Official Statistics

National Diet and Nutrition Survey 2019 to 2023: report

Published 11 June 2025

Executive summary

Introduction

The National Diet and Nutrition Survey (NDNS) is designed to assess the diet, nutrient intake and nutritional status of the general UK population. Participants aged 18 months and over living in private households are selected each year from all 4 UK countries. The sample is designed to be nationally representative. The survey data is used by UK governments to monitor progress towards achieving diet and nutrition objectives and to develop food and nutrition policies.

This report presents data for 2019 to 2023. Data collection was carried out between October 2019 and July 2023. Data collection was suspended between March and October 2020 due to the COVID-19 pandemic. The survey consisted of:

• an interview to collect background information
• recording food and drink consumption on 4 non-consecutive days using an online tool (Intake24)
• physical measurements including height and weight
• a physical activity questionnaire
• a urine sample to assess iodine levels
• a blood sample to assess nutritional status biomarkers (the amount of some nutrients in blood)

In this report, results are presented for 7 standard NDNS age groups:

• 18 months to 3 years
• 4 to 10 years
• 11 to 18 years
• 19 to 64 years
• 65 to 74 years
• 75 years and over

Results are also presented by sex for all age groups except 18 months to 3 years. Due to small sample sizes, the 2 oldest age groups have been combined for reporting results for nutritional status biomarkers in blood and urine (65 years and over).

This report includes:

• food and drink consumption, nutrient intake and nutritional status by age and sex - this includes the percentage of people meeting UK dietary recommendations and the contribution of food groups to energy and nutrient intakes
• consumption of food and drink from the out of home sector (such as cafes, restaurants, takeaways) and contribution to energy and nutrient intakes
• food and drink consumption, nutrient intake and nutritional status by:
  • household income
  • Index of Multiple Deprivation (IMD) for England only (IMD ranks how disadvantaged a person’s neighbourhood is based on a range of factors, including income, employment, education, health, crime and housing)

Household food security (that is access to sufficient safe and nutritious food) was collected in 2022 to 2023 (fieldwork year 15) only. No analysis of food consumption or nutrient intake by household food security was carried out for this report due to small numbers.

Commentary in this report focuses on foods, drinks and nutrients chosen on the basis of their public health importance and/or because population intakes are not meeting dietary recommendations. More results can be found in the data tables accompanying this report. The absence of commentary for a food, drink or nutrient does not indicate that it is not of public health interest.

Method changes

From 2019 (fieldwork year 12) the dietary assessment method used in the survey was changed from a paper diary to Intake24 (an online tool) completed on 4 non-consecutive days over a 2 week period. An evaluation of the method change has shown step changes in estimates for some foods. So, it is not possible to compare food consumption and nutrient intakes from 2019 to 2023 with those from previous years.

A change to the methods for transporting blood samples to the laboratory for analysis has also meant that results from 2019 to 2023 cannot be compared with previous results for some analytes. So, comparison with NDNS data for previous years (before 2019 to 2023) is only presented for urinary iodine.

Main findings for 2019 to 2023: food consumption

Average food consumption figures in this section include participants who did not consume the food or drink on any of their recording days. These are non-consumers. If non-consumers are excluded, then average consumption will be higher.

Fruit and vegetables

Most participants did not meet the UK government recommendation to eat at least 5 portions of a variety of fruit and vegetables each day.

On average, children aged 11 to 18 years ate 2.8 portions of fruit and vegetables a day. Less than 1 in 10 children aged 11 to 18 years (9%) met the ‘5 A Day’ recommendation.

On average, adults consumed 3.3 to 3.7 portions per day (depending on age). Less than 1 in 5 adults (17%) met the 5 A Day recommendation.

Consumption for adults and older adults was lower than previous published figures. This is likely to be partly due to the dietary method change. It may also reflect a real reduction in fruit and vegetable consumption, due to a combination of lack of availability during the COVID-19 pandemic and cost of living pressures. Other data sources have suggested this.

Red and processed meat

The UK government also recommends that people who eat more than 90g of red or processed meat a day should cut down to 70g.

In 2019 to 2023, average consumption of red and processed meat was below 70g per day in all age and sex groups. Men aged 19 to 64 years ate the most red and processed meat. On average, men aged 19 to 64 years ate 66g per day but about a quarter (27%) ate more than 90g per day.

Sugar-sweetened soft drinks

The highest average consumption of sugar sweetened soft drinks was by children aged 11 to 18 years (124 millilitres (mls) per day). The highest consumers in this age group drank 606mls per day. Average consumption by younger children (4 to 10 years) was under half this amount (54mls per day, highest consumers 371mls per day). Among adults, average consumption was highest in men aged 19 to 64 years (108mls per day; highest consumers 768mls per day).

Main findings for 2019 to 2023: energy and nutrient intakes

The Estimated Average Requirement (EAR) for energy is the amount of calories that will meet the needs of about half the general population. The EAR varies by age and sex. With the exception of children aged 18 months to 3 years, average energy intakes were below the EAR. This is likely to be due to underreporting and is a common problem in diet surveys. Underreporting of energy intake in the NDNS was around 30% overall.

Saturated fats

The UK government recommends that no more than 10% of energy comes from saturated fats. For 2019 to 2023, average intakes exceeded the recommendation. Saturated fats provided 12.5% of energy for children (18 months to 18 years) and 12.6% for adults. Overall, 85% of children and 82% of adults did not meet the recommendation.

Free sugars

The UK government recommends that no more than 5% of energy comes from free sugars. Overall, less than 1 in 10 children (9%) and less than 1 in 5 adults (19%) met the recommendation. Average intakes were 10.5% of energy for children and 10% for adults. Girls aged 11 to 18 years had the highest intake (12% of energy).

Fibre

Most people do not meet UK government fibre recommendations (at least 30g per day for adults and 15 to 25g per day for children depending on their age). For younger children, 78% of those aged 18 months to 3 years and 86% of those aged 4 to 10 years did not meet the recommendation for their age group. For children aged 11 to 18 years and for adults, 96% did not meet the fibre recommendation.

Main findings for 2019 to 2023: nutritional status

Vitamin D status

Vitamin D status is assessed by measuring the amount of 25-hydroxyvitamin D in blood. Low vitamin D status (defined as 25-hydroxyvitamin D concentration less than 25 nanomoles per litre (nmol per litre) increases the risk of poor bone and muscle health. In 2019 to 2023, the proportions of age groups that had low vitamin D status were:

• 10% of children aged 4 to 10 years
• 23% of children aged 11 to 18 years
• 18% of adults aged 19 to 64 years
• 12% of adults aged 65 years and over

Folate status

Folate status is assessed in 2 ways (red blood cell (RBC) folate and serum folate) and both are reported in the NDNS. RBC folate reflects long term folate status. A concentration less than 305nmol per litre indicates folate deficiency.

In 2019 to 2023, the proportions of age groups that had low RBC folate were:

• 12% of children aged 11 to 18 years
• 4% of adults aged 19 to 64 years
• 2% of adults aged 65 years and over

Folate is particularly important in helping to prevent neural tube defects forming in the fetus in early pregnancy. For women of childbearing age (16 to 49 years), 83% had an RBC folate concentration less than 748nmol per litre, below which there is increased risk of pregnancy affected by neural tube defects.

Iodine status

Iodine is measured in urine. Urinary iodine concentration (UIC) for most age and sex groups was above the threshold of 100 micrograms (µg) per litre, indicating adequate population iodine status. However, there was evidence of insufficient population iodine status for girls aged 11 to 18 years and women of childbearing age (16 to 49 years). UIC in these groups was below 100µg per litre.

Since the measurement of urinary iodine was introduced into NDNS in 2013, there has been a significant year-on-year decrease in UIC. Between 2013 to 2023, UIC decreased by 29% for girls aged 11 to 18 years and by 25% for adults aged 19 to 64 years.

Diets and household income in the UK

Participants in higher income households in 2019 to 2023 were closer to meeting some dietary recommendations. However, where diets failed to meet recommendations, this was consistent across the range of income.

There was a slight increase in intake of fruit and vegetables with increasing income for all age and sex groups. Fibre intake also increased, on average, with increasing income across all age and sex groups.

There was little variation in energy (calorie) intakes by income for most population groups. This was also the case for percentage energy from free sugars. An exception was for boys aged 4 to 10 years and adults aged 65 to 74 years, where percentage energy from free sugars tended to go down with increasing income.

Folate status (RBC folate concentration) of women of childbearing age (16 to 49 years) tended to increase with increasing income. No association was seen for other population groups.

For children and adults aged 19 to 64 years, vitamin D status increased with increasing household income. Among adults aged 65 years and over, vitamin D status decreased with increasing income.

Diets, IMD and equivalised household income quintiles for England only

Analysis by IMD was done for participants in England only. People in the most deprived IMD group in England tended to have poorer diets, although this was not consistently seen across all age and sex groups. More differences were seen by IMD than by household income in England. Overall, where diets failed to meet dietary recommendations, this was the case across all IMD and income groups in England.

Food and drinks from the out of home sector

Almost three-quarters (72%) of participants reported buying food or drink from the out of home sector (for example cafes, pubs, takeaways) in the last 7 days, with most making purchases 1 to 2 times a week. Children aged 11 to 18 years (81%) and adults aged 19 to 64 years (77%) were most likely to report buying food or drink from the out of home sector.

In their online food and drink consumption record (covering up to 4 days), around half of participants (54%) recorded at least one eating occasion (snack, meal or drink) where the food had come from the out of home sector, the most popular being ‘fast food’ or ‘take-away’. For participants who reported eating ‘out of home’, almost a quarter (23%) of their energy (calorie) intake came from these eating occasions. This was also true for their intake of other nutrients, including saturated fat, free sugars and sodium.

1. Introduction

Background

The National Diet and Nutrition Survey (NDNS) was set up in 2008 with a continuous programme of fieldwork, designed to assess the diet, nutrient intake and nutritional status of the general population aged 18 months and over living in private households in the UK. Fieldwork ran continuously from 2008 to 2023, apart from a 7-month suspension during the COVID-19 pandemic.

NDNS is jointly funded by the Department of Health and Social Care (DHSC) and the UK Food Standards Agency (FSA).

The NDNS phase for this report (fieldwork years 12 to 15, 2019 to 2023) was carried out by a consortium of the National Centre for Social Research (NatCen) and the Medical Research Council (MRC) Epidemiology Unit at the University of Cambridge (Epidemiology Unit). Fieldwork in Northern Ireland was carried out by the Northern Ireland Statistics and Research Agency (NISRA) and NatCen.

NDNS provides the only source of nationally representative UK data on the types and quantities of foods consumed by individuals, from which estimates of nutrient intake for the population are derived. Analysis of blood and urinary biomarkers provides data on the population’s nutritional status.

UK governments use NDNS results to monitor progress towards achieving their diet and nutrition objectives. For example, to monitor progress towards achieving a healthy, balanced diet as shown in the Eatwell Guide. The results also inform policy development.

NDNS is an important source of evidence underpinning the work of the Scientific Advisory Committee on Nutrition (SACN) and its advice to UK governments on nutrition related issues.

The FSA uses the food consumption data to assess exposure to chemicals in food, as part of the risk assessment and communication process in response to a food emergency. It is also used to inform negotiations on setting regulatory limits for contaminants.

Time period for this report

The results presented in this report are for data collected from October 2019 to July 2023 (referred to as 2019 to 2023). This period corresponds with fieldwork years 12 to 15 of the survey which started in 2008. The survey is organised by fieldwork year. The report refers to fieldwork years when describing some aspects of the methodology. Delays in fieldwork due to the COVID-19 pandemic meant that there was some overlap between fieldwork years 13, 14 and 15.

Content of this report

This report presents an overview of food consumption, nutrient intake and nutritional status for the UK population from 2019 to 2023. This report is accompanied by 7 sets of data tables, corresponding to chapters 3 to 9 of the report, as outlined below. The data tables are available at the report publication page.

Chapter 2 provides detail of the methodology and survey design, including the sample design and size. It also describes the changes in dietary assessment methodology and in the processing and transport of blood samples since the 2016 to 2019 report.

Chapter 3 provides response rates over the fieldwork period (2019 to 2023) and presents characteristics of participants including body mass index (BMI), ethnicity and food security (tables 3.1 to 3.3).

Chapter 4 presents data on foods consumed and the percentage of the population meeting the main food based dietary recommendations (tables 4.1 to 4.18). Commentary in this chapter covers fruit and vegetables, red and processed meat and sugar-sweetened soft drinks.

Chapter 5 presents data on nutrient intakes, including the percentage of the population meeting dietary recommendations, and the contribution of food groups to nutrient intakes (tables 5.1 to 5.68). Commentary in this chapter focuses on energy, saturated fatty acids, free sugars and fibre.

Chapter 6 presents nutritional status biomarkers (tables 6.1 to 6.17). Commentary in this chapter covers blood markers of folate and vitamin D status and urinary iodine.

Chapter 7 presents food consumption, nutrient intakes and nutritional status biomarkers by equivalised household income for the UK as a whole (tables 7.1 to 7.27).

Chapter 8 presents data by IMD and equivalised household income for England only (tables 8.1 to 8.29).

Chapter 9 focuses on food and drink from the out of home sector (tables 9.1 to 9.7).

There is background information on the survey, including the sample design and methods, in the NDNS 2019 to 2023 appendices. The questionnaires, participant information leaflets and consent forms are also in the appendices. There is also:

• questionnaire data on physical activity in appendix V
• details of additional variables for nutrients and biomarkers not included in this report in appendix T (PDF 267KB) (these variables will be available at the UK Data Service)
• more information on the statistical analysis methods used in appendix U (PDF 233KB)
• more information on underreporting in appendix X (PDF 1.04MB)

There is a full list of appendices in the Appendices section at the end of this report.

There is also a separate report on the change to the methods for transporting blood samples for processing and analysis. See the blood sample transport evaluation report for more information.

Age groups and sample

The NDNS sample is drawn from all 4 UK countries and is designed to be nationally representative.

Results are presented for the age groups:

• 18 months to 3 years (sex combined)
• 4 to 10 years
• 11 to 18 years
• 19 to 64 years
• 65 to 74 years
• 75 years and over

Due to small sample sizes, the 2 oldest age groups have been combined for reporting results for blood and urine biomarkers of nutritional status (65 years and over).

2. Methodology

This chapter provides an overview of the methodology for NDNS 2019 to 2023 (fieldwork years 12 to 15). Full details of the methods and fieldwork documents are provided in appendices A to BB. Data collection was carried out between October 2019 and July 2023.

NDNS is a cross-sectional survey (that is data collected from participants at a single point in time), designed to be representative of the UK population. NDNS was affected by the COVID-19 pandemic, and from March 2020 the survey methodology and protocols had to be adapted in line with government advice and regulations. This included the suspension of fieldwork between March and October 2020. Further detail is provided in section B.1.2 of appendix B (PDF 373KB).

Sample design

The survey aimed to collect data from a UK representative sample of 1,000 people per year with:

• 500 adults aged 19 years and over
• 500 children aged 18 months to 18 years

The Northern Ireland government funded an increase to the sample size in Northern Ireland. Additional addresses were selected in Northern Ireland. These extra cases are included, with appropriate weighting, in the data set on which this report is based. Results for Northern Ireland will be reported separately by the Food Standards Agency in Northern Ireland.

For each fieldwork year, the sample was taken from the ‘small users’ list in the Post Office Postcode Address File (PAF). These are addresses (or delivery points) which receive fewer than 25 articles of mail a day. To be cost effective, the addresses were clustered into primary sampling units (PSUs), small geographical areas based on postcode sectors, randomly selected from across the UK. A list of addresses was randomly selected from each PSU. A reserve sample was taken at the same time as the main sample, to be used (in full or in part) if achieved numbers fell short of the target.

Selected addresses were randomly allocated to one of 2 address types: ‘basic addresses’ and ‘young person addresses’.

‘Basic addresses’ made up around one-third of the selected addresses. In these, up to 2 adults and 1 child were selected.

‘Young person addresses’ made up around two-thirds of the issued sample. In these, up to 2 children (each from a different NDNS age group: 18 months to 3 years, 4 to 10 years, 11 to 18 years) and no adults were selected.

This design was used to achieve (as far as possible) equal numbers of adults and children in the sample. Further details of the sample design are provided in appendix B.

Fieldwork

Information describing the purpose of the survey was posted to all selected addresses. This was followed by an interviewer visiting each address to recruit participants in the eligible age ranges.

At each address, the interviewer recorded the number of households. In cases where there were 2 or more households, the interviewer randomly selected one household to be invited to take part. From each selected household, the interviewer randomly selected up to 2 adults and 1 child, or 2 children from different age groups, to take part in the survey. These are known as participants.

The survey was carried out in 2 stages. The first stage was a Computer Assisted Personal Interview (CAPI) with each participant, carried out by the interviewer. In the case of a child aged 10 years or under, the CAPI was carried out with their parent or guardian (defined as a person with legal responsibility for the child). See appendix D (PDF 1.02MB) for the content of the CAPI.

Participants were then asked to self-complete 4 24-hour dietary recalls, a physical activity questionnaire and have their height and weight measured. Urine samples were requested from participants aged 4 years and over (see appendix M (PDF 196KB) for detail).

Participants completed 24-hour dietary recalls using an online dietary assessment method, Intake24^{[footnote 1]}. See appendix A (PDF 275KB) for details of the dietary assessment. They completed the first recall on the day of the interview. Participants received the invitation to complete their next dietary recall by text or email 2 to 6 days (randomly allocated) after completing their previous dietary recall. If a participant did not complete their dietary recall on the requested day, they received up to 4 reminders sent over the next 9 days to prompt them to complete it, always requesting completion of the recall for the preceding day.

Participants were invited to complete an online physical activity questionnaire at the end of their third dietary recall. The invitation was reissued at the end of recall 4 if it had not been completed by then. See appendix V for details of the adult and child questionnaires.

Participants who took part in the CAPI and completed at least one of 4 requested dietary recalls were classified as ‘productive’ and were invited to take part in the second stage of the survey. This involved a visit from a biomedical fieldworker to take a blood sample and further physical measurements. See appendix I for details of the biomedical fieldworker visit and appendix P for details of blood sample collection and processing.

In years 2019 to 2022, a sub-sample of participants aged 4 years and over was invited to take part in a doubly labelled water (DLW) sub-study for assessment of misreporting by measurement of total energy expenditure. See appendix X for more information about misreporting.

From March 2020, fieldwork was suspended due to COVID-19 restrictions. When fieldwork restarted in October 2020, a remote protocol was put in place so that interviewers did not enter participants’ homes but instead carried out interviews over the phone. Some elements of the survey were suspended until face-to-face interviewing resumed. This included the fieldworker visits to collect blood and urine samples which resumed in November 2021. See appendix B for more details.

Methodological changes and considerations for data interpretation

Change in the dietary assessment method and analytical approach

For the first 11 years of NDNS fieldwork (2008 to 2019), dietary assessment was based on a paper diary completed by participants over 4 consecutive days. The design aimed to achieve an appropriate balance of weekdays and weekend days across the sample. Diaries were reviewed by interviewers and foods and portions were retrospectively coded by trained coders.

In 2018, the survey moved to an online dietary data collection method with automated coding to enable increased cost efficiency, to provide opportunities for improving data quality, and giving potential to scale up the survey in the future. Following a review and evaluation of online automated tools, Intake24 was selected to replace the paper food diary in NDNS from 2019. Intake24 is a web-based, multiple pass 24-hour recall tool and there are further details about it in appendix A.

The change from paper diary to online 24-hour recall was accompanied by a move to collecting dietary data on non-consecutive days, rather than consecutive days. The aim remained to achieve an appropriate balance of weekdays and weekend days across the sample. Collecting dietary data on non-consecutive days provided the opportunity to calculate ‘usual intakes’ which is the accepted method to estimate population habitual nutrient and food intakes. This is a statistical procedure which uses the whole data set to eliminate day-to-day variation in individual intakes (within-person random error) (see appendix U for more information about calculating usual intakes).

For previous NDNS years (2008 to 2019) which used the paper diary, nutrient and food intakes were calculated using the ‘day average’ method (where intakes from each recording day are simply averaged to get a daily intake for each individual). When calculating ‘day average’ nutrient and food intakes, the variance of the usual group intake is inflated by day-to-day variation in individual intake, resulting in inflated estimates of the prevalence of low or high intakes.

Collecting repeated 24-hour recalls on non-consecutive days, it is possible to eliminate the within-person variability of the data to obtain an estimate of the population usual intake distribution (Souverein and others, 2011). The main effect of shifting to calculating ‘usual intakes’ is to reduce the extremes of the distribution, drawing them closer to the mean. This enables a better estimation of ‘percentiles’ or ‘proportions above or below a threshold’. The ‘day average’ method is now known to overestimate proportions above or below a threshold.

The ‘usual intake’ method was applied to all nutrients and most foods in the 2019 to 2023 data set. For less frequently consumed foods, for example oily fish, the ‘usual intake’ method could not be applied because there was insufficient data to estimate within-person variance. In these situations, the ‘day average’ method was used.

Introducing Intake24 and moving to calculating ‘usual intakes’ has made it possible to include all participants with at least one recall in the dietary data set. Participants who completed only 1 or 2 recalls can be included because day-to-day variation estimates from participants of similar ages with 3 or 4 recalls can be applied. This differs from the protocol in 2008 to 2019 using the paper diary, where only participants who completed 3 or 4 diary days were included (a diary with fewer than 3 days was considered incomplete).

A formal evaluation of the dietary method change was carried out using data collected from 2019 to 2023 in 3 stages. A report was published for each stage:

• stage 1 report (September 2021)
• stage 2 report (November 2023)
• stage 3 report (May 2025)

The evaluation found that step changes (that is changes that are different from what the year 1 to 11 trend data would suggest) in intakes of some foods and nutrients were observed at the point when Intake24 replaced the food diary in 2019. Where step changes occurred, it was not possible to be confident in the continuation of the time series. So, the time trend for dietary data has not been published in this report. There is also no statistical comparison between 2019 to 2023 combined and the last report for 2016 to 2019 combined. Further details can be found in the stage 3 evaluation report. The changes in dietary assessment method and analytical approach should be taken into account if comparing the 2019 to 2023 data with previous NDNS data.

The evaluation notes that much of the data collection period for 2019 to 2023 coincided with the COVID-19 pandemic and rising inflation. These events have resulted in changes to the availability and price of food with likely consequences for lifestyle and eating behaviours for many people (Hoenink and others, 2024; Food Foundation, 2022; OHID, 2025).

The evaluation concluded that while some of the differences in food and nutrient intakes seen in the evaluation were likely to be attributable to the diet method changes (as the differences were observed before the pandemic), it cannot be ruled out that some of the changes are real rather than methodological. More years of data are needed and it is too early to say whether or not the step changes seen represent a sustained change in long-term trends.

Change in participant selection within a household

At the time the dietary assessment method was changed to Intake24, the number of adult and child participants per household was increased to improve cost efficiency.

In 2008 to 2019 (fieldwork years 1 to 11), 1 adult and 1 child were selected in around one-third of sampled addresses (referred to as ‘basic addresses’) and 1 child (no adult) in around two-thirds of sampled addresses (referred to as ‘young person addresses’).

For 2019 to 2023 (fieldwork years 12 to 15), up to 2 adults and 1 child were selected from around one-third of addresses and up to 2 children (each from a different NDNS age group: 18 months to 3 years, 4 to 10 years, 11 to 18 years) and no adults were selected from the remaining two-thirds of addresses.

Analysis carried out as part of the stage 3 evaluation suggests that the new participant selection model resulted in some clustering of dietary data within a household, although the degree of clustering is unlikely to be great enough to reduce confidence in the survey estimates.

Change in blood sample transport and processing

Changes were made to the blood sample transport and processing protocols in 2021 (fieldwork year 13) with the aim to simplify fieldwork logistics, improve cost efficiency and standardise sample processing. The blood sample is collected from the participant into a number of tubes. Prior to 2021, one blood tube for each participant was sent by first class post at ambient temperature (that is without cooling) to the NDNS central laboratory (MRC Epidemiology Unit, Cambridge) to be used for full blood count and whole blood folate analysis. The remaining blood tubes were processed in a local laboratory within 2 hours of collection.

Informed by a pilot study (Jones and others 2021) the decision was taken to change the blood transport protocol so that all blood samples were posted in an insulated box with cold packs using an overnight delivery service direct to the MRC Epidemiology Unit. To evaluate the impact of the change in specimen transport on biomarker concentrations, the different transport protocols were compared in a nested study within NDNS between November 2021 and March 2022, in parallel with year 13 fieldwork. Overall, both the pilot and nested studies showed that the posting protocol provided valid results for all NDNS biomarkers, and the change in protocol had little impact on the measured concentrations of most biomarkers. These results are described in more detail in the blood sample transport evaluation report. Details of the blood and urine sample collection, processing and analysis methods are provided in appendices M, P and Q of this report.

However, for a few biomarkers, potential differences between the 2 sample processing protocols may have contributed to differences in biomarker concentration before and after the method change. This raises implications for data interpretation and should be considered when using the results data and comparing results over time (pre and post- method change). To minimise risk of misinterpretation of the data, statistical analyses of the comparison between 2019 to 2023 (years 12 to 15) combined, the last report for 2016 to 2019 (years 9 to 11) combined, and the time trend analysis for 2008 to 2023 have not been presented in this report. Further explanation can be found in the blood sample transport evaluation report.

Since the protocol for processing urine samples has not changed, urinary iodine time trends are included in this report.

Underreporting of energy intake

The misreporting of energy intake (generally underreporting) is a known issue for all dietary surveys and studies (Poslusna and others, 2009). In NDNS the DLW biomarker has been used to measure total energy expenditure in a sub-sample of participants to assess the extent of misreporting. The results of the most recent DLW sub-study, carried out from 2019 to 2022 (available in the stage 2 evaluation report), showed that energy intake was underreported by around 30% on average. There is evidence that there is less underreporting in young children than in older children and adults, but other differences between population subgroups are less clear. Previous studies carried out when diet was assessed using the food diary found very similar levels of underreporting overall. There is no evidence that the move to Intake24 has resulted in a higher level of underreporting, although the level remains substantial.

It is not known whether the components of energy intake (protein, fat, carbohydrate, alcohol) are misreported equally or differentially, for example whether fat intake may be over or underreported to a greater degree than protein. Energy and nutrient intakes presented in this report have not been adjusted to take account of misreporting. Further details are provided in appendix X.

Weighting the NDNS data

It is necessary to apply weighting factors to the data collected in NDNS for 2 reasons, which were to:

• minimise any bias in the observed results which may be due to differences in the probability of households and individuals being selected to take part
• attempt to reduce non-response bias

The survey design meant that adults living in households with more than one other adult, and children in households with one or more other children were less likely to be selected than adults or children in single adult or child households.

Also, there were a number of stages in the survey where it was possible for participants to drop out. If the people who refused to participate at a particular stage were systematically different from those who took part, then the sample would be biased.

Weighting factors were used to correct for both these cases. There were 2 stages to the weighting scheme. The first was to create a set of design weights to correct for the unequal selection probabilities. The second was to create weighting factors to adjust for non-response.

The sample design includes an adjustment for selecting more addresses in Northern Ireland. All of the addresses in Northern Ireland, and so participants, are weighted down as a result. This means that weights were applied to put the 4 countries into their correct population proportions in the UK so that, for example, the percentage of the NDNS sample in Northern Ireland is the same as the percentage of the UK population that is in Northern Ireland.

Further details about the weighting strategy are provided in appendix BB.

3. Response and sociodemographic characteristics

Response

Fieldwork periods

Tables 3.1 to 3.3 show household and individual-level response rates for 2019 to 2023 combined and for the individual fieldwork years (years 12 to 15).

Due to the impact of the COVID-19 pandemic (see chapter 2), cases are unevenly distributed across years and are skewed towards the later years when restrictions were lifted and in-home interviewing could resume. See appendix B for further detail.

Figures shown in this chapter are based on unweighted data and so represent the actual number of individuals who took part in NDNS.

Household-level response

For 2019 to 2023 (fieldwork years 12 to 15), 21,794 addresses were contacted of which 42% were eligible for household selection. Ineligible addresses included vacant or derelict properties and institutions. ‘Young person’ addresses that were screened out because they did not contain any children in the eligible age range (18 months to 18 years) accounted for 79% of the total ineligible addresses.

Household selection was carried out at 98% of eligible addresses. The individuals in the remaining 2% of addresses refused to participate before the household selection could be carried out (table 3.1).

In total, 32% (2,925) of selected households were ‘productive’, which means at least one selected participant completed the stage 1 CAPI questionnaire and the first dietary recall. Over the fieldwork years response declined:

• 41% in 2019 to 2020
• 36% in 2020 to 2021
• 32% in 2021 to 2022
• 30% in 2022 to 2023

These response rates are lower than in the previous NDNS phase, which were:

• 50% in 2016 to 2017
• 45% in 2017 to 2018
• 47% in 2018 to 2019

NDNS, in common with all UK surveys, has experienced a decline in response rates over time and a sharper decline during the COVID-19 pandemic.

Individual-level response

In total, 4,370 individuals (2,296 adults and 2,074 children) from 2,925 households completed the stage 1 CAPI questionnaire. Analyses in this report (including response rates for subsequent stages or components of the survey) are based on these 4,370 individuals. Response to survey stages, overall and by fieldwork year, are shown in tables 3.2 and 3.3.

Of the individuals completing the stage 1 CAPI questionnaire:

• 94% (4,089) went on to complete at least 1 dietary recall
• 81% (3,551) completed at least 2 dietary recalls
• 75% (3,277) completed at least 3 dietary recalls
• 71% (3,107) completed all 4 dietary recalls

There was some variation between fieldwork years in the percentage of individuals completing the stage 1 CAPI questionnaire who went on to complete at least one recall:

• 97% in 2019 to 2020
• 92% in 2020 to 2021
• 91% in 2021 to 2022
• 94% in 2022 to 2023

The percentage of individuals completing 4 recalls was 79%, 79%, 71% and 68%. There was no difference between males and females nor between adults and children in stage 1 questionnaire and dietary recall completion rates.

Overall, 70% of invited participants completed the physical activity questionnaire. Completion rates were higher among adults than children (82% of invited adults completed the physical activity questionnaire compared with 56% of invited children).

Participants aged 4 years and over who were interviewed in-person were asked to provide a urine sample and 87% did so (90% of adults, 82% of children). Collection of urine samples was suspended from March 2020 to November 2021 when face-to-face fieldworker visits were not possible due to the COVID-19 pandemic.

Participants were invited to take part in a second stage, a visit by a biomedical fieldworker, primarily to obtain a blood sample. Biomedical fieldwork was particularly affected by the COVID-19 pandemic as face-to-face visits were suspended between March 2020 and November 2021. This meant that biomedical fieldworker visits were delayed for many participants, so the gap between stage 1 and 2 was much longer than in other fieldwork years (between 1 and 22 months and an average of 8 months, in contrast to an average of less than 2 months in other years).

Overall, 34% of the core UK sample who completed the stage 1 CAPI questionnaire were visited by a biomedical fieldworker (45% of adults and 22% of children). However, visit rates were lower for the fieldwork years most affected by the pandemic.

In total, 29% of the core UK sample completing the stage 1 CAPI questionnaire provided a blood sample. Adults were more likely to give a blood sample than children (40% of adults did so, compared with 17% of children). Younger children were less likely to give a blood sample than older children or adults: 7% of children aged 18 months to 3 years and 13% of 4 to 10 year olds provided a blood sample, compared with 24% of 11 to 18 year olds and between 39% and 44% aged 19 years and over.

Sociodemographic characteristics of participants

This section describes the sociodemographic and health-related lifestyle characteristics of the NDNS sample for 2019 to 2023, using data collected during the CAPI interviews. Data is shown in tables 3.4 to 3.10.

Body mass index (BMI)

Interviewer-measured height and weight was sought from participants aged 2 years and over who were interviewed in person. Self-reported height and weight measurements were collected from participants who were interviewed by phone. Height and weight measurements were used to calculate BMI, defined as a person’s weight in kilograms divided by the square of the person’s height in metres (kg/m²).

The self-reported height and weight measurements for adults (aged 19 years and over) were adjusted using prediction equations based on the Health Survey for England (HSE) methodology report. The adjustment takes into account the tendency of people to underestimate their weight and overestimate their height. In the HSE methodology report, mean BMI from self-reported height and weight was 1kg/m² and 1.1kg/m² lower than interviewer-measured mean BMI among men and women, respectively. As the self-reported values have been adjusted, they have been combined with measured values for reporting.

The child (aged 18 years and under) self-reported height and weight measurements were not adjusted, as the HSE adjustment method was developed only for adult data, so these measurements have not been reported. Interviewer-measured heights and weights are reported below.

Adults

Table 3.4 shows mean BMI and corresponding BMI status in adults, by age group and sex, defined according to the World Health Organization (WHO) BMI classification (for more information, see A healthy lifestyle - WHO recommendations) as shown in table A.

Table A: BMI classification

BMI (kg/m2)	Description
Less than 18.5	Underweight
18.5 to less than 25	Healthy weight
25 to less than 30	Overweight but not obese
30 to less than 40	Obese grades I and II
40 or more	Morbidly obese (grade III)

There were no differences in mean BMI by sex, with men aged 19 years and over having a mean BMI of 28.4kg/m² and women 28.2kg/m². There were also no differences in mean BMI by age group, with means of 28.1kg/m² to 28.5kg/m² across the broad age and sex groups. A higher percentage of men were living with overweight or obesity (74% in men and 63% in women) but a slightly higher percentage of women than men were living with obesity (32% in women and 29% in men) (see table 3.4).

Mean BMI figures for adults in NDNS were slightly higher than HSE 2021 figures (which were all based on adjusted self-reported heights and weights). Among men in NDNS, 74% were living with overweight or obesity compared with 69% in HSE 2021. For women, the figures were 63% and 59%, respectively.

Children

The data for children aged 18 months to 18 years are reported in table 3.5 for interviewer-measured height and weight.

For population monitoring purposes, a child’s BMI is classed as overweight or obese where it is on or above the 85th centile or 95th centile respectively, based on the British 1990 (UK90) growth reference data. The population monitoring cut offs for overweight and obesity are lower than the clinical cut offs (91st and 98th centiles for overweight and obesity) used to assess individual children. This is to capture children in the population in the clinical overweight or obesity BMI categories and those who are at high risk of moving into the clinical overweight or clinical obesity categories. This is outlined in National Child Measurement Programme: operational guidance.

There was very little difference in BMI between girls and boys when based on interviewer- measured height and weight. Similar proportions of boys and girls were:

• living with overweight (14% and 15% respectively)
• living with overweight and obesity combined (36% and 35% respectively)
• living with obesity (21% and 20% respectively)

The most recent HSE data on children for comparison was HSE 2019. The proportion of girls living with obesity appears to be higher in NDNS than in HSE 2019 (also based on interviewer measurements) (20% in NDNS, 13% in HSE), although the age groupings are different. There were 32% of boys and 28% of girls living with overweight and obesity in HSE 2019.

Ethnicity

The ethnicity data presented in table 3.6 shows participants were:

• 86% White
• 6% Asian
• 4% Black
• 3% mixed or multiple ethnic background
• 2% other ethnic background

Census figures on ethnicity are not available for the UK as a whole, but these figures are broadly similar to the latest England and Wales 2021 Census figures which were:

• 82% White
• 9% Asian
• 4% Black
• 3% mixed ethnic groups
• 2% other ethnic groups

Vegetarian and vegan diets

Participants interviewed in fieldwork years 12 to 14 were asked whether they would describe themselves (or their child) as vegetarian, vegan or neither (see table 3.7). Of the participants:

• 3% reported that they were vegetarian
• less than 0.5% reported following a vegan diet

In year 15, participants could describe themselves (or their child) from an expanded range of categories. Of the participants:

• 2% reported that they were vegetarian
• less than 0.5% reported that they were vegan
• 2% reported that they were pescatarian
• 4% described themselves as ‘mainly vegetarian or vegan (occasionally eating meat)’
• 3% reported ‘following a religious practice for eating’

For more information, see table 3.8.

Socio-economic classification

Participants were assigned a National Statistics socio-economic classification (NS-SEC) based on the employment of the ‘household reference person’ for their household. NS-SEC is an Office for National Statistics standard classification. Categories are assigned based on a person’s occupation, whether employed, self-employed, or supervising other employees. Responses are shown in table 3.9.

Equivalised income

The adult completing the household questionnaire was asked to estimate the total household income. This was then used to derive equivalised household income, using the McClements scoring system which takes into account the number and ages of household members (McClements, 1977). Responses were applied to all individual participants in a household, including children.

To enable comparisons, the range of incomes is split equally into tertiles (thirds). As can be seen in table 3.10, the proportion of participants in each equivalised household income tertile was similar:

• 34% in the lowest tertile
• 34% in the middle tertile
• 32% in the highest tertile

However, a higher proportion of children (38%) than adults (31%) were in the lowest tertile and conversely, a higher proportion of adults (37%) than children (28%) were in the highest tertile.

Index of Multiple Deprivation

IMD is a measure of area deprivation, based on 37 indicators, across 7 domains of deprivation. IMD is a measure of the overall deprivation experienced by people living in a neighbourhood, although not everyone who lives in a deprived neighbourhood will be deprived themselves. To enable comparisons, areas are classified into quintiles (fifths) according to their IMD score. Quintiles of IMD are used to give an area-level measure of socio-economic status, as compared with the household-level measure of equivalised household income.

There is no UK wide IMD. Each country in the UK produces its own version using similar methodologies. Differences in the indicators used, the time periods covered, and the sizes of their small areas mean that it is not possible to make direct comparisons between these indices. Results in this report are based on English Indices of Deprivation 2019 and are shown for England only.

The IMD was applied to all individual participants in a household, including children. Overall, 17% of the sample in England was in the most deprived quintile and 23% was in the least deprived. The difference was most pronounced for adults, 15% of whom were in the most deprived quintile and 23% were in the least deprived. For children, the proportions were 20% and 23% respectively (see table 3.11).

Food security

In 2022 to 2023 (fieldwork year 15), the adult who completed the household questionnaire was asked a series of questions about their household’s experiences with accessing and consuming adequate food over the last 12 months. These questions were based on a set of questions on food security developed in the USA. Participants whose response to these questions suggested they may be ‘food insecure’ were asked further questions relating to the last 30 days. Responses were used to derive a food security score which was applied to all individual participants in a household, including children. Results are shown in table 3.12.

The results showed that:

• 81% lived in households with high food security
• 10% lived in marginally food secure households
• 6% in households with low food security
• 4% in households with very low food security

Children were more likely than adults to live in households with low or very low food security. None of the older participants (aged 65 years and over) lived in a household with very low food security.

The findings are in line with those reported in Family Resources Survey: financial year 2022 to 2023, which asked people interviewed between April 2022 and March 2023 about their experiences in the previous 30 days (83% high food security, 7% marginal food security, 5% low food security and 5% very low food security).

NDNS found a lower proportion of food insecure households than the FSA’s ‘Food and You 2 Wave 6 survey’ for England, Wales, and Northern Ireland, which asked participants completing the survey between October 2022 to January 2023 about their experience over the last 12 months. More information can be found in the Food and You 2: Wave 6 Key Findings.

While these other surveys used the same questions, caution should be taken when comparing results with NDNS due to differences in time periods and how the data was analysed.

4. Foods consumed

Reported consumption data by NDNS food groups is presented in tables 4.1 and 4.2 for 2019 to 2023 (fieldwork years 12 to 15). For this report, the NDNS food grouping structure has been reviewed and updated to better reflect current public health priorities and monitoring requirements. Further details can be found in appendix R.

In tables 4.1 and 4.2, all composite dishes, including homemade dishes and manufactured products, are assigned to a food group based on the main components of the dish. For example, beef lasagne has been assigned to ‘beef products and dishes’ and vegetable lasagne has been assigned to ‘vegetable products and dishes’.

Tables 4.1a to 4.1c show mean consumption of NDNS food groups for the total survey population (including non-consumers, who did not report consumption from a particular food group in any of their recalls).

Tables 4.2a to 4.2c show mean consumption of NDNS food groups for consumers only and the percentage of consumers. Consumption figures for these tables have not been calculated using the ‘usual intakes’ method because it was not possible to apply this method to all food groups. Instead, consumption has been calculated using the ‘day average’ method (see chapter 2 and appendix U for more details of this method). No commentary is provided on these tables.

Tables 4.3 to 4.18 present consumption of selected foods for the total survey population (including non-consumers) for 2019 to 2023 and for the previous survey years. Consumption figures for fruit, fruit juice, vegetables, meat and fish in these tables are based on disaggregated data. They include only the contribution of relevant ingredients from composite dishes (both homemade dishes and manufactured products) but exclude the other components of those dishes. For example, the beef component of a beef lasagne is reported under red and processed meat and the vegetable component is reported under vegetables. Consumption of sugar-sweetened soft drinks, sugar and chocolate confectionery, biscuits, buns, cakes and pastries, and crisps and savoury snacks is based on the NDNS food group.

The commentary below focuses on:

• fruit and vegetables
• red and processed meat
• sugar-sweetened soft drinks

The absence of commentary for the other foods does not indicate that there is no public health interest.

In tables 4.3 to 4.18, for both disaggregated and non-disaggregated data, intakes have been calculated using the ‘usual intakes’ method for 2019 to 2023. As this was not possible for oily fish and sugar confectionery, intakes for these products have been calculated using the ‘day average’ method. For foods reported as NDNS food groups, for example sugar-sweetened soft drinks, the values in tables 4.3 to 4.18 calculated using ‘usual intakes’ will differ slightly from those in tables 4.1a to 4.1c calculated using the ‘day average’ method.

No statistical comparisons have been performed to analyse differences between the latest data (2019 to 2023) and the previous set of estimates (2016 to 2019 (fieldwork years 9 to 11)) and no time trend analysis is included in this report. This is due to the change from paper diary to online 24-hour recalls in 2019, which means that results for 2019 to 2023 cannot be directly compared with results from previous years (see chapter 2).

Evidence from the DLW sub-study shows that there continues to be substantial underreporting of energy intake in NDNS. This is common to all dietary surveys. It is likely that there are differences in the extent to which different foods are underreported but it is not possible to say which foods are underreported to a greater or lesser extent. See chapter 2 and appendix X for more information on underreporting.

Fruit and vegetables

The current recommendation is to consume at least 5 portions of a variety of fruit and vegetables per day, as outlined on the NHS page 5 A Day: what counts?. For people aged 11 years and over, a portion is 80g to give a minimum target of 400g per day. While children 10 years and under are also recommended to eat at least 5 portions of a variety of fruit and vegetables a day, no portion size has been set. The number of portions of fruit and vegetables consumed per day has been calculated from disaggregated data for adults and children aged 11 to 18 years based on the 5 A Day criteria. See appendix A for details.

Table 4.3a shows that mean consumption of fruit and vegetable portions for 2019 to 2023 was below the recommendation for all adult age and sex groups and children aged over 11 years.

On average, children aged 11 to 18 years consumed 2.8 portions per day and 9% met the 5 A Day recommendation.

On average, adults aged:

• 19 to 64 years consumed 3.3 portions per day
• 65 to 74 years consumed 3.7 portions per day
• 75 years and over consumed, 3.6 portions per day

This meant 17% of all adults met the 5 A Day recommendation.

Consumption for adults and older adults was lower than previously published figures. This is likely to be partly due to the dietary method change. It may also reflect a real reduction in fruit and vegetable consumption due to a combination of lack of availability during the COVID-19 pandemic and cost of living pressures during this data collection period. Other data sources, including the Department for Environment, Food and Rural Affairs report Family Food 2023, and the Office for Health Improvement and Disparities (OHID) report ‘Changes in food and drink purchasing behaviour and the impact on diet and nutrition: 2021 to 2023’ suggest that households bought less fruit and vegetables during this period (OHID, 2025). See chapter 2 and the stage 3 evaluation report.

Red and processed meat

The current recommendation is that adults with high intakes of red and processed meat (of 90g or more per day) should consider reducing their intakes to no more than 70g per day, as detailed on the NHS page Meat in your diet.

Table 4.8 shows the mean daily intake of red and processed meat based on disaggregated data. See appendix R for details. Mean intake for 2019 to 2023 was below 70g per day in all age and sex groups. Intake was higher in men than women, with the highest mean intake in men aged 19 to 64 years (66g per day). In this age group 27% of men consumed more than 90g of red and processed meat per day compared with 6% of women.

Sugar-sweetened soft drinks

In 2015, the SACN report Carbohydrates and health recommended that the consumption of sugar-sweetened beverages should be minimised by both children and adults.

Table 4.13 shows that for 2019 to 2023 the highest mean consumption of soft drinks with added sugar was in children aged 11 to 18 years (124g per day, equivalent to 124mls per day). The highest consumers in this age group (97.5th percentile) drank 606mls per day. Children aged 18 months to 3 years and 4 to 10 years consumed on average 33mls per day and 54mls per day, respectively. The highest consumers drank 280mls and 371mls per day, respectively. Among adults, consumption was highest in men aged 19 to 64 years (with the mean at 108mls per day and the highest consumers drank 768mls per day). See appendix U for an explanation of 97.5th percentile.

5. Nutrient intakes

Intakes of energy and macronutrients are presented in tables 5.1 to 5.13 and intakes of micronutrients are presented in tables 5.27 to 5.47 for 2019 to 2023 (fieldwork years 12 to 15) and for previous survey years. The percentage contribution of NDNS food groups to energy and macronutrient intakes are presented in tables 5.14 to 5.24 and micronutrients in tables 5.48 to 5.68. See appendix R for more information about NDNS food groups.

Evidence from the DLW sub-study shows that there was substantial underreporting of energy intake in adults and older children, about 30% overall for adults and children. This is common to all dietary surveys and was seen in previous NDNS data. It is not possible to say whether the macronutrient components of energy intake are underreported equally or differentially nor the impact on micronutrient intakes. See chapter 2 and appendix X for more detail.

No statistical comparisons have been performed to analyse differences between the latest data (2019 to 2023) and the previous set of estimates (2016 to 2019 for years 9 to 11 (combined)) and no time trend analysis is included in this report. This is due to the change in methods from paper diary to online 24-hour recalls in 2019 which means that results for 2019 to 2023 cannot be directly compared with results from previous years. See chapter 2.

The commentary in this chapter focuses on a selection of nutrients of public health interest, which are:

• energy
• free sugars
• saturated fatty acids
• fibre

It also combines data from the tables noted above, including:

• describing current intakes for 2019 to 2023
• comparisons with recommendations
• the contribution of food groups to intake

The absence of commentary for a nutrient does not indicate that there is no public health interest or concern about intakes.

SACN statement on expressing fat and carbohydrate recommendations recommended that the dietary reference values (DRVs) for fats and carbohydrates should be expressed as a percentage of energy intake excluding energy from ethanol (alcohol). Macronutrient intakes in this chapter are expressed as a percentage of energy excluding ethanol for comparison with the DRVs. The tables also present intakes as a percentage of total energy.

Energy

Mean daily intakes of total energy and energy excluding ethanol for 2019 to 2023 are shown in tables 5.1a to 5.1d.

Mean total energy intakes for children in kilocalories per day (kcals per day) were:

• 1,188 kcals per day for children aged 18 months to 3 years
• 1,469 kcals per day for children aged 4 to 10 years
• 1,808 kcals per day for boys aged 11 to 18 years
• 1,538 kcals per day for girls aged 11 to 18 years

Mean energy intakes for adults in kcals per day were:

• 1,883 kcals per day for men aged 19 to 64 years
• 1,491 kcals per day for women aged 19 to 64 years
• 1,738 kcals per day for men aged 65 to 74 years
• 1,371 kcals per day for women aged 65 to 74 years
• 1,659 kcals per day for men aged 75 years and over
• 1,423 kcals per day for women aged 75 years and over

Mean daily intakes of total energy exceeded the EAR in children aged 18 months to 3 years but were below the EAR in the other age groups. However, it is unlikely that average energy intakes are actually below recommendations and underreporting of energy intakes is likely to contribute to these findings. The DLW sub-study showed evidence of underreporting particularly in adults and children aged 11 years and over. It is likely that there are people who underreport and overreport in all age groups including younger children, with younger children more likely to overreport and older children and adults more likely to underreport. See appendix X for more information on underreporting.

The per cent contribution of food groups to total energy intake for 2019 to 2023 are shown in table 5.14. The ‘cereals and cereal products’ food group was the largest source of energy for all age groups, contributing on average:

• 39% for children aged 18 months to 3 years
• 43% to 44% for children aged 4 to 18 years
• 38% to 39% for adults aged 19 years and over

‘Milk and milk products and alternatives’ was the second largest contributor to energy intake for children aged 18 months to 3 years (18%).

‘Meat and meat products’ was the second largest contributor to energy intake for children aged 11 to 18 years and adults aged 19 to 64 years (15% to 16%).

Children aged 4 to 10 years and adults aged 65 years and over got a similar proportion of energy from ‘milk and milk products and alternatives’ and ‘meat and meat products’ (12% to 14%).

Saturated fats

Intakes of saturated fats are expressed as a percentage of energy intake excluding ethanol (alcohol) throughout this section. This has been shortened to percentage of energy intake for simplicity.

The government recommendation is that the population average contribution of saturated fats to energy intake should be no more than 10%, as recommended in the SACN report Saturated fats and health. This recommendation applies to all adults. It also applies in full for children aged 5 years and over. It does not apply to children aged 2 years and under.

Mean daily intakes for saturated fats for 2019 to 2023 are shown in tables 5.5a to 5.5d. Mean intake exceeded the recommendation in all age groups to which the recommendation applies (tables 5.5c and 5.5d). In children, mean intake was 12.5% of energy intake. In adults, mean intake was 12.6%. Adults aged 75 years and over had the highest mean intake of saturated fats among adults (13.7% of energy intake).

The recommendation for saturated fats intake was met by:

• 15% of children aged 4 to 10 years
• 16% of children aged 11 to 18 years
• 19% of adults aged 19 to 64 years
• 17% of adults aged 65 to 74 years
• 7% of adults aged 75 years

The per cent contribution of food groups to saturated fat intake in 2019 to 2023 are shown in table 5.17.

The ‘cereals and cereal products’ group was the largest contributor for adults (28% to 31%) and for children aged 4 to 18 years (36% to 37%). This was mainly the contribution from sandwiches (which also includes spread and fillings), pizza, sweet biscuits and cakes. ‘Milk and milk products and alternatives’ made the largest contribution for children aged 18 months to 3 years (38%). ‘Meat and meat products’ contributed 12% to 19% to saturated fat intake across all the age groups.

Free sugars

Intakes of free sugars are expressed as a percentage of energy intake excluding ethanol throughout this section. This has been shortened to percentage of energy intake for simplicity.

The government recommendation is that for those aged one year and over free sugars should provide a population average of no more than 5% of energy, as recommended in the SACN reports Carbohydrates and health and Feeding young children aged 1 to 5 years.

Mean daily intakes for free sugars for 2019 to 2023 are shown in tables 5.11a to 5.11d. In all age groups, mean intake of free sugars exceeded the recommendation. In the 11 to 18 years age group, mean intakes were around double the recommendation (10.9% energy intake for boys and 12% for girls). In the 4 to 10 years and 19 to 64 years age groups, mean intakes were also around double the recommendation (10.2% and 10.4% of energy respectively).

For children, the recommendation for intake of free sugars was met by:

• 27% of those aged 18 months to 3 years
• 8% of those aged 4 to 10 years
• 5% of those aged 11 to 18 years

For adults, the recommendation for intake of free sugars was met by:

• 17% of those aged 19 to 64 years
• 30% of those aged 65 to 74 years
• 15% of those aged 75 years and over

The per cent contribution of food groups to free sugars for 2019 to 2023 are shown in table 5.23.

For all age groups, the largest contributor was ‘cereals and cereal products’ providing 29% to 40% of free sugars intake, with the highest in young children and older adults. This was mainly from the sub-group ‘buns, cakes, pastries and fruit pies’, and from the sub-groups ‘sweet biscuits’ and ‘breakfast cereals’.

For adults, ‘sugar, preserves and confectionery’ was the second largest contributor to free sugars intake, providing 24% to 26%.

For children, ‘non-alcoholic beverages’ was the second largest contributor to free sugars intake, providing 19% to 27%.

The contribution of fruit juice and smoothies to free sugars intakes was:

• 12% in the 18 months to 3 years age group
• 15% in the 4 to 10 years age group
• 13% in the 11 to 18 years age group

The contribution of soft drinks with added sugar to free sugars intakes was:

• 4% in the 18 months to 3 years age group
• 6% in the 4 to 10 years age group
• 13% in the 11 to 18 years age group

Fibre

As recommended in the SACN report Carbohydrates and health, the government recommendation is that the population average intake of fibre for adults should be 30g per day. The recommendations are proportionally lower for children:

• 25g per day for children aged 11 to 16 years
• 20g per day for children aged 5 to 11 years
• 15g per day for children aged 2 to 5 years

For the purposes of reporting the 18 months to 3 years age group, the recommendation has been applied to the whole group, including those 2 years and under.

Mean daily intakes for fibre for 2019 to 2023 are shown in tables 5.12a and 5.12b.

In all age groups, mean intake of fibre was below the recommendations.

For children, mean daily intakes of fibre were:

• 12.5g for those aged 18 months to 3 years
• 14.5g for those aged 4 to 10 years
• 15.4g for those aged 11 to 18 years

For adults, mean daily intakes of fibre were:

• 16.4g for those aged 19 to 64 years
• 16.9g for those aged 65 to 74 years
• 16.4g for those aged 75 years and over

For children, the fibre intake recommendation was met by:

• 22% aged 18 months to 3 years
• 14% aged 4 to 10 years
• 4% aged 11 to 18 years

For adults, 4% met the fibre recommendation.

The per cent contribution of food groups to fibre intake for 2019 to 2023 are shown in table 5.24.

‘Cereals and cereal products’ were the main source of fibre for all age groups, contributing between 43% and 50% to average daily intakes. ‘Vegetables and vegetable products and dishes’ was the second major contributor to fibre intakes (19% to 25%). ‘Fruit’ provided 18% of fibre intakes for children aged 18 months to 3 years and 6% to 12% for the other age groups.

6. Nutritional status biomarkers

Statistics are presented in tables 6.1 to 6.17 for 2019 to 2023 (fieldwork years 12 to 15) for blood and urine analytes chosen for nutritional and public health interest. These include blood biomarkers of:

• anaemia and iron
• vitamin B12
• riboflavin
• vitamin B6
• folate
• vitamin C
• vitamin D
• selenium
• zinc
• cholesterol

It also includes urinary iodine.

No statistical comparisons have been performed to analyse differences between the latest data (2019 to 2023) and the previous set of estimates (2016 to 2019 for years 9 to 11 (combined)) and no time trend analysis is included in this report. This is because, from 2021 (fieldwork year 13) onwards, the protocol for the transport and processing of blood samples was changed and this may have affected the ability to compare between years (see chapter 2 and the blood sample transport report).

The methods for transporting and processing spot urine samples remained unchanged. So, a time trend analysis for urinary iodine is included in this report for 2013 (the start of urinary iodine assessment) to 2023.

Appendix Q describes the laboratory analytical methods for all biomarkers measured in NDNS as well as quality control data for the same period. Appendix U provides information on the analytical approach for generating the statistics and, for iodine, time trend analysis. The laboratory analytical methods for all biomarkers were the same as those used for the 2016 to 2019 data.

The commentary in this chapter focuses on a selection of nutrients of public health interest, which are:

• vitamin D
• folate
• iodine

The absence of commentary for a nutrient does not indicate there is no public health interest.

Vitamin D

Vitamin D is produced in the skin when it is exposed to sunlight containing ultraviolet B (UVB) radiation. In the UK, during the winter months, sunlight containing UVB radiation is limited and so dietary sources are essential. However, vitamin D is found in a limited number of foods and so the UK government advises everyone to take a daily vitamin D supplement (10µg) between October and March. Some population groups, for example those who have no or very little sunshine exposure, are recommended to take a daily supplement containing vitamin D (10µg) throughout the year.

Data on vitamin D intakes including and excluding the contribution from supplements and the percentage of participants taking a supplement containing vitamin D are shown in tables 5.32a to 5.32d.

Vitamin D status (the amount available to the body) is measured by the concentration of 25-hydroxyvitamin D (25(OH)D) in blood. In the UK, risk of poor musculoskeletal health is considered to increase at 25(OH)D concentrations below 25nmol per litre and forms the basis for UK dietary recommendations for vitamin D. This measure of vitamin D status reflects the availability of vitamin D in the body from both skin synthesis and from dietary sources. Appendix Q provides details of the analytical method for 25(OH)D.

The proportion of participants who had a serum 25(OH)D concentration less than 25nmol per litre were:

• 18% of adults aged 19 to 64 years
• 12% of adults aged 65 years and over
• 10% of children aged 4 to 10 years
• 23% of children aged 11 to 18 years

Consistent with previous NDNS, the prevalence of deficiency in adults aged 19 to 64 years appeared marginally greater (3 percentage points) in men than women, although these differences were not tested for statistical significance.

For a more detailed breakdown of vitamin D status, see table 6.9.

Due to the limited availability of UVB radiation of the correct wavelength during the winter months in the UK, production of biologically relevant quantities of vitamin D in the skin cannot occur. So, 25(OH)D concentration shows a strong seasonal pattern in the UK population. Table 6.15 shows 25(OH)D concentration and prevalence of serum concentration below 25nmol per litre by season for each age and sex group.

For each group, 25(OH)D concentration was lowest in January to March and highest in July to September. Mean concentrations in April to June and October to December were similar to each other. The percentage of the population with serum 25(OH)D concentrations below 25nmol per litre was highest in January to March and ranged between 21% and 38% in the age and sex groups 11 years and over, falling to between 0% and 10% in the summer months (July to September). Note that changes with seasonality do not affect interpretation of vitamin D NDNS data because NDNS covers the entire calendar year.

Folate

Folate is a general term for a number of related compounds crucial to metabolic systems including the formation of red blood cells. Folate deficiency may lead to megaloblastic anaemia (a type of anaemia in which the bone marrow produces abnormally large red blood cells that do not function normally) in adults and children. Inadequate folate is a risk factor for:

• neural tube defects (a group of birth defects affecting the brain and spinal cord)
• other poor health outcomes in early pregnancy during fetal development

Folate is found in green leafy vegetables and beans. Folic acid is the synthetic form of folate that is used to fortify foods, such as breakfast cereals, and in micronutrient supplements. While most people should be able to obtain enough folate to meet the body’s requirements from their diet, the government recommends that pregnant women, or women who could get pregnant, should take a 400µg folic acid supplement from before pregnancy until the 12th week of pregnancy. For more information, see the NHS page Pregnancy, breastfeeding and fertility while taking folic acid.

Folate status can be assessed with a number of different biomarkers. Red blood cell (RBC) folate represents folate status during the previous 3 to 4 months, mirroring tissue folate stores. As such, RBC folate is less sensitive to short-term fluctuations in folate intake or metabolism and provides a better longer-term marker of population folate status. Serum folate concentration responds to recent changes in folate intake and provides information on folate status over the short term.

Red blood cell (RBC) folate

RBC folate is calculated from:

• whole blood folate
• serum folate
• haematocrit

For more information on the analytical methods used and on calculating RBC folate, see appendix Q.

Based on haematological indicators and risk of megaloblastic anaemia, the threshold for folate deficiency assessed with RBC folate is 305nmol per litre (Institute of Medicine, 1998). The threshold of 748nmol per litre for RBC folate in women of childbearing age (aged 16 to 49 years) is the level below which there is increased risk of neural tube defects (Tinker and others, 2015).

In 2019 to 2023, RBC folate concentration was less than 305nmol per litre in:

• 12% of children aged 11 to 18 years
• 4% of adults aged 19 to 64 years
• 2% of adults aged 65 years and over
• 7% of women of childbearing age (16 to 49 years)

For a more detailed breakdown of RBC folate concentrations, see table 6.6.

More than twice the percentage of girls aged 11 to 18 years (17%) had RBC folate concentration less than 305nmol per litre compared with boys (7%) of the same age.

The proportion of women of childbearing age (16 to 49 years) with RBC folate concentration below the neural tube defect threshold of 748 nmol per litre was 83%.

These findings provide a better reflection of RBC folate status than findings from earlier survey years due to the change in blood processing and transport methods and resulting shorter delay in sample processing. This has resulted in lower proportions of each population group with an RBC concentration less than the 305nmol per litre threshold compared to the previous NDNS report. Similarly, a lower proportion of women of childbearing age had an RBC concentration below the neural tube defect threshold, but the proportion remained high.

Serum folate

In NDNS, serum folate concentration is calculated from the sum of individual folate forms, including folic acid (see appendix Q for information on the analytical methods). Based on risk of megaloblastic anaemia, a serum folate concentration below 13nmol per litre indicates possible deficiency and 7nmol per litre indicates clinical deficiency (WHO, 2015). There is no internationally accepted serum folate threshold for risk of neural tube defects.

In 2019 to 2023, serum folate concentration was less than 7nmol per litre in:

• 8% of children aged 4 to 10 years
• 24% of children aged 11 to 18 years
• 14% of adults aged 19 to 64 years
• 7% of adults aged 65 years and over

For a more detailed breakdown of serum folate concentrations, see table 6.7.

Iodine

Iodine is required for the production of thyroid hormones that are essential for healthy cells. In children, insufficient iodine intake is a risk factor for reduced growth and cognitive function. Population iodine status is assessed by measuring urinary iodine concentration (UIC). Most iodine consumed in the diet is excreted in urine and UIC reflects and fluctuates with recent dietary intake. When measured in populations, these fluctuations even out and median UIC can be used to assess population iodine dietary supply and so provide a population-level indication of iodine status.

In NDNS, urinary iodine concentration was introduced into NDNS in 2013 (fieldwork year 6). There were no processing or analytical method changes related to urinary iodine measurement in 2019 to 2023. See appendix Q for details of the analytical method used for urinary iodine.

As outlined in the WHO guidance on iodine deficiency, criteria for adequate population iodine status is a median UIC between 100µg per litre and 199µg per litre and fewer than 20% of the population below 50µg per litre. Median UIC less than 100µg per litre indicates insufficient population iodine intake.

In 2019 to 2023, girls aged 11 to 18 years and adults (both men and women) aged 19 to 64 years had median UIC less than 100µg per litre (see table 6.16).

Of girls aged 11 to 18 years, 29% had UIC below 50µg per litre and the median UIC was 95µg per litre whereas in boys in the same age group, 12% had UIC below 50µg per litre and the median UIC was 141µg per litre.

In adults aged 19 to 64 years and women of childbearing age (16 to 49 years) 25% and 30% respectively had UIC below 50µg per litre, with median UIC of 89 µg per litre and 82µg per litre, respectively.

Median UIC in children aged 4 to 10 years was 144µg per litre and 13% had UIC below 50µg per litre.

Time trend analysis (table 6.17) indicated a significant average decrease in UIC in girls aged 11 to 18 years of 3.4% (confidence interval (CI) 0.7, 6.0) per year between 2013 and 2023. Similarly, in adults aged 19 to 64 years there was a significant average year-on-year decrease in UIC of 2.9% (CI 1.9, 3.9), and this was the same for both men and women. These represent decreases of 29% for girls aged 11 to 18 years and 25% for adults aged 19 to 64 years over 10 years (2013 to 2023). See appendix U for more information on confidence intervals.

7. Equivalised income and foods, nutrients and nutritional status

Equivalised household income analysis for 2019 to 2023 (fieldwork years 12 to 15) has been carried out on foods, nutrients and markers of nutritional status as listed in chapter 1 and results are presented in worksheets 7.1 to 7.27.

Equivalisation is a standard methodology that adjusts household income to account for different demands on resources, by considering the household size and composition. The adult completing the household questionnaire was asked to estimate the total household income, before any deductions (such as for tax). This included income from:

• earnings
• self-employment
• benefits
• pensions
• interest from savings

For the equivalised income analysis, the average change in each variable per £10,000 increase in equivalised household income was estimated (using the slope of the regression line on the plots in worksheets 7.1 to 7.27) from a linear regression model along with the 95% confidence interval. £10,000 was selected as a convenient increment to assess the size of the change in each variable, but it should not be interpreted as having any diet-related meaning.

Appendix U provides a full explanation of the analytical approach. Due to differences in the variation of the data points or sample size within each of the age and sex groups (see table 9.0), there are instances for some foods, nutrients and blood analytes where larger slopes were not statistically significant, whereas smaller slopes were statistically significant. Where there are limited data points within an age and sex group (for example, oily fish) interpreting the magnitude of change per £10,000 increase should be done cautiously.

The commentary in this chapter describes the findings for some foods, nutrients and biomarkers of public health interest. It focuses on the magnitude of change, and upward or downward trends where these are considered nutritionally meaningful, rather than statistical significance. Statistical significance is indicated by the confidence intervals set out in brackets in the text.  

The text in this report does not describe the actual group mean for each income decile. Trends in arithmetic mean are reported as ‘change per £10,000’ where the dietary data were normally distributed and could be analysed without transformation. Where the dietary data was skewed and needed to be log-transformed before analysis, the trends in geometric mean are reported as ‘percentage change per £10,000’. There is no commentary where a regression line cannot be fitted (where most of the data is zero, where there is a clear non-linear relationship, where the number of data points within an age and sex group was less than 30). See appendix U for more explanation of the statistical analysis in this chapter.

Distribution of equivalised income data

On inspection, NDNS equivalised income data appears positively skewed, but a log transformation resulted in a negative skew. So, no transformation was applied to the income data prior to the regression analysis. Investigation of high income responses indicated that they did not unduly affect the regression slope. For more information see appendix U.

Foods

For trends in ‘sugar-sweetened soft drinks’, biscuits and ‘buns, cakes and pastries’, there was a high proportion of non-consumers in some or all age and sex groups, so ‘percentage of consumers’ and ‘intakes for consumers only’ are presented instead of population intakes. This is because the regression analysis of population intakes is highly influenced by zero values which can be misleading.

Fruit and vegetables (5 A Day portions)

5 A Day portions have been calculated for the following age groups:

• 11 to 18 years
• 19 to 64 years
• 65 to 74 years
• 75 years and over

Refer to chapter 4 for more details.

Worksheet 7.1 shows that with increasing household income, the number of 5 A Day portions consumed per day increased for all age and sex groups, although in most age groups this change was small (0.1 portions per day per £10,000 increase in equivalised income on average). The largest increase was seen in men aged 65 to 74 years, with a 0.3 portions per day increase (about 24g of fruit and vegetables) (CI 0.2, 0.4) for every £10,000 increase in equivalised income.

The figures in worksheet 7.2 show that the proportion achieving the 5 A Day recommendation was low regardless of income. However, for most age and sex groups (where it was possible to estimate income trends), small increases in the proportion meeting the 5 A Day recommendation were seen with increasing household income.

The largest increase was for men aged 65 to 74 years, where the proportion meeting the recommendation increased by 6 percentage points (CI 6, 7) for every £10,000 increase in equivalised income.

Sugar-sweetened soft drinks

Worksheet 7.6 shows that there was little change in the percentage consuming sugar-sweetened soft drinks with increasing household income for most age and sex groups.

The largest change was for girls aged 4 to 10 years where the percentage consuming sugar-sweetened soft drinks decreased by 4 percentage points (CI 2, 6) for every £10,000 increase in equivalised income.

There was a downward trend in quantities of sugar-sweetened soft drinks being consumed with increasing income among:

• children aged 18 months to 3 years
• boys aged 4 to 10 years
• women aged 19 to 64 years (worksheet 7.7)

Biscuits and cereal bars

Worksheet 7.12 shows that for males, there was an upward trend in the percentage consuming biscuits (sweet and savoury) and cereal bars with increasing household income. The exception was men aged 75 years and over, where the percentage of consumers decreased with increasing income.

It was not possible to estimate income trends for some female age groups (as fewer females reported consuming biscuits and cereal bars). But the percentage of female consumers aged 75 years and over, as with males, decreased with increasing income.

There was no consistent pattern across age and sex groups for quantities of biscuits and cereal bars being eaten with increasing income (worksheet 7.13).

Buns, cakes and pastries 

Worksheet 7.14 shows that for most age and sex groups, the percentage of consumers of buns, cakes and pastries increased with increasing household income. The largest increases were seen among:

• children aged 18 months to 3 years
• women aged 65 to 74 years
• men aged 75 years and over

Each of these increased by 4 percentage points (CI 2, 6, CI 0, 7 and CI 0, 8 respectively) for every £10,000 increase in equivalised income.

There was no consistent pattern across age and sex groups with increasing income for the quantities consumed, and the differences were small (worksheet 7.15).

Nutrients 

Macronutrient intakes in this chapter are expressed as a percentage of energy intake excluding energy from ethanol (alcohol), as outlined in the SACN statement on expressing fat and carbohydrate recommendations. This has been shortened to ‘percentage of energy intake’ for simplicity.

Energy 

Worksheet 7.18 shows that for children there was little change in total energy intake with increasing household income.

In most adult age and sex groups, total energy intake increased on average with increasing household income. Changes were small with the largest seen in men aged 65 to 74 years, where total energy intake increased by 40kcal per day (CI 12, 67) for every £10,000 increase in equivalised income.

Free sugars

The figures in worksheet 7.20 show the majority of free sugars intakes exceeded the recommendation of no more than 5% of energy across the range of equivalised income for all child age groups and for adults aged 19 to 64 years, as indicated by the regression line.

For adults aged 65 to 74 years and women 75 years and over, the regression line indicated that people in households with the highest equivalised incomes would be more likely to meet the recommended intake.

Free sugars intake as a percentage of energy fell with every £10,000 increase in equivalised income for:

• boys aged 4 to 10 years (0.3 percentage points (CI 0.1, 0.5))
• adults aged 65 to 74 years (0.4 percentage points (CI 0.1, 0.6))
• women 75 years and older, (0.3 percentage points (CI 0.0, 0.6))

Fibre

The figures in worksheet 7.21 show that for boys and girls aged 11 to 18 years and all adult age and sex groups, the majority of fibre intakes were below recommendations (see chapter 5).

For all age and sex groups, fibre intake increased on average with increasing income. The largest increase in fibre intake was 0.8g per day (CI 0.3, 1.2) for every £10,000 increase in equivalised income seen in men aged 65 to 74 years.

For children, while the increases with income were smaller (0.3 to 0.4g per day), the regression line indicated that those in higher equivalised income households would be more likely to meet recommended intakes.

Sodium

Worksheet 7.22 shows that for both children and adults, changes in sodium intake with increasing household income were small and not in a consistent direction. Intakes are based on estimated sodium content of foods consumed. They do not include sodium from salt added by participants during cooking or at the table. 

Nutritional status biomarkers

Red blood cell (RBC) folate

Worksheet 7.26 indicates no association between equivalised income and RBC folate concentration in children, adult men or older women. In women of childbearing age (16 to 49 years), the data suggest that RBC folate concentration was 2% (CI 0, 5) higher for every £10,000 increase in equivalised income.

Vitamin D

Worksheet 7.27 shows that in adults aged 19 to 64 years, vitamin D status (25-hydroxyvitamin D concentration) increased by 1 nmol per litre (CI 0, 2) for every £10,000 increase in equivalised income.

Among women aged 19 to 64 years, the increase was 2 nmol per litre (CI 1, 3) for every £10,000 increase in equivalised income. A similar magnitude of increase was observed in children between 4 and 18 years.

In contrast to the other age and sex groups, for adults aged 65 years and over, and men in particular in this age group, the regression analysis indicated decreases in vitamin D status for every £10,000 increase in equivalised income.

8. Index of Multiple Deprivation and foods, nutrients and nutritional status

IMD is a measure of the overall deprivation experienced by people living in a neighbourhood, although not everyone who lives in a deprived neighbourhood will be deprived themselves. IMD is assigned to an area rather than to a specific address or a household. Quintiles of IMD are used to give an area-level measure of socioeconomic status, as opposed to the household-level measure of equivalised household income. Further details on IMD and equivalised household income are given in chapter 3.

This chapter presents quintile analysis by age group for IMD and equivalised household income for 2019 to 2023 to compare the associations with diet and nutritional status. Analysis in this chapter is for England only. Versions of the IMD variable for Scotland, Wales and Northern Ireland will be available in the archived data set.

The term ‘England Index of Multiple Deprivation’ has been abbreviated throughout the chapter to ‘EIMD’ and ‘equivalised household income quintile’ has been abbreviated to ‘income quintile’.

Statistics for EIMD and income quintiles are presented in worksheets 8.1 to 8.29. Due to small numbers, results for foods and nutrients are presented for men and women combined for age groups:

• 18 months to 3 years
• 4 to 10 years
• 11 to 18 years
• 19 to 64 years
• 65 years and over

IMD quintile 1 contains the most deprived areas and quintile 5 contains the least deprived areas. Equivalised income quintile 1 is the group with the lowest equivalised household income and quintile 5 is the group with the highest equivalised household income. For the nutritional biomarkers RBC folate and 25-hydroxyvitamin D, the number of participants aged 18 months to 18 years was too small to be subdivided into quintiles and these groups are not presented in the tables.

The commentary in this chapter describes the findings for foods, nutrients and markers of nutritional status, selected for their public health interest. It describes intakes and status in relation to IMD and equivalised income quintiles and any clear patterns, for example where there is an increase or decrease across the quintiles.

No statistical testing has been carried out as there is no link between statistical significance and nutritional meaningfulness and to avoid issues with multiple comparisons. Appendix U provides a full explanation of the analytical approach.

If comparing EIMD and income quintile, note that the former is an area-level measure of deprivation while income is based on the household’s characteristics. For example, households with a relatively high equivalised income may live in areas included in the most deprived quintile and conversely, households with a low equivalised income might live in a less deprived area.

Foods

For all foods, mean consumption is for the total survey population in England (that is including non-consumers, those who did not report consumption from that food group in any of their recalls). As the numbers in some of the quintiles for children aged 18 months to 3 years are less than 50 they are not included in the commentary.

Fruit and vegetables (5 A Day portions)

5 A Day portions have been calculated for the age groups:

• 11 to 18 years
• 19 to 64 years
• 65 years and over

Worksheet 8.1a shows that for all 3 age groups, mean consumption of 5 A Day portions was lowest in the most deprived EIMD quintile, which were:

• 2.4 portions per day in children aged 11 to 18 years
• 2.8 portions per day in adults aged 19 to 64 years
• 2.6 portions per day in 65 years and over

For children aged 11 to 18 years and adults aged 65 years and over, mean consumption of 5 A Day portions was highest in the least deprived EIMD quintile, which were:

• 3.3 portions per day in children aged 11 to 18 years
• 4.3 portions per day in adults aged 65 years and over

For adults aged 19 to 64 years mean consumption of 5 A Day portions in quintiles 3, 4 and 5 were very similar.

For children aged 11 to 18 years and adults aged 65 years and over, mean consumption of 5 A Day portions by income quintile was, as seen with EIMD, lowest in the lowest income quintile and highest in the highest income quintile (worksheet 8.1b).

For adults aged 19 to 64 years mean consumption was highest in EIMD quintile 5 but similarly low in quintiles 1, 2 and 3.

For both EIMD and household income, mean consumption was below 5 A Day for all quintiles in all age groups.

Worksheet 8.2a shows that the proportion achieving the 5 A Day recommendation was low in all EIMD quintiles but tended to be lower in the more deprived quintiles. For children aged 11 to 18 years and adults aged 19 to 64 years, the percentage meeting the recommendation was lowest in the most deprived EIMD quintile (6% and 11%) and highest in the least deprived EIMD quintile (13% and 21%). For adults aged 65 years and over, the percentage meeting the recommendation was lowest in EIMD quintile 2 (16%) and highest in EIMD quintile 4 (30%).

The percentage of adults aged 65 years and over who achieved the 5 A Day recommendation in the highest income quintile (33%) was more than double that in the lowest income quintile (15%) (worksheet 8.2b). No clear pattern by income quintile was observed in the other 2 age groups.

Sugar-sweetened soft drinks

For children aged 4 to 10 years and 11 to 18 years, the percentage consuming sugar-sweetened soft drinks was highest in the most deprived EIMD quintile (38% and 60%) (worksheet 8.6a). For adults, there was no clear pattern by EIMD quintile in the highest percentage consuming sugar-sweetened soft drinks. There was no clear pattern across the age groups of the EIMD quintile with the lowest percentage of consumers.

Mean consumption of sugar-sweetened soft drinks was highest in the most deprived EIMD quintile for all age groups except those aged 65 years and over (worksheet 8.7a). For children aged 11 to 18 years there was a clear decrease in consumption from the most deprived to the least deprived EIMD quintile.

For equivalised income, there was no consistent pattern across the age groups in the percentage consuming sugar-sweetened soft drinks nor the amounts consumed by quintile (worksheets 8.6b and 8.7b).

Biscuits and cereal bars

For all age groups, the percentage of consumers of biscuits (sweet and savoury) and cereal bars was generally higher in the less deprived EIMD quintiles (worksheet 8.12a). There was no clear pattern across the age groups in the percentage consuming biscuits and cereal bars by income quintile (worksheet 8.12b).

No clear pattern was observed across the age groups in the mean consumption of biscuits and cereal bars by EIMD nor income quintiles (worksheets 8.13a and 8.13b).

Buns, cakes and pastries

Worksheet 8.14a shows the percentage consuming buns, cakes and pastries was highest in the least deprived EIMD quintile for all age groups except those aged 19 to 64 years. The percentage of consumers was lower in the more deprived quintiles.

For children aged 4 to 10 years there was a marked difference in the percentage of consumers in the least deprived quintile (74%) compared with the other quintiles (44% to 53%). There was no clear pattern in the percentage consuming buns, cakes and pastries by income quintile (worksheet 8.14b).

Mean consumption of buns, cakes and pastries was lowest in the most deprived EIMD quintile for all age groups except those aged 11 to 18 years (worksheet 8.15a). For adults aged 65 years and over there was an upward trend in consumption moving from the most deprived to the least deprived EIMD quintile.

There was no clear pattern in the amount of buns, cakes and pastries consumed by income quintile (worksheet 8.15b).

Nutrients

Macronutrient intakes in this chapter are expressed as a percentage of energy intake excluding energy from ethanol (alcohol), as outlined in the SACN statement on expressing fat and carbohydrate recommendations. This has been shortened to percentage of energy intake for simplicity. As the numbers in some of the quintiles for children aged 18 months to 3 years are less than 50 they are not included in the commentary.

Energy

Worksheet 8.18a shows that for adults aged 65 years and over, total energy intake was lowest in the most deprived EIMD quintile (1,339 kcal per day) and highest in the least deprived EIMD quintile (1,643 kcal per day). For the other age groups, mean total energy intakes were similar across the EIMD quintiles.

For equivalised income, total energy intake was lowest in the lowest income quintile (1,534 kcal per day) and highest in the highest income quintile (1,800 kcal per day) for adults aged 19 to 64 years (worksheet 8.18b). For adults aged 65 years and over, total energy intake was also highest in the highest income quintile but similar in the other 4 quintiles. For children, mean total energy intakes were similar across the income quintiles.

Free sugars

Worksheet 8.20a shows that for all age groups, mean free sugars intake exceeded the recommendation of no more than 5% of energy across the EIMD quintiles. In all age groups, mean intake of free sugars as a percentage of energy was highest in the most deprived EIMD quintile. There was no consistent pattern for the lowest intake of free sugars by EIMD quintile.

For adults, free sugars intake was highest in the lowest income quintile (worksheet 8.20b). For adults aged 65 years and over, there was a 2.8 percentage point difference in free sugars as a percentage of energy between the lowest income quintile (10.4%) and the highest income quintile (7.6%). There was no clear pattern by income quintile in children.

Fibre

Mean fibre intakes were below recommendations in all EIMD and income quintiles for all age groups.

For children aged 11 to 18 years and for adults there was an upward trend in mean fibre intake from the most deprived to the least deprived EIMD quintile (worksheet 8.21a). This was most marked in adults aged 65 years and over where there was a 5.2 grams (g) per day difference in mean consumption between the most deprived EIMD quintile (13.6g per day) and the least deprived (18.8g per day).

For most age groups, mean fibre intake was very similar across income quintiles 1 to 4 and was slightly higher in the highest income quintile (worksheet 8.21b). For adults aged 65 years and over, mean fibre intake was 20.3g per day in the highest income quintile compared with 16.7g per day or less in the other 4 quintiles.

Sodium

Intakes are based on sodium content of foods consumed. They do not include sodium from salt added by participants during cooking or at the table.

Worksheet 8.22a shows that for adults aged 65 years and over, mean sodium intake was lowest in the most deprived EIMD quintile (1,464mg per day) and highest in the least deprived EIMD quintile (1,629mg per day). For the other age groups, there was no clear pattern in mean sodium intake by EIMD quintile.

Mean sodium intake for adults aged 65 years and over was similar across the income quintiles (worksheet 8.22b). For adults aged 19 to 64 years, mean sodium intake was lowest in the lowest income quintile (1,607mg per day) and highest in the highest income quintile (1,873mg per day). For the other age groups, there was no consistent pattern in mean sodium intake by income quintile.

Nutritional status biomarkers

For nutritional status biomarkers, numbers of participants are less than 50 in all quintiles for the 3 child age groups and in some quintiles for adults aged 65 years and over. So, commentary is restricted to adults aged 19 to 64 years.

Red blood cell (RBC) folate

Worksheet 8.26a shows that in adults aged 19 to 64 years, the lowest RBC folate concentration was observed in the most deprived EIMD quintile but there was no consistent pattern with other EIMD quintiles.

For equivalised income, in adults aged 19 to 64 years, mean RBC folate concentrations were very similar in quintiles 1 to 3 and were slightly higher in quintiles 4 and 5 (worksheet 8.26b).

The percentage of adults aged 19 to 64 years with RBC folate below 305nmol per litre (the clinical threshold for folate deficiency) was highest in the most deprived EMID quintile (10%) and in income quintile 2 (9%) (worksheets 8.27a and 8.27b). In the least deprived quintile for both EIMD and income, the percentage of adults aged 19 to 64 years with RBC folate concentration below 305nmol per litre was 2%.

Vitamin D

For adults aged 19 to 64 years, the mean serum concentration of 25-hydroxyvitamin D (25(OH)D) was lowest in the more deprived EMID quintiles 1 and 2 (worksheet 8.28a). A similar pattern was observed by income quintile (worksheet 8.28b).

In the most deprived EMID quintile, 34% of adults aged 19 to 64 years had a 25(OH)D concentration below 25nmol per litre (threshold for deficiency in respect to bone health) compared with between 9% and 17% in the other 4 quintiles (worksheet 8.29a). A similar pattern was observed when considering income quintiles with the lower income quintiles 1 and 2 having a higher proportion of adults aged 19 to 64 years with 25(OH)D concentration below 25nmol per litre (worksheet 8.29b).

9. Food and drink from the out of home sector

Data on purchasing and consumption of food and drinks from the out of home sector is presented in tables 9.1 to 9.7. The data was collected in the stage 1 CAPI questionnaire and in Intake24.

‘Out of home sector’ refers to any outlet where food or drink is prepared in a way that means it is ready for immediate consumption, on or off the premises (for example a cafe, pub or takeaway).

Data from the CAPI questionnaire

As part of the stage 1 CAPI questionnaire, participants were asked how often they bought food and drink from the out of home sector. Participants were shown a list of places where they might buy food or drink to eat on the premises, elsewhere, or on the go, and asked if they had bought food or drinks from these places in the last 7 days.

As shown in table 9.1, 72% of participants reported buying food or drink from the out of home sector in the last 7 days:

• 42% bought a drink (on its own or with food)
• 38% bought lunch
• 36% bought an evening meal
• 36% bought snacks
• 15% bought breakfast

Purchases from the out of home sector were most often made 1 to 2 times a week. Less than 4% reported buying food or drink every day in the last 7 days. Children aged 11 to 18 years (81%) and adults aged 19 to 64 years (77%) were most likely to report buying food or drink from the out of home sector.

Participants aged 16 years and over and in full or part-time employment were asked what they usually did about meals when they were at work. Overall, 59% said they brought food from home and 22% said they ate food from a work canteen or cafe or shop (table 9.2). Participants aged 16 to 18 years were more likely to eat food from a work canteen or cafe or shop (41%) compared with those aged 19 to 64 years (22%) and those aged 65 years and over (16%).

Participants aged 18 years and under and in full-time education were asked what they usually did for lunch on a school or college day. For participants aged 4 to 15 years, the majority reported having a cooked school lunch (58% of those aged 4 to 10 years and 48% of those aged 11 to 15 years - see table 9.3). For older children aged 16 to 18 years:

• 38% reported bringing a packed lunch from home
• 34% reported having a cooked school or college lunch
• 12% reported buying lunch from a shop or cafe

Ninety per cent of children aged 4 to 10 years reported having nowhere at school to buy drinks or snacks (separately from meals). For children aged 11 to 18 years, who were more likely to be able to buy drinks and snacks at school, 50% of those aged 11 to 15 years and 59% of those aged 16 to 18 years reported buying drinks or snacks.

The following proportions reported having some sort of free or subsidised food or drink at school:

• 52% of children aged 4 to 10 years
• 22% of children aged 11 to 15 years
• 14% of those aged 16 to 18 years

Also, 7% of children aged 4 to 10 years reported having a free school lunch not as part of universal free school meals.

To note, universal free school meals were available in England to all children in full-time education from reception to year 2 (ages 4 to 7 years) and to all primary school children in Scotland (primary 1 to primary 3 since 2015 and primary 4 to 5 since 2021) during the data collection period. But universal provision of free school meals was not available in Northern Ireland nor Wales during this time.

Data from Intake24

This section provides new estimates of the contribution of the out of home sector to energy and nutrient intakes based on information collected in Intake24 on the source of foods and drinks consumed. It is not possible to compare this to previous NDNS estimates. For each eating occasion (meal, snack or drink) recorded in Intake24, participants were asked to record where they had bought or obtained most of the food and drink for that occasion (table 9.4). Around half of participants (54%) recorded at least one eating occasion where the food had come from the out of home sector, including:

• 23% from a fast food or takeaway outlet
• 17% from a sit-down restaurant or pub
• 17% from a cafe, coffee shop, sandwich bar or deli

The age groups most likely to report at least one occasion where their food had come from the out of home sector were:

• 4 to 10 years - 58%
• 11 to 18 years - 60%
• 19 to 64 years - 57%

The age groups least likely to report at least one occasion where their food had come from the out of home sector were:

• 18 months to 3 years - 44%
• 65 to 74 years - 40%
• 75 years and over - 41%

These estimates, based on up to 4 recalls, are lower than the figure from CAPI which asked about the last 7 days.

Table 9.5 presents the same data as a percentage of eating occasions for each age group. Eating occasions in which food had been mostly bought from the out of home sector (including food delivery services and takeaways) accounted for 10% of all eating occasions overall, 11% for the 19 to 64 years age group and 12% for the 11 to 18 years age group.

Overall, food and drink consumed at out of home eating occasions contributed 12% of total energy intake (table 9.6). This increased to 23% when only those participants who reported consumption from the out of home sector were included (table 9.7). Similar proportions were seen for other nutrients including saturated fats, free sugars and sodium. Food and drink from the out of home sector contributed more to the energy and nutrient intakes of those aged 11 to 18 years and 19 to 64 years than of those in the younger and older age groups. When looking at out of home consumers only, 25% of average energy intake came from food and drink from the out of home sector for the 11 to 18 years age group and 24% for the 19 to 64 years age group.

Acknowledgements

This report was prepared by:

• Caireen Roberts, Kerry Jones, David Collins and Polly Page (MRC Epidemiology Unit)
• Beverley Bates (NatCen)
• Gillian Swan and Jo Nicholas (OHID)

The authors would like to thank all who gave up their time to take part in the NDNS. They would also like to acknowledge the professionalism and commitment of the fieldworkers who worked on the survey.

The authors would also like to thank everyone who contributed to the work behind this report and its production, particularly the following people.

Colleagues at the MRC Epidemiology Unit: Suzanna Abraham, Jean Adams, Birdem Amoutzopoulos, Rachel Barratt, Auguste Boge, Soren Brage, Carol Dorling, Anila Farooq, Jackie Foreman, Nita Forouhi, Dani Ghosh, Dan Griffiths, Lewis Griffiths, Stefanie Hollidge, Lukas Horch, Steve Knighton, Albert Koulman, Amanda McKillion, Sarah Meadows, Anna Melachrou, Lydia Moore, Steph Moore, Angela Mulligan, Samuel Odeyemi, Elise Orford, Tolulope Osunnuyi, Damon Parkington, Ivan Poliakov, Ann Prentice, Kirsten Rennie, Antonia Smith, Toni Steer, Tessa Strain, Rebecca Stratford, Tabasum Tabasum, Kirsty Trigg, Michelle Venables, Ella Westdrop and Nick Wareham.

Colleagues at NatCen: Steve Edwards, Suzanne Hill, Jess Melling, Simrit Roop, Dhru Shah, Mari Toomse-Smith.

Barbara Muldoon, Michael Guiney and colleagues at the Northern Ireland Statistics and Research Agency for organising and carrying out interviewer fieldwork in Northern Ireland.

Colleagues at Centers for Disease Control and Prevention laboratories, Atlanta for carrying out whole blood folate analyses.

Colleagues at the Core Biochemical Assay Laboratory and Pathology Department, Addenbrook’s Hospital, Cambridge University Hospitals for blood analyses.

Colleagues at the Trace Element Laboratory, University Hospital Southampton.

Elaine Gunter (Specimen Solutions, Limited Liability Company) for an independent quality review of laboratory procedures and analyses.

Members and former members of the NDNS Project Board: Susan Fairweather-Tait, Mairead Kiely, Julie Lovegrove, Hilary Powers and Sian Robinson (Scientific Advisory Committee on Nutrition), Jenny Mindell and Jayne Woodside (independent).

Colleagues at OHID: Adrienne Cullum, Paul Niblett and Celia Sabry-Grant.

Colleagues at FSA: Robin Clifford and Joseph Shavila.

Colleagues who worked on the survey at Public Health England (up to 2021): Mark Bush, Louis Levy and Danielle Weiner.

Colleagues at Food Standards Scotland: Gillian Fraser, Fiona Comrie and Gillian Purdon.

Colleagues at the Food Standards Agency in Northern Ireland: Emily Chan, Naomi Davidson, Aoibheann Dunne and Brídín Nally.

Colleagues at the Welsh Government: Sarah Rowles and Chris Roberts.

This research was also supported by the National Institute for Health and Care Research (NIHR) Cambridge Biomedical Research Centre (funding award number: NIHR203312).

References

Food Foundation. The Broken Plate 2023 (viewed 1 May 2025).

Hoenink JC, Garrott K, Jones NRV, Conklin AI, Monsivais P and Adams J. Changes in UK price disparities between healthy and less healthy foods over 10 years: An updated analysis with insights in the context of inflationary increases in the cost-of-living from 2021. Appetite 2024: volume 197, article 107290.

Jones KS, Meadows SR, Chamberlain K, Parkington DA, Collins D, Page P and Koulman A. Delayed processing of chilled whole blood for 24 hours does not affect the concentration of the majority of micronutrient status biomarkers. The Journal of Nutrition 2021: volume 151, issue 11, pages 3,524-3,532.

McClements LD. Equivalence scales for children. Journal of Public Economics 1977: volume 8, issue 2, pages 191-210.

OHID, 2025. Changes in food and drink purchasing behaviour and the impact on diet and nutrition: 2021 to 2023 (viewed 5 June 2025).

Poslusna K, Ruprich J, de Vries JHM, Jakubikova M and van’t Veer P. Misreporting of energy and micronutrient intake estimated by food records and 24 hour recalls, control and adjustment methods in practice. British Journal of Nutrition 2009: volume 101, issue 2, pages S73-S85.

Souverein O, Dekkers A, Geelen A, Haubrock J, de Vries JH, Ocké MC, Harttig U, Boeing H and van ‘t Veer P. Comparing four methods to estimate usual intake distributions. European Journal of Clinical Nutrition 2011: volume 65, supplement 1, pages S92-S101.

Tinker SC, Hamner HC, Ping Qi Y and Crider KS. U.S. women of childbearing age who are at possible increased risk of a neural tube defect-affected pregnancy due to suboptimal red blood cell folate concentrations, National Health and Nutrition Examination Survey 2007 to 2012. Clinical and Molecular Teratology 2015: volume 103, issue 6, pages 517-526.

Appendices

The report’s appendices can be found at NDNS 2019 to 2023 appendices. All the appendices are listed below.

• Appendix A: dietary data collection and processing

• Appendix B: methodology

• Appendix C: stage 1 participant documents

• Appendix D: stage 1 (including CAPI)

• Appendix F: physical activity questionnaire

• Appendix G: stage 2 participant documents

• Appendix H: consent forms

• Appendix I: stage 2 (including CAPI)

• Appendix J: dietary feedback and GP letter examples

• Appendix L: measurement protocols

• Appendix M: spot urine methodology

• Appendix P: blood sample collection and processing and priority order of blood analytes (tables)

• Appendix Q: blood and urine sample analysis methods

• Appendix R: food groups and food groups tables

• Appendix T: topics in report and UKDS

• Appendix U: statistical methods

• Appendix V: physical activity and accompanying data tables

• Appendix X: misreporting – doubly labelled water study

• Appendix BB: sampling and weighting and accompanying data tables

1. Intake24 (UK Locale, System Version 3, 2019, Cambridge University). Intake24 was developed by Newcastle University, originally with funding from Food Standards Scotland and is licenced under the Open Government Licence. The tool is now maintained and developed in collaboration by Cambridge University, Monash University (Australia) and Newcastle University (UK). The version of Intake24 used for NDNS is provided and adapted by the University of Cambridge, based on the original, with technical advisory input from Newcastle University. ↩