Trends in dietary intake and physical activity level in female students (1988 to 2003) after excluding under-reporters, using six different methods to identify under-reporters

Abstract

Background - Exclusion of under-reporters (UR) have affected conclusions in some studies. Objective - To investigate trends in non-under-reporters (NUR) using different methods to identify/exclude UR. Design. Analysis of recorded energy (EI) and macronutrient intakes and activities in 887 female university students, 1988 to 2003. Physical activity level (PAL) and energy expenditure (EE) were determined using a factorial method. Methods used to identify UR were; (A) EI:EE <0.76; (B) Goldberg method (I) using subjects classified into three activity categories (low, PAL=1.56; medium, PAL=1.64; high, PAL=1.82); (C) Goldberg method (I) using PAL=1.55; (D) EI:BMR <1.1; (E) EI:BMR <127; (F) EI:EEp <0.70, where EEp was EE predicted from an equation. Trends were determined using linear regression of median data for NUR for each year. Outcomes - The number ofNUR using each method (A-F) and the statistical significance of trends (P value) in dietary intake and PAL between 1988 and 2003 are shown in the table.