The toxicokinetic process is complex and dynamic and may vary based on demographic variables such as age, sex, or race/ethnicity or may change with diet, coexposures (e.g., environmental chemicals, alcohol, tobacco, and medications), and certain medical conditions. The variability in the toxicokinetic process makes interpretation of biomonitoring data inherently complex. It is difficult to gauge the differences in the magnitude of exposures among individuals based upon biological measurements alone because their metabolism may play a key factor in these differences. Measurement of urinary biomarkers allows direct assessment of exposure and dose for some pesticides. For example, 2,4-D is largely unmetabolized, and > 95% is eliminated in urine (Sauerhoff et al. 1977
). Multiple elimination routes and variable metabolism can complicate the measurement and interpretation of biomarkers for other chemicals. For example, atrazine has been shown to have multiple urinary metabolites (Buchholz et al. 1999
). Large interindividual differences in the formation of the metabolites, and changes of the metabolite profile at different times after exposure further complicate interpretation of exposure and dose using urinary atrazine biomarkers (Buchholz et al. 1999
). Other pesticides may undergo significant elimination through feces and sweat that might not be accounted for in urine collections. Understanding intra- and interpersonal variability in metabolism (e.g., the ability to appropriately oxidize molecules using cytochrome P450 enzymes or detoxify activated molecules with paraoxonase activity) and excretion will allow better assessment of the uncertainty in concentration measurements and dose estimations. Regardless, the interpretation of biological data remains a complex process and should be made with caution.
Interpretation of biological measurements can be confounded in several ways. The farmworker may have been exposed to the pesticide of interest in the days before the monitored activity. In this case, the biomarker level may not be at a baseline before sample collection in the study. In other cases, the farmworker may be exposed to the chemical in days subsequent to the monitored activity, potentially interfering with results from multiday postapplication sample collections. Correction of pre- or postactivity concentrations may be difficult or impossible unless the other exposure scenarios are well defined and the uptake and elimination kinetics well understood. Where feasible, studies of exposure resulting from specific pesticide handling or work tasks should be performed with no other handling or work task within several days before or after the monitored activity. If that is not feasible, information about the time, duration, and amount of handling and work tasks should be collected through questionnaires to allow better interpretation of the monitored activity.
Interpretation of biological measurements can also be complicated if the metabolite of interest can be formed by different parent pesticides. For example, 1-naphthol is a metabolite of both naphthalene (used as a moth repellent and also a polycyclic aromatic hydrocarbon from combustion processes) and carbaryl (Shealy et al. 1997
). In this case, it may be important to obtain information regarding potential exposures to the other chemicals. The most specific biomarker available for a particular chemical should be used whenever possible to simplify interpretation. Emerging work also suggests that people can be exposed to the metabolite of a chemical in the environment (Morgan et al. 2005
; Wilson et al. 2003
) or through the diet (Lu et al. 2005
). If the metabolite is absorbed in the body, then distinguishing between exposure to the parent chemical and the metabolite may create uncertainty in the interpretation of the measurement. Measurements to assess potential exposures to metabolites should be performed to determine whether the metabolite is present in the worker’s environment in sufficient quantities to interfere with the biomonitoring study.
Successful interpretation of biological measurements often requires collecting important information from the farmworker regarding the activities resulting in pesticide exposures. Information about the start and completion time for important activities is needed in order to place the exposure in correct relation to the biological sample collection timing. Information about the work task can be used to assess differences between farmworker measurement results. Such information might include use of specific equipment or procedures and the number of times activities were repeated. Information about other potential exposure factors may be collected to aid interpretation of results. Such factors might include use of personal protective equipment or engineering controls, hygiene activities and timing, food consumption or smoking during the work period, and possible spills or equipment leaks during pesticide handling (Quandt et al. 2006
There are several issues regarding interpretation of urinary measurement results. For spot samples, the concentration measurement may not be representative of elimination over longer time periods because of short-term volume and excretion rate differences. For 24-hr samples, there can be a high degree of intra- and interindividual variability in 24-hr urine volume, making uncorrected comparisons of concentrations across days and between people difficult to interpret. Researchers often fail to consider that the 24-hr collection period does not necessarily translate directly to a 24-hr excretion period—a period that may be several hours shorter or longer depending on void times. Studies relying on single 24-hr samples should take the excretion period into account when comparing results among applicators.
Urinary creatinine is often used to adjust for urine volume in biomonitoring for exogenous chemical exposures. For some chemicals, such corrections have been shown to reduce uncertainties (Barber and Wallis 1986
). However, for some biomonitoring studies the creatinine corrections have failed to improve, or have actually increased, uncertainties (Berlin et al. 1985
). There are two reasons why creatinine corrections may not be appropriate for pesticide biomonitoring. First, there is a wide range of normal creatinine excretion in healthy adults, making it difficult to compare creatinine-adjusted results between people (Alessio et al. 1985
; Barr et al. 2005b
). Second, relatively large intraindividual daily variability in creatinine excretion has been reported, suggesting that even creatinine corrections between 24-hr periods for the same individual may not be appropriate (Greenblat et al. 1976
). The underlying physiologic process of creatinine formation and excretion can be dependent on several factors, including age, gender, diet, exercise, muscle mass, and underlying disease (Boeniger et al. 1993
). These processes may not be the same as those governing metabolism and excretion of pesticides.
Other parameters such as specific gravity or osmolality have been suggested as a way to adjust for variable urinary volumes. Another approach using excretion rates has not been widely reported but may deserve consideration and testing. Under this approach, the excretion rate is calculated for the urine sample and used in analysis rather than the concentration. To calculate the excretion rate information about the total urine void volume, the start and end times for collection of the urine sample, and the time of the previous void before starting to collect the urine sample must be known. This approach has the potential advantages of accounting for all of the excretion during a given time period and eliminating the urine volume issue. This approach might not be appropriate if pesticide or pesticide metabolite excretion is found to be dependent on internal urine production rates.