Summary and conclusions, Miljøstyrelsen

Sundhedsmæssige problemer ved brug af mikrobiologiske bekæmpelsesmidler i væksthuse

Summary and conclusions

Microbiological pest control has been introduced in greenhouses as a supplement or substitution for chemical insecticides and fungicides. In Denmark the types most often used are the insecticides Bacillus thuringiensis and Verticillium lecanii. Trichoderma harzianum is used as a fungicide acting as a competitor to the plant pathogenic fungi. Development of this new technology enables evaluation of the health risk.

The aim of this study has been an evaluation of the risk of developing type I allergy and inflammatory lung diseases due to exposure to microbiological pesticides during work in greenhouses.

In 1996 a survey was made on the use of microbiological pesticides in greenhouses in Funen County, Denmark. 193 interviews were carried out with greenhouse owners employing 2,206 full-time workers, rising to 2,756 workers in the peak season. 59% of the greenhouses comprising 69% of the area under glass and employing 78% of the full-time workers used biological pest management. In 27% of the greenhouses both microbiological pesticides and macrobiological pest management (predator mites, nematodes etc.) were used, while 13% exclusively used microbiological pesticides in the form of bacteria and fungi. Most prevalent was the use of the species: Bacillus thuringiensis, Trichoderma harzianum, and Verticillium lecanii.

In the follow up study 31 greenhouse companies were included to cover the use of the different products. At the start they had 773 full-time workers and produced almost exclusively ornamental plants, more than 100 different species with marguerites, chrysanthemum, and roses as the most frequent. The persons were exposed to Bacillus thuringiensis, Trichoderma harzianum, or Verticillium lecanii. In addition, a group without exposure to microbiological pesticides was included. The fungus Paecilomyces fumosoroseus was introduced at the time of the study and therefore included.

In the first round of the study in 1997-1998, 456 persons were included. Of these, 316 were re-examined in the follow up 1 year later, at which time 123 persons were included. This gives a cross section of 579 persons (32% males and 68% females) aged between 16 and 67 years with average seniority of 9.7 years in the trade. The group of persons was followed for three years until 2001.

At the annual examinations, the persons were interviewed about working conditions, focusing on exposure to microbiological pesticides, and health conditions, focusing on symptoms related to allergy and inflammatory upper and lower respiratory tract diseases. A spirometry and bronchial challenge test was carried out using histamine chloride, as well as a skin-prick test with standard inhalatory allergens. Blood samples were taken for analysis for IgE antibodies against the total of 8 products of the four types of pesticides.

In the cross-sectional material of 579 persons, the prevalence of a total of 24 symptoms, covering irritation in eyes and airways, varied between 9% and 31% (itching in the eyes). 6.6% of the subjects reported they had asthma while about 20% reported one or more symptoms of rhinitis.

In greenhouses using Bacillus thuringiensis products the prevalence of itching in the eyes at work increased, and a frequency of more than once a week was observed. The persons who handled Bacillus thuringiensis product had more chest tightness than the rest, while in the follow up an increased incidence of itching in the eyes was seen in greenhouses using Bacillus thuringiensis. Trichoderma harzianum in greenhouses was related to coughing, difficulty of breathing, itching in the nose, and more unspecific annoyance. The use of Verticillium lecanii was not related to any symptoms and the effects of Paecilomyces fumosoroseus could not be evaluated due to a low number of persons exposed.

There was no difference in the lung function and in histamine sensitivity between subjects exposed to the products and those not exposed to the products.

There were measurable IgE antibodies against the different products. The prevalence of antibodies above the detection limits was highest for the Verticillium lecanii and the Bacillus thuringiensis products, while antibodies against Trichoderma harzianum and Paecilomyces fumosoroseus were less frequent. There was no clear relationship with the measures of individual exposure. In particular, no sensitisation was observed in the highly exposed persons who had handled and sprayed the products. During the one-year follow up, the levels of antibodies for the individual were stable and there was no sign of increased sensitisation.

The incidence of new symptoms was about 10% and only in a few symptoms were related to the measurements of exposure or sensitisation. Due to the relatively few new symptoms in the short follow up, the significance of this part of the study for detecting exposure-response relationship is low.

In the cross sectional material, persons with one or more positive prick tests as an indicator of atopy showed higher prevalence of eye and nose symptoms and diagnosed asthma than the non-atopics. Furthermore, the incidence of new symptoms during follow-up in the group of atopic persons was considerably higher than among the non-atopic persons, and persons with house dust mite allergy were over represented in the group that left the study between the first and second examination. This could indicate a "healthy worker effect" due to the different exposures in the greenhouses. The evaluation of health effects should therefore primarily be based on measures of incidence and the study has therefore been extended for three years to achieve a higher significance.

In conclusion, the study showed a relatively high prevalence of symptoms among greenhouse workers, and there was a limited relationship to estimates of exposure to Bacillus thuringiensis and Trichoderma harzianum, but not to Verticillium lecanii or Paecilomyces fumosoroseus products. A few effects in the follow up were seen too, while there was no effect on physiological measurement. The significance of the one-year study, however, was low and the "healthy worker effect" detected supports the notion that emphasis should be on measuring incidence in a longer follow up.

A limited number of significant effects of the microbiological pesticides were seen in this study. Nevertheless, altering procedures and using personal protective equipment during handling and spraying may limit exposure. The possible effects on persons exposed during re-entry activities need more investigation of both the real exposure and the dose-response relationship.