Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Pesticide contamination of workers in vineyards in France


In order to build tools to quantify exposure to pesticides of farmers included into epidemiological studies, we performed a field study in Bordeaux vineyards during the 2001 and 2002 treatment seasons to identify parameters related to external contamination of workers. In total, 37 treatment days were observed in tractor operators corresponding to 65 mixing operations, 71 spraying operations and 26 equipment cleaning. In all, four operators with backpack sprayers and seven re-entry workers were also monitored. We performed both detailed observations of treatment characteristics on the whole day and pesticide measurements of external contamination (dermal and inhalation) for each operation. The median dermal contamination was 40.5 mg of active ingredient per day for tractor operators, 68.8 mg for backpack sprayers and 1.3 mg for vineyard workers. Most of the contamination was observed on the hands (49% and 56.2% for mixing and spraying, respectively). The median contribution of respiratory route in the total contamination was 1.1%. A cleaning operation resulted in a 4.20 mg dermal contamination intermediate between a mixing (2.85 mg) and a spraying operation (6.13 mg). Farm owners experienced higher levels than workers and lower contaminations were observed in larger farms. The contamination increased with the number of spraying phases and when equipment cleaning was performed. Types of equipment influenced significantly the daily contamination, whereas personal protective equipment only resulted in a limited decrease of contamination.


Many epidemiological studies have suggested the possible role of pesticides in the occurrence of cancers (Blair et al., 1992; Dich et al., 1997; Acquavella et al., 1998), neurological diseases (Colosio et al., 2003) or reproductive disorders (Cocco, 2002). Exposure assessment represents a critical concern in these studies (Blair and Zahm, 1990, 1993; Garcia et al., 2000; Hamey, 2001). Indeed, quantitative approaches appear essential to assess the consistency of the results, to study dose–response relationships and to provide data for risk assessment. Exposure assessment includes the determination of its probability (what proportion of individuals are exposed to pesticides?), its frequency (how often do they get exposed?) and its intensity (what is the level of their exposure?). Questionnaires are the main source of data for epidemiological studies and over- or underestimation of risks largely depend on their accuracy. Intensity of exposure is usually considered as a function of area treated (or farm size) and/or the number of years of pesticide use. The assumption that these questionnaire data can be used as surrogates for exposure levels in any agricultural setting requires further validation.

The main objective of the PESTEXPO study in France is to build tools in order to assess exposure to pesticides in various agricultural settings for epidemiological purposes. Field observations intend to identify parameters likely to affect the intensity of exposure and to be ascertained by questionnaires in epidemiological studies. The objective of the first step of the study, presented in this paper, was to obtain specific dermal and inhalation exposure levels for different tasks in vineyards and to determine parameters related to the daily contamination of workers.

Material and methods


The study took place during the 2001 and 2002 spraying seasons (from March to August) in the Gironde area (southwestern France), which is planted with 127,000 ha vineyards (73% of the farms of the area) (Agreste).

Volunteers included in the study were selected through several sources: subjects identified with the help of agricultural bodies (Chambre d'Agriculture, Institut Technique de la Vigne et du Vin), agricultural cooperatives, the local branch of the French health and welfare department for agricultural workers (Mutualité Sociale Agricole) or randomized from the area phone book. All of them were informed in January and February of the study and scheduled to phone the team study the day before the first treatment involving dithiocarbamates.

Sampling Strategy in Operators and Backpack Sprayers

Dithiocarbamates (including mancozeb, maneb, metiram-zinc and propineb) were used as markers of exposure because they were considered representative of pesticide use in vineyards regarding their heavy use in vine-growing (from 2 to 3 kg/ha), their use since the early 1960s and their powder or granule formulation. Several approaches have been taken to assess dermal exposure (Fenske, 1993; van Hemmen and Brouwer, 1995). Direct methods consist in the quantification of the amount of pesticides on the workers' skin or clothing, using patches or the whole body technique (Soutar et al., 2000). Hand exposure is assessed by handwashing or with the use of cotton gloves (OECD). As most of the pesticide application in vineyards takes place during the warm season (from April to August), workers do not commonly wear coveralls and gloves during treatments.

As pesticides may also enter the body through the respiratory tract, potential inhalation exposure was assessed by measuring the concentration of the compounds at the workers' breathing zones.

For these reasons and also as biomonitoring was conducted concurrently (data not shown), the patch and handwash techniques were selected for this study.


They were made with 10 cm × 10 cm layers of surgical cotton gauze, as recommended for dusts and dried particles, and backed with an aluminum foil. They were placed at 11 locations on the skin of the worker under the clothing before each mixing, spraying and equipment cleaning (Figure 1). Following each phase, patches were removed, individually identified, packed in aluminum foil and stored on ice before and during transport to the laboratory and finally transferred to the deep freeze (−20°C) before determination of the pesticide amount. Contamination of each part of the body was obtained by multiplying the concentration on the patches (in microgrammes per cm2) by the surface area of the part, mentioned in Figure 1. For equipment cleaning, only three patches were placed: one on each forearm and one on the head. Dermal contamination during one task was the sum of the values for the different parts of the body.

Figure 1

Patches layout and surface of the sampled body part.


After each phase, 750 ml water was poured slowly over the hands of the worker, who was instructed to rub his hands together. The wash water was collected in a disposable aluminum tray, homogenized by mixing, before 60 ml were collected for analysis.


Dithiocarbamates were collected on quartz fiber filter (Millipore® MF support pads, AP10 37 mm) set in cassettes (Millipore®, 37 mm), which were held to the farmer's shoulder and connected to a personal air sampling pump (Panametrics®), with a sampling rate of 1 l/min. Cassettes were not capped during sampling in order to measure all particles whatever their size. Pump flow measurements were pre- and post-calibrated using the Gilibrator® Air Flow Calibration System. The air sampling times covered the whole working phases (with separate measures for mixing and spraying); afterwards filters were capped individually, stored on ice before and during transport and transferred to deep freeze until analysis. The inhalation dose was calculated from the concentration in the air for a lung ventilation of 15 l/min.

Sampling Strategy in Vineyard Workers

In workers involved in vineyard tasks other than pesticide spraying, pesticide exposure was also studied with patches (replaced at lunch break) located at the same 11 areas as applicators by air sampling with the same devices (one filter for the whole period of work) and handwashing at midday and at the end of the day, that is to say 22 patches, two handwashings and one filter for a whole day of work.

The total pesticide dermal exposure of the workers was calculated from the measured pesticide concentrations on the patches and handwashing samples. The pesticide amounts measured in the different body parts were added to give the amount of pesticides for the whole body. Finally, the total exposure was calculated as the sum of the body, hands and inhalation.

Analytical Methods

Determination of Dithiocarbamates in Patches, Handwashing Water and Quartz Fiber Filters

The determination of dithiocarbamates was performed indirectly by measuring its degradation product in hot and acidic conditions, namely carbon disulfide (CS2). In this aim, correlations were established between two different concentrations of the individual dithiocarbamates (mancozeb, maneb, metiram-zinc and propineb), 100 p.p.b. and 500 p.p.b., and the measured concentrations of CS2 to determine the degradation yield.

About 2 g of patches (or 2 ml of water, or a whole quartz fiber filter with a correction factor to normalize to a 2 g weight) were introduced in a 22 ml head space vial, in which were added 200 μl methanol, 2 ml purified water and 3 ml stannous chloride (5 g/l in 5 M hydrochloric acid). The vial was immediately sealed and deposited in the head-space auto-sampler, where it was heated and pressurized. The gas phase was analyzed with a Hewlett Packard GC/MS system 5890/5972, including a methylpolysiloxane SPB 1 (60 m × 0.32 i.d., 5 μm film thickness, Supelco, Bellafonte PA, USA) capillary column. The MS conditions were as follows: ionization by electron impact (70 eV); selected ion monitoring (SIM) of m/z 76 and 78 (quantitation ion is underlined).

With each serie of samples, calibration standards were prepared using blank matrixes (patches, water or quartz fiber filters) spiked at 10, 50, 100, 500 and 1000 p.p.b. CS2. Each serie also included two internal quality controls prepared with blank matrixes spiked at 50 and 500 p.p.b., respectively, with another stock solution. Using external calibration, the method was linear over the range 10–1000 p.p.b. (mean correlation coefficient r=0.996), the interdays precision CV was less than 20% and the interdays inaccuracy less than 15%. The limit of quantitation (LOQ) of the method was 10 p.p.b., and the limit of detection 5 p.p.b. When the result was below the LOQ, the level was assumed to be equal to the LOQ.

Parameters Related to the Daily Contamination

Values of daily contamination were log-transformed in order to get normally distributed variables. The role of the parameters potentially related to the daily contamination was analyzed by linear regressions and calculation of the Pearson correlation coefficient for continuous variables. For categorical variables, analysis of variance was performed. The following parameters were considered in relation with the total daily contamination: worker status (farm owner versus worker), worker experience in pesticide treatments (duration in years), worker education level (no degree versus any degree in agriculture), vineyard acreage of the farm, acreage treated with pesticides on the observation day, period of the year (spring versus summer), pesticide formulation (powder versus granules), quantity of active ingredient handled on the day, number and total duration of mixing phases, number and total duration of spraying phases, equipment cleaning, types of equipment (straddling, between-row tractor : see Figure 2), presence of a cab on the tractor, clothing (tee-shirt or shorts versus coverall), gloves wearing, technical problems during treatment such as spillages, overflow, pipes blocking (none versus one or more at any phase).

Figure 2

Pictures of the main types of spraying equipment in vineyards.



Figure 3 presents the different levels of the study population: 25 farms were enrolled in the study in 2001 and 2002, corresponding to 38 different subjects: eight of them participated twice and one of them three times. This resulted in 48 observation days during the two seasons: nine in 2001 and 39 in 2002. Observations were scattered in the different parts of the Bordeaux vine-growing area. Among the 48 observations, 37 corresponded to operators with tractors, four to operators with backpack sprayers and seven to workers involved in vineyard tasks. All 38 volunteers were men except for two women (three observations in vineyard work), with a mean age of 43.7 years. All applicators had previous experience in the spraying of pesticides: in average 20.3 years for tractor operators and 11.2 years for backpack sprayers. Of them, 18 were farm owners (47.4%) and 20 were farm workers (52.6%). The educational level was low (no degree) for 16 subjects (43.2%), medium or high for 21 subjects (56.3%). In all, three of the backpack sprayers were in the highest educational level and were well-trained in the test of new products for agrochemical companies.

Figure 3

Presentation of the numbers at the different levels of the study.

Characteristics of the Observation Days

(Table 1) Two-thirds of the observations took place during the spring (N=30), the others during the summer (N=18). The minimal and maximal temperatures during spraying or vineyard tasks averaged 19.5°C (SD 6.0) and 24.7°C (SD 6.2), respectively. A quarter of the values of the minimal temperature were below 15°C and a quarter of the maximal values above 29°C. Observations lasted about 7.25 h for operators with tractors, 2.5 h for backpack sprayers and nearly 9 h for vineyard workers. Dithiocarbamates used were mancozeb (73.2% of the observation days), metiram zinc (18.3%), maneb (7.1%) and propineb (1.4%).

Table 1 Characteristics of the observation days and median of the total contamination observed that day.

Characteristics of Treatment Phases for Operations with Tractors

Mixing Operations

(Table 2) In total, 65 mixing operations were observed. The duration of a single phase of mixing averaged 19 min. A mean quantity of 10.9 kg of active ingredient was used per operation, corresponding to the opening and pouring of 2.8 containers in average (mainly bags) and leading to a mean volume mixture of about 900 l. Liquid formulations were used only four times, all others were solid: powders (N=22, 33.8%), or granules (N=39, 60%). Technical problems were observed in 19 (29.2%) of the mixing operations. They consisted in spillages due to containers falling down (N=6), overflow (N=4), excessive foaming (N=3), tank or pipes being blocked (N=3) and any other type of mechanical problems (N=3).

Table 2 Main characteristics of the working phases for applicators with tractor sprayers.

Spraying Operations

In total, 71 spraying phases were observed with a mean duration of almost 2 h (116 min). The formulated pesticides (powders, granules and liquids) were mixed with water, with a mean volume of the tank sprayer of 929 l. The mean concentration of active ingredient in the mixture was 14.6 g/l, ranging from 1.4 to 40 g/l. The volume applied was on average 137 l/ha, ranging from 60 to 250 l/ha. A quarter of the observations concerned straddling tractors; others were ordinary tractors driving between rows. Spraying was performed with a four side closed cabin 43 times (60.6%) and without or with an open cabin 28 times (39.4%). Technical problems were observed 17 times (23.9 %), corresponding nine times to the unblocking of a nozzle, and other times to mechanical problems (N=4) or external events (phone call, customer coming in, etc.) (N=4).

Equipment Cleaning

Equipment cleaning was observed for 26 tractor operators. The mean duration of the cleaning was 18 min. Mixture left in the tank at the end of spraying was 21 l in average ranging from 0 to 100 l. A technical problem was noticed corresponding to the overflow of the bucket used when emptying the mixture from the tank.

Protective Equipment

(Table 3) Operators wore gloves on 40 mixing operations (61.5%) and masks on 28 (43.1%). They had bare forearms or lower legs, respectively, in 26.2% and 16.9% of mixing phases. They wore gloves on seven spraying operations (9.9%) and masks on two (2.8%). They had bare forearms or lower legs in 35.2% and 26.8% of spraying phases, respectively. During equipment cleaning, gloves were worn by 11 operators (40.7%) and more than a quarter had bare forearms and/or bare lower legs.

Table 3 Protective equipments in workers during the various phases.

Characteristics of Treatment Phases for Operation with Backpack Sprayers

In all, four operators with backpack sprayers were monitored, two of them (A and B) on the same day, the same farm and the same task but with different personal protective equipment. Each operator mixed and sprayed once. The quantity mixed was 320 g for A and B, 108 g for C and 525 g for D, corresponding to a concentration of 24.6, 5.4 and 35 g/l of mancozeb in the mixture, respectively. Mixing was performed under extractor hood twice (operators B and C), without gloves and coverall only for A and never with a mask. The mixing duration averaged 36 min. All formulations were granules.

Average spraying time was 32 min. All operators but operator A wore a mask, a coverall and gloves. They sprayed a mean area of 0.1 ha. Two cleaning operations were monitored (A and B), lasting 25 min in both cases. Operator A had no protective equipment and wore a tee-shirt and shorts, whereas B wore a mask, gloves, a coverall and boots (Table 3).

Characteristics of Work in the Vineyard

In all, four vine workers were observed on a single farm in 2002, corresponding to seven observations. Of them, three workers were observed on the same day, on parcels located near vineyards treated with mancozeb on that day (observations E, F and G). The same three workers as well as a fourth were observed later on in the season in reentry conditions (observations H to K): the plots they were working in had been treated 5 days before by mancozeb. On the two observation days, they were involved in various tasks: quality and disease checking (E and H), lowering or bending wires that support vine branches (F and I), removing leaves at the vine stock bottom (J), forming single buds (G and K) (Table 3).

Levels of Contamination

Contamination on the Whole Day

The median of the actual dermal contamination on the whole day of treatment was 40.5 mg, for tractor operators (range: 0.4–3358 mg), 68.8 mg for backpack sprayers (range: 1.0–236.8) and 1.3 mg for vineyard workers (range: 0.6–2.6). For inhalation, the median of the daily contamination was 0.44 mg for tractor operators, 0.56 mg for backpack sprayers and 0.01 mg for vineyard workers. Therefore, in the 19 observations, where respiratory measurements were available all along the mixing and spraying tasks, the respiratory exposure averaged 2.9 % of the total contamination (range 0.02–16.3%). The average contribution to daily dermal exposure was 51.7% and 22.5% for mixing/loading and application, respectively. For the 25 operators, for whom contamination during equipment cleaning was determined, the respective contributions of spraying, mixing and cleaning in the total contamination varied with individuals. The task which contributed the most to the dermal exposure was spraying in 13 operators, cleaning in eight subjects and mixing in four subjects.

Dermal Contamination for Each Task

In tractor operators, the median of the dermal contamination during a single operation was 2.85 mg for mixing, 6.13 mg, for spraying and 4.20 mg for cleaning (Figure 4). In total, 10% of the operations resulted in values exceeding 37.2 mg for mixing, 50.1 mg for spraying and 52.2 mg for cleaning. The maximal values (564 mg for a mixing operation and 463 mg for a spraying operation) were observed in the same operator on two different observation days. This operator used a tractor without a cabin, wore a protective coverall but no gloves. No technical problem was observed during the tasks he performed. For cleaning, the maximal value was 152 mg, corresponding to a 10 min operation including the draining of 50 l of mixture in a yard, and the washing of a straddling tractor with a hose, with no personal protective equipment.

Figure 4

Distribution of the dermal contamination of workers during mixing-loading and application with tractors (minimum, 25e percentile, median, 75e percentile, maximum).

Detailed contributions of each body part in the dermal contamination during mixing and spraying are presented in Figure 5a and b. Hands contributed half of the dermal contamination in spraying (49.0%) and slightly more in mixing (56.2%). Forearm contribution ranked second for mixing (13%), slightly higher than legs (12.3%) and head (10.8%). However, in spraying, forearm contribution was only 7.4%, lower than legs (16.0%) and head (17.3%).

Figure 5

(a) Mean contribution of each body area to the actual dermal contamination (tractor applicators). (b) Mean contribution of each body area to the actual dermal contamination of backpack sprayers and vineyard workers (whole day of work).

Individual levels for backpack sprayers and vine workers exposure are presented in Table 4 for dermal and respiratory contamination during each operation.

Table 4 Dermal contaminations of workers during mixing and spraying with backpacks and during vinework (mg of active ingredient).

Correlation between Observed Parameters and Daily Contamination

(Table 5) Among worker characteristics, only his status was related to the contamination: farm owners experienced significantly higher levels (P=0.01). The number of years of experience in pesticide treatments did not influence the level of contamination (ρ=−0.10, P=0.6), neither did the educational level (P=0.98). The total vineyard acreage was significantly inversely correlated to the contamination of the workers (ρ=−0.5, P=0.003), whereas the area treated on the day of observation was not correlated (ρ=0.11, P=0.5). The contamination increased significantly with the number of spraying phases (ρ=0.36, P=0.03) but not with their duration (ρ=0.22, P=0.19). Neither the number (ρ=0.25, P=0.13) nor the duration of the mixing phase (ρ=0.13, P=0.45) influenced the daily contamination. Equipment cleaning significantly increased the contamination (P=0.007). Even if the highest values were observed predominantly with granules and powders rather than with liquids or dispersible bags, the contamination on the whole day was not correlated to the type of formulation (P=0.61). The quantity of active ingredient handled on the day tended to increase the contamination but not significantly (ρ=0.26, P=0.12). The lowest contamination was observed with straddling tractors, the highest with tractor mounted equipment; the trailed sprayers led to intermediate levels. The difference in contamination between the three types of equipment was significant (P=0.02). A higher contamination was measured for the tractor without a cab or with open cabs (P=0.02). The influence of the clothing (wearing a tee-shirt or shorts versus a coverall or long sleeves plus trousers) on the daily contamination of the whole body was limited (P=0.96), as well as wearing gloves (P=0.2). Technical problems reported in any task of the treatment did not significantly influence the contamination (P=0.8).

Table 5 Role of observed parameters in the daily contamination (qualitative variables).


Field observations in the Bordeaux area provided original data on task characteristics and on the levels of pesticide contamination for tractor operators, backpack sprayers and workers in vineyards. They suggested that the daily intensity of the contamination was related neither to the treated acreage nor to the duration of the operation, but was negatively correlated with the size of the farm, the status of the worker, the type of equipment, the number of spraying phases and the presence of a cleaning.

Even if all areas of Gironde and all types of equipment were represented in our study, the sample was constituted of volunteers and consequently could not be considered fully representative of the population of Gironde vine workers. Actually, the mean size of the farms in our study (47 ha) exceeded the mean size of the farms in Gironde (13 ha). It is likely that real levels of contamination are even higher as subjects accepting to participate in such a study were more aware of prevention and had better conditions of work.

Sampling strategies we used were one of those recommended by OCDE (Chester, 1993). The patch method was sometimes criticized because of possible nonuniform deposition. In our study, this limitation was balanced by the large number and large size of the patches (1100 cm2) corresponding to 5.8% of the total body area, and by the high number of measurements (average: 50 by individual). Moreover, the deposition of solid formulations is expected to be uniform than with liquids, which are likely to splash during mixing.

The handwashing technique is usually considered to underestimate the quantity of pesticides as pesticides could be embedded in the skin, and could penetrate during the time of the operations. However, Hsu et al. (1988) describe that for most pesticides 90% of the total amount of pesticide was removed during the first rinse and this inconvenience should be limited by the repetition of washing after each operation (five washes per day in means in our study). Other techniques were proposed like rinsing with ethanol or propanol (Geno et al., 1996), but authors agree to say that the removal is mostly explained by the mechanical actions. Efficiency of handwash sampling for mancozeb was reviewed by Brouwer et al. (2000) and was shown to be 66% and 81% in two studies using propanol for rinsing and 86% in a study using water and soap for rinsing.

We also chose the patch and handwashing methods so that the operator was comfortable with respect to temperature, and then minimized interference with usual conditions of work. Moreover, forcing workers to wear a coverall and gloves would have added an artificial barrier to the dermal penetration and would have disturbed the biological monitoring.

We also measured the total deposition of pesticides on the filters. It can be argued that the penetration of particles in the respiratory tract varies with particle size and that we should have used size selective techniques. We decided to measure particles independently of their size as oral uptake of nonrespirable particles deposited in the mouth, the nasal cavity or the upper airways may also be of importance contributing to the total contamination. However, the contribution of the respiratory route appears limited, as in other previous studies (Dowling and Seiber, 2002).

Levels of daily contamination appeared rather consequent in our study when comparing to regulatory levels. For mancozeb, the no observable adverse effect level (NOAEL) is 5 mg/kg/day conducting to an acceptable daily intake (ADI) of 0.05 mg/kg/day, that is 3 mg per day for a 60 kg individual. Assuming a 10% penetration of mancozeb through the skin, the half of all subjects in our study and 60% of the tractor operators exceeded the ADI despite wearing gloves were quite often when mixing. Moreover, spraying was mainly performed with tractors equipped with closed cabins. Technical problems occurred in about a quarter of the operations, and a tee-shirt or shorts were currently worn. However, these circumstances were considered as usual conditions of work by operators, and they did not appear to explain the major part of the contamination. Although dithiocarbamates are now generally applied twice a year in Bordeaux vineyards, they were used up to 10 times a year till the 1990s against mildew. For vine workers, the exposure lasts longer as they are in contact with the vines for several days after each treatment.

Differences were observed in the contamination according to the tasks. Spraying resulted in the highest contamination, but not proportionally to its duration: indeed, spraying corresponded to 54% of the contamination but to 80% of the treatment duration. Equipment cleaning, which corresponded to the shorter operation (7% of the duration), resulted in a median concentration intermediate between spraying and mixing operations. The part of the contamination during equipment cleaning was even responsible for the majority of exposure in some subjects (up to 80% of the daily contamination), despite the fact that measurements during cleaning were performed with a simplified sampling strategy (measurements only for hands, forearms and head), and consequently underestimate the contamination. This underestimation can be estimated to 20% to 30% when considering that arms, trunk and legs accounted for 27.9% of the dermal contamination while spraying and 22.7% while mixing. Specific contribution of contamination during cleaning was observed for the first time in the present study.

The contamination of the hands was the highest in all tasks performed, as expected from other studies. Even for subjects with gloves the contamination was quite high. This result is certainly explained by the fact that subjects were left free either to use or not to use gloves and to choose their own gloves, possibly old ones. Brouwer et al. (2001) emphasized that the reduction of external exposure is not always linearly related to increase in protection and he proposes several explanations for this: (i) “proper use” of the gloves is not always achieved under usual conditions of work, because of discomfort, lack of knowledge, economical assertions, (ii) a breakthrough of the protective material occurs after a lag time, in relation with the importance of the pesticide loading, which could especially explain high levels of contamination at the end of the day during cleaning phase, (iii) some local parameters like moisture of the skin or sprinkling with water (rain, water during cleaning) can influence the permeation.

Our study identified some parameters to be associated to the whole day contamination. Unlike usually considered in epidemiological studies, the contamination did not increase with the acreage, or with the duration of the treatment. Moreover, larger was the farm lower was the contamination, which can be explained by better equipment and better hygiene in larger farms. As workers were observed in larger farms, this might also explain that they were less contaminated than farm owners. The type of equipment appears to influence contamination in a significant manner. Straddling tractors resulted in significantly lower contamination than between rows ones. This might be explained by the distance between the operator and the pesticide cloud and/or the vegetation, by the differences in the cabin closing, by a shorter cleaning, and also by the fact that straddling tractors were mainly observed in larger farms and were driven by farm workers.

In the present analysis, a technical problem was defined as any unexpected event occuring during treatment and likely to increase the exposure of the vine worker. No association was observed between the occurrence of these events as a whole and the daily dermal level of contamination. However, further analysis of our data will determine the influence of specific events (nozzle unblocking, overflow, etc) on the contamination during specific tasks (mixing, spraying, cleaning), and of specific parts of the body (hands, trunk, legs, arms, etc).

Our results cannot be easily compared to others. Indeed, to our knowledge, only one study in the scientific literature reported exposures for subjects in vineyards, but the conditions were experimental: only one row was treated by each tractor, which cannot represent the contamination on a whole day of treatment (Coffman et al., 1999). Even in other settings available data in real conditions of use are scarce and the sample sizes are limited, rarely exceeding 10en subjects or replicates.

Our data on contamination by dithiocarbamates should enable extrapolations to other pesticides as external exposure is mainly affected by nonchemical factors which should not be specific to the compound. This is why we plan other studies in vineyards with other makers of exposure to verify this assumption. Folpet was used as another marker of exposure in 2003 and sampling is currently under analysis. We also plan to compare the parameters affecting the exposure in vineyards and in open fields to face differences and similarities like the role of equipment, of work habits, etc.


  1. Acquavella J., Olsen G., Cole P., Ireland B., Kaneene J., Schuman S., and Holden L. Cancer among farmers: a meta-analysis. Ann Epidemiol 1998: 8 (1): 64–74.

    CAS  Article  Google Scholar 

  2. Agreste. Recensement Agricole. 2000 Chiffres locaux. Tableaux essentiels par région. [cited 2004-05-17]. Avalaible from : URL: 〈〉.

  3. Blair A., and Zahm S. Methodologic issues in exposure assessment for case-control studies of cancer and herbicides. Am J Ind Med 1990: 18: 285–293.

    CAS  Article  Google Scholar 

  4. Blair A., and Zahm S.H. Patterns of pesticide use among farmers: implications for epidemiologic research. Epidemiology 1993: 4 (1): 55–62.

    CAS  Article  Google Scholar 

  5. Blair A., Zahm S.H., Pearce N.E., Heineman E.F., and Fraumeni J.F. Clues to cancer etiology from studies of farmers. Scand J Work Environ Health 1992: 18: 201–215.

    Article  Google Scholar 

  6. Brouwer D.H., Boeniger M.F., and VanHemmen J. Hand wash and manual skin wipes. Ann Occup Hyg 2000: 44 (7): 501–510.

    CAS  Article  Google Scholar 

  7. Brouwer D.H., Marquart H., and van Hemmen J.J. Proposal for an approach with default values for the protection offered by PPE, under European new or existing substance regulations. Ann Occup Hyg 2001: 45 (7): 543–553.

    CAS  Article  Google Scholar 

  8. Chester G. Evaluation of agricultural worker exposure to, and absorption of, pesticides. Ann Occup Hyg 1993: 37 (5): 509–523.

    CAS  PubMed  Google Scholar 

  9. Cocco P. On the rumors about the silent spring. Review of the scientific evidence linking occupational and environmental pesticide exposure to endocrine disruption health effects. Cad Saude Publica 2002: 18: 379–402.

    Article  Google Scholar 

  10. Coffman C.W., Obendorf S.K., and Derksen R.C. Pesticide deposition on coveralls during vineyard applications. Arch Environ Contam Toxicol 1999: 37 (2): 273–279.

    CAS  Article  Google Scholar 

  11. Colosio C., Tiramani M., and Maroni M. Neurobehavioral effects of pesticides: state of the art. Neurotoxicology 2003: 24 (4,5): 577–591.

    CAS  Article  Google Scholar 

  12. Dich J., Zahm S., Hanberg A., and Adami H. Pesticides and cancer. Cancer Causes Control 1997: 8: 420–433.

    CAS  Article  Google Scholar 

  13. Dowling K., and Seiber J. Importance of respiratory exposure to pesticides among agricultural populations. Int J Toxicol 2002: 21: 371–381.

    CAS  Article  Google Scholar 

  14. Fenske R.A. Dermal exposure assessment technique. Ann Occup Hyg 1993: 37 (6): 687–706.

    CAS  PubMed  Google Scholar 

  15. Garcia A., Orts E., Esteban V., and Porcuna J. Experts' assessment of probability and level of pesticide exposure in agricultural workers. JOEM 2000: 42 (9): 911–916.

    CAS  PubMed  Google Scholar 

  16. Geno P.W., Camann D.E., Harding H.J., Villalobos K., and Lewis R.G. Handwipe sampling and analysis procedure for the measurment of dermal contact with pesticides. Arch Environ Contam Toxicol 1996: 30: 132–138.

    CAS  Article  Google Scholar 

  17. Hamey P.Y. The need for appropriate use information to refine pesticide user exposure assessments. Ann Occup Hyg 2001: 45: S69–S79.

    CAS  Article  Google Scholar 

  18. Hsu J., Wheeler H.J., Camann D., Schattenberg H.R., Lewis R.G., and Bond A. Analytical methods for detection of nonoccupational exposure to pesticides. J Chromatogr Sci 1988: 26 (4): 181–189.

    CAS  Article  Google Scholar 

  19. OECD Environmental Health and Safety publications. Series on Testing and Assessment no 9. Guidance document for the conduct of occupational exposure to pesticides during agriculltural application. Paris 1997. [cited 2004-05-17]. Avalaible from :URL: 〈〉.

  20. Soutar A., Semple S., Aitken R.J., and Robertson A. Use of patches and whole body sampling for the assessment of dermal exposure. Ann Occup Hyg 2000: 44 (7): 511–518.

    CAS  Article  Google Scholar 

  21. van Hemmen J.J., and Brouwer D. Assessment of dermal exposure to chemicals. Sci Total Environ 1995: 168: 131–141.

    CAS  Article  Google Scholar 

Download references


This study was supported by funding from the Association pour la Recherche contre le Cancer (réseau ARECA) and the Ligue contre le Cancer (Comités de la Gironde et du Calvados) . We thank all the vine workers who participated in PESTEXPO and the agricultural bodies for their interest in the study. We aknowledge Dr Pierre-Gérard Pontal for valuable discussions on methodology and for comments on the manuscript.

Author information



Corresponding author

Correspondence to Isabelle Baldi.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Baldi, I., Lebailly, P., Jean, S. et al. Pesticide contamination of workers in vineyards in France. J Expo Sci Environ Epidemiol 16, 115–124 (2006).

Download citation


  • pesticide
  • occupational exposure
  • vineyard
  • dermal contamination
  • inhalation.

Further reading


Quick links