general enquiries on this form should be made...

General enquiries on this form should be made to:Defra, Science Directorate, Management Support and Finance Team,Telephone No. 020 7238 1612E-mail: [email protected]

SID 5 Research Project Final Report

SID 5 (Rev. 3/06) of 21

NoteIn line with the Freedom of Information Act 2000, Defra aims to place the results of its completed research projects in the public domain wherever possible. The SID 5 (Research Project Final Report) is designed to capture the information on the results and outputs of Defra-funded research in a format that is easily publishable through the Defra website. A SID 5 must be completed for all projects.

This form is in Word format and the boxes may be expanded or reduced, as appropriate.

ACCESS TO INFORMATIONThe information collected on this form will be stored electronically and may be sent to any part of Defra, or to individual researchers or organisations outside Defra for the purposes of reviewing the project. Defra may also disclose the information to any outside organisation acting as an agent authorised by Defra to process final research reports on its behalf. Defra intends to publish this form on its website, unless there are strong reasons not to, which fully comply with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.Defra may be required to release information, including personal data and commercial information, on request under the Environmental Information Regulations or the Freedom of Information Act 2000. However, Defra will not permit any unwarranted breach of confidentiality or act in contravention of its obligations under the Data Protection Act 1998. Defra or its appointed agents may use the name, address or other details on your form to contact you in connection with occasional customer research aimed at improving the processes through which Defra works with its contractors.

Project identification

1. Defra Project code PS2610

2. Project title

The TEMPEST Study - Epidemiology of the association between pesticides and Parkinson's Disease

3. Contractororganisation(s)

University of Aberdeen, University Office, Regent Walk, Aberdeen, AB24 3FX, Scotland, UK.

Institute of Occupational Medicine, Research Avenue North, Riccarton, Edinburgh, EH14 4AP, Scotland, UK.

54. Total Defra project costs £ 121,409(agreed fixed price)

5. Project: start date................ 01 March 2006

end date................. 01 March 2007

SID 5 (Rev. 3/06) of 21

6. It is Defra’s intention to publish this form. Please confirm your agreement to do so...................................................................................YES NO (a) When preparing SID 5s contractors should bear in mind that Defra intends that they be made public. They

should be written in a clear and concise manner and represent a full account of the research project which someone not closely associated with the project can follow.Defra recognises that in a small minority of cases there may be information, such as intellectual property or commercially confidential data, used in or generated by the research project, which should not be disclosed. In these cases, such information should be detailed in a separate annex (not to be published) so that the SID 5 can be placed in the public domain. Where it is impossible to complete the Final Report without including references to any sensitive or confidential data, the information should be included and section (b) completed. NB: only in exceptional circumstances will Defra expect contractors to give a "No" answer.In all cases, reasons for withholding information must be fully in line with exemptions under the Environmental Information Regulations or the Freedom of Information Act 2000.

(b) If you have answered NO, please explain why the Final report should not be released into public domain

Executive Summary7. The executive summary must not exceed 2 sides in total of A4 and should be understandable to the

intelligent non-scientist. It should cover the main objectives, methods and findings of the research, together with any other significant events and options for new work.

Parkinson’s disease (PD) is a neurodegenerative brain disease that generally develops late in life and affects 0.1% of the general population. In most cases the cause of the illness is unknown. Many studies have shown a modest increased risk of PD with pesticide use but no individual pesticide has been implicated consistently. However, a number of studies have not found a significant association between pesticide exposures and PD. It is important both for pesticide users and regulatory authorities to be able to identify those pesticides that increase the risk of PD. Better pesticide estimates offer a method to identify those pesticides, if any, that may increase the risk of this disease and would increase the power of case-control studies to detect associations between specific groups of pesticides and Parkinson’s disease.

The aim of this study was to better characterise pesticide exposures by the creation of a pesticide task exposure matrix (TEM) for Scotland for the period from 1945 to 2005. Professional and amateur uses were to be considered, including veterinary medicines (e.g. sheep dips) and biocides that contain the relevant active ingredients.

A task list was generated of tasks involving pesticide exposure in Scotland. Pesticides (insecticides, herbicides and fungicides) which had been identified for use in tasks were then coded using the following classification system. Insecticides were grouped into four main types: organochlorines (OC); organophosphates (OP); carbamates; and others. Herbicides were classified by their mode of action, using the Herbicide Resistance Action Committee (HRAC) system which has 24 groups. To avoid an unmanageably large number of herbicide categories in subsequent analyses herbicides were further limited to three commonly used HRAC herbicide groups (C1, D, O) and the remaining 21 groups were collapsed into a fourth group, ‘Other’ herbicides. Fungicides were categorised into carbamates (which includes dithiocarbamates) and ‘other’.

Two occupational hygienists assessed the likely probability and intensity of exposures for each pesticide group for a given use. A task-exposure matrix was then produced listing pesticide groups against the likely exposure to pesticide when used in that particular task (e.g. spraying barley). Once this matrix was produced it was applied to existing data gathered in Scotland as part of a study of the genetic, environmental and occupational risk factors for Parkinson’s disease (the Geoparkinson Study) and the resulting exposure estimates employed in a re-analysis of that dataset.

Statistical re-analysis of the Geoparkinson data (limited to Parkinson’s disease cases and all controls) adjusted for smoking and for familial history of the disease, and then added each of the pesticide inhalation exposures to the regression model. We then repeated these analyses for the skin exposure estimates. We concluded that there was no evidence for a relationship between risk of PD and exposure to any or all of the pesticides.

This project has successfully developed a TEM for pesticide usage in Scotland and then applied this

SID 5 (Rev. 3/06) of 21

TEM to an existing dataset drawn from the Geoparkinson study to establish proof of concept. The outcome of the re-analysis was not unexpected given the sample size: there was no evidence of an association between any pesticide group and Parkinson’s disease. Conceptually the TEM developed in this project offers improved pesticide exposure estimation and could be employed in any future case-control study of PD and pesticide exposure in Scotland.

Project Report to Defra8. As a guide this report should be no longer than 20 sides of A4. This report is to provide Defra with

details of the outputs of the research project for internal purposes; to meet the terms of the contract; and to allow Defra to publish details of the outputs to meet Environmental Information Regulation or Freedom of Information obligations. This short report to Defra does not preclude contractors from also seeking to publish a full, formal scientific report/paper in an appropriate scientific or other journal/publication. Indeed, Defra actively encourages such publications as part of the contract terms. The report to Defra should include: the scientific objectives as set out in the contract; the extent to which the objectives set out in the contract have been met; details of methods used and the results obtained, including statistical analysis (if appropriate); a discussion of the results and their reliability; the main implications of the findings; possible future work; and any action resulting from the research (e.g. IP, Knowledge Transfer).

SID 5 (Rev. 3/06) of 21

Development of a task - exposure matrix (TEM) for pesticides usage in Scotland and its application to occupational data from an existing case-control study of pesticide exposure and Parkinson’s disease.

Dick FD, Miller BG, MacCalman L, Semple SE, Ritchie P, van Tongeren M, Sherriff D.

Background

Parkinson’s disease (PD) is a neurodegenerative brain disease that generally develops late in life and affects 0.1% of the general population. PD develops once enough nerve cells in the area of the brain that controls movement have died. The loss of these cells in the basal ganglia affects movement and leads to the sufferer experiencing tremor, rigidity and falls. In most cases the cause of the illness is unknown. Medical researchers have found it hard to identify which factors trigger the death of these cells and at what time of life these triggers may act. Many studies have shown a modest increased risk of PD with pesticide use but no individual pesticide has been implicated consistently. However, some studies have not found a significant association [Koller 1990, Jimenez-Jimenez 1992, Taylor 1999, Behari 2001] and others (limited to fungicides or herbicides) have found negative associations [Brown 2006]. It is important both for pesticide users and regulatory authorities to be able to identify those pesticides, if any, that increase the risk of PD. Better pesticide estimates offer a method to identify those pesticides or groups of pesticides that may increase the risk of this disease and would increase the power of case-control studies to detect associations between specific groups of pesticides and Parkinson’s disease. The main aim of this study was to better characterise pesticide exposures by the creation of a pesticide task -exposure matrix for Scotland from 1945 to 2005.

1.0 Introduction

1.1 Agricultural use of pesticides in Scotland

1.1.1. Arable farming

In the UK, pesticides were first used for crops in the 19th century. Herbicide use began in the 1940’s and by the 1960’s was universal (Potts 1986). In the 1970’s cereal production was highly profitable and this led to increased pesticide use, which until then had been uneconomic except for weed control. However, in the last decade the weight of pesticides applied has reduced due to; more active agents, lower pesticide concentrations, use of several low level applications instead of a full rate application and improved spraying technology. Currently, arable farming’s use of pesticide in Scotland accounts for approximately 1360 tonnes of active ingredient (AI) per year with cereals accounting for 70% (957 tonnes AI) of all pesticides used in this sector. However, soft fruit (10.20 kg AI/ha) and potatoes (8.03 kg AI/ha) use the greatest weight of active ingredient per hectare.

Table 1: Annual arable farming use of pesticides in Scotland: weight of active ingredient (kg AI) applied per hectare of crop.

1Pesticide Usage in Scotland. Arable Crops 2002. Scottish Agricultural Science Agency (SASA)2Pesticide Usage in Scotland. Vegetables for Human Consumption 2003. Scottish Agricultural Science Agency (SASA)

SID 5 (Rev. 3/06) of 21

Crop Hectares (ha) under cultivation kg AI kg AI/haCereal

(Barley, Wheat, Oats)1448,718 957,388 2.14

Rape1 36,401 98,973 2.71Potatoes1 29,689 238,417

(excludes dessicants)8.03

Vegetables2 10,594 42,139 3.97Soft fruit3 1,828 18,642 10.20

Total 527,230 1,355,559 2.57

3 Pesticide Usage in Scotland. Soft Fruit Crops 2001. Scottish Agricultural Science Agency (SASA)

SASA reports on pesticide usage in Scotland are reproduced on the Central Science Laboratory’s website (http://www.csl.gov.uk/index.cfm)

1.1.2. Livestock farming

Pesticides are used in livestock farming for control of parasites such as warble fly in cattle and mites in chickens. Their major use, however, is in sheep farming, where flocks are treated for sheep scab and other infestations. These veterinary medicines for sheep are mainly applied as dips (organophosphates, pyrethroids) or pour-ons (e.g. triazine, pyrethroids). According to the Scottish Agricultural Science Agency’s pesticide usage in Scotland survey of 1996 (http://www.csl.gov.uk/index.cfm) approximately 11 million sheep were treated by dipping and 3.5 million were treated with pour-on insecticides in that year. Pesticide exposures to shepherds, crofters and farmers during dipping are infrequent, given the seasonal nature of sheep dipping, but may be substantial. (Buchanan et al., 2001)

1.2 Amenity use of pesticides in Scotland

Within Scotland, it is estimated that the amenity sector’s use of pesticides accounts for 4-6% of all pesticide use (Watkins, 2002). The major amenity uses are:

Industrial ( roads, railways, rights of way etc.) ~ 35% Grass ( golf courses, parks, etc.) ~30% Public health and pest control operatives (private and local authority) ~ 20% Forestry and horticulture ~ 15%

Within the UK, these uses accounted for 1300 tonnes of active ingredient (formulated product = 7035 tonnes). With the exception of private sports clubs employing professional groundsmen or green keepers, contractors apply the majority of pesticides in this sector. The introduction of more effective molecules, improvement in application techniques and the withdrawal of some of the older pesticides, has reduced the amount of pesticide being used.

1.3 Hobby gardeners

An estimated 0.25 to 1% of all pesticide use by active ingredient is by hobby gardeners. Watkins (Watkins, 2002) estimates that hobby use accounted for 4 tonnes of active ingredient in 1999 (formulated product = 185 tonnes). Of this:

92% was applied as herbicide 7% was applied as insecticide 1% applied as fungicide

Nowadays products for amateur gardeners differ from those used by the agricultural and amenity sectors in several ways; garden products are normally formulated as a granular or dilute solution; the majority of pesticides are supplied as ready to use products; all garden products must be suitable for use without the need for protective clothing.

The Geoparkinson study indicated that 47% of the Scottish sample had been exposed to pesticides as a result of gardening activity. By extrapolation from the 2001 census (General Register Office for Scotland, 2002), almost 2 million adults in Scotland will have been exposed in this way.

1.4 Pesticide exposure estimation

Ideally, pesticide exposure estimates would take account of all possible routes of exposure and the main factors influencing exposure throughout an individual’s lifetime. Pesticides may be absorbed by inhalation, ingestion (including mucociliary clearance of inhaled particles/droplets) or, in some cases, through the skin i.e. dermal absorption. Dermal (skin) exposure is a significant exposure route for some pesticides such as the organophosphates. The pesticide formulation (e.g. powder, granules, liquid concentrate, pre-mixed solution), packaging, method of application, worker training, use of personal protective equipment (PPE) and prevailing weather can all influence exposure. One major challenge is that many people cannot recollect which agents they have used making it difficult to establish an association between a specific agent and PD.

SID 5 (Rev. 3/06) of 21

http://www.csl.gov.uk/index.cfm


Additionally, some pesticide users have applied a range of different agents, sometimes as mixtures, making it even more difficult to implicate specific agents. Exposures to pesticide active ingredients may arise not only through work but also in the home e.g. insecticides for human parasites, flea treatments on family pets, exposure to timber treatment agents, use of garden pesticides and the ingestion of pesticide-contaminated water or food. This last exposure is difficult to quantify. Bio-monitoring captures an individual’s total pesticide dose and not dietary intake alone. Although good data exists for typical pesticide residues in foods applying this to an individual’s dietary intake is challenging due to variation in peoples’ behaviour with regard to food purchasing decisions (e.g. cultivation methods, supplier, geographical origin which might influence pesticide residues) and preparation (e.g. washing or peeling of fruit).

1.5 Pesticide exposure estimation methods employed in previous PD studies

Direct measurements of pesticide exposure or absorption are the exception rather than the rule in studies of the health effects of pesticide exposure (Perry et al., 2006). Cross-sectional or cohort studies can measure pesticide exposures in a subset of participating subjects and then use these figures to validate exposure estimates for the cohort as a whole. Such measurements, however, are rarely available for retrospective studies and so PD researchers must rely on estimates of pesticide exposure of varying precision (Brown et al., 2006). Some PD studies have employed simple metrics such as ever/never exposed to pesticides (Chan et al., 1998; Behari et al., 2001) or job title (Tuchsen 2000) while others have relied on self-reported duration and/or frequency of exposure (Semchuk et al., 1992; Wong 1991). Employing a dichotomous measure such as any exposure versus no exposure can lead to exposure misclassification so reducing a study’s statistical power to detect an association where one exists. Self-reported exposure has been shown to be reliable in respect of broad categories of pesticides recall but it is poorer for specific agents (Engel et al., 2001). Some studies have explored use of estimates of exposure to specific pesticides (Hertzman et al., 1994; Liou et al., 1997) with varying results. One study used possession of a pesticide user license as a marker of exposure. (Baldereschi et al., 2003) A more detailed approach to pesticide exposure estimation used a job-exposure matrix modified by subjective exposure estimation techniques (Semple et al., 2004).

1.6 Job-exposure matrix (JEM)

One approach to pesticide exposure estimation is the job-exposure matrix (JEM) which lists a range of occupations on one axis and potential exposures on the other: the cells of the matrix are then populated with estimates of the probability, intensity or frequency of each exposure. The job-exposure matrix is a system for utilising “expert judgement”, exposure measurement data or other information, where some value of exposure is assigned to a job title or industry grouping. Job-exposure matrices have been used in previous studies of farmers and horticulturists. (Young et al., 2004) The JEM is especially useful in situations where a subject’s recollection of their work history is better than their recall of specific exposures, such as pesticides. Some authors have noted that a JEM, developed for another study or population may not be applicable in a new setting or population (van Tongeren et al., 2002). Where a JEM is developed for a specific worksite the level of exposure information is likely greater than is possible for a generic JEM developed for use in population studies (Teschke et al., 2002). One limitation of JEMs is the misclassification of jobs due to large variability of exposure within the job titles (Miligi and Masala, 1991). This is very likely in population-based studies where job titles are sometimes very general and grouped across regions and agricultural facilities. Job titles are often composed of many tasks and there is variability of exposure between those tasks. Commonly interviewer or self administered questionnaires are used to collect data on occupational or non-occupational exposures.

1.7 Task-exposure matrix (TEM)

A further refinement of the JEM, the Task Exposure Matrix (TEM) allows sub-division of a job title into constituent tasks. The TEM aims to reduce the variability of assessments based on a job title by identifying the tasks involved in the job being described and assessing the exposure due to these tasks. Some workers with similar job titles will report different task duties and hence exposures. In farming, for example, an individual would perform different tasks depending on the crop, animal or land area characteristics of that farm. The TEM also has the ability to incorporate changes in tasks over time due to technological advances or alterations in pesticide usage. There is therefore a strong justification for the development of exposure matrices which are based on the individual tasks of a job rather than the job itself i.e. a task-exposure matrix (TEM).

1.8 Background to the study - the Geoparkinson study

SID 5 (Rev. 3/06) of 21

1.8.1 Geoparkinson dataThe Geoparkinson study was conducted between 2000 and 2004. (Dick et al., 2007) This was a multi-centre case-control study of 959 prevalent cases of parkinsonism (of whom 767 had Parkinson’s disease) and 1989 age and gender balanced controls. Subjects were recruited in five participating countries (Scotland, Italy, Sweden, Romania and Malta). An objective of this European Union funded study was to investigate whether exposure to chemicals increased the risk of developing parkinsonism or Parkinson’s disease. Cases were defined as having Parkinson’s disease or parkinsonism using the UK Brain Bank clinical diagnostic criteria (Hughes et al., 1992).

1.8.2 Scottish Geoparkinson dataThe Scottish arm of the study involved 620 people (202 cases and 418 controls), of whom 290 reported pesticide use.

1.8.3 Assessment of exposure in the Geoparkinson study

As part of the assessment of exposure in the Geoparkinson study (Semple et al., 2004), a detailed occupational history was obtained from each participant by interviewer-administered questionnaire, with details of every job held for at least six months. This history included the job title and the start and end dates for that job. For each job there was a series of questions regarding exposure to pesticides. A positive response to one of these questions prompted the administration of an exposure-specific questionnaire to gather more detail on; number of hours per day performing task, working conditions (e.g. application methods), symptoms and possible exposure concentrations. A similar process was performed for non-occupational (hobby) exposures. Where amateur uses of pesticides were indicated for gardening it was assumed that the subject had commenced this hobby at age 40. This approach generated:

A brief description of the tasks and work methods involved in the job; An average duration of exposure to the substance/pesticide/material in terms of hours per day; An average frequency of contact with the material in terms of days per year; An indication of the quantity of the material used.

The purpose of these exposure-specific questionnaires was to gather additional data beyond job titles to assist in task exposure reconstruction.

1.8.4 Magnitude of pesticide exposure

During the analysis of data for the Geoparkinson study the following exposure metrics (Semple et al., 2004) were produced:

Metric 1. The total number of years of exposure to pesticide. Metric 2. Cumulative exposure (CE): derived by multiplying the estimated exposure intensity, the

exposure frequency (number of days per year), a frequency adjustment (to take account of the number of hours per day of exposure), and the exposure duration (number of years of contact within that job). This total exposure figure for each job was expressed in OEL (occupational exposure limit) years, where one OEL year was defined as being equivalent to exposure at the then current UK airborne OEL (HSE 2002) for a pesticide typically used in that task for 8 hours per day for 240 working days per year. The CE was then the sum of these individual values from each job title. The top ten rankings for CE from the Scottish Geoparkinson dataset are shown in Appendix 1.

Metric 3. Average annual intensity (AAI) of exposure: calculated by dividing the CE by the number of years of exposure to that material. Expressed as a fraction of the OEL, it indicates the equivalent average daily exposure level of the worker during employment involving exposure. The ten highest rankings for AAI from the Scottish Geoparkinson dataset are shown in Appendix 2.

Although the majority of pesticide exposures were hobby uses (e.g. applying herbicide to garden paths), the highest AAI’s and CE’s were related to agricultural activities such as crop spraying or sheep dipping.

1.9 Study aims

The main aim of this study was to construct a task-exposure matrix for specific pesticide groups in Scotland covering the period 1945 to 2005.

A secondary aim of the study was to establish proof of concept for the TEM. The task-exposure matrix for pesticide use in Scotland was applied to pesticide exposure information from the Scottish arm of the

SID 5 (Rev. 3/06) of 21

Geoparkinson study. The new exposure estimates generated using the TEM were then used in a re-analysis of the Scottish Geoparkinson dataset.

2.0 Methods2.1 Steps in the development of the task-exposure matrix

The development of the task-exposure matrix involved a number of steps as shown in the flow chart in Appendix 3. First, broad categories of pesticide use were identified from the tasks reported in the Scottish Geoparkinson dataset. A list was then generated of tasks involving pesticide exposure in Scotland. The next step was to determine the typical pesticides used in each task in each decade that it was used (denoted as ‘eras’ in this report). This involved a literature search of published and grey literature and expert interviews. Given the very large number of potential agents we collapsed these into ten groups of pesticides. A pesticide usage database was then created using the task list and the typical pesticide groups employed in those tasks. We also gathered information regarding the concentration of pesticides used in agricultural applications and incorporated this into the pesticide usage database. Information regarding pesticide application methods by decade for each task was compiled. This led to the construction of a matrix with a total of 81 job-tasks further sub-divided into a final list of 248 task-pesticide combinations. Where usage data for a pesticide group indicated use within a given decade an exposure assessment was then performed independently by two occupational hygienists for dermal and inhalational exposures separately.

2.2 Categories of pesticide use

Broad categories of pesticide exposure were identified from the tasks reported in the Scottish Geoparkinson dataset. Analysis of 290 exposure questionnaires from the Geoparkinson dataset indicated that there were three main categories of pesticide exposure: hobby exposure, occupational agricultural (arable or livestock) exposure, and occupational amenity exposure (Table 2). It was decided that these three categories of pesticide use should form the basis of the task list for the Tempest study.

Table 2: Main categories of pesticide exposure identified in the Scottish Geoparkinson data set

Category of exposure* Example of task No. of subjects reporting exposure

Hobby Gardening 237Occupational Agriculture Crop spraying,

sheep dipping62

Occupational Amenity Park maintenance 18

*Subjects may have been exposed to more than one category of exposure.

2.3 Task list: tasks involving pesticide exposure in Scotland

A task list was then generated of tasks involving pesticide exposure in Scotland. The list was initially populated with all pesticide tasks identified in the Scottish Geoparkinson pesticide questionnaires. The task list was then expanded to include other tasks, not reported in the Geoparkinson study but which would have exposed the user to pesticides (Appendix 4). This job and task list was refined by two occupational hygienists (SS and MvT) to produce a finalised task list of 81 individual tasks divided into 8 task groups (table 3).

Table 3: Task codings for the eight task groups

Task Group Code DescriptionAL Agricultural livestockF ForestryQ AquacultureAF Arable farmingG Domestic gardeningAM Amenity useD Domestic indoor

SID 5 (Rev. 3/06) of 21

P Other occupational (not elsewhere classified)

2.4 Determining the typical pesticides used for any given task

The next step was to determine the typical pesticides used for any given task in a given decade in Scotland.

2.4.1 Identifying historical changes in pesticide use

There were a number of limitations in the Scottish Geoparkinson dataset particularly around task descriptions. In some cases these task descriptions were very limited, e.g. “sprayed fields”, but gave no further information regarding the crop treated. This meant that it was not possible to identify the pesticide group used as this was influenced by crop and decade. Although there has been very little change in the hectarage of crops grown in Scotland over the last 60 years, the type of crops has changed considerably (see Appendix 5). Agricultural practices and crops cultivated may change over several years reflecting market conditions: therefore, to establish the probability of which crops were likely to have been treated, sample year data for 1950, 1960, 1970 and 1980 were obtained from the Analytical Services Statistics Branch, Scottish Executive’s Environment and Rural Affairs Department (SEERAD). (D. Rowley, personal communication, 2006) Sample year data for 1990 and 2000 were obtained from Central Science Laboratory pesticide usage reports, available on-line at http://www.csl.gov.uk/index.cfm. This, in the absence of Geoparkinson pesticide questionnaire detail, was then used to identify which crops were probably treated during a given decade (see Appendix 5). The pragmatic approach was taken of selecting a ‘sample’ year for each decade, or ‘era’, to avoid the dataset becoming so large as to be unmanageable. Although permanent and temporary grassland constitute the majority of cultivated land mass, pesticide applications to them are rare (average 0.10 kg.AI/ha), (Appendix 5). In view of this grassland treatments were excluded.

2.5 Literature review

To identify information on pesticide usage in Scotland a comprehensive search strategy for identifying English language literature was conducted using Medline, Embase, Agricola and Scopus databases. The keywords employed were “Pesticides”, “Insecticides”, “Fungicides”, Herbicides”, “Farming”, “Sheep”, “Salmon” “Amenity”, “Gardening”, “Application”, “Exposure”, “Patterns”, and “Change”. Limited literature was found. A search of secondary sources and grey literature was also performed. This included government reports, surveys, periodicals, bulletins and codes of practice (n= 187). Where gaps in literature were identified, relevant experts were approached for their opinion. These interviews were conducted by the administration of semi-structured questions, either face to face or by telephone. Where possible the opinion of more than one expert was sought: the list of experts who contributed is given in Appendix 6.

2.6 Pesticide groups

Pesticides (insecticides, herbicides and fungicides) which had been identified for use in tasks were then coded using the following classification system:

2.6.1 Insecticides:We grouped insecticides into four main types: organochlorines (OC); organophosphates (OP); carbamates; and others. The ‘other’ pesticide grouping consisted primarily of botanicals and pyrethroids. Full details of these categories with examples are provided in table 4.

2.6.2 Herbicides There are over 1000 herbicides in existence. Unlike the insecticides which can be broadly classified into four groups based on their chemical structure, the coding of herbicides presented a greater challenge. On the advice of a pesticide use expert, herbicides were classified by their mode of action, using the HRAC system (Herbicide Resistance Action Committee, 2005). This method of classification has the advantage of having only 24 groups as compared to the Weed Science Society of America classification (Mallory-Smith and Retzinger, 2003) which has 27. Both systems are similar. However, to avoid an unmanageably large number of herbicide categories in subsequent analyses, with consequent reduction in statistical power, herbicides were further limited to four groups. We selected three commonly used HRAC herbicide groups (C1, D, O) and the remaining 21 groups were collapsed into a fourth group, ‘Other’ herbicides as shown in Table 4. Group C1 included atrazine which has previously been shown to be neurotoxic in rats (Rodriguez et al. 2005). Group D included paraquat which has been linked to Parkinson’s disease (Hertzman, 1990, Liou et al. 1997).

SID 5 (Rev. 3/06) of 21


2.6.3 FungicidesFungicides were categorised into carbamates (which includes dithiocarbamates) and ‘other’. Dithiocarbamates are the largest group of fungicides in terms of quantity sold. In addition they are one of the older groups of fungicides and their continued use is a reflection of their effectiveness and relative low cost. Examples of dithiocarbamates include mancozeb and maneb: the latter agent has been implicated in the development of parkinsonism in one case report (Meco et al., 1994).

Table 4: Pesticide groups included in the study

Pesticide category Examples in this group

InsecticidesOrganochlorines (OC) Aldrin, DDT, dieldrin, chlordane

Organophosphates (OP) Malathion, parathion, diazinon, methylparathion

Carbamates Carbaryl, carbofuran, methylcarbamate

Other insecticides Permethrin, lambda-cyhalothrin, rotenone, nicotineHerbicidesC1 Simazine, bromacil, atrazine, lenacil, terbacil

D Diquat, paraquatO 2,4-D, dichlorprop, MCPA, mecoprop, fluroxypyr, clopyralidOther herbicides Glyphosate, linuron, diuron, dinoseb.

Fungicides

Carbamate fungicides Mancozeb and maneb

Other fungicides Copper sulphate, benzimidazole

2.7 Pesticide usage database

The task list and pesticide groups were merged to produce a pesticide usage database for each task based on the likely pesticide group used in a given decade for a specified task. To enable accurate exposure estimations the concentration of pesticides used in applications was determined so far as possible. The wealth of information in Pesticide Usage Reports for Scotland published on-line by the Central Science Laboratory facilitated the compilation of detailed data for agriculture from the 1960’s onwards. This information was then incorporated into the pesticide usage database. Information concerning the concentration of pesticide for other applications (hobby and amenity use) could not be obtained. Details from the pesticide usage database were then used to populate the task-pesticide matrix. For each era we assigned a probability of a particular pesticide group being used for the given task (total probability for all pesticide groups always summed to 100%). This is presented in Appendix 7 and includes such pesticide uses as fly control in amenity and domestic settings, wet rot and woodworm treatments (but not residence in a treated home) and insecticide treatments in carpet manufacture.

2.8 Pesticide application methods

Information regarding methods of pesticide application by decade (i.e. era) for each task was compiled as a result of literature searches and expert advice. Details of methods of application are shown in Appendix 8.

2.9 Exposure reconstruction

The intensity of exposure for all tasks across the seven eras was then assessed independently by two occupational hygienists (SS and MvT) with expertise in exposure reconstruction techniques. The assessments of intensity were carried out on a four point semi-quantitative scale (none, low, medium and high) for exposure by both inhalation and dermal exposure routes. The range of exposures for these categories is presented in tables 5 and 6. The estimates of the intensity of exposure were predominantly made based upon information on: active ingredients, amount used per hectare, likely application technique and assumed use of personal protective equipment, including protective clothing and respiratory protection.

SID 5 (Rev. 3/06) of 21

Table 5: inhalation exposure guidance for pesticides

Inhalation exposure guidance

Airborne concentrations while performing that taskAssessment Description

0 No exposure1 0-5% WEL* very low2 6-25% WEL low3 26-100% WEL medium4 >100% WEL high

* WEL – Workplace exposure limits published by HSE in EH40/2005 (ISBN 0717629775).

Table 6: dermal exposure guidance for pesticides (Cattani et al., 2001)

Dermal exposure guidance

Active ingredient skin loading (whole body) rate while performing that taskAssessment Description

0 No exposure - also used when material does not have an Sk notation*1 0-1 mg/hr very low2 1-10 mg/hr low3 11-100 mg/hr medium4 >100 mg/hr high

* Sk notation – ‘…some substances have the ability to penetrate intact skin and become absorbed into the body, thus contributing to systemic toxicity; these substances are marked … with an ‘Sk’ notation’. Workplace exposure limits EH40/2005, page 43, HSE 2005.

The two hygienists’ independent assessments agreed completely on 52% (n=875) of assessments, with a further 41% (n=698) differing by just 1 banding. 8% (n=132) of assessments differed by 2 or 3 bandings. No assessments were 4 bandings apart. Where there was a difference of more than 1 banding (n=132) a consensus was reached between the assessors. This produced a TEM that had complete agreement in 53% (n=893) of assessments and +/-1 banding for the remaining 47% (n=812) of assessments. Our final agreed TEM used the highest banding from the assessors where there remained a difference of 1 banding and these maximum values were employed for statistical analysis.

Finalised inhalation and dermal exposures were distributed as shown in table 7. Note that the maximum values were employed for statistical analysis.

Table 7: showing the distribution of exposure intensity, as assessed by each hygienist, and the maximum values used as the finalised values for inhalation and dermal exposures

Summary exposure dataIntensity SS % MVT % Max. %

Dermal 0 253 30 254 30 253 301 62 7 98 12 33 42 225 27 152 18 188 223 227 27 190 22 192 234 79 9 152 18 180 21

Total 846 846 846

Inhaled 0 40 5 0 0 0 01 250 30 418 49 245 292 369 44 200 24 301 363 132 16 181 21 226 27

SID 5 (Rev. 3/06) of 21

4 55 7 47 6 74 9Total 846 846 846

2.10 Temporal trends in pesticide exposures

Control of pesticide exposures has improved across the seven decades that the TEM spans. In addition to changes in pesticide type there have been significant changes to application methods, first driven by efficiency in application of what can be an expensive material and more recently by concerns relating to human and environmental health. Evidence from analyses of the HSE National Exposure Database indicates that there has been a decrease in exposure levels of most chemicals (not specifically pesticides) between 2-10% per year (Creely et al., 2006: HSE RR460). Our exposure intensity estimates reflect these changes. For example, the percentage of those assessments in the highest two exposure intensity bandings (3 and 4) falls from 54% of all assessments in the 1950 era to 1% in the 2000 era.

2.11 Assumptions made in TEM for the eight main tasks across the seven eras

In addition to the temporal trends of decreasing exposure discussed in section 2.10 above we also made a number of assumptions concerning application methods, the use of personal protective equipment and personal knowledge of health hazards that would influence worker behaviour and consequent exposure levels. In general there was a shift away from hand-held application methods in the arable farming sector over time with increasing field size and use of tractor mounted booms to deliver pesticides. Ventilated tractor cabins with air filtration have become increasingly wide-spread. Similarly, protective clothing (gloves, overalls, boots) became the norm for those involved in applying large quantities of pesticide in the arable, livestock and amenity sectors through the 1980s to the present day. Application methods have shown least change in the hobby sector although formulations have changed, concentrations have decreased and the potential for dermal uptake has fallen.

2.12 Application of TEM to occupational histories

The TEM contained coded exposure intensities and proportions of time exposed for 248 tasks1 assessed, indexed by task, the seven eras 1940s – 2000s, and nine pesticide classes assessed. Individual histories contained entries for tasks, with start and finish dates. To calculate exposures, it was necessary to allocate exposures to the times in tasks from histories, and accumulate these assessments. In the TEM, each entry for intensity of pesticide exposure was coded 0-4, representing ranges. These were assigned nominal mid-range values (Table 8).

Table 8: assigned nominal mid-range values for intensity of pesticide exposure

Level code

Exposure routeInhalation Dermal

0 0 01 2.5 0.52 15 5.53 62.5 554 250 500

Inhalation values are a percentage of the airborne UK WEL for that pesticide. Dermal values are the active ingredient (AI) skin loading (whole body) rate while performing that task (mg/hr).

For each task in an individual exposure history, start and finish dates were used to calculate the amount of contributing time in each decade. For each decade, time was multiplied by mid-range intensity inhalation intensity and proportion, and these were summed over all tasks and decades to produce an individual cumulative inhalation exposure to each class of pesticide. The same calculation was repeated for individual dermal exposures to the classes, this time using the mid-range dermal intensities and proportions (Table 6). For both inhalation and dermal exposures, Average Annual Intensities (AAIs) were estimated by dividing the cumulative exposures by the total contributing time in years. Cumulative exposures for all pesticides were also calculated, by summing across classes. All calculations were performed by reading MS Excel spreadsheets into the statistical package GenStat (Payne, 2004), and using its facilities for manipulating tables indexed by multiple factors.

2.13 Statistical methods

To identify and quantify any exposure-response relationship between Parkinson’s disease and pesticide exposure, the data were analysed by logistic regression methods (Cox and Snell, 1989), using the statistical software package GenStat (Payne 2004). Models included adjustment for binary coefficients representing

SID 5 (Rev. 3/06) of 21

smoking history (questionnaire Q13: ever versus never) and family history of Parkinson’s disease (Q27G), both of which were highly significant in all models. Other adjustments used in the original Geoparkinson analysis (Dick et al., 2007) were not significant, and were not employed.

With smoking and family history of PD in the logistic model predictor, each of the exposure variables was tested for the significance of its inclusion, evaluating the change in deviance as a chi-square statistic. Cumulative inhalation and dermal exposures to each of the pesticide types, and to the total of all pesticides combined, were each fitted in turn. The analyses were repeated with exposures recalculated as Average Annual Intensity (AAI). Carbamate insecticides and carbamate fungicides were combined into a single group (carbamates).

1. Two tasks identified were rat/mouse control and rabbit control using gas/warfarin. As the mechanism of action of these agents was so dis-similar to other pesticides they were removed from the JEM giving 248 tasks in the final TEM.

3.0 Results3.1 Application of TEM to occupational histories

GEOP-derived cumulative exposure

0 200 400 600 800 1000 1200 1400 1600TEM

PE

ST-

deriv

ed in

hala

tion

cum

ulat

ive

expo

sure

0

1000

2000

3000

4000

5000

0.01 0.1 1 10 100 1000 10000

0.01

0.1

1

10

100

1000

10000

Figure 1: Comparison of TEMPEST-derived inhalation exposure estimates with estimates used in the Geoparkinson study; (a) measurement scale, (b) log scale.

Figure 1(a) compares the individual cumulative inhalation exposure estimates based on the final TEM values with the original Geoparkinson pesticide exposure estimates (which combined inhalation and dermal exposures: see section 4.3) based on the individual assessments reported by Semple (Semple et al., 2004). Outside the upper and right axes the distributions of the individual values are shown in rug-plots, with an outward-facing tick for each value. There are a few points where one or other estimate is high and agreement seems poor. Figure 1(b) shows the same data, on a log scale (so omitting all zero exposures), again with rug-plots. Here the impression is much more of a systematic trend between the two measures, with a fair spread around a common ratio.

SID 5 (Rev. 3/06) of 21

GEOP-derived cumulative exposure

0 200 400 600 800 1000 1200 1400 1600

TEM

PE

ST-

deriv

ed d

erm

al c

umul

ativ

e ex

posu

re

0

2000

4000

6000

8000

10000

12000

14000

0.01 0.1 1 10 100 1000 10000

1e-2

1e-1

1e+0

1e+1

1e+2

1e+3

1e+4

1e+5

(a) (b)

Figure 2: Comparison of TEMPEST-derived dermal exposure estimates with estimates used in the Geoparkinson study; (a) measurement scale, (b) log scale.

As discussed in section 4.2, the cumulative exposure assessments used in the Geoparkinson study included consideration of inhalation and dermal exposure routes, while TEMPEST did not attempt to combine the uptake from these routes in to a unified metric. Figure 2 (above) shows the comparison of TEMPEST assessments of dermal skin loadings and the Geoparkinson cumulative exposure (inhaled and dermal uptake expressed as inhalation exposure equivalents).

3.2 Re-analysis of Geoparkinson data focussing on maximum exposure analysis

Table 9 shows the 418 controls, 170 Parkinson’s disease cases, and 32 parkinsonism cases, split by whether or not their exposure to each group of pesticides was estimated as greater than zero. It is clear that analyses of the PD cases, as defined by the UK Parkinson’s Disease Society Brain Bank clinical diagnostic criteria, (Hughes et al., 1992) have much greater power than analyses of the parkinsonism cases to show a relationship with exposure, if one exists.

Table 9: showing numbers exposed to each pesticide group by case-control status

SID 5 (Rev. 3/06) of 21

Case-control statusControl Parkinson’s

disease ParkinsonismExposure > 0 No Yes No Yes No Yes

Pesticide group* Insecticides OC 327 91 119 51 27 5

Insecticides OP 329 89 117 53 27 5Carbamates 335 83 128 42 25 7Insecticides ‘other’ 326 92 121 49 27 5Herbicides C1 307 111 119 51 20 12Herbicides D 296 122 116 54 20 12Herbicides O 299 119 111 59 18 14Herbicides ‘other’ 295 123 116 54 20 12Fungicides ‘other’ 375 43 152 18 24 8

All pesticides 240 178 84 86 12 20

* see section 2.6 for an explanation of the pesticide groups.

Logistic regression analysis of the Parkinson’s disease cases adjusted for smoking and for familial history of the disease, and then added each of the exposures, as continuous variables, to the baseline regression model. Goodness of fit in a logistic regression model is measured by a quantity known as 'deviance' (essentially a weighted sum of squares). The extent to which including a regression predictor term, such as an exposure variable, improves the model is quantified by comparing the deviance achieved with and without that term in the model. Table 10 shows the change in model deviance achieved by addition of each inhalation exposure variable while Table 11 shows the change in model deviance by addition of each dermal exposure variable.

Table 10: showing change in model deviance achieved by the addition of each inhalation exposure variable.


The statistical significance of deviance change statistics is tested by treating them as chi-square statistics. Here each has one degree of freedom, and the critical value for significance is 5% i.e. 3.84. None of the statistics in either tables 10 and 11 comes close to this critical value, and we conclude that there is no evidence for a relationship between risk of PD and exposure to any or all of the pesticides.

Table 11: showing change in model deviance achieved by the addition of each dermal exposure variable.

Pesticide group* Deviance changeInsecticides OC 0.1Insecticides OP 0.3

SID 5 (Rev. 3/06) of 21

Pesticide group* Deviance changeInsecticides OC 0.2Insecticides OP 2.3Carbamates 0.6Insecticides ‘other’ 0.3Herbicides C1 0.4Herbicides D 0.0Herbicides O 0.8Herbicides ‘other’ 0.0Fungicides ‘other’ 1.0

All pesticides 0.4

Carbamates 0.3Insecticides ‘other’ 3.7Herbicides C1 0.0Herbicides D 0.3Herbicides O 1.3Herbicides ‘other’ 0.0Fungicides ‘other’ 2.3

All pesticides 0.4


Similar analyses were carried out using exposures estimated as average annual intensities (AAIs). The results again showed no strong evidence of an effect on PD risk from any or all of the inhalation or dermal exposures.

4.0 Discussion4.1 Task-exposure matrix: strengths and limitations

The task-exposure matrix for pesticide usage in Scotland has a number of strengths and limitations. It is acknowledged that the information on likely agents employed in tasks in earlier decades (1940s-1960s) was limited and in places was wholly based on ‘expert’ reports. However for later decades a greater volume of published material was available on which to base estimates of exposure. In addition work practices and pesticide selection may have varied across the agricultural regions of Scotland. Thus expert reports, while the best available evidence for some sectors, may not be wholly reliable. The construction of the matrix required that expert decisions be made as to the most likely agent employed, the likely application method and the probable quantities used. While these decisions were informed, so far as possible, by available data on pesticide usage, application methods and crops grown, expert assessments, in particular for earlier decades, may have resulted in some degree of exposure misclassification in the TEM. Strengths of this approach are that explicit, pesticide class specific, (semi-)quantitative exposure estimates are provided in the matrix and, subject to available funding, could be validated.

It should be emphasised that the task-exposure matrix was constructed for application in Scotland and therefore may not be directly applicable to other countries with different work practices, crops and pests.

The development of the task-exposure matrix has highlighted the information that should be gathered during any future case-control studies. The TEM can be used as a template for developing a questionnaire to collect data that will allow more targeted and precise exposure reconstructions.

4.2 Comparison of Geoparkinson versus TEMPEST exposure metrics

It is important to recognise the differences between the exposure assessment methods used in the original Geoparkinson study (Dick et al., 2007) and those employed by applying the TEM for pesticides generated as part of the present work. The Geoparkinson study used a simple JEM to quantify exposure to any pesticide and then allowed modification of this base estimate using any relevant data that was extracted from the exposure interview. (Semple et al., 2004) These ‘exposure modification factors’ included descriptions of the use of personal protective equipment, details of the quantities of pesticide used and any report of symptoms related to pesticide use. The exposure estimates generated in the present study were more detailed in that they assigned particular pesticide types to a task and then gave exposure intensities for each pesticide class for both dermal and inhalation routes. Importantly, however, they did not take into account any individual data from the original interview to modify the TEM-derived exposure values. This would have required further work in terms of extraction and coding the data from the original questionnaires. While most of the original assessments did not provide detailed data on ‘exposure modification factors’, the small percentage that did usually provided important evidence of either very good control or extremely high exposure levels. For those few cases the differences between the TEM-derived estimate and the original exposure assessment may be relatively large.

Note that the initial Geoparkinson assessments produced a combined exposure intensity that took in to consideration exposure by dermal and inhalation routes and expressed the intensity as an inhalation exposure equivalent value. The present study does not do this and produces dermal exposure loadings separate from inhalation exposure concentrations. Methods of estimating the uptake of given dermal exposures require

SID 5 (Rev. 3/06) of 21

information on frequency of loading, washing and changes of clothing. Combining inhalation and dermal exposures for different pesticide classes was beyond the remit of this work.

4.3 Reanalysis of the Geoparkinson dataset: proof of concept

Dick et al. (2007) reported results for the five countries participating in the Geoparkinson study. Analysis of the Scottish data alone, treating the previous estimate of lifetime exposure to all pesticides as a continuous variable, gave a relative risk for PD of 1.003 per OEL.yr, just significant at 5%. Our analyses with the revised TEM-based exposure estimates, for all groups combined, fell well short of even that level of significance. The difference in results could be due to recall bias, chance or perhaps to the more generic nature of the exposure assessment in the new work, leading to more individual exposure misclassification. Note that we did not combine dermal and inhalational exposures (as was done for the original Geoparkinson analysis) thereby possibly reducing the power to detect a difference between cases and controls, should one exist. Table 2 shows that, of all the exposures listed by subjects, only 80 occurred in an occupational setting, with the remainder occurring during domestic and/or hobby applications, with very much lower levels of exposures. The power of analyses to relate PD risk to total pesticide exposure in this population was therefore always low. In the case of exposures to individual groups of pesticides, the numbers were even smaller, and the power correspondingly lower.

4.4 Potential applications of the task-exposure matrix

The task-exposure matrix if used in association with a questionnaire designed to obtain key data on tasks undertaken and factors influencing exposure (era, duration of use, method of application, hectares treated, exposure controls etc.) could be employed in a cohort study of Parkinson’s disease and pesticide usage in occupationally exposed groups such as agricultural workers or pest control operatives. While technically feasible such a study would be expensive as a very large cohort numbering many thousands of workers would be necessary to generate sufficient cases of PD for study. For example, the Agricultural Health Study, a US prospective cohort study of pesticide applicators and their spouses, involves more than 89,000 people http://www.aghealth.org/index.html. A more efficient design for the study of relatively rare diseases is a case control study: the TEM could be used in a case-control study although the sample size would need to be considerably greater than that of the Scottish arm of the Geoparkinson study to achieve adequate power. Were similar task-exposure matrices for pesticide usage developed for other countries then this would facilitate larger, and so more powerful, international studies of PD and pesticide exposures.

4.5 Future work

This project has successfully developed a task-exposure matrix for specific pesticide categories in Scotland for the period 1945-2005 and then applied this matrix to an existing dataset drawn from the Geoparkinson study (Dick et al., 2007) to establish proof of concept. The outcome of the re-analysis was not unexpected given the sample size: there was no clear evidence of an association between any pesticide group and Parkinson’s disease. This likely reflects the relatively small sample studied (418 controls, 170 cases of PD). One area for future work is the validation of the task-exposure matrix against measured exposures in field conditions. Conceptually the task-exposure matrix for pesticides developed in this project offers improved pesticide exposure estimation and could be employed in any future case-control study of PD and pesticide exposure in Scotland.

References

The Agricultural Health Study http://www.aghealth.org/index.html

Baldereschi M, Di Carlo A, Vanni P, Ghetti A, Carbonin P, Amaducci L, Inzitari D. 2003. Lifestyle-related risk

factors for Parkinson’s disease: a population-based study. Acta Neurol Scand 108:239-244.

Behari M, Srivastava AK, Das RR, Pandey RM. 2001. Risk-factors of Parkinson’s disease in Indian patients. J

Neurol Sci 190:49-55.

Brown TP, Rumsby PC, Capelton AC, Rushton L, Levy LS. 2006. Pesticides and Parkinson’s disease – is there a

link? Environ Health Perspect 114:156-64.

SID 5 (Rev. 3/06) of 21

http://www.aghealth.org/index.html

http://www.aghealth.org/index.html

Buchanan D, Pilkington A, Sewell C, Tannahill SN, Kidd MW, Cherrie B, Hurley JF. 2001. Estimation of

cumulative exposure to organophosphate sheep dips in a study of chronic neurological health effects among

United Kingdom sheep dippers. Occup Environ Med 58:694-701

Cattani M, Cena K, Edwards J, Pisaniello D. Potential dermal and inhalation exposure to chlorpyrifos in

Australian pesticide workers. Ann Occup Hyg. 2001 Jun;45(4):299-308.

Central Science Laboratory. Pesticide usage survey reports.

http://www.csl.gov.uk/newsAndResources/resourceLibrary/articles/puskm/index.cfm

Chan DKY, Woo J, Ho SC, Pang CP, Law LK, Ng PW, Hung WT, Kwok T, Hui E, Orr K, Leung MF, Kay R. 1998.

Genetic and environmental risk factors for Parkinson’s disease in a Chinese population. J Neurol Neurosurg

Psychiatry 65:781-784.

Cox DR, Snell EJ.1989. Analysis of binary data (2nd edition). London: Chapman and Hall.

Creely KS, van Tongeren M, While D, Soutar AJ, Tickner J, Agostini M, de Vocht F, Kromhout H, Graham M,

Bolton A, Cowie H, Cherrie JW. 2006. Trends in inhalation exposure: mid 1980’s till present. HSE Research

Report 460. http://www.hse.gov.uk/research/rrpdf/rr460.pdf

Dick FD, De Palma G, Ahmadi A, Scott NW, Prescott GJ, Bennett J, Semple S, Dick S, Counsell C, Mozzoni P,

Haites N, Bezzina Wettinger S, Mutti A, Otelea M, Seaton A, Söderkvist P, Felice A. 2007. Environmental risk

factors for Parkinson's disease and parkinsonism: the Geoparkinson study. Occup Environ Med (published online

1st March 2007).

Engel LS, Seixas NS, Keifer MC, Longstreth WT Jr, Checkoway H. 2001. Validity study of self-reported pesticide

exposure among orchardists. J Expo Anal Environ Epidemiol 11:359-68.

General Register Office for Scotland 2002. Scotland’s Census 2001. 2001 Population Report Scotland. General

Register Office for Scotland, Edinburgh. ISBN 1-874451-66-4.

Herbicide Resistance Action Committee (HRAC). 2005. Classification of Herbicides According to Mode of Action

http://www.plantprotection.org/HRAC/

Hertzman C, Wiens M, Bowering D, Snow B, Calne D. 1990. Parkinson’s disease: a case-control study of

occupational and environmental risk factors. Am J Ind Med 17: 349-355.

Hertzman C, Wiens M, Snow B, Kelly S, Calne D. 1994. A case-control study of Parkinson’s disease in a

horticultural region of British Columbia. Mov Disord 9:69-75.

Hughes AJ, Daniel SE, Kilford L, Lees AJ. 1992. Accuracy of clinical diagnosis of idiopathic Parkinson's disease:

a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 55:181-184.

SID 5 (Rev. 3/06) of 21

HSE. 2002. EH40/2002 Occupational exposure limits. EH40 2002; 7-28.

HSE. 2005 Workplace exposure limits. EH40/2005.

Jimenez-Jimenez FJ, Mateo D, Gimenez-Roldan S. 1992. Exposure to well water and pesticides in Parkinson's

disease: a case-control study in the Madrid area. Mov Disord 7:149-52

Koller W, Vetere-Overfield B, Gray C, Alexander C, Chin T, Dolezal J, Hassanein R, Tanner C. 1990.

Environmental risk factors in Parkinson’s disease. Neurology 40:1218-1221.

Liou HH, Tsai MC, Chen CJ , Jeng CS, Chang JC, Chen SY, Chen RC.1997. Environmental risk factors and

Parkinson's disease: a case-control study in Taiwan. Neurology: 48:1583-1588.

Mallory-Smith CA, Retzinger EJ. 2003. Revised classification of herbicides by site of action for weed resistance

management strategies. Weed Technology 17:605–619.

Meco G, Bonifati V, Vanacore N, Fabrizio E. 1994. Parkinsonism after chronic exposure to the fungicide maneb

(manganese ethylene-bis-dithiocarbamate). Scand J Work Environ Health: 20:301-5.

Payne RW (ed). 2004. The guide to GenStat Release 8. Rothamsted: Lawes Agricultural Trust.

Perry MJ, Marbella A, Layde PM. 2006. Nonpersistent pesticide exposure self-report versus biomonitoring in farm

pesticide applicators. Ann Epidemiol: 16:701-7.

Rodriguez VM, Thiruchelvam M, Cory-Slechta, DA. 2005. Sustained Exposure to the Widely Used Herbicide

Atrazine: Altered Function and Loss of Neurons in Brain Monoamine Systems EHP:113: 6.

Semchuk KM, Love EJ, Lee RG. 1992. Parkinson’s disease and exposure to agricultural work and pesticide

chemicals. Neurology; 42:1328-1335.

Semple SE, Dick F, Cherrie JW. 2004. Exposure assessment for a population-based case-control study

combining a job-exposure matrix with interview data. Scand J Work Environ Health: 30:241-248.

Taylor CA, Saint-Hilaire MH, Cupples LA, Thomas CA, Burchard AE, Feldman RG, Myers RH. 1999.

Environmental, medical, and family history risk factors for Parkinson's disease: a New England-based case

control study. Am J Med Genet 88:742-9.

Teschke K, Olshan Af, Daniels JL, De Roos AJ, Parks CG, Schultz, Vaughan TL. 2002. Occupational exposure

assessment in case-control studies: opportunities for improvement. Occup Environ Med ;59:575-594.

Tuchsen F, Jensen AA. 2000. Agricultural work and the risk of Parkinson's disease in Denmark, 1981-1993.

Scand J Work Environ Health. 26:359-62.

SID 5 (Rev. 3/06) of 21

Van Tongeren M, Nieuwenhuijsen MJ, Gardiner K, Armstrong B, Vrijheid M, Dolk H, Botting B. 2002. A job-

exposure matrix for potential endocrine-disrupting chemicals developed for a study into the association between

maternal occupational exposure and hypospadias. Ann Occup Hyg. 46:465-77.

Watkins, S. 2002. World Non-Agricultural Pesticide Markets: 2nd ed. Agrow reports, PJB Publications Ltd.,

Richmond, Surrey, UK.

Wong GF, Gray CS, Hassanein RS, Koller WC. 1991. Environmental risk factors in siblings with Parkinson’s

disease. Arch Neurol: 48:287-289.

Young HA, Mills PK, Riordan D, Cress C. 2004. Use of a crop and job specific exposure matrix for estimating

cumulative exposures to triazine herbicides among females in a case-control study in the Central Valley of

California. Occ Env Med 61:945-951.

References to published material9. This section should be used to record links (hypertext links where possible) or references to other

published material generated by, or relating to this project.

SID 5 (Rev. 3/06) of 21

general enquiries on this form should be made...

Documents