부산대학교 도서관

The associations of adiposity and smoking with endometrial cancer risk, although well established, are not fully understood. Most previous studies have relied on relatively crude measures of exposures and were not large enough to examine reliably these associations by population characteristics and tumour subtypes. This thesis aims to characterise in unprecedented detail the associations of both adiposity and smoking with endometrial cancer risk. The analyses were performed using data from UK Biobank and the Million Women Study. Adjusted hazard ratios (HRs) for the associations of different measures of adiposity and smoking with endometrial cancer risk were estimated using Cox regression models. Analyses involving 135,110 postmenopausal women in UK Biobank demonstrated that body mass index (BMI) and more specific indices of overall adiposity obtained from bioimpedance analysis (such as fat mass) were broadly similar in their associations with endometrial cancer risk (HR per 1-SD BMI = 1.71, 95%CI 1.61-1.82; fat mass = 1.73,1.63-1.85). These associations were mediated, in part, by inflammation and alterations in free testosterone levels. Central adiposity was associated with a small increase in risk of endometrial cancer independently of BMI (HR per 1-SD waist to hip ratio = 1.13, 1.01-1.26). Data from up to 674,971 postmenopausal women in the Million Women Study showed that the risk associated with adiposity was more pronounced in never and past HRT users than current users. The association between adiposity and risk was significantly stronger for type I than type II tumours. Analyses involving up to 195,446 UK Biobank women and 743,114 Million Women Study participants demonstrated that current and past smoking reduced the risk of endometrial cancer by 27% and 11%, respectively, compared with never smoking. The risk reduction was dose-dependent, consistent across tumour subtypes and persisted for more than 20 years after smoking cessation. These results suggest that both overall and central adiposity contribute to endometrial cancer risk and provide support for the hypothesis that adiposity influences endometrial cancer risk through pathways involving sex hormones and inflammation. The findings also indicate that smoking confers long-term protection against endometrial cancer risk and highlight the need for further research to elucidate the mechanisms underlying the reduced risk of endometrial cancer among smokers.