Skip to main content
. 2010 Nov 16;104(1):193–197. doi: 10.1038/sj.bjc.6605996

Table 3. Binary analysis.

        Univariate
Multivariate *
Cancer type Exposed Case Control OR 95% CI OR 95% CI
Glioma No 706 1317 1     1    
  Yes 67 185 0.66 0.49 0.89 0.59 0.42 0.81
Colorectal No 5574 10 543 1     1    
  Yes 658 1467 0.85 0.77 0.94 0.84 0.75 0.94
Brain No 505 1013 1     1    
(excl glioma) Yes 94 151 1.26 0.95 1.67 1.00 0.72 1.38
Breast No 8651 16 834 1     1    
  Yes 1642 3262 0.98 0.92 1.05 0.97 0.91 1.04
Lung No 5555 10 992 1     1    
  Yes 982 1522 1.30 1.19 1.42 1.14 1.02 1.28
Prostate No 6861 13 112 1     1    
  Yes 670 1217 1.06 0.96 1.17 0.94 0.84 1.04
All No 27 841 53 790 1     1    
  Yes 4112 7801 1.03 0.99 1.07 0.93 0.89 0.97

Abbreviations: OR=odds ratio; CI=confidence interval. Patients with repeat prescriptions for tricyclics were defined as exposed. *Corrected for smoking status, diagnosis of depression, alcohol use and BMI. NSAID use corrected for in colorectal cancer only.