The home environment and asthma symptoms in childhood: two population based case-control studies 13 years apart.

BACKGROUND Prevalence surveys of asthma and/or wheezing among all children aged between 7 1/2 and 8 1/2 attending state and private schools in the London Borough of Croydon were conducted in February 1978 and February 1991. Two population based case-control studies drawn from the survey responders were used to investigate the association between childhood wheeze and characteristics of the home environment and to assess whether changes in these characteristics between 1978 and 1991 may have contributed to an increase in the population prevalence of wheeze among school children. METHODS Information on exposure to potential indoor environmental risk factors was obtained from parents by home interview and compared between cases-that is, children with frequent (> or = 5) or in-frequent (1-4) attacks of asthma or wheezing in the past 12 months- and controls, with adjustment for study. Changes in exposure over time were assessed by comparing control groups. RESULTS Between 1978 and 1991 the population prevalence odds of wheeze increased by 20% (OR 1.20; 95% CI 1.04 to 1.39). Change in parental smoking, gas cooking, pet ownership, and central heating did not appear to explain the rise. Use of non-feather pillows was positively associated with childhood wheeze even after adjusting for other risk factors and after re-coding from non-feather to feather cases thought to have changed pillow in response to symptoms (OR 1.54; 95% CI 1.13 to 2.10). The proportion of control children reportedly using non-feather pillows was 44% in 1978 and 67% in 1991. CONCLUSIONS Increased use of non-feather pillows was the only domestic indoor exposure studied which appeared to explain a modest rise in prevalence of wheeze from 1978 to 1991. Our analysis attempts to address behavioural change in response to the child's symptoms but an artifact arising from lifelong avoidance of feather bedding in atopic families cannot be entirely discounted.