American Association for Clinical Chemistry
Better health through laboratory medicine
May 2008 Clinical Laboratory News: Toxic Metals

CLN Banner Logo

May 2008: Volume 34, Number 5

Toxic Metals
Why Hair Analysis Deserves Another Look
By Erlo Roth, MD and David W. Quig, PhD

Long, lustrous hair is more than a marker of good health. Hair is also an excretory tissue that can provide a chronological record of bioavailable trace elements in the body. In fact, detection of toxic metals in hair actually predates detection in blood or urine. Hair analysis as a tool in clinical toxicology was first reported in 1929, and over the years, forensic scientists have used the presence of mercury, arsenic, lead, and thallium in hair as evidence for the cause of death. And, just recently, scientists using today’s highly sensitive and reliable instruments confirmed the presence of toxic levels of lead in the hair of both Ludwig van Beethoven and President Andrew Jackson.

Analysis of elements in hair represents a valuable and inexpensive screen for excessive exposure to toxic elements. With growing concern about environmental pollution and recent news about the presence of certain toxic metals in toys, interest in hair analysis is increasing. While it should not be considered a stand-alone diagnostic test, when used in conjunction with patient symptoms and other laboratory tests, elemental analysis of hair is a reasonably accurate and precise technology that can determine long-term exposure to various toxic elements. Here we describe how the science of hair testing has evolved and some emerging applications of the methodology.

Science or Quackery?

Over the years, serious criticisms of hair analysis as both a diagnostic and an epidemiologic tool have been raised. One of the most widely cited articles, “Commercial hair analysis: science or scam?” was published by Stephen Barrett in 1985 (1). After sending samples of hair from two presumably healthy teenagers to 13 independent laboratories, Barrett reported major concerns in three areas: interlaboratory variability in results and reference ranges; lack of proficiency testing; and promotion of food supplements by six of the laboratories. He declared that managing a variety of diseases and nutrient imbalances based on hair analysis for trace elements was “unscientific, economically wasteful, and probably illegal.”

While Barrett’s survey revealed a lack of standardization and potential conflicts of interest for some of the laboratories, the study had its own flaws. The hair sample came from more than one donor, was not taken from close to the scalp, and was up to 15 cm long. These factors cast doubt on the homogeneity of the pooled sample, which is an important requirement for proficiency testing (PT) samples. This error may have occurred because Barrett, a psychiatrist, may not have been familiar with analytical chemistry principles.

Barrett also missed the opportunity to compare the precision of each lab. A table published in the article did not provide the numeric results or data regarding the precision of the individual laboratories, although all were sent four samples. Furthermore, Barrett did not report the true value of the pooled hair sample. The data, therefore, did not rule out acceptable precision and accuracy of an individual lab’s analysis.

In 2001, Seidel et al. performed a similar survey of six independent laboratories, but used one individual’s hair samples that had been cut close to the scalp (2), as is now the recommended practice. They found that three of the six labs promoted brand-name food supplements and also observed variability both in the results and methods of sample preparation. Based on their findings, the researchers concluded “that healthcare practitioners refrain from using such analyses to assess individual nutritional status or suspected environmental exposures.” But like Barrett, the researchers did not test the pooled hair sample for homogeneity or determine the true value of the sample prior to distributing it. In the same issue of the Journal of the American Medical Association, an editorial criticized nutritional counseling or therapy based upon hair analysis.

When this later study was conducted, CLIA’88 was in place, and some of the worst laboratories had been closed. The main flaw of the study, however, was its failure to recognize that two of the labs, both of which performed large volumes of testing and did not promote food supplements, had good agreement in their reference ranges in 20 of the 31 results. The widely discrepant results of other laboratories most likely reflected quality problems, but by condemning all laboratories that perform hair analysis, the researchers threw out the proverbial baby with the bath water.

Recent Advances in Hair Analysis Technology

Much has changed, both analytically and in the use of the information derived from the analyses, in the field of elemental analysis of hair in recent years. The instruments have become significantly more sensitive, precise, specific, accurate, and stable. Inductively-coupled plasma/mass spectrometry (ICP-MS) is emerging as the method of choice, although labs are slowly migrating to it, likely due to the high cost of the instrumentation.

ICP-MS works by aspirating a liquid sample into a high-temperature argon–ion plasma. In this plasma, ions are formed from the elements in the sample. The mass spectrometer identifies the elements in the sample based on their mass and quantifies the amount based upon comparison of the intensity of each mass to known standards and controls. Table 1 shows typical detection limits of the methodology for some of the common toxic metals.

Table 1
Typical Detection Limits for ICP-MS









Molybdenum .010
Chromium .010
Vanadium 0.07
Lithium 0.10

*parts per billion

A look at the CAP 2007 PT event for toxic metals reveals some interesting data. Of the labs participating in the blood lead PT, only 19 used ICP-MS, 130 used anodic stripping voltametry, and 120 used atomic absorption spectrometry (AAS). For blood cadmium, 11 of 30 labs used AAS and 19 used ICP-MS. For cadmium at a mean concentration of 0.85 µg/L the CV for AAS was 29.5%, but only 13.9% for ICP-MS. Even for a concentration of 38.32 µg/L, the CV for AAS was 8.9%, versus 4.0% for ICP-MS, demonstrating the superior precision of the latter.

Improved Standardization and Proficiency Testing

The other developments that have improved the quality of hair testing include standardization of sample preparation methods and PT. Following a 1998 publication from the International Atomic Energy Agency (3), methods of hair analysis became more standardized. The following year, the Institut National de Santé Publique du Québec began offering a PT program for elemental analysis of hair called QMEQAS, which currently includes 23 elements. About 16 laboratories participate regularly, and this number is growing steadily toward the traditional statistical minimum of 20. Data from this study, summarized in Table 2, show that the analytic precision of four elements in hair compares favorably with that of the same elements in blood. It also shows the accuracy of the participants in relation to the spiked value.

Table 2
Precision of Hair Analysis vs. Blood Analysis
*Data from QMEQAS 07H-01
^ Data from NY Department of Health Event # 2, 2007

These data demonstrate that both accuracy and precision of hair testing are as good as or better than those of many widely used clinical laboratory assays, particularly certain immunoassays. For comparison, the 2007 CAP Ligand Assay Surveys K/KN-C showed a CV of 29.1 % for all results of Free T3 of sample K-11, and the mean ranged from 5.77 pg/mL to 15.65 pg/mL between analytical systems. That same survey revealed mostly double digit figures for the CVs of individual methods for RBC folate and a CV of 254.8% for the PSA Ratio.

This analytical progress led the U.S. Environmental Protection Agency to state that “…if hair samples are properly collected and cleaned, and analyzed by the best analytical methods, using standards and blanks as required, in a clean and reliable laboratory by experienced personnel, the data are reliable”(4).

The Excretory Physiology of Hair

In the last decade, scientists have also learned much about the physiology of hair. It acts somewhat like a chromatography column, reflecting the excretion of elements and drugs over time. This property makes hair a more useful sample for evaluating long-term exposure than blood and even urine. The concentration of many trace elements in blood is transient and related to their supply in the previous hours or days, whereas in hair these elements remain for months. Drug screening companies, like Psychemedics (Acton, Mass.) have successfully exploited this property of hair. The company holds a patented hair analysis technology for drugs of abuse that has been approved by the State of New York, although the state still does not allow hair testing for heavy metals in living persons.

Once elements are incorporated into growing hair there is no exchange back into the body. Therefore, their concentration in hair is usually far greater than in blood or urine. Not only does hair provide a longer record of exposure over time to some toxic metals, but metals are also concentrated in hair during its growth. For example, researchers demonstrated that the concentration of methyl mercury in human hair is up to 300-times higher than that in blood (5).

Consistent with this finding is the fact that the predominant form or species of a toxic element that accumulates in hair correlates well with its half-life in blood. For example, the predominant form of mercury in hair is methyl mercury. Its half-life is about 60–70 days in blood, primarily in the red blood cells. In contrast, inorganic mercury accumulates in hair to a much lesser extent and has a half-life in blood (primarily in plasma) of only about 2–4 days. Inorganic arsenic, which has a longer half-life in blood than the organic, dietary arsenic species, is the dominant form of arsenic that accumulates in hair. This makes sense since in order for an element to become incorporated into hair it must first circulate in blood.

The Beethoven Case

A 4-year analysis of a lock of Ludwig van Beethoven’s hair suggested that lead poisoning could be the cause of certain ailments suffered by the famous composer. High lead concentrations in Beethoven’s hair were found in independent analyses concluded in 2000 by McCrone Research Institute (Chicago, Ill.) and Argonne National Laboratory (Argonne, Ill.), in which washing of the sample was performed at our laboratory.

The data provided evidence that Beethoven had plumbism or lead poisoning, which may have caused his life-long illnesses, impacted his personality, and possibly contributed to his death. Conducted on a lock of hair that was snipped after Beethoven died in 1827 at the age of 56, the analysis found a concentration of lead 100 times the level normally found in hair today.

Interpretation of Test Results, Clinical Utility

Reference ranges and the interpretation of hair analysis results are still somewhat of a challenge, due to the newness of the technology and the fact that metals intoxication is notorious for its late and non-specific symptoms. The standard method of extrapolating lethal doses in animals is obviously inappropriate when the objective in humans is prevention. Based on more than 40,000 results, our laboratory uses the upper 97.5th percentile method of determining reference ranges according to CLSI C28-A2. But these limits are still in flux, as they are for blood lead levels, where experts are now calling for the upper limit of tolerance to be lowered from 10 µg/L to 5 µg/L.

Although knowledge about the clinical usefulness of elemental analysis in hair is still in its infancy, some significant diagnostic applications have already been recognized, especially for methyl mercury. Several studies have also reported positive correlations between human hair mercury levels and fish consumption among various ethnic groups (5,6).

Researchers have also applied hair testing technology to pregnant women and children. A study of hair from pregnant women showed that higher mercury levels were associated with a higher risk of preterm delivery (7) and lower cognition in their infants (8). Another recent study using the National Health and Nutrition Examination Survey (NHANES) data set concluded that hair testing has been valuable in determining the incidence of excessive exposure to methyl mercury in women (9).

Some intriguing work involving people with autism has also been published. Authors of a recent study concluded that the relationship between blood mercury and hair mercury levels is significantly reduced in persons with autism compared with non-autistic persons, particularly at high levels of mercury. They go on to suggest that impairment of the hair’s excretory function is involved in autistic behavior (10).

Wider recognition of the utility of hair analysis is also emerging. The Agency for Toxic Substances and Disease Registry of CDC convened a “Hair Analysis Panel Discussion”, which concluded that “…hair analysis can be useful for simply identifying or confirming exposures” (11). And in 2001, the EPA established a benchmark for methyl mercury in maternal hair of 11 ppm, which is equivalent to 46 to 49 µg/L in maternal blood (12).

Setting Aside Misconceptions

As hair testing gains wider acceptance, laboratorians will benefit from understanding its advantage in analyzing long-term exposure to toxic metals. It is widely recognized that blood levels of toxic elements do not accurately reflect levels of net retention, especially when significant exposures occurred in the distant past.

In fact, some physicians already rely upon hair testing as a non-invasive screening test. Goyer and colleagues determined that the relationship between blood lead and urinary lead excreted with EDTA treatment is nonlinear, in that arithmetic increases in blood lead are associated with exponential increases in post-EDTA urinary lead excretion (13). This finding is the basis for the current evaluation protocol used by physicians who order hair analysis. If hair levels are elevated, they perform a provocation or mobilization test preceded and followed by 24-hour urine levels of the toxic element(s) in question. If this challenge significantly increases the urinary excretion of that element, a therapeutic chelation is indicated.

Elemental analysis of hair is now a reasonably accurate and precise technology in ethical laboratories. Recent scientific evidence soundly supports the methodology, and critics who continue to say that environmental exposure to toxic elements should not be evaluated by hair analysis have failed to acknowledge the new findings. Today, many reputable institutions, such as the CDC, ARUP (Salt Lake City, Utah) and Mayo Medical Laboratories (Rochester, Minn), perform hair analyses. As with any samples sent to reference laboratories, physicians and laboratorians should use objective criteria for evaluating the quality of the testing service, such as the results of proficiency testing and whether the lab also analyzes other more widely used matrices such as blood and urine.

The time has come to set aside outdated opinions and explore the full potential of this analytical tool for detecting specific toxins in our increasingly polluted environment. Elements such as lead, mercury, cadmium, manganese, and arsenic are poisonous, and people must exercise care when working in environments that may expose them to these toxic metals. By just being aware that one’s hair concentration is significantly higher than the 97th percentile of the population, individuals may be more inclined to take preventive measures and reduce their risk of damaging their health.


Barrett S. Commercial hair analysis: science or scam? JAMA 1985; 254: 1041-1045.

Seidel S, Kreutzer R, Smith D, McNeel S and Gilliss D. Assessment of commercial laboratories performing hair analysis. JAMA 2001; 285: 67–72.

Ryabukin YS. Activation analysis of hair as an indicator of contamination of man by environmental trace element pollutants. IAEA Report, IAEA/RL/50, Vienna; 1978.

USEPA, Report 600/4-79-049.

Phelps R W, Clarkson T W, Kershaw T G et al. Interrelationships of blood and hair mercury concentrations in a North American population exposed to methyl mercury. Arch Environ Health 1980; 35: 161–168.

Malm O, Branches FJ, Akagi H, et al. Mercury and methyl mercury in fish and human hair from the Tapajós river basin, Brazil. Sci Total Environ 1995; 175: 141–50.

Xue F, Holzman C, Rahbar MH, Trosko K, Fischer L. Maternal fish consumption, mercury levels, and risk of preterm delivery. Environ Health Perspect. 2007; 115: 42–7.

Oken E, Wright RO, Kleinman KP, Bellinger D, Amarasiriwardena CJ et al. Maternal fish consumption, hair mercury, and infant cognition in a U.S. Cohort. Environ Health Perspect. 2005; 113: 1376–80.

Allen BC, Hack CE, Clewell HJ. Use of Markov Chain Monte Carlo analysis with a physiologically-based pharmacokinetic model of methyl mercury to estimate exposures in U.S. women of childbearing age. Risk Anal. 2007: 27(4): 947–959.

DeSoto MC, Hitlan RT. Blood levels of mercury are related to diagnosis of autism: a reanalysis of an important data set. J Child Neuro 2007; 22: 1308–1311.

U.S. EPA Integrated Risk Information System (IRIS), 2001; US EPA online resource:

Goyer RA, Cherian MG, Jones MM et al. Role of chelating agents for prevention, intervention, and treatment of exposures to toxic metals. Environ Health Perspect 1995; 103: 1048–52.


Erlo Roth, MD, is Medical Director of Doctors’ Data, Inc. in St. Charles, Ill. and of Biosafe Laboratory in Chicago, Ill. 




David Quig, PhD, is Vice President, Scientific Support for Doctor’s Data, Inc. where he conducts studies pertaining to the effects of heavy metal and chemical toxicity on nutrition and metabolism and advises medical practitioners about the interpretation of laboratory test results and treatment options for their patients. Email: