Q. My wife has chronic health problems. She had a hair analysis done that shows various toxic exposures, nutritional deficiencies and possible allergies. What's your take?

A. Hair analysis may be useful in testing for DNA, methyl mercury and perhaps arsenic. But in 1985, Dr. Stephen Barrett submitted identical samples to different labs, which reported such widely differing levels of minerals such as cadmium, chromium, lead and tin that he titled his article "Commercial Hair Analysis: Science or Scam?"

In 2001, Sharon Seidel and others from the Environmental Investigations Branch of the California Department of Health decided to see if things had improved and gave hair samples from a healthy woman to the six "CLIA-certified" labs responsible for 90 percent of hair analyses in the U.S. Again, reported levels varied widely — by tenfold for 12 minerals. And the labs gave conflicting advice — one identified her as a "fast metabolizer" and advised avoidance of vitamin A and an increase in purines and dairy products, and another lab called her a "slow metabolizer" and advised taking vitamin A and avoiding purines and milk. The investigation team concluded that hair analysis was unreliable and recommended that doctors not use the test to assess nutritional status or suspected environmental exposures.

Similarly, in June 2004, the Hong Kong Medical Journal described three sick children thought to have heavy-metal poisoning. Hair-sample analyses showed toxic levels of various heavy metals, so chelation therapy was offered. But none of the children had a history of exposure or specific clinical features of heavy-metal poisoning. Subsequent blood tests for heavy metals were found to be normal, disproving the diagnosis of heavy-metal poisoning. The doctors likewise questioned the validity of hair analysis.

Shampoos, dyes, permanents and external contamination all affect hair mineral levels. And no one agrees on measurement methods (like pre-washing before hair analysis) or on normal values for various hair minerals, so laboratory quality control is impaired. For instance, acetone, a common pre-washing agent, removes bromine and calcium while adding zinc, copper and mercury. Manganese, arsenic and lead levels differ between whites and blacks. Antimony content varies considerably in different areas of the scalp. Hair aluminum increases with age, while other elements decrease. Finally, data linking hair mineral levels to their effects in the organs are largely unavailable. These are all major barriers to accurate hair analysis, and advice based on these analyses may be incorrect or even harmful to one's health.

Dr. Landis Lum is a familypractice physician for Kaiser Permanente and an associate clinical professor at the University of Hawai'i John A. Burns School of Medicine. Send questions to Prescriptions, Island Life, The Honolulu Advertiser, P.O. Box 3110, Honolulu, HI 96802; islandlife@honoluluadvertiser.com; or fax 535-8170. This column is not intended to provide medical advice.