National Diet and Nutrition Survey: Assigning mixed dishes to food groups in the nutrient databank

The National Diet and Nutrition Survey (NDNS) gathers information on the dietary habits and nutritional status of the UK population aged 1.5 years upward. Mixed dishes, being a mixture of components with varying proportions, prove problematic for categorising and reporting consumption-unlike basic foods, which can be classified into specific food groups relatively easily. For both purchased and homemade mixed dishes, it is advantageous to have a consistent method for assigning dishes to appropriate food groups, while retaining continuity with previous NDNS surveys to enable investigation of trends over time. Historically two main methods have been used to classify mixed dishes in the NDNS databank: either grouping by meat or fish content, or grouping by the main food component. Neither method is entirely satisfactory. In this study, selected foods were tabulated by both methods to determine which approach should be used in future work. Neither method proved entirely satisfactory alone, and in order to maintain consistency and continuity for the main survey of the new NDNS rolling programme, it was decided to take into account elements from both methods and to place comparable recipes together in the same food groups. A varied approach encompassing food names, proportions of ingredients and case-by-case judgement is the most appropriate way to classify mixed food dishes in a nutrient databank. (C) 2009 Elsevier Inc. All rights reserved.