Abstract

The authors investigated the hypothesis that exposure to pesticides could be related to central nervous system disorders in a prospective cohort study of 1,507 French elderly (1992–1998). Lower cognitive performance was observed in subjects who had been occupationally exposed to pesticides. In men, the relative risks of developing Parkinson’s disease and Alzheimer’s disease for occupational exposure assessed by a job exposure matrix were 5.63 (95% confidence interval: 1.47, 21.58) and 2.39 (95% confidence interval: 1.02, 5.63), respectively, after confounding factors were taken into account. No association was found with having a primary job in agriculture or with environmental pesticide exposure, nor was an association found in women. These results suggest the presence of neurologic impairments in elderly persons who were exposed occupationally to pesticides.

Received for publication January 31, 2002; accepted for publication September 11, 2002.

Mortality due to neurodegenerative diseases is expected to increase 119–231 percent in the United States between 1990 and 2040, depending on the population model used (13). A recent meta-analysis on the relation between pesticide exposure and Parkinson’s disease revealed a positive association (4), even if no dose-response relation has been established and no specific type of pesticide identified. Other studies have found delayed cognitive impairments in relation to pesticide exposure (58), which could be predictive of dementia before the clinical diagnosis of disease (9). However, few studies have dealt with the relation between Alzheimer’s disease and pesticide exposure (1012). The aim of the present study was to search for an association between lifelong cumulative exposure to pesticides and neurodegenerative diseases in an elderly cohort.

MATERIALS AND METHODS

Study population

The study population was that of the PAQUID Study, a cohort study involving elderly people aged 65 years or older at inclusion in 1987 who were living at home or in an institution in Gironde, southwestern France. The methodology of the PAQUID Study has been described elsewhere (13). Figure 1 presents the dates of each step, the numbers of subjects under follow-up, and the times of collection of the data used in our analysis.

Assessment of pesticide exposure

Occupational exposure

Detailed occupational histories were generated for all subjects from specific questionnaires that had been filled in during a face-to-face interview at the 5-year follow-up. All jobs were coded using the classification of the Institut National de la Statistique et des Etudes Economiques (14). The likelihood of and level of exposure to pesticides (including insecticides, herbicides, and fungicides) with reference to the job codes were assessed by a panel of six experts blinded to neurologic status, and the median of the experts’ judgments was used. The assessment finally provided a job exposure matrix (table 1). Experts considered that workers on and owners of small farms had experienced higher exposure, since they had mixed and sprayed pesticides themselves using older and less efficient equipment. Cumulative occupational exposure to pesticides was calculated as the sum of exposures incurred in all calendar periods, exposure itself being the product of the period duration and the level indicated by the job exposure matrix. Quartiles of the distribution of the cumulative index values were considered.

Environmental exposure

Since the median duration of residence in the same district was 40 years (interquartile range, 20–63 years), we hypothesized that place of residence at baseline was an indicator for environmental exposure to pesticides, and we established two variables: rural residency and residency in a district with vineyards.

Rural residency. The classification of the Institut National de la Statistique et des Etudes Economiques was used to code the place of residence at baseline. According to this classification, the definition of an “urban area” relied on the continuity of the housing (buildings less than 200 m apart) and the gathering of 2,000 inhabitants or more.

Residency in a district with vineyards. The proportion of surface area planted with vineyards was calculated for each district of residency at baseline. We used 5 percent coverage as the threshold for considering that a district was planted with vineyards.

Neurologic assessment

Cognitive impairment

Neuropsychological tests on standardized questionnaires were administered at each subject’s home by psychologists who were supervised by a researcher in neuropsychology and were not aware of the study hypothesis. We considered here results from the Mini-Mental State Examination (MMSE), which is the sum of subscores that measure orientation to time and place, the recording of three words, language, and visual construction (15). Depression, which influences neuropsychological performance, was evaluated with the French version of the Center for Epidemiologic Studies Depression Scale (16).

Parkinson’s and Alzheimer’s diseases

A simple algorithm based on a standardized questionnaire classified the subjects as either suspected or not suspected of having dementia. Thereafter, subjects who met the Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised, criteria for dementia were evaluated by a neurologist, who filled in information on the National Institute of Neurological and Communicative Disorders and Stroke–Alzheimer’s and Related Disorders Association criteria and the Hachinski score (17, 18). Cases were definitively classified by considering the results of jointly available complementary examinations. At baseline screening and 5-year follow-up, diagnosis of Parkinson’s disease was ascertained by a two-phase design (19). In the first stage, the following two validated questions were used to screen for Parkinson’s disease: “Do your arms or legs shake at rest?”; “Do you experience slowness or stiffness in your movements?”. All subjects who gave positive answers to both questions and/or were taking anti-Parkinson’s disease drugs (including anticholinergic medications, dopamine agonists, and monoamine oxidase inhibitors) were visited at home by a trained neurologist for confirmation or exclusion of a diagnosis of Parkinson’s disease. At 8 and 10 years of follow-up, the possibility of Parkinson’s disease was explored solely by the question, “Do you have Parkinson’s disease?”.

Other variables

The questionnaire also collected information on history of smoking and sociodemographic factors: age, sex, and educational level considered in two classes (whether or not a subject had received the “Certificat d’Études,” a primary school diploma, which was previously found in PAQUID to be the best educational threshold for predicting cognitive impairment (20)).

Analysis

Analyses were performed separately in men and women for the effect of each occupational or environmental pesticide exposure. First, because information on occupational exposure was collected at the 5-year follow-up, MMSE results obtained at that time were compared according to past pesticide exposure to explore the cross-sectional association with general cognitive functioning. Since there was no normative reference value in our population, the MMSE threshold used was the 25th percentile of the distribution of the scores. Potentially confounding factors associated with MMSE performance in univariate analysis with p values less than 0.25 were retained in the multivariate models.

We also explored the association between each pesticide exposure variable and incident cases of diagnosed Parkinson’s disease or Alzheimer’s disease between 5-year follow-up and 10-year follow-up, based on the Cox proportional hazard models in which the time scale was the age of the subjects. Thus, age was not entered as an explanatory variable in the model. Indeed, several authors have advocated that in the study of age-associated diseases, the appropriate time scale for survival models is age rather than time since the baseline survey (21, 22). We used a Cox model with delayed entry to take into account the left-truncation process (subjects entered the sample only if they had not developed a neurodegenerative disease prior to their age at entry). Dates of Alzheimer’s and Parkinson’s disease occurrence were estimated to lie at the center of the interval between the follow-up period in which the disease was observed and the follow-up period in which the subject was considered Alzheimer’s and Parkinson’s disease-free.

RESULTS

Study population

The sample was initially composed of 1,122 men (40.2 percent) and 1,670 women. At the 5-year follow-up point, 623 subjects had died, 614 refused to be visited, and 48 could not be contacted, while 1,507 were still alive and could be visited. They were predominantly women (61.2 percent) (table 2). The mean age was 78.6 years (standard deviation, 6.3). Most of the subjects were office employees (37.0 percent) or blue-collar workers (21.3 percent). A high level of education was more represented than a low level (72.6 percent).

Pesticide exposure assessment

Nineteen job titles were estimated to involve nonnull pesticide exposure (table 1), corresponding to 320 subjects (21.2 percent). A cumulative exposure index could be calculated for 228 subjects (71.3 percent). In others, dates of the ending or beginning of jobs were missing. The median exposure duration was 28 years, and the median delay since the end of exposure was 20 years. Having a primary job in agriculture was reported by 173 of the subjects (11.5 percent). One subject out of four was living in a rural district (26.5 percent), and 414 were living in a district planted with vineyards. The median duration of residence at baseline was 54.5 years for subjects with Alzheimer’s disease and 47 years for subjects with Parkinson’s disease.

Cross-sectional analysis of MMSE performance at 5-year follow-up

Performances on the MMSE were significantly associated with educational level, age, and depressive symptoms in both men and women. Each additional year of age increased the risk of scoring low by 13 percent. A lower educational level was a risk factor for scoring low, by a factor of 5.9 in women and a factor of 10.6 in men. Depressive symptoms also had a negative impact on performance, with risks of 2.4 in women and 3.9 in men.

Taking these factors into account, the risk of scoring low on the MMSE remained higher among subjects with pesticide exposure (figure 2). For occupational exposure determined by the job exposure matrix, the increase was significant, with an odds ratio of 1.45 (95 percent confidence interval (CI): 1.04, 2.02); it was almost significant for men separately (odds ratio = 1.61, 95 percent CI: 0.90, 2.86) but was not significant for women (odds ratio = 1.07, 95 percent CI: 0.66, 1.75). Risk of scoring low on the MMSE was also higher among subjects who reported working primarily in agriculture. Living in a district planted with vineyards tended to be associated with an increase in risk of scoring low, but not significantly.

Incidence of neurodegenerative diseases between the 5- and 10-year follow-ups

Parkinson’s disease

Twenty-four cases of Parkinson’s disease (10 in men and 14 in women) arose between the 5-year and 10-year follow-ups, corresponding to an incidence of 5 per 1,000 person-years.

Eight cases occurred in occupationally exposed subjects (8.9 cases per 1,000 person-years) and 16 in nonexposed subjects (4.1 cases per 1,000 person-years) (p = 0.07).

In men, univariate analysis showed a significant association between Parkinson’s disease and occupational exposure as determined by the job exposure matrix, with a relative risk of 6.0 that remained significant at 5.6 (95 percent CI: 1.5, 21.6) after adjustment for smoking and educational level (table 3). The adjusted relative risk in men increased from 5.3 (95 percent CI: 0.6, 49.3) in the first quartile to 5.7 (95 percent CI: 0.5, 60.3) in the second quartile and 10.9 (95 percent CI: 1.7, 70.3) in the third quartile. In the fourth quartile, no case was exposed. Results were not significant for other pesticide exposures.

In women, there was no significant association between Parkinson’s disease and any pesticide exposure variable.

Alzheimer’s disease

We analyzed 96 incident cases of Alzheimer’s disease between the 5- and 10-year follow-ups (71 in women and 25 in men). The incidence of Alzheimer’s disease in our study was 21 per 1,000 person-years. Twenty-six cases occurred in exposed subjects (30.7 cases per 1,000 person-years) and 70 in nonexposed subjects (18.9 cases per 1,000 person-years) (p = 0.02).

In men, univariate analysis showed a significant association between Alzheimer’s disease and occupational exposure, with a relative risk of 2.9 that remained significant after adjustment for education and smoking (relative risk = 2.4, 95 percent CI: 1.0, 5.6) (table 3).

The adjusted relative risk in men increased from 1.2 (95 percent CI: 0.2, 9.5) in the first quartile to 3.8 (95 percent CI: 0.8, 17.1) in the second quartile and to 3.9 (95 percent CI: 0.8, 17.6) in the third quartile. The relative risk was 2.6 (95 percent CI: 0.6, 12.2) in the fourth quartile. Results were not significant for other pesticide exposure variables or in women.

DISCUSSION

In this cohort of French elderly, we found an association between past occupational exposure to pesticides and low cognitive performance, together with an increase in the risk of developing Alzheimer’s disease or Parkinson’s disease. The fact that only occupational exposure was related to neurologic outcomes (MMSE score, Alzheimer’s disease, and Parkinson’s disease) and that the relation appeared exclusively in men is consistent with our knowledge of pesticide use in vineyards, since pesticide treatment tasks (i.e., pesticide mixing and application) are performed almost exclusively by males.

Unfortunately, because of industrial and trade interests, information on the use of specific pesticides in defined geographic areas like Gironde is not available. We characterized the pattern of pesticide use in Gironde by analyzing 60 treatment calendars collected from farm owners in the region in 1992 and 1993. The number of different pesticides used on any given farm each year ranged from three to 23. Dithiocarbamates accounted for 37 percent of the organic substances applied (7.2 kg/hectare/year) and folpet for 26 percent (5 kg/hectare/year). Organophosphates and carbamates corresponded to 1 percent and 2 percent, respectively, and were sprayed at mean doses of 230 g and 390 g per hectare per year.

From the present study, we cannot provide definitive conclusions about the specific pesticides responsible for the effects observed. However, these findings allow hypotheses to be drawn about the pesticide exposures currently incurred in vineyards, since one particularity of this agricultural setting is the predominance of fungicide use. Our assessment of Parkinson’s disease at 8 and 10 years of follow-up relied solely on answers to questionnaires, which could have induced a differential bias. However, we do not think such misclassification could be related to pesticide exposure, because the hypothesis of this association was revealed neither to the subjects nor to the interviewers. In any case, the incidence of Parkinson’s disease observed in our study is consistent with that seen in other studies (23).

For Alzheimer’s disease, the diagnosis was based on neurologic examination at each follow-up of the cohort. The incidence rates in the PAQUID Study did not differ from those observed in other European cohorts (24).

The difference in risk between men and women is noteworthy. In case-control studies of Parkinson’s disease, a risk of approximately 2 is usually found, which corresponds to the risk we would find when pooling men and women.

Our results are coherent with those of other studies of Parkinson’s disease but raise new questions regarding differences between the sexes in pesticide-related Parkinson’s disease. They indicate that conclusions might depend largely on exposure assessment. They throw new light on the possible relation with Alzheimer’s disease, as previous studies failed to demonstrate an association. They suggest that cognitive disturbances persist in occupationally exposed subjects, even long after work cessation. Therefore, attention should be paid to the possibility of long-term neurologic effects of pesticide use in agricultural settings where fungicides account for much of the occupational exposure.

ACKNOWLEDGMENTS

This study was supported by grants from the following organizations: Fondation de France, Novartis Pharma, Axa Insurance Group, Caisse Nationale d’Assurance Maladie, Caisse Primaire d’Assurance Maladie de Dordogne, Conseil Général de la Dordogne, Conseil Général de la Gironde, Conseil Régional d’Aquitaine, Danone, Ministère de la Recherche et de la Technologie, Mutualité Sociale Agricole de Gironde et de Dordogne, Mutuelle Générale de l’Education Nationale, Société Péchiney, 2010 Média, and Caisse de Retraite Interentreprise, Capimmec, Institut du Cerveau, Direction Régionale des Affaires Sanitaires et Sociales.

The authors thank members of the PAQUID study team involved in cohort set-up and data collection, especially C. Helmer for her helpful advice. They also thank Professor F. Tison for his role in Parkinson’s disease diagnosis.

Correspondence to Dr. Isabelle Baldi, Laboratoire Santé Travail Environnement, Institut de Santé Publique d’Epidémiologie et de Développement, 146 rue Léo Saignat, 33076 Bordeaux Cedex, France (e-mail: isabelle.baldi@isped.u-bordeaux2.fr).

FIGURE 1. Design of the PAQUID Study, Gironde, France, 1987–1998.

FIGURE 1. Design of the PAQUID Study, Gironde, France, 1987–1998.

FIGURE 2. Risk of scoring low on the Mini-Mental State Examination (threshold: 25th percentile) in the whole study population and separately in men and women, according to occupational exposure to pesticides (job exposure matrix), having a main job in agriculture, residency in a district planted with vineyards, and rural residency, PAQUID Study, 1992–1998. Results are expressed as odds ratios adjusted for age, education, and depressive symptoms.

FIGURE 2. Risk of scoring low on the Mini-Mental State Examination (threshold: 25th percentile) in the whole study population and separately in men and women, according to occupational exposure to pesticides (job exposure matrix), having a main job in agriculture, residency in a district planted with vineyards, and rural residency, PAQUID Study, 1992–1998. Results are expressed as odds ratios adjusted for age, education, and depressive symptoms.

TABLE 1.

Titles of jobs involving a nonnull level of exposure to pesticides, PAQUID Study, Gironde, France, 1992–1998

Job title Exposure level 
Wine grape grower or fruit grower on <7 hectares 
Worker in wine grape growing or fruit growing 
Farmer on <20 hectares 
Truck farmer or horticulturist on <1.5 hectares 
Wine grape grower or fruit grower on 7–20 hectares 
Wine grape grower or fruit grower on 20–40 hectares 
Worker in truck farming or horticulture 
Poultry or mixed animal breeder on <10 hectares 
Farmer on 20–40 hectares 
Gardener 
Carpenter 
Farm worker 
Herbivore breeder on <10 hectares 0.5 
Farmer on >40 hectares 0.5 
Furniture craftsperson 0.5 
Veterinarian 0.5 
Technician in agriculture 0.5 
Fire fighter 0.5 
Farm or forest machine driver 0.5 
Breeding worker 0.5 
Job title Exposure level 
Wine grape grower or fruit grower on <7 hectares 
Worker in wine grape growing or fruit growing 
Farmer on <20 hectares 
Truck farmer or horticulturist on <1.5 hectares 
Wine grape grower or fruit grower on 7–20 hectares 
Wine grape grower or fruit grower on 20–40 hectares 
Worker in truck farming or horticulture 
Poultry or mixed animal breeder on <10 hectares 
Farmer on 20–40 hectares 
Gardener 
Carpenter 
Farm worker 
Herbivore breeder on <10 hectares 0.5 
Farmer on >40 hectares 0.5 
Furniture craftsperson 0.5 
Veterinarian 0.5 
Technician in agriculture 0.5 
Fire fighter 0.5 
Farm or forest machine driver 0.5 
Breeding worker 0.5 
TABLE 2.

Sociodemographic characteristics of the study population at the 5-year follow-up according to pesticide exposure (assessed by a job exposure matrix), PAQUID Study, Gironde, France, 1992–1998

 Not exposed  Exposed  Total 
 No.  No.  No. 
Sex         
Men 454 38.2  131 40.9  585 38.8 
Women 733 61.8  189 59.1  922 61.2 
Mean age (years) 78.4 (6.3)*   79.2 (6.4)   78.6 (6.3)  
Main job         
Housewife 124 10.5  20 6.3  144 9.6 
Blue-collar worker 219 18.5  101 31.7  320 21.3 
Office employee 477 40.4  79 24.8  556 37.0 
Farmer† 14 1.2  87 27.3  101 6.7 
Craftsperson or shopkeeper 179 15.2  20 6.3  199 13.3 
Executive 168 14.2  12 3.8  180 12.1 
Educational level         
Certificat d’Études‡ 254 21.4  159 49.7  413 27.4 
Higher level 933 78.6  161 50.3  1,094 72.6 
 Not exposed  Exposed  Total 
 No.  No.  No. 
Sex         
Men 454 38.2  131 40.9  585 38.8 
Women 733 61.8  189 59.1  922 61.2 
Mean age (years) 78.4 (6.3)*   79.2 (6.4)   78.6 (6.3)  
Main job         
Housewife 124 10.5  20 6.3  144 9.6 
Blue-collar worker 219 18.5  101 31.7  320 21.3 
Office employee 477 40.4  79 24.8  556 37.0 
Farmer† 14 1.2  87 27.3  101 6.7 
Craftsperson or shopkeeper 179 15.2  20 6.3  199 13.3 
Executive 168 14.2  12 3.8  180 12.1 
Educational level         
Certificat d’Études‡ 254 21.4  159 49.7  413 27.4 
Higher level 933 78.6  161 50.3  1,094 72.6 

* Numbers in parentheses, standard deviation.

† Farm workers (n = 72) constituted 4.8% of the total population.

‡ Degree formerly obtained at the end of primary school education in France.

TABLE 3.

Incidence of Parkinson’s disease and Alzheimer’s disease between the 5- and 10- year follow-ups, PAQUID Study, Gironde, France, 1992–1998

 Parkinson’s disease (n = 24 incident cases)  Alzheimer’s disease (n = 96 incident cases) 
 Crude RR† Adjusted‡ RR 95% CI†  Crude RR Adjusted‡ RR 95% CI 
Occupational exposure        
Men 6.05 5.63 1.47, 21.58  2.86 2.39 1.02, 5.63 
Women 0.81 1.02 0.22, 4.82  1.23 0.89 0.49, 1.62 
Main job in agriculture        
Men 2.35 1.62 0.31, 8.63  1.75 1.32 0.43, 4.10 
Women 0.67 0.81 0.10, 6.40  1.15 0.85 0.40, 1.86 
Rural residency        
Men 1.76 1.45 0.38, 5.49  0.96 0.74 0.30, 1.86 
Women 1.20 1.31 0.40, 4.30  0.93 0.76 0.44, 1.31 
Residency in a district planted with vineyards        
Men 0.59 0.46 0.09, 2.29  1.00 0.84 0.35, 2.01 
Women 0.78 0.87 0.24, 3.19  1.03 0.88 0.52, 1.50 
 Parkinson’s disease (n = 24 incident cases)  Alzheimer’s disease (n = 96 incident cases) 
 Crude RR† Adjusted‡ RR 95% CI†  Crude RR Adjusted‡ RR 95% CI 
Occupational exposure        
Men 6.05 5.63 1.47, 21.58  2.86 2.39 1.02, 5.63 
Women 0.81 1.02 0.22, 4.82  1.23 0.89 0.49, 1.62 
Main job in agriculture        
Men 2.35 1.62 0.31, 8.63  1.75 1.32 0.43, 4.10 
Women 0.67 0.81 0.10, 6.40  1.15 0.85 0.40, 1.86 
Rural residency        
Men 1.76 1.45 0.38, 5.49  0.96 0.74 0.30, 1.86 
Women 1.20 1.31 0.40, 4.30  0.93 0.76 0.44, 1.31 
Residency in a district planted with vineyards        
Men 0.59 0.46 0.09, 2.29  1.00 0.84 0.35, 2.01 
Women 0.78 0.87 0.24, 3.19  1.03 0.88 0.52, 1.50 

* Cox models with left-truncation.

† RR, relative risk; CI, confidence interval.

‡ Relative risks were adjusted for smoking and education.

References

1.
Lilienfeld D, Perl D. Projected neuro-degenerative disease mortality among minorities in the United States 1900–2010.
Neuroepidemiology
 
1994
;
13
:
179
–86.
2.
Lilienfeld D, Perl D. Projected neuro-degenerative disease mortality in the United States 1900–2010.
Neuroepidemiology
 
1993
;
12
:
219
–28.
3.
Schulte PA, Burnett CA, Boeninger MF, et al. Neurodegenerative diseases: occupational occurrence and potential risk factors.
Am J Public Health
 
1996
;
86
:
1281
–8.
4.
Priyadarshi A, Khuder SA, Schaub EA, et al. A meta-analysis of Parkinson’s disease and exposure to pesticides.
Neurotoxicology
 
2000
;
21
:
435
–40.
5.
Cole D, Carpio F, Julian J, et al. Neurobehavioural outcomes among farm and nonfarm rural Ecuadorians.
Neurotoxicol Teratol
 
1997
;
19
:
277
–86.
6.
Rosenstock L, Keifer M, Daniell WE, et al. Chronic central nervous system effects of acute organophosphate pesticide intoxication.
Lancet
 
1991
;
338
:
223
–7.
7.
Stephens R, Spurgeaon A, Calvert IA, et al. Neuropsychological effects of long-term exposure to organophosphates in sheep dip.
Lancet
 
1995
;
345
:
1135
–9.
8.
Steenland K, Cedillo L, Tucker J, et al. Thyroid hormones and cytogenetic outcomes in backpack sprayers using ethylenebis(dithiocarbamate) (EBDC) fungicides in Mexico.
Environ Health Perspect
 
1997
;
105
:
1126
–30.
9.
Fabrigoule C, Rouch I, Taberly A, et al. Cognitive process in preclinical phase of dementia.
Brain
 
1998
;
121
:
135
–41.
10.
Baldi I, Filleul L, Mohammed-Brahim B, et al. Neuropsychologic effects of long-term exposure to pesticides: results from the French Phytoner Study.
Environ Health Perspect
 
2001
;
109
:
839
–44.
11.
Gauthier E, Fortier I, Courchesne F, et al. Environmental pesticide exposure as a risk factor for Alzheimer’s disease: a case-control study.
Environ Res
 
2001
;
86
:
37
–45.
12.
The Canadian Study of Health and Aging: risk factors for Alzheimer’s disease in Canada.
Neurology
 
1994
;
44
:
2073
–80.
13.
Dartigues JF, Gagnon M, Barberger-Gateau P, et al. The PAQUID epidemiological program on brain ageing.
Neuroepidemiology
 
1992
;
11(suppl 1)
:
14
–18.
14.
Institut National de la Statistique et des Etudes Economiques. Nomenclature des professions et catégories socio-professionnelles. Paris, France: Institut National de la Statistique et des Etudes Economiques, 1994.
15.
Folstein M, Folstein S, McHugh P. “Mini-Mental State”: a practical method for grading the cognitive state of patients for the clinician.
J Psychiatr Res
 
1975
;
12
:
189
–98.
16.
War Department, Adjudant General’s Office. Army Individual Test Battery: manual of directions and scoring. Washington, DC: War Department, Adjudant General’s Office, 1944.
17.
McCann SJ, LeCouteur DG, Green AC, et al. The epidemiology of Parkinson’s disease in an Australian population.
Neuroepidemiology
 
1998
;
17
:
310
–17.
18.
Wade JP, Mirsen TR, Hachinski VC, et al. The clinical diagnosis of Alzheimer’s disease.
Arch Neurol
 
1987
;
44
:
24
–9.
19.
Tison F, Dartigues JF, Dubes L, et al. Prevalence of Parkinson’s disease in the elderly: a population study in Gironde, France.
Acta Neurol Scand
 
1994
;
90
:
111
–15.
20.
Letenneur L, Gilleron V, Commenges D, et al. Are sex and educational level independent predictors of dementia and Alzheimer’s disease? Incidence data from the PAQUID Project.
J Neurol Neurosurg Psychiatry
 r
1999
;
66
:
177
–83.
21.
Korn E, Graubard B, Midthune D. Time-to-event analysis of longitudinal follow-up of a survey: choice of the time-scale.
Am J Epidemiol
 
1997
;
145
:
72
–80.
22.
Commenges D, Letenneur L, Joly P, et al. Modelling age-specific risk: application to dementia.
Stat Med
 
1998
;
17
:
1973
–88.
23.
Elbaz A, Bower J, Maraganore D, et al. Risk tables for parkinsonism and Parkinson’s disease.
J Clin Epidemiol
 
2002
;
55
:
25
–31.
24.
Launer L, Andersen K, Dewey M, et al. Rates and risk factors for dementia and Alzheimer’s disease: results from EURODEM pooled analyses. EURODEM Incidence Research Group and Work Groups.
Neurology
 
1999
;
52
:
78
–84.