Skip to main content

Table 1 Summary/review of 17 major repeated cross-sectional and longitudinal surveys conducted in England over the January, 1970 to December, 2018 period

From: A critical review of diet-related surveys in England, 1970-2018

Repeated cross-sectional surveys
  Living Cost and Food Survey (LCFS) Actives Lives Survey (ALS)
Survey Background The LCFS (formerly known as the Expenditure and Food Survey (EFS) prior to 2008) is the UK’s premier household expenditure survey, which captures information on the spending patterns and cost of living across the UK. The ALS, which replaced the Active People Survey in November, 2015, is a sport and recreation survey which measures physical activity levels of over 198,000 persons living across England.
Survey Design and Methodology Annual repeated cross-sectional survey. Sample selected using multi-stage stratified random sampling with clustering. Household addresses with small user postcodes are randomly selected from the Royal Mail’s postcode address file (PAF). Face-to face interviews (individual and household questionnaires administered) and 2-week self-reported expenditure diaries completed by all members of the household, aged 16 years and over. Simplified expenditure diaries are completed by children 7 to 15 years old. Annual repeated cross-sectional survey. Multi-stage stratified random sample. Each year, approximately 198,250 persons are targeted for inclusion in the survey. Household addresses are randomly selected from the Royal Mail’s postcode address file (PAF) and letters sent inviting up to two adults (16 years and over) per household to complete a questionnaire online or via post (for persons without internet access). Participants are rewarded with a £5 voucher from a range of retailers. During the adult survey, persons are asked if there are any 14–15-year olds in their household. Children aged 14–15 who ae interested and receive parental consent to participate in the study are contacted and asked to complete a young person questionnaire.
Target population and level of geography covered Families/households within the UK (England, Scotland, Northern Ireland and Wales). Data for England are available at the national and Government Office Region (GOR) level. Local authority level data can be made available upon request and approval by the UK Data Service. Individuals 14 years and older living in England during the 2015–2016 and 2016–2017. Data are available for the Government Office Region (GOR), County Sport Partnerships, Counties and Local Authority District level. The survey was designed to achieve a minimum annual sample size of 500 for each local authority, with the exception of the City of London and Isles of Scilly, in which the target sample size was 250.
Type of dietary assessment used Household food expenditure data captured in the Family Food Module of the survey are used as a proxy measure for food consumption. Single 24-h screener/brief/shortened instrument (fruit and vegetable only) completed online or via post.
Primary users of diet-related data Academics/Researchers and several Governmental Departments. The Family Food Module of the LCFS is primarily used by the Department for Environment Food and Rural Affairs (Defra) to monitor food consumption and to produce the annual Family Food Report (a report which provides estimates of nutrient content and statistics on household food purchases by food type). Academics/Researchers, Local Authorities, Public Health England (PHE)
Data Accessibility/Availability Data accessible through the UK Data Service. Data currently available for the 2008–2017/18 period Data accessible through the UK Data Service. Data currently available for the 2015–2016 and 2016–2017 survey periods.
Types of variables captured Socio-demographic information (age, sex, occupation, education), GOR, local authority level geography, data garnered from 2-week expenditure diary (expenditure on energy, bills, utilities and food). Socio-demographic information (age, sex, employment) and health measures such as obesity and fruit and vegetable consumption over a 24-h period.
Cost to access Not applicable Not applicable
Key features/potential benefits 1. Nationally representative annual survey with relatively large sample size (approximately 5000 households each year) 2. Two (2) week expenditure diaries (completed by each member of the household 16 years and over) detailing purchased quantities of food and drink are used to estimate food consumption in England. 3. Possible to make comparisons between low and high-income households. 1. Large sample size and a nationally representative sample of the English population. 2. Although focused on sport and recreation, the survey also includes data on fruit and vegetable consumption. 3. The availability of local authority data makes it possible to analyse dietary consumption below the regional (GOR) level.
Key considerations 1. Difficult to compare data prior to 2008 as a different survey methodology was used for the previous EFS. 2. Survey designed to capture household expenditure on food and quantities of food and drink purchased. The survey does not capture foods actually consumed by individuals. 1. The survey only captures self-reported fruit and vegetable consumption over a single 24-h period. 2. Difficult to compare data prior to 2015 as a different survey methodology was used for the previous Active People Survey.
  National Diet and Nutrition Survey (NDNS) Health Survey for England (HSE)
Survey Background The NDNS was originally established in 1992 as a series of four separate cross-sectional surveys, capturing information on: children ages 1 ½ -4 ½ years (1992–1993), young people 4–18 years old (1997), adults 19–64 years old (2000–2001) and persons 65 years and over (1994–1995). In 2008, the new NDNS Rolling Programme (RP) was introduced as a nationally representative repeated cross-sectional survey which captures information on the type and quantity of foods and beverages consumed by 1000 persons (500 adults and 500 children) annually in the UK. The HSE is an annual survey used to monitor and assess changes in the overall health and lifestyle of persons living within England.
Survey Design and Methodology Annual repeated cross-sectional survey. Multi-stage stratified random sample. Face-to face interviews conducted with respondents to capture food preparation, smoking and drinking habits. Self-completed 4-day food diaries are completed by persons 12 years and older and parents and/or carers are asked to complete food diaries for children 11 years and younger. Anthropometric measurements and blood and urine samples collected via nurse interview. Annual repeated cross-sectional survey. Multi-stage stratified random sample. Face-to face interviews, self-completed questionnaires and a follow-up nurse visit carried out to collect anthropometric measurements and blood samples.
Target population and level of geography covered Individuals 1 ½ years and older, residing in private households in the UK. Data for England are available at the national and Government Office Region (GOR) level. Adults (defined as persons 16 years and older) and children (0–15 years old) living in private households in England. Data available at the national, Government Office Region (GOR) and Strategic Health Authorities level. Local authority level data only available upon request and approval by NatCen Social Research at a cost.
Type of dietary assessment used Four (4) day food diary Food frequency questionnaire (FFQ) used prior to 2009. Single 24-h screener/brief/shortened instrument (fruit and vegetable only) used since 2009.
Primary users of diet-related data Academics/Researchers, policymakers, UK Health Departments, Scientific Advisory Committee on Nutrition’s (SACN), Food Standards Agency (FSA) and several Governmental Departments. Academics/Researchers, policymakers, the Department of Health & Social Care, Public Health England (PHE), NHS England, other NHS bodies, Local Authorities, charities and voluntary organisations. Data used to track the national achievement of the 5-A-Day, fruit and vegetable target.
Data Accessibility/Availability Data for the NDNS RP are accessible through the UK Data Service. Data currently available for the 2008–2016/17 period (survey wave 1–9). Data are accessible through the UK Data Service. Data currently available for the 1991–2017 period.
Types of variables captured Socio-demographic information (age, sex, occupation, education), GOR and all foods and beverages consumed over a 4-day period. Socio-demographic information (age, sex, occupation, education), GOR, general health, height and weight measurements and fruit and vegetable consumption.
Cost to access Not applicable No cost to access GOR level data but lower level geography (e.g. local authority level) can be accessed at a minimum cost of £1000.
Key features/potential benefits 1. Availability of annual food consumption data at the national level and 2. Detailed information available on all foods and beverages actually consumed by individuals over a 4-day period using the food diary method. 1. Nationally representative annual survey with large sample size of approximately 10,000 individuals (8000 adults and 2000 children). 2. Data captured could be used to explore relationships between diet (specifically fruit and vegetable consumption), obesity and associated chronic diseases.
Key considerations 1. Relatively small annual sample size compared to larger cohort studies which employ methods which are less tedious than the food diary method. 2. Difficult to compare data prior to 2008 with NDNS RP data, as a different survey methodology was used previously. This makes it difficult for comparisons to be made across the survey waves and for changes in diet to be assessed over time. 1. Significant changes (e.g. the complete omission of the fruit and vegetable module in the 2012 survey wave) have been made to the type of diet questions asked, which makes it difficult for comparisons to be made across the survey waves and for changes in diet to be assessed over time.
Food and You Survey
Survey Background Food and You is a random probability survey commissioned by the Food Standards Agency (FSA) every 2 y. The survey captures information on public attitudes and self-reported knowledge as it relates to food safety, production and other food-related issues.  
Survey Design and Methodology Bi-annual repeated cross-sectional survey. Multi-stage stratified random sample. Face-to face interviews conducted with adults, defined as persons aged 16 years and over.  
Target population and level of geography covered Adults (16 and over) residing in private households the UK. Data for England are available at the national and Government Office Region (GOR) level.  
Type of dietary assessment used Food frequency questionnaire (FFQ) conducted at each wave of the survey  
Primary users of diet-related data Academics/Researchers, policymakers and several Governmental Departments, particularly the Food Standards Agency (FSA)  
Data Accessibility/Availability Data are accessible through the UK Data Service. Data currently available for the five survey waves completed to date: 2010, 2012, 2014, 2016 and 2018.  
Types of variables captured Socio-demographic information (age, sex, occupation, education, household income), GOR, frequency of consumption of foods such as beef, poultry, burgers, ready meals, diary, fruits and vegetables.  
Cost to access Not applicable  
Key features/potential benefits 1. Nationally representative survey with sample size of about 3000–3500 individuals every 2 years
2. Besides data collected via FFQs, the survey also captures respondents’ knowledge of current dietary recommendations and perceptions of what constitutes a healthy and balanced diet.
 
Key considerations 1. Changes made to diet-related questions asked over the years, makes it difficult for comparisons to be made across the survey waves and for changes in diet to be assessed over time.  
Longitudinal surveys
  Southampton Women’s Survey (SWS) Born in Bradford (BIB)
Survey Background The SWS was established between 1998 and 2002 with the primary aim of measuring non-pregnant women aged 20–34 years living in Southampton (England) and to follow-up members of the cohort who subsequently became pregnant. The study’s major aim was to examine the effect of diet and lifestyle factors on the health of mothers and their children throughout the lifecourse. BIB is a study which tracks the health of over 13,500 children (and their parents) born at the Bradford Royal Infirmary between March 2007 and December 2008. The study tracks the health of these children from pregnancy throughout childhood and into adulthood.
Survey Design and Methodology Longitudinal Birth Cohort study. Pre-pregnancy home visits were made to 12,583 non-pregnant women (who were 20–34 years old during the 1998–2002 period) who resided in Southampton, England and surrounding areas. Pre-pregnancy food diaries were completed by participants and face-to-face interviews and blood samples were taken by a research nurse. Follow-up nurse visits were made to 3158 women who became pregnant and delivered a live born child; blood samples taken, and follow-up interviews conducted. Participants were asked to keep a food diary during early and late pregnancy. Follow-up surveys were conducted when children were 6 and 12 months and 3, 6–7, 8–9 and 11–13 years old. Longitudinal Birth Cohort study. Women who planned to be give birth during the 2007–2011 period were recruited and baseline data on socio-economic status, ethnicity and family trees, diet, physical and mental health were collected from 12,453 women at 26–28 weeks of pregnancy. Baseline data were also collected from 3448 partners of recruited mothers. Follow-up self-administered questionnaires were completed by partners at 6 and 12 months. Follow-up home visits were made with 2 sub-groups within the cohort when children were 6, 12, 18 months and 2, 3 and 4 years old to collect information on growth trajectories, risk factors for childhood obesity and exposures to asthma and atopy. Follow-up waves are heavily dependent on the level of funding available.
Target population and level of geography covered 12,583 non-pregnant women aged 20–34 years during the 1998–2002 period, living in Southampton (South East of England) and surrounding areas; 3158 women who became pregnant and delivered a live born child subsequent to recruitment and their children. Pregnant women (26–28 weeks) who delivered babies at the Bradford (North England) Royal Infirmary, fathers of the children and the children born to recruited mothers.
Geographical area captured: Bradford (North of England)
Type of dietary assessment used Interviewer administered FFQ and 24-h recall conducted at each survey wave, food diaries completed by mothers at pre-pregnancy, early pregnancy and when children were 3 years old and 24-h diet recalls administered when children were 6 months old. Food frequency questionnaire (FFQ) at each wave of the survey.
Primary users of diet-related data Academics/Researchers, Local Authority. Academics/Researchers, Local Authority, National Health Service (NHS).
Data Accessibility/Availability Data accessible through the MRC Lifecourse Epidemiology Unit, University of Southampton. Data available for women before pregnancy (1998–2002) and during early and late pregnancy. Data for children are available for 6 and 12 months, 3, 6–7, 8–9 and 11–13 years old. Data (and details regarding survey data currently available) accessible through the Bradford Institute for Health Research.
Types of variables captured Socio-demographic information (age, sex, occupation, employment, education), general diet, dietary changes and a 100-point FFQ asking the frequency of consumption in the last 3 months of fruits, vegetables, potatoes, rice, soft drinks, dairy, bread and a host of other foods across the various food groups. Socio-demographic information (age, sex, occupation, employment, education) and a more than 100-point FFQ asking the frequency of consumption in the last 2–3 months of fruits, vegetables and a host of other foods across the various food groups.
Cost to access Not applicable Not applicable
Key features/potential benefits 1. Food consumption data available for a wide variety of foods. 2. Cohort study data can be used to track changes over time. 3. Availability of pre- and post-pregnancy data. 1. Food consumption data available for a wide variety of foods 2. Cohort study data can be used to track changes over time. 3. Bradford has a large ethnic community and so the study captures ethnic minority groups which are usually underrepresented
Key considerations 1. Study not representative of English population; only focuses on Southampton (South of England). 2.The study only focuses on women and their children over time. 3. Complete data on children are not available for the entire cohort at each age of follow-up. 1.Study not representative of English population; only focuses on Bradford (North of England) 2. Changes made to diet-related questions across the survey waves may make it difficult to make comparisons over time. 3. Follow-up waves are heavily dependent on the level of funding available.
  Understanding Society British Cohort Study 1970 (BCS70)
Survey Background Understanding Society is an annual large-scale, multi-topic longitudinal cohort study established to understand social and economic changes in the UK at the individual and household level. The BCS70 is a large national longitudinal birth cohort study which tracks over 17,000 persons born in England, Scotland and Wales in a single week in 1970. The study has gathered information related to the health, social, economic and educational development of participants.
Survey Design and Methodology Annual Longitudinal/panel/cohort study. Multi-stage stratified random sample. The first wave was conducted in 2009 when over 40,000 households were selected. Since then, follow-up interviews have been conducted with the same individuals every 12 months. At each survey wave, one member of the household is asked to complete a household questionnaire and each person 16 years and older is interviewed and asked to complete a separate (self-completed) questionnaire. Members of the household aged 10–15 years (young people) are also asked to complete a separate (self-completed) paper or web-based/online questionnaire. Web-based surveys were introduced in wave 7 (2016) of the survey. Longitudinal Birth Cohort study. All children born in England, Scotland and Wales in 1970 were recruited and eight follow-up surveys have been conducted to date. Follow-up interviews were done when children were 5, 10, 16, 26, 30, 34, and 42 years of age (in 2012). Although data are not currently available, a follow-up survey was conducted at age 46 (in 2016) and information is currently being processed. In the 2004 study (age 34) cohort members were given a basic skills (numeracy and literacy) assessment test and a self-completion questionnaire to complete.
Target population and level of geography covered Individuals living within over 40,000 households in the UK. Data for England are available at the national and Government Office Region (GOR) level. Local authority level, Westminster Parliamentary Constituencies, Local Education Authorities and Travel to Work Areas are available upon request and approval by the UK Data Service under its Special License Agreement. Children born in England, Scotland and Wales in a single week in 1970.
Type of dietary assessment used Short food frequency screener/brief instrument which primarily captured fruit and vegetable consumption. 4-day food diary and a 24-h diet recall included in 1986 wave of survey. Online diet diary also included in the 2016 wave, when respondents were 46 years old.
Primary users of diet-related data Academics/Researchers. Academics/Researchers.
Data Accessibility/Availability Data accessible through the UK Data Service. Data currently available for the 2009–2018 period (survey wave 1–9) Data accessible through the UK Data Service. Data currently available for the 1975–2016-18 survey period.
Types of variables captured Socio-demographic information (age, sex, education, family, social life), self-reported health, type of milk, bread usually consumed, daily and weekly consumption of fruits and vegetables. Socio-demographic information (age, sex, occupation, education). Consumption of fruits, vegetables, meat, dairy, soup, potatoes, biscuits, crisps, fizzy drinks, sweets and ice-cream consumed over a 24-h period. All foods consumed over a 4-day period in 1986 (paper-based food diary) and in 2016 (online food diary) when respondents were 46 years old.
Cost to access Not applicable Not applicable
Key features/potential benefits 1. Large sample size, nationally representative and conducted annually. 2. Cohort study data can be used to track changes over time. 1. Large sample size and nationally representative 2. Cohort study data can be used to track changes over time. 3. Detailed information on all foods consumed by participants over several days were captured in food diaries conducted in the 1986 and 2016 wave of the survey.
Key considerations 1. Very few diet-related questions included in the study (fruit and vegetable consumption, dairy, bread). Questions posed in the main questionnaire primarily focused on the type of bread and milk consumed and portions of fruits and vegetables consumed in a typical week. 2. Differences in the number and types of diet-related questions asked across survey waves could make it difficult for comparisons to be made over time. 1. Food diary data for the 1986 and 2016 wave are being cleaned and the expected date of release is undetermined 2. Changes made to diet-related questions across survey waves could make it difficult for comparisons to be made over time.
  Avon Longitudinal Study of Parents and Children (ALSPAC) UK Women’s Cohort Study (UKWCS)
Survey Background ALSPAC also known as the Children of the 90s Study, is a study which tracks the health and well-being of 14,400 families living within the Bristol area. The UKWCS is a large-scale cohort study which explores the relationship between diet (including foods, nutrients and supplements) and health outcomes such as cancer, cardiovascular disease and obesity amongst over 35,000 middle aged women in the UK.
Survey Design and Methodology Longitudinal Birth Cohort study. Study posters were disseminated, and local community midwives discussed the study with women with expected deliveries between April 1991 and December 1992. Persons who contacted the study team were included in the study. Baseline data were captured during pregnancy and follow-up assessments carried out when children were 4 weeks to 24 years of age. Self-completed postal questionnaires were completed by mothers, children and teachers (of children) and clinical assessment visits were carried out at different stages of the study. Longitudinal Cohort study. Direct mail questionnaires were sent by the World Cancer Research Fund to persons, particularly women, living in England, Scotland and Wales, listed on direct mailing lists. Female survey responders aged 35–69, who self-identified as vegetarian or non-red meat eaters were included in the study. Baseline data were collected during the 1995–1998 period and follow up (known as phase 2 of the study) was done during the 1999–2002 period. Several sub-studies have been carried out over the years. For instance, an iron status sub-study in 2000–2002, a snacking study in 2006 and a pilot study to test a web-based 24-h dietary assessment tool in 2014.
Target population and level of geography covered All women pregnant during 1990–1992, who resided in Bristol/Avon Health Authority and surrounding areas, their partners and all children born out of these pregnancies.
Geographical area captured: Bristol and surrounding areas (South West of England)
Middle aged women (aged 35–69 at recruitment) living in England, Scotland and Wales, who self-reported as being vegetarian or non-red meat eaters.
Geographical area captured: England, Scotland and Wales and English regions. Regions included in the study’s data set can be easily converted to Government Office Region (GOR) categories
Type of dietary assessment used Food frequency questionnaires (FFQs). Food diaries were completed by parents when children were 7, 10 and 13 years of age. Food frequency questionnaire (FFQ); a 4-day food diary (completed during the follow up study in 1999–2002) and a 24-h web-based diet recall assessment pilot in 2014.
Primary users of diet-related data Academics/Researchers. Academics/Researchers.
Data Accessibility/Availability Data (and details regarding survey data currently available) accessible through the University of Bristol Data (and details regarding survey data currently available) accessible through the Consumer Data Research Centre
Types of variables captured Socio-demographic information (age, sex, occupation, employment, education), consumption of fruits, vegetables and a host of other foods which vary across the survey waves. Socio-demographic information (age, sex, occupation, education), food consumption data captured from FFQs and food diaries conducted at different survey waves.
Cost to access Minimum cost of £2715 to access Not applicable
Key features/potential benefits 1. Large sample size. 2. Cohort study data can be used to track changes over time. 1. Large sample size. 2. Cohort study data could be used to track changes over time.
3. Availability of food diary data provides detailed information on all foods consumed by participants.
Key considerations 1. Costly to access. 2. Study not representative of English population; only focuses on Bristol and surrounding areas (South West of England). 3. Changes made to diet-related questions across survey waves could make it difficult for comparisons to be made over time 1. Food diaries completed in phase 2 of the study (1999–2002) and diaries completed during the 2014 online pilot study were still being processed at the time of this assessment. As such, these data are not available, and the date of release is undetermined
2. Study not representative of the English population. Participants were mostly vegetarian, middle aged, middle class, white women who volunteered to be a part of the study during the late 1990s 3. Changes made to diet-related questions across the survey waves may make it difficult to make comparisons over time.
  Whitehall II Study Millennium Cohort Study (MCS)
Survey Background The Whitehall II study is a cohort study conducted to assess the causes of social inequalities in health in England. The MCS is a large national longitudinal birth cohort study which tracks 19,000 children born in the UK during 2000–2001, from childhood into adulthood.
Survey Design and Methodology Longitudinal Cohort study. A cohort of 10,308 middle-aged persons (3413 females and 6895 males, aged 35–55 years old) who worked in the London offices of 20 Whitehall departments in 1985–1988 were included in the study. During the 2015–2016 period, research clinics were established in London, Bristol, Birmingham and Liverpool to allow persons (especially retired persons) now living within these and surrounding areas to be a part of the study and reduce the level of attrition. Members of the cohort were invited to attend a clinic research screening every 5 years and a postal survey sent to participants between clinic phases. Overall, data has been collected over 12 waves, from 1985 to 1988 to 2015–2016 Longitudinal Birth Cohort study. Multi-stage stratified random sample. The sample consisted of all children born (live births) over 12 months (from 1 September 2000 in England and Wales and for 59 weeks from 22 November 2000 in Scotland and Northern Ireland). Six surveys have been conducted to date, capturing information when children were 9 months and 3, 5, 7, 11 and 14 years of age (in 2015). Although data are currently unavailable, the 7th wave was conducted in 2018 captures children at age 18. A combination of data collection methods has been used. These include face-to-face interviews, self-completed questionnaires; psychological measurements, observation; time use diaries and physical measurements.
Target population and level of geography covered Middle-aged persons who worked in the London offices of 20 Whitehall departments in 1985–1988. Children born in the UK (England, Scotland, Northern Ireland and Wales) during 2000–2001. Data for England are available at the national and Government Office Region (GOR).
Type of dietary assessment used Food frequency questionnaire (FFQ) Food frequency questionnaire (FFQ)
Primary users of diet-related data Academics/Researchers. Academics/Researchers.
Data Accessibility/Availability Data accessible through the University College London. Data available for waves 1–12 (1986–2016) Data accessible through the UK Data Service. Data currently available for the 2001–2015 survey period.
Types of variables captured Socio-demographic information (age, sex, occupation, employment, retirement, education, income), self-reported health and frequency of consumption in the last 12 months of fruits, vegetables, meat, fish, soups, sauces, spreads, eggs, dairy products, fats, bread, pasta, potato, rice, sweets and snacks were consumed. Socio-demographic information (age, sex, occupation, employment, education of parents), consumption of fruits and vegetables and other foods such as bread, milk, sugary drinks and fast foods.
Cost to access Not applicable Not applicable
Key features/potential benefits 1. Food consumption data available for a wide variety of foods. 2. Fairly large sample size across the 12 waves (10,308 in 1985–1988 to 5632 in 2015–2016). 3. Cohort study data can be used to track changes over time. 1. Large sample size and nationally representative. 2. Cohort study data can be used to track changes over time. 3. Children were asked to state their consumption of fruits and vegetables and other foods such as bread, sugary drinks and fast food at age 14.
Key considerations 1. Study not representative of English population. Study focused on middle-aged civil servants. 2. Changes made to diet-related questions across the survey waves may make it difficult to make comparisons over time. 3. Based on the current age-group of participants, the study is now primarily focused on issues surrounding population ageing. 1. Cohort members are still very young, which currently limits the assessment of diet by age/over lifecourse.
2. Changes made to diet-related questions across survey waves could make it difficult for comparisons to be made over time.
  European Prospective Investigation into Cancer and Nutrition (EPIC Norfolk/Oxford) UK Biobank
Survey Background EPIC is a large cohort study which aims to examine diet as a risk factor for cancer and other chronic diseases amongst over 80,000 middle aged persons in the UK. The UK Biobank is a large-scale longitudinal study which follows 500,000 middle-aged persons across the UK to investigate the association between diet and a range of diseases such as cancer, heart disease, stroke, diabetes and dementia.
Survey Design and Methodology Longitudinal Cohort study. EPIC Oxford: 65,000 persons from the general population were recruited between 1993 and 1999 via EPIC nurses in GP practices in Greater Manchester, Oxfordshire and Buckinghamshire, England. Postal questionnaires were also sent to members of the Vegetarian Society of the UK and Vegan Society, and study information distributed through health magazines and shops, to capture persons located across the entire UK. Follow-up surveys were conducted 5, 10 and 15 years later. EPIC Norfolk: Invitations were sent to all 40–79-year olds on collaborating GP listings. Over 30,000 persons within Norwich and surrounding areas (East of England) were recruited over the 1993–1997 period. Participants were followed up at 18 months, 3, 13 and 20 years after recruitment. A combination of data collection methods was used for both studies (nurse interview to collect anthropometric measurements and blood samples, self-completed questionnaires (on physical activity) and record linkages via hospital diagnoses, death certification and cancer registration Longitudinal Cohort study. Population-based registers such as those held by the National Health Service (NHS) were used as a sampling frame to identify persons living within proximity to study assessment centres. Each assessment centre aimed to recruit as many persons within the target population. Baseline data (for the 2006–2010 period) were collected at assessment centres, where self-reported baseline questionnaires were used to collect health and lifestyle-related data and interviews conducted to collect physical measurements and biological samples. A follow up survey was conducted in 2011–2012.
Target population and level of geography covered EPIC Oxford: Men and women 35 years and over (at recruitment) who lived in Greater Manchester, Oxfordshire and Buckinghamshire in England and vegetarians /vegans located across the UK. EPIC Norfolk: Men and women aged 40–79 (at recruitment) who lived in Norwich and surrounding towns and rural areas. Middle-aged males and females (persons aged 40–69 during the 2006–2010 period) who lived within a 10-mile radius of 35 study centres strategically located across England, Wales and Scotland.
Type of dietary assessment used Food frequency questionnaire (FFQ) and a 7-day food diary (completed at recruitment and at the 2nd wave of the study) Food frequency questionnaire (FFQ) with foods related to increased cancer risk conducted at baseline. Web-based 24-h recall repeated on four occasions over a 16-month period.
Primary users of diet-related data Academics/Researchers. Academics/Researchers.
Data Accessibility/Availability The EPIC Oxford study is accessible through the University of Oxford and EPIC Norfolk through the University of Cambridge. Details on current data availability accessible from both institutions. Data accessible through the UK Biobank. Data available (at the time of assessment) for the 2006–2010 (baseline) and 2011–2012 period.
Types of variables captured Socio-demographic information (age, sex, occupation, education), food consumption data captured from FFQs and food diaries conducted at different survey waves. Socio-demographic information (age, sex, employment) and fruits, vegetables, meat, dairy and a host of other foods consumed (total of over 200 foods) over a 24-h period.
Cost to access Not applicable Minimum £2000 to cover application and data access cost. Possibly reduced cost of £500 for research students (subject to review and approval).
Key features/potential benefits 1. Large sample size. 2. Cohort study data can be used to track changes over time. 3. Availability of food diary data (at recruitment and wave 2) which provides detailed information on foods consumed by participants. 1. Large sample size. 2. Cohort study data can be used to track changes over time. 3. Detailed information on foods consumed by participants over repeated days (repeated 24-h diet recalls).
Key considerations 1. Study not representative of English population. Focused on middle-aged persons living in Norwich, Greater Manchester, Oxfordshire and Buckinghamshire who were in some instances selected via purposive sampling. 2. 50% of participants were vegetarians/vegans. 3. Changes made to diet-related questions across the survey waves may make it difficult to make comparisons over time 1. Study not representative of English population. Focused on middle-aged persons from less-deprived areas (based on the target population). 2. The baseline survey captured some aspects of diet consumption but was not as comprehensive as the 2011–2012 survey wave. 3. Differences in the number and types of diet-related questions asked across survey waves could make it difficult for comparisons to be made over time. 4. Relatively high cost to access data. 5. Lengthy application process and possible lag time for approval.
  British Regional Heart Study (BRHS) British Women’s Heart and Health Study (BWHHS)
Survey Background The BRHS is a cohort study, established in 1978–1980, which explores the factors associated with heart disease, hypertension and stoke amongst 7735 middle-aged men (40–59 years at recruitment) recruited from General Practices (GPs) in 24 towns in England, Scotland and Wales. The BWHHS is a cohort study, established in 1999 as a complement to the BRHS. The study follows 4286 women, aged 60 years and over (at recruitment) from 24 General Practices (GPs), in 23 towns in England, Scotland and Wales
Survey Design and Methodology Longitudinal Cohort study. Almost 8000 middle-aged men who were selected at random from one GP in each of the 24 towns, were examined over the 1978–1980 period. Self-completed health and lifestyle questionnaires and clinical assessments/examinations (inclusive of anthropometric measurements) completed at baseline (1978–80). Follow-up self-completed questionnaires were completed in 1985,1992,1996,1998–2000, 2003,2005,2007,2010–12, 2014,2015, 2016,2017 and 2018. A review of GP records (including all hospital and clinic correspondence) was also conducted bi-annually. A clinical re-examination was done in the 1998–2000 wave. Participants were also given a self-completed activity survey questionnaire and asked to wear an activity monitor and keep a 3-day activity diary in 2010, 2011, 2012, 2013, 2015 and 2017. Longitudinal Cohort study. Almost 8000 middle-aged women were randomly selected from 24 GPs, in 23 towns from 1999 to 2000. Self-completed health and lifestyle questionnaires, and nurse administered interviews and medical examinations were completed at baseline (1999–2000). A review of GP records (including all hospital and clinic correspondence) was completed at baseline and in 2002, 2004, 2007, 2011–12 and 2016–17. Self-completed health and lifestyle questionnaires were completed in 2003, 2007 and 2010–2011. Participants were also given a self-completed activity survey questionnaire and asked to wear an activity monitor/belt and keep a 3-day activity diary in 2010–2011.
Target population and level of geography covered Middle-aged men aged 40–59 years (at recruitment) who resided in 24 towns across England, Scotland and Wales. Middle-aged women aged 60 years and over (at recruitment) from 23 towns across England, Scotland and Wales.
Type of dietary assessment used Food frequency questionnaire (FFQ) Food frequency questionnaire (FFQ)
Primary users of diet-related data Academics/Researchers. Academics/Researchers.
Data Accessibility/Availability Data accessible through University College London Data accessible through University College London
Types of variables captured Socio-demographic information (age, sex), health status, consumption of fruits and vegetables, fish, meat, bread and a host of other foods which vary across the survey waves. Socio-demographic information (age, sex), consumption of fruits, vegetables, cheese, milk, red meat and other foods which vary across the survey waves.
Cost to access Unknown (Information inaccessible at time of assessment). Unknown (Information inaccessible at time of assessment).
Key features/potential benefits 1. Cohort study data can be used to track changes over time. 2. Data captured could be used to explore relationships between diet, cardiovascular disease and associated chronic diseases. 1. Cohort study data can be used to track changes over time. 2. Data captured could be used to explore relationships between diet, cardiovascular disease and associated chronic diseases.
Key considerations 1. Study not representative English population. Study only captures middle-aged men from 24 towns across sections of Scotland, England and Wales. 2. Differences in the number and types of diet-related questions asked across survey waves could make it difficult for comparisons to be made over time. 3. Based on the current age-group of participants, the study is now primarily focused on issues surrounding population ageing. 1. Study not representative English population. Study only captures middle-aged women from 23 towns across sections of Scotland, England and Wales. 2. Differences in the number and types of diet-related questions asked across survey waves could make it difficult for comparisons to be made over time.