National Nutrition and Physical Activity Survey methodology

Latest release
Reference period
2023
Released
5/09/2025
Next release Unknown
First release
Release date and time
05/09/2025 11:30am AEST

Overview

Scope

Includes:

  • usual residents in Australia aged 2+ years living in private dwellings
  • urban and rural areas in all states and territories, excluding very remote parts of Australia and discrete Aboriginal and Torres Strait Islander Communities.

Geography

The data available includes estimates for Australia.

Source

The National Nutrition and Physical Activity Survey conducted by the Australian Bureau of Statistics.

Collection method

  • Face-to-face interview with an ABS Interviewer
  • 24-hour dietary recall data collected face-to-face with an ABS Interviewer or via an online interview
  • Some physical activity and sleep data was collected on a voluntary basis via an accelerometer.

Concepts, sources and methods

History of changes

Full history of changes

About this survey

Overview

The 2023 National Nutrition and Physical Activity Survey (NNPAS) is a component of the wider Intergenerational Health and Mental Health Study (IHMHS) funded by the Australian Government Department of Health, Disability and Ageing.

The 2023 NNPAS was conducted from January 2023 to March 2024. Data was collected from approximately 8,800 households around Australia.

The survey focused on nutrition, physical activity and sleep, and the roles they play in health and wellbeing. Topics collected as part of this survey included:

  • general health (including selected health conditions and smoking status)
  • dietary intake
  • dietary supplements
  • physical activity
  • sedentary and sleep behaviour.

Self-reported height and weight were also collected. In addition to the self-reported measurements, respondents could voluntarily provide blood pressure, height, weight and waist measurements.

Some topics were included for the first time in the 2023 NNPAS, including questions about gender and sexual orientation, oils and fats used when cooking, and collection of physical activity and sleep data using an accelerometer.

The survey also collected a standard set of information about demographic topics such as age, sex at birth, country of birth, main language, employment, education and income.

The 2023 NNPAS is generally comparable with the 2011-12 NNPAS. Key changes to the survey in 2023 compared with the 2011-12 NNPAS are detailed in the Content changes section. This includes updates to the classification of long-term health conditions (for more details, refer to Health conditions) and to the AUSNUT food and dietary supplement classification.

How the data is collected

Scope

The scope of the survey included:

  • all usual residents in Australia aged 2 years and over living in private dwellings
  • both urban and rural areas in all states and territories, except for very remote parts of Australia and discrete Aboriginal and Torres Strait Islander communities
  • members of the Australian permanent defence forces living in private dwellings and any overseas visitors who have been working or studying in Australia for the last 12 months or more, or intend to do so.

The following people were excluded:

  • visitors to private dwellings
  • overseas visitors who have not been working or studying in Australia for 12 months or more, or do not intend to do so
  • members of non-Australian defence forces stationed in Australia and their dependants
  • non-Australian diplomats, diplomatic staff and members of their households
  • people who usually live in non-private dwellings, such as hotels, motels, hostels, hospitals, nursing homes and short-stay caravan parks (people in long-stay caravan parks, manufactured home estates and marinas are in scope)
  • people in Very Remote areas
  • discrete Aboriginal and Torres Strait Islander communities
  • households where all Usual Residents are less than 18 years of age.

Collection method

Households could complete the first part of the survey, which collected basic demographic information about all usual residents of the household, via an online form or face-to-face interview with an ABS Interviewer. The individual questionnaires were completed via face-to-face interview only.

Dietary intake data was collected using a 24-hour dietary recall instrument called Intake24. Data was collected for two days of data, at least eight days apart. The first day’s data ('Day 1') was collected during the survey interview, while the second day’s data ('Day 2') was collected via online form or as a second face-to-face interview.

If respondents were not capable of answering for themselves, due to illness/injury, cultural reasons, language problems or young age, an appropriate person from the household could be asked the questions on their behalf. This was known as a proxy interview. 

School-aged children aged 5 years and over and adults could volunteer to wear an accelerometer device for seven days. At the completion of the 2023 NNPAS, trained interviewers explained the accelerometer study and provided a device to respondents who agreed to participate. A pre-paid satchel was provided for return of the device at the end of the wear period. Participation was voluntary, and respondents could withdraw at any time. To cover expenses for travel associated with returning the device, respondents were provided a reimbursement of $25, paid via gift card.

Respondents could also provide blood and urine samples which were collected at Sonic Healthcare Australia Pathology collection clinics, their subsidiaries or via a home visit using standard operating procedures for phlebotomy collection. These samples were tested for specific biomarkers for chronic disease and nutrition status. People aged 12 years and over were asked to provide a blood and a urine sample, while children aged 5–11 years were asked to provide a urine sample only. Further information is available in the Health Measures Survey 2022-24 section of the Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

Sample design

Households were randomly selected to participate in the survey. One adult aged 18 years and over and one child aged 2–17 years were randomly selected to complete individual questionnaires.

If the randomly selected child was aged 2–14 years, a parent/guardian answered the questions on the child’s behalf.

If the randomly selected child was aged 15–17 years, parental/guardian consent was sought for the selected child to answer the questions. Where consent was not given, a parent/guardian answered the questions on the selected person’s behalf.

Proxy interviews were accepted for selected adults who were unable to answer for themselves due to language difficulties, significant long term illness or disability.

Response rates

There were 8,817 fully responding households for the Day 1 component of the survey, a response rate of 54.9%. The following tables summarise the response rates achieved for the sample approached across Australia.

Response rates, Australia

Day 1 response rates of initial sample
 Number of households
(no.)
Proportion of households
(%)
Selected households18,313100.0
Sample loss(a)2,25512.3
Selected households after sample loss16,05887.7
  1. Sample loss includes situations such as selected persons away for enumeration period, no adult in household, vacant or derelict dwelling, or dwelling converted to non-dwelling or holiday home.
Day 1 response rates after sample loss
 Number of households
(no.)
Proportion of households
(%)
Selected households after sample loss16,058100.0
     Fully/adequately responding households8,81754.9
     Not adequately responding households  
          Full/part refusal2,25414.0
          Full/part non-contact3,64622.7
          Other(a)1,3418.4
          Total not adequately responding7,24145.1
  1. Includes situations such as death or illness; work, health and safety concerns; instrument issues; and incomplete records.
Day 2 response rates of initial sample
 Number of persons
(no.)
Proportion of persons
(%)
Total fully/adequately responding persons from Day 111,199100.0
Sample loss(a)4303.8
Selected Day 1 persons after sample loss10,76996.2
  1. Includes technical errors and situations where an adult respondent’s Day 1 interview was conducted by proxy with the respondent not present.
Day 2 response rates after sample loss
 Number of persons
(no.)
Proportion of persons
(%)
Selected Day 1 persons after sample loss10,769100.0
     Fully/adequately responding9,03283.9
     Not adequately responding  
          Full/part refusal1681.6
          Full/part non-contact1,51814.1
          Other510.5
          Total not adequately responding1,73716.1

Day of the week of interview and dietary recall

Interviews were conducted on all days of the week; however, the proportion of people who responded varied across each day. Monday was the most common day for Day 1 and the second most common day for Day 2 (Thursday the most common), and Sunday the least common day for both days of interview. 

The first day of 24-hour dietary recall consumption was the day prior to interview (from midnight to midnight). For example, if an interview was conducted on Monday, the dietary recall tool would collect consumption for Sunday. The second recall was done at least 8 days later, and, where possible, on a different day to Day 1. The distribution of each interview day is shown in the table below. 

Proportion of responses for day of the week of interview, by dietary recall day, NNPAS 2023
Day of week of interviewProportion of week (%)
Day 1Day 2
Monday21.517.6
Tuesday14.516.9
Wednesday14.716.2
Thursday15.117.8
Friday12.913.1
Saturday15.610.1
Sunday5.88.3
Total100.0100.0

Many data items are impacted by the day of the week, such as food types collected in the 24-hour dietary recall and sleep. Therefore, broad day of the week of interview (Day 1) was incorporated as a weighting benchmark to adjust for any bias in the data. For further details, refer to Estimation methods in the How the data is processed section.

Accelerometer

Of the 10,687 respondents aged 5 years and over (after sample loss is taken into account), 4,199 were fully/adequately responding in the accelerometer study, a response rate of 39.3%.

Accelerometer response rates of initial sample
 Number of persons
(no.)
Proportion of persons
(%)
Total fully/adequately responding persons from Day 111,199100.0%
Sample loss(a)5124.6%
Accelerometer study opt-in after sample loss4,61041.2%
          Fully/adequately responding(b)4,19937.5%
          Not adequately responding4113.7%
Accelerometer study opt-out6,07754.3%

(a) Includes technical error, children aged 2–5 years and not yet attending school and situations where an adult respondent’s Day 1 interview was conducted by proxy with the respondent not present.

(b) Met minimum wear time of 48 hours or more. Further restrictions were applied to some data. See ‘Minimum wear time’ section.

Content

The survey collected the following content:

  • Demographics – age, sex, gender and sexual orientation, country of birth, main language spoken, marital status
  • Household details – type, size, household composition, Socio-Economic Indexes for Areas (SEIFA), geography
  • Labour force status
  • Educational attainment
  • Household Income
  • Migrant and Visa status
  • Long-term health status relating to diabetes, hypertension, kidney disease
  • Risk factors such as tobacco smoking and physical activity
  • 24-hour dietary recall
  • Specific dietary information such as food avoidance, consumption of oils, fats, salt, and dietary supplements
  • Physical and sedentary activity
  • Sleep behaviours
  • Self-reported height and weight
  • Physical Measures – blood pressure, height, weight and waist.

The 2023 NNPAS uses the Standard for Sex, Gender, Variations of Sex Characteristics and Sexual Orientation Variables, 2020. Data from this survey are typically presented using the Sex at birth variable. When a small number of responses are recorded in any output category, outputs may be suppressed or combined into other categories due to confidentiality and statistical issues. A small number of people in the study reported having a term other than male or female recorded as their sex at birth. Estimates for people whose sex at birth is neither male nor female are not able to be output as a separate category, but they are included in the estimates for total Persons. 

See the Data Item List for full details of content collected in the 2023 NNPAS.

Survey materials

A copy of the questionnaire, prompt cards, measurements card and dietary recall materials are available on request by emailing client.services@abs.gov.au or calling 1300 135 070.

How the data is processed

Estimation methods

As only a sample of people in Australia were surveyed, results needed to be converted into estimates for the whole population. This was done through a process called weighting:

  • Each person or household is given a number (known as a weight) to reflect how many people or households they represent in the whole population.
  • A person or household’s initial weight is based on their probability of being selected in the sample. For example, if the probability of being selected in the survey was one in 45, then the person would have an initial weight of 45 (that is, they would represent 45 people).

The person and household level weights are then calibrated to align with independent estimates of the in-scope population, referred to as ‘benchmarks’. The benchmarks use additional information about the population to ensure that:

  • people or households in the sample represent people or households that are similar to them
  • the survey estimates reflect the distribution of the whole population, not the sample.

Benchmarks align to the estimated resident population (ERP) at June 2023 which was 10,386,915 households and 25,392,728 people aged 2 years and over (after exclusion of people living in non-private dwellings, very remote areas of Australia and discrete Aboriginal and Torres Strait Islander communities).

The day of the week the interview was conducted was also incorporated as a benchmark because people tend to eat differently on weekdays and weekends. Interview day was grouped into weekdays (Tuesday-Friday, reflecting food eaten from Monday-Thursday) and weekends (Saturday-Monday, reflecting food eaten from Friday-Sunday) to best capture typical eating patterns. Labour force status was also considered, as there is a known relationship between day of interview and labour force status.

Survey estimates for the second dietary recall day were not weighted separately and represent a sub-sample of the main survey.  

Sample counts and weighted estimates for both Day 1 and Day 2 interviews are presented in the tables below.

Sample counts and weighted estimates, Australia

Day 1
Age group (years)Persons in sampleWeighted estimate
Males (no.)Females (no.)Persons (no.)Males (‘000)Females (‘000)Persons (‘000)
2-4236220456463.1434.0897.1
5-9367369736843.5791.91,635.4
10-14347344692822.6777.21,600.0
15-19338304644888.8761.21,652.3
20-24201220421796.6796.11,592.8
25-29271305577917.7914.71,833.4
30-34332383715890.6994.51,885.1
35-393844558391,008.7982.81,991.4
40-44358408767849.8889.11,751.6
45-49327362689793.0831.41,624.5
50-54315349664782.5823.21,605.7
55-59294305599751.1764.81,515.9
60-64367420787728.4790.01,518.4
65-69343408751619.0684.01,303.0
70-74311362673535.3557.31,092.6
75-79264306570437.0471.0908.0
80-84161204365291.0321.6612.6
85 years and over97157254148.2224.7372.9
Total all ages5,3135,88111,19912,567.212,809.425,392.7
Day 2
Age group (years)Persons in sampleProportion of Day 1 weighted estimate
Males (no.)Females (no.)Persons (no.)Males (%)Females (%)Persons (%)
2-411911022944.545.344.9
5-929830159982.480.281.3
10-1429126855984.076.580.4
15-1926323850379.479.679.5
20-2414717932675.579.777.6
25-2921724746580.779.480.1
30-3426130656780.183.181.7
35-3930738869580.885.082.8
40-4429834464385.885.685.8
45-4926829756582.985.384.1
50-5425228553781.177.479.2
55-5924426250681.486.684.0
60-6429935565481.484.382.9
65-6929735264985.682.584.0
70-7426630056687.178.182.5
75-7922124146285.876.581.0
80-8413715929686.575.380.6
85 years and over7613521180.089.485.7
Total all ages4,2614,7679,03280.680.480.5

Age standardisation

Age standardisation is a way of allowing comparisons between two or more populations with different age structures to remove age as a factor when examining relationships between variables. Age standardisation can also be used to control for age when examining relationships between variables that are age-dependent within a single population. For example, when assessing the relationship between body mass index (BMI) and sweetened beverage consumption, it can be difficult to ascertain if the relationship is due to age or another underlying factor. This is because age is independently correlated with both BMI and sweetened beverage consumption. By removing age as a confounding factor, age standardisation allows for a clearer understanding of the relationship between the variables.

Proportions quoted in commentary in this publication are not age standardised, however, proportions presented in Table 21 for sweetened beverages include age standardised rates. Data are age standardised to the 2001 Australian population.

Accuracy

Show all

How the data is released

Release strategy

This release presents estimates of nutrient intakes, food groups consumed, dietary behaviours, and food security for 2023. Commentary presents analysis by age groups, sex and selected population characteristics.

Data Cubes (spreadsheets) in this release present tables of estimates, proportions and their associated measures of error. A data item list is also available. Additionally, a concordance between AUSNUT 2011-13 and AUSNUT 2023 is available on the FSANZ website.

Detailed microdata (including nutrition and physical activity) will be available in DataLab later in 2026 for users who want to undertake interactive (real time) complex analysis of microdata in the secure ABS environment.

Future releases include:
•    Australian Dietary Guidelines (ADG) analysis
•    Basic microdata.

To discuss how ABS can cater for your specific needs or data requirements, contact us via the Consultancy Request Form.

Confidentiality

The Census and Statistics Act 1905 authorises the ABS to collect statistical information and requires that information is not published in a way that could identify a particular person or organisation. The ABS must make sure that information about individual respondents cannot be derived from published data.

To minimise the risk of identifying individuals in aggregate statistics, a technique called perturbation is used to randomly adjust cell values. Perturbation involves small random adjustment of the statistics which have a negligible impact on the underlying pattern. This is considered the most satisfactory technique for avoiding the release of identifiable data while maximising the range of information that can be released. After perturbation, a given published cell value will be consistent across all tables. However, adding up cell values in Data Cubes to derive a total may give a slightly different result to the published totals. 

Content changes

The following table summarises the main content changes applied in the 2023 NNPAS compared with the 2011-12 survey. For full details of available data items, refer to the Data Item List.

The 2023 NNPAS is considered to generally be comparable to the 2011-12 NNPAS. However, due to the time between the NNPAS surveys, there have been numerous changes to the content. These changes are mostly due to updates to relevant nutrition and physical activity guidelines, updates to demographic standards (e.g. country of birth, occupation, industry) and the addition of content based on user needs.

2023 NNPAS content changes (changes between 2023 NNPAS and 2011-12 NNPAS)
TopicChanges
Overall
  • Instrument design and layout changes introduced to improve usability.
Age
  • Added additional question for those respondents aged 5 years to determine whether attending school. This information was used for sequencing into various nutrition and physical activity modules to align with guidelines.
Gender and sexual orientation
Visa status
  • New question module and associated outputs.
Country of birth of parents
  • New Māori/Pacific Islander descent question for comparability across other household health surveys.
Education
  • Additional questions, minor question wording and sequencing updates to improve comparability across other household surveys.
Employment
  • Removed workers compensation and shift work questions.
  • Minor question and sequencing updates to improve comparability across household surveys.
Smoking
  • Removed questions about irregular smokers and history of smoking.
24-hour dietary recall
  • Different instrument (Intake24) used to collect 24-hour dietary recall.
  • Day 2 recall completed online or face-to-face, previously by telephone-interview.
  • Updated AUSNUT food and dietary supplement classification – for more information, refer the Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.
  • Minor question and sequencing updates to improve collection of dietary supplement intakes
Dietary behaviour
  • Updated questions and output items to improve capture of reasons for food avoidance.
  • Expanded questions regarding diet or eating pattern followed, including new questions regarding vegan or vegetarian diets.
  • New question to collect the main type of oil or fat used to prepare meals.
  • Amended questions for salt added when cooking or preparing food.
  • Removed questions on main source of tap water and self-reported serves of fruit and vegetables.
Food security
  • More extensive module added to collect information on food security status of household.
  • Population updated to only include proxy interview when selected respondent is present.
Cardiovascular health
  • Minor question and sequencing updates to improve comparability across household surveys.
Life stages
  • Module updated to only collect data relevant to nutritional intake (asking if respondent pregnant or breastfeeding).
  • Changed population to respondents aged 18 to 50 years who reported sex at birth as female or another term. In 2011-12 NNPAS, population was respondents aged 10 years and over who reported sex at birth as female.
Pre-school aged child physical activity and sedentary behaviour
  • Population updated to include children aged 5 years who have not yet started at school to align with physical activity guidelines. Previously only included children aged 2-4 years.
  • Questions on types of devices updated to reflect technology changes. Time used was collected in ranges rather than the specific amount of time.
  • New questions and output items on time secured in car seats or prams.
School-aged child physical activity and sedentary behaviour
  • Population updated to only include children aged 5 years old who have started school, as well as 6-17-year-olds. Previously included all children aged 5-17 years.
  • Major update to sedentary behaviour questions and sequencing to reflect technology changes and simplify questions for respondents.
  • Questions about time spent doing homework were not asked in 2023.
Child sleep behaviour
  • Population increased to include children aged 2-4 years.
  • Questions on typical sleep, homework done before device use, tv off during mealtimes, device use while supervised, or with time restrictions were not collected in 2023.
  • New questions added for use of devices in hour before bed, sleep quality and naps taken.
Adult physical activity
  • Major updates to question module.
  • Strength and toning activities done and method of transport used to get places collected.
  • Other physical activity questions replaced by accelerometer.
Adult sleep
  • Questions on typical sleep replaced by question on sleep quality.
Self-perceived body mass
  • Added questions regarding self-reported height and weight.
  • Removed questions regarding satisfaction with weight.
Physical measurements (height, weight, waist and blood pressure)
  • Due to COVID-19, the procedures for collecting physical measurements have been adapted to account for increased hygiene and social distancing measures, including a move to collection via self-measurements only (rather than via ABS Interviewers) and use of single use waist measurement tape.
  • Imputation of physical measures for respondents with missing values was not completed in the 2011-12 NNPAS but is used in the 2023 NNPAS.
  • Removal of measurements for respondents who self-reported pregnancy.
  • Population for blood pressure updated to only include respondents aged 18 years and over. Previously included respondents aged 5 years and over.
Accelerometer
  • New accelerometer device replaced the pedometer used in 2011-12 NNPAS.
  • Population updated to remove 5-year-olds who are not yet attending school. 
Biomedical opt in
  • Respondents aged 18 years and over were also invited to participate in the Australian Health Biobank (AHB) (as well as the National Health Measures Survey). 
Self-assessed health
  • Not collected in 2023 National Nutrition and Physical Activity Survey.
Adult sedentary behaviours
  • Not collected in 2023 National Nutrition and Physical Activity Survey.

Dietary behaviour

Food avoidance

Respondents were asked if they avoided any foods due to food intolerances, allergies, or for cultural, religious and ethical reasons. If so, they were then asked which foods they avoided, see table below.

Show classification

Diet and eating patterns

The 2023 NNPAS questionnaire asked respondents 15 years of age and older whether they follow any kind of diet or eating pattern, and whether they describe themselves as vegetarian or vegan. Many respondents reported following a diet or eating pattern and the reasons for doing so. Those describing themselves as vegetarian or vegan were considered to be following a diet or eating pattern. 

Show classification

Oils and fats

The main type of oil or fat used to cook dishes containing vegetables, meat, chicken or seafood was asked of respondents 18 years and over. Persons in the household aged 2-17 years were then given the responses of the selected adult in household. Oils and fats used in baking were excluded.

Responses to these questions are used to determine the oil and fat content of home-prepared foods consumed in the 24-hour dietary recall. More information is available in the Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

Salt

The design of the questionnaire has been updated since the 2011-12 NNPAS for collection of salt use in cooking or preparing food. 

In 2023, selected persons aged 18 years and over were asked about salt use when cooking and preparing meals. Where a child was present in the household, the responses of the selected adult were applied to the selected child. In 2011-12, however, all selected persons aged 2 years and over were asked these questions directly, so the responses of the child could differ from the responses of the adult.

In both 2011-12 and 2023 surveys, all selected persons in a household were asked about salt use at the table. 

Information on how salt is estimated in the 24-hour dietary recall is available in the Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

Food security

In this survey, a household’s food security status was based on whether one or more members of the household had enough food, or money to buy the food, needed for an active, healthy life at all times in the last 12 months. This was assessed using a set of 10 questions, known as the Adult Food Security Survey Module, developed by the United States Department of Agriculture (USDA). 

  • The questions were asked of the selected adult in the household on behalf of all members of the household.
  • The specific experiences of children in the household do not form part of this measure.
  • Minor modifications were made to the wording used in some questions to improve their ability to be understood and interpreted in the Australian context.

Households were asked three questions to establish whether there were any indicators of food insecurity. If they answered “yes” to any question, households were asked more questions to establish the severity of food insecurity. Households were given 1 point for each “yes” answer, to give a total score between 0 and 10.

Around 2.0% of households had at least one missing response and required responses to be imputed. Missing values were imputed based on the household’s responses to other questions using the USDA’s direct imputation method. A “yes” answer was imputed only if a household indicated it had experienced all the less severe types of food insecurity and also experienced a more severe type of food insecurity. A small proportion of households that did not answer all the first three questions did not have missing values imputed and were not classified. This equates to around 0.7% of households.

Households were then classified as being food secure or being food insecure (marginal, moderate or severe) based on their total score. Following consultation with Australian food security experts, the ABS adopted Statistics Canada’s approach to classifying households. This differs from the USDA approach by:

  • classifying a score of 2 as moderate food insecurity instead of marginal food insecurity, and
  • classifying marginal food insecurity as food insecure instead of food secure. 

The food security status assigned to a household may not reflect the experience of all individuals within the household.

Health conditions

A long-term health condition was defined as a medical condition (illness, injury or disability) which was current at the time of interview and had lasted, or was expected to last, 6 months or more.

Some reported conditions were assumed to be long-term, including diabetes, rheumatic heart disease, heart attack, angina, heart failure and stroke. Diabetes, rheumatic heart disease, heart attack, angina, heart failure and stroke were also assumed to be current. Respondents could report multiple health conditions.

Any reported health conditions that did not meet this definition were excluded from estimates, e.g. a person may have been told that they had a health condition in the past, but it is no longer current or expected to last 6 months or more.  Conditions that were not considered to be current and long term can be analysed using the data item Condition Status (CONDSTAT) on the survey microdata.

The classification hierarchy is based on the 10th revision of the International Classification of Diseases (ICD). A mini version of the classification used in the 2022 National Health Survey (NHS) was created for the 2023 NNPAS, as respondents were only asked about cardiovascular conditions, diabetes and kidney disease.

See the Data Item List for full details of the conditions classification used in the 2023 NNPAS. 

Self-reported physical activity, sedentary behaviour and sleep

Physical activity, sedentary behaviour and sleep data collected as part of the 2023 NNPAS are assessed against Australia’s Physical Activity and Sedentary Behaviour Guidelines, 2014. For further information about the guidelines by age group, see Australia's Physical Activity and Sedentary Behaviour Guidelines.

Note these guidelines also include sleep recommendations for children.

As was the case in 2011-12 NNPAS, the collection of this data was different depending on the respondent’s age. 

The age groups in 2011-12 were:

  • Pre-school aged children (2-4 years old)
  • School-aged children (5-17 years old)
  • Adults (18+).

The age groups in 2023 were:

  • Pre-school aged children (2-4 years old and 5 years old not attending primary school)
  • School-aged children (6-17 years old and 5 years old attending primary school)
  • Adults (18+).

Note: For respondents who were 5 years of age, a parent/guardian was asked if their child attended school. Children attending school were treated as school-aged children, while those not attending were treated as pre-school aged children. This is in line with Australia's Physical Activity and Sedentary Behaviour Guidelines.

In addition to the questionnaire data, information was directly measured by an accelerometer. See Directly Measured Physical Activity and Sleep.

Self-reported physical activity

Self-reported sedentary behaviour

Self-reported sleep

Directly measured physical activity and sleep (accelerometer)

What is an accelerometer?

Accelerometers are a common type of sensor used to study human movement. They are wearable devices that measure linear acceleration – the change in a person’s speed (velocity) per unit time. The international unit for acceleration is meters per second squared (\(m/s^2\)). Acceleration is often described in relation to gravity, where \(1g = 9.81 m/s^2\)

A key advantage of accelerometers is their ability to detect movement with a high degree of precision on three axes (x, y, z). They also avoid the measurement errors that occur when people are asked to recall their physical activity.

This study used an Axivity AX3 accelerometer. The device contains an ADXL345 tri-axial accelerometer manufactured by Analog Devices. It recorded acceleration at 100Hz with a dynamic range of +/- 8g. This means the device measured acceleration 100 times per second and could capture movement up to eight times gravity in either direction.

The device was worn on the wrist of the dominant hand, where possible. This was done to improve comparability across respondents (by using the same wrist for everyone) and to align with other population accelerometer studies, such as the UK Biobank[1]. The device was able to be worn in the shower, while playing non-contact sports and in bed.

It may have been taken off if:

  • wearing the device was unsafe or uncomfortable (for example, playing contact sports)
  • the respondent exceeded an underwater depth of 1.5m
  • the device caused skin irritation. 

How accelerometer data is processed

Acceleration thresholds are applied to the raw data to classify the level of activity at each moment in the day. A statistical model is then used to determine whether the respondent is asleep (the sleep period) or awake. 

The analysis considers several time periods, including: 

  • midnight to midnight (calendar day)
  • midday to midday (to analyse overnight sleep)
  • wake-up time to wake-up time (different per individual).

The chart below shows adults’ average hours per day by activity type. When interpreting results, it’s important to remember that if someone spends more time in one activity type, less time will be spent in other types of activity.

The accelerometer data was processed with GGIR (version 3.2-6)[2], and R (version 4.4.1)[3]. GGIR is a software package designed to process multi-day raw accelerometer data for physical activity and sleep[4-6]. As the accelerometer moves, the average acceleration (or speed) of the device can be calculated. This is measured in milligravity (mg) units. 

Outputs were later combined with survey data (e.g. age, sex, geography) and survey weights to produce estimates. Population estimates for the accelerometer study were not weighted separately and represent a sub-sample of the main survey.  Analysis with the NNPAS survey weights indicate that while some bias is likely present, it is small and within the margin of error. This bias is mainly caused by different participation rates by age group. The ABS recommends interpreting age-based differences with caution, particularly comparing estimates of children to adults.

Further information about how the analytical software models each activity type is noted in the section below. A full list of the parameters used in GGIR and detail on Concepts, sources and methods will be available in May 2026. Researchers wishing to analyse the data using different parameters may do so in the DataLab.

Directly measured physical activity

Directly measured sleep

Device wear time rates

Respondents were asked to wear the device for seven 24-hour periods (168 hours). However, not all respondents wore the device for the entire duration, for example:

  • 20.6% of respondents wore the device for the full 168 hour wear time
  • 92.9% of respondents wore the device for at least 48 hours. 

Graphs below show the proportion of the sample by age and by the number of hours or number of nights the device was worn. Older age groups had higher adherence to the wear time instructions than younger age groups. For example, comparing different groups by the minimum wear threshold of 48 hours shows:

  • 96.4% persons aged 65 and over met the minimum wear threshold
  • 92.1% of adults aged 18–64 years met the threshold
  • 89.7% of children aged 5–17 years met the threshold.

Number of hours the device was worn, by age group

Number of nights worn, by age group

Wear time rates, by age group

Data were analysed in relation to the amount of time respondents wore the device. The aim was to ensure the estimates produced were as accurate as possible, while still maintaining population representativeness for the Australian population. 

Average daily physical activity estimates were relatively stable from 48 hours of wear time or greater – see Graphs below.

Average minutes of daily inactivity by wear time

Average minutes of daily light physical activity by wear time

Average minutes of daily moderate and vigorous physical activity by wear time

Among participants who had already met the minimum 48 hour wear time requirement, the average amount of sleep changed very little when different sleep‑threshold rules were applied. See table below.

Average length of sleep period, by met minimum wear time

Handling missing data

Missing data were imputed to ensure that each respondent had a complete 7-day (168 hours) record. When data were missing for a particular time point, an average value from the same time point on other valid wear days was imputed. Imputation rates for each age group are noted below.

Imputation rates by age group and weekday/weekend, proportion of hours in wear period
  5–17 years18–64 years65 years and overTotal 5 years and over
Proportion of imputed hours in wear period, weighted (%)
Weekdays20.020.915.019.5
Weekend days23.322.817.921.9
All days20.921.415.820.2

Note: There is a statistically significant difference between weekday and weekend day imputation rates for persons 5 years and over only. 517 years, 1864 and 65 years and over were not statistically different.

This imputation approach supports the production of population-level estimates and reliable weekly averages. However, it may not fully reflect the activity or sleep patterns of individual respondents, particularly at times of day where wear patterns are uneven or biased. 

As noted in the earlier section on wear time rates, the ABS applied wear time criteria to ensure estimates in the NNPAS: Measured physical activity and sleep release remained robust and representative. Researchers working with microdata should take care when interpreting results if they do not apply the same minimum wear time criteria. To support this, flag variables have been included in the survey microdata so users can identify and exclude imputed values if needed.

Comparison of physical activity methods

Comparing device-measured physical activity with self‑reported activity often gives different results. For example, accelerometers record all movement, whereas questionnaires rely on what respondents remember and chose to report.

People may forget everyday movement, like taking the stairs or walking between rooms. Self‑reported methods can also be affected by memory errors or by respondent bias. 

For adults, measured moderate and vigorous physical activity (MVPA) was higher than self-reported estimates in the National Health Survey. Measured MVPA was substantially higher for females compared to males, suggesting females may under‑report their activity more than males. 

For children, self-reported physical activity was collected in NNPAS. Here, self-reported MVPA in children was 22 minutes per day higher than the accelerometer estimate. This may be due to parents reporting on their behalf. It may also be due to accelerometers being taken off for contact sports and during other activities due to discomfort.

Self‑reported length of sleep in NNPAS was generally higher than measured sleep. This difference likely reflects the way each method defines sleep. Questionnaire responses capture the time a person reports being in bed with the lights out. Accelerometer measures only classify sleep once movement falls below a defined threshold. This difference should also be considered when interpreting estimates of sleep efficiency.

Physical measures

In the 2023 NNPAS, voluntary measurements of height, weight and waist circumference were collected from respondents aged two years and over, whilst voluntary blood pressure measurements were also collected from respondents aged 18 years and over. Measurements were not provided by respondents who advised they were pregnant. The collection method is unchanged since the 2011-12 NNPAS. These measurements provide information on body size (using Body Mass Index (BMI)), and the risk of developing chronic disease, and high blood pressure amongst the Australian population.

Body Mass Index (BMI)

BMI is a simple index of weight-for-height that is commonly used to classify underweight, normal weight, overweight and obesity. It is calculated from height and weight information, using the formula weight (kg) divided by the square of height (m):

 \(\normalsize B M I =\frac{w e i g h t[kg]}{h e i g h t ^ 2[m]^2}\)

To produce a measure of the prevalence of underweight, normal weight, overweight or obesity in adults, people aged 18 years and over were classified based on their BMI score as recommended by the World Health Organization’s BMI Classification. The BMI categories for children take into account the age and sex of the child. For a detailed list of the cut-offs see Appendix 4 in the National Health Survey: Users’ Guide, 2017–18 (cat. no. 4363.0).

Non-response rates

Physical measurements have a relatively high rate of non-response due to their voluntary and sensitive nature. To correct for the high rate of non-response, imputation of values for those who did not have measurements collected was used to achieve estimates of physical measurements for the whole population. 

Non-response rates for physical measurements were higher in the 2023 NNPAS than the 2011-12 NNPAS. 

Non-response rates for height and/or weight, by age
 2011-12 NNPAS2023 NNPAS
Height and/or weight%%
Children (2-17 years)18.344.9
Adults (18 years and over)15.735.5

The non-response rate for waist measurements in 2023 was 33.9% for adults and 45.5% for children. The non-response rate for blood pressure measurements (taken for adults only) in 2023 was 31.5%.

The higher non-response rates in 2023 could in part be due to the trend of declining participation in physical measurements. The COVID-19 pandemic may also have had an effect. The procedures for collecting physical measurements in the 2023 NNPAS were adapted to include increased hygiene and social distancing measures, and respondents needed to take their own measurements rather than ABS interviewers taking the measurements.

Self-reported height and weight

In addition to the voluntary measured items, respondents in the 2023 NNPAS were also asked to self-report their height and weight measurements (respondents who advised that they were pregnant were not asked to self-report as they are not applicable to the BMI population for analysis). This provides valuable information about height and weight which can be used in assisting in the imputation for those with missing values.

How imputation works

In the 2023 NNPAS, missing values were imputed using the 'hot decking' imputation method. In this method, a record with a missing response (the 'recipient') receives the response of another similar record (the 'donor'). Several characteristics were used to match recipients. For adults, they were:

  • age group
  • sex
  • part of state (capital city and balance of state)
  • self-perceived body mass (underweight, acceptable, or overweight)
  • whether or not has high cholesterol (as a long-term health condition)
  • self-reported BMI category (calculated from self-reported height and weight).

For example, a female recipient aged 35–39 years who lives in a capital city, has a self-reported BMI category of overweight (calculated using self-reported height and weight), has a self-perceived body mass of healthy and has high cholesterol will match to a donor record who has the same profile (female, 35–39, self-reports as overweight, etc.).

For BMI, 86.2% of imputed records matched to a donor record using all variables. The remaining 13.8% were matched using fewer variables.

For children aged 2-17 years, the following variables were used:

  • single year of age
  • sex
  • part of state
  • self-reported BMI category. 

Self-perceived body mass and cholesterol data were not collected for children aged 2-17 years so could not be used as an imputation variable.

Sex was not used as a variable to match recipients to donors for people with a Transgender and Gender Diverse experience. People with a Transgender and Gender Diverse experience did not act as donors during the imputation process. Imputation was not performed for people who self-reported pregnancy as they are not applicable to the BMI population for analysis. 

Basal Metabolic Rate (BMR)

Basal metabolic rates (BMR) are the amount of energy needed for a minimal set of functions necessary for life over a defined period. BMR is expressed as kilojoules (kJ) per 24-hours and is calculated using age, sex and weight (kg). 

BMR can be used to estimate the amount of energy a person needs to maintain a healthy body weight. The amount of physical activity a person does impacts the amount of energy their body requires. Active people will require larger amounts of energy, whereas people with a more sedentary lifestyle will require less. No adjustment has been made to BMR for activity levels or health status.

The ratio of energy intake to BMR (EI:BMR) is used for determining low and high energy reporters in persons aged 10 years or more. See Under-reporting in the 24-hour dietary recall section.

There are several accepted methods for estimating BMR. The ABS used the Mifflin-St Jeor method to estimate BMR, which replaced the Schofield method (Schofield, 1985) used in the 2011-12 NNPAS. For further information see Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

24-hour dietary recall

Dietary intake information was collected using a 24-hour dietary recall tool called Intake24 which was adapted for use in Australia in conjunction with Monash University and Food Standards Australia New Zealand. Intake24 was developed, and validated using a doubly labelled water study, by the University of Newcastle in the UK[11]. The tool captured information on the food and beverage intakes for all respondents on the day prior to interview, from midnight to midnight. Where possible, at least 8 days after the first interview, respondents were followed up to provide a second 24-hour dietary recall either via a web form or a second face-to-face interview. 

The purpose of the 24-hour dietary data collection is to estimate total amounts of food, beverages and dietary supplements consumed by the Australian population to assess dietary behaviours and the relationship between diet and health. Data were merged with the AUSNUT 2023 files to produce estimates of energy, macronutrient, vitamin and mineral intakes, and Australian Dietary Guideline (ADG) food group serves.

The Food and nutrient release provides an estimate of energy and nutrient intake based on the Day 1 recall only. The Usual Nutrient Intakes release provides an estimate of usual nutrient intake based on Day 1 and Day 2 recall data. The usual intakes of Australian Dietary Guidelines (ADG) food groups will be available in a subsequent release.

Further information on the 24-hour dietary recall methodology is available in the Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

Dietary recall data quality

Respondent error

Processing error

Under-reporting

Usual intakes

Usual intakes are an estimate of what people ‘usually’ eat, as opposed to what they reported eating on a single day. For example, someone might eat fewer serves of vegetables on some days of the week compared to others. Usual intakes account for the day-to-day variation (within-person) to create a long-term average. Usual intakes can be estimated for both food types and nutrients consumed. 

Usual food and nutrient intakes are compared to the Australian Dietary Guidelines (ADG) and Nutrient Reference Values (NRV) recommendations. The ADGs provide advice about the amount and kind of foods to eat. The NRVs provide advice about the amount of essential nutrients (e.g. protein, calcium, vitamin D) to consume. 

The recommendations are designed to present public health advice to the general population. People who need nuanced advice on food or nutrient intakes (due to their specific circumstances) may need to consult a health practitioner who can consider their individual needs.

As the ABS reports on different age groups to the published NRV recommendations, adjustments were made during processing. For NRVs reported in this publication, see IHMHS: Concepts, Sources and Methods – Food and nutrient collections.

Sample

This survey was based on two separate 24-hour dietary recalls. This is because a single 24-hour dietary recall shows what someone ate on one day, but it doesn’t capture how intake changes from day-to-day. A second recall helps measure this variation.

The final sample for usual intakes was 97.6% of all Day 1 respondents (n = 10,929). Of the usual intakes sample, 80.7% also completed a Day 2 dietary recall (n = 8,814).

Some populations are excluded from the analysis due to low sample counts and availability of comparable recommendations. Detail on this is highlighted in the sections below.

Persons who were excluded based on sex or gender characteristics

Persons who were excluded based on pregnant or breastfeeding status

How the usual intake data is processed

The National Cancer Institute (NCI) method was used to estimate these usual intakes. The NCI method uses the combined intake records for a group to estimate the group’s usual intake distribution[19]. Day 2 records for the group are used to estimate and remove within-person (or day-to-day) variation from the group's Day 1 intakes, which narrows the tails of the distribution. Because the method relies on pooled group data for its calculations, it does not produce usual intake estimates for individual respondents. 

The IHMHS: Concepts, Sources and Methods details how the ABS uses the NCI method, including:

  • model types and covariates applied,
  • adjustments to Nutrient Reference Values (NRVs), and
  • principles for comparing against the guidelines.

This method has been validated on other nutrition survey data[19]. It was also used by the ABS for the 2011–12 NNPAS. 

Dietary supplements

Interviewers recorded the Australian Register of Therapeutic Goods Administration (TGA) identification number of each dietary supplement taken by the respondent in 24 hours prior to interview. The AUST L numbers were assigned to listed medicines including vitamins, minerals, and herbal and homoeopathic products. These included supplements containing ingredients that are nutrients, such as multivitamin or fish oil products, and exclude products intended for inhalation or use on the skin.

For dietary supplements without an AUST L code, interviewers recorded details of the supplement which were coded in processing. Up to 15 different supplements were recorded per person. Data were merged with the AUSNUT 2023 files to produce estimates of macronutrient, vitamin and mineral intakes for each supplement.

Further information on dietary supplement collection and processing can be found in the Intergenerational Health and Mental Health Study: Concept, Sources and Methods.

Dietary recall food and supplement classification

Food and dietary supplement consumption patterns can be described using several approaches. These provide different types of information, the use of which will depend on the purpose. In this study, data on consumption patterns is presented by the:

  • AUSNUT 2023 major, sub-major, and minor food groups
  • Discretionary food flag
  • Australian Dietary Guidelines food groups 

For more information about these classifications and how they are used, see Intergenerational Health and Mental Health Study: Concepts, Sources and Methods.

Data in the Food and Nutrients release are reported by the AUSNUT food groups and discretionary food flag. Examples of the foods and dietary supplements included in the AUSNUT major food groups can be found below. 

Example of AUSNUT major food groups

Example of Australian Dietary Guidelines major food groups

Footnotes

  1. Doherty, A., Jackson, D., Hammerla, N., Plötz, T., Olivier, P., Granat, M. H., et al. 2017. Large scale population assessment of physical activity using wrist worn accelerometers: The UK Biobank study. PLoS ONE, 12(2). https://doi.org/10.1371/journal.pone.0169649
  2. van Hees, V., Migueles, J., Fang, Z., Zhao, J., Heywood, J., Mirkes, E., Sabia, S. 2025. GGIR: Raw Accelerometer Data Analysis. R package version 3.2-6. https://doi.org/10.5281/zenodo.1051064   https://CRAN.R-project.org/package=GGIR
  3. R Core Team. 2024. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/
  4. Migueles, J. H., Rowlands, A. V., Huber, F., Sabia, S., van Hees, V. T. 2019. GGIR: A research community–driven open source R package for generating physical activity and sleep outcomes from multi-day raw accelerometer data. Journal for the Measurement of Physical Behaviour, 2(3), 188–196. https://doi.org/10.1123/jmpb.2018-0063
  5. van Hees, V. , Fang, Z., Langford, J., Assah, F., Mohammad, A., da Silva, I., Trenell, M., White, T., Wareham, N., Brage, S. 2014. Autocalibration of accelerometer data for free-living physical activity assessment using local gravity and temperature: An evaluation on four continents. Journal of Applied Physiology, 117(7), 738–744. https://doi.org/10.1152/japplphysiol.00421.2014
  6. van Hees, V., Sabia, S., Anderson, K., Denton, S., Oliver, J., Catt, M., Abell, J., Kivimäki, M., Trenell, M., Singh-Manoux, A. 2015. A novel, open access method to assess sleep duration using a wrist-worn accelerometer. PLoS ONE, 10(11). https://doi.org/10.1371/journal.pone.0142533
  7. Hildebrand, M., van Hees, V., Hansen, B., Ekelund, U. 2014. Age group comparability of raw accelerometer output from wrist- and hip-worn monitors. Medicine & Science in Sports & Exercise, 46(9), 1816–1824. https://doi.org/10.1249/mss.0000000000000289
  8. Hildebrand, M., Hansen, B., van Hees, V., Ekelund, U. 2016. Evaluation of raw acceleration sedentary thresholds in children and adults. Scandinavian Journal of Medicine & Science in Sports, 27(12), 1814–1823. https://doi.org/10.1111/sms.12795
  9. Patterson, M. R. 2020. Development of an algorithm to count steps from 24hr wrist accelerometry data. Verisense-Toolbox, ShimmerEngineering. https://github.com/ShimmerEngineering/Verisense-Toolbox/tree/master/Verisense_step_algorithm
  10. van Hees, V., Sabia, S., Jones, S., Wood, A., Anderson, K., Kivimäki, M., Frayling, T., Pack, A., Bucan, M., Trenell, M., Mazzotti, D., Gehrman, P., Singh-Manoux, B., Weedon, M. 2018. Estimating sleep parameters using an accelerometer without sleep diary. Scientific Reports, 8(1). https://doi.org/10.1038/s41598-018-31266-z
  11. Foster, E., Lee, C., Imamura, F., Hollidge, S. E., Westgate, K. L., Venables, M. C., … Brage, S. 2019. Validity and reliability of an online self-report 24-hr dietary recall method (Intake24): a doubly labelled water study and repeated-measures analysis. Journal of Nutritional Science, 8, e29. doi:10.1017/jns.2019.20 https://pmc.ncbi.nlm.nih.gov/articles/PMC6722486/
  12. Rumpler, W., Kramer, M., Rhodes, D. Moshfegh, A. J., and Paul, D.R. 2008. Identifying sources of reporting error using measured food intake. Eur J Clin Nutr 62, 544–552. https://doi.org/10.1038/sj.ejcn.1602742
  13. Brassard, D., Laramée, C., Robitaille, J., Lemieux, S., & Lamarche, B. (2020). Differences in population-based dietary intake estimates obtained from an interviewer-administered and a self-administered web-based 24-h recall. Frontiers in Nutrition, 7, 137. https://doi.org/10.3389/fnut.2020.00137
  14. Macdiarmid, J., and Blundell, J. 1998. Assessing dietary intake: Who, what and why of under-reporting, Nutrition Research Reviews, 11, pp 231-253. doi:10.1079/NRR19980017. http://www.ncbi.nlm.nih.gov/pubmed/19094249
  15. Goldberg, G. R., Black, A. E., Jebb, S. A., Cole, T. J., Murgatroyd, P. R., Coward, W. A., and Prentice, A. M. 1991. Critical evaluation of energy intake data using fundamental principles of energy physiology: 1. Derivation of cut-off limits to identify under-recording. European Journal of Clinical Nutrition, 45(12), 569–581. https://pubmed.ncbi.nlm.nih.gov/1810719/
  16. Black, A.E. 2000. The sensitivity and specificity of the Goldberg cut-off for EI:BMR for identifying diet reports of poor validity. European Journal of Clinical Nutrition, 54, 395-404. https://pubmed.ncbi.nlm.nih.gov/10822286/
  17. Gibson, R.S. 2005. Chapter 5: 'Measurement errors in dietary assessment', Principles of Nutritional Assessment Second Edition, Oxford University Press, p.168.
  18. National Health and Medical Research Council 2025. 'Public Consultation on 2025 Methodological Framework for the Review of Nutrient Reference Values', last accessed 12 March 2026.
  19. National Cancer Institute 2013. ‘Usual Dietary Intakes: The NCI Method’, last accessed 05 February 2026.

Glossary

Show all

Back to top of the page