INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The prevalence of respiratory symptoms, asthma, chronic bronchitis, and respiratory function impairment among farmers is considered to be high (1). An excess of asthma symptoms has been observed, especially in grain and swine workers (4). Moreover, farming could be one of the occupations with the highest risks of asthma in nonatopic subjects (8). Few studies on asthma and atopy have been conducted in dairy farming. Studies by Iversen and coworkers (9, 10), Choudat and coworkers (11) and Rylander and coworkers (12) suggested the prevalence of asthma and/or bronchial hyperreactivity to be higher in dairy and swine farmers than in the general population. Pham and colleagues, in French farmers (both dairy and grain), made the same observations (13). Recently, Johnston and coworkers, in Swedish farmers, observed an important increase in the prevalence of asthma in recent years; this prevalence was about 10% in 1995 (14). Bronchial hyperresponsiveness in dairy farmers does not appear to be due to past episodes of farmer's lung disease or sensitization to thermophilic actinomycetes, but is probably caused by the occupational environment of dairy farming (15). By contrast, the prevalence of atopy does not appear to be higher in dairy farmers than in the nonfarming population (2, 16, 17).

The Doubs province is a damp, semi-mountainous dairy farming region in France with small family farms. Occupational exposure in this area is mainly related to hay, microorganisms, and cow allergens because the use of pesticides, fertilizers, ensilage, cereals, or feed additives is very rare. Studies in this region and in this farming context have suggested an excess of chronic bronchitis and bronchial obstruction (18), but asthma has not been specifically analyzed. With the main aim of determining whether this type of farming activity causes asthma and allergy, we evaluated the prevalence of respiratory symptoms and IgE-mediated allergy, and respiratory function status in a sample of dairy farmers and a control group of nonexposed subjects living in the same rural area. The allergological part of this study showed that the prevalence of IgE-mediated allergy was the same in the two groups (17). The results related to respiratory symptoms and function are presented in this report.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This cross-sectional study was undertaken with the cooperation of the Doubs Mutualité Sociale Agricole (MSA) (Agricultural Health Insurance Mutual), the medical unit of which organizes free quinquennial physical examinations for all affiliated members (farmers and agricultural administrative workers) in the province. The examinations were carried out near the subjects' homes in premises supplied by the localities concerned. The study took place in 1994 and involved six districts located at a low-altitude area (250 to 750 m) in the Doubs province.

Population

The target population from these six districts included approximately 1,900 farmers of both sexes, age 16 to 60 yr. Farmers involved in anything other than dairy production were excluded, leaving 1,412 dairy farmers, a quarter of whom (353) were randomly selected to participate in the study. The control group included 189 subjects, all nonexposed administrative workers from the six districts who worked in various agricultural branches (MSA, agricultural banks, Provincial Department of Agriculture, and Regional Chamber of Agriculture).

All subjects from both groups were contacted individually and informed of the objectives of the study as well as its practical procedure. The study protocol was approved by the institutional review board of the MSA and patients gave written informed consent. Each subject accepting to participate was asked to the specific examination session which took place during the customary free quinquennial MSA examinations. Half of the subjects randomly selected from those who did not agree to come to the specific examination were contacted by phone in order to obtain demographic and medical information and the reasons for their refusal.

Subjects were asked to fill in a medical and an occupational questionnaire, undergo spirometry and a bronchodilatation test, have skin prick tests, and a blood test to measure total IgE and Phadiatop.

Questionnaires

The questionnaires were sent to the subjects 10 d before their medical examination, collected during the medical examination, and reviewed in the presence of the subject concerned. The medical questionnaire was an adapted French version of the long version of the European Community Respiratory Health Survey questionnaire (19). The main items in respiratory status were: personal and family history of respiratory disease and allergy, asthma (self-reported and confirmed by a doctor), attack of asthma, wheezing or whistling in the chest, being woken by attacks of shortness of breath or coughing, morning cough or morning sputum in the winter, chronic bronchitis, and dyspnea.

History of allergy covered nasal allergies including hay fever, eczema or any kind of skin allergy, or allergy to insect stings or bites. Self-reported asthma was defined by a positive answer to the question, "Have you ever had asthma?" Chronic bronchitis was defined as cough and expectoration for 3 mo of the year or more for at least two consecutive years. Dyspnea was defined as shortness of breath when hurrying on the level or walking up a slight hill. Nonsmokers were defined as those having smoked on average less than one cigarette, one cigar, or one pipe a day for a year. Current smokers smoked this amount or more, and ex-smokers had stopped smoking at least 1 mo before the time at which they filled out the questionnaire.

The occupational questionnaire was prepared by the authors in collaboration with engineers and technicians of the Provincial Department of Agriculture. Questions concerned the size of the farm and of the herd, the method of storing and drying hay, the type of tasks regularly performed (milking, foddering, etc.) and their relationship with the symptoms.

Respiratory Function Tests

Respiratory function tests were performed according to the American Thoracic Society Recommendations (20). A portable pneumotachograph (Autospiro Minato AS 500; Medical Science Company Ltd., Osaka, Japan) was used to measure slow vital capacity (VC), forced expiratory volume in one second (FEV₁) and forced midexpiratory flow (FEF_25-75). The spirometer was calibrated daily for atmospheric pressure, hygrometry, and temperature, and periodically with a 1.5-L syringe. A minimum of three adequate measurements was required for each subject. The best value was selected after correction to BTPS. Subjects with a FEV₁/VC ratio less than 90% of reference values had a bronchodilatation test with 400 µg salbutamol as a powder (Ventodisks; Laboratoires Glaxowellcome, Paris, France). A bronchodilator response was obtained when the improvement in FEV₁ and/or FVC was both larger than 12% predicted and exceeded 200 ml (21). Values were expressed as a percentage of the European Community for Steel and Coal (ECSC) reference values, calculated in relation to sex, age, and height (22).

Allergological Tests

The method for allergological tests has been detailed previously (17). Briefly, skin prick tests (SPT) were performed on the volar surface of the forearm for the following allergens: Dermatophagoides pteronyssinus, Acarus siro, cow dander, cat dander, mixed grass pollen, mixed betulaceae pollen, and an extract of mixed moldy and nonmoldy hay from farms in the Doubs (Laboratoire des Stallergènes, Fresnes, France). Total IgE were measured by the Microparticular Enzymatic Immuno Assay (MEIA; IM x-IgE; Abbott, Rungis, France). IgE were categorized into two levels using 180 KUI. L¹ as the cutoff for the upper limit of the normal range at the laboratory that carried out the assays. Analysis of serum IgE antibodies against a mixture of relevant inhalants allergens present on a solid phase was performed using the Phadiatop test with the capsulated hydrophilic carrier method (Phadiatop/Cap system; Pharmacia Diagnostic AB, Uppsala, Sweden). The results of this allergological evaluation have already been published in part (17) and therefore will not be detailed in the current report.

Data Analysis

Quantitative variables were expressed as mean with standard deviation. Qualitative variables were compared by the chi-square test and quantitative variables by analysis of variance. When used as a continuous variable, the serum IgE level was transformed logarithmically (log₁₀) before statistical analysis because of its approximately normal log distribution. Because age, sex, and smoking were statistically different between the two groups, a logistic regression model was used to estimate odds ratios for dairy farming adjusted for age (< 35 yr, 35- 49, > 49), sex (female, male), and smoking (nonsmokers, ex-smokers, current smokers smoking less than 11 pack-years, current smokers smoking more than 10 pack-years). A logistic regression model was also used to test interactions between exposure (the fact of being a farmer) and smoking on respiratory symptoms. In this model, smoking was considered in two classes (current and ex-smokers [smokers ever] versus nonsmokers). A multiple linear regression model was used to correlate respiratory function parameters and exposure (dairy farmers versus control subjects), age (continuous variable), sex (male versus female), and smoking (four classes). Data analysis was performed using the BMDP statistical software package. Values of p  0.05 were regarded as significant. Assumption for residual normality was assessed by normal probability plot of the residuals.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Description of the Population and Response Rate

Two hundred sixty-five of 353 farmers and 149 of 189 control subjects qualified for inclusion in the study, giving a response rate of 75.1% and 78.8%, respectively. All 149 control subjects underwent all of the explorations, whereas three of the 265 farmers were not given allergological tests. The main characteristics of the subjects in both groups are shown in Table 1. The farmers were shown to be younger and to smoke less than the control subjects.

The 44 farmers and 20 control subjects randomly selected from the 128 nonrespondents did consent to be interviewed by telephone. The nonrespondents were comparable to the respondents in terms of smoking, sex ratio, and average age. The reasons given were, for the most part, either lack of time, or lack or absence of interest in the MSA health examinations. Two of the 44 farmers (4.5%) and 1 of the 20 control subjects (5%) were self-reported asthmatics. Four of the 44 farmers (9%) and 1 of the 20 control subjects (5%) had a history of wheezing.

Respiratory Symptoms

Table 2 shows the prevalence of atopy, asthma, and respiratory symptoms. The prevalence of atopy was identical in both groups. Fourteen farmers (5.3%) and 5 control subjects (3.4%) had self-reported asthma and 4 farmers and 2 control subjects had attacks of asthma within the last year. These differences were not significant. Asthma was confirmed by a doctor in 13 of 14 farmers and in all control subjects. Fifteen farmers, 7 men and 8 women, were considered as being asthmatic according to the following criteria: asthma reported by the subject (n = 14) and/or attacks of wheezing with breathlessness at rest (n = 1). Symptoms were closely linked to occupation in 7 of 15 cases. By order of prevalence, the occupational tasks were: fodder feeding (6), meal feeding (3), preparing straw litter (3), milking (2), presence in the barn (2), and haymaking (1).

The prevalence of all the respiratory symptoms was consistently higher in the exposed group. The differences were nonsignificant for asthma-related symptoms (attack of asthma, attack of shortness of breath at rest, woken by cough, woken by shortness of breath), but were highly significant for chronic bronchitis, chronic cough, chronic sputum and wheezing.

Table 3 presents the prevalence of the symptoms stratified by smoking status in two classes (nonsmokers and smokers ever) in exposed and nonexposed subjects. It shows that the prevalence of the symptoms not specifically related to asthma (wheezing, morning cough, morning phlegm, chronic bronchitis) increased proportionally from the nonsmoker control group to the smoker exposed group, that the prevalence of these symptoms in exposed nonsmokers was higher than in nonexposed smokers, and that the effect of smoking was more pronounced in exposed subjects. These observations did not apply to asthma and asthma-related symptoms. The effect of exposure on chronic bronchitis and related symptoms was of the same magnitude or higher than that of smoking. This presentation (Table 3) suggests an additive or a "more than additive" effect of exposure and smoking, but analysis of interaction by logistic regression showed a positive interaction (i.e., a synergistic effect) only for chronic cough (p = 0.05).

Respiratory Function

Twenty farmers and four control subjects did not have spirometry or were excluded from spirometric parameter analysis in accordance with the criteria used. With the exception of the FEV₁/VC ratio, which was lower in the farmer group, there was no statistically significant difference in respiratory function between the two groups (Table 4). Bronchodilatation tests were carried out when there was bronchial obstruction defined by a FEV₁/VC ratio < 90% of reference values. This occurred in 52 farmers and 21 control subjects. Only 45 farmers and 13 control subjects, those who were off respiratory medications, were tested. The bronchial obstruction was rarely reversible since only 5 of 45 farmers had a positive bronchodilatation test.

Multivariate analysis using multiple linear regression was carried out to analyze the factors that influence respiratory function. This analysis showed smoking to have an influence on all respiratory function parameters (p < 0.01 for vital capacity and FEV₁; p < 0.0025 for the FEV₁/VC ratio). It also showed the effect of exposurebeing a farmeron the FEV₁/ VC ratio (p < 0.025) (Table 4). The same regression analysis after replacing smoking in four classes by smoking in pack-years as a continuous variable, gave the same results.

Allergological Tests

Results of allergological tests in the two groups, previously reported, are not shown here.

Asthma and asthma-related symptoms were positively and significantly correlated to indicators of IgE-mediated allergy; this was not the case for "nonspecific symptoms" such as chronic cough, chronic sputum, and chronic bronchitis. Ten of the 15 asthmatic farmers had a personal or family history of atopy. Ten showed either a positive SPT, a positive Phadiatop test, or a high level of total IgE; two of the 10 tested positively for all three of these indicators of IgE-mediated allergy. The prevalence of positive indicators of IgE-mediated allergy decreased with age, but the correlation was statistically significant only with respect to the Phadiatop test (p < 0.05). No influence was observed with respect to sex or smoking.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study shows dairy farmers in the Doubs to have an excess of all the respiratory symptoms studied. The difference in comparison with the control group was weak and nonsignificant with respect to asthma and asthma-related symptoms, and strong for chronic bronchitis and related symptoms (cough, phlegm, wheezing). The effect of exposure on these symptoms was generally more pronounced than that of smoking. These respiratory problems were allied to a moderate bronchial obstruction which could not be reversed by bronchodilatation; they do not seem to be explained by IgE-mediated allergy.

Analysis of the results shows the prevalence of asthma and of asthma symptoms to be only slightly higher in farmers. We retrospectively determined that the power of the comparison tests for the prevalence of self-reported asthma in the two groups was 26%. This low value is explained by the low prevalence of the disease and by the size of the samples. It is therefore possible that with larger samples, notably for the control group, significant differences would have been observed. Nevertheless, the sample size used does seem large enough to allow us to consider this segment of agricultural workers not to be at high risk of asthma. The control subjects were agricultural administrative employees with no current occupational exposure: they lived and/or worked in villages or small towns with little or no industrial or environmental pollution. Many of them (approximately 20% according to an estimation by the MSA) nevertheless belonged to farming families and could have been exposed during their childhood or youth to agricultural antigens. However, it is unlikely that this factor significantly affected our results. Indeed, the prevalence of asthma in this control group (3.4%) was often lower than previously published values in the general population in different European countries (23) and very clearly lower than figures reported for three big French cities in the European Community Respiratory Health Survey (28).

The prevalence of asthma in the farmers studied (5.3%) falls within figures previously published on this segment of the agricultural population, that is, between 3% (29) and 12% (9). The latter figure concerns older farmers. Many studies have been carried out in large industrial-like agricultural complexes such as grain silos or swineries, and therefore could have been subject to self-selection bias. In our farming population, family farms are handed down from father to son. It is possible, nevertheless that the most "healthy" of the farmer's children tends to take over the family farm. Moreover, in our farming region, farmers rarely leave their farms for health reasons. An inquiry was conducted at our request by the MSA. This inquiry showed that over the past 10 yr, only nine farmers in the Doubs province had left farming due to occupational respiratory diseases, only two of whom were from the districts studied. It is therefore highly unlikely that our results were influenced by a selective mobility of farmers with asthma symptoms. Consequently, a healthy worker effect is unlikely in our study.

Our study also confirms the risk of chronic bronchitis in dairy farmers. Similar results have already been observed in studies carried out in the same province (18, 30) as well as in another French province (11), Denmark (9), Finland (31), the United States (32), and Italy (33). Chronic bronchitis is prevalent in approximately 9% of all farmers in the Doubs province (17). It was higher (12 to 14%) in districts located at an altitude between 700 and 1,100 meters where rainfall and exposure to microorganisms are higher. In our study, the 6.4% prevalence at low altitudes echoed the results we published in 1993 (30) on neighboring districts. In our current study, the difference with the control group appeared very significantly after adjustment for age, sex, and smoking, with an odds ratio over 10. It is generally believed that, in grain workers and in animal farmers, the adverse effect of exposure is smaller than, or of the same magnitude as, that of smoking (34), and that exposure and smoking have an additive effect (34, 35). The intensity of the exposure effect on chronic bronchitis in our study was higher than it was in the previously cited studies. The effect of exposure on chronic bronchitis and related symptoms (wheezing, morning cough, morning phlegm) was greater than or of the same magnitude as that of smoking (Table 3). Presentation of the symptoms stratified by smoking classes suggests an additive or a "more than additive" effect of smoking and exposure, but analysis of interaction demonstrated a synergy only for morning cough. Larger samples might have allowed us to observe a significant synergistic effect for the other symptoms (chronic bronchitis and related).

In addition, farmers had a moderately but significantly lower FEV₁/VC ratio than the control subjects. The effect of smoking on the respiratory function parameters was, however, larger than that of exposure. This moderate obstruction appeared to be permanent as it was only rarely reversed with 400 µg of salbutamol. This absence of reversibility of the airflow obstruction is further evidence against asthma as a cause of the symptoms in this population. Our results show chronic bronchitis and related symptoms to be a greater problem than asthma in this region and in this agricultural context.

Fifteen farmers were found to be asthmatic. In 2 of the 15 cases, asthma definitely resulted from IgE-mediated mechanisms. In 7 of 15 cases, symptoms were directly linked to occupation. In these seven patients, only three showed positive skin prick tests, a high level of total IgE, or a positive Phadiatop test. Although our sample of asthmatic subjects was far too small to allow conclusions to be drawn, it does not appear that asthma in dairy farmers derives first and foremost from an allergy. Similar findings in a recent Swedish study showed only one-third of the asthmatic farmers tested to have allergy-related asthma (14). Prevalence of IgE-mediated allergy was similar in both groups (17); about 15% of subjects had a positive Phadiatop test and 35% at least one positive skin prick test. Similarly, there was no difference between groups for sensitization to the different allergens tested. Moreover, no statistically significant correlation was found between the allergological tests and the indicators of exposure in farmers (17). Asthma and related symptoms were significantly associated with indicators of IgE-mediated allergy. This was not the case for chronic bronchitis and related symptoms which are the symptoms significantly related to farming.

In conclusion, our study shows that working on dairy farms is associated with an excess of respiratory symptoms which is weak for asthma and related symptoms and high for chronic bronchitis and related symptoms. The effect of exposure on the latter symptoms is greater than that of smoking or of the same magnitude. This excess of symptoms in dairy farmers is associated with a very moderate chronic bronchial obstruction; it does not appear to be closely related to IgE-mediated allergy. Chronic bronchitis is probably a greater problem than asthma in this agricultural context.