same thing as an epidemiological screening of an entire population, but rather provides data that can justify more extensive studies or more intensive assessments of toxicity for the purpose of designing and administering health care.   Nor does such a procedure obviate the danger from contamination “by external contact with lead sources,” [10] though this problem is probably more likely in urban populations (especially where cultural practices include exposure of head hair to substances containing lead).  

Another example of this use of head hair analyses will be helpful.  In an American university, a number of students exhibited signs of previously undiagnosed learning disabilities of types that have been associated with neurotoxicity of such heavy metals as lead and manganese.[11]  Such an association is plausible because these toxins interfere with key neurotransmitters including dopamine and serotonin that play a central role in cognition and behavior,    At the same time, testing of the water supplies delivered to this university revealed manganese levels at approximately 2.4 times the Maximum Contaminant Level (MCL) established by the U.S. Environmental Protection Agency.

To determine whether harmful effects of this water supply should be further explored as well as to explore treatments of the affected students and justify the expenditure of funding remediation of the water system, head hair samples were taken from four students with clearly characterized behavioral problems.  These were submitted for analysis to Doctor’s Data Analysis, Inc. of St. Charles, Illinois, a laboratory known for its high level of standards which has frequently published research findings in peer-reviewed journals.[12]    In one case, moreover, a student had developed Chronic Fatigue Syndrome after one year at the university and it was judged useful to secure head hair testing from both parents to check the possibility that the toxins had been absorbed from the home environment.

In this small sample, all four cases revealed that manganese levels were at dangerously high levels (over the 68^th percentile of national averages, and in some cases over 95^th percentile).    For each of the four affected students, a different mixture of other toxic heavy metals was also at dangerous levels, so that the principal common factor was the same chemical found in the local water supply.  For the one student with Chronic Fatigue Syndrome whose parents were also tested, manganese and other toxins were not at dangerous levels in either parent, confirming that exposure to manganese was more likely to have occurred at the university after enrollment than at home.

In no way could one pretend that analysis of four students in an institution with an enrollment of over 4,000 was a “sample” of the entire population.  Even so, the procedure of analyzing the head hair of six individuals had a two-fold effect.  First, for the four students concerned, head hair analysis made it possible to begin remediation (e.g., consultation with a Naturopathic Doctor for herbal and dietary treatments, such as anti-oxidants, capable of reducing the severity of symptoms).  Second, the existence of this data greatly impressed those who were engaged in seeking the municipal decision to renovate the local water system.    Conversation with public officials in other communities confirms that, given fears of suit for toxic tort, a relatively small number of cases that are well characterized may be important as the first step – but surely not the only or last step – in remediating both individual and community problems.

As this example should indicate, a relatively small number of head hair analyses (where possible over 10 to 15) is adequate to confirm the existence of potential problems requiring further assessment and treatment.  Whatever the possible confounding factors in individual assays of head hair, as with any other biomedical test, the issue concerns the probability that results would occur by chance as well as the existence of reasonable explanations for the results observed.    In particular, where the presence of a toxin in the environment is associated with dangerously high levels of the same toxin in a large proportion of those tested, testing error seems less likely to be an explanation of the findings.  Where there is co-occurrence of a number of toxins also found in the environment, as occurred in the study of the Oujé Bougoumou Cree, the chance that the result is due to experimental error becomes extremely small.

IV. LINKING ENVIRONMENTAL DATA WITH HEAD HAIR ANALYSES AMONG THE OUJE BOUGOUMOU CREE

The importance of head hair as evidence of uptake of multiple toxins from the environment is greatly reinforced by a consideration of the probable effects of different modes of exposure for various toxins.    Among the Oujé Bougoumou Cree, who rely on food from the affected environments, this factor is particularly important.   In fish, the item in the food chain extensively analyzed for toxins, the most frequently elevated element was mercury, with levels ranging from 0.10 to 0.46 mg/Kg and above 0.24mg/Kg at seven of ten sites (Report, Table 6).    By contrast,  cadmium was under 0. 1 mg/Kg in fish at nine of the ten measurement sites, both arsenic and chromium were an unspecified level under 0.5 mg/Kg at nine of the sites, and copper varied from 1.2 to less than 0.5 mg/Kg depending on location.    That this difference would be reflected in different patterns of uptake for mercury than for the other heavy metals is particularly likely because, as one expert has emphasized, “Mercury toxicity is a retention toxicity that builds up during years of exposure.”[13] As a result, mercury is the toxin most likely to be consistently absorbed from the food chain – and this hypothesis can be used to test the validity of the testing procedure.

If mercury is primarily absorbed by the Oujé Bougoumou through eating fish, head hair assays should consistently show a strong correlation between the age of the individual sampled and higher levels of mercury in head hair.   Insofar as cadmium, arsenic and chromium  are absorbed from contact with soil or water, which are more dangerous in some places than others, levels of these toxic metals should vary primarily due to the place where each individual lives.  In contrast, mercury levels should be correlated with the age of the person tested since the longer an individual eats fish containing mercury, the more mercury can be expected to be absorbed.  If so, are the data which we report consistent with these hypotheses?   Assessment of the association between age and mercury levels in head hair not only provides a quick means to check the hypothesized role of toxins in the food chain, but also can serve as a test of data reliability (since a correlation between head hair and each subject’s age is not likely to be an artifact of poor analytical methodology).

A statistical test of the predicted association between subject’s age and the head hair levels of mercury as contrasted with arsenic, cadmium, or chromium confirms the hypothesis.    For mercury, there is a strong correlation between head hair levels and age (r = .658).  While children under age 18 had mercury levels below the reference level for that age (2.0 mg/g,  the level of the toxin exceeded by less that 5% of comparable populations studied), all but 4 individuals under age 40 and everyone tested over age 40 exceeded the adult reference level  (3.0 mg/g).     The graphic display of the data makes clear – whether for the full sample (Figure1) or for males and females separately Figures 2a-b) -- the older the individual tested, the greater the likelihood of dangerously high mercury uptake.    

This pattern, while predicted for food intake, is not likely to be characteristic of toxins absorbed from soil or water if pollution differs from one location to another.    For the latter exposures, each individual’s residence and location of activity are more likely to influence toxic uptake than age.  Consistent with this hypothesis, the correlations between age and head hair levels of arsenic, cadmium, chromium, and copper are all much lower than the correlation between age and mercury (Table 1).    Moreover, since mercury interferes with the body’s natural detoxification mechanisms (notably metallothionine), one would predict that where the association between one of these other toxins and age is highest (chromium, r = .426), there is an even higher correlation between chromium and mercury levels in each individual (r = .499).  Similarly, correlations between cadmium and mercury levels (r = .497) or between arsenic and mercury levels (r = .393) are substantially higher than the correlation between either of these geographically varying toxins  age (cadmium, r = .233; arsenic, r = .215).  Such patterns, consistent with the greater role of fish and other food items in mercury toxicity, provide reasonable confirmation that head hair analysis is not totally confounded by unreliable laboratory methods.

Although these data concern only a small sample of Oujé Bougoumou Cree.   head hair analysis of a limited portion of a local population is only the first step in the exploration of health and behavioral deficits due to environmental exposures to toxins.      It is rare to have a localized study comparing toxins from the environment (sediment and water in this case) and the food chain (fish) to individual uptake (head hair),     Even so, this is only the first step in a comprehensive evaluation and treatment that should proceed if the preliminary head hair analysis gives plausible reasons for concern.  A separate statement provides the outline of such a follow-up, which includes comprehensive head hair testing for the entire Oujé Bougoumou population, multiple biochemical assays, including urine, blood, and head hair, for those with dangerous findings in the first screening test, and specialized biochemical intervention to treat the effects of toxins confirmed by these assays. 

TABLE 1: Correlation of heavy metals in head hair and age

                        Arsenic             Copper             Chromium         Cadmium          Mercury

Full  sample (n=23)

Correlation between      0.2146               0 .0019                        0.2334              0.2334              0.658

Age and Element

Correlation between

Level of Mercury and

Element                        0,3926               -.1221               0.4993              0;4573

Fig. 1: Age and Mercury Levels – Entire Oujé Bougoumou Sample

Mercury

Levels

Age

Total sample of 23 (sexes combined). 

Correlation of age and mercury level: r = .658

Solid Bar: reference range 2.0 µg/g to age 18, 3.0µg/g over 18.

Note that virtually all those over the age of 18 are above the reference level for Americans.

NOTE:  THESE FINDINGS ARE STRONGLY CONSISTENT WITH THE PATHWAY OF MERCURY FROM THE FOOD CHAIN (FISH) TO THE OUJE BOUGOUMOU CREE.    They confirm the need for a TOTAL sample of the entire population, with a control for amounts of fish consumed and other indications of exposure to mercury.

Fig. 2a: Age and Mercury Levels  in Oujé Bougoumou Males

Mercury

Level

(µg/g)

Age

Reference levels: 2.0 µg/g for ages 0-18, 3.0 for ages above 18.

Total sample: 15 males, of whom only 5 – ages 15, 18, 19, 26, and 38 – were under the “reference” range.    The correlation of age and mercury level (r = .693) is reinforced by visual evidence that all males over the age of 40 and 25% of those under that age had head hair levels of mercury far in excess of the reference range established for American populations.

Fig. 2b: Age and Mercury Levels  in Oujé Bougoumou Females

Mercury

Level

(µg/g)

AGE

Reference levels: 2.0 µg/g for ages 0-18, 3.0 for ages above 18.

Total sample: 8 females, of whom only 3 – ages12, 12, and 14 – were under the “reference” range.    While the number is too small to make the correlation of age and mercury level significance, it is evident that all females over the age of 20 had head hair levels of mercury far in excess of the reference range established for American populations.

NOTES