Prepared testimony of Robert I. Krieger (Personal Chemical Exposure Program, Department of Entomology, University of California, Riverside 92521) concerning non-occupational human pesticide exposure.

Introduction. Non-occupational uses of pesticides on lawns, gardens, and in homes are normally associated with low levels of unintentional human exposure. These chemical exposures are not illnesses, but result from unintended residue contacts (air, water, diet, treated surfaces) which result in absorption, distribution, metabolism, and clearance in urine and feces of small amounts of pesticides or their degradation products. The continual absorption and elimination of natural and synthetic chemicals from our bodies protects our vital processes from adverse effects resulting from accumulation of toxic amounts of both natural and synthetic chemicals.

Responsible risk assessment requires the best possible human pesticide exposure estimates. With respect to residential exposures of children and parents, available information is limited in scope and does not imply excessive risk.

Current regulatory concerns about non-occupational, residential exposures are about 10 years old. In 1985 Berteau and Mengle of the California Department of Health Services and Maddy of the Department of Food and Agriculture conducted preliminary review of some pesticides used indoors. They noted several cases (6) from the California Pesticide Illness Surveillance System in which illness were reported after structural pest control. The cases more likely resulted from sensory responses (especially bad odor) and confusion than from systemic toxicity based upon pesticide use practices and the nature and time to onset of symptoms. Subsequently hypothetical exposure estimates (sometimes greater than presumed toxic levels) were developed for infants, children and adults following label use for propoxur, DDVP, and chlorpyrifos (Berteau et al.,1989).

Considerable attention and effort by scientists in academia, regulatory agencies, and the industrial sector followed publication of default indoor scenarios (Berteau et al.,1989). During the same interval preparation of formal pesticide risk assessments became common practice (NRC/NAS 1984), and a premium was placed upon human exposure data. The Food Quality Protection Act of 1996 brings increased attention to monitoring human indoor and outdoor postapplication pesticide exposures.

This current effort to establish non-occupational pesticide exposures has been pursued with extremely limited knowledge of the actual sample frame of human exposure. As a result the numerous methods for monitoring indoor air and surfaces that have been developed to assess exposure, have an uncertain, if any, relationship to the magnitude and dynamics of residential pesticide exposure.

Monitoring. Since normal use is without adverse effects, only advanced analytical procedures makes it possible to confirm the magnitude and distribution of non-occupational pesticide exposures. The recent risk remediation activities following illegal use of methyl parathion showed a range (non-toxic) of urinary metabolites of derived from exposure the parent insecticide and its primary metabolite. Action levels varied by a factor of 4 between children (<1 year) and persons 5 years-of-age and older based upon EPA Risk Management Criteria (Grissom et al., Personal communication). The median metabolite levels resulting from methyl parathion misuse (18.4 ug/ml urine) are greater than background levels (0.7 ug/ml urine) reported by Hill et al. (1995). Similarly, background levels of organophosphate metabolites are routinely less than 1 ug equivalent of parent insecticide per kg body weight (Krieger, unpublished).

Human Monitoring Following Indoor Insecticide Foggers. Experimental evaluation of the exposure potential of indoor foggers was initiated shortly before Berteau et al. (1989) published their alarming default exposure estimates (mg/kg). A carefully controlled, 20-minute series of structured, high-contact, activities were selected for use by persons wearing cotton suits (socks, gloves, and union suits). High contact and efficient transfer of pesticide to the cotton suit was used to represent daily indoor skin exposure. The experimental protocol has remained virtually unchanged during the past10 years (Ross et al., 1988), and a recent manuscript details validation studies (Krieger et al., In review).

A series of urine monitoring and experimental studies which are in progress at the University of California, Riverside, have begun to define a real-world sample frame which will represent actual human experience. Urine monitoring of persons living in residences which have received routine use of pesticide foggers and area sprays has been performed. These uses probably represent worst-case residential exposures and they are an invaluable means of demonstrating factors such as time and source strength which can limit residential pesticide exposure.

Preliminary review of these data reveals that children may absorb 3-10 times more insecticide than their parents who spend similar amounts of time indoors. Absorbed dosages of both are well below toxic levels. The normal daily range of human indoor organophosphate exposures is 0.1-10 ug/kg rather than 100- to 1000-times higher as predicted by unvalidated "screening levels" calculated from environmental monitoring data (Berteau et al., 1989; Fenske et al., 1990 and 1998; Gurunathan et al., 1998).

Of additional importance is the duration of exposure following indoor pesticide use (Figure 2). Daily chlorpyrifos exposures were estimated by urine biomonitoring for TCP metabolite. Seven family members (aged 18 to 88) lived 16-24 hours per day in the 10 rooms of their 1,500 square foot, two-story home in southern California. Six flea foggers which each discharged 1.8 g chlorpyrifos depositing about 5 ug/cm² on the carpet. Following a 3-day study in 1996 in which TCP clearance increased each day for 3 days, urine specimens were collected for a more extended period in a 1997 study (Figure 2). Urine clearance of TCP was greatest during the first week (Nolan et al., 1984) and slowly returned to control levels as a result of the insecticide persistence indoors. The results were not predicted by our previous air and surface monitoring (Ross et al. 1990; 1992) or air levels reported in the literature (Figure 3). Environmental levels decline much more rapidly (hours) than human exposure (days).

Available air and surface pesticide levels following foggers or broadcast spray use do not predict pesticide exposures of residents. Crack and crevice treatments produce much lower exposure since direct contact with surface residues is substantially reduced. This generalization is confirmed by Shurdut et al. (In review) reporting very low exposures of urine monitored residents following crack and crevice organophosphate applications.

Modeling. Worst-case assumptions regarding pesticide availability and absorption have been used to estimate human exposure without reference to actual levels of absorption and clearance. In recognition of the importance of skin absorption and inhalation, suggested means to estimate human exposure by Berteau et al. (1989), Fenske and co-workers (1990; 1998), and Lioy and colleagues (Gurunathan et al., 1998) result in unreasonably high estimates of pesticide exposure, particularly for children. Regardless, these untested estimates have had substantial regulatory impact and may have groundlessly alarmed persons who view themselves to be at undue risk.

Many pathways of potential exposure and important activity patterns which might influence exposure can be described, e.g., recent studies by the U. S. EPA, but the pathways remain hypothetical due to uncertainties concerning inherent default assumptions and lack of measured human exposures. The results of EPA methods development and environmental monitoring will not immediately move us closer to estimates of human exposure until the sample frame is defined by more extensive biological monitoring.

Management. Adoption of risk assessment as a critical risk management tool provides the manufacturing, regulatory and public health sectors an opportunity to enhance public confidence in the safe use of chemical technologies. There is no substance for which a safe dose can not be determined.

Premature or inaccurate overestimates of health effects of pesticides based upon hazard identification rather than upon the best possible appraisal of their potential to have harmful effects on people are irresponsible and inconsistent with public health and regulatory experience. When risks are negligible statements such as "highest risk" and "safer" do not communicate with a confused public who are increasingly becoming conditioned to view themselves as victims rather than as benefactors of chemical technologies.

Conclusion. Home, turf, and garden uses of pesticides to protect health and property, expose residents to residues in air and on surfaces. The amount of the residue (micrograms per square centimeter) is low relative to levels which would impact health. Resulting exposures of children and adults are low but sufficient to be detected with modern analytical procedures. Pesticide levels decline more slowly indoors than outdoors so human exposures resulting from indoor applications use may be more prolonged than those occurring outdoors. Since exposures occur over a period of weeks, short term residue measurements (typically for 1 or 2 days) of air, surface wipes, or dislodgeable levels are inadequate for risk assessment because they decline much more readily than absorbed dose. At present methods are not available to reliably transform environmental measurements into estimates of absorbed dose.

Human monitoring data representing non-occupational or residential pesticide exposures are remarkably limited and presently inadequate for establishing an estimate of aggregate exposure. The reported magnitude of exposure based upon biological monitoring prompts confidence in the overall safety of pesticide use rather than alarm over excessive or harmful human exposure. Normal exposures are less than toxic levels including the lowest adverse effect level (LOAEL), less than the no adverse effect level (NOAEL), and less than regulatory reference doses derived by reducing NOAELs by traditional uncertainty factors representing interpersonal (10) and interspecies (10) variability.