Response to the NRCD Petition 

Exposures

Exposure through maternal milk:

NRDC claims that EPA’s aggregate exposure was inadequate because the Agency failed to consider that ‘‘[t]here is evidence of exposure to 2,4-D through maternal milk” citing Sturtz et al. (2000) study finding that 2,4-D is excreted in maternal milk in rats, and NRDC thereby inferred potentially significant exposures to the nursing neonates. NRDC claims 2,4-D residues were found in the stomach content, blood, brain and kidney in 4-day neonatal rats.

Sturtz et.al. has very limited value for human risk assessment. The 2,4-D treatment rates to the lactating rats were 50, 70 and 100 mg/kg bw/day, doses in excess of saturation of renal clearance of 2,4-D (vanRavenzwaay et.al., 2003; Saghir et al., 2006; 2009). The toxicity observed in the resulting non-linear pharmacokinetic range does not inform human risk assessment (EPA RED), because nonlinear doses are exceedingly unlikely to occur in humans. Thus Sturtz is not useful in assessing any potential human risk from substantially lower occupational and general population exposures to 2,4-D, including infants.

In the development of the 2,4-D RED, EPA was well aware of the potential for 2,4-D to be transferred to nursing offspring through milk and previously addressed NRDC’s concerns: “EPA is aware, as a result of animal feeding studies using exaggerated doses, that 2,4-D may be present in milk. It is not surprising that the study relied upon by NRDC suggests that 2,4-D is transmitted in breast milk given the massive doses of 2,4-D in that study of 50, 70 and 700 milligrams/kilogram of body weight/day (mg/kg/day). By comparison, EPA estimates that the maximum dietary exposure from food to human females ages 13-50 is 0.01018 mg/kg/day and the average exposure is 0.000642 mg/kg/day (USEPA Memorandum, 2002). These values range from 4,900 to 1 million times lower than the values in the cited rat study. Further, EPA’s manner of doing risk assessment for infants is protective of any pesticide exposure to infants from human breast milk because the exposure values EPA assumes for pesticides in cow’s milk greatly exceed the values that could be present in breast milk.” (US Federal Register 2005 at 46706, 46735).

The EPA estimates of average human 2,4-D exposures described above are demonstrably reasonable. Recent human biomonitoring of farm families, including non-applicator spouses and children during a period of high 2,4-D use (spring time farm application period), established a geometric mean dose of 0.00008 mg/kg/day for the spouses and 0.00022 mg/kg/day for all children (Alexander et.al., 2007). These very low-level exposures have been confirmed in an EPA-conducted biomonitoring study of 135 pre-school children and their adult care-givers from randomly selected households in North Carolina and Ohio. Despite detection of 2,4-D in samples of carpet dust from most of the surveyed households, the highest level of exposure found in a single child was 0.00028 mg/kg/day, while the 50th percentile value in the children was ten-fold lower (Morgan et.al, 2008).

For both the occupational and indoor residential exposures characterized above (Alexander et.al., 2007; Morgan et.al., 2008), a low potential to adversely impact human risk has been supported in a recent analysis by Aylward and Hays (2008). Using animal and human pharmacokinetic information, these investigators calculated Biomonitoring Equivalent (BE) values for 2,4-D exposure, which were defined as urine or blood concentrations of 2,4-D expected to result from human exposure in relation to existing health-based exposure guidelines. For 2,4-D, urine and plasma BE values were derived for exposures to both chronic and acute reference doses (RfD). Both the occupational and general population urine concentrations reported in Alexander et.al (2007) and Morgan et.al. (2008) are well below even the lowest BE calculated for the EPA chronic RfD, and even further below BE’s estimated for acute exposures, i.e., values appropriate for reference to occupational or short-term residential exposures. Urinary biomonitoring samples reflect aggregate exposure inclusive of both occupational and other incidental exposure sources, e.g., house-dust, surface contamination from outdoor to indoor track-in, etc. Thus, these data also indicate that the presence of 2,4-D in household dust or exposure scenarios as noted in the Nishioka et.al. study (1996) cited by NRDC or in Morgan et.al. (2008) does not translate to absorbed doses of 2,4-D likely to present an adverse human health risk. In summary, the presence of a detection in dust does not equate risk. Biomonitoring and bioavailability (transfer rate) studies correlate a risk factor between the levels in dust and actual exposure.

NRDC also claimed that, “EPA failed to include any lactational exposure in its aggregate risk assessment.” That statement is not correct. The language used in the EPA 2,4-D Health Effects Division risk assessment states in section 4.2.2 -- Acute Dietary Exposure and Risk, ‘The acute dietary assessment was slightly refined as the following fairly conservative assumptions were made: … Note that ½ of the average LOD from PDP monitoring data was used as the milk exposure value because no milk samples contained detectable 2,4-D residues over several years of PDP sampling.” (USEPA Memorandum 2004b)For the 2,4-D risk assessment, EPA assumed that 2,4-D would be present in milk at 0.004 ppm for both acute and chronic exposure (despite it being non-detectable in PDP sampling). EPA’s aggregate exposure assessment was protective for all children, including nursing infants.

NRDC further alleges “Since the completion of 2,4-D reregistration, additional studies have been published that confirm the lactational exposure and identify adverse effects in the offspring,” and cites the 2006 report of Sturtz, et al., (2006). However, the decreased pup body weights reported in litters of dams treated with 15 mg/kg/day of 2,4-D in the diet have not been replicated in another recent rat dietary life-stage study conducted under Good Laboratory Practice (GLP) conditions (Marty et.al., 2009). Marty and co-investigators (2009) demonstrated that 2,4-D significantly decreased pup body weights only when dams were treated with 2,4-D equal to or greater than 800 ppm in the diet continuously from pre-breeding through completion of lactation on postnatal day 21 (equivalent to 52, 61, and 71-117 mg/kg/day during pre-breed, gestation and lactation periods respectively). Pup body weights were not affected in litters from dams similarly treated with 400 ppm (25, 28 and 37-57 mg/kg/day, respectively). Importantly, these investigators also demonstrated that the body weight decreases noted at 800 ppm occurred under conditions in which 2,4-D pharmacokinetic behavior in the lactating dams was distinctly non-linear, thus rendering these responses of limited relevance to human risk assessment (Saghir et.al., 2009).

NRDC adds a minor discussion on potential effects on the brain of neonatal rats exposed lactationally to 2,4-D, citing recent 2006 and 2007 papers by the same research group in Argentina. These studies were conducted at high doses of 2,4-D (well above pharmacokinetic nonlinearity), and included use of an unrepresentative route and mode of administration (intraperitoneal) which would further amplify nonlinear pharmacokinetic performance (Garcia et.al., 2004; Ferri et.al., 2007;;Garcia et.al., 2006).

The Industry Task Force II on 2,4-D Research Data has completed the in-life portion of a GLP one-gen reproduction study that examines potential life-stage impacts of dietary 2,4-D administration to rats on reproductive, endocrine, developmental neurotoxicity and immunotoxicity responses. As previously noted, the Task Force is conducting the study to meet the requirements of EPA’s Data Call-In Notice. The Task Force anticipates that the final study report for this study will be submitted to the Agency in 2009.

Allegation: Underestimated dermal absorption factor

For dermal absorption risk assessment NRDC claims that EPA used a dermal absorption factor of 10 percent which is inappropriate considering the synergistic effects of other exposures.

Among the studies EPA considered to establish the dermal absorption risk factor was a Feldmann and Maibach (1974) human study that showed absorption of 5.8% ± 2.4 (3.4-8.2%) of 2,4-D applied to the forearm skin. This study was an extreme test using 14C 2,4-D applied with an acetone vehicle. Acetone tends to denature the skin, a process in which proteins or nucleic acids of the skin lose their 3D-structure, thus allowing for increased potential dermal absorption. The skin was not protected and area not washed for 24 hours for maximum absorption.

Recently, Ross, et.al. (2005) reviewed and summarized numerous dermal absorption studies and determined that the human percutaneous absorption of 2,4-D has been well characterized. These investigators examined five published studies using human subjects, including Feldmann and Maibach (1974), and concluded these studies exhibited remarkable reproducibility across a span of three decades and multiple laboratories, formulations, and methods. They also noted that while it is considered rare to have even two human dermal studies available to assess dermal absorption of a specific chemical product, having five 2,4-D studies was exceptional. These human data provide valuable perspective for characterizing the variability (CV=60%) and central tendency (mean = 5.7% dermal absorption) across the published studies, and supported the conclusion that the central tendency value, which was in close agreement with the finding of Feldmann and Maibach (1974), can be used with confidence in human risk assessment. Ross et al. further observed that the human dermal absorption determinations predicting a low percentage of 2,4-D dermal absorption were in excellent agreement with findings from worker biomonitoring studies in which measured internal doses were very low and ranged from 2.0-5.2 µg/kg body weight. In addition to the comprehensive analysis of Ross and co-workers, EPA has previously acknowledged that, although the literature denotes a 2,4-D dermal absorption of 5.8%, a 10 percent dermal absorption factor was selected as a protective value given study variation (USEPA 2005).

Allegation: Dermal absorption enhanced by other factors

NRDC alleges that the dermal absorption of 2,4-D may be elevated in the presence of several interacting factors such as use of sunscreen, alcohol consumption and the insect repellent DEET. EPA has specifically responded to this allegation in the past as follows:,

“With regard to the dermal absorption value used and the use of insect repellents and/or sunscreens, in reality, those same farmworkers describe[d]… as not having the opportunity to change their work clothes and/or shower, most likely would not be using these potential enhancer products either. [For the Moody et.al., 1992, study specifically examining DEET]… the conclusion that the data demonstrate a difference between exposure with [14±4.5%] and without [10±11.5%] DEET is not supported given the magnitude of the standard deviation. What can be gleaned from the study is that significant exposure can occur from hand contact, and taking measures to limit dermal exposure; e.g., washing after exposure and or the use of chemical-protective gloves, is recommended.”

Concerns for potential interacting substances to impact dermal absorption has been further mitigated by personal protective equipment (PPE) changes implemented by EPA in the 2005 re-re-registration document (RED) of 2,4,-D (USEPA 2005, p.114). The RED describes the PPE requirements for liquids, wettable powder, formulations in water-soluble packages and water-dispersible granules as:

“All mixers, loaders, applicators, flaggers, and other handlers must wear:  long-sleeved shirt and long pants, shoes and socks, plus chemical resistant gloves, when applying postharvest dips or sprays to citrus, applying with any handheld nozzle or equipment, mixing or loading, cleaning up spills or equipment, or otherwise exposed to the concentrate, chemical resistant apron when applying postharvest dips or sprays to citrus, mixing or loading, cleaning up spills or equipment, or otherwise exposed to the concentrate.”

The requirement to use chemical resistant gloves in demonstrated conditions of intensive exposure potential has in fact been demonstrated to significantly mitigate 2,4-D exposures. The EPA RED decision regarding glove use in high-potential exposure conditions is consistent with a robust peer-reviewed literature which has repeatedly demonstrated that use of protective rubber gloves alone dramatically reduces the total absorbed dose 2,4-D (Alexander et.al., 2007; Arbuckle et.al., 2002; Hines et.al., 2001; Harris et.al., 2002). The focus on hand protection is also consistent with a report that concluded contact with hands accounted for 80-90% of potential cumulative worker exposure to 2,4-D (Grover et.al.,1986). Thus, in addition to EPA’s previous comments addressing why the proposed chemical interactions are unlikely to impact dermal absorption, it is also clear the potential impacts of such postulated interactions would be significantly reduced by the EPA RED- mandated call for use of chemical-resistant gloves in high-intensity exposure conditions.

Allegation: Rubber glove permeability is enhanced by DEET and sunlight

NRDC infers that glove use will not afford adequate dermal protection by citing a study (Moody and Nadeau, 1992) as concluding that simultaneous exposure of rubber gloves to “DEET and sunlight” rendered them “highly permeable” to 2,4-D penetration. NRDC then postulated that if enhanced penetration occurred it would be associated with increased dermal absorption due to glove occlusion of the skin.

Moody and Nadeau (1992) examined total 14C-ring labeled 2,4-D permeability across a 1.5 cm2 cut-out of natural rubber latex gloves incubated in laboratory apparatus at 37º C for 48 continuous hours. This study was primarily intended as an hypothesis-generating experiment and does not provide any definitive conclusions for enhanced risk of 2,4-D dermal penetration associated with glove use for the following reasons:

1. although the NRDC petition infers that DEET alone enhanced 2,4-D permeability, the authors concluded DEET did not significantly impact 2,4-D penetration through the rubber (2.4 ± 1.8% with DEET vs 3.2 ± 3.46 without DEET);
2. in what can only be regarded as preliminary evidence (n = 2), the permeability of 2,4-D was increased to 6.2 ± 0.73% when the rubber was simultaneously exposed to a UV lamp emitting UVA (not “sunlight”; non-UVA control 0.3 ± 0.14% 2,4-D permeation);
3. neither DEET alone (no effect on permeation) or UVA alone rendered the rubber “highly permeable” to 2,4-D despite the relatively severe experimental conditions;
4. only permeation of total radioactivity was measured and thus the authors themselves could not exclude that the postulated enhanced permeation with UVA-cotreatment was represented by 2,4-D photolysis product(s); and
5. the authors stated the field relevance of the observations remained to be established.

As noted above, however, several well-conducted studies have demonstrated that use of rubber gloves in fact significantly reduced 2,4-D dermal absorption in actual field conditions. These exposure reductions occurred even when no special precautions were taken to assure use of undamaged gloves (Alexander et.al., 2007; Arbuckle et.al., 2002; Hines et.al., 2001; Harris et.al., 2002).

Allegation: Dermal penetration implications of occlusion and/or soaking of 2,4-D into clothing not adequately assessed

As summarized above, repeated field studies have confirmed that protection (“occlusion”) of skin by use of rubber gloves dramatically reduces 2,4-D dermal absorption.

In a comprehensive study, Lavy, et al. (1987) and co-workers examined the overall absorption of 2,4-D in four types of intensive exposure 2,4-D forestry applications: backpack, injection bar, hypohatchet, and hack-and-squirt. For each of the four types of exposure scenarios, 2,4-D absorption was evaluated in four crews of 20 workers each under two exposure conditions: 1) workers wore usual clothing but no additional PPE and followed normal work habits; and 2) workers were issued new leather gloves and leather boots for spraying and were also required to use all feasible precautions to reduce exposure (e.g., use neoprene gloves for mixing/filling; wash hands before rest periods; bathe and change clothes as soon as possible after work; etc.). For all types of applications except backpacks, use of new gloves, boots and other precautions reduced overall 2,4-D absorption. Use of the additional exposure precautions did not, however, impact 2,4-D exposures in backpack applicators. Although 2,4-D exposure was highest in these applicators and was attributed to the observation that the clothing of both groups often was saturated with spray, dew or perspiration, it is important to note that even despite the extreme exposure conditions associated with forestry backpack application the average applicator exposures averaged 0.0876 – 0.0980 mg/kg. These exposures were not considered health threatening by the authors in that they estimated a margin of exposure of 272 relative to the referent animal toxicity NOEL of 24 mg/kg/day. Thus, a low dermal absorption potential for 2,4-D was demonstrated even under conditions of intensive exposure such as that represented by saturation of clothing, Agricultural and lawn care applicator scenarios also have been shown to result in dermally absorbed doses that are significantly less than forestry backpack applicators. For example, the geometric mean exposure to 2,4-D farm applicators has been reported as 0.00246 mg/kg/day (2.46 µg/kg/day; Alexander et.al., 2007), while professional lawn care applicators exposure did not exceed 0.006 mg/kg/day (6 µg/kg/day; Yeary, 1986) and homeowners applying liquid 2,4-D formulation with no gloves or other PPE precautions (and evidence of direct spills of concentrate on bare skin during applications) experienced a maximum 2,4-D dose of 0.0071 mg/kg (7.1 µg/kg; Harris et.al., 1992). Importantly, homeowners using recommended PPE such as rubber gloves and/or using granulated formation experienced mostly non-detectable exposure (Harris et.al., 1992).

Allegation: When tracked indoors, 2,4-D persists in carpets for up to one year after a single turf application.

NRDC states that 2,4-D tracked into homes persists in carpets for up to one year at concentrations of 0.5 µg/gm (Nishioka et al., 1996). Nishioka et al., 1996 calculated that value for carpet dust from short-term carpet dust samples collected in a tracking study. The value of 0.5 µg/gm is higher than the average 2,4-D concentration of 0.156 µg/gm found in carpet dust samples collected in Ohio homes of 2,4-D applicators (Morgan et al. 2008) and substantially higher than the average level of 0.0475 µg/gm found in similar applicator’s homes in North Carolina. Morgan et al., 2008 reported exposure levels of spouses and children of the applicators and found the maximum exposure for a child to be 0.00028 mg/kg/day while the 50th percentile exposure in children was ten fold lower than that. These exposures are substantially below the oral reference dose for 2,4-D.

Next: Conclusion