Article Text
Abstract
Background Childhood overweight and obesity are linked to increased risk of chronic disease in later life and are a target for UK Government action. Development of short food frequency questionnaires with low participant burden could aid evaluation of initiatives to improve dietary quality in pre-school and primary school-aged children.
Methods The UK National Diet and Nutrition Survey (NDNS) is a general population survey in which carers of children from randomly-selected households were asked to complete a four-day food diary. Diaries were coded and intake frequencies derived for groups of nutritionally similar foods. Principal component analysis (PCA) was performed on frequencies of consumption separately for children of preschool age (1–3 years) and primary school age (4–11 years). Reduced-item diet scores were calculated by multiplying coefficients for the 12 most characteristic foods at each age by every individual’s standardised reported frequency of consumption in each age group. Diet scores calculated for both the full and 12-item scores were standardised. Willing participants provided a blood sample and beta carotene status was measured by high performance liquid chromatography.
Results 1,069 pre-schoolers and 2,565 primary school-aged children provided dietary data in the first eight years of the NDNS (2008–2016). PCA of intake frequencies of 129 food groups revealed a prudent dietary pattern in both age groups, characterised by high intakes of salad vegetables, tap water, fruit, and nuts and seeds, and low intakes of soft drinks, purchased chips, manufactured coated chicken and turkey products, crisps and savoury snacks, and purchased burgers and kebabs. The 12-item prudent diet scores were calculated based on foods with the six highest and six lowest coefficients. Spearman’s correlation between the full and 12-item prudent diet scores was 0.86 amongst preschoolers and 0.84 amongst primary school-aged children. The mean differences between the full and 12-item prudent diet scores were both 0.00SDs with Bland-Altman limits of agreement -1.05 to 1.05SDs in preschoolers and -1.10 to 1.10 SDs in primary school-aged children. Correlations were seen between the full prudent diet scores and beta-carotene status (rs=0.23 for preschoolers and rs=0.30 for primary school-aged children); these were only slightly attenuated for the 12-item prudent diet scores (rs=0.19 for preschoolers and rs=0.26 for primary school-aged children).
Conclusion Short 12-item prudent diet scores based on foods that characterise a prudent dietary pattern describe healthy diets at young ages. Age-specific 12-item FFQs derived from these items offer appropriate tools to describe diets while reducing participant burden.