Study Examines Effectiveness of Dietary Recall Method

By Jim Core
May 14, 2003

A human nutrition research procedure called the "dietary recall method" can be used effectively to gauge people's food consumption, according to a new study by Agricultural Research Service scientists in Beltsville, Md. In the study, 133 adults recalled what they ate within 10 percent of the actual amounts of food they consumed.

ARS scientists at the agency's Beltsville Human Nutrition Research Center (BHNRC) conducted the study to test the accuracy of what's called the "5-step multiple pass method," a dietary recall procedure developed by the U.S. Department of Agriculture to assess food consumption. ARS is USDA's chief scientific research agency.

Congress requires USDA to survey the food intake of Americans. Under that mandate, BHNRC's Food Surveys Research Group (FSRG) conducts the "What We Eat in America" survey. Using the survey data, researchers examine diets as factors in disease or malnutrition, and they help public policy officials make decisions concerning food safety and food fortification.

In the recall study, Joan Conway, a research chemist with BHNRC's Diet and Human Performance Laboratory, and FSRG collaborators found that participants overestimated actual intakes of energy, protein, carbohydrate and fat by less than 10 percent. Previous studies of other 24-hour dietary recall methods found significant under-reporting of food intake.

Participants ate all meals and snacks for one day in the BHNRC human study facility. The amount of food they ate was recorded by a dietitian. A structured telephone interview was conducted the next day using the 5-step multiple pass method. During the interview, researchers asked participants what they ate the day before. They were given measuring aids including the USDA Food Model Booklet, measuring cups and spoons. The food they remembered eating was compared to the actual recorded amount.

U.S. Department of Agriculture
 


This article comes from Science Blog. Copyright © 2004
http://www.scienceblog.com/community