Table 4

Prostate cancer-related physical symptoms—late disease patients

Stage III/IV—any GleasonUnivariate modelMultivariate model 1*Multivariate model 2†
Weighted proportionOROROR
(NI vs RoI)(NI vs RoI)(NI vs RoI)
Ongoing side effectRoI (%)NI (%)(RoI as baseline)(RoI as baseline)(RoI as baseline)
Urinary incontinence22.215.90.650.660.88
(0.44 to 0.97) p=0.035(0.44 to 0.99) p=0.047 (0.55 to 1.41) p=0.591
Loss of libido51.6‡64.7‡1.68‡1.61‡1.32
(1.22 to 2.31) p=0.001(1.16 to 2.23) p=0.005(0.92 to 1.90) p=0.129
Erectile dysfunction66.966.40.951.091.29
(0.68 to 1.33) p=0.784(0.77 to 1.55) p=0.623(0.87 to 1.89) p=0.202
Bowel problems14.221.71.601.401.19
(1.07 to 2.39) p=0.021(0.90 to 2.16) p=0.133(0.75 to 1.87) p=0.458
Breast changes (gynaecomastia)9.4‡23.3‡2.80‡3.09‡2.30‡
(1.81 to 4.32) p<0.001(1.94 to 4.91) p<0.001(1.41 to 3.73) p=0.001
Hot flashes18.8‡41.1‡2.95‡2.79‡2.33‡
(2.08 to 4.18) p<0.001(1.95 to 3.99) p<0.001(1.55 to 3.51) p<0.001
Fatigue24.6‡39.0‡1.93‡1.71‡1.53
(1.39 to 2.70) p<0.001(1.20 to 2.44) p=0.003(1.05 to 2.23) p=0.028
  • Results are weighted by country, age at diagnosis and time since diagnosis.

  • *Logistic regression model adjusted for current age, number of comorbidities, time since diagnosis and method of diagnosis.

  • †Logistic regression model adjusted for above plus prostatectomy, external beam radiotherapy, brachytherapy and hormone therapy records with missing treatment or method of diagnosis dropped from all models (n=12). Significant difference at p<0.05 but with Bonferroni correction applied.

  • ‡Significant difference between countries with RoI as baseline.

  • NI, Northern Ireland; RoI, Republic of Ireland.