Biological samples are complex in nature. Because some may be difficult to obtain and may be available only in small amounts, rigorous protocols for collecting these samples are needed. In addition, it is crucial that the specimens used for biomonitoring truly reflect the composition of the original sample. Maintaining a sample’s representativeness starts when separating samples into specimens, generally performed to reduce repeated thaw/freeze cycles, thus minimizing potential contamination and degradation (National Research Council 2006
). The sample must be fully thawed (if frozen before) and mixed before making aliquots, and care must be taken to ensure the correct labeling of each specimen.
Further, the concept of representativeness may be of particular interest in the case of samples collected from infants, young children, and pregnant women and in situations where cross-contamination of the specimen with other tissues/fluids can occur. For example, urine collected from a woman during her period could be tainted with blood, and contamination of seminal fluid with urine cannot be ruled out. In these situations, we recommend that the potential for contamination be noted. Furthermore, guidance for the collection of samples to minimize potential cross-matrix contamination in such situations is needed. Amniotic fluid, cord blood, and meconium are promising matrices for assessing prenatal exposures, a period when humans are highly susceptible to potential adverse health effects from exposure to certain chemicals. If cross-contamination of the specimen occurs, biomarkers measured in meconium and in amniotic fluid would reflect exposure not only during gestation but also during the neonatal period or during delivery, particularly for ubiquitous chemicals or those commonly present in medical settings. Therefore, it is critical that the personnel responsible for collecting the samples appropriately document all events related to the collection and communicate them to the study principal investigator and the analytical laboratory personnel.
Cross-contamination of amniotic fluid with the mother’s blood during delivery might affect primarily the amniotic fluid concentrations of persistent chemicals that are normally measured in blood or serum/plasma. By contrast, cross-contamination of meconium or another matrix with urine would likely have a bigger impact for nonpersistent chemicals, which are metabolized and eliminated primarily in the urine, than for persistent chemicals that undergo rather limited urinary excretion. To minimize the potential impact of cross-contamination of meconium/feces with urine, additional measures for standardizing the collection procedures, such as avoiding the use of diapers containing meconium/feces that also appear to be wet, can be implemented (Calafat and Needham 2008
). Although measuring chemicals in complex biological matrices is analytically possible (Needham et al. 2005
), because of potential uncertainties during collection, interpreting the concentrations of biomarkers in matrices with relatively high potential for cross-contamination should be conducted cautiously.
One other consideration that may affect the representativeness of a given sample relates to the collection of urine by using absorbent materials (e.g., diapers). First, one must ensure that these materials do not contain the target chemical. Second, unlike urine collected directly in a urine cup, bag, or similar container, these specimens need to be extracted from the absorbent material before their analyses (Lee and Arbuckle 2009
). As expected, the urine, other urinary biomolecules or solutes, and both conjugated and free urinary species of the target chemicals will be only partially recovered, and the composition of the extracted urine will change. The extraction efficiency of a given compound relates to its aqueous solubility, which strongly depends on its chemical structure—which determines its physicochemical properties (e.g., lipophilicity, ionizability)—and on the nature of the solution (i.e., urine), which is affected by pH, ionic strength, temperature, and other solutes (Kerns et al. 2008
In general, the recovery from absorbent materials of the urinary conjugates of a chemical will be higher than that of the less hydrophilic free species. Because organic chemicals are excreted mostly as urinary conjugates, interpretation of biomonitoring results should not be affected considerably provided an adequate extraction of the conjugates exists. Nonetheless, because of the differential extraction losses, exposures to organic chemicals, if estimated from the concentrations of free and conjugated (i.e., total) species in urine collected from absorbent materials, may be somewhat underestimated, whereas the fraction of the chemicals excreted as conjugates may be overestimated. Therefore, urine sampling methods from infants and young children (Lee and Arbuckle 2009
) should be examined for their potential impact in the exposure assessment process. Other methods that do not require using absorbent materials should be evaluated.