Early onset of breast cancer in British black women

Abstract

British Journal of Cancer (2008) 98, 2011. doi:10.1038/sj.bjc.6604415 www.bjcancer.comPublished online 10 June 2008& 2008 Cancer Research UKSir,We read with interest the article by Bowen et al (2008) reportingdifferences between age of onset, type of breast cancer and survivalin British black and white women in East London. They reportedthat, compared to white women, black women presented onaverage 21 years younger, that tumours in younger black womenwere more likely to be aggressive and that survival was pooreramong black women despite similar treatment and socioeconomicstatus.However, we were surprised by the authors’ statement that therehave been no previous published data on the patterns of breastcancer in British black women. We recently reported data in thisjournal on mortality from various cancers including breast cancerby country of birth in the United Kingdom (Wild et al, 2006). Weused country of birth in our analysis because death certificates inthe United Kingdom do not specify ethnicity. Although werecognise that country of birth does not necessarily reflectethnicity, it is a reasonable proxy. For example 85% of womenborn in the Caribbean and West Indies identified themselves as‘Black Caribbean’ in the 2001 Census (Office for National Statistics,2001).We reported higher breast cancer mortality, with standardisedmortality ratios (SMR) and 95% confidence intervals (CI) amongwomen born in West Africa of 132 (105–163) and in North Africaof 132 (96–176). This was an unexpected finding not reported inprevious similar analyses of mortality by country of birth and wewere interested to see the potential explanations offered by thefindings of Bowen et al.Secondly, we would like to offer an additional explanation forthe observed age difference at presentation in the British blackwomen. The national Census indicates that the British blackpopulation in the United Kingdom is younger in comparison to thewhite population (Office for National Statistics, 2001). This mayinfluence the mean age at presentation of disease. The authorshave sought to compare the age structure in the two populations;however, because broad age groups (16–59 years and 60 years ormore) were used, important variations in age structure mayremain unnoticed. It is a common mistake to deduce from themean age of presentation that the disease occurs earlier. Estimatesof age-specific rates are essential to assess whether true differencesexist in age at presentation.Finally, in our study (Wild et al, 2006) we did not find asignificant excess of breast cancer mortality for women born in theWest Indies (SMR 92 [80–106]), the majority of whom are likely tobe black. There was a much larger number of breast cancer deaths(208) for women born in the West Indies over the 3-year period ofthe study than the 85 and 43 breast cancer deaths among womenborn in West and North Africa respectively. This suggests thatthere may be heterogeneity of breast cancer risk among Britishblack women. ‘British black’ is likely to be too broad a term for thestudy of ethnic differences and ethnicity should be further definedwhere possible (Agyemang et al, 2005).