Baseline characteristics by allocation group (n = 7775).

Values are absolute numbers (percentages) for categorical variables. *Tested for differences in characteristics in intervention and control group using multinomial regression adjusted for randomisation stratum. Characteristics established using postal questionnaires during pregnancy included: smoking during pregnancy (yes, no), maternal atopy (yes, no), maternal ethnicity (Dutch, non-Western, other-Western) and maternal educational level. The Dutch Standard Classification of Education was used to categorise women’s self-reported highest education qualification [21]: low (less than 4 years of high school), middle (college), and high (Bachelor’s degree, Master’s degree). Data on net household income were available at the 2nd year of life. Birth weight (grams) and gestational age at birth (weeks) were obtained from medical records. Postnatal factors were established using questionnaires and included: breastfeeding ever at age 0–6 months (yes, no); keeping pets (yes, no) at the 1st year of life; ETS exposure at home (yes, no) measured at age 0–6 months; respiratory tract infections (yes, no) and wheezing (yes, no) at the 1st year of life.