Chapter 17 Hair Mineral Analysis
Introduction
In recent years, several attempts have been made to standardize HA testing techniques,1 but laboratories still do not have an agreement on procedures for handling a hair sample.2–4
HA has been used successfully to test for drug abuse,34–41 and studies have been performed in several locations examining the validity of using HA to test drug use before the reinstatement of driving licenses.42,43 However, drug metabolites are not found on most commercial HAs.
HA appears to offer potential as a correlating diagnostic tool in a few of the listed conditions, although hair mineral patterns should not be used exclusively for diagnosis. For example, the high hair sodium values in infants with cystic fibrosis show very little overlap with those in controls, and one study demonstrated that children with learning disabilities can be diagnosed with 98% accuracy because of a consistent pattern of high hair values of cadmium, manganese, and chromium in conjunction with low values of lithium and cobalt.5 The usefulness of HA as a research tool can hardly be questioned. However, significant controversy exists about using the method for the clinical diagnosis of diseases other than heavy metal toxicity and as an indicator of nutritional status. Moreover, multiple studies question the inter- and intra-laboratory accuracy of HA.44–48 The difference between research and clinical use is significant. Although it may be of interest in a research setting in which patients with various skin conditions have lower mean levels of hair magnesium than controls, the two groups overlap so much that the procedure is diagnostically useless.18 Moreover, even when an altered HA pattern was associated with a disease, generally there have been few investigations as to whether mineral supplementation would affect the clinical condition or even revert the hair mineral pattern to normal.
Wide overlap of values in disease and control groups is frequently the case with HA, resulting in excessive false-positives and false-negative results. For example, mentally retarded patients have been found to have lead, sodium, and potassium hair values approximately twice those of controls,13 but the standard deviations are extremely large for some minerals (for sodium, 1644.71 ± 1814.93 versus 744.43 ± 1987.0018; and for potassium, 870.15 ± 1009.19 versus 408.35 ± 689.99), and the large overlap greatly reduces the clinical use of HA. In addition, many of the altered HA patterns associated with the diseases previously listed are as yet unconfirmed.
Specific Minerals
Calcium and Magnesium
Hair calcium and magnesium values were found to be elevated in patients with fibromyalgia.49 High hair calcium content has been associated with a reduced risk of coronary heart disease.50 Low hair calcium content has been found in the last trimester of pregnancy,51 but hair calcium content increased in response to supplementation during pregnancy.52 Low hair magnesium levels have been reported in autistic children, children with attention deficit hyperactivity disorder, patients with various skin disorders, and patients with several types of leukemia, whereas high levels have been reported in conjunction with dyslexia and Prader-Willi syndrome.12,17,18,53–55 The meaning of these associations remains unknown for the most part.
A 2001 study concluded that analysis of hair calcium and phosphorus content was of value as a complementary detection tool in abnormalities of bone metabolism.56 Elevations in both calcium and magnesium were correlated with a low dietary calcium/magnesium ratio in one study, suggesting that this finding may be indicative of an induced hyperparathyroidism,57 but that hypothesis remains unproven. Hair calcium and magnesium also vary in response to the hardness and pH of the water in which the hair is usually washed.50 Supplementation of dietary magnesium has been reported to increase hair magnesium levels in deficient children.58 Nonetheless, in one study of congenital hypomagnesemia, researchers concluded that hair magnesium level was not a useful tool in monitoring mineral status because the values were higher in affected subjects than in subjects who were not deficient.59
Chromium
Hair chromium content is low in insulin-dependent diabetics,20 although there is much overlap with normal persons. Chromium levels decrease with age,22 but the meaning of this change remains unknown. Hair and tissue chromium levels vary greatly during pregnancy, being very high during the first few months of normal pregnancy and subsequently decreasing.60,61 Late in pregnancy, hair chromium content typically becomes low,62 suggesting deficiency.34 However, high hair chromium value in pregnancy is associated with low birth weight infants.63 In patients with gestational diabetes, hair chromium content is high early in pregnancy but decreases in late pregnancy.61 Increasing dietary intake of chromium has been linked to increasing hair chromium values,64 but supplemental chromium does not seem to alter hair levels.60,65
Copper
Oral contraceptive use is associated with decreased hair copper and increased serum copper.66 High hair copper levels are associated with being female, lactation, idiopathic scoliosis, and pregnancy in some, but not all, studies.67–70 Surprisingly, conditions that affect systemic copper status have been shown not to affect hair levels. Copper deficiency,71 Wilson’s disease,72,73 and cirrhosis74 do not significantly alter hair copper content. Hair copper levels also vary with geographic location.75 However, fur and liver copper values have been found to correlate in rats,74 and one study reported that supplemental copper raises hair copper levels.76 Hair color also has been found to influence the levels of copper in the hair.70,77
At this time, hair copper measurement appears unreliable for clinical application.67
Manganese
Levels of hair manganese in mothers of infants with congenital malformations and their offspring were significantly lower in one study,8 which may allow maternal hair manganese levels to be used as an indicator of the risk for malformations. Both nonsignificantly altered78 and normal levels of manganese79 have been reported in patients with epilepsy.78 Hair manganese values have been reported to be elevated in people with violent behavior, with varying levels of significance.79,80 Evidence that manganese supplementation affects behavior does not appear to exist at present.
Selenium
Levels of selenium in well water and hair show good correlation.81 High hair selenium levels are seen in toxicity,82,83 and low levels are seen in deficiency.84 Low hair selenium values have been reported in babies with neural tube defects and their mothers,9 as well as in patients with phenylketonuria.85 Tissue selenium values reflect short-term variations in intake,40 and hair levels of the element rise significantly after supplementation. 86–88
Sodium and Potassium
It is generally accepted, even by proponents of HA, that hair sodium and potassium do not reflect dietary status. High elevations of hair sodium may be diagnostic in cystic fibrosis,21 but require confirmation. A relatively low sodium/potassium ratio has been reported in celiac disease.88 Although many HA advocates cite low hair sodium and potassium as indicative of “adrenal exhaustion,” the only (preliminary) study exploring this subject reported that hair sodium and potassium did not correlate with adrenal function.89 Except for cystic fibrosis, hair sodium and potassium appear to hold little promise for clinical use.
Zinc
Hair zinc levels have received more research attention than any other mineral. Low hair zinc has been associated with zinc deficiency, anorexia nervosa, hyperactivity, gender, age, atherosclerosis, β-thalassemia, vegetarianism, lung cancer, leukemia, celiac disease, epilepsy in males, epilepsy in general, short stature in childhood, poverty, insulin-dependent diabetes mellitus, neural tube defects, and during pregnancy.* It also has been reported in neonates if the time between pregnancies is short.108
Because a few of these conditions have been associated with potential zinc deficiencies and supplemental zinc has been shown to increase hair zinc levels, practitioners who use HA often rely on hair zinc as an indicator of zinc status. However, one trial reported that hair zinc levels declined after supplementation.†
Other factors that affect hair zinc levels also interfere with the clinical use of this tool. Shampooing and dying affect hair zinc levels,111 as does the sex of the subject,111–115 age,70 and hair growth rate. Malnourished children have shown both low105,116,117 and high118 hair zinc.
Poor correlations between hair zinc and height,119 weight, and zinc consumption also have been reported,76,112 although one study reported a correlation among hair zinc, weight, and zinc consumption.119 Another study found that obese people of both sexes had higher hair zinc than those of normal weight and that there was a correlation between the degree of obesity and higher hair zinc levels.120
Although low hair zinc levels have been reported in patients with insulin-dependent diabetes mellitus,109 generally there is considerable overlap between cases and controls. To further complicate the picture, in a study of female children of non–insulin-dependent diabetes mellitus parents, hair zinc was found to be significantly higher compared with women with no family history of non–insulin-dependent diabetes mellitus.121
