Background: Few prospective studies have sufficient power to investigate the relationship between obesity and lethal prostate cancer (PC) in detail or within population subgroups, such as by smoking status. This study aimed to investigate the associations of pre-diagnostic body mass index (BMI) with risk of progression from time of PC diagnosis to fatal outcome, to assess effect modification by smoking status and to evaluate the joint effect of BMI and smoking on PC survival.

Methods: We conducted a survival analysis of 10,206 PC patients from the NCI Breast and Prostate Cancer Cohort Consortium (BPC3). BMI and smoking status were estimated before PC diagnosis. Deaths among PC patients were categorized into deaths from PC and other causes. We conducted the analysis in 3 parts according to exposures: BMI (18-22.9 kg/m2, 23-24.9 kg/m2. 25-27.9 kg/m2, 28-29.9 kg/m2, 30-34.9 kg/m2, 35 kg/m2+), smoking (Never, former, current) and their joint effect, respectively. Each analysis was performed on prostate cancer specific mortality and other mortality separately. Competing-risks regression model was used to take into account of other causes of death. For the above analyses, two models were used. Model 1 controlled for age at diagnosis, drinking status, diabetes status, cohort, duration between baseline and PC diagnosis; Model 2 further adjusted for PC stage and Gleason score at diagnosis in addition to model 1.

Results: There were 1028 deaths from PC and 1919 other deaths among all PC patients during an average follow up of eight years. Higher BMI was associated with higher risk of death from PC and other causes; compared to BMI (23-24.9 kg/m2) group, men with BMI 35+ kg/m2 had HR=1.83, 1.16-2.91 for PC death and HR=1.49, 1.04-2.14 for other deaths. The positive trend was apparent only among never and former smokers but not current smokers. Similar effect was found among men whose BMI measured >5 years but for men whose BMI was measured <=5 years before PC diagnosis, only those in the highest category (BMI 35+ kg/m2) was associated with higher risk of dying from PC. Former smokers as well as current smokers have significant higher risk of progressing to death after PC diagnosis regardless BMI with significant interaction between BMI and smoking.

Conclusions: Higher BMI and smoking history (former and current smokers) are possible risk factors for dying from PC among all PC cases in the BPC3 cohort. The effect of BMI on PC survival is modified by smoking status.

Citation Format: Changzheng Yuan, Jorge Chavarro, Sara Lindström, Peter Kraft, Jing Ma. Prediagnostic body-mass index, smoking, and prostate cancer survival. [abstract]. In: Proceedings of the Twelfth Annual AACR International Conference on Frontiers in Cancer Prevention Research; 2013 Oct 27-30; National Harbor, MD. Philadelphia (PA): AACR; Can Prev Res 2013;6(11 Suppl): Abstract nr B49.