Table 2. Factors associated with smoking before pregnancy and quitting during pregnancy of the study population in the Greek RHEA birth cohort study, 2007–2008.
ORa | 95% CI | P valueb | ||
Factors associated with smoking before pregnancy (Current/ex-smokers vs. nonsmokers) | ||||
Paternal age | Change per year | 0.96 | 0.95–0.97 | <0.001 |
Paternal ethnicity | Greek | 1.00 | ||
Non-Greek | 0.41 | 0.23–0.74 | 0.003 | |
Education level | High | 1.00 | ||
Mid/low | 1.40 | 1.06–1.85 | 0.018 | |
Residence | Urban | 1.00 | ||
Rural | 0.68 | 0.49–0.92 | 0.013 | |
Paternal smoking status | Nonsmoker | 1.00 | ||
Smoker | 2.90 | 2.25–3.75 | <0.001 | |
Factors associated with continued smoking during pregnancy (Current vs. ex-smokers) | ||||
Paternal age | Change per year | 0.98 | 0.97–0.99 | 0.002 |
Paternal ethnicity | Greek | 1.00 | ||
Non-Greek | 3.03 | 0.95–9.70 | 0.062 | |
Paternal smoking status | Nonsmoker | 1.00 | ||
Smoker | 1.76 | 1.16–2.67 | 0.008 | |
Parity | Primipara | 1.00 | ||
Multipara | 1.72 | 1.13–2.61 | 0.011 |
aBackward logistic regression analysis was conducted with the following factors in step 1: maternal and paternal age, level of education (high vs. mid/low), ethnicity (Greek vs. immigrant), paternal smoking status, parity (primipara vs. multipara), place of residence (urban vs. rural), and maternal occupational status (working vs. unemployed or on leave). In the first regression analysis an OR >1 indicates women more likely to be smokers; in the second regression analysis, an OR >1 indicates women who continued to smoke during pregnancy and did not quit.
bP values based on 2-sided tests, with variables excluded when P > 0.1. Statistical significance was defined as P < 0.05.