Objective: This study assesses whether a simple count of food items and food groups can predict the nutritional adequacy of the diet in an economically poor country. Design: A three-day weighed record of children. Setting: Koutiala town, in Southeastern Mali. Subjects: Seventy-seven children, 13-58 months of age. One child was excluded owing to an extraordinarily low food variety. Intervention: The study was conducted in April-August 1995. Data from this study were used to create two different indices: Food Variety Score (FVS), a simple count of food items, and Dietary Diversity Score (DDS), a count of food groups. Mean Adequacy Ratio (MAR) was calculated as an indicator for nutrient adequacy, and used to validate FVS and DDS. Results: Mean (s.d.) FVS was 20.5 (3.8) and mean (s.d.) DDS was 5.8 (1.1). A positive correlation was found both between FVS and MAR (Pearson 0.33, P<0.001) and DDS and MAR (Pearson 0.39, P<0.001). With cut-off points for FVS at 23 and for DDS at 6, the indices have high ability to identify those with a nutritionally inadequate diet. MAR increased with increasing FVS and DDS. FVS needs to be at least 15 or DDS at least 5 to give a satisfactory MAR. Conclusion: Although a simple count of food items or food groups cannot give a full picture of the adequacy of the nutrient intake, the results from this study show that the food scores can give a fairly good assessment of the nutritional adequacy of the diet, particularly if combined. Such indicators are important for identification of vulnerable groups in areas where people normally eat from a shared bowl, which makes detailed dietary intake studies difficult, time consuming and expensive. Sponsorship: The Norwegian Universities' Committee for Development Research (NUFU), the Norwegian Research Council and the Nordic Africa Institute funded the project. | 
