sex and nativity), household SHS exposure had a significant (p=0.043) total effect, non-significant (p=0.133) natural direct effect, and marginally insignificant (p=0.058) natural indirect effect, which were jointly mediated by asthma status and adolescent’s self-reported smoking status, with a proportion of mediated risk to atopic dermatitis of 29.6%.
After accounting for the influences of the pre-exposure covariates, causal mediation analysis using the IOW approach showed that SHS exposure at home had a non-significant NDE (OR NDE =1.40; 95% CI: 0.90–2.17, p=0.133), marginally significant NIE (OR NIE =1.12; 95% CI: 1.00–1.26, p=0.058) jointly mediated through asthma and self-reported smoking status with a proportion of 29.6% mediated effect on the likelihood of atopic dermatitis ( Table 2 ).