Table 2:

Summary of recommended biological specimens to evaluate dietary mercury exposure

Specimen	Measurement recommended?	Exposure period	Comments/limitations
Blood*	Yes	Last 3 mo	Most common, accurate measure of postnatal methylmercury exposure10 Hard to ascertain if there have been acute exposure events that may explain elevated blood level10 Possibly affected by hematocrit because mercury binds to hemoglobin47
Hair†	Yes	About 1 mo for every cm of hair	Less invasive and sample is easily preserved10 Well correlated with blood measurement Approximate delay of 20 d between exposure and excretion10 May be analyzed by cm for a retrospective portrayal of mercury exposure over time10 Limited usefulness when hair is short May be affected by permanent hair treatments, natural hair colour and structure48
Fingernail or toenail	No		Few reference values10 Exposure period hard to establish Limited usefulness in clinical practice
Urine	Not for methylmercury		May be used to confirm recent exposure to inorganic mercury10
Plasma	No		Very low mercury fraction in plasma49
Cord blood at birth	Yes	Last wk of gestation	Considered best available measure of prenatal mercury exposure10 Few reference values10 Mercury level is about 1.5 higher than in maternal blood at delivery15

^*If patient’s hair is long enough and resources permit, combine with measurement in hair sample (analysis by cm if possible) to clarify whether exposure is long term or acute.

^†There may be important differences in blood-to-hair mercury ratio between people; singular use of hair sample and of hair-to-blood concentration conversion of 250 proposed by the World Health Organization should be discouraged when establishing individual risk in a clinical setting.