Discussion

In the present study we have collected detailed dietary intake data from adults of Black Caribbean and West African ethnicity residing in the UK and described for the first time the influence of ethnicity and indicators of acculturation such as length of UK residence on the nutritional composition of the diets. Our results show that in adults of Caribbean and West African ancestry having a long UK residence is associated with greater dietary acculturation and increasing intakes of energy, fat and Na; a nutrient profile known to have detrimental health impacts⁽ Reference Keys ^{23 )}. Furthermore, we have presented detailed information for the foods that contribute to these changing nutrient profiles and described a diet rich in traditional foods in West African adults compared with a more Westernised or acculturated diet in Caribbean adults who have been resident in the UK longer.

There has been relatively little work performed to describe the dietary intake and nutritional composition of the diets of African and Caribbean communities in the UK or the effects and extent of dietary acculturation. Traditionally, African-Caribbean diets have been characterised by lower total and saturated fat contents and higher total carbohydrate content compared with that of White European adults⁽ Reference Vyas, Greenhalgh and Cade ^{6 ,} Reference Goff, Griffin and Lovegrove ^{24 )} and the UK national average⁽ Reference Gregory, Foster and Tyler ^{25 )}, which is believed, at least in part, to contribute to the cardioprotective lipid profile⁽ Reference Miller, Kotecha and Wilkinson ^{26 )} and low rates of IHD in these communities⁽ Reference Wild and McKeigue ^{27 )}. However, emerging evidence suggests this cardioprotection is being lost in subsequent generations; in the USA relative risk rates of myocardial infarction are now greater in young African-American adults compared with Euro-Americans⁽ Reference Jolly, Vittinghoff and Chattopadhyay ^{28 )} and in the recent CHASE study of children in England the lipid profile of the children of Black Caribbean ancestry was no different from that of the White European children⁽ Reference Donin, Nightingale and Owen ^{13 )}. Historical work by Sharma et al.⁽ Reference Sharma, Cade and Jackson ^{10 )} comparing the dietary intake of a small cohort of traditional and migrant Jamaican adults described distinct differences in fat and saturated intakes⁽ Reference Sharma, Cade and Jackson ^{10 )} and in further studies comparing first-generation Caribbean migrants and second-generation UK-born migrants increases in fat and saturated fat were associated with being born in the UK⁽ Reference Sharma, Cade and Riste ^{8 )} suggesting a clear impact of migration, birthplace and UK residence on dietary intake. These studies are approaching 20 years old and the importance of the African-Caribbean community to the UK population has continued to increase. Furthermore, ethnicity is an evolving concept and it is now recognised that the traditional ‘African-Caribbean’ classification of ethnicity includes many distinct cultures that newer classifications try to recognise; it is important that a more current assessment of the dietary intakes of these communities is performed.

The participants in the present study were of Black Caribbean or Black West African ethnicity which was self-defined by the participants and confirmed by parental and grandparental birth places. The demographic profile of the ethnic groups differed: the Caribbean participants were older, more likely to be born in the UK and had been resident in the UK for longer than the West African participants, which is representative of national data^{( 1 )} and reflects earlier large-scale migration from the Caribbean islands in the 1950s compared with the countries of West Africa in the 1980s. In our study we have shown that the acculturation related to being resident in the UK for a prolonged duration is associated with significantly greater intakes of energy, fat and Na. In a recent analysis of adults of West African origin, Anderson et al.⁽ Reference Anderson, Younger and Heald ^{29 )} proposed that increasing protein, fat and saturated fat intakes were associated with increased risk of type 2 diabetes, such that for every 1 % increment in energy from fat and saturated fat, diabetes risk increased by 5 and 16 %, respectively⁽ Reference Anderson, Younger and Heald ^{29 )}. A further important morbidity in people of African and Caribbean origin is hypertension, which is believed to contribute to high rates of stroke, and may occur due to greater salt sensitivity⁽ Reference Richardson, Freedman and Ellison ^{30 )}. Our data suggest that dietary changes occur in adults born in the UK or in those with a prolonged residence that could further drive the development of type 2 diabetes and hypertension; this is particularly concerning given the already high prevalence of these diseases in these communities⁽ Reference Mindell and Zaninotto ^{2 )}.

The detrimental effects of an ‘urbanised’ lifestyle are recognised in driving the burden of chronic diseases and are therefore a target for both public health and health-care management programmes. Explicit inequalities in health-care provision are recognised to impact on minority ethnic groups in the UK and elsewhere, and a need to improve the cultural sensitivity of health care as well as develop targeted prevention strategies is recognised⁽ Reference Lakhani ^{14 )}. The present study provides a detailed analysis of the foods that are important in the diets of people of Black Caribbean and West African ethnicity, which is important for better informing health-care practitioners who work with these communities. We have shown that diets of West African adults rely significantly more on traditional foods and dishes compared with a relatively Westernised diet in Caribbean adults. The present study is the first to describe important foods consumed in the West African diet and to demonstrate the contribution they make to the nutritional profile of the diet; for example, traditional ‘one-pot’ stews and soups and traditional starches such as yam, cassava and fufu. Our data also demonstrate which non-traditional foods and dishes are prevalent in the diet of these Caribbean and West African adults; for example, biscuits, confectionery, processed meats, sugar-sweetened beverages and breakfast cereals. Similarly we have presented an analysis of the foods consumed by the Caribbean participants for whom traditional foods and dishes are important but less so than for the West African participants. Sharma et al.⁽ Reference Sharma, Cade and Jackson ^{10 )} produced an FFQ specifically for populations of African-Caribbean ancestry and described the foods and dishes that were important to this community. Interestingly, many of the traditional Caribbean foods that featured heavily in the diet almost 20 years ago, such as curried mutton, homemade West Indian soup and hard dough bread, no longer feature in the lists we have produced and are replaced by Westernised foods such as confectionery, savoury snacks, biscuits and sugar-sweetened beverages. In more recent analyses of the diets of African-American adults, Sharma et al.⁽ Reference Sharma, Wilkens and Shen ^{11 ,} Reference Sharma, Cao and Arcan ^{31 )} demonstrated the importance of foods such as burgers, sodas and very little reliance on traditional foods, which is in agreement with our findings. In the Caribbean diet we found that sugar-sweetened beverages were the fifth most important contributor to energy intake, which is alarming considering recent evidence from a meta-analysis demonstrating the detrimental impact that sugar-sweetened beverage consumption has on body weight⁽ Reference Malik, Pan and Willett ^{32 )} and may in part explain the higher BMI of the Caribbean participants. This information assists in informing health-care practitioners of the dietary practices of these communities and enables identification of foods that may need to be targeted or encouraged in advising patients on dietary modifications, ultimately enabling care to be more culturally sensitive and individualised.

The strengths and limitations of the study warrant consideration. The study benefits from a more sensitive classification of ethnicity than has previously been used; studies from the USA use ‘African-American’ and older studies from the UK use ‘African-Caribbean’ to describe all people of African ancestry; in using these collective terms these studies describe a heterogeneous population and fail to recognise distinct cultural practices. In our study we have described ethnicity more specifically and as a result have presented data that are of direct relevance to health-care practitioners who care for members of these communities and for the development of culturally sensitive public health strategies. The dietary data were collected using a series of repeated structured 24 h recall assessments which allowed us to record detailed information on traditional foods and recipes and portion sizes; most of the studies in the literature have used a single 24 h recall⁽ Reference Sharma, Cade and Jackson ^{10 ,} Reference Donin, Nightingale and Owen ^{12 )} or FFQ⁽ Reference Vyas, Greenhalgh and Cade ^{6 ,} Reference Mennen, Jackson and Sharma ^{7 ,} Reference Sharma, Wilkens and Shen ^{11 ,} Reference Anderson, Younger and Heald ^{29 )} that are unlikely to have captured such detail. Our first assessment was performed in person to enable detailed information on portion size consumption to be collected; furthermore, ethnically sensitive photographic aids were used to record portion sizes accurately. Our interviews were lengthy (45–60 min) in order that we captured full details of recipes and home-prepared dishes. These were analysed as new entries in our nutritional composition database as the original database contains very few traditional Caribbean or African foods. We entered more than 420 new recipes or foodstuffs into the database and are confident that these had a significant impact on our findings, as confirmed by the contribution these foods and dishes made to the top ten food sources of nutrients. The total contribution that each of our top ten food lists made to total intake is comparable to that in previous studies⁽ Reference Sharma, Cade and Jackson ^{10 ,} Reference Sharma, Wilkens and Shen ^{11 )} and gives us confidence that our food groupings, made using an adaptation of the groupings in the UK National Diet and Nutrition Survey, were appropriately sensitive.

Rates of under-reporting with the 24 h recall methodology are similar to those of other diet assessment methodologies⁽ Reference Poslusna, Ruprich and de Vries ^{33 )}. In our study we estimated that under-reporting may have been present in approximately 37 % and 51 % of the Caribbean and West African participants, respectively; this is slightly higher than the general average of 30 %⁽ Reference Poslusna, Ruprich and de Vries ^{33 )} but not markedly different from Mennen et al.’s study of migrant and traditional African-Caribbean communities which showed that under-reporting is higher in migrant communities⁽ Reference Mennen, Jackson and Sharma ^{7 )}. We appreciate that our participants are not representative of the wider UK Caribbean and West African communities; however, we believe our study is the biggest study to date to collect detailed food group and nutrient data through dietary records as opposed to FFQ. Furthermore, we are aware of other limitations: our participants were recruited predominantly from London and the surrounding areas and may not therefore be nationally representative, but it is recognised from national records that London is the most important region in the UK with up to 80 % of the African and Caribbean community residing within it^{( 1 )}. Our sample size was not sufficient to explore gender differences within and between ethnic groups, which may be an important line of investigation in future work⁽ Reference Parackal, Smith and Parnell ^{34 )}. All dietary assessment methodologies have limitations associated with misreporting and misinterpretation; however, we are confident that our data have not been more greatly impacted by these limitations than other studies of this type as our data are comparable to other studies⁽ Reference Sharma, Cade and Riste ^{8 ,} Reference Mennen, Jackson and Cade ^{35 )}. In analysing the nutritional composition of the diets of these communities we are limited to reporting the principal macronutrients, Na and fibre for which we had complete data. While other nutrients are of interest we acknowledge that, due to our nutritional composition database containing very few Caribbean or West African foods, our analysis included a significant proportion of data taken from food labels which are usually limited to macronutrients, Na and fibre. We appreciate the multidimensional nature of dietary acculturation⁽ Reference Satia-Abouta, Patterson and Neuhouser ^{9 )}; our quantitative methods did not enable us to explore the complexity of this phenomenon in our communities. Finally, we did not collect data on any biomedical markers of health and are not able to draw any firm conclusions on the effects of our findings on the health of our participants. It is important that further studies are performed to understand the impact of dietary acculturation on chronic disease risk in these communities.