Intended for healthcare professionals


Reduced risk of wheezing in children using feather pillows is confirmed

BMJ 1997; 314 doi: (Published 15 February 1997) Cite this as: BMJ 1997;314:518
  1. David Strachan, Reader in epidemiologya,
  2. Iain M Carey, Statisticiana
  1. a Department of Public Health Sciences, St George's Hospital Medical School, London SW17 0RE

    Editor—T J Kemp and colleagues' study of levels of mite allergen in pillows1 was prompted by the unexpected findings of our case-control study in which feather pillows were associated with a reduced risk of severe childhood asthma.2 The cases and controls in our analysis were drawn from a larger survey, and a different subsample of the survey respondents underwent skin prick testing with extracts of Der p I, mixed grass pollen, and cat fur, as described elsewhere.3 The subsample included groups at high and low risk of hay fever, as predicted from family size and structure.3 The questionnaire used in our case-control study was posted to the parents of the 727 children who were skin tested, and replies were received for 617 (85%). We present here the results for 523 children, whose parents denied having made alterations to the child's bedroom (as defined previously2).

    Table 1 shows the proportion of children sleeping on a feather pillow by history of wheezing and reaction to Der p I on skin prick testing. Only 30 children with a weal <1 mm in diameter to Der p I had a visible weal (⁥1 mm) to either grass pollen or cat fur, so the results are similar if categorised by any positive skin prick reaction.

    Table 1

    Number (percentage) of children who used feather pillows by history of wheezing and sensitisation to house dust mite*

    View this table:

    Use of a feather pillow was similar among the 353 children with no history of wheezing and 73 children with inactive wheeze in 1991 but was significantly reduced among the 97 children with wheeze in the past year in 1991. The odds ratio comparing all current wheezers to never wheezers, and feather pillows to non-feather pillows, was 0.46 (95% confidence interval 0.26 to 0.81) (P=0.006). When 26 more severe wheezers were excluded the odds ratio was 0.56 (0.30 to 1.04) (P=0.07). The inverse association of use of a feather pillow with current wheezing was of similar strength among children with and without skin prick reactions to Der p I.

    These observations on essentially new data confirm our earlier findings of a reduced risk of wheezing, and particularly severe wheezing, in children using feather pillows.2 Differences in nocturnal exposure to mite allergen, as proposed by Kemp and colleagues, are unlikely to explain the effect of pillows, which is of similar magnitude in children with and without sensitisation to house dust mite. The observation that levels of mite allergen are no higher in feather pillows than in synthetic substitutes, however, contributes to the equipoise necessary before a more rigorous evaluation of different types of pillow by controlled clinical trial is contemplated.


    1. 1.
    2. 2.
    3. 3.