(View plain language abstract)
Diet is an important modifiable risk factor for obesity, which in turn is a major risk factor for the development of many cancers. It is difficult to elucidate the exact role of nutrition in obesity development, and subsequent cancer development due to the complexity of diet. The use of dietary patterns, reflecting total diet, is a complementary approach to studying diet-health associations that may prove important in understanding the aetiology of obesity and avenues for intervention.
Derive dietary patterns from the Avon Longitudinal Study of Parents and Children (ALSPAC); examine, longitudinally, the relationship between these patterns and changes in body fat/lean mass.
Dietary information has been collected on several occasions throughout childhood using Food Frequency Questionnaires (FFQs) and 3-day diet diaries at 7, 10 and 13 years of age. Dietary patterns assessed by Principal Components Analysis (PCA) and Cluster Analysis (CA) are the primary exposures. Total body fatness and lean tissue mass have been assessed at 9, 11, 13 and 15 years of age by Dual Energy X-Ray Absorptiometry (DXA).
A comprehensive set of dietary patterns have now been derived from FFQs and diet diaries. PCA patterns from FFQs tend to be: ‘Processed’, ‘Traditional’ and ‘Health conscious’ and from diet diaries ‘Health aware’, ‘Traditional’ and ‘Packed Lunch’. CA patterns from diet diaries were consistently ‘Processed’, ‘Healthy’, ‘Traditional’ and ‘Packed Lunch’. A number of issues have been tackled: a) PCA and CA derived patterns were comparable from both sources; b) Using food groups expressed as weights provided the best input data for those dietary patterns derived from diet diaries; c) Patterns were associated with nutrient intakes in the expected way; d) The ‘healthy’ pattern derived from cluster analysis was the most stable over time; e) The ‘processed’ pattern derived from the FFQ at age 3 appears to be important for predicting fat mass at age 9 in girls; f) Boys who were in an ‘unhealthy’ cluster at least twice between 7 and 13 years of age based on CA had the highest fat mass index (FMI) at the age of 15, for girls those who were in a ‘healthy’ cluster at all time points had lower mean BMI and FMI.
The methodological aspects of dietary pattern analyses are considerable and often overlooked by other researchers. An unhealthy diet associated with foods high in fat, sugar and salt has been linked to increasing fat mass in adolescents.
Obesity is a major risk factor for cancer. Diet is particularly important for the development of obesity, but specific aspects of the diet that contribute to obesity have been difficult to define. We do not eat individual foods/nutrients in isolation, so it may be useful to look at overall patterns of diet in order to better understand the causes of obesity. Our project will derive dietary patterns at various time points through childhood and examine any associations with body fatness up to the age of 15.
Participants have been weighed, measured and undergone body composition scans (measuring body fat and lean mass) four times between the ages of 9 and 15 years. Dietary intakes have been assessed at several time points since birth using a food frequency questionnaire (FFQ). In addition, 3-day diaries were collected at the ages of 7, 10 and 13 years. Dietary patterns have been examined using the statistical methods of Principal Components Analysis (PCA) and Cluster Analysis (CA). PCA uses the correlations that exist between foods to create patterns of food intake – scores are calculated for each pattern for each person; CA groups people together based on their similarities in food intake. Both types of patterns are labelled according to the foods that are most highly associated with them.
There are different types of diets consumed by children and adolescents, which can be crudely described as healthy or unhealthy based on their association with nutrient intakes. We have shown that an unhealthy diet (i.e. one associated with foods high in fat, sugar and salt) is linked to slightly greater mean body-fat levels in adolescents.