Background, purpose and research design by qws18475


									1         Background, purpose and research design

This chapter describes the background to the National Diet and Nutrition Survey (NDNS) of
adults aged 19 to 64 years, its main aims and the overall sample and research designs and
methodologies. The next chapter covers response to the survey and the appendices give a
more detailed account of the various methodologies for the different components of the

Results from this NDNS will be published in four volumes with a separate summary volume.
They will cover food and nutrient intake data derived from the analyses of dietary reports,
and data on nutritional status from physical measurements, including anthropometric data,
blood pressure, physical activity and the analyses of the blood and urine samples1.

1.1       The National Diet and Nutrition Survey Programme
The National Diet and Nutrition Survey programme is a joint initiative between the Food
Standards Agency and the Department of Health (DH). The programme was established in
1992 by the Ministry of Agriculture, Fisheries and Food (MAFF) and DH, following the
successful completion and evaluation of the benefits of the first survey of this type, of the diet
and nutritional status of British adults aged 16 to 64 years carried out in 1986/87 (1986/87
Adults Survey)2. MAFF’s responsibility for the NDNS programme transferred to the Food
Standards Agency on its establishment in April 2000.

The NDNS programme aims to provide comprehensive, cross-sectional information on the
dietary habits and nutritional status of the population of Great Britain. The results of the
surveys within the programme are used to develop nutrition policy at a national and local
level and to contribute to the evidence base for Government advice on healthy eating.

The NDNS programme is intended to:

      • provide detailed quantitative information on the food and nutrient intakes, sources of
          nutrients and nutritional status of the population under study as a basis for
          Government policy;
      • describe the characteristics of individuals with intakes of specific nutrients that are
          above and below the national average;
      • provide a database to enable the calculation of likely dietary intakes of natural
          toxicants, contaminants, additives and other food chemicals for risk assessment;

      • measure blood and urine indices that give evidence of nutritional status or dietary
         biomarkers and to relate these to dietary, physiological and social data;
      • provide height, weight and other measurements of body size on a representative
         sample of individuals and examine their relationship to social, dietary, health and
         anthropometric data as well as data from blood analyses;
      • monitor the diet of the population under study to establish the extent to which it is
         adequately nutritious and varied;
      • monitor the extent of deviation of the diet of specified groups of the population from
         that recommended by independent experts as optimum for health, in order to act as a
         basis for policy development;
      • help determine possible relationships between diet and nutritional status and risk
         factors in later life;
      • assess physical activity levels of the population under study; and
      • provide information on oral health in relation to dietary intake and nutritional status.

The NDNS programme is divided into four separate surveys planned to be conducted at
about three-yearly intervals. Each survey is intended to have a nationally representative
sample of a different population age group: children aged 1½ to 4½ years; young people
aged 4 to 18 years; people aged 65 years and over, and adults aged 19 to 64 years. The
Reports of the NDNS of children aged 1½ to 4½ years, of people aged 65 and over, and of
young people aged 4 to 18 years were published in 1995, 1998 and 2000 respectively3, 4, 5.

1.2      The need for a survey of adults
The last national survey of diet and nutrition in adults was the Dietary and Nutritional Survey
of British Adults carried out in 1986/87, thereafter referred to as the 1986/87 Adults Survey.
The changes in eating habits and lifestyles noted in that survey have continued throughout
the intervening years.        Increasing numbers of people are travelling and taking holidays
abroad, and with increased multi-culturism this has led to a greater variety of foods available.
Increasing demands on people’s time and longer working hours have led to greater demand
and availability of pre-prepared and convenience foods. There has also been an increase in
eating out of the home. There is a need, therefore, to assess the impact of such changes on
diet and nutrition among adults, to update the findings of the 1986/87 Adults Survey and to
complete the NDNS cycle by conducting a survey on adults aged 19 to 64 years.

One of the major uses of the NDNS data is for food chemical risk assessment. The
availability of up-to-date data on food consumption is important to ensure that estimates of
dietary exposure to food chemicals are as accurate as possible.

The Food Standards Agency and DH commissioned the Social Survey Division of the Office
for National Statistics (ONS) and the Medical Research Council Human Nutrition Research,
Cambridge (HNR) to carry out this survey of adults. Staff at HNR were responsible for
obtaining ethics approval for the survey from the Multi-centre Research Ethics Committee
(MREC) and National Health Service Local Research Ethics Committees (LRECs). They
were also responsible for recruiting the blood takers (phlebotomists), and dealing with those
aspects of the survey concerned with the venepuncture procedure and urine samples, and
for the analysis of the blood and urine samples that were collected. A survey doctor was
employed by HNR principally to liaise with and deal with questions from LRECs, to provide
support for ONS fieldworkers and the phlebotomists in the event of any medical problem
arising, to report all clinically significant blood results and blood pressure along with any
abnormal blood pressure and blood results to the respondent and the respondent's GP (if
appropriate).   The survey doctor was also available to answer any questions from
respondents on the venepuncture, urine collection and blood pressure procedures6.
Professor Angus Walls from the University of Newcastle-Upon-Tyne Dental School provided
training and support in the oral health component.      ONS, as the lead contractor, was
responsible for all other aspects of the dietary and oral health components of the survey,
including sample and survey design, recruitment and training of fieldworkers, data collection
and analysis.

1.3    The aims of the survey
The survey was designed to meet the overall aims of the NDNS programme in providing
detailed information on the current dietary behaviour, nutritional status and oral health of
adults living in private households in Great Britain.

The survey design needed to incorporate methods for collecting detailed information on the
respondent’s household circumstances, general dietary behaviour and health status, on the
quantities of foods consumed, and on physical activity levels, anthropometric measures,
blood pressure levels and blood and urinary analytes. Additionally an oral health component
was needed to collect information on oral health behaviour and on the number of teeth and
amalgam fillings7.

1.4    The sample design and selection

A nationally representative sample of adults aged 19 to 64 years living in private households
was required.     It was originally estimated that an achieved sample of about 2,000
respondents was needed for analysis and to ensure comparisons could be made with the
1986/87 Adults Survey.

As in previous surveys in the NDNS series, fieldwork was required to cover a 12-month
period, to cover any seasonality in eating behaviour and in the nutrient content of foods, for
example, full fat milk. The 12-month fieldwork period was divided into four fieldwork waves,
each of three months duration8. The fieldwork waves were:

       Wave 1: July to September 2000
       Wave 2: October to December 2000
       Wave 3: January to March 2001
       Wave 4: April to June 2001

Where there was more than one adult between the ages of 19 and 64 years living in the
same household, only one was selected at random to take part in the survey. As well as
reducing the burden of the survey on the household, and therefore reducing possible
detrimental effects on co-operation and data quality, this reduces the clustering of the sample
associated with similar dietary behaviour within the same household and improves the
precision of the estimates.

The sample was selected using a multi-stage random probability design with postal sectors
as first stage units. The sampling frame included all postal sectors within mainland Great
Britain, and selections were made from the small users’ Postcode Address File. The frame
was stratified by 1991 Census variables.

A total of 152 postal sectors was selected as first stage units, with probability proportional to
the number of postal delivery points, and 38 sectors were allocated to each of the four
fieldwork waves. The allocation took account of the need to have approximately equal
numbers of households in each wave of fieldwork, and for each wave to be nationally
representative. From each postal sector 40 addresses were randomly selected9.

Eligibility was defined as being aged between 19 and 64 and not pregnant or breastfeeding
at the time of the doorstep sift. The diet and physiology of pregnant or breastfeeding women
is likely to be so different from those of other similarly aged women as to possibly distort the
results. Further, as the number of pregnant or breastfeeding women identified within the

overall sample of 2000 would not be adequate for analysis as a single group, it was decided
that they should be regarded as ineligible for interview.

A more detailed account of the sample design is given in Appendix D. True standard errors
and design factors for the main classificatory variables used in the analysis of the survey
data are given in each of the individual volumes.

1.5     The components of the survey
These were as follows:

      • an initial face-to-face interview using computer assisted personal interviewing
        methods (CAPI) to collect information about the respondent’s household, their usual
        dietary behaviour, including foods avoided and reasons for doing so, use of salt at the
        table and in cooking, the use of artificial sweeteners and consumption of herbal teas,
        smoking and drinking habits, their health status, their use of fluoride and dietary
        supplements, herbal remedies and medicines, socio-economic characteristics, and
        for women in defined age groups, use of contraceptives, menopausal state and use
        of hormone replacement therapy;
      • a seven-day weighed intake dietary record of all the food and drink consumed by the
        respondent both in and out of the home;
      • a record of the number of bowel movements the respondent had over the seven-day
        dietary recording period;
      • a seven-day physical activity diary collected over the same period as the dietary
      • anthropometric measurements: standing height, body weight, waist and hip
      • blood pressure measurements;
      • 24 hour collection of urine;
      • if consent was given, a venepuncture procedure to collect a sample of blood for
        analysis of nutritional status indices;
      • a short post-dietary record interview, using CAPI, to collect information on any
        unusual circumstances or illness during the period which might have affected eating
      • self-completion Psychological Restraint Questionnaire (Eating Habits questionnaire)
        to assess under-reporting asked at post-dietary record interview;
      • self count of teeth and amalgam fillings;

      • a face-to-face interview, using CAPI, to collect information on the respondent’s oral
        health behaviour7;
      • collection of a sample of tap water from the respondent’s home for analysis of

While the aim was to achieve co-operation with all the various components, the survey
design allowed for the respondent to participate in only some components.

Ethics approval was gained for the feasibility and mainstage survey from a Multi-centre
Research Ethics Committee (MREC) and National Health Service Local Research Ethics
Committees (LRECs) covering each of the 152 sampled areas (see Appendix N for further
details of the ethics approval procedures).

As a token of appreciation a gift voucher for £10 was given to the respondent if the dietary
record was kept for the full seven days11. Each respondent was also given a record of his or
her anthropometric and blood pressure measurements. Results of a number of the blood
analyses were also reported to the respondent at approximately 6 weeks and 12 months
after the interview (see Appendix L).

Copies of the fieldwork documents and the interview questions are given in Appendix A.

Feasibility work carried out between September and December 1999 by the Social Survey
Division of ONS and the Medical Research Council Human Nutrition Research tested all the
components of the survey and made recommendations for revisions for the mainstage. For a
subgroup of the feasibility study sample the validity of the dietary recording methodology was
tested using the doubly labelled water methodology to compare energy expenditure against
reported energy intake. Further details of the design and results of the feasibility study are
presented as Appendix C.

The results of the feasibility study need to be regarded with some caution. Restrictions
placed on recruitment procedures by the MREC resulted in a much reduced response rate.
It is, therefore, possible that those who did co-operate in the feasibility study were
characteristically different from the general population, for example, in that they were more
interested in their diet and had more time to give to the survey.

1.6     Fieldwork

Over the fieldwork period a total of 88 ONS interviewers worked on the survey, the majority
working in at least two waves. All the interviewers working on the survey had been fully
trained by the Social Survey Division of ONS and most had experience of working on other
surveys in the NDNS programme, or of other surveys involving record keeping such as the
National Food Survey (NFS)12.

Each interviewer attended a five-day residential briefing before starting fieldwork.            The
briefing was conducted by research and other professional staff from the Social Survey
Division of ONS, from HNR, and staff from the Food Standards Agency and DH. Professor
Angus Walls from Newcastle-Upon-Tyne Dental School instructed interviewers on the
rationale and protocol for the self-count of teeth and amalgam fillings. Prior to the residential
briefing each interviewer was required to keep and code his or her own three-day weighed
intake record. Following the residential briefing all interviewers were required to complete a
post-briefing exercise. This involved asking a friend or relative to complete a three-day
weighed intake diary, and the interviewer coding the diary. Successful completion of this
exercise was a requirement for beginning fieldwork.

At the briefing interviewers were trained in all aspects of the survey and received individual
feedback from the nutritionists on their record-keeping and coding. The main components
covered by the training were:

    • the sample and selecting the respondent;
    • obtaining consents;
    • the questionnaire interview, in particular how to deal with certain ‘sensitive’ topics;
    • completing the weighed intake dietary record;
    • checking, probing and coding the dietary record;
    • collecting the physical activity information;
    • techniques for making the anthropometric measurements and measuring blood
    • the record of bowel movements;
    • the 24-hour urine sample;
    • collecting the tap water sample;
    • the procedures for obtaining a blood sample;
    • the oral health interview, in particular instructions on completing the self-count of
       teeth and amalgam fillings.

Emphasis was placed on the need for accuracy in recording and coding and in measurement
techniques. Practical sessions gave interviewers the opportunity to practice the
anthropometric measurements, coding food items, completing and checking diaries, and the
self-count of teeth and amalgam fillings.

Phlebotomists attended for the last two days of the residential briefings (see Appendix N).

In addition to the residential briefings, written instructions were provided for all interviewers
and for the phlebotomists who would be taking the blood samples. Interviewers working on
non-sequential fieldwork waves were recalled for a one-day refresher briefing to maintain the
accuracy of diary and brand coding and anthropometric and blood pressure measurement

In order that appropriate official bodies and personnel were informed about the nature of the
survey, letters were sent by ONS, prior to the start of fieldwork, to Chief Constables of Police,
Directors of Social Services and Public Health and to Chief Executives in Health Authorities
with responsibility for one or more of the selected fieldwork areas (postal sectors). The letters
gave information on when and where the survey would take place, what was involved in the
survey and asked that appropriate personnel at a more local level be informed. Copies of
these letters are reproduced in Appendix B.

In keeping with SSD normal fieldwork procedures, a letter was sent to each household in the
sample in advance of the interviewer calling, telling them briefly about the survey (see
Appendix A).

1.7    Plan of the report
Given the wealth of data collected in this NDNS, it was decided to publish the findings in a
number of separate topic reports rather than one substantive report. This has the advantage
of making some data available much earlier than it would otherwise be, and allows those with
specific interests to select the volume(s) most appropriate for their needs.

These methodology chapters and appendices have not been published as a separate
volume. They appear here on the Food Standards Agency website and a summary is
included in each published volume. The next chapter in this report gives response data for
the various components in the survey and describes the characteristics of the responding
sample. This report then describes the methodologies and procedures used in the survey,
including the seven-day weighed intake record (Appendix F), the physical activity diary

(Appendix I), anthropometry and blood pressure measurements (Appendix J), obtaining the
urine and tap water samples (Appendix P) and the venepuncture procedure (Appendix O).

Details of the weighing and recording procedures and subsequent coding and editing of the
dietary records are given, including details of the procedures for collecting information about
items consumed out of the home. The purpose and choice of anthropometric measurements
made and the techniques and instruments used are reported. The reasons for the choice of
blood pressure monitor are discussed and the protocol for taking the measurements is
described. The purpose of the venepuncture procedure and the protocol is described. An
account of the laboratory processing procedures and the quality control methods and data
are given in Appendix O. Appendix P explains the reasons why a 24-hour urine collection
was made and gives details of the equipment used.

The substantive results from the survey are presented in four separate volumes, with a fifth
summary volume. The first three volumes are primarily concerned with food and nutrient
intake data derived from the analyses of the dietary records and the results are presented for
different socio-demographic groups in the overall responding sample, for example by age
group, sex, region and household receipt of certain state benefits. In all volumes the data
presented are based on the samples of respondents co-operating with the relevant aspect of
the survey rather than those who completed all components.

The first volume covers the types and quantities of foods consumed by the different socio-
demographic groups.       The second volume reports on energy intakes, intakes of
carbohydrates, protein and alcohol and of fats and fatty acids. The third volume reports on
average daily intakes of vitamins and minerals, from food sources alone and from all
sources, including any dietary supplements being taken. The chapter on minerals also
includes results from the analyses of the urine samples. Throughout the second and third
volumes actual intakes are compared with dietary reference values, where appropriate.

The fourth volume covers physical measurements, that is the anthropometric data and
derived indices, blood pressure measurement and the analyses of the blood samples. The
anthropometric data (height, weight, waist and hip circumferences, and derived indices) and
blood pressure data are compared with measurements recorded on other surveys. Other
characteristics of the respondent associated with the anthropometric measurements and
blood pressure measurements are assessed in regression analyses. The results from the
analyses of the samples of blood are presented and, where relevant, the associations
between dietary intakes and blood levels are examined, for example plasma vitamin C with

fruit and vegetable consumption. The fourth volume also includes information on the physical
activity results from the physical activity diaries.

In each volume, where appropriate, results are compared with those from other surveys
including the 1986/87 Adults Survey (see Appendix S for a summary).

A fifth volume will provide a summary of the findings in the other four substantive results

Inevitably, given the volume of data collected in the survey and the potential range of
analyses, the individual volumes can only present initial findings. They are therefore largely
concerned with providing basic descriptive statistics for the variables measured and their
association with social, demographic and behavioural characteristics of the sample
population. It has only been possible to present a limited amount of data on the associations
between the dietary, physiological, biochemical and activity data.

Like previous surveys in the NDNS programme, a copy of the survey database, containing
the full data set will be deposited with The Data Archive at the University of Essex following
publication of the final summary volume. Independent researchers who wish to carry out their
own analyses should apply to the Archive for access13.

References and endnotes

        The volumes in the series cover:
        (i)     Types and quantities of foods consumed, to be published Autumn 2002;
        (ii)    Macronutrient intakes (energy, protein, carbohydrates, fats and fatty acids and
                alcohol), to be published early 2003;
        (iii)   Micronutrient intakes (vitamins and minerals, including analysis of urinary analytes), to
                be published Spring 2003;
        (iv)    Nutritional status (blood pressure, anthropometry, blood analytes and physical
                activity), to be published Summer 2003;
        (v)     Summary report, providing a summary of the key findings from the four volumes, to be
                published Autumn 2003.
        Gregory J, Foster K, Tyler H, Wiseman M. The Dietary and Nutritional Survey of British
        Adults. HMSO (London, 1990).
        Gregory JR, Collins DL, Davies PSW, Hughes JM, Clarke PC. National Diet and Nutrition
        Survey: children aged 1½ to 4½ years. Volume 1: Report of the diet and nutrition survey.
        HMSO (London, 1995).

     Finch S, Doyle W, Lowe C, Bates CJ, Prentice A, Smithers G, Clarke PC. National Diet and
     Nutrition Survey: people aged 65 years and over. Volume 1: Report of the diet and nutrition
     survey. TSO (London, 1998).
     Gregory JR, Lowe S, Bates CJ, Prentice A, Jackson LV, Smithers G, Wenlock R, Farron M.
     National Diet and Nutrition Survey: young people aged 4 to 18 years. Volume 1: Report of the
     diet and nutrition survey TSO (London, 2000).
     Further details of the role and responsibilities of the survey doctor are given in Appendix O.
     Unlike the other NDNS surveys respondents were not asked to participate in a full dental
     examination. The oral component comprised an oral health interview and a self tooth and
     amalgam filling count. More details are provided in Appendix R.
     Because in some cases fieldwork extended beyond the end of the three-month fieldwork
     wave or cases were re-allocated to another fieldwork wave, cases have been allocated to a
     wave for analysis purposes as follows. Any case started more than four weeks after the end
     of the official fieldwork wave has been allocated to the actual quarter in which it started. For
     example, all cases allocated to Wave 1 and started July to October 2000 appear as Wave 1
     cases. Any case allocated to Wave 1 and started in November 2000 or later appears in a
     subsequent wave; for example a case allocated to Wave 1 which started in November 2000 is
     counted as Wave 2. All cases in Wave 4 (April to June 2001) had been started by the end of
     July 2001.
     Initially 30 addresses were selected within each postal sector. Results from Wave 1 indicated
     a higher level of age-related ineligibles than expected and a much lower response rate. In
     order to increase the actual number of diaries completed and to give interviewers enough
     work an additional 10 addresses were selected for Waves 2, 3 and 4.
     Analysis of the fluoride from the tap water samples will not be reported on in any of the four
     volumes of this NDNS.
     Gift vouchers were from WH Smith Ltd.
     Department for Environment, Food & Rural Affairs. National Food Survey 2000. TSO (London,
     For further information about the archived data contact:
              The Data Archive
              University of Essex
              Wivenhoe Park
              Essex CO4 3SQ
              Tel: (UK) 01206 872001
              Fax: (UK) 01206 872003


To top