Predictors of Smoking Cessation Among College Students in a Pragmatic Randomized Controlled Trial

An effective strategy to quit smoking should consider demographic aspects, smoking-related characteristics and psychological factors. This study examined potential predictors of smoking cessation in Spanish college students. A total of 255 college student smokers (18-24 years old), recruited to a cessation trial (Spain, 2013-2014), comprised an observational cohort. The main outcome was biochemically verified (urine cotinine) abstinence at the 6-month follow-up. Baseline potential predictors included socio-demographic, smoking-related and psychological variables (Fagerström Test for Nicotine Dependence (FTND), expired monoxide level (CO), intention to quit, previous quit attempts, participation in previous multicomponent programmes and confidence in quitting). Logistic regression models were used to identify potential predictors, the area under the ROC curve (AUC) was used to discriminate the capacity of the predictors and the Hosmer-Lemeshow goodness-of-fit test was used to assess model calibration. After 6 months of follow-up, variables related to high nicotine dependence, FTND and expired CO levels were associated with lower odds of quitting smoking (OR = 0.69 [95% CI 0.54-0.89] and 0.84 [0.77-0.92], respectively). Furthermore, being prepared to change (OR = 3.98 [1.49-10.64], p = 0.006) and being confident to quit (OR = 4.73 [2.12-10.55], p < 0.001) were also potential predictors of smoking cessation. The model that combined all these variables had the best predictive validity (AUC = 0.84 [0.78-0.91], p = 0.693) and showed good predictive capacity (χ² = 10.36, p = 0.241). Findings highlight that, in this population of college student smokers, having a lower level of nicotine dependence, being prepared to quit and having the confidence in the ability to quit were associated with smoking cessation, and these factors had good predictive capacity.