Premorbid diet in relation to survival from prostate cancer (Canada)

Objectives: To examine the associations between prediagnostic energy, fat, and vitamin A intake and survival from prostate cancer. Methods: Two hundred and seven cases of prostate cancer from Toronto and 20 1 cases from Vancouver provided diet histories at diagnosis between 1989 an d 1992 and were followed for survival from prostate cancer. After exclusion s for various reasons, 263 cases (135 from Toronto, 128 from Vancouver) wer e analyzed in Cox proportional hazards models. Results: Following adjustments for clinical stage, histologic grade, and ot her factors, significantly lower risks of dying from prostate cancer in the highest compared with the lowest tertiles of monounsaturated fat intakes w ere observed in each city and in the combined city analyses (combined citie s: hazard ratio [HR] = 0.3; 95% confidence interval (CI) = 0.1-0.7). Surviv al from prostate cancer was significantly better for cases in the highest t ertile of energy intake in Toronto (HR = 0.1; CI = 0.01-0.6) in contrast to that in Vancouver where these cases did relatively worse (HR = 2.6; CI = 0 .6-10.7). Other nutrients were either not consistently or not significantly associated with prostate cancer survival in the two cities. Conclusions: This bi-center cohort study observed a consistent and signific ant inverse association between the premorbid intake of monounsaturated fat and risk of death from prostate cancer. The inconsistent results for energ y intake between cities could potentially be attributed to non-respondent b ias in Toronto.