Although most attention has focused on possible dietary risk or protective factors for prostate cancer as the most likely environmental risk factor category to explain the racial/ethnic variation in incidence and the impact of migration on risk modification, other factors have also been evaluated over the past few decades. Among these, cigarette smoking and a history of any type of sexually transmitted disease are among the most reproducible. As there are no highly suspected carcinogens to the prostate found in cigarette smoke or any direct evidence of an infectious etiology of prostate cancer, it has been proposed that both of these risk factors might be indices of an androgenic profile associated with prostate cancer development.27 Smokers have higher circulating testosterone levels than nonsmokers.28 While there is no direct link between androgen levels and sexually transmitted diseases, androgen levels may be modestly correlated with indices of sexual activity.29
The substantial literature on smoking and prostate cancer has been reviewed,30 with the overall conclusion that smokers have only a very modest, if any, increase in risk but that risk may be somewhat higher for fatal than incident disease. The reasons for this possible risk differential remain obscure.
Among the various parameters of sexual activity indices and prostate cancer risk that have been evaluated (e.g., age at first intercourse, frequency of intercourse, marital status), the only one that has been consistently linked to prostate cancer risk is a history of a sexually transmitted disease.31 While this has suggested to some a possible infectious etiology, the epidemiology does not strongly support this premise. For example, Ross and colleagues32 showed that celibate priests do not have a low risk of prostate cancer; and unlike cervical cancer (known to be caused by a sexually transmitted infectious agent), prostate cancer is not inversely related to socioeconomic status, nor is risk elevated in Latinos. An alternative hypothesis is that such a history is an index of sexual activity, which in turn is correlated with underlying androgen levels.
Other factors have been suggested to increase prostate cancer risk but with highly inconsistent results. Alcohol consumption, which is of interest primarily because of its impact on circulatory steroid hormone levels, was recently the subject of a meta-analysis involving 33 epidemiological studies that had assessed alcohol consumption and prostate cancer risk.33 Alcohol consumers overall had a risk of prostate cancer of 1.05 compared to non-drinkers, but risk increased to 1.21 among men consuming at least four drinks per day.
Obesity and indices of body size have also been of interest in terms of prostate cancer risk. Body mass index and height are of interest mechanistically in terms of both possible impact on steroid hormone levels and possible relationships with insulin-like growth factors (see below, Insulin-like Growth Factor Signaling Pathways).
No consistent pattern has emerged from these studies in terms of either height, weight, body mass index, or other indices of body size.34
In addition to these possible nondietary risk factors, two nondietary preventive factors have been suggested for prostate cancer. Lee et al.35 summarized the 23 studies to date that have examined a role of physical activity in prostate cancer risk.35 They calculated a summary relative risk of 0.8 for the most active versus the least active physical activity categories across this range of studies. The quality of physical activity assessment varied substantially across these studies, and none was designed specifically to assess this relationship. One suggested mechanism for a protective effect of physical activity is via modulation of androgen levels.36
Several recent studies have suggested that regular use of nonsteroidal antiinflammatory drugs may reduce risk of prostate cancer, but no definitive study has been done.37 The probable mechanism for any preventive efficacy is via inhibition of cyclooxygenase 2, an inducible enzyme with proangiogenic, prometastatic, and an-tiapoptotic properties.38
Was this article helpful?