Skip to main content
. Author manuscript; available in PMC: 2012 Oct 15.
Published in final edited form as: Biol Psychol. 2011 Aug 2;88(1):141–146. doi: 10.1016/j.biopsycho.2011.07.006

Table 3.

Results of Cox regression models examining the association between pain measures and smoking relapse. The omnibus model of the second step significantly improved when pain ratings (during or after) CPT was added to the first model including gender and nicotine dependence (χ2 > 4.5, ps < .05). In addition, the pain ratings (during or after) predicted smoking relapse after controlling for other variables (ps < .05). Note. χ2 = chi-square test for the overall model; Δ χ2 = chi-square test for the change from the previous step; B = regression coefficients; SE = standard error; HR = hazard ratio; CI = confidence interval; CPT = cold pressor task; FTND = Fagerström Test of Nicotine Dependence; MPQ = McGill Pain Questionnaire.

χ2 Δ χ2 B SE HR 95% CI
Pain ratings
(during CPT)
Step 1 5.96*   ----  
Gender −.03 .34 0.97   0.50 – 1.88
FTND .20 .08 1.23* 1.05 – 1.44
Step 2 10.49*   4.54*
Pain .02 .01 1.02* 1.001 – 1.03

Pain ratings
(after CPT)
Step 1 5.96*   ----  
Gender −.03 .34 0.97   0.50 – 1.88
FTND .20 .08 1.23* 1.05 – 1.44
Step 2 11.97** 6.18*
Pain .03 .01 1.03* 1.01 – 1.05

MPQ
Step 1 5.29   ----  
Gender .07 .34 1.08   0.55 – 2.10
FTND .20 .08 1.22* 1.04 – 1.43
Step 2 6.78   1.31 
MPQ .02 .02 1.02   0.98 – 1.07
*

p < .05.

**

p < .01.