| Abstract |
Studies evaluating the nutritional status of individuals living with HIV infection have consistently reported a strong association between malnutrition and mortality. In resource poor countries, malnutrition frequently worsens the prognosis of HIV-infected individuals and nutrition interventions may be the most cost effective treatments available to some populations. Accurate dietary assessment is essential in evaluating the potential role of nutrition interventions in HIV/AIDS outcomes and requires evaluation of food intake and subsequent conversion of foods consumed to nutrient intakes and dietary patterns. Our goal is to develop and evaluate a food composition database for assessment of dietary intake data from international studies investigating the role of nutrition in HIV infection and to make the nutrition database available to investigators within the CFAR network. 24-hour dietary recalls will be collected from subjects enrolled in studies in Vietnam, Argentina and India. All 24-hour recalls will be coded centrally at Tufts University and analyzed in a computerized nutrient database, the Nutrition Data System (NDS) 2005. Because the NDS nutrient database is limited in the number and type of international foods that it contains, and the nutrient content of foods can differ among countries, it is necessary to supplement the NDS with additional foods. Using food composition tables from each of the three countries, foods will be entered as a recipe, with the "ingredients" consisting of the nutrients per 100 grams edible portion of the food. Actual recipes from the three country sites, where available, will be entered using the same method as for foods contained in the food composition tables. Ingredients will replace nutrients, and the serving size in volume or weight and the yield of the recipe are entered. As each food or recipe is entered into the database, it will assign a specific food category so that cluster analysis may also be used in analyzing the dietary data. In this study, we will also compare estimated nutrient intake data from 24-hour recalls done on NDS and the same 24 hour recall analysis done locally using the Indian database (it is the most complete). We will compare estimated nutrient intakes between the two methods in four ways 1) median of individual differences in intakes 2) correlation coefficients 3) quintile ranking and 4) intake falls below the Dietary Reference Intakes (DRI).
|
