Discussion

In this study we have shown a strong correlation between the country-specific female OC use and incidence of prostate cancer among worldwide, continent and even intra-European nations. This correlation appeared specific to OC as no association was demonstrated with other forms of contraception such as intrauterine devices, condoms or vaginal barriers. Furthermore, prostate cancer mortality was also associated with OC use when examined globally. The correlation to OC use was independent of GDP as a measure of a country's wealth, and strongest in Europe.

This study represents the first systematic analysis of associations between OC use and prostate cancer. It is an ecological study and thus has, as with all correlational studies, significant limitations with respect to causal inference.18 As such, it must be considered hypothesis generating.

There are several plausible explanations for this association. Prostate cancer has been associated with sexual transmission. Although no particular infectious agent has been identified, recent interest in the xenotropic murine leukaemia virus-related virus and its discovery in semen has raised this as a possible candidate.17 19 Clearly more studies are needed. We would hypothesise, however, that if sexual activity were the explanation for the above observations, similar outcomes would be noted for other forms of contraception and that one could even assume a protective effect. As we do not have individual level data, these hypotheses are not testable and would require a long latency period.

Another plausible explanation for the association between OC use and prostate cancer is the potential environmental impact of OCs. The last two decades have witnessed growing scientific concerns and public debate over the potential adverse effects that may result from exposure to a group of chemicals that have the potential to alter the normal functioning of the endocrine system in wildlife and humans. These chemicals are typically known as endocrine disturbing compounds (EDCs). Temporal increases in the incidence of certain cancers (breast, endometrial, thyroid, testis and prostate) in hormonally sensitive tissues in many parts of the industrialised world are often cited as evidence that widespread exposure of the general population to EDCs has had adverse impacts on human health. OCs in use today can potentially act as EDCs as they frequently contain high doses of ethinyloestradiol, which is excreted in urine without degradation. This can then end up either in the drinking water supply or passed up the food chain.11 OCs were made publicly available in the 1960s, and have been widely used since the 1980s, hence the exposure to these substances, even in small quantities, may be chronic enough (20–30 years) to have a clinically significant effect.

There are limited epidemiological data that have examined associations between prostate cancer and exposure to environmental EDCs. These are largely derived from occupational exposures, and many lack internal exposure information. In one retrospective cohort epidemiology study of Canadian farmers linked to the Canadian National Mortality Database, a weak but statistically significant association between acres sprayed with herbicides and prostate cancer deaths was found.20 Multigner et al21 have recently demonstrated that environmental exposure to chlordecone, an organochlorine insecticide with well defined oestrogenic properties, increases the risk of prostate cancer. Studies on workers in Germany22 and the USA23 showed a small but statistically significant excess in prostate cancer mortality, based on a limited number of cases. Other studies have failed to demonstrate this association.24–26 All former studies looked at occupation exposure to high concentrations in pesticides; however, in our study we speculate that low concentrations in drinking water supply may cause PCa, due to the more chronic everyday exposure. Furthermore, environmental EDCs may affect the unborn child in the state of organogenesis and cause significant genetic or epigenetic malformations.

In contrast, several recent studies have demonstrated that PCa may not be related to endogenous androgens. The Endogenous Hormones and Prostate Cancer Collaborative Group, analysing5 18 prospective studies of 3886 men with PCa and 6438 control subjects, found no associations between PCa risk and serum concentrations of testosterone, calculated free testosterone, dihydrotestosterone, dehydroepiandrosterone sulphate, androstenedione, androstanediol glucuronide, oestradiol or calculated free oestradiol. However, this study investigated serum hormonal levels. EDCs may increase the risk of PCa by affecting tissue levels or causing genetic or epigenetic changes that may not be found using serum levels. Li Tang et al6 studied the association between repeat polymorphisms of three key oestrogen-related genes (CYP11A1, CYP19A1, UGT1A1) and risk of prostate cancer in the Prostate Cancer Prevention Trial. The results indicate that repeat polymorphisms in genes involved in oestrogen biosynthesis and metabolism may influence risk of PCa. Further studies are needed to determine the role of EDCs in PCa.

Some may argue that our results only reflect screening and treatment patterns for prostate cancer, with the more developed countries having both a higher use of OCs and a higher incidence of prostate cancer. Unfortunately data on worldwide screening tendencies or prostate specific antigen (PSA) use is unavailable. However, we included a multivariable analysis controlling for GDP per capita. GDP refers to the market value of all final goods and services produced in a country in a given period. GDP per capita is often considered an indicator of a country's standard of living. In our multivariable analysis, OC use was associated with both incidence and mortality, even when controlling for GDP. We believe this analysis has strengthened our hypothesis considerably; however, additional confounding does exist and should be explored in future studies. Finally, we cannot report the true levels of EDCs in the water supply and food chain. We hope such data will be available in the near future.

In conclusion, we have demonstrated a significant correlation between OC use and prostate cancer incidence and mortality. Classic case–control and cohort studies may not reveal this association as we are hypothesising an environmental effect. Tissue correlation and environmental studies are encouraged.