The proposed analyses are exploratory in nature and will use simple methods in an attempt to idenitfy groups of children who follow certain dietary patterns throughout childhood. Dietary patterns have been obtained using Principal Components Analysis at the ages of 3, 4, 7 and 9 years in the ASLPAC children (1-3). For each child a continuous score was obtained for each dietary pattern. It is envisaged that there will be children who consistently score highly on individual patterns, e.g. the 'health conscious' or 'processed' patterns over time. On the other hand there are likely to be children whose diets have changed, for example, they may have initially scored highly on the 'processed' pattern and low on the 'health conscious' pattern in early childhood but reversed this trend through mid-childhood. Identifying such groups may be useful for examining the longitudinal associations between diet and a wide variety of health outcomes, by offering an alternative dietary exposure.

The preferred method for assessing trajectories of dietary patterns throughout childhood would be to use a technique such as latent profile or latent trait modelling. However, specialised software is required and time does not allow adequate learning of this application: It is anticipated that preliminary results will be presented at the ICDAM7 conference in June 2009 (abstract deadline is November 13th). We therefore propose to apply cluster analysis to the dietary pattern scores. This technique places individuals into distinct, non-overlapping groups based on the similarity of their dietary pattern scores. We will also investigate grouping children "by hand", based on quintiles of dietary pattern scores and compare the results obtained.

Depending on the results we hope to investigate the associations between a wide variety of variables and the 'trajectories' obtained, these will include social and parental lifestyle factors, feeding difficulties and other behaviours. This exploratory analysis will also set the scene for future analyses using the optimal methods of latent profile/trait modelling.

1. K North, P Emmett and the ALSPAC Study Team. Multivariate analysis of diet among three-year old children and associations with socio-demographic characteristics. European Journal of Clinical Nutrition 2000; 54: 73-80.3.

2. K Northstone, P Emmett and the ALSPAC Study Team. Multivariate analysis of diet in children at four and seven years of age and associations with socio-demographic characteristics. European Journal of Clinical Nutrition 2005; 59: 751-760.4.

3. K Northstone, PM Emmett. Are dietary patterns stable throughout early and mid-childhood? A birth cohort study. British Journal of Nutrition 2008. AOP: doi: 10.1017/S0007114508968264.5.