Eating by number: Undergraduate dietary practices and the impact of perceived, ideal and actual BMI

Background: Approximately 127 million Americans face an increased risk of morbidity and mortality resulting from excess body weight, while eating disorders are connected with tobacco, alcohol, and marijuana use; delinquency; high-risk sexual activity; and suicide attempts, making it increasingly clear that the U.S. faces a costly and perplexing nutritional paradox. Health policy and programming must strike a balance, encouraging healthy weights without glorifying unrealistic body ideals. Promoting accurate self-assessment of BMI is essential to this task, as research demonstrates that discordance between actual and perceived BMI can result in either increased incidence of eating disorders or dismissal of weight-lose interventions. Methods: Participants were recruited from four introductory and two upper-level undergraduate anthropology courses at the University of South Florida. Freelisting and pilesort methodologies were used in two focus groups to elicit and group commonly consumed food items, which informed the creation of a food frequency questionnaire (FFQ). The FFQ was administered to the larger group along with a BMI Silhouette Matching Test (BMI-SMT). Height and weight measurements were recorded and randomly selected participants asked to maintain journals. Analysis: ANTHROPAC and SPSS were used to (1) assess the level of variation between ideal, perceived, and actual BMI measurements as influenced by demographic factors such as gender and (2) compare measures of cognitive dissonance to FFQ data and cultural consensus models to explore eating strategies associated with various understanding of body shape. The extent to which misconceptions of BMI contribute to negative eating behaviors is assessed and programming recommendations are discussed.