[States and predictors of postpartum smoking relapse: a population based longitudinal study].

Objectives　Although more than half of women who smoke attempt to stop smoking after conception, many relapse after delivery. We conducted a population-based longitudinal study to identify the predictors of postpartum smoking relapse.Methods　Participants were expectant mothers living in Nagoya city, Japan, who notified Health Centers of their pregnancy from April 2014 to March 2015. A self-administered questionnaire was given to the expectant mothers that includes individual factors in the pregnancy: expectant mother's smoking status, age, marital status, experience of parturition, (mother's and father's) occupations, infertility treatment reception, feelings when pregnancy was confirmed, plans to return to parents' house for delivery, accessibility to help with childcare, household member(s) smoking in the same room, alcohol consumption, and depressive symptoms continuing more than 2 weeks.　We followed their smoking status at their children's "3-month-old health check-up" (3 months) and "1-year-and-6-months-old health check-up" (18 months) held in Health Centers until March 2017.　The data were analyzed using a combination of Chi-square or Fisher's exact test and logistic regression modeling. The analyses were conducted separately in primiparas and multiparas in addition to all expectant mothers.Results　Participants were 24,413 mothers; 18,041 were followed up at 3 months and 14,163 at 18 months.　Of the 18,041 mothers at 3 months, 1,031 primiparas and 695 multiparas stopped smoking when they confirmed pregnancy; 89 (8.6%) primiparas and 107 (15.4%) multiparas relapsed at 3 months. Of the 14,163 mothers at 18 months, 789 primiparas and 568 multiparas stopped smoking when they confirmed pregnancy; 155 (19.6%) primiparas and 174 (30.6%) multiparas relapsed smoking at 18 months.　As a result of logistic regression modeling, "multiparas," "younger (<25 years old)," "not married (only in multiparas)," "no plan to return to mother's parent's house for delivery," "household member(s) smoking in the same room (only in primiparas)," and "depressive symptoms (only in all mothers and primiparas)" were the predictors of postpartum smoking relapse at 3 months. "Multiparas," "not married (only in all mothers)," "no help with childcare (only in all mothers)," and "household member(s) smoking in the same room" were the predictors of postpartum smoking relapse at 18 months.Conclusion　More mothers relapsed with smoking after 3 months than before 3 months. The predictors of postpartum smoking relapse differed between 3 and 18 months. Support to continue smoking cessation was needed for each mother at an appropriate time not only in pregnancy but also after delivery.