a1 Department of Genetics, University of North Carolina at Chapel Hill, NC, USA
a2 Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden
a3 Department of Psychology, University of Southern California, CA, USA
a4 Department of Psychiatry and Nutrition, University of North Carolina at Chapel Hill, NC, USA
a5 Department of Psychiatry, Transdisciplinary Tobacco Use Research Center, University of Pennsylvania, PA, USA
Background We investigated 12 variables and their interactions as correlates of smoking cessation among regular smokers in the population-based Swedish Twin Registry (STR).
Method Detailed information on tobacco use and personal characteristics were available from 14 715 male and female twins aged 42–64 years who participated in a screening of the population-based STR and reported being regular smokers in their lifetime. A two-stage analytic design was used to examine correlates of smoking cessation. The sample was split at random and significant main effects and interactions identified in the testing set were examined in the validation set. Hazard ratios (HRs) and 95% confidence intervals (CIs) describe the association between correlates and smoking cessation.
Results Twelve main effects were significantly associated with smoking cessation in the testing set; eight were confirmed in the validation set. Of the nine interactions identified in the testing set, none remained significant when evaluated in the validation set after Bonferroni correction. HRs were highest for Swedish oral smokeless tobacco (snus) use (HR 2.70, 95% CI 2.30–3.20), >11 years of education (HR 1.57, 95% CI 1.43–1.73) and being married or cohabitating (HR 1.51, 95% CI 1.39–1.63). Although not statistically significant after Bonferroni correction, snus use also appeared important in the context of interactions, where lower nicotine dependence score, higher socio-economic status (SES) and greater body size were associated with smoking cessation only among participants who never used snus.
Conclusions Snus use was the strongest independent correlate of smoking cessation. Further studies should investigate the mechanism of this association.
(Received May 11 2007)
(Revised October 17 2007)
(Accepted October 20 2007)
(Online publication December 10 2007)
c1 Address for correspondence: Dr H. Furberg, Department of Genetics, CB#7264, 4109D Neurosciences Research Building, University of North Carolina, Chapel Hill, NC 27599-7264, USA. (Email: firstname.lastname@example.org)