An Exploratory Analysis of the Scottish Diet 2001-2009 Using Household Purchase Data

  • Karen Louise Barton

    Student thesis: Doctoral ThesisDoctor of Philosophy


    Robust, longitudinal measures of national food and nutrient intake are key for informing and monitoring public health policy on diet related health. In order to estimate food and drink consumption and assess diet quality in Scotland, household food purchase data, from 2001 to 2009, were analysed. Data from the Expenditure and Food Survey / Living Costs and Food Survey were adjusted for waste (and other appropriate factors), weighted to the Scottish population, and sampling methods were taken into account. Analyses focused on variations in intake by year, Scottish Index of Multiple Deprivation (SIMD), urban/rural residence and comparison to national dietary targets.

    There were few significant changes in consumption of foods or nutrients over the nine year period. Small increases were found in consumption of mean daily fruits and vegetables (259 to 279g), brown/wholemeal bread (18.2 to 21.4g) and breakfast cereal (19.6 to 23.2g). A decrease was found in the percentage of food energy from saturated fat (15.5 to 15.1%) and non-milk extrinsic sugars (NMES) (15.5 to 14.8%), concurrent with a reduction in whole milk and soft drink consumption, respectively. Mean diet quality index (DQI) score also improved (36.0 to 37.6%).

    In comparison with the most deprived SIMD quintile, the least deprived had the highest mean daily consumption of fruit and vegetables (200, 348g), brown/wholemeal bread (17.2, 26.5g), breakfast cereal (15.7, 27.0g), oil-rich (20.8, 39.9g) and white fish (77.2, 112g), the lowest consumption of total bread (105, 91.5g), and percentage of food energy from NMES (15.5, 14.3%) for 2007 to 2009. The most deprived had the highest energy density (737, 696kJ/100g) and the lowest diet DQI over the nine year period (e.g. 33.6, 41.4% for 2007 to 2009). Diet inequalities remained similar over the nine year period. With the exception of fish consumption, remote rural areas had a higher mean consumption of all target foods compared to urban areas, following adjustment for confounding variables. Principal component analysis identified 4, 5 and 6 dietary patterns explaining the highest amount of variance for each 3-year period respectively. Those explaining most variance, which were similar for each time period, were characterised as 'takeaway/eaten out', 'healthy with fruit and vegetables', and 'traditional'.

    None of the Scottish dietary targets were achieved by 2009 indicating the need for more effective public health policy. The small changes detected between 2001 and 2009 are encouraging but highlight the time taken for even modest changes in diet to occur and the impact of new policy initiatives outlined in more recent years has yet to be seen. It is of particular concern that foods targeted for increased consumption are significantly lower in the most deprived groups of the population.
    Date of Award2014
    Original languageEnglish
    SponsorsScottish Government
    SupervisorWendy Wrieden (Supervisor) & Annie Anderson (Supervisor)

    Cite this