Pesticide Illness, Farm Practices, and Neurological Symptoms among Farm Residents in Colorado

Environmental Research Section A 90, 89}97 (2002) doi:10.1006/enrs.2002.4398

Pesticide Illness, Farm Practices, and Neurological Symptoms among Farm Residents in Colorado Lorann Stallones* and Cheryl Beseler*Department of Psychology and -Department of Environmental Health, Colorado State University, Fort Collins, Colorado 80523-1876 Received October 30, 2001

INTRODUCTION The purpose of this study is to evaluate the association between pesticides and neurological symptoms among a population exposed to organophosphate chemicals as a result of agricultural use. Chronic sequelae of acute pesticide poisoning from organophosphate compounds include a variety of neurological symptoms including restlessness, irritability, and trouble sleeping. Individuals who have had an acute pesticide poisoning have been reported to suffer a wide range of neurological symptoms that occur from weeks to months after the initial episode. Data for this study came from a cross-sectional survey of farmers and their spouses conducted in an eight-county area in northeastern Colorado. Neurological characteristics were assessed to determine their relationship with previously reported pesticide-related illnesses. Symptoms that were signiAcantly associated with a previous poisoning were difAculty concentrating [OR 2.07, 95% conAdence interval (CI) 1.22, 3.50]; relatives noticing person had trouble remembering things (OR 2.54, 95% CI 1.47, 4.39); making notes to remember things (OR 2.18, 95% CI 1.20, 3.97); Anding it hard to understand the meaning of newspapers, magazines, and books (OR 1.90, 95% CI 1.01, 3.60); felt irritable (OR 1.84, 95% CI 1.08, 3.12); felt depressed (OR 2.82, 95% CI 1.65, 4.81); had heart palpitations without exertion (OR 2.83, 95% CI 1.22, 6.54); sleeping more than usual (OR 3.58, 95% CI 1.95, 6.58); difAculty moving Angers or grasping things (OR 2.08, 95% CI 1.06, 3.24); and headaches at least once a week (OR 1.85, 95% CI 1.06, 3.24). Stepwise regression was used to identify the best explanatory model of pesticide-related illness. Variables that were associated with increased odds of illness were being male, being depressed, sleeping too much, and using crop organophosphates.  2002 Elsevier Science (USA) Key Words: neurological symptoms; pesticides.

Organophosphates are highly toxic compounds readily absorbed through the skin, mucous membranes, gastrointestinal tract, and respiratory tract. The toxic effects are caused by an inhibition of cholinesterase, an enzyme that inactivates the neurotransmitter acetylcholine. Acetylcholine then accumulates at the cholinergic synapses, resulting in peripheral and central nervous system effects (1, 2). Red blood cell and plasma levels of cholinesterase are indicators of exposure to organophosphates. In a study of Minnesota farmers, researchers found that 43% of fumigators experienced a 20% drop in red blood cell cholinesterase activity and 36% experienced a 20% or greater drop in plasma cholineseterase activity (3). Impairments resulting from pesticide poisoning include both cognitive and physical sequelae. Acute effects include nausea, vomiting, abdominal pain, numbness and/or tremors in the extremities, fatigue, headaches, excessive salivation, diarrhea, generalized weakness, respiratory problems, and blurred vision (1, 2). Acute psychological effects include anxiety, depression, irritability and restlessness (1, 2). Additionally, neuropsychological effects, such as dif7culty concentrating, problems 7nding words, impairments of memory, and decreased alertness, may result following an acute exposure (1, 2). Exposure to organophosphates may have chronic, long-term effects and has been linked to delayedonset peripheral neuropathies, primarily affecting the extremities (4), neuropsychological changes (5, 6), and neurobehavioral changes (4, 6, 7). The delayed reaction may be the result of the permanent inhibition of several different serine hydrolases (8). Serine hydrolases constitute a large class of enzymes which hydrolyze ester, peptide and amide bonds. This group includes a neuronal membrane protein, neuropathy target esterase (NTE), the 89 0013-9351/02 $35.00  2002 Elsevier Science (USA) All rights reserved.

90

STALLONES AND BESELER

physiological substrate of which is unknown (9). NTE undergoes reversible inhibition, as does cholinesterase, but differs in that molecular rearrangement results in covalent modi7cation of an active site serine residue creating permanent inhibition of the enzyme (10). The irreversible inhibition may not be a direct cause of the neuropathy, but a gain of function by the modi7ed protein may result in toxicity (10). In a 1992 study of migrant farmworkers, 21 workers who had experienced two documented acute exposures were evaluated with a neuropsychological battery, medical history questionnaire, and an anxiety and depression scale (6). The investigators found the exposed group to be signi7cantly more impaired on measures of motor speed and coordination, visuospatial memory, anxiety, depression, and physical symptoms. Savage et al. (5) also found individuals with a history of organophosphate poisoning to perform signi7cantly worse on tests of intellectual functioning, academic skills, 8exibility of thinking, abstraction, and motor speed and coordination. Rosenstock et al. (7) studied a group of 36 agricultural workers with a history of hospitalization for acute organophosphate poisoning and found that their performance on the neurological and neuropsychological batteries was signi7cantly inferior to that of controls. The workers were evaluated an average of 2 years postpoisoning, and exposure during the 3 months prior to testing was uncommon. The poisoned group had poorer performance scores on 7ve of six subtests evaluating verbal attention, visual memory, and visuomotor and motor functions. In addition, the exposed group performed signi7cantly worse on neuropsychological tests of visuomotor sequencing, problem solving, and visual attention (7). Steenland et al. (4) reported the results of an evaluation of 128 individuals with suspected and con7rmed organophosphate poisoning in California. These individuals differed signi7cantly from controls on tests of sustained visual attention and two mood scale tests. Those individuals with con7rmed poisoning also differed signi7cantly from controls on a test of vibrotactile sensitivity and the symbol digit test. The purpose of this paper is to assess the relationship between self-reported pesticide poisoning, farming activities, and neurological symptoms in a large-population-based study of farm residents in northeastern Colorado. METHODS

Population The study population was selected using a multistage area sample of farms in an eight-county area of

TABLE 1 Response Rates by Primary Sampling Unit (Farm) by County, Colorado Eight-County Survey, 1992}1997

County Sedgwick Phillips Yuma Logan Washington Morgan Weld Larimer Total

No. of farms contacted

No. of farms enrolled

29 57 180 297 187 137 801 189

35 77 121 126 138 91 202 45

21 36 77 65 76 46 122 36

60 47 64 52 55 51 60 80

1877

835

479

57

No. of ineligible residencesa

% participation

a

Rural homes, not operating farms.

northeastern Colorado. The eight-county area included 47% of the total agriculturally employed population in Colorado. Colorado has been mapped using township/range units (TRUs), which consist of 36 sections of land area, each covering 23,040 acres or 36 square miles of land. The United States Public Land Survey (USPLS) began in 1785 and covers all land that was not settled by the time of the of7cial government survey, including all of the state of Colorado (11). The number of TRUs sampled from each of the eight counties was selected based on the average acreage of the farms in the county, the probability of a farm operator being a resident on the farm, and the probability of the land being in agricultural use. TRUs were randomly selected from within the study counties. Sections within the TRUs were than randomly selected and farm dwellings were identi7ed within each of the sampled sections. Sections consist of 640 acres of land, or 1 square mile. Farms were located within the sampled sections using rural directories in six counties and property value assessment 7les in two counties where rural directories were unavailable. Farm residents, operators and their spouses, were recruited for the study between 1992 and 1997. A total of 761 individuals representing 479 farms were enrolled in the study. Farm contacts were made directly by hired, trained interviewers from the counties where the study was conducted. The initial contact was made either by phone or by letter. Table 1 contains the response rate by farms. The primary sampling unit was the farm; if the farm operator was willing to participate, the farm was included in the survey. Not all spouses participated and, in a few cases, two principal operators for one farm were included when they

PESTICIDES AND NEUROLOGICAL SYMPTOMS

self-identi7ed as such. No further follow-up was made with those who refused to participate; however, comparing farming operations in the eight counties with those who participated in the survey indicated there were no differences in the type of farming operation between refusals and participants. Interview Enrolled participants were asked to complete a personal interview that took from 45 min to 2 h depending on the complexity of individuals' personal histories. Interviews were conducted during the slower seasons for the type of agriculture in the area, between October and early March during the study period, 1992}1997. Interviews were conducted in the homes of participants. Written documentation of informed consent was obtained from each of the study participants. The study was conducted with approval of the procedures from the Human Research Committee at Colorado State University. The gender of the interviewer was not matched to the gender of the respondents. All interviews were conducted in English, there were no known Spanish-speakingonly farmers in the eight-county area where the survey was conducted. Interviewers had a paper questionnaire for each participant. Separate questionnaires were used for the primary operator and the spouse of the operator. Farm characteristics (e.g., acres in production, primary agricultural activities, type of pesticides used on the farm) were asked of the primary operator, but not the spouse. Interviewers conducted separate interviews for the spouse and the operator. There were no sensitive questions on the questionnaire, therefore no attempt was made to conduct interviews in the absence of other family members, the choice was left to study participants. No exclusion criteria were used, as the farm operator had to be actively involved in farming and therefore was assumed to be capable of responding to the questions. Similarly, the spouses who participated were healthy and capable of understanding the questions asked. Data were collected on the individual characteristics of the respondents and on the farms where the respondents worked. The number of families residing on the farm, the crop value, the types of crops grown, the number of acres of cropland in use, the number of workers on the farm, and how many hours the farm operator worked off the farm were recorded during the interview. Information concerning pesticides used on the farm and how they were applied was obtained from the respondents. Respondents were asked what protective equipment was used when preparing and

91

applying pesticides. Possible responses included gloves, coveralls, dust masks, respirator, rubber boots, and closed-cab tractors. The farms represented in this study grew corn, sorghum, wheat, barley, oats, soybeans, rye, millet, dry edible beans except limas, sun8ower seeds, potatoes, sugar beets, hay, alfalfa hay and livestock. General health questions included a history of coronary heart disease, hypertension, respiratory symptoms and disease, allergies, stroke, diabetes, cirrhosis, multiple sclerosis, Parkinson’s disease, and cancer. Respondents were asked whether they had ever become ill from any exposure to pesticides, whether they had sought medical treatment due to the illness, and whether a physician had diagnosed pesticide poisoning as a result of their illness. Respondents were asked how often they had experienced a neurological symptom in the past month. Twenty-four neurological symptoms were assessed because they have been associated with exposure to organic solvents, heavy metals and pesticides (12). The exposure of primary interest in this study was the use of agricultural pesticides. The occurrence of symptoms was based on the following scale: (1) not at all, (2) a little, (3) moderately, (4) quite a bit, and (5) extremely. The neurological symptoms included in the questionnaire are shown in Table 3. Odds ratios with 95% con7dence intervals were estimated from a univariate analysis for reporting a neurological symptom and experiencing pesticide-related illness. A Mantel}Haenszel
3 test for trend was calculated for those neurological symptoms that showed an elevated odds ratio. Pesticides were listed in the questionnaire by trade name and coded as dichotomous variables based on their active ingredient. Types of crop pesticides used on the farms included terbufos, chlorpyrifos, carbamate, and other organophosphates. Livestock pesticides included chlorpyrifos, phosmet, dichlorvos, carbamate, and other organophosphate. Herbicides included atrazine, alachlor, 2,4-D, carbamate, and other herbicides. The percentage of farm residents using these groups of pesticides was determined and a univariate analysis was performed to obtain estimated odds ratios for using the pesticide and reporting a pesticide-related illness. Questions concerning methods of pesticide application asked included whether the pesticide had been aerially applied, incorporated as granules into the soil, applied by tractor or pulled sprayer, added to the water system in a center-pivot irrigation system, or utilized by some other method. Respondents were asked whether they personally mixed, loaded

92

STALLONES AND BESELER

TABLE 2

RESULTS

Characteristics and Comparison of Male, Female, and Combined Farm Residents in the Study Population, Colorado, 1992}1997

Characteristic

Males % (n) Females % (n) (n"460) (n"301)

Age in years (30 2.6 (12) 30}39 19.2 (88) 40}49 29.6 (136) 50}59 20.5 (94) 59 28.1 (129) Married Yes 86.9 (399) No 13.1 (60) Race White 98.3 (451) Black 0.7 (3) Asian, Paci7c 0.4 (2) Islander Perceived general health Excellent 27.3 (125) Very good 42.6 (195) Good 23.8 (109) Fair/poor 6.3 (29) High school graduate Yes 85.1 (314) No 14.9 (55) Illness due to pesticide exposure Yes 12.3 (56) No 87.7 (398) Substantial income decline Yes 26.7 (121) No 73.3 (332) Current smoker Yes 10.7 (49) No 89.3 (410) Involved in farm work Yes 98.8 (458) No 0.2 (1)

Population % (n) (n"761)

5.0 (15) 19.6 (59) 26.9 (81) 22.3 (67) 26.3 (79)

3.6 (27) 19.3 (147) 28.7 (218) 21.2 (161) 27.3 (208)

96.4 (290) 3.7 (11)

90.7 (690) 9.3 (71)

98.7 (297) 0.3 (1) 0.0 (0)

98.4 (749) 0.5 (4) 0.3 (2)

23.7 (71) 43.7 (131) 27.7 (83) 5.0 (15)

25.8 (196) 43.0 (327) 25.4 (193) 5.8 (44)

94.0 (282) 6.0 (18)

90.4 (687) 9.6 (73)

4.4 (13) 95.6 (284)

9.2 (69) 90.8 (684)

23.7 (70) 76.4 (226)

25.6 (192) 74.4 (558)

10.0 (30) 90.0 (271)

13.8 (79) 86.2 (495)

83.4 (251) 16.6 (50)

93.3 (710) 6.7 (51)

or applied the pesticide. These responses were coded as ‘‘yes’’ or ‘‘no’’ in the analysis. Odds ratios and 95% con7dence intervals for pesticide-related illnesses and method of application of pesticides were calculated in a univariate analysis. Variables found signi7cant in the univariate analyses were included in a multivariate logistic regression analysis modeling the probability of having reported a pesticide-related illness. Stepwise logistic regression was used to verify the best model. SAS software version 8.2 (SAS Institute, Inc., Cary, NC) was used for the statistical analysis. Variables found signi7cant in univariate analysis were included in a multivariate conditional regression analysis.

Characteristics of the respondents in the study population are shown in Table 2. The mean age of the study population was 50.7 years ranging from 24 to 85, with a standard deviation of 13.6. Descriptions of the farms where 761 respondents lived and/or worked are shown in Table 3. Nearly 64% of the respondents lived on single-family farms and greater than 96% of the farms in this study population had operators who resided on the farm. Nearly 57% did not have paid employment outside the farm. Eighty-one farms (10.7%) did not have cropland in use at the time of the survey. Table 4 contains a description of the pesticide usage on the study farms.

TABLE 3 Selected Farm Characteristics of 761 Farm Residents in the Study Population, Colorado, 1992}1997 Characteristic of farm Number of families Single family Two families Three families Four or more families Crop value $0}40,000 $40,000}99,000 Greater than $100,000 Acres of cropland in use None 1}200 210}600 601}1400 Greater than 1400 Number of assistants working on farm last 12 months 0 1}5 6}10 Greater than 10 Number of assistants who were seasonal/migrant workers 0 1}5 6}10 Greater than 10 Number of farms with operators who reside on the farm Number of days operators/respondents work off the farm None 1}49 50}99 100}149 150}199 Greater than 200 days

%

63.8 17.6 12.2 6.4

(n)

(486) (134) (93) (49)

26.8 (201) 25.8 (193) 47.4 (355) 10.7 22.2 28.6 24.7 13.8

(81) (169) (217) (188) (105)

7.1 60.0 25.1 7.8

(54) (456) (191) (59)

18.5 66.5 11.0 4.0

(131) (471) (78) (28)

96.5 (735)

56.8 9.9 5.3 4.6 5.3 17.4

(432) (75) (40) (35) (40) (132)

93

PESTICIDES AND NEUROLOGICAL SYMPTOMS

TABLE 4 Numbers and Percentages of Pesticides Used by Residents in the Study Population (nⴝ759) During a 12}Month Period, and Use of Personal Protective Equipment during Application, Colorado, 1992}1997

Type of pesticide

% (n) using this pesticide

Speci7c types of pesticides mixed and applied by farm residents

Number of pesticide applications in 12 months mean (SD)

Herbicide Yes Personally mixed and applied herbicide No protective equipment useda No

74.8 (568) 44.8 (340)

Atrazine Alachlor 2,4}D

17.1 (129) 7.2 (54) 37.1 (279)

1.3 (1.2) 1.2 (0.6) 1.5 (1.3)

20.8 (155) 7.5 (56) 11.4 (85)

1.2 (1.3) 1.3 (1.3) 1.6 (3.5)

0.9 (7) 4.4 (33) 37.1 (276)

6.3 (8.1) 5.6 (9.3) 2.6 (5.2)

5.9 (28) 25.2 (191) Crop insecticide

Yes Personally mixed and applied pesticide No protective equipment useda No

52.2 (396) 29.5 (224)

Terbufos Chlorpyrifos Other OP/carbamate

3.8 (10) 47.8 (363) Livestock insecticide

Yes Personally mixed and applied pesticidea No

51.3 (389) 48.9 (342)

Phosmet Dichlorvos Other OP/carbamate

48.7 (370)

Note. Data include means and standard deviations for the number of applications of pesticide during the 12-month period. a Study participants who reported not using any personal protective equipment when applying at least one herbicide or crop insecticide. All subjects using at least one livestock insecticide reported using personal protective equipment.

Sixty-nine respondents in the study population reported a pesticide-related illness. A total of 761 individuals responded to the neurological questions. Only one individual reported having a seizure, so this symptom was not included in the analysis. Conditional logistic regression analysis, controlling for age and gender, showed increased odds for reporting reoccurring neurological symptoms in those who had experienced a pesticide-related illness (Table 5). Symptoms that were signi7cantly associated with a previous poisoning were dif7culty concentrating; relatives noticing person had trouble remembering things; making notes to remember things; 7nding it hard to understand the meaning of newspapers, magazines, and books; feeling irritable; having heart palpitations without exertion; sleeping more than usual; dif7culty moving 7ngers or grasping things; and headaches at least once a week. Questions concerning dif7culty concentrating; having relatives notice memory lapses; making notes to remember things; feeling depressed, having heart palpitations; sleeping more than is usual; and having dif7culty moving 7ngers or grasping things showed odds ratios greater than 2 with con7dence intervals that did not include 1. A
2 test for trend showed that having trouble remembering things was highly signi7cant

(
2 22.40, P(0.0001) (Table 6). Sleeping more than usual and feeling depressed were also highly associated with having had a pesticide illness (Table 6). The odds of an acute pesticide-related illness were increased in those applying a crop organophosphate (OP), but not in those applying a livestock OP or herbicides. Of the 266 respondents who applied a livestock OP, only 2 reported a pesticide illness. Due to the signi7cant association of crop OPs with pesticide illness, only crops were examined for associations with pesticide exposure. Analyzing whether a respondent applied a crop OP showed an increased odd of 1.76 for having a pesticide-related illness (Table 7). Those who applied an OP, excluding the 61 individuals who used the chlorpyrifos Lorsban, had an odds ratio of 2.45. A separate analysis of the 159 individuals who applied Counter showed an odds ratio of 1.96 for reporting a pesticide-related illness. Examining the type of organophosphate applied showed that a terbufos, Counter, and other organophosphates, including Cygon, Diazinon, DiSyston, Dyfonate, Malathion, Penncap, Pounce and Thimet, were signi7cantly associated with reporting an acute pesticide illness. Of the 14 individuals who applied Thimet, 7 reported a pesticide-related illness (Fisher’s exact P-value of 0.00013).

94

STALLONES AND BESELER

TABLE 5

TABLE 6

Neurological Symptoms among Colorado Farm Residents, 1992}1997

Mantel}Haenszel
2 test for trend, with P Values, for Neurological Symptoms SigniAcantly Associated with Acute Pesticide}Related Illness, Colorado Farm Residents, 1992}1997

Neurological symptom

Odds ratio

95% CI


2 for trend P value

Neurological symptom Tired more easily than expected for the activity you do? Felt light headed or dizzy? Had diculty concentrating? Been confused or disoriented? Had trouble remembering things? Have your relatives noticed that you have trouble remembering things? Had to make notes to remember things? Found it hard to understand the meaning of newspapers, magazines and books that you have read? Felt irritable? Felt depressed? Had heart palpitations even when not exerting yourself? Been sleeping more than is usual for you? Had diculty falling asleep? Been bothered by lack of coordination or loss of balance? Had any loss of muscle strength in your legs or feet? Had any loss of muscle strength in your arms or hands? Had diculty moving your 7ngers or grasping things? Had numbness or tingling in your 7ngers lasting more than a day? Had numbness or tingling in your toes lasting more than a day? Had headaches at least once a week? Had diculty driving because you felt dizzy or tired even though you'd slept enough? Have you had a lower tolerance for alcohol (takes less to get drunk)?

1.26 1.23 2.07 2.09 1.34 2.54

0.75}2.12 0.55}2.73 1.22}3.50 0.91}4.82 0.80}2.25 1.47}4.39

2.18

1.20}3.97

1.90 1.84 2.82

1.01}3.60 1.08}3.12 1.65}4.81

2.83

1.22}6.54

3.58 1.45

1.95}6.58 0.84}2.53

2.34

0.97}5.65

0.98

0.40}2.43

1.36

0.67}2.74

2.08

1.06}4.09

1.16

0.34}3.99

2.61 1.85

1.00}6.78 1.06}3.24

2.18

0.79}6.05

2.41

0.86}6.79

Note. The table shows adjusted odds ratios and 95% con7dence intervals estimated from conditional logistic regression analyses for neurological symptoms in 69 cases of acute pesticide-related illness and 692 farm residents not reporting a pesticide}related illness. The analyses were adjusted for age and gender.

Pesticide application practices were examined to ascertain whether any particular methods increased the odds of having a pesticide-related illness (Table 8). Pesticide application involving a tractormounted or pulled sprayer and the incorporation of granules into the soil showed an increased odds of having a pesticide-related illness, with con7dence intervals not including 1. These odds ratios were elevated over those of mixing and applying pesticides, which also showed an elevated odds ratio for

Had dif7culty concentrating? Been confused or disoriented? Had trouble remembering things? Have your relatives noticed that you have trouble remembering things? Had to make notes to remember things? Found it hard to understand the meaning of newspapers, magazines and books that you have read? Felt irritable? Felt depressed? Been sleeping more than is usual for you? Had any loss of muscle strength in your legs or feet? Had dif7culty moving your 7ngers or grasping things? Had headaches at least once a week? Had dif7culty driving because you felt dizzy or tired even though you'd slept enough? Have you felt high from chemicals you use at work? Have you had a lower tolerance for alcohol (takes less to get drunk)?

9.20 6.10 9.27

0.0024 0.0135 0.0023

22.40 10.54

(0.0001 0.0012

6.35 4.89 14.77 14.42

0.0117 0.0270 0.0001 0.0001

4.37

0.0365

11.40 5.83

0.0007 0.0157

4.36

0.0368

5.04

0.0248

0.50

0.4787

a pesticide-related illness. Counter was the primary terbufos applied to crops and was signi7cantly associated with application of granules to the soil (
2 568.73, P(0.0001). Counter, soil granules, and tractor application were highly correlated. Of the 159 individuals who applied Counter, 147 did so by applying granules to soil. Most of the respondents applying an OP loaded and applied the pesticides themselves, but did not mix the pesticide. A logistic regression analysis showed gender, being depressed, sleeping too much, and using an TABLE 7 Odds Ratios and 95% ConAdence Intervals for Type of Pesticide Used and Having an Acute Pesticide-Related Illness in Colorado Farm Residents, 1992}1997

Type of pesticide Total crop organophosphates Livestock organophosphates Crop terbufos (Counter) Crop chlorpyrifos Other crop organophosphate

% (n)

Odds ratio

95% CI

30.6 (232) 35.1 (266) 20.8 (159) 8.0 (61) 10.5 (80)

1.76 0.99 1.96 0.90 2.45

1.06}2.92 0.59}1.66 1.14}3.36 0.35}2.34 1.29}4.64

95

PESTICIDES AND NEUROLOGICAL SYMPTOMS

TABLE 8

TABLE 10

Odds Ratios for Certain Characteristics of Pesticide Application and Having a Pesticide-Related Illness in Colorado Farm Residents, 1992}1997

Odds Ratios and 95% ConAdence Intervals Estimated from a Logistic Regression Analysis Modeling the Probability of Having Had a Pesticide-Related Illness Using Characteristics of Pesticide Use and Reported Neurological Symptoms in Colorado Farm Residents, 1992}1997

Pesticide characteristic

Odds ratio

% (n)

Aerially applied No 93.2 (712) 1.00 Yes 6.8 (52) 1.08 Granules incorporated into soil No 76.6 (585) 1.00 Yes 23.4 (179) 2.04 Tractor-mounted or pulled sprayer No 95.8 (732) 1.00 Yes 4.9 (32) 2.98 Backpack or hand sprayer No 99.5 (760) 1.00 Yes 0.5 (4) 3.34 Added to water in center-pivot irrigation system No 195.9 (733) 1.00 Yes 4.1 (31) 2.06 Mix pesticide No 96.5 (737) 1.00 Yes 3.5 (27) 0.39 Load pesticide No 80.4 (614) 1.00 Yes 19.6 (150) 1.56 Apply pesticide No 76.6 (585) 1.00 Yes 23.4 (179) 1.63

95% CI

0.42}2.82

1.21}3.44

1.24}7.16

0.34}32.55

0.76}5.56

0.06}2.73

1.16}2.08

1.19}2.22

organophosphate pesticide, other than chlorpyrifos, to be signi7cant factors in modeling the probability of reporting a pesticide illness (Table 9). Feeling depressed and sleeping more than is usual were important neurological symptoms in the model showing a twofold odds ratio (Table 10). Using an organophosphate rather than another pesticide was signi7cantly associated with having had a pesticiderelated illness. Covariates that were assessed that were not found to be signi7cantly associated with pesticide-illness included years of schooling, alcohol consumption, farm income, and age (data not shown). TABLE 9 Conditional Logistic Regression Analysis of Having Had a Pesticide-Related Illness and Characteristics of Pesticide Use and Neurological Symptoms in Colorado Farm Residents, 1992}1997

Characteristic

Regression coecient

Standard error


2

P value

Gender Depressed Sleep too much Use OP

!1.29 0.84 1.11 0.50

0.33 0.29 0.33 0.28

15.01 8.59 11.45 3.30

0.0001 0.0025 0.0032 0.0137

Characteristic

Odds ratio

95% CI

Gender Depressed Sleep too much Use OP

0.27 2.39 3.09 2.34

0.14}0.52 1.36}4.20 1.62}5.89 1.17}4.66

DISCUSSION

In support of previous reports (Table 11), pesticide illnesses were associated with neurological symptoms in this population of agricultural residents in Colorado. Of the 24 neurological symptoms assessed, 14 were signi7cantly associated with having reported a pesticide-related illness. Of the 14 symptoms, each demonstrated a trend indicating more of a problem among the pesticide-poisoned individuals. Those who were poisoned reported a response that indicated an increased level of dif7culty resulting from all 14 of the neurological symptoms assessed. Organophosphates applied to crops were found to be associated with reporting a pesticide-related illness in this study population. Herbicides and organophosphates applied to livestock were not signi7cantly associated with reporting a pesticide-related illness. Certain OPs, the application method used, and the nature of the contact farmers had with the chemical were risk factors for pesticide-related illnesses. These activities included granule application to the soil, use of tractor-mounted sprayers or pulled sprayers, and involvement with loading or applying pesticides. Counter, a terbufos, was highly associated with applying granules to soil using a tractor or pulled sprayer. It may be that the application methods, together with the speci7c organophosphate being applied, resulted in the pesticide illnesses reported. Counter is a diethoxy compound with "S on the phosphoryl group. It is an indirect inhibitor of acetylcholinesterase because the thio group must 7rst be oxidized to become a direct inhibitor, but it is always enzymatically activated upon exposure and becomes a direct inhibitor of acetylcholinesterase. Dimethoxy compounds react more rapidly than diethoxy compounds, but the diethoxy compounds are considered more toxic because acetylcholinesterase, when inhibited by a diethoxy compound, shows much slower spontaneous reactivation, creating

96

STALLONES AND BESELER

TABLE 11 Neurological Symptoms Associated with Organophosphorous Pesticides Studies with similar 7ndings (Ref. no.)

Neurological Symptom Changes in cognitive function Had diculty concentrating? Been confused or disoriented? Had trouble remembering things? Have your relatives noticed that you have trouble remembering things? Had to make notes to remember things? Found it hard to understand the meaning of newspapers, magazines and books that you have read? Mood disorders Felt irritable? Felt depressed? Been sleeping more than is usual for you? Had diculty falling asleep? Fatigue Tired more easily than expected for the activity you do? Felt light}headed or dizzy? Had diculty driving because you felt dizzy or tired even though you had slept enough? Had heart palpitations even when not exerting yourself? Peripheral nerve symptoms Had any loss of muscle strength in your legs or feet? Had any loss of muscle strength in your arms or hands? Had numbness or tingling in your 7ngers lasting more than a day? Had numbness or tingling in your toes lasting more than a day? Had diculty moving your 7ngers or grasping things? Been bothered by lack of coordination or lack of balance? Had headaches at least once a week?

5 7 13 14, 15 6 4, 16

4 16

16 17

7a 14 4a 18a 16 19a

a

As indicated by vibrotactile sensitivity measurements.

a longer period of acetylcholinerstase inhibition. Chlorpyrifos, commonly known as Lorsban and Dursban, is a diethyl compound with a"S on the phosphoryl group. It shows a toxicity similar to that of other organophosphates, but was not signi7cantly associated with reporting a pesticide-related illness. Cygon and Malathion are dimethyl compounds with the "S on the phosphoryl group. Dyfonate is a fonofos, contains an arylthio on the phosphoryl group, and is considered very toxic. Thimet is a diethyl phorate and is also considered highly toxic. Most of the pesticides associated with a pesticiderelated illness were diethyl compounds with the "S group or those, such as Thimet and Dyfonate, considered to be highly toxic. Of the 61 individuals applying Lorsban, 32 of them applied it as granules to the soil, but 95% of those applying Counter did so by applying it to the soil. Twenty-one individuals applied Lorsban by an irrigation system. The application of granules to soil may result in the inhalation of dust and a greater risk of exposure than applications using water. Lorsban and Counter are both applied to corn and

sugar beets. Since corn is so commonly grown in this population, it would not be expected to be signi7cantly associated with any particular type of soil pesticide. Counter is more often applied to Irish potatoes and Lorsban is more often applied to wheat crops and alfalfa hay. Independent risk factors for pesticide-related illnesses in this population included gender, with males being at higher risk than females, two neurological symptoms, being depressed and sleeping too much, and using an organophosphate pesticide on the farm. Overall, these 7ndings support much of the research that has been conducted relating pesticide poisoning with chronic neurological sequelae. In addition, it extends the associations to speci7c farming practices, including commodities and pesticide application practices that increase the farm resident’s risk of a pesticide-related illness. Limitations of the study are those related to all cross-sectional surveys. The data were obtained as self-reports without veri7cation from physicians of the pesticide poisoning episodes. In addition, the time sequence of exposure

PESTICIDES AND NEUROLOGICAL SYMPTOMS

and outcome is 8awed. Information on pesticides used on the farms came from the previous year; pesticide poisoning re8ected cumulative lifetime incidence. Nor did the time frame for assessing the neurological symptoms parallel the current exposure to pesticides. Nonetheless, the study 7ndings are similar to those of all previous reports of neurological symptoms and past experience with pesticide poisonings. In addition, the current study provides information related to speci7c farm practices, including application practices and agricultural products. ACKNOWLEDGMENTS This work was funded by the Centers for Disease Control and Prevention, the National Institute of Occupational Safety and Health, U04/CCU806060, and the National Center for Injury Prevention and Control, R49/CCR811509. Its contents are solely the responsibility of the authors and do not necessarily represent the of7cial views of the Centers for Disease Control.

REFERENCES Hayes, A. L., Wise, R. A., and Weir, F. W. (1980). Assessment of occupational exposure to organophosphates in pest control operators. Am. Ind. Hyg. J. 41, 568}575. Mearns, J., Dunn, J., and Lees}Haley, P. R. (1994). Psychological effects of organophosphate pesticides: A review call for research by psychologists. J. Clin. Psychol. 50, 286}293. Garry, V. E., Kelly, J. T., Spafka, J. M., Edwards, S., and Grif7th, J. (1994). Survey of health use characterization of pesticide appliers in Minnesota. Arch. Environ. Health 49, 337}343. Steenland, K., Jenkins, B., Ames, R. G., O'Malley, M., Chrislip, D., and Russo, J. (1994). Chronic neurological sequelae to organophosphate pesticide poisoning. Am. J. Public Health 84, 731}736. Savage, E. P., Keefe, T. J., Mounce, L. M., Heaton, R. K., Lewis, J. A., and Burcar, B. J. (1988) Chronic neurological sequelae of acute pesticide poisoning. Arch. of Environ. Health 43, 38}45.

97

Reidy, T. J., Bowler, R. M., Rauch, S. S., and Pedrozza, G. I., (1992). Pesticide exposure neuropsychological impairment in migrant farm workers. Arch. Clin. Neuropsychol. 7, 85}95. Rosenstock, L., Keifer, M., Daniell, W. E., McConnell, R., and Claypole, K. (1991). Chronic central nervous system effects of acute organophosphate pesticide intoxication. Lancet 338, 223}227. Dodson, G., and Wlodawer, A. (1998). Catalytic triads their relatives. Trends Biochem. Sci. 23, 347}352. Glynn, P. (1999). Neuropathy target esterase. Biochem. J. 344, 625}631. Atkins, J., Glynn, P. (2000). Membrane association of critical residues in the catalytic domain of human neuropathy target esterase. J. Biol. Chem. 275, 24477}24483. Greenhood, D. (1964). ‘‘Mapping.’’ Univ. Chicago Press, Chicago. Johnson, B. L. (1987). ‘‘Prevention of Neurotoxic Illnesses in Working Populations.’’ pp. 169}214. Wiley, New York. Kaplan, J., Kessler, J., Rosenberg, N., Pack, D., and Schaumburg, H. (1993). Sensory neuropathy associated with Dursban (chlorpyrifos) exposure. Neurology 43, 2193}2196. McConnell, R., Keifer, M., and Rosenstock, L. (1994). Elevated tactile vibration threshold among workers previously poisoned with methamidofos other organophosphate pesticides. Am. J. Ind. Med. 25, 325}334. Stephens, R., Spurgeon, A., Calvert, I., Beach, J., Levy, L., and Berry, H. (1995). Neuropyschological effects of long-term exposure to organophosphates in sheep dip. Lancet 345, 1135}1139. Steenland, K., Dick, R., Howell, R., Chrislip, D., Hines, C., Reid, T., Lehman, E., Laber, P., Krieg, E., and Knott, C. (2000). Neurologic function among termiticide applicators exposed to chlorpyrifos. Environ. Health Perspect. 108, 293}300. Burns, C., Cartmill, J., Powers, S., and Lee, K. (1998). An update of the morbidity experience among employees potentially exposed to chlorpyrifos. Occup. Environ. Med. 55, 65}70. Cole, D., Carpio, F., and Leon, N. (1998). Assessment of peripheral nerve function in an Ecuadorian rural population exposed to pesticides. J. Toxicol. Environ. Health 55, 77}91. Stokes, L., Stark, A., Marshall, E., and Narang, A. (1995). Neurotoxicity among pesticide applicators exposed to organophosphates. Occup. Environ. Med. 52, 648}653.