Skip to main content
. Author manuscript; available in PMC: 2024 Feb 7.
Published in final edited form as: Acta Oncol. 2023 Feb 7;62(1):89–99. doi: 10.1080/0284186X.2023.2173536

Table 2.

Hazard ratios (HRs) for suicide with 95% confidence intervals (CIs). Model 1: age-adjusted model. Model 2: additive model including information on previous use of antidepressants. Model 3: interaction model, estimating the effect of prostate cancer on suicide risk depending on the status of previous depression.

Model 1: Age-adjusted model Model 2: Previous use of antidepressants (additive model) Model 3: Interaction model
n Events PYRS HR 95% CI HR 95% CI HR 95% CI
Cancer-free men 357,384 1239 3,084,849 1.00 Reference 1.00 Reference
Prostate cancer 37,527 108 237,638 1.05 (0.85–1.26) 1.02 (0.84–1.25)
Age at diagnosis, per year 1.05 (1.04–1.06) 1.05 (1.04–1.06)
No previous use of antidepressants 358,696 1134 3,069,137 1.00 Reference
Previous use of antidepressants 36,215 213 253,351 2.24 (1.93–2.59)
Cancer-free men with previous depression 32,402 187 231,889 1.00 Reference p = 0.1*
Prostate cancer patients with previous depression 3813 26 21,462 1.39 (0.92–2.10)
Cancer-free men without previous depression 324,982 1052 2,852,960 0.44 (0.34–0.57)
Prostate cancer patients without previous depression 33,714 82 216,177 0.47 (0.40–0.54)
*

p-value of analysis of variance between the additive model (Model 2) and the interaction model (Model 3).