|Impact of Pesticide Use on Health in Developing Countries (IDRC, 1993)|
|Part I: Epidemiology of pesticide poisoning - practical research experience in developing countries|
Instituto Nacional de Salud, Bogota, Colombia
Paraquat, a herbicide of the bipyridil group, has been used in Colombia with increasing frequency since its registration in 1969. This study aimed to discern the effects of paraquat on agricultural workers. A questionnaire and a medical examination (including lung-function tests) were used to determine exposure. The results indicate that there is a positive association between the use of paraquat (at high levels of exposure), smoking habits, and chronic obstructive pulmonary disease.
Paraquat (1,1'-dimethyl-4,4'-bipyridylium dichloride) has been used as a broad-spectrum herbicide for nearly 30 years. Since its registration in Colombia in 1969, paraquat has been used principally for the control of weeds in cotton, sugarcane, potato, corn, and tomato cultivation. It is also used as a desiccator to facilitate the harvesting of rice, sugarcane, soybeans, corn, and sorghum crops and as a defoliant in the cultivation of bananas, cacao, oil palms, citrus, and other fruits. In nonagricultural areas, paraquat is used as a nonspecific contact herbicide.
Paraquat has been classified by the World Health Organization (WHO) as a class 11 (moderately hazardous) compound. In experimental animals, the systemic administration of this substance results in progressively degenerative and potentially lethal lesions in the lungs. In addition to degenerative lung fibrosis, observed effects in humans range from dermal toxicity and irritation of the skin and mucous membranes (Hearn and Keir 1971) to corrosive gastrointestinal effects, renal tubular damage, and liver dysfunction (Grant et al. 1980). At present, there is no antidote for paraquat poisoning and treatment is perforce symptomatic.
Paraquat is not destroyed by the biological processes of plants or mammals. In mammals, absorbed paraquat is largely eliminated unchanged by the kidneys because of its low molecular weight and nonvolatility. However, because of its cationic nature, most of the compound filtered by the kidneys is reabsorbed by the body before it can be eliminated in the urine. This high rate of retention results in toxic effects for unusually low doses of paraquat. The renal tubular damage caused by the compound's redox cycling properties rapidly renders the kidney incapable of removing it from the blood, thereby increasing the harmful effects on bodily functions.
Paraquat is selectively toxic to the lungs, where it is found to accumulate (Levin et al. 1979). Humans are the most susceptible species to paraquatinduced lung damage, which is manifested as a loss of lung capacity and function. The accompanying liver impairment adds to a general loss of vitality. Brain damage has also been observed after paraquat intoxication.
The marked increase in the use of paraquat in Colombia (560 kL in 1980,698 kL in 1981, 1 052 kL in 1982, and 2 483 kL in 1987) prompted the government to investigate its effects on the health of people working with the substance or indirectly exposed to it in the environment. It was expected that those most exposed would be found to suffer a pulmonary fibrosis type of pneumopathy (a condition in which the elastic tissue of the lungs is damaged and replaced by fibrotic, nonelastic tissue), as well as restricted pulmonary function leading to impaired respiration. After a review of regional data on paraquat consumption, target crops, and dosages, the municipality of Carmen de Viboral in the district of Antioquia was selected for study.
Antioquia is an agricultural region where paraquat use is atypically high, representing 10% of the national total. The working population of Carmen de Viboral had allegedly been trained in the use of paraquat by local manufacturers and the Secretariat of Agriculture of Antioquia. However, investigation revealed that only the heads of farms had received training even though an average of nine other people per farm were involved in plantation work using paraquat. Information obtained from the regional health service indicated a high prevalence of paraquat-related health problems in Antioquia, and these figures could be extrapolated to other agricultural regions of the country with a fair degree of accuracy.
Carmen de Viboral is located in the eastern part of Antioquia and covers an area of 448 km². Of its 24 657 inhabitants, 43% live in the municipality while 57% form the rural population. The entire municipal agricultural area is devoted to potato, bean, and corn cultivation. The crop area constitutes 32% of the township land; the rest is used for grazing or forestry. In contrast to other areas, where itinerant workers spray the crops, the agricultural workers in Carmen de Viboral are a stationary population.
The purpose of this study was to assess the frequency and severity of the pathologies associated with unrestricted use of the herbicide paraquat. The specific objectives were:
· To determine the prevalence of respiratory pathologies in
the population of Carmen de Viboral;
· To investigate the presence and measure the concentration of paraquat in the blood and urine of the population;
· To assess the possible and actual levels of exposure in the population by measuring the concentration of paraquat in the air during spraying; and
· To establish an association between different exposure levels and the physiopathological manifestations observed in the population.
Materials and methods
The study population consisted of 21 932 inhabitants of the rural township of Carmen de Viboral. Due to geographical and socioeconomic differences, 2 725 inhabitants of this area were not included in the sampling.
Census and other information were obtained from the Ministry of Health, together with maps and plans from the Department of National Statistics Administration. All data were verified by the investigators. One in every four houses was chosen for the survey. The sample of 941 homes consisted of 5 483 people, all of whom were interviewed. For medical examination, 896 people (a random subsample of about 20% of the original population) aged 7 years and older (except women more than 3 months pregnant) were selected. For the pulmonary function study, this sample was further reduced to 485 by including only men between the ages of 15 and 70 years.
Between April and July 1986, a questionnaire was administered to obtain information on sociodemographic status, occupational or other exposure to paraquat, and general health (particularly regarding the respiratory system). Variables likely to influence the effect of paraquat, such as smoking habits and previous exposure to other lung risk factors, were also determined. In consultation with medical personnel at the San Juan de Dios University Hospital in Bogota and on the basis of current knowledge of the effects of paraquat on humans, the American Thoracic Society's 1978 revision of the Medical Research Council's questionnaire for respiratory function was adopted for this study. The questionnaire was pretested in a pilot study and refined accordingly.
All 896 subjects in the subsample underwent a general medical examination during this same period. Lung function was evaluated by spirometry. Thoracic x-ray examination was not performed for ethical reasons. According to the values for the components of the test, lung function was classified as follows:
· Restrictive = forced vital capacity (FVC) < 80% of the predicted value; forced expiratory volume per second (FEV 1.0) divided by FVC 2 75%; and residual volume (RV) or total lung capacity (TLC) or both < 120% of the predicted value;
· Obsructive = FVC > 80% of the predicted value; FEV 1.0/FVC < 75%; and RV > 80% of the predicted value; and
· Normal = FVC 2 80% of the predicted value; FEV 1.0/FVC 2 75%; RV 80-120% of the predicted value; and TLC 80-120% of the predicted value.
Based on an exposure-to-paraquat score obtained from the questionnaire, blood and urine samples were collected during December 1986 and March 1987 from 603 paraquat users. The concentration of paraquat in these samples was then determined.
During the paraquat spraying period, December 1986 to March 1987, air in the breathing zone of 14 workers was collected and the concentration of paraquat measured. Sampling was done with the aid of personal sampling pumps that passed the air over collection filters. Paraquat was extracted from the filters and concentrations determined by the National Institute of Occupational Safety and Health analytical method S294: paraquat (NIOSH 1979). Samples were analyzed for paraquat content at the National Institute of Health, Bogota, with a 10% external control carried out in the laboratory of toxicology of the University of Laval, Quebec, Canada.
Because data were generated from a random sample of the population, it is possible to compare the group of people that displays the outcome of interest (disease) with the group that does not, according to the presence or absence of exposure to paraquat in a 2 x 2 table. The estimate of the risk of exposed people relative to unexposed people is the cross product of the entries in this table. It represents a relative risk or odds ratio, for which variances, standard errors, confidence limits, and significance tests (chi squared) were calculated.
Of the 5483 people interviewed, 61.2% (3 357 people) lived in the rural area of Carmen de Viboral and 38.8% (2 126 people) in the municipality. By age distribution the largest group consisted of people 15-44 years of age, followed by those under 15 years. In general,11.3% of the population was found to be illiterate; 18.7% of this group consisted of men who lived in the rural area. Of the rural working population, 86.3% worked in agriculture or with cattle and 2.4% performed activities related to the production of ceramics. In the township, the corresponding proportions for these two groups of activities were 26.0% and 13.6%, respectively.
Of the population interviewed, 11.0% (603 people) used paraquat: 4.4% (94 people) of the urban population and 15.2% (509 people) of the population in the rural area. The distribution of paraquat users by age and level of exposure (high and low) was determined by the questionnaire (Table 1).
Of the 603 paraquat users surveyed, six were women; four in the 25-34 years age group and two 35-44 years of age. One woman was in the high-exposure group.
Commonly, 94.5% of the paraquat users applied the herbicide with sprayers; 98.4% used knapsack-type sprayers. Of these, 2.8% considered their spraying equipment to be in poor condition, 21.4% in regular condition, and 75.8% in good condition. However, 71.2% of the users reported that paraquat wet their bodies during spraying. Generally, from the answers reported in the questionnaire, the safety instructions provided by the producers or recommended by international health agencies were not properly followed in Carmen de Viboral.
Questions about present and previous exposure to dust, fumes, mists, gases, or vapours revealed that 6.7% of the full-time workers were exposed to dust in the workplace in the year preceding the study. One-third of these workers were from the rural area.
Among those 10 years old and above, 21% were regular smokers, consuming at least one cigarette every day for the past 5 years or more. Of all smokers, 80% had smoked for 5 years or more. Of the total population surveyed, 33.9% of the men and 11.4% of the women were regular smokers.
When asked whether they had experienced illness during the 2 weeks before the study, 17% of those interviewed gave positive replies. Of the problems cited, 7.2% were related to the respiratory tract in both the rural and urban areas. With regard to duration of illness, 62.5% had been ill for less than 15 days; 22.7% between 15 days and 3 months; and 10.1% for at least I year. The major complaints associated with respiratory problems were coughing, runny nose, fever, chills, phlegm, and dyspnea.
Respiratory diseases were also diagnosed by a physician (Table 2). Chronic bronchitis accounted for 12.8% of the responses, asthma 2.7%, and tuberculosis 0.2%. Official data for the region confirmed the prevalence of tuberculosis.
Compared with nonusers, a higher prevalence of colds that became localized in the chest was observed in users exposed to high levels of paraquat. This difference was not significant when the values for high- and low-exposure users were combined and compared with nonusers. This phenomenon can be explained by differences in the handling of paraquat, i.e., some users apply paraquat correctly, whereas others do not.
In the subsample examined, chronic bronchitis, one of the diseases of the chronic obstructive pulmonary disease (COPD) group and the pathology of greatest concern, was more prevalent in paraquat users. Chronic bronchitis was also more prevalent in both smoking and nonsmoking paraquat users compared with nonusers in both categories (Table 3). The highest prevalence was among paraquat users who smoked. The relative risk or odds ratio for paraquat users who smoked was 7.47 (95% Cl = 2.9-19). From this, an additive effect of paraquat use and smoking can be postulated.
The attributable risk or maximum proportion of chronic bronchitis that can be attributed to paraquat use in the study sample corresponds to 84 cases per 1 000 people. Consequently, for the population (21 932) from which the study sample was taken, it can be extrapolated that there are 1 116 cases of chronic bronchitis in the community resulting from paraquat use.
Because the reference values used in Colombia for lung function tests are for people 15-70 years of age, these tests were done only for the 486 men in this age group (Table 4). Women were excluded from this analysis because only six used paraquat.
Paraquat users in Carmen de Viboral suffered from pulmonary obstructive lung disease in greater proportion than nonusers. The figures displayed give users of paraquat an odds ratio for COPD of 3.11 (95% Cl = 1.41-6.78).
Controlling for exposure to dust and other irritants, 15.6% of paraquat users who were smokers were found to have COPD compared with no smoking nonusers (Table 5). The odds ratio of developing COPD for paraquat users who smoked was 16.6 (95% Cl 0.9-298). Statistically significant differences were not observed for COPD or restrictive pulmonary disease among paraquat users and nonusers who were nonsmokers (Table 5).
Air sampling tests, in which filters from 14 people sampled at the breathing level under normal working conditions were analyzed, showed the concentration of paraquat to be consistently below 0.5 mg/m³, the acceptable level of exposure. All samples of blood and urine taken from users of paraquat in this study gave negative results.
Those exposed to a high level of paraquat consistently displayed a higher prevalence of symptoms related to colds that affected the lower respiratory tract. Subsequent medical examination showed that this symptom complex was a good indicator of the presence of COPD. Clinically diagnosed chronic bronchitis was the predominant disease category with a higher prevalence among paraquat users than nonusers.
The pulmonary function tests clearly demonstrated the statistically significant occurrence of COPD among paraquat users. The prevalence of COPD was higher among paraquat users who smoked compared with nonsmoking users, but the statistical significance was weak, probably because of the small sample size.
There were no significant differences with regard to restrictive lung disease between users and nonusers of paraquat. The largest proportion of those with restrictive disease were among those not using paraquat; exposure to dust and other irritants may have contributed to the condition. However, only two paraquat users reported previous dust exposure and their pulmonary function tests were within normal limits.
The occurrence of COPD in this occupational group was unexpected. This is the first time such an association has been described; poisoning by paraquat more commonly produces restrictive lung disease. It is not known to what extent duration and levels of exposure may have influenced the study findings. Less than 5 years has elapsed since paraquat was introduced in Carmen de Viboral and the levels of observed exposure were within suggested limits.
Acknowledgment - This research was carried out with the assistance of a grant from the International Development Research Centre, Canada.
Grant, H.C.; Lantos, P.L.; Parkinson, C. 1980. Cerebral damage in paraquat poisoning. Histopathology, 4, 185-195.
Hearn, C.E.D.; Keir, W. 1971. Nail damage in spray operators exposed to paraquat British Journal of Industrial Medicine, 28, 399-403.
Levin, P.J.; Klaff, L.J.; Rose, A.G.; Forguson, A.D. 1979. Pulmonary effects of contact exposure to paraquat a clinical and experimental study. Thorax, 34,150-160.
NIOSH (National Institute for Occupational Safety and Health). 1979. Manual of analytical methods (2nd ed.). US Department of Health, Education and Welfare, Cincinnati, OH, USA. Vol. 5, S294, pp. 79-141.