ENVIRONMENTAL RESEARCH 59, 22%237 (1992)

Organophosphate Pesticide Exposure in a Group of Washington State Orchard Applicators CATHERINE K A R R , * ' t P A U L DEMERS,*':~ L U C I O G . COSTA,* W I L L I A M E. DANIELL,*':~ SCOTT BARNHART,*':~ MARY MILLER,'~'~: G E N E GALLAGHER, *A S. W . H O R S T M A N , * ' § D A V I D E A T O N , * AND L I N D A ROSENSTOCK*':~

Departments of ~Medicine and *Environmental Health, University of Washington, Seattle, Washington 98014; §Department of Preventive Medicine, University of Kentucky, Lexington, Kentucky 40506; and ~Safety and Health Assessment and Research Program, State of Washington Department of Labor and Industries Received May 15, 1992 As part of a study to investigate the potential for organophosphates to cause chronic neurologic sequelae, we assessed the pesticide exposure experience of a group of Washington State apple orchard applicators. Seasonal monitoring of cholinesterase activity for 48 regular organophosphate applicators and a control group of 40 slaughterhouse workers was performed. A subset of the pesticide applicators participated in an in-depth exposure assessment. This involved observation of spraying activities during 1 spray day, as well as cholinesterase monitoring and dermal exposure assessment using a fluorescent tracer in the pesticide formulation. Comparison of seasonal red blood cell cholinesterase change in pesticide workers according to exposure level, characterized by frequency of pesticide spraying and protective equipment use, showed lower cholinesterase levels among higher exposed groups compared to lesser exposed groups. In-depth exposure assessment revealed exposure primarily on the head and hand regions. Subclinical changes (less than 15% inhibition) in red cell cholinesterase correlated well with dermal exposure calculations. This study suggests that cholinesterase monitoring may be a useful biological marker for even subclinical organophosphate pesticide effects. © 1992AcademicPress, Inc.

INTRODUCTION Estimates suggest that 80,000 cases of pesticide-related illness occur annually in U.S. farmworkers (Coye et al., 1986a,b). At highest risk are workers who handle pesticides directly, such as applicators, mixers, and loaders. Organophosphate insecticides are the most frequently implicated (Murphy, 1986). Organophosphate poisoning is the result of acetylcholinesterase inhibition in the peripheral and central nervous system. Central nervous system (CNS) symptoms have been reported as sequelae to cases of acute poisoning (Rosenstock et al., 1990) and symptoms have been observed in association with chronic exposure to levels of organophosphates insufficient to cause evidence of symptomatic acute intoxication (Rosenstock, 1987). For occupational exposures, dermal absorption is recognized as the most important route of exposure and a variety of methods to determine exposure by this route have been employed. The patch technique has been used widely in exposure 1Deceased. 229 0013-9351/92 $5.00 Copyright© 1992by AcademicPress, Inc. All rightsof reproductionin any form reserved.

230

KARR ET AL.

assessments alone or in combination with other techniques (Franklin et al., 1981; Niss et al., 1983; Atallah et al., 1982; Weisskopf et al., 1988). The patch technique involves the placement of patches at points close to body parts that come into contact with the pesticide and subsequent extrapolation of patch data to exposed areas. Recently, a variety of tracer materials, often fluorescent dyes that can be visualized under long-wavelength ultraviolet light (black light), combined with a skin scrub procedure, patching, or video imaging, have demonstrated potential value for quantitative dermal assessment (Franklin et al., 1981; Fenske, 1986). Traditionally, blood cholinesterase determinations have been the most widely employed method of biological monitoring of organophosphate exposure. This method has served as a basis for monitoring worker exposure, establishing reentry intervals, assessing changes in application practices, evaluating personal protective equipment, and evaluating field application equipment (Coye et al., 1986a,b). An important limitation to the use of blood cholinesterase activity is the potentially great interindividual variation, which compromises meaningful interpretation of postexposure measurements without a baseline measurement. Inter- and intralaboratory variations also pose difficulties with the use of cholinesterase measurements (Coye et al., 1986b). In addition, the sensitivity of cholinesterase activity to low or chronic levels of exposure may be poor, as dermal absorption of low levels of organophosphates by fieldworkers was insufficient to cause a significant decrease in blood cholinesterase levels (Franklin et al., 1981; Knaak et al., 1979). This study was designed to provide information on the pesticide exposure experience of a group of Washington State apple orchard applicators. The assessment of whether the pesticide exposure caused decrements in neuropsychological performance in this population is described in a companion article (Daniell et al., 1992). Washington farms employed 110,700 farm workers during peak employment in 1990 (WA State Employment Security Department, 1991). Apple production uses more pesticides than any other crop in the United States (Ferguson, 1985). The relation between cholinesterase monitoring data and other measures of exposure were investigated. Exposure was assessed using information on work practices, questionnaire data designed to estimate exposure, and calculated dermal exposure based on fluorescent tracer skin swabs. METHODS

This study utilized a population of 57 regular applicators of organophosphates working in the apple orchard region of Washington's Yakima valley. Recruitment involved contact with growers to secure participation. The applicators of participating growers were then recruited for voluntary participation. The study population and a comparison cohort of slaughterhouse workers were enrolled in a larger study of chronic pesticide exposure and neurobehavioral effects (see companion article). Follow-up data were available for 48 applicators and 40 comparison workers, providing the basis for the study results presented here. Data on the use of protective equipment, work history, and work habits (particularly with respect to pesticide use), brief medical history, and demographic characteristics were available for these applicators from prespray season (Janu-

O R G A N O P H O S P H A T E P E S T I C I D E EXPOSURE

231

ary-February 1988) and postspray season (August-October 1988) questionnaire results. In addition, prespray season and postspray season 10-ml blood samples were drawn for each participant, collected in two EDTA tubes (5 ml each), placed in an ice-cooled chest and delivered to the University of Washington for both red blood cell and plasma cholinesterase activity analysis. The preseason (baseline) samples were collected prior to the first spraying of the season and at least 3 months after the last spraying of the previous year. Postseason samples were collected at least 3 weeks after the last pesticide application of the season. Thus, observed decreases in cholinesterase activity would reflect the cumulative inhibitory effect from moderate chronic pesticide exposure over the spray season or from some interval peak exposure. When utilizing a single baseline measurement for comparison, diagnosis of clinically significant inhibition requires red blood cell cholinesterase inhibition of at least 15 and/or a 20% decrease in plasma activity (Coye et al., 1986a,b). Each sample was analyzed within 24 hr of collection, using the method of Ellman et al. (1961). The prespray season baseline blood samples were also sent for analysis to a commercial laboratory where the same protocol for cholinesterase determination was employed; for analytical reasons, baseline measurements of the commercial laboratory are reported here. Postseason blood samples were analyzed only at the University of Washington. Because of interlaboratory differences in red blood cell cholinesterase determinations, changes in red blood cell and plasma cholinesterase are presented as changes adjusted to the comparison population. Adjustment involved subtracting the mean cross-season difference in cholinesterase within the control group from individual post season values. Seven study subjects and the orchards that employed them participated in an in-depth exposure assessment during a single workday of the 1988 summer spray season. These applicators were observed on the first day of spraying during the commencement of a particular pesticide cover (a complete cycle of pesticide application to a crop). This ensured that exposure measurements reflected the observed spray day and not previous spraying episodes. Time between pesticide applications is usually at least 3 weeks. Observations focused on the spray rig type and application rate, use of protective equipment during pesticide mixing, loading, and application, type and amount of pesticides applied, and total time spent spraying. All except one of the applicators sprayed 50 or 35% wettable powder azinphos-methyl (Guthion) alone or in combination with 78% liquid phosphamidon. The other applicator sprayed 50% wettable powder chlorpyrifos and 78% liquid phosphamidon. Azinphos-methyl, phosphamidon, and chlorpyrifos are insecticides used in control of the coddling moth, an important apple pest. The EPA has classified azinphos-methyl and phosphamidon as Class I pesticides, those with highest toxicity by oral and/or dermal LD50. Chlorpyrifos is rated as Class II, the group with next highest acute toxicity. Dermal exposure was assessed for the subset of participants utilizing a fluorescent tracer/skin scrub and swab technique developed in a 1987 pilot study. Improvements were developed and described by Gallagher (1989). In brief, a soluble fluorescent tag, sodium fluorescein dye, was added to the pesticide formulation at

232

KARR ET AL.

the time of dilution with water and mixing of the wettable powders. H o m o g e n e o u s mixing was assumed. U p o n spraying cessation, each applicator was asked to strip to his underwear and put on a pair of prewashed, nonfluorescent dark blue athletic shorts. The subjects were then given protective glasses and examined under long wavelength ultraviolet (black light) where visualization of fluorescent exposed areas was possible. All areas that showed fluorescence were estimated in size and recorded. Fluorescent areas were then swabbed with alcohol, saved in separate specimen containers, and brought to the laboratory for fluorimetric analysis. Blood samples were drawn in the morning before spraying had c o m m e n c e d and as soon as possible after spraying had ceased for that day, in a similar manner to the baseline samples.

RESULTS Pesticide applicator mean percentage adjusted change during the Spray season for red blood cell cholinesterase was - 3 . 9 % and for plasma cholinesterase was + 0.7%. The applicator group was further analyzed by three groups with increasing exposure: (1) an incidental exposure group of four applicators who act primarily as supervisors or managers and do not usually handle pesticides directly, (2) pesticide applicators who sprayed less than 10 days during the 1988 season, and (3) applicators who sprayed more than 10 days. Those who spray pesticides more regularly exhibited a dose-dependen; ~reater mean decrease in red bloc, d cell cholinesterase than those who report incidental or no exposure (Table 1). The relation between the use of different types of protective equipment and c h o l i n e s t e r a s e c h a n g e was investigated by dividing pesticide w o r k e r s w h o sprayed 10 days or more during the 1988 season into two subgroups representing two levels of protection: (1) those who used a rainsuit and respiratory protection TABLE 1 MEAN SEASONALCHANGEIN RED BLOODCELL(RBC) CHOLINESTERASELEVELSBY PESTICIDE EXPOSUREGROUP

Pesticide exposure group Slaughterhouse workers (n = 40) Pesticide workers (n = 48) Supervisors (n = 4) Applicators (n = 44) 10 Days spraying (n = 27) Enclosed cabs (n = 9)C Rainsuit with respiratory protection (n = 16)

Preseason RBC ChE mean (U/liter)

Postseason changea mean (SD)

Percentage changeb

7653 7970 8271 7943 8195 7785 7578

0 (782) -314 (957) 552 (777) -393 (940) -315 (754) -442 (1051) - 154 (1245)

Reference 3.9% +6.6% -4.9% -3.8% -5.7% -2.0%

7891

-725 (789)

-9.2%

a Postseason changes adjusted for interlaboratory variation. Change after adjustment for mean change in the control group (+ 1412). b Percentage change from preseason level after adjustment. c Two of the applicators who sprayed >10 days are not listed under enclosed cabs or rainsuit use. Information on protective equipment use was unavailable for these participants.

O R G A N O P H O S P H A T E P E S T I C I D E EXPOSURE

233

and (2) those who s p r a y e d f r o m within an enclosed cab type of rig. F o r red blood cell cholinesterase, applicators who sprayed from within an enclosed cab exhibited a lower m e a n change (Table 1). M e a n changes in plasma cholinesterase do not display the same pattern (Table 2). T h e s e v e n participants involved in the in-depth exposure a s s e s s m e n t spent an average of 22.5 days involved in spraying during the previous season (1987); the other applicators spent an average of 26.4 days. The extent of reported use of hats, gloves, r u b b e r boots, glasses or goggles, cartridge or full-face respirators, and p a p e r m a s k s was similar for both groups. All seven applicators o b s e r v e d during the spray operation used tractor-pulled orchard air blast equipment with the o p e r a t o r seated approximately 2 m in front of the spray rig nozzles. Application rates ranged f r o m 0.8 to 2.0 lb/acre. Spraying c o m m e n c e d in the early morning and continued as long as t e m p e r a t u r e or wind conditions permitted, b e t w e e n 2.5 and 6 hr. In addition to pulling the spray rig during the spray operation, loading and mixing of pesticides into the spray rig tanks were p e r f o r m e d by all of the applicators each time a tank was refilled. The o b s e r v e d use of protective equipment during the application and loading/ mixing in the tanks is presented in Table 3. Total estimated dermal exposure for six individuals ranged from 19 to 1235 txg (Table 4). Applicator 2, who was the only applicator o b s e r v e d not to w e a r gloves during both the application and mixing operations, had the highest estimated exposure. Ninety percent of his exposure was on his hands. The second highest estimated exposure, 340 jxg, was o b s e r v e d for applicator 1 who reported a mechanical problem with his spray rig. When he turned and looked o v e r his shoulder to investigate the problem, his left cheek was struck by b l o w b a c k of the pesticide being sprayed, resulting in 80% of his e x p o s u r e on his head. F o r all applicators, the majority of o b s e r v e d e x p o s u r e was to either the head or hands. TABLE 2 MEAN SEASONAL CHANGE IN PLASMA CHOEINESTERASE LEVELS BY PESTICIDE EXPOSURE GROUP

Pesticide exposure group Slaughterhouse workers (n = 40) Pesticide workers (n = 48) Supervisors (n = 4) Applicators (n = 44) 10 Days spraying (n = 27) Enclosed cabs (n = 9)c Rainsuit with respiratory protection (n = 16)

Preseason Plasma ChE mean (U/liter)

Postseason changea mean (SD)

Percentage changeb

2845 3045 3126 3038 3118 2988 2913

0 (288) 23 (447) 215 (260) 5 (458) 13 (508) 0 (434) -67 (332)

Reference +0.7% +6.9% +0.2% +0.4% -0.0% -2.3%

3077

39 (511)

- 1.3%

a Postseason changes adjusted for interlaboratory variation. Change after adjustment for mean change in the control group (+ 99). b Percentage change from preseason level after adjustment. c Two of the applicators who sprayed >10 days are not listed under enclosed cab or rainsuit use. Information on protective equipment was unavailable for these participants.

234

K A R R E T AL,

TABLE 3 PERSONAL PROTECTIVE EQUIPMENTa USED BY DERMAL EXPOSURE ASSESSMENT PARTICIPANTS Subj. No.

H e a d protection

H a n d protection

1

H a r d hat

L a t e x gloves

2

Rubberized hood b

None

3

Rubberized hood

L e a t h e r gloves b

4

Helmet b

Butyl gloves b

5

Cotton cap

L a t e x gloves

6

Cotton cap

L a t e x gloves

7

Cotton cap

Latex gloves

Respiratory protection Half face respirator with dual organic cartridges Full face respirator with dual organic cartridges Half face respirator with dual organic cartridge b Powered air purifying respirator with helmet Half face respirator with dual organic cartridge Half face respirator with dual organic cartridge Half face respirator with dual organic cartridge

a All applicators wore a rubberized rain j a c k e t and pants. b R e m o v e d during mixing/loading.

Red cell and plasma cholinesterase determinations were made for samples collected in the morning before spraying commenced and at the end of spraying for that day for each of the seven applicators involved in the in-depth exposure assessment. None of the applicators exhibited a clinically significant (> 15%) decline in red cell or plasma cholinesterase at the end of the spray day observation. However, for the six applicators who also had quantitative dermal exposure data, measured changes in red blood cell cholinesterase were inversely proportional to the estimated amount of dermal exposure (r -= -0.84) (Table 5). Red blood cell cholinesterase change was also correlated with other exposure measures including time spent spraying and kilograms of active ingredient sprayed (r = - 0 . 8 2 and r = - 0.78, respectively). A plot of dermal recovery as log normal values yields a strong correlation, r = - 0 . 9 5 , with red cell cholinesterase change. In contrast,

TABLE 4 DISTRIBUTION OF ESTIMATED DERMAL EXPOSURE AND CHOLINESTERASE (ChE) ACTIVITY CHANGE a Subj. No. 1 2 3 4 6 7

Distribution of e x p o s u r e (%) Head 80 9 100 78 91

Neck

Hands

17 1

2 90 78 22 9

Change in C h E (%) Other 1 22

Total (txg)

Plasma

RBC

340 1235 54 24 34 19

+ 0.2 +6.0 + 1.8 + 0.4 - 1.8 +4.0

- 6.0 - 10.0 - 3.5 0.4 - 2.0 +4.0

a Subject 5 is omitted because no usable analytic data on dermal exposure were obtained.

235

O R G A N O P H O S P H A T E PESTICIDE EXPOSURE

TABLE 5 CORRELATION BETWEEN CHOLINESTERASE CHANGE AND ONE DAY EXPOSURE MEASUREMENTS FOR DERMAL EXPOSURE ASSESSMENT PARTICIPANTS

Exposure measurement Calculated dermal exposure (n = 6) Log normal calculated dermal exposure (n = 6) Hours spent spraying (n = 7) Kg. Active ingredient applied (n = 7)

RBC cholinesterase r (P-value) -0.84 - 0.95 -0.82 -0.78

(.018) (.002) (.012) (.019)

Plasma cholinesterase r (P-value) +0.67 + 0.45 +0.39 +0.73

(.071) (. 184) (.184) (.030)

there was no correlation between plasma cholinesterase change and the exposure classifications. DISCUSSION Visualization and dermal recovery of fluorescent tracer indicated that pesticide exposure primarily involved the head and hand regions. Other studies of agricultural work activities have shown that a large percentage of dermal exposure is attributable to hand exposure. For example, in one study, hand exposure was 52 to 99% of the total dermal exposure in mixer-loaders and 41-64% of total exposure in applicators (Franklin et al., 1985). Biological monitoring of cholinesterase is generally regarded as a useful tool for surveillance for clinically significant exposures to organophosphate insecticides. Its value in monitoring lower levels of exposure which may still nonetheless be clinically important is unclear. We found subclinical changes in red cell cholinesterase in applicators correlated well with estimated dermal exposure. The utility of red blood cell cholinesterase as an indicator of acute, low exposure may be underestimated. The utility of cholinesterase monitoring for chronic, low-level exposure to pesticides is uncertain. The body may be able to compensate for progressively increasing levels of acetylcholine (Gallo and Lawryn, 1991). Plasma cholinesterase is rapidly replaced, whereas red cell depression may persist; the latter is more likely the better indicator of chronic effect (Gallo and Lawryn, 1991). In this study, comparisons of seasonal red blood cell cholinesterase changes in pesticide workers according to exposure level, characterized by frequency of pesticide spraying and protective equipment used, all suggest that the cholinesterase levels of higher exposed groups were slightly more depressed than those of lesser exposed groups. This study has some other important limitations which should be borne in mind when reviewing its results. The small numbers of workers studied is especially important given intraindividual variations in cholinesterase activity. The magnitude of the across-season change in RBC cholinesterase was small and not statistically significant, but the internal analyses by levels of exposure demonstrated effects in the expected direction. The measures used in this study to indicate exposure, including days of pesticide use, type of protective equipment used, hours of spraying, and kilograms of active ingredient sprayed, are only imprecise indicators of the true exposure to

236

KARR ET AL.

organophosphate pesticides. Even the dermal recovery technique, which yields seemingly precise numbers, is an experimental technique which has not yet been fully validated and should be considered only semiquantitative (Gallager, 1989). Nonetheless, misclassification resulting from the use of these rough surrogates of exposure should, on average, serve to minimize any true relationships present. Despite this, our study found a relatively consistent relationship between all measures of exposure and RBC cholinesterase activity. Although the investment required for an in-depth exposure assessment is relatively high, the ability to extrapolate from a small to a larger group of workers could further increase the potential yield of such an effort. The validity of such extrapolation is supported by our finding similar self-reports of protective equipment use and pesticide work habits between the subgroup of applicators involved in the in-depth exposure assessment and the larger group of forty-three pesticide workers. In addition, valuable recommendations for improvements in work practices to protect applicator health may result. For example, in this study, protective equipment removal from the head and hands was observed during tank filling when the concentrated form of the pesticide is handled. The use of fluorescent tracer, allowing worker visualization of personal exposure, is an excellent and impressive tool for worker education. The generalizability of these findings may be limited, however. The orchards and their applicators were voluntary participant populations; the validity of extrapolation to other nonvoluntary participant populations is not known and it is quite likely that these subjects represent a group with better than average health and safety practices. For example, 100% of the subjects observed in this study were equipped with rubber rainsuits and respiratory protection for mixing and applying. This is in striking contrast to a Washington State farm worker survey which found 52% of 59 farmworkers who reported mixing or applying pesticides never received any protective clothing or equipment (Mentzer and Villalba, 1988). In conclusion, cholinesterase monitoring may be a useful biological marker for even subclinical organophosphate pesticide effects. This monitoring will have particular importance if further studies confirm the suspicion that chronic effects may arise from repeated, low-level occupational exposures to organophosphate insecticides. REFERENCES Atallah, Y. H., et al. (1982). Exposure of pesticide applicators and support personnel to O-ethyl O-(4-nitrophenyl)phenylphosphonothionate(EPN). Arch. Environ. Contam. Toxicol. 11, 219-225. Coye, M. J., et al. (1986a). Biological monitoring of agricultural workers exposed to pesticides. I. Cholinesterase activity determinations. J. Occup. Med. 28(8), 619-627. Coye, M. J., et al. (1986b). Biological monitoring of agricultural workers exposed to pesticides. II. Monitoring of intact pesticides and their metabolites. J. Occup. Med. 28(8), 631-636. DanieU, W. E., Barnhart, S., Demers, P., Costa, L. G., Eaton, D. L., Miller, M., and Rosenstock, L. (1992). Neuropsychological performance among agricultural pesticide applicators. Environ. Res. 59, 217-228. Ellman, G. L , et al. (1961). A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem. Pharmacol. 7, 88-95. Fenske, R. A. (1986). A video-imaging technique for assessing dermal exposure. II. Fluorescent tracer testing. A m . Ind. Hyg. Assoc. J. 47(12), 771-775.

ORGANOPHOSPHATE PESTICIDE EXPOSURE

237

Ferguson, W. L. (1985). "Pesticide Use on Selected Crops: Aggregated Data, 1977-80," Ag. Info. Bull. No. 494, pp. 2-10. Economic Research Service, USDA. Franklin, C. A., et al. (1981). Correlation of urinary pesticide metabolite excretion with estimated dermal contact in the course of occupational exposure to guthion. J. Toxicol. Environ. Health 7, 715-731. Franklin, C. A., et al. (1985). "Occupational Exposure to Pesticides and Its Role in Risk Assessment Procedures Used in Canada," ACS Symposium Series No. 273, pp. 429--444. Am. Chem. Soc., Washington, DC. Gallagher, E. (1989). Masters thesis, University of Washington Department of EnvironmentalHealth. Gallo, M. A., and Lawryn, N. J. (1991). Organic phosphorous pesticides. In "Handbook of Pesticide Toxicology" (W. J. Hayes and E. R. Laws, Eds.), pp. 926-946. Academic Press, San Diego. Knaak, J. B., et al. (1979). Alkyl phosphate metabolite levels in the urine of field workers giving blood for cholinesterase test in California. Bull. Environ. Contam. Toxicol. 21, 375-380. Mentzer, M., and Villalba, B. (1988). "Pesticide Exposure and Health: A Study of Washington Farmworkers." Evergreen Legal Services, Granger, WA. Murphy, S. D. (1986). Toxic effects of pesticides. In "Toxicology" (C. Klaassen et al., Eds.), pp. 527. MacMillan, New York. Niss, H. N., et al. (1983). Exposure of spray applicators and mixer-loaders to chlorbenzilate miticide in Florida citrus growers. Arch. Environ. Contam. Toxicol. 12, 477-482. Rosenstock, L. (1987). Clinical management of pesticide poisoning. In "Toxicology of Pesticides: Experimental, Clinical, and Regulatory Perspectives" (L. G. Costa et al., Eds.), pp. 197-206. Springer-Verlag, Berlin/Heidelberg/New York. Rosentock, L., et al. (1990). Chronic neuropsychological sequelae of occupational exposure to organophosphate insecticides° A m . J. Ind. Med. 18, 321-325. Washington state Employment Security Department (1991). "Agriculture, Forestry, and Fishing Employment in Washington State," July. Weisskopf, C., et al. (1988). Personnel exposure to diazinon in a supervised pest eradication program. Arch. Environ. Contam. Toxicol. 17, 201-212.