Objectives: To develop the methods for assessment of food and nutrient intake using standardized food frequency questionnaires (FFQ) in three African origin populations from Cameroon, Jamaica and Caribbean migrants to the United Kingdom.
Design: Cross-sectional assessment of diet from a representative sample in each site, using either a 2-day food diary or a 24-h recall method to determine foods for inclusion on the food frequency questionnaire.
Setting: A rural and urban site in Cameroon, Evodoula and Cite Verte in Yaounde, respectively; a district in Kingston Jamaica; African-Caribbeans living in central Manchester, UK.
Subjects: Aged 25-79 years, 61 from the Cameroonian urban site, 62 from the village site; 102 subjects from Jamaica (additional analysis on a subsample of 20): 29 subjects from Manchester, UK.
Main outcome measures: Foods contributing to nutrients in each site to allow the development of a FFQ.
Results: A high response rate was obtained in each site. Comparison of macronutrient intakes between the sites showed that carbohydrate was the most important contributor to energy intake in Jamaica (55%) and the least in rural Cameroon. In rural Cameroon, fat (mainly palm oil) was the most important contributor to energy intake (44%). Manchester had the highest contribution of protein to energy (17%). Foods contributing to total energy, protein, fat and carbohydrate were determined. In rural Cameroon, the top 10 food items contributed 66% of the total energy intake compared to 37% for the top 10 foods in Manchester. Foods contributing to energy were similar in Jamaica and Manchester. Cassava contributed 44% of the carbohydrate intake in rural Cameroon and only 6% in urban Cameroon. One FFQ has been developed for use in both sites in Cameroon containing 76 food items. The FFQ for Jamaica contains 69 foods and for Manchester 108 food items.
Conclusion: Considerable variations exist within sites (Cameroon) and between sites in foods which are important contributors to nutrient intakes. With careful exploration of eating habits it has been possible to develop standardized, but locally appropriate FFQs for use in African populations in different countries.