The purpose of descriptive epidemiology is to identify risk factors that may provide aetiological insights. In the case of prostate cancer, explaining the marked variations in disease incidence by ethnicity present the greatest challenge. Comparisons of population-based registries consistently show that Black men in the USA have the greatest age-adjusted incidence, with a 60 times greater risk than men in Shanghai, China for the period 1988–92 . In the period 1996–2000, Black men in the USA had an age-adjusted incidence rate 1.6 times that of White American men, of 272 per 100 000 men per year . The highest recorded age-adjusted incidence rate of 304 per 100 000 has since been reported in black men living in Jamaica .
There has, however, been minimal research to determine whether ethnic differences in prostate cancer incidence exist in the UK. The African-Caribbean UK population are predominantly first-generation migrants and it would be expected that they would have higher incidence rates of prostate cancer than white men. The only published incidence data for West Indian men comes from the Office for National Statistics Longitudinal Study, which is based on a 1% sample of the population of England and Wales. This has reported a doubling of the standardized incidence ratios compared with all men (2.2, 95% CI 1.0–4.1), but this ratio was imprecise and based on only nine cases, and ethnicity was based on country of birth and may therefore have been misclassified . This is supported by a recent report from East London which noted a 3.5 times greater risk for Black than White men .
What accounts for these differences in risk is unclear. Access to and uptake of healthcare, frequency of PSA testing or other diagnostic tests, and completeness of registry reporting will differ across populations. The magnitude of the observed differences between countries may therefore not reflect true differences, but is unlikely to completely explain these observations. Furthermore, if Black men within countries are less likely to have PSA testing, as in general they are poorer and may therefore have less access to healthcare, then any Black-White differences will be an underestimate of the true rates of latent disease.
Various biological explanations have been proposed to explain the markedly elevated rates for Black men, assuming they are not artefact. These include ethnic differences in: (a) androgen level exposure in utero and in adulthood; (b) components of the IGF axis; and (c) polymorphic alleles in candidate hormonal genes, such as the androgen and vitamin D receptors, and the SRD5A2 gene that encodes the enzyme 5α-reductase type 2, which converts testosterone to its active metabolite dihydrotestosterone. So far, inconclusive results and methodological limitations make it difficult to draw any firm conclusions about these hypotheses. Whilst genetic explanations appear attractive, there have been large increases in the incidence rates for low-risk countries over the last decade, which are less affected by trends in PSA screening . These have occurred in parallel with the increasing adoption of a western-type diet and lifestyle suggesting the importance of modifiable environmental factors.
Whether race itself is an independent predictor of survival after a diagnosis of prostate cancer has also been the subject of considerable debate. Population-based cancer registry data from the Surveillance, Epidemiology, and End Results Program of the National Cancer Institute, which collects data on ≈ 10% of the USA population, has consistently reported that Black men with prostate cancer in the USA have a worse survival than White men, even after adjusting for stage at presentation . However, this important dataset has limited information on important potential confounders such as socio-economic status and comorbidity. In addition, adjusting for stage alone may be inadequate, as studies suggest that Black men present with a higher PSA level and a higher Gleason grade  than White men with the same disease stage, leaving the possibility that observed differences reflect residual confounding. Smaller scale but more detailed hospital-based studies could be informative. In a recent systematic review of 17 hospital-based studies, most (11 of 17) concluded that race was not an independent predictor of treatment failure, whilst the others found a worse prognosis for Black men . It is hard to reconcile these discordant findings, but they most likely reflect methodological differences in population samples, completeness of data, length of follow-up, multivariable modelling techniques and chance variation.
Surprisingly there are currently no UK data on survival by ethnicity. An analysis of routine mortality data by place of birth, for the period 1970–85, showed a 70% higher mortality rate for Caribbean immigrants, however, as mortality is determined by both incidence and case fatality, it is not possible to directly infer whether this reflects differences in survival or not .
If Black men with prostate cancer have a poorer prognosis than white men then several possible explanations must be considered. They may have: (a) more advanced disease and higher grade of disease at presentation; (b) sociocultural differences in access and uptake of healthcare, resulting in delayed presentation or different care; (c) differential clinical management resulting in less aggressive investigation and/or treatment; (d) increased comorbidity and reduced life-expectancy; or (e) a more aggressive disease phenotype, attributable to environmental and/or genetic risk factors.
Studying ethnic differences in the incidence and survival of prostate cancer may provide important aetiological clues that may not only explain the increased risk for Black men, but also explain why other subgroups of the population are at an increased or decreased risk. If Black men are a higher risk group and have a more aggressive disease, then this may be important in relation to early diagnosis. Alternatively, if any observed differences reflect differential access or provision of treatment, then the organization of health systems will need to consider how it can prevent barriers to appropriate specialist care and so provide greater equity in healthcare delivery.