Pöördumine

WHO methods for estimates of physical activity in adolescents

aged 11-17 years and over 2010-2025

Background document

Department of Noncommunicable Diseases and Mental Health

November 2025

2

CONTENTS

1. Introduction ........................................................................................................................................ 3

2. Data identification, access and inclusion .......................................................................................... 3

2.1 Data inclusion criteria .................................................................................................................. 4

2.2 Data identification ........................................................................................................................ 5

2.3 Data access and processing .......................................................................................................... 5

3. Bayesian hierarchical ordinal regression .......................................................................................... 6

References .............................................................................................................................................. 7

3

1. Introduction

The World Health Organization (WHO) is currently updating country, regional and global estimates of

physical activity in adolescents aged 11-17 years.

WHO’s recommendation on adolescent physical activity changed in 2020 (1). Prior to 2020, the

recommendation was for adolescents to undertake at least 60 minutes of moderate-to-vigorous physical

activity (MVPA) on every day of the week. The 2020 update stated that adolescents should undertake an

average of 60 minutes of MVPA per day (1). Globally, most survey data on adolescent physical activity ask

adolescents to report the number of days in the past week that they undertook at least 60 minutes of MVPA.

WHO has previously used these data to estimate trends in the percentage of adolescents not meeting the

previous guideline, i.e., reporting fewer than 7 days of 60 minutes of MVPA per week (2). The available

surveillance data cannot be used to directly assess whether adolescents meet the new recommendation of

an average of 60 minutes per day. However, these data can be used to describe the distribution of activity

across days and to assess trends over time. In its current update, WHO has estimated trends in the number

of days (0 to 7) of at least 60 minutes of MVPA per week, by country. These estimates will serve to provide

an update on adolescent activity levels and trends until surveillance data that can monitor whether

adolescents meet WHO’s current guidelines are widely collected. These new estimates will be published in

WHO information products, such as the WHO NCD Portal (ncdportal.org).

WHO has estimated trends between 2010 and 2025 of the distribution of the number of days per week on

which adolescents aged 11-17 report doing at least 60 minutes of MVPA in 195 countries and territories. The

analysis included two steps:

1) Identifying data sources; accessing and extracting data; and systematically assessing data against

inclusion criteria;

2) Applying a statistical model to estimate the distribution of the number of days per week of at least

60 minutes of MVPA and its uncertainty by sex, age, country, and year for adolescents aged 11-17

years.

In this analysis, levels of physical activity are based on self-report data assessing physical activity undertaken

at school, for leisure, or for transportation. Estimates were made for all countries, including those for which

no data were included.

2. Data identification, access and inclusion

Our data search and access strategy was designed to obtain as many sources as possible while ensuring that

the sources were representative of the population at the national level or at least covered three areas within

4

the country. We accessed data in two forms: 1) anonymised individual-level self-report data when available,

and 2) summary statistics, namely percentage of adolescents by days of activity reported, age in years, and

sex.

2.1 Data inclusion criteria

All data sources were population-based surveys of adolescents. We included data sources if:

 A question on the number of days per week of at least 60 minutes of MVPA was asked. Acceptable

response options include: 0 days, 1 day, . . ., 7 days; or tick boxes for Monday through Sunday (see

example in Box). Recall periods of past 7 days or typical/usual week were included;

 Total MVPA must include activity at school, active travel, and active leisure/play. These domains may

be explicit or implicit in the question preamble or examples given (see example in Box);

 Data were collected from the general or school-going adolescent populations through probabilistic

sampling using a defined sampling frame, and are representative of a national or defined

subnational population covering at least three regions within a country;1

 data were collected from the general population aged 11-17 years, or at least 3 years of this age

group (e.g., 13-15-year-olds);

 statistics by single day of reported activity (0 days, 1 day . . . 7 days), by single year of age and by sex

were reported or could be calculated from individual record data;

 data were collected in or after 2008;

1 Data sources covering one or two areas that includes at least 75% of the country’s population or a UK home nation (e.g., Scotland) were also included.

Box. Example of single-item question used in the Global school-based student health survey (GSHS). Physical activity is any activity that increases your heart rate and makes you breathe hard. Physical activity can be done in sports, playing with friends, walking to school, or in physical education class. Some examples of physical activity are running, biking, dancing, and football.

1. During the past 7 days, on how many days were you physically active for a total of at least 60 minutes per day? ADD UP ALL THE TIME YOU SPENT IN ANY KIND OF PHYSICAL ACTIVITY EACH DAY.

A. 0 days B. 1 day C. 2 days D. 3 days E. 4 days F. 5 days G. 6 days H. 7 days

5

 data were from a WHO member state or one of three included territories; and

 total survey sample size was at least n=100.

We excluded data sources with the following characteristics:

 the average daily minutes of MVPA were collected (this includes questionnaires designed for adults

such as IPAQ or GPAQ or domain-specific questions that must be aggregated to obtain total MVPA),

 activity was measured using a device such as an accelerometer, or

 data sources that excluded migrants when collected in countries where migrants comprised more

than 40% of the youth population (Kuwait, Qatar, United Arab Emirates).

Our last data update was in November, 2025.

2.2 Data identification

We used several strategies to identify data for inclusion, including:

 data from previous WHO estimates (2);

 data from the Global school-based student health surveys (GSHS);

 data from the Health Behaviour in School-Aged Children (HBSC) study;

 surveys provided in the 2023 WHO Country Capacity Survey (3) in response to a query about the

latest nationally representative source of adolescent physical activity data;

 surveys cited in the Active Healthy Kids Report Cards (4);

 surveys identified by a targeted online Google search using each country name, terms for physical

activity and national survey (translated as appropriate) checking the first page of hits; and

 other data shared with WHO, including during previous consultations with countries.

We aim to include data sources shared by countries during the present country consultation, provided they

meet the inclusion criteria listed in section 2.1.

2.3 Data access and processing

Once a survey was identified for inclusion, we obtained individual-level data or summary statistics. For each

data source accessed as individual-level data, we computed the percentage of students reporting 0-7 days of

at least 60 minutes of MVPA, by number of days reported, single year of age and sex, applying sample

weights if available. School-based surveys may include adolescents outside of the intended age range

because sampling occurs at the class level, and classes include adolescents of different ages. We included

data reported by students within the intended age range. In addition, we included data on younger or older

students if the number of students included was at least 50% the average sample size of the ages included

by design. We applied two additional quality criteria to data accessed at the individual-level: we excluded

6

the data if age or sex were missing for more than 20% of observations, and if physical activity data were

missing for more than 20% of observations. Finally, we combined data accessed as summary statistics and

data accessed at the individual level into a single database.

3. Bayesian hierarchical ordinal regression

Our aim was to estimate the distribution of the number of days per week on which adolescents report doing

at least 60 minutes of MVPA for every country, year (2010-2025), and single year of age (11-17 years), and

for males and females. All analyses were done separately for males and females.

We used Bayesian hierarchical ordinal regressions, which use all available data to make estimates for each

country-year-age-day unit. The ordinal regression, implemented as a cumulative logit (ordered logistic)

model (in brms notation, we used the cumulative family with flexible thresholds (5)), estimates the

distribution of days adolescents report doing at least 60 minutes of activity; the overall shape of the

distribution is fitted separately for each analysis region. In addition, the percentage of adolescents reporting

0 or 7 days is fitted separately using country random-effects, allowing for country-specific inflation/deflation

of these key indicators. The mean number of days per week (the predictor in the ordinal regression) was

specified as a function of country, region, age, and year. Country estimates were nested hierarchically within

analysis regions, meaning that estimates for each country were informed by data from that country, if

available, and by data from other countries, especially those in the same region with data in similar time

periods. The hierarchical model shares information to a greater degree where data are nonexistent or

weakly informative (i.e., have large uncertainty), and a lesser degree in data-rich countries and regions.

Trends over time were also modelled hierarchically for each country, nested within its analysis region. Age

patterns were modelled as a linear function that varied by region. In addition, we included global age and

year trends. The model included a variance term that accounted for unobserved design factors (sample

design, season, etc.) that lead to additional variability in the data beyond that expected due to sample size.

We fit the models (one for males and one for females) using custom stan code (6) based on the brms

package in R (5).

7

References

1. Bull FC, Al-Ansari SS, Biddle S, Borodulin K, Buman MP, Cardon G, et al. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br J Sports Med. 2020 Dec;54(24):1451–62.

2. Guthold R, Stevens GA, Riley LM, Bull FC. Global trends in insufficient physical activity among adolescents: a pooled analysis of 298 population-based surveys with 1·6 million participants. Lancet Child Adolesc Health. 2020 Jan 1;4(1):23–35.

3. World Health Organization. Assessing national capacity for the prevention and control of noncommunicable diseases: report on 2023 global survey [Internet]. Geneva, Switzerland: World Health Organization; 2025. Available from: https://www.who.int/publications/i/item/9789240105997

4. Aubert S, Barnes JD, Demchenko I, Hawthorne M, Abdeta C, Abi Nader P, et al. Global Matrix 4.0 Physical Activity Report Card Grades for Children and Adolescents: Results and Analyses From 57 Countries. J Phys Act Health. 2022 Nov 1;19(11):700–28.

5. Bürkner PC. brms: An R Package for Bayesian Multilevel Models Using Stan | Journal of Statistical Software. J Stat Softw [Internet]. 2017 Aug 29 [cited 2025 Nov 20];80(1). Available from: https://www.jstatsoft.org/article/view/v080i01

6. Stan development team. RStan: the R interface to Stan [Internet]. 2025. Available from: https://mc- stan.org/

WHO methods for estimates of physical activity in adolescents

aged 11-17 years and over 2010-2025

Background document

Department of Noncommunicable Diseases and Mental Health

November 2025

2

CONTENTS

1. Introduction ........................................................................................................................................ 3

2. Data identification, access and inclusion .......................................................................................... 3

2.1 Data inclusion criteria .................................................................................................................. 4

2.2 Data identification ........................................................................................................................ 5

2.3 Data access and processing .......................................................................................................... 5

3. Bayesian hierarchical ordinal regression .......................................................................................... 6

References .............................................................................................................................................. 7

3

1. Introduction

The World Health Organization (WHO) is currently updating country, regional and global estimates of

physical activity in adolescents aged 11-17 years.

WHO’s recommendation on adolescent physical activity changed in 2020 (1). Prior to 2020, the

recommendation was for adolescents to undertake at least 60 minutes of moderate-to-vigorous physical

activity (MVPA) on every day of the week. The 2020 update stated that adolescents should undertake an

average of 60 minutes of MVPA per day (1). Globally, most survey data on adolescent physical activity ask

adolescents to report the number of days in the past week that they undertook at least 60 minutes of MVPA.

WHO has previously used these data to estimate trends in the percentage of adolescents not meeting the

previous guideline, i.e., reporting fewer than 7 days of 60 minutes of MVPA per week (2). The available

surveillance data cannot be used to directly assess whether adolescents meet the new recommendation of

an average of 60 minutes per day. However, these data can be used to describe the distribution of activity

across days and to assess trends over time. In its current update, WHO has estimated trends in the number

of days (0 to 7) of at least 60 minutes of MVPA per week, by country. These estimates will serve to provide

an update on adolescent activity levels and trends until surveillance data that can monitor whether

adolescents meet WHO’s current guidelines are widely collected. These new estimates will be published in

WHO information products, such as the WHO NCD Portal (ncdportal.org).

WHO has estimated trends between 2010 and 2025 of the distribution of the number of days per week on

which adolescents aged 11-17 report doing at least 60 minutes of MVPA in 195 countries and territories. The

analysis included two steps:

1) Identifying data sources; accessing and extracting data; and systematically assessing data against

inclusion criteria;

2) Applying a statistical model to estimate the distribution of the number of days per week of at least

60 minutes of MVPA and its uncertainty by sex, age, country, and year for adolescents aged 11-17

years.

In this analysis, levels of physical activity are based on self-report data assessing physical activity undertaken

at school, for leisure, or for transportation. Estimates were made for all countries, including those for which

no data were included.

2. Data identification, access and inclusion

Our data search and access strategy was designed to obtain as many sources as possible while ensuring that

the sources were representative of the population at the national level or at least covered three areas within

4

the country. We accessed data in two forms: 1) anonymised individual-level self-report data when available,

and 2) summary statistics, namely percentage of adolescents by days of activity reported, age in years, and

sex.

2.1 Data inclusion criteria

All data sources were population-based surveys of adolescents. We included data sources if:

 A question on the number of days per week of at least 60 minutes of MVPA was asked. Acceptable

response options include: 0 days, 1 day, . . ., 7 days; or tick boxes for Monday through Sunday (see

example in Box). Recall periods of past 7 days or typical/usual week were included;

 Total MVPA must include activity at school, active travel, and active leisure/play. These domains may

be explicit or implicit in the question preamble or examples given (see example in Box);

 Data were collected from the general or school-going adolescent populations through probabilistic

sampling using a defined sampling frame, and are representative of a national or defined

subnational population covering at least three regions within a country;1

 data were collected from the general population aged 11-17 years, or at least 3 years of this age

group (e.g., 13-15-year-olds);

 statistics by single day of reported activity (0 days, 1 day . . . 7 days), by single year of age and by sex

were reported or could be calculated from individual record data;

 data were collected in or after 2008;

1 Data sources covering one or two areas that includes at least 75% of the country’s population or a UK home nation (e.g., Scotland) were also included.

Box. Example of single-item question used in the Global school-based student health survey (GSHS). Physical activity is any activity that increases your heart rate and makes you breathe hard. Physical activity can be done in sports, playing with friends, walking to school, or in physical education class. Some examples of physical activity are running, biking, dancing, and football.

1. During the past 7 days, on how many days were you physically active for a total of at least 60 minutes per day? ADD UP ALL THE TIME YOU SPENT IN ANY KIND OF PHYSICAL ACTIVITY EACH DAY.

A. 0 days B. 1 day C. 2 days D. 3 days E. 4 days F. 5 days G. 6 days H. 7 days

5

 data were from a WHO member state or one of three included territories; and

 total survey sample size was at least n=100.

We excluded data sources with the following characteristics:

 the average daily minutes of MVPA were collected (this includes questionnaires designed for adults

such as IPAQ or GPAQ or domain-specific questions that must be aggregated to obtain total MVPA),

 activity was measured using a device such as an accelerometer, or

 data sources that excluded migrants when collected in countries where migrants comprised more

than 40% of the youth population (Kuwait, Qatar, United Arab Emirates).

Our last data update was in November, 2025.

2.2 Data identification

We used several strategies to identify data for inclusion, including:

 data from previous WHO estimates (2);

 data from the Global school-based student health surveys (GSHS);

 data from the Health Behaviour in School-Aged Children (HBSC) study;

 surveys provided in the 2023 WHO Country Capacity Survey (3) in response to a query about the

latest nationally representative source of adolescent physical activity data;

 surveys cited in the Active Healthy Kids Report Cards (4);

 surveys identified by a targeted online Google search using each country name, terms for physical

activity and national survey (translated as appropriate) checking the first page of hits; and

 other data shared with WHO, including during previous consultations with countries.

We aim to include data sources shared by countries during the present country consultation, provided they

meet the inclusion criteria listed in section 2.1.

2.3 Data access and processing

Once a survey was identified for inclusion, we obtained individual-level data or summary statistics. For each

data source accessed as individual-level data, we computed the percentage of students reporting 0-7 days of

at least 60 minutes of MVPA, by number of days reported, single year of age and sex, applying sample

weights if available. School-based surveys may include adolescents outside of the intended age range

because sampling occurs at the class level, and classes include adolescents of different ages. We included

data reported by students within the intended age range. In addition, we included data on younger or older

students if the number of students included was at least 50% the average sample size of the ages included

by design. We applied two additional quality criteria to data accessed at the individual-level: we excluded

6

the data if age or sex were missing for more than 20% of observations, and if physical activity data were

missing for more than 20% of observations. Finally, we combined data accessed as summary statistics and

data accessed at the individual level into a single database.

3. Bayesian hierarchical ordinal regression

Our aim was to estimate the distribution of the number of days per week on which adolescents report doing

at least 60 minutes of MVPA for every country, year (2010-2025), and single year of age (11-17 years), and

for males and females. All analyses were done separately for males and females.

We used Bayesian hierarchical ordinal regressions, which use all available data to make estimates for each

country-year-age-day unit. The ordinal regression, implemented as a cumulative logit (ordered logistic)

model (in brms notation, we used the cumulative family with flexible thresholds (5)), estimates the

distribution of days adolescents report doing at least 60 minutes of activity; the overall shape of the

distribution is fitted separately for each analysis region. In addition, the percentage of adolescents reporting

0 or 7 days is fitted separately using country random-effects, allowing for country-specific inflation/deflation

of these key indicators. The mean number of days per week (the predictor in the ordinal regression) was

specified as a function of country, region, age, and year. Country estimates were nested hierarchically within

analysis regions, meaning that estimates for each country were informed by data from that country, if

available, and by data from other countries, especially those in the same region with data in similar time

periods. The hierarchical model shares information to a greater degree where data are nonexistent or

weakly informative (i.e., have large uncertainty), and a lesser degree in data-rich countries and regions.

Trends over time were also modelled hierarchically for each country, nested within its analysis region. Age

patterns were modelled as a linear function that varied by region. In addition, we included global age and

year trends. The model included a variance term that accounted for unobserved design factors (sample

design, season, etc.) that lead to additional variability in the data beyond that expected due to sample size.

We fit the models (one for males and one for females) using custom stan code (6) based on the brms

package in R (5).

7

References

1. Bull FC, Al-Ansari SS, Biddle S, Borodulin K, Buman MP, Cardon G, et al. World Health Organization 2020 guidelines on physical activity and sedentary behaviour. Br J Sports Med. 2020 Dec;54(24):1451–62.

2. Guthold R, Stevens GA, Riley LM, Bull FC. Global trends in insufficient physical activity among adolescents: a pooled analysis of 298 population-based surveys with 1·6 million participants. Lancet Child Adolesc Health. 2020 Jan 1;4(1):23–35.

3. World Health Organization. Assessing national capacity for the prevention and control of noncommunicable diseases: report on 2023 global survey [Internet]. Geneva, Switzerland: World Health Organization; 2025. Available from: https://www.who.int/publications/i/item/9789240105997

4. Aubert S, Barnes JD, Demchenko I, Hawthorne M, Abdeta C, Abi Nader P, et al. Global Matrix 4.0 Physical Activity Report Card Grades for Children and Adolescents: Results and Analyses From 57 Countries. J Phys Act Health. 2022 Nov 1;19(11):700–28.

5. Bürkner PC. brms: An R Package for Bayesian Multilevel Models Using Stan | Journal of Statistical Software. J Stat Softw [Internet]. 2017 Aug 29 [cited 2025 Nov 20];80(1). Available from: https://www.jstatsoft.org/article/view/v080i01

6. Stan development team. RStan: the R interface to Stan [Internet]. 2025. Available from: https://mc- stan.org/