Role of hormonal and other factors in human prostate cancer.

American men have a lifetime risk of about 18% for prostate cancer diagnosis. Large international variations in prostate cancer risks and increased risks among migrants from low- to high-risk countries indicate important roles for environmental factors. Major known risk factors include age, family history, and country/ethnicity. Type 2 diabetes appears to reduce risk, while high birth weight and adult height are linked to increased risk of aggressive prostate cancer. Limited evidence supports an association with a history of sexually transmitted infections. A previous meta-analysis of eight cohort studies indicated no associations with plasma androgen, estrogen, or sex hormone binding globulin (SHBG) levels. However, there were dose-response relationships with baseline plasma testosterone levels in two studies that adjusted for other serum hormones and obesity. Finasteride (a drug that blocks testosterone activation) reduced prostate cancer risk by 25%. Low-frequency genes linked to familial prostate cancer only explain a small fraction of all cases. Sporadic cases were linked to relatively common polymorphisms of genes involved in (1) androgen synthesis, activation, inactivation and excretion, (2) hormone and vitamin D receptors, (3) carcinogen metabolism, and (4) DNA repair. Epidemiologic evidence supports protective roles for dietary selenium, vitamin E, pulses, tomatoes/lycopene, and soy foods, and high plasma 1,25-dihydroxyvitamin D levels. There is inadequate evidence that vegetables, fruit, carotenoids, and vitamins A and C reduce risk and that animal fat, alpha-linoleic acid, meat, coffee, and tea increase risk. Two major cohort studies found dose-response relationships with dietary calcium intake. Total dietary energy intake may enhance risk. Limited evidence supports a protective role for physical activity and elevated risk for farmers and other men with occupational pesticide exposure, particularly to organochlorine compounds and phenoxy herbicides. There is inadequate evidence for a relationship with alcohol or smoking. Most known or suspected external risk factors may act through hormonal mechanisms, but our review found little supporting evidence, and substantial further research is needed.