Biomonitoring Study

 • Title 

 • Abstract 

 • Introduction 

 • Methods 

 • Results 

 • Discussion 

 • Conclusions 

 • References 

 • Tables 

Discussion

Available biomonitoring data for 2,4-D in both the general and agricultural populations indicate that current uses and practices suggest exposures that are below the acceptable exposures identified by the U.S. EPA. A “margin of safety” is the ratio between the exposure guidance value and measured exposure. In this analysis, the exposure guidance value (RfD) was converted to a BERfD value for comparison to the measured biomarker concentrations. General population values indicate a margin of safety compared to the BERfD of approximately 200 at the central tendency and greater than 50 at the upper percentiles of exposure. In turn, the BERfD is 100-fold below the BEPOD, which is the biomarker concentration associated with chronic intake in humans at the animal-tohuman extrapolated POD. The conclusion of a substantial margin of safety holds whether comparisons are made using volume or creatinine-adjusted concentrations. Median or average urinary 2,4-D concentrations for applicators are consistently below the BE values associated with occupational exposure targets set by U.S. EPA (2004); however, evidence exists for exceptions near the occupational BE target value in a few individuals from the studied occupationally exposed populations. Biomonitoring data for spouses and children of applicators on the day following use of 2,4-D also fall below the BE values associated with general population acute exposure RfDs set by U.S. EPA (2004).

Other studies have reported related biomonitoring data. Arcury et al. (2007) studied children from North Carolina farm worker families in 2004. Multiple pesticides (or metabolites) were measured in urine samples from these children (1 to 6 years of age). The median 2,4-D concentration was below the limit of detection (LOD) of 0.2 μg/L (42% of the 60 sampled children had detectable concentrations of 2,4-D, but the range of detected concentrations was not reported). Garry et al. (2001) measured urinary 2,4-D in small numbers of forestry applicators who used a variety of methods to apply the herbicide. Backpack sprayers had the highest measured urinary concentrations during time periods of use, with a median of 160 μg/L and a range up to 1700 μg/L (n=7). Other modes of application such as use of boom sprayers or aerial applications resulted in lower urinary 2,4-D concentrations, with all measured values below 500 μg/L for boom sprayers and below 100 μg/L for other modes. These values are consistent with the concentrations observed in farm applicators from the Alexander et al. (2007) study, and are also below the occupational BERfD presented in Table 1.

The evaluation presented here is based on BE values derived from the U.S. EPA risk assessment of 2,4-D (U.S. EPA 2004). However, the Canadian PMRA has also recently estimated acceptable daily exposures to 2,4-D (PMRA 2007). The derived acute and chronic reference doses are based on the same underlying data as used by the U.S. EPA, with similar or identical choices of POD. However, the PMRA assessment generally applied total uncertainty factors approximately 3-fold lower than those applied by U.S. EPA, resulting in exposure estimates that are approximately 3-fold greater than those set by the U.S. EPA. Thus, the BEPOD values associated with the PMRA risk assessment would be essentially identical to those for the corresponding U.S. EPA exposure guidance values. Although BE values were not specifically derived based on the PMRA assessments, corresponding urinary BE values would be approximately 3-fold higher than those derived based on the U.S. EPA RfDs. BE values corresponding to the PMRA Acute RfD (ARfD) values for acute exposure in the general population and in females of reproductive age equal to 1,000 and 4,000 μg/L, respectively (2,000 and 7,000 μg/g creatinine). The BE value corresponding to the PMRA acceptable daily intake (ADI) for chronic exposure would be 700 μg/L (1,000 μg/g creatinine). Thus, reliance on the PMRA risk assessment does not change the overall conclusion of a substantial margin of safety under the various exposure scenarios.

Uncertainties and Limitations.

BE values are derived based on expected average concentrations (either volume based or creatinine-adjusted) in urine under conditions consistent with the underlying exposure guidance value (chronic or acute exposure conditions). Some variability in concentration is expected due to use of spot urine samples, inter-individual variability in creatinine excretion rates, and variability in urinary volume due to hydration status. Morgan et al. (2004, 2008) investigated the variability of 2,4-D concentrations among spot urine samples (i.e., first morning void, after lunch, and before bedtime) collected over the course of 48 hours from 28 adults and 28 children. The maximum measured spot urine value was within a factor of 3 of the mean value in 53 of the 56 individuals, consistent with previous assessments of variability among spot samples (see, for example, Scher et al. 2007).

2,4-D is relatively short-lived, with a urinary half-life on the order of one day, so for an individual in the general population, a single measurement does not characterize longterm exposure. However, the NHANES urinary data for 2,4-D are representative of the United States population and samples were collected at various times through the year. NHANES data would be expected to capture indications of higher exposures if they were occurring with any frequency, unless such variations were highly seasonal and geographically isolated. Urinary concentration data from Morgan et al. (2004, 2008) collected from two different geographical regions of the United States (NC and OH) over the course of a year suggest somewhat higher exposures than reflected in the NHANES dataset, but both sets indicate general population exposures far below health-based exposure guidance values.

A notable deficit in the available data for the general population pertains to residential uses of 2,4-D. Unlike exposures to 2,4-D users in agricultural populations, systematic evaluations of domestic use of the chemical are not available. These episodic exposures would not likely be captured in the NHANES or Morgan et al. (2008) data. To the extent that domestic applications do not result in exposures greater than those resulting from agricultural applications, human exposures should be within the margin of safety demonstrated by these existing study data. More research is needed to understand the domestic usage patterns of 2,4-D in residential settings and the resulting potential human exposures to this herbicide in the United States and Canada.

The RfD values derived by the U.S. EPA are based on non-cancer endpoints. 2,4-D has also been assessed for potential carcinogenic effects. Non-Hodgkin lymphoma (NHL) was associated with herbicides and 2,4-D in a series of case-control studies initiated more than 20 years ago (Hoar et al. 1986; Zahm et al. 1990). Subsequent case-control and cohort studies have not confirmed these early observations (Burns et al. 2001; DeRoos et al. 2003; Hartge et al. 2005; Pearce et al. 1989; Schroeder et al. 2001; Woods et al. 1987). Recent reviews of NHL (Alexander et al. 2007) and 2,4-D (Garabrant and Philbert 2002) have concluded that the epidemiologic evidence remains “scant” and unsupportive for this association.

BE values are screening values and are not intended for use as definitive measures of risk for individuals. They do not represent a bright line between safe and unsafe levels, but rather allow evaluation of biomonitoring data in a public health risk context consistent with the existing risk assessment for 2,4-D (LaKind et al. 2008). Biomarker concentrations below the BERfD indicate a low priority for risk assessment follow-up, while concentrations in excess of the BERfD but below the BEPOD indicate a medium priority for risk assessment follow-up. Values in excess of the BEPOD indicate a high priority for risk assessment follow-up. Risk assessment follow-up may include examination of the underlying risk assessment, exposure pathway investigations, or other risk management activities (LaKind et al. 2008). Acute RfDs and the corresponding BE values are targeted at isolated, single-day exposures, and are only appropriate for use in evaluating biomonitoring data when there is specific knowledge of a potential acute exposure. The biomonitoring data reviewed here for both members of the general population and applicators generally falls into the range of low priority for risk assessment follow-up, according to the guidelines for BE communication (LaKind et al. 2008).

Next: Conculsions