In medicine, the definition "normal" has several meanings. It is used to distinguish a "normal" or healthy person from the abnormal or unhealthy individual, and in this context, we refer to "normal" iron levels as values that reflect good health, when in fact "normal" iron concentrations can be found in the presence of disease.

In the absence of disease or disease symptoms, a person is medically and legally considered normal or healthy; however, a person without symptoms of disease does not necessarily enjoy good or even optimal health.

In the laboratory, "normal" is used to describe a set of laboratory results that is based on statistics. For many analytes such as serum iron or aluminum, whole blood lead or urine mercury, reference levels have been established by the Center for Disease Control (CDC) . For other analytes less known in conventional medicine, including the important blood chromium or urine nickel, reference values have not been endorsed. This applies for most elemental reference ranges in hair, with arsenic being one of the exceptions.

In some cases, instrumentation dictates that in spite of existing reference ranges, a new set of values needs to be established. A difference in instrument sensitivity and detection limits is generally the reason. For instance, while MS ICP provides detection limits in the PPB range, traditional ICP-AES measurement of blood lead concentrations are more difficult due to instrumentation limitations and interference. However, the experienced analytical chemist can statistically and mathematically correct some of these problems by carefully evaluating the known concentrations of standard solutions. Thus, the establishment of a specific set of reference values that is based on existing laboratory conditions and variables, provides a valid rational for establishing "laboratory-specific" reference values. While the "normal" range of such "laboratory-specific" reference value deviates from official or otherwise used "normal" ranges, analytical data that is outside that range reflects abnormal conditions. It provides reliable data and is accepted laboratory procedure.

When no reference values are officially available or recommended by the CDC or other government agency, the laboratory has the responsibility to establish reference ranges by following standard laboratory procedures. Since 1994, ALL laboratories engaged in testing human samples have to provide licensing inspectors with sufficient statistical data that evidences the statistical and mathematical basis that led to the establishment of a specific set of reference values. In fact, all reference values were established in this manner. The only difference is that ‘established’ reference ranges have demonstrated to define "normal" for a longer time, and more clinical data is available to support them.

2. Statistical Basis:

For the establishment of reference ranges, the Gaussian distribution is generally used, although additional statistical and mathematical measurements are taken into account. The following figures demonstrate a particular distribution curve with specific mathematical and statistical properties:

    REFERENCE RANGE = 30 - 50 ppm

Valkovic V. Human Hair Vol I & II. CRC Press 1988