Identifying eating patterns in male and female undergraduates using cluster analysis

Although there is substantial research on eating patterns, little effort has been paid to developing a classification of eating behavior applicable to the general population, rather than to people seeking help for obesity or eating disorders. Using cluster analysis, this study identified six types of eating patterns among normal volunteers. One hundred and sixteen females and 70 males completed a questionnaire concerning weight history, food intake patterns, use of satiation cues, and attitudes toward weight gain. Subjects also completed the Restraint Scale (Herman & Polivy, 1975). Height and weight were measured. Factor analysis reduced the questionnaire to nine internally reliable and meaningful scales; these were then entered into a K-means cluster analysis of subjects. Of the six clusters, two represented mild forms of disordered eating, two could be considered to represent more regulated eating styles, and two were distinguished by differential sensitivity to internal satiation cues. Construct validity of clusters was explored against gender, degree of overweight and scores on the Restraint Scale. Discussion focuses on the values of a cluster analytic technique to identify multidimensional patterns of food intake.