SELF-REPORTED DIETING - HOW SHOULD WE ASK - WHAT DOES IT MEAN ASSOCIATIONS BETWEEN DIETING AND REPORTED ENERGY-INTAKE

Objectives: To determine whether self-reports of dieting, to control w eight are associated with reported energy intake, how this association varies with the phrasing of questions on dieting behaviors, and wheth er this association differs by educational level and weight status amo ng adult male and female respondents. Methods: The study population in cluded 996 women and 227 men, aged 20-45, who volunteered to participa te in a weight gain prevention trial. Participants completed surveys a t baseline regarding their dieting behaviors and nutritional intake. R esults: The association between self-reported dieting and energy intak e varied according to the phrasing of the questions assessing dieting behaviors. Multi-item scales and nonambiguous single-item questions (e .g.,''current dieting'') were more strongly associated with reported e nergy intake than more general single-item questions (e.g., ''doing an ything to lose weight''). Overweight dieters reported lower energy int ake than overweight nondieters. Among nonoverueight persons, associati ons between dieting and energy intake were not significant The associa tion between dieting and energy intake did not differ by educational l evel among women. Among men, dieting predicted lower energy intake in those with low educational levels, although the number of men with low educational levels who were dieting was small. Reported dieting was n ot associated with energy intake among men with higher educational lev els. Discussion: Associations between self-reports of dieting and repo rted energy intake vary according to the phrasing of specific question s about dieting, gender, education, and weight status. These factors s hould be taken into account in the design of instruments for measuring these behaviors and in the interpretation of results, especially acro ss studies using different methodologies. (C) 1997 by John Wiley & Son s, Inc.