|dc.description.abstract||Analysing dietary patterns is an alternative yet complementary approach to examining intakes of single nutrients, and can provide insight into relationships between the whole diet and risk of disease. Although several food frequency questionnaires have been validated for their ability to identify dietary patterns in adult populations, similar analyses do not appear to have been performed in toddlers. The aim of this thesis was to evaluate the relative validity and reproducibility of a food frequency questionnaire for identifying dietary patterns in toddlers aged 12-24 months, and for ranking toddlers into categories of dietary pattern scores.
We recruited 160 toddlers aged 12-24 months, and administered the food frequency questionnaire on two occasions, approximately five weeks apart. Five days of weighed food records per participant were collected on randomly assigned days over this same period. To identify dietary patterns, principal component analysis was conducted separately on data from the first and second administrations of the food frequency questionnaire, and from the five-day weighed food records. Relative validity was assessed by comparing the dietary patterns derived from the first administration of the food frequency questionnaire to those derived from the weighed food record. Cross-classification and Pearson correlation coefficients were used to evaluate relative validity. Reproducibility was assessed by comparing the dietary patterns derived from the first and second administrations of the food frequency questionnaire, using Pearson and intraclass correlation coefficients.
Three major dietary patterns were characterised from the principal component analysis that were similar across both administrations of the food frequency questionnaire, and the weighed food records. One pattern labelled “sweet foods and hot chips” was associated with high intakes of ‘sweet foods’, ‘hot chips and roast potato and kumara’, ‘spreads’, ‘processed meat’, ‘sweet drinks’, and ‘nutritive drinks’. The second pattern named ‘vegetables and meat’ was associated with high intakes of ‘vegetables’, ‘meat’, ‘eggs and beans’, and ‘fruit’; and a third pattern labelled “milk and fruit” was associated with high intakes of ‘milk and milk products’, and ‘fruit’, and inversely associated with ‘breast milk’ and ‘infant formula and toddler milk’. The first administration of the food frequency questionnaire correctly classified 41.8 - 50.3% of toddlers into the same quartile of pattern score as the weighed food record, for the three patterns. Pearson correlation coefficients between the first food frequency questionnaire and the weighed diet record ranged from 0.56-0.68 for the three patterns. For the reproducibility analysis, Pearson correlation coefficients and intraclass correlation coefficients between the first and second administrations of the food frequency questionnaire ranged from 0.71-0.72 for all three dietary patterns.
These results indicate that the food frequency questionnaire has acceptable relative validity. It is able to identify dietary patterns in 12-24 month old New Zealand toddlers, and has a good ability to rank toddlers into categories of dietary pattern score. The patterns identified by the first administration of the food frequency questionnaire were highly reproducible.||