Dust samples were weighed and placed, unsieved, into glass Petri dishes and macerated in 70% lactic acid. A few drops of 1% (w/v) aqueous lignin pink were added to stain the mites, and the dishes placed on a hotplate at 45 ℃ for up to 3 h. The fluid was then poured off into a fresh dish. The first dish was scanned twice under a stereobinocular microscope and the mites removed from the remaining fluid with a finely drawn Pasteur pipette. The pouring process was repeated with the second dish, and so on, until all the fluid had been examined. Mites were washed in distilled water and mounted in Faure's medium using the frozen block technique. Mean extraction efficiency was 81%. Mites were identified and counted under a compound microscope and were recorded as 'damaged' (i. e. dead) or 'intact' (i. e. live)using morphological criteria.

3. Assessment of damp in homes: Homes were assessed visually and graded as "damp" based on the presence of two or more of the foltowing features: presence of mould growth on walls and ceilings, strong smell of mould, peeling and water-stained wallpaper and moist plasterwork, condensation on windows, walls and ceilings of rooms other than bathrooms or kitchens,  and signs ot direct water penetration. Homes graded as "dry" hạd none of these features.

4. Statistical analyses: Mite population densities were expressed as numbers of mites per 0.1 g of dust. Medians were used as the measurement of central tendency, with interquartile ranges. Analysis of differences in mite population density was performed using the Mann-Whitney U-test. Analysis of differences in frequency of damp in homes, and numbers of people in each mite exposure class was performed using the x² test with Yates' correction. Relative risk was calculated as described by Armitage.

责任编辑：admin