Potential self-selection bias in a nested case-control study on indoor environmental factors and their association with asthma and allergic symptoms among pre-school children.

Selection bias means a systematic difference between the characteristics of selected and non-selected individuals in epidemiological studies. Such bias may be introduced if participants select themselves for a study. The present study aims at identifying differences in family characteristics, including health, building characteristics of the home, and socioeconomic factors between participating and non-participating families in a nested case-control study on asthma and allergy among children. Information was collected in a baseline questionnaire to the parents of 14,077 children aged 1-6 years in a first step. In a second step 2,156 of the children were invited to participate in a case-control study. Of these, 198 cases and 202 controls were finally selected. For identifying potential selection bias, information concerning all invited families in the case-control study was obtained from the baseline questionnaire. Results show that there are several possible biases due to self-selection involved in an extensive study on the impact of the home environment on asthma and allergy among children. Factors associated with participating were high socioeconomic status of the family, more health problems in the case families, and health-related lifestyle factors, such as non-smoking parents. The overall conclusion of this study is that there are selection biases involved in studies that need close cooperation with the families involved. One solution to this problem is stratification, i.e. investigating associations between exposures and health in the same socioeconomic strata.