Childhood Obesity Prevalence: What the Survey Data Shows, Country by Country

A parent sits in a clinic room holding a printout that says their nine-year-old is in the 96th BMI percentile. The number is exact. What it means for that specific child is not. This is a sourced digest of childhood obesity prevalence data by country: the measured rates from named surveys and the reference standards that produce them. Every figure below names the survey, the country, the year, and the issuing body. Every figure states whether it uses the CDC 2000 growth charts 95th percentile, the WHO growth standards, or the IOTF thresholds, because a child classified as obese under one reference may fall into a different category under another. Prevalence is a snapshot of a population at a point in time. It is not a prediction for an individual child. Where the data come from self-report rather than measured height and weight, the prevalence figure will read lower than the true rate, and that gap is stated.

Growth Standards For Children Under Five

The World Health Organization publishes two growth references. For children aged 0-5 years, the WHO child growth standards from the 2006 Multicentre Growth Reference Study are prescriptive: they describe how children should grow under optimal conditions, based on 8,440 breastfed infants from Brazil, Ghana, India, Norway, Oman, and the USA. The obesity threshold is a BMI-for-age more than two standard deviations above the median (>+2 SD). Severe obesity sits at >+3 SD.

The 5-19 Years Reference And Its Limits

For ages 5-19, the WHO Reference 2007, a reconstruction of the 1977 NCHS/WHO reference, uses the same +2 SD threshold. These are the figures most often cited in global reports such as the Global Burden of Disease study and they are the standard for cross-national comparisons. A 2023 WHO report on European childhood obesity used the +2 SD cut-off and reported that roughly one in four European adolescents were overweight including obesity, with secular trends showing no decline in the last decade. The WHO data come from cross-sectional survey submissions by member states. Quality depends entirely on whether the country used measured or self-reported height and weight. Countries that rely on self-report understate prevalence by an estimated 0.4-1.0 kg/m² on average in young people.

The United States uses the CDC 2000 growth charts. These are descriptive, not prescriptive: they show how US children actually grew between 1963 and 1994, drawn from NHANES survey data. Obesity is defined as at or above the 95th percentile of the BMI-for-age chart for ages 2-20. Severe obesity means a BMI at or above 120% of the 95th percentile, or at or above 35 kg/m², whichever is lower. The most recent NHANES data (2017-2020, published 2021) put US obesity prevalence at 19.7% for ages 2-19, with the highest rates in adolescents aged 12-19 (22.2%) and in Black and Hispanic populations. The CDC updates these figures through continuous NHANES cycles. Any number over a year old is a historical snapshot. A key failure mode: the CDC 95th percentile does not align with the WHO +2 SD threshold. At age 10, the WHO obesity cut-off is slightly lower than the CDC threshold for boys. A boy classified as obese under WHO standards may register only as overweight under CDC rules. Always check which growth chart the survey used before comparing rates across countries.

England: The NCMP Annual Census

The secular trend in childhood obesity, the change in prevalence across decades, is best documented in high-income countries with repeated cross-sectional surveys. In England, the National Child Measurement Programme has measured every pupil in Reception (age 4-5) and Year 6 (age 10-11) annually since 2006. The 2023/24 data, published by NHS England, show obesity prevalence of 9.6% in Reception and 22.1% in Year 6. The overweight including obesity figure is 22.1% in Reception and 36.6% in Year 6. These are measured heights and weights, not self-reported, and the sample frame is the entire state-educated population, giving narrow confidence intervals. The trend since 2006 is a slow upward drift, with a sharp spike during the COVID-19 pandemic that has not fully reversed.

United States And Canada: Diverging Paths

In the United States, the NHANES surveys allow a longer view. Obesity prevalence in ages 2-19 rose from 5.2% in 1971-1974 to 19.7% in 2017-2020, with the steepest increases in the 1990s and early 2000s. A Canadian study using the Canadian Health Measures Survey showed a stable prevalence from 2004 to 2015 at about 13% using the WHO thresholds, suggesting the secular increase is not universal across all high-income countries. The key caveat: comparing trends requires using the same cut-off and the same measurement method across every survey point. A shift from self-reported to measured data within a country will create an artificial jump in prevalence. That jump is a methodological artefact, not a real change.

England And Australia

England: The NCMP is the gold standard for national surveillance. Reception (age 4-5) obesity 9.6%, overweight including obesity 22.1% (2023/24, NHS England). Year 6 (age 10-11) obesity 22.1%, overweight including obesity 36.6%. Measured height and weight. Uses UK90 growth charts for the overweight and obesity thresholds (≥85th and ≥95th percentile), which differ from both CDC and WHO cut-offs. The UK90 reference population is British children from 1978-1990. A child at the 95th percentile today is likely heavier than a child at the 95th percentile in the 1990 reference. Scotland, Wales, and Northern Ireland run their own programmes but publish comparable figures. Australia: The National Health Survey uses self-reported height and weight for ages 2-17, which understates prevalence. The 2020-21 survey reported 25% of young people aged 5-17 as overweight including obese using WHO thresholds. The measured data from the 2011-13 Australian Health Survey, the most recent with direct measurement, put the figure at 27%.

Japan, Brazil, And India

Japan: The School Health Survey measures pupils at ages 6, 9, 12, and 14. Using the Japanese Society for the Study of Obesity thresholds, obesity prevalence in 2021 was about 11% for boys and 9% for girls aged 6-14. Japan uses a country-specific definition that does not match WHO or CDC cut-offs. Direct comparison with other countries is misleading. Brazil: The Pesquisa Nacional de Saúde (2019) measured young people aged 5-19 and reported 14% obesity using WHO thresholds, with a strong socioeconomic gradient: obesity is more prevalent in higher-income urban households, the reverse of the pattern seen in the UK and US. India: The Comprehensive National Nutrition Survey (2016-18) used WHO thresholds and reported 5% obesity in ages 5-19, with wide variation by state and urbanicity. Indian national guidelines recommend a lower BMI threshold for obesity (≥25 rather than ≥30) for adults, but paediatric cut-offs remain the WHO +2 SD level.

The IOTF Alternative Standard

The IOTF thresholds are an alternative standard used in many research studies. Published in 2000 by Cole et al. in the BMJ, the IOTF thresholds are centile curves that pass through adult BMI 25 and 30 at age 18, based on 97,876 young people from Brazil, Great Britain, Hong Kong, Netherlands, Singapore, and the USA. Their advantage is consistency across countries: a child classified as obese by IOTF in Brazil and in the UK sits on the same conceptual scale. Their disadvantage is that they are less sensitive than the CDC or WHO cut-offs. They identify fewer children as obese. A study comparing all three definitions in a single UK sample found that IOTF obesity prevalence was about two-thirds of the CDC or WHO prevalence in the same group. Any literature review that cites childhood obesity prevalence without naming the threshold is incomplete.

The most common error when reading childhood obesity prevalence data by country is treating the number as a universal truth rather than a survey-specific estimate tied to a particular growth reference, measurement method, and year. A child classified as obese under the CDC 2000 growth charts may not be classified as obese under the WHO growth standards or the IOTF thresholds. A prevalence figure from a self-reported survey is not comparable to one from a measured survey. A figure from 2019 is not the same as a figure from 2023. The only honest answer to 'What is the childhood obesity rate in country X?' is a sentence that names the survey, the year, the body that published it, and the cut-off used. Any statistic that omits those four things is not a fact. It is a floating number that will mislead the reader.