The Scottish Government

« Previous | Contents | Next »

Listen

REPORT OF THE WORKING GROUP ON MONITORING SCOTTISH DIETARY TARGETS

4. Surveys Taking Place in Scotland that Include Dietary Methodology

4.1 Review of Existing Surveys

4.1.1 The Working Group on Monitoring Scottish Dietary Targets commissioned a review of diet and health surveys currently taking place in Scotland and the UK (Wrieden W., Peace H., Armstrong J., 2003). Information from the surveys was extracted and collated under the headings listed below and a report was produced for the workshop in September 2003.

Headings Used to Extract Information from Surveys

• Survey / Study Objective

• Location

• Date

• Age group, gender and status (e.g. healthy, pregnant etc) of participants

• Number in sample (including Scottish sample number)

• Method used to obtain sample

• Dietary Methodology

• Any other comments

• Future of the Survey

4.1.2 The most relevant information and a summary of the advantages and disadvantages of each survey with respect to monitoring the Scottish Dietary Targets are included in this report in paragraphs 4.2.1- 4.8.9. The summary tables from the report can be found in Appendix 3 and the full report is available from the FSA library, Aviation House, London or www.food.gov.uk.

4.2 The National Diet and Nutrition Survey

4.2.1 The National Diet and Nutrition Survey (NDNS) was initially funded by the Ministry for Agriculture, Fisheries and Food (MAFF) and the Department of Health (DoH). Since 2000 the NDNS programme has been funded jointly by FSA and the DoH.

4.2.2 To date there have been five surveys carried out as part of the NDNS programme. The full objectives of the programme can be found in the reports of the individual surveys:

• The Dietary and Nutritional Survey of British Adults (Gregory J. et al., 1990)

• National Diet and Nutrition Survey: Children Aged 1- to 4- (Gregory J. et al., 1995)

• National Diet and Nutrition Survey: People Aged 65 Years and Over (Finch S. et al., 1998)

• National Diet and Nutrition Survey: Young People Aged 4-18 Years (Gregory J. & Lowe S., 2000)

• National Diet and Nutrition Survey: Adults Aged 19-64 Years (Henderson L. & Gregory J., 2003)

4.2.3 Each survey takes around five years from planning to publication and the time between collection of data on each age group does not reflect the speed of dietary changes. This time lag limits the capacity to track changes over time and assess trends in detail.

4.2.4 The NDNS uses a four to seven day weighed food intake and has been analysed to give nutrient intakes. The data can also be used to measure progress towards food targets.

4.2.5 The Scottish sample was 123 in the most recent NDNS (Henderson L. & Gregory J., 2003). This is inadequate for detailed analysis of the results by age or socio-economic group within Scotland, and is too small to compare Scotland to the rest of the UK. Therefore it is unsuitable for monitoring progress towards the targets in 2005. For future monitoring purposes, an enhanced sample for Scotland in any future surveys should be considered.

4.2.6 Following a recent review of the NDNS programme by the FSA, it is possible that a year on year rolling programme with cumulative sampling covering all population groups will be adopted. This survey design would provide more timely data and flexibility to respond quickly to new requirements, whilst allowing comparison with previous surveys. The continuous nature of a rolling programme would also provide an opportunity to develop a national identity for the programme in each of the four countries of the UK. This may help to increase the low response rate, which was 47% for the UK in the recent survey of adults (Henderson L. & Gregory J., 2003). The rolling programme would also allow the addition of targeted groups, such as the elderly and ethnic minorities, to be studied.

4.3 The Low Income Diet and Nutrition Survey

4.3.1 The LIDNS is a UK-wide survey that is funded by the FSA, with the enhanced Scottish sample funded by FSAS.

4.3.2 A feasibility study was completed in July 2002 and the main stage fieldwork for LIDNS began in November 2003. The survey uses multiple pass 24-hour recall dietary methodology which reduces respondent burden and under-reporting.

4.3.3 Statistical advice given to FSAS during the planning of the survey was that the minimum sample size required for inter-country comparisons between Scotland and the rest of the UK was 400. To carry out the same in depth analysis within Scotland as within England would require a sample size of 3000 which could not be justified. FSAS were advised that nothing would be gained from a sample size between 400-3000 in Scotland and therefore the minimal figure of 400 was chosen.

4.3.4 Although LIDNS would be capable of measuring the food and nutrient based targets, the survey sample is not representative for Scotland. Therefore LIDNS will not provide suitable data for monitoring progress towards the Scottish Dietary Targets.

4.4 National Food Survey / Expenditure and Food Survey

4.4.1 The NFS was established in 1940 and was replaced by the EFS in 2001. The data for the EFS is collected continuously from the whole of the UK and is reported annually (as was the NFS). The EFS is the joint responsibility of Defra and The Office for National Statistics for Great Britain (ONS).

4.4.2 The EFS uses till receipts and two-week food consumption diaries for each household member aged 7 and over to gather data. The data is presented per person, based on an average of all the household members weighted for the gender and age composition of the household. This means that the survey cannot be used to give separate information on children.

4.4.3 The EFS data is representative as the population of mainland Scotland is sampled (including north of the Caledonian Canal which the NDNS does not cover but the LIDNS does). However, as the data is collected at household level the EFS does not provide information about sub-groups such as children. In 2000 the number of Scottish households included in the main NFS was 548 (1320 individuals), similar to the number of households surveyed (586) in the recent EFS 2002-03. A detailed description of the EFS structure and response rate (58% for the 2002-03 survey) can be found in: 'Family Spending A Report of the 2002-03 Expenditure and Food Survey' (ONS, 2004).

4.4.4 EFS Family Food for 2001-02 and 2002-03 may be accessed on the following web addresses respectively:

( http://statistics.defra.gov.uk/esg/publications/efs/2002/default.asp). ( http://statistics.defra.gov.uk/esg/publications/efs/2003/default.asp).

The 2002-03 Report can also be obtained from The Stationary Office (Defra, 2004).

4.4.5 The data can be used to monitor progress towards all the nutrient and food-based targets in the Scottish Diet Action Plan with the exception of the NME sugar target for children.

4.4.6 It is possible that in 2008 the EFS will become part of a new continuous population survey. It is envisaged that this will lead to improved estimates of food and nutrient intakes and it is likely that some basic health outcomes will be included. Additional contextual information will be gathered so it may be possible to extract information for sub-samples. This work will be carried out by ONS and could potentially be used for monitoring the Scottish Dietary Targets beyond 2005.

* NFS data was summed for the years 1989-1991 to provide the baseline food and nutrient data presented in the Scottish Diet Report (Scottish Office 1993). Selected recommendations contained within the Scottish Diet Report were then carried forward into the SDAP.

4.5 The Scottish Health Survey

4.5.1 There have been two Scottish Health Surveys to date in 1995 and 1998 (Dong W. & Erens B., 1997 and Shaw A. et al., 2000), and fieldwork for the third SHS began in 2003. The survey covers a nationally representative sample of the population of approximately 13,000 subjects, and collects a range of information on health trends. Children were included in the sample for the first time in 1998 and the sample size was 3900. The sample size for the SHS 2003 will be similar to 1998.

4.5.2 The SHS Eating Habits Module asks about the frequency of consumption of a limited selection of foods aggregated into 25 food groups, including bread, breakfast cereal and fruit and vegetables. Due to the limited nature of the questionnaire it does not provide sufficient quantitative information to directly monitor progress towards the Scottish Dietary Targets.

4.5.3 The Working Group explored the possibility of utilising the current SHS data to obtain a measure of the nutrient-based SDAP targets but decided against this as:

• Some major food categories are missing from the eating habits module e.g. composite dishes such as pizza, quiche, soups and sauces.

• Detail regarding the type of food eaten is recorded for only four of the twenty-five foods groups.

• The portion size is given for less than half of the twenty-five foods groups included.

• There is a change in the food descriptions used each year for some foods e.g. vegetables, baked beans, tuna.

• There is a change in the fruit and vegetable consumption time frame in the 2003 survey.

4.5.4 A review process for the fourth SHS has recently begun and the timescale and elements of the survey are under discussion. The consultation process will assess demand and need for the SHS, review how extensively previous survey data has been utilised, determine views from local NHS Boards and ascertain what information is collected at local level. The process will also establish whether the current format should continue, whether there is demand for a more frequent survey or whether the focus of the survey should change on a rotational basis. These factors will affect the timescale of future studies and it is unlikely that there will be another survey before 2007 and certainly not before 2005.

4.5.5 The large representative sample size makes the SHS a robust framework on which to attach a dietary assessment methodology capable of monitoring the Scottish Dietary Targets. However the SHS questionnaire is currently at full capacity, so any additions to the existing eating habits module would be at the expense of other topics and would potentially increase the total cost of the survey. Competing priorities for the SHS will need to be carefully considered.

4.5.6 One possible approach would be to remove the current limited eating habits module from a future SHS and replace it with a more comprehensive dietary intake questionnaire. It was recognised that, realistically, a subset of the Scottish Health Survey participants may be asked to complete a comprehensive dietary intake questionnaire, though ideally this would be done for all participants.

4.5.7 It should be noted that if a completely revised eating module were considered some of the longitudinal comparisons with previous SHS would be lost.

4.6 Local Health and Lifestyle Surveys

4.6.1 The fifteen NHS Boards of Scotland carry out periodic Local Health and Lifestyle Surveys of a sample of their population (some have only done one or two while others have done several). These surveys gather information on the health-related knowledge, attitudes and behaviour of the population to assist with local planning. There is a large quantity of data available although it has not been collated and analysed for the whole of Scotland.

4.6.2 Many of the surveys incorporate core questions developed by the Scottish Needs Assessment Programme (SNAP), including some on dietary intake. However, there are inconsistencies between the Health Boards with respect to the dietary questions which makes it difficult to compare data between NHS Boards and to aggregate survey data across Scotland. Whilst some of the surveys may be helpful in monitoring progress towards the food-based targets none can measure the nutrient-based targets.

4.7 Market Research

4.7.1 There are several market research companies in the UK. TNS were invited to give a presentation to the September meeting of the Working Group.

4.7.2 TNS operates Great Britain's leading continuous consumer panel, collecting data since 1974. The TNS Super-Panel was launched in 1991 and consists of 15,000 households with around 1350 in Scotland. The sample is demographically and regionally balanced to give a representative picture of the marketplace. The households record all purchases they bring into the home using scanners and data is downloaded twice weekly to TNS via electronic terminals in the house.

4.7.2 The sample is recruited on a volunteer basis and selected to meet demographic and regional targets. Incentives for taking part are low (points for buying items from a brochure or music vouchers depending on age group) which raises questions about respondent bias in such a panel. It is potentially quite positive that respondents are prepared to do so much for so little apparent reward (apart from the 'common good').

4.7.3A wide range of standard and special analysis can be carried out on this data, including Scotland only analysis, but while the quantity of data is large, the quality may be a problem (as indicated above). The food-based targets could be measured but not the nutrient-based targets. Portion size is based on consumer input and there are problems with recording ingredients and home cooked foods.

4.7.4 TNS also manage the Family Food Panel Complete, which is the UK's largest and only continuous monitor of food and drink consumption. The panel is comprised of 11,000 individuals in 4,200 households who record all food and drink consumed for two weeks every six months. The survey also records the weight and height of respondents which can be used to look at body mass index (BMI) and obesity. Lifestyle and attitude questions can also be linked to show if people's diets are reflecting what their attitudes suggest.

4.7.5 The FSA recently initiated a research study (Elliot P. et al., 2003) to examine whether commercially available data sets on household food consumption could be used to carry out nutritional surveillance. The results of this feasibility study suggest that it would be possible to monitor food purchasing patterns reliably at household level using TNS data and these could be used to inform nutritional surveillance. As the data currently stands it would not be possible to correlate food purchases with health data.

4.7.6 The Working Group also noted the possibility of purchasing tailored questions on specific foods of interest from the Market Research UK Omnibus e.g. FSAS has already purchased questions on the consumption and purchase of fish.

4.8 Other Surveys

4.8.1 The surveys described in this section, while not sufficiently robust for monitoring the Scottish Dietary Targets, may provide useful additional proxy indicators or contextual information relevant to monitoring the Scottish Dietary Targets.

4.8.2 The Scottish Health Education Population Survey (HEPS) is an annual survey that was established in 1996 by NHS Health Scotland (there is a one year gap in data for 2000) with the aim of providing national level data on health related knowledge, attitudes, motivations and behaviours. HEPS contributes to monitoring the performance of health education / promotion in Scotland and provides information for planning future public health strategies.

4.8.2 There are two survey waves, in March and September each year, with a yearly total sample size of 1800 adults aged 16-74. The response rate is generally around 72%.

4.8.3 The questions relating to diet cover awareness of the recommended fruit and vegetable intake, changes to diet in the context of a healthy lifestyle, changes made to the diet in the last year, barriers to healthy eating and the frequency of eating five food groups:

• fruit, salad and vegetables not including potatoes

• starchy food e.g. bread, rice, pasta, potatoes

• sugar and foods containing a lot of sugar e.g. cakes, sweets, soft drinks

• fatty or fried foods e.g. crisps, chips

• fish (not fried).

4.8.4 HEPS has recently undergone a period of review and the decision has been taken to continue with the survey as it provides trend data on key indicators. The questions included will be thoroughly reviewed but it would be preferable to add questions rather than alter what is there so longitudinal trend data can still be obtained.

4.8.5 The Health and Health Behaviours of Scottish School Children survey is part of a WHO study and is managed by the Child and Adolescent Health Research Unit at the University of Edinburgh. It samples children aged 11, 13 and 15 and in 1998 the total sample size was 5632. The survey includes questions on dietary behaviours and asks about the frequency of consumption of bread, milk (and type consumed), fruit, vegetables, processed meat, chips, crisps, sweets, cakes, sugary drinks and coffee. This survey runs every four years with the next planned for 2006.

4.8.6 The Scottish School Adolescent Lifestyle and Substance Use Survey (SALSUS) is by design focused on smoking, drinking and drug use but does contain questions on healthy eating, body image and dieting. There are a few contextual questions and some which ask about the frequency of consumption of fruit and vegetables, sweets, cakes, biscuits, pastries, crisps, chips / other fried potatoes, soft drinks and alcohol. This survey was carried out in 2000, 2004 and there are plans for another survey in 2006.

4.8.7 The Scottish Executive Education Department (SEED) has the lead for a new survey - the Longitudinal Survey of Children. It is envisaged that the survey will collect data over a period of up to twenty years, tracking a number of cohorts of children. The aim will be to meet a wide variety of research purposes through a core set of questions supplemented by modules targeting specific issues. At the time of writing (2004) the questionnaires are being drafted. Piloting will take place towards the end of the year and fieldwork will begin in early spring 2005.

4.8.8 Four independent random population surveys were carried out in Scotland in 1986, 1989, 1992 and 1995 by the Cardiovascular Epidemiology Unit, University of Dundee, as part of the WHO MONItoring trends and determinants in CArdiovascular Disease (MONICA) programme. The surveys provide food and nutrient data on over 6000 Scots aged 25-74. The surveys used a 65 item semi-quantitative FFQ that was updated for specific new items as necessary. Care must be taken when comparing data from different years. No further dietary surveys are planned for Scotland with respect to the MONICA project.

4.8.9 The Aberdeen Prospective Osteoporosis Screening Study (APOSS), designed and run by the University of Aberdeen Department of Medicine and Therapeutics, began in 1990 with a sample of 5118 perimenopausal, early and late menopausal women aged 45-54. Diet at baseline was assessed using a validated FFQ designed specifically for the study of nutrients associated with bone health. It contains ninety-eight foods organised into twenty sections and contains sub-sections on both present and past food intake (up to twelve and thirty years previously). Additional sections of the questionnaire are designed to collect data on dietary supplement use, dietary habits and physical activity. This project is ongoing.

« Previous | Contents | Next »