Certain risk factors for prostate cancer are well identified: age, family history of this cancer, ethnic origin (e.g. Afro-Caribbean origin) and large size. But apart from obesity and occupational exposure to pesticides, few modifiable risk factors have been identified, although several are suspected. This is why a team of Chinese researchers conducted a literature review to evaluate them. This was published in Plos Medicine.
The selected publications were identified until September 2023. They brought together 92 meta-analyses of prospective observational epidemiological studies and 64 Mendelian randomization studies, which seek to establish links between a condition and genetic factors.
The meta-analyses found 45 significant associations between cancer and risk factors, most of which had already been found in previous studies, including consumption of sugary drinks, processed meats, dairy products, vitamin D and eggs increasing the risk and that of soy and selenium reducing it. No association was noted between cancer and a medical history.
Mendelian randomization studies have noted associations between cancer and the height of individuals, as well as tobacco consumption and physical activity, but these associations are not very robust. Combining meta-analyses and Mendelian randomization studies, the most significant associations are physical activity, height and smoking, but here too they are not robust.
The authors highlight that compared to previous work, their study included more potential risk factors (notably pre-existing illnesses, treatments received, surgical history). In addition, for the first time she brought together meta-analyses of observational studies and Mendelian randomization studies.
They recognize several limitations to their work. Most of the work has focused on Western populations living in rich countries, which prevents the generalization of their conclusions to other populations on the planet. Often, the association of a risk factor with a cancer stage is not explicit. For example, smoking is correlated with the incidence of prostate cancer, but studies differ regarding its effects on aggressive prostate cancer. This is why they recommend harmonization of the severity criteria between the two types of studies. Finally, they recommend taking with caution the results of Mendelian randomization studies which are not supported by observational epidemiological studies or biological arguments.
