Diet History Questionnaire: Development of the DHQ Nutrient Database

The nutrient and food group database, created for analyzing the DHQ, is based on national dietary intake data from the 1994-96 US Department of Agriculture's Continuing Survey of Food Intake by Individuals (CSFII). These 24-hour dietary recall data were used to decide which foods to include on the DHQ and what the portion sizes should be.

The CSFII foods were placed into food groups consistent with line items on the DHQ. For example, there are many different individual foods that are subsumed by the line item, "lasagna, stuffed shells, stuffed manicotti, ravioli, or tortellini." The CSFII data were analyzed to provide a mean nutrient or food group value, by gender and portion size, based on all reported intakes of the various lasagna, stuffed shells, etc., reported on 24-hour recalls in CSFII. This produced a single nutrient or food group value by portion size and gender for each food on the DHQ.

Nutrient values were not computed separately by age group because our research (cited below) showed that separate nutrient values by age group did not improve estimates. However, for investigators interested in adding age-specific values in the database, the Diet*Calc software has the flexibility to allow you to do so. For a more detailed description of the method used to create the values for the DHQ database, please see:

Subar AF, Midthune D, Kulldorff M, Brown CC, Thompson FE, Kipnis V, Schatzkin A. An evaluation of alternative approaches to assigning nutrient values to food groups in food frequency questionnaires. Am J Epidemiol 2000;152;279-86.

Nutrients from Nutrition Data System for Research

The methods used to add nutrients such as carotenoids and tocopherols to the DHQ database are the result of the efforts of Dr. Lori Beth Dixon at New York University and Thea Zimmerman at Westat. In the DHQ database such nutrients are clearly marked in the variable labels as "(NDS-R based)". The methods used to add NDS-R nutrients to the DHQ database are described in:

Dixon LB, Zimmerman TP, Kahle LL, Subar AF. Adding carotenoids to the NCI Diet History Questionnaire Database. J Food Comp Anal 2003;16:269-280.

To summarize, however, these values are the result of matching 1994-96 CSFII food codes to similar foods in the nutrient database of the Nutrition Data Systems for Research (NDS-R) from the University of Minnesota, which has nutrient values not available from the USDA Survey Nutrient Database. The values for the DHQ were then computed using the databased approach described in:

Subar AF, Midthune D, Kulldorff M, Brown CC, Thompson FE, Kipnis V, Schatzkin A. An evaluation of alternative approaches to assigning nutrient values to food groups in food frequency questionnaires. Am J Epidemiol 2000;152;279-286.

Addition of Glycemic Load

The Glycemic Index (GI) is a measure of the glycemic effect of carbohydrate in a particular food compared to an equivalent amount of carbohydrate in a standard amount of glucose or white bread. The GI is the ratio of the area under the curve for the glycemic response to the test food to the area under the curve for the glycemic response to glucose. The Glycemic Load (GL) of a serving of a specific food is simply the product of its GI (divided by 100) and the grams of carbohydrate from a single serving of that food. It is important to note that a food with a high GI may not always have a high GL. This can happen if the food has very little carbohydrate (for example, meat) or if the food is consumed in small quantities.

The methods to add GL to the DHQ are the result of the efforts of Andrew Flood at the University of Minnesota in conjunction with Amy Subar at NCI and Steve Hull and Thea Zimmerman of Westat. We assigned published GI values¹ to each of the individual CSFII foods consistent with line items on the DHQ. In cases where CSFII foods did not correspond tightly to foods with published GI values, we used decision criteria to assign GI values. We then calculated gender and serving size-specific GL for each of the 225 food groups using a weighted mean method published by Subar et al². (The Diet*Calc software incorporates these GL values in the latest nutrient database. As a reference, we also provide users with the overall GI values by DHQ food group to provide a sense of the values upon which the GL values are based.) Quality assessments were made to help evaluate the success of this method for assigning GL values. A publication on this method is in press at the Journal of the American Dietetic Association.

In addition, we provide two files. The first is a listing of GI values for individual CSFII food codes that provides GI values for those foods consumed by adults and queried on the DHQ or other FFQs used at NCI. It is organized by CSFII food code. The second, a DHQ-only version, is organized by DHQ food groups consistent with line items on the questionnaire. It includes links to GI values published by Foster-Powell et al¹. We also provide both files together as an MS Excel file. In all these files, the foods listed as having a null value are those to which we did not assign a GI value because they were not important contributors to a food group for which we were calculating a GL value.

1. Foster-Powell K, Holt SH, Brand-Miller JC. International table of glycemic index and glycemic load values: 2002. Am J Clin Nutr 2002;76(1):5-56.

2. Subar AF, Midthune D, Kulldorff M, Brown CC, Thompson FE, Kipnis V, Schatzkin A. An evaluation of alternative approaches to assigning nutrient values to food groups in food frequency questionnaires. Am J Epidemiol 2000;152(3):279-86.