Estimated portion sizes in a school-aged population.

OBJECTIVE To produce study-specific portion sizes for 11-year-old children in a population-based birth cohort and to compare these study-specific portion sizes with previously published children's portion sizes, to assess their relevance today. DESIGN Two multiple-pass 24 h dietary recalls were taken. The Food Standard Agency's photographic food atlas was used to quantify intakes. Study-specific food portion sizes were calculated for each food group. Portion sizes were calculated for all children and separately for boys and girls. The nutrient intake from the 24 h dietary recalls was analysed using study-specific and published portion sizes for individual participants. Agreement was assessed using Pearson's correlation, intra-class correlation coefficients and the Bland-Altman method. SETTING Birth cohort study, UK. SUBJECTS Children (mean age 11.3 years, n 264) and parents/guardians. RESULTS A total of 124 food portion sizes were calculated. Differences in portion weights between boys and girls were seen only for seven food items. There was a significant positive relationship (P < 0.001) between intakes of each nutrient as determined by the two sets of portion sizes. Correlation coefficients ranged from 0.77 (protein) to 0.98 (β-carotene). The intra-class correlation coefficients showed good agreement between nutrient intakes determined by the study-specific and published portion sizes (P < 0.001). CONCLUSIONS Nutrient intakes calculated using portion sizes from our population were similar to those calculated from portion size data collected in a national survey, despite being collected over a decade later. The present study adds to the small amount of evidence regarding portion sizes in UK children and shows agreement with previously published paediatric portion sizes.