Multivariate analysis of diet in children at four and seven years of age and associations with socio-demographic characteristics

Eur J Clin Nutr. 2005 Jun;59(6):751-60. doi: 10.1038/sj.ejcn.1602136.


Aim: We have previously reported on distinct dietary patterns obtained from principal components analysis (PCA) of food frequency questionnaires from 3-y-old children. In this study, we repeat these analyses at 4 and 7 y of age.

Design: As part of regular self-completion questionnaires, the primary source of data collection in the Avon Longitudinal Study of Parents and Children, parents were asked to record the frequency of consumption of 57 different food types for their children. A total of 9550 subjects (68% of original cohort) were available from the 4-y data sweep and 8286 (59%) from the 7-y.

Methods: Distinct dietary patterns were identified at each age using PCA. These were then related to social and demographic characteristics of the parent and child.

Results: Three dietary patterns were established cross-sectionally at both 4 and 7 y of age to best describe the types of diet being consumed. One component at both time points described a diet based on 'junk'-type foods with high-fat and sugar content, processed and convenience foods. A second described the 'traditional' British diet based on meat, potatoes and vegetables. The final, a 'health-conscious' pattern was associated with vegetarian style foods, rice, pasta, salad and fruit. At both time points, the 'junk' pattern was significantly more likely in white children, where maternal education level was low and where the child had more siblings. The 'traditional' pattern was more likely in girls, where the mother had a partner and in nonvegetarians (both mother and child). The 'health-conscious' pattern was more likely with increasing levels of education and increasing maternal age.

Conclusion: We have demonstrated consistent dietary patterns in cross-sectional analyses at two ages in these children with similar socio-demographic associations evident at each age. Future analyses will track these dietary patterns over time and form a basis for the study of a variety of childhood outcomes.

MeSH terms

  • Adult
  • Child
  • Child Nutritional Physiological Phenomena*
  • Child, Preschool
  • Cohort Studies
  • Cross-Sectional Studies
  • Diet / economics
  • Diet / standards*
  • Diet Surveys
  • Educational Status
  • England
  • Family Characteristics
  • Female
  • Humans
  • Longitudinal Studies
  • Male
  • Mothers / education
  • Mothers / psychology*
  • Multivariate Analysis
  • Socioeconomic Factors*
  • Surveys and Questionnaires