We would like to thank Dr Preece for his letter. He raises the issue that loss of symptomatic workers during follow up does not explain the absence of a decline in lung function in workers who worked with laboratory animals for more than four years, and concludes that lung function decline in short term employed workers is not sustained. We think that this interpretation of our data is somewhat overoptimistic and is based on the assumption that lung function changes are similar in those lost to follow up and those available for follow up, and that short term and long term employed workers can be compared. In our paper we analysed lung function decline separately for long term and short term employed workers as there were strong indications that those working with laboratory animals for more than four years were less likely to experience adverse health effects due to contact with laboratory animals. Workers lost to follow up were more often sensitised to laboratory animals and more often reported allergic and chronic respiratory symptoms. In addition, long term employed workers were less likely to become sensitised to laboratory animals during follow up (unpublished data). Long term effects in the recently employed workers should therefore not be inferred from (the lack of) effects in this survivor population.

We also do not agree that an excess decline in FEV₁ in the order of 80 ml/y for exposed and sensitised short term employed workers can be qualified as a “small” effect. From a public health point of view such a decline implies more than a tripling of the normal decline due to aging, and is larger than the excess decline found in populations of heavy smokers.^1–3 The estimated decline in FEV₁ for symptomatic short term employed workers was −22 ml/y, which is similar in magnitude to the normal decline due to aging. Although this estimate was not statistically significant, we think this is mainly an issue of statistical power.

The fact that we found no additional lung function decline due to allergen exposure in sensitised long term employed workers may also be related to qualitative or quantitative aspects of exposure. The exposure measure we used to characterise exposure during follow up was rather crude and might have failed to capture relevant differences in exposure level among the long term employed workers. On a final note, we would like to point out that the models we used are best suited to identify aetiological factors and should not be used for predictive purposes. An article on diagnostic modelling, explicitly dealing with laboratory animal allergy has been published in this journal quite recently.⁴ and is recommended reading for the interested occupational physician.