Mediation And Moderation Analysis On The Association Of Perception Of Smoking Risk With Successful Smoking Cessation

BackgroundTobacco use continues to be the leading global cause of preventable death. As the world’s largest tobacco manufacturing and consuming country, China has a high prevalence of tobacco smoking. Moreover, in China, the rural population whose smoking rate and smoking amount are both higher than the urban people is in the majority. Therefore, in our country, especially in rural areas, to develop effective tobacco control programs is imperative and urgent.Although tobacco control contains several aspects, promoting smokers to quit smoking is one of the main aspects. Leventhal’s Self Regulation Model (SRM) suggests that individuals’perceptions of a health problem, including its causes, influence their perceived control over the problem and their attempts to cope with it. Moreover, many domestic and foreign researches indicated that level of perception of smoking significantly affected smoking cession. However, how perception of smoking affected smoking cession and whether there are some factors affect the association of perception of smoking with smoking cessation and how these factors work were rarely reported at home and abroad. Therefore, in our study, we discussed the mediating effect of smoking amount and moderating effect of age, occupation, age of starting smoking and marital status on the association of perception of smoking with smoking cessation to provide evidence needed to better inform education for tobacco control intervention.Objectives1. To detect the mediating effect of smoking amount on the association of perception of smoking with smoking cessation.2. To detect the moderating effect of age, occupation, age of starting smoking and marital status on the association of perception of smoking with smoking cessationMethodsA face-to-face interview was conducted in9villages located in three counties in Shandong province of China. A community-based case-control study was conducted with adult spontaneous successful smoking quitters as the cases, and adult spontaneous failed smoking quitters as the controls; The score of perception of smoking risks was categorized into tertiles. The smoking amount was evaluated by pack-year, which was also categorized into tertiles and scored inversely. According to the WHO division standard of age, we divided age into youth, middle age and old age; we divided occupation into work at home, work outside and others; we divided education into below junior high school, junior high school and above junior high school; we divided reasons for quitting smoking into illness or fear of being ill and other; The KHB, a Stata’s user-written program in which the logistic regression model was estimated using the decomposition method proposed by Karlson, Holm, and Breen, was performed to assess the mediation; and we assessed the moderation by conducting linear odds model. Propensity score as the covariate in the regression model was used to adjust the potential confounding in mediation analysis; and propensity score was categorized into tertiles and as the covariate in the model was used to adjust the potential confounding in moderation analysis. Main results1. After adjusting age, age to start smoking, occupation, education, marital status, reasons for quitting smoking by using propensity score, the small negative effect of smoking amount on the association of perception of smoking risks with spontaneous successful smoking cessation was observed (P=0.001, OR=1.034,95%CI:1.013-1.056), the mediation effect of smoking amount is24.29%among the total effect. The dose-response analysis showed that the mediation effects of smoking amount in the groups with middle and high perception of smoking risks were9.95%and39.05%among their total effect, respectively.2. After controlling potential confounding, the moderation effect of age, occupation was not significant (P=0.308and P=0.711). And the moderation effect of age to start smoking, marital status was significant (P=0.038and P=0.003)Conclusions1. This study shows that smoking amount has only small negative effect on the association of perception of smoking risks with spontaneous successful smoking cessation, but in contrast, high level of perception of smoking risks has the major positive effect on spontaneous successful smoking cessation. Increasing smokers’ knowledge of smoking risks should increase the probability of successful smoking cessation.2. This study shows that the association of perception of smoking risks with spontaneous successful smoking cessation is affected by age of starting smoking and marital status, but not affected by age and occupation. When smokers didn’t start smoking at early age or married, whether smokers’age is young or old and occupation type, improving smokers’ knowledge of smoking risks should increase the probability of successful smoking cessation.