Limitations of a home characteristics questionnaire as a predictor of indoor allergen levels.

We examined associations between reported home characteristics and levels of indoor allergens in 499 homes of a birth cohort of children of allergic/asthmatic families from the Boston area. The risk of having dust mite allergen (Der f 1 or Der p 1) > or = 2 microg/g on the bedroom floor was highest in houses, but 16% of apartments had levels this high. Compared with that from smooth floors, dust from carpeted bedroom floors had 2.1 times the risk of dust mite allergen > or = 10 microg/g, but levels exceeded 2 microg/g in 32% of uncarpeted floors. The risk of dust mite allergen > or = 10 microg/g of dust from plastic-encased mattresses was 0.5 times that for cloth mattresses, but levels in 21% of plastic-encased mattresses were > or = 2 microg/g. In 93% of homes with and 3% of homes without cats, Fel d 1 was > or = 8 microg/g, but 34% of homes without cats had Fed d 1 > or = 1 microg/g. Cockroach allergen (Bla g 1 or Bla g 2) was > or = 2 U/g of kitchen dust in 26% of apartments and only 6% of houses, and it was detected in 48% of homes with no reported signs of cockroaches in the previous 12 mo. Home characteristics reporting is a relatively weak predictor of the absence of allergen, and exposure can occur even when no cats are present, beds are encased in plastic, no carpets are present, and no signs of cockroaches have been reported.