Behavioral Characteristics of Diet: Developing Survey Instruments for Ethnically Diverse Populations

Melissa Nelson, Ph.D.
Division of Epidemiology & Community Health, University of Minnesota

The purpose of this proposed research is to validate and refine new survey tools for assessing specific aspects of dietary intake and related characteristics among African American, Latino and Caucasian adolescents, building upon on-going TREC projects at UMN and USC. Brief dietary screeners and related survey instruments are needed to facilitate research investigating behavioral characteristics of energy balance, particularly among ethnically diverse populations. Rather than assessing specific nutrient intake per se, these new survey tools provide detailed assessments of intake of specific food types (e.g., fast food, sweetened beverages) as well as other behavioral characteristics (e.g., meals purchased away from home, school meals, family meal patterning) which may be particularly influential in the development of obesity and important targets for obesity prevention strategies.

The specific aims of this proposal are:

  1. to test the validity of short dietary survey instruments in assessing beverage intake and fast food consumption among ethnically diverse populations of adolescents in the Minneapolis and Los Angeles metro regions. The criterion measure for comparison in the validation will be 24-hour recalls and/or diet records, which are currently being collected through on-going TREC research,
  2. to test the reliability of a 41-item questionnaire assessing a variety of dietary behavior characteristics, including beverage and fast food intake, as well as meal patterning, food purchasing away from home, school meals, and family meal patterns,
  3. to use findings from validity and reliability testing in refining these survey instruments to improve their suitability for use among ethnically diverse populations, and
  4. to descriptively assess. these important obesity-relevant characteristics of eating behavior among the African American, Latino and Caucasian adolescent populations, and to assess the construct validity of the new survey measures by evaluating the associations between these behavioral characteristics and overall nutrient intake (e.g., caloric intake) and body composition measures (e.g.,. BMI, DEXA, BOD POD).