British Journal of Nutrition

Full Papers


Associations between the portion sizes of food groups consumed and measures of adiposity in the British National Diet and Nutrition Survey

Mary T. Kellya1, Kirsten L. Renniea1a2, Julie M. W. Wallacea1, Paula J. Robsona3, Robert W. Welcha1, Mary P. Hannon-Fletchera1 and M. Barbara E. Livingstonea1 c1

a1 Northern Ireland Centre for Food and Health (NICHE), University of Ulster, Cromore Road, Coleraine BT52 1SA, UK

a2 Unilever Corporate Research, Colworth House, Sharnbrook, Bedfordshire MK44 1LQ, UK

a3 School of Public Health, University of Alberta, Edmonton, AB, Canada T6G 2G3


The objective of the present study was to examine the associations between the portion sizes of food groups consumed with measures of adiposity using data from the National Diet and Nutrition Survey of British adults. Seven-day weighed dietary records, physical activity diaries and anthropometric measurements were used. Foods eaten were assigned to thirty different food groups and analyses were undertaken separately for men and women. The median daily portion size of each food group consumed was calculated. The potential misreporting of dietary energy intake (EI) was identified using the following equation: EI − estimated energy requirements × 100 = percentage of under-reporting (UR) of energy needs. Multinomial logistic regression (adjusted for age, social class, physical activity level and UR) was used to determine the portion sizes of food groups most strongly associated with obesity status. Few positive associations between the portion sizes of food groups consumed and obesity status were found. However, UR was prevalent, with a median UR of predicted energy needs of 34 and 33 % in men and women, respectively. After the adjustment was made for UR, more associations between the food groups and obesity status became apparent in both sexes. The present study suggests that the true effect of increased portion size of foods on obesity status may be masked by high levels of UR. Alternatively, these data may indicate that an increased risk of obesity is not associated with specific foods/food groups but rather with an overall increase in the range of foods and food groups being consumed.

(Received February 06 2008)

(Revised August 05 2008)

(Accepted August 05 2008)

(Online publication October 10 2008)


c1 Corresponding author: Professor M. Barbara E. Livingstone, fax +44 28 70323023, email


Abbreviations: EER, estimated energy requirements; EI, energy intake; NDNS, National Diet and Nutrition Survey; UR, under-reporting; WAL, waist action level