Background: Studies of the association between indoor allergen exposure and the development of allergic diseases have often measured allergen exposure at one point in time.
Objective: We investigated the variability of house dust mite (Der p 1, Der f 1) and cat (Fel d 1) allergen in Dutch homes over a period of 8 years.
Methods: Data were obtained in the Dutch PIAMA birth cohort study. Dust from the child's mattress, the parents' mattress and the living room floor was collected at four points in time, when the child was 3 months, 4, 6 and 8 years old. Dust samples were analysed for Der p 1, Der f 1 and Fel d 1 by sandwich enzyme immuno assay.
Results: Mite allergen concentrations for the child's mattress, the parents' mattress and the living room floor were moderately correlated between time-points. Agreement was better for cat allergen. For Der p 1 and Der f 1 on the child's mattress, the within-home variance was close to or smaller than the between-home variance in most cases. For Fel d 1, the within-home variance was almost always smaller than the between-home variance. Results were similar for allergen levels expressed per gram of dust and allergen levels expressed per square metre of the sampled surface. Variance ratios were smaller when samples were taken at shorter time intervals than at longer time intervals.
Conclusion: Over a period of 4 years, mite and cat allergens measured in house dust are sufficiently stable to use single measurements with confidence in epidemiological studies. The within-home variance was larger when samples were taken 8 years apart so that over such long periods, repetition of sampling is recommended.