Children and Teenagers: BMI Percentiles, Growth Charts and What They Mean

You have a BMI result for your child, or you have been told they sit at a particular centile, and you need to know what that number does and does not mean. The answer is not a diagnosis. A child's BMI is read against age- and sex-specific growth charts that show where they sit on a reference distribution of other children measured in the same way. This hub routes you to the right guide depending on what you are trying to do: understand a percentile result, read a growth chart, make sense of a teenage BMI shift during puberty, or talk to a child about weight without causing harm. Every threshold and statistic here is attributed to the issuing body with the year. You will not find a diet plan, a calorie target, or a weight-loss instruction for a child. It points you toward a clinician for any personalised assessment.

This guide covers children, teenagers, BMI, growth charts, and centiles. It explains the three reference frameworks in play, the CDC 2000 growth charts for the US, the WHO Child Growth Standards for under-5s, and the UK-WHO growth charts used by the Royal College of Paediatrics in Britain, and what each one is designed to do.

• CDC child BMI categories (US, 2000): Underweight: <5th percentile; Healthy weight: 5th to <85th percentile; Overweight: 85th to <95th percentile; Obesity: ≥95th percentile; Severe obesity: ≥120% of 95th percentile or BMI ≥35 kg/m² (whichever is lower). Source: CDC Clinical Growth Charts.
• WHO growth standards (0-5 years): Prescriptive standards from the WHO Multicentre Growth Reference Study (2006). Describe how children should grow under optimal conditions. Based on 8,440 healthy, breastfed infants from six countries.
• UK-WHO growth charts (2-18 years): NICE guideline CG189 (2014, updated 2022) recommends UK90 growth charts. Clinical cut-offs: Overweight ≥91st centile; Obesity ≥98th centile; Underweight ≤2nd centile. Population monitoring cut-offs: Overweight ≥85th centile; Obesity ≥95th centile.
• Child BMI is not adult BMI: Adult cut-offs (underweight <18.5, normal 18.5-24.9, overweight 25-29.9, obese ≥30) are never applied to children. Paediatric BMI is always interpreted as a percentile or z-score against age and sex.

A BMI of 22 means something completely different for an 8-year-old girl than for a 40-year-old man. For a child, the raw number is meaningless until it is plotted on a growth chart that accounts for age and sex. The CDC 2000 growth charts for ages 2-20 are built from NHANES data collected between 1963 and 1994, representing how US children grew during that period. The WHO Child Growth Standards for ages 0-5, published in 2006, are prescriptive: they describe how children should grow under optimal conditions, based on the WHO Multicentre Growth Reference Study of 8,440 healthy, breastfed infants from Brazil, Ghana, India, Norway, Oman, and the USA. The UK-WHO growth charts merge these two frameworks for UK clinical use, following NICE guideline CG189.

The thing most parents get wrong is treating a centile as a health grade. A child at the 90th centile is not diagnosed as having excess weight; the centile describes where that child sits on a reference distribution. The 85th and 95th percentiles are screening thresholds, not clinical diagnoses. Two children at the same centile can have very different body compositions. BMI for children is a screening tool at population scale, not a diagnostic for an individual body.

CDC Growth Charts And The 2022 Extension

The United States uses the CDC 2000 BMI-for-age growth charts for children aged 2-20. The data come from the National Health and Nutrition Examination Survey (NHANES) for 1963-1994, with approximately 31,000 children. In December 2022, the CDC published Extended BMI-for-Age Growth Charts that add percentiles up to the 99.99th percentile for children with severe obesity, based on NHANES 1988-2016 data from about 30,000 children. This update addresses a real problem: the original 2000 charts compressed severely obese children into the 99th percentile, making it impossible to track changes at the extreme.

WHO Standards And The UK Compromise

The WHO Child Growth Standards for ages 0-5 use a different approach. The Multicentre Growth Reference Study selected healthy, breastfed infants from six countries to describe optimal growth. For ages 5-19, the WHO 2007 reference merges the MGRS data at age 5 with the 1977 NCHS/WHO reference for older children. The WHO thresholds use standard deviations, not centiles: thinness is below -2 SD, overweight above +1 SD, and obesity above +2 SD. At age 19, the adult WHO cut-offs take over.

In the United Kingdom, the Royal College of Paediatrics recommends UK90 growth charts for children aged 2-18, based on Cole et al. (1995) using pooled data from 11 British surveys between 1978 and 1993. The clinical cut-offs are different: overweight starts at the 91st centile, and obesity at the 98th centile. The population monitoring cut-offs used by the National Child Measurement Programme in England use the 85th and 95th centiles instead. This creates exactly the confusion you would expect: a child classified as 'overweight' for monitoring may not be 'overweight' for clinical purposes.

What A Centile Actually Measures

A centile is a rank on a reference distribution. If your child is at the 80th centile for BMI, that means 80% of children of the same age and sex in the reference population have a lower BMI, and 20% have a higher one. That is all it means. It is not a measure of body fat, metabolic health, or fitness. It is not a diagnosis of a health condition.

Body Fat, Muscle And Misclassification

The relationship between a BMI centile and actual body fat varies. A meta-analysis by Javed et al. (2015) covering 53,521 children aged 0-18 found that BMI has a pooled sensitivity of 0.73-0.82 and specificity of 0.93-0.95 for detecting excess adiposity using DXA as the reference. That means BMI misses a significant fraction of children with excess body fat, and it also labels some children as overweight who have normal body fat. The false positive rate is higher in athletic children with high lean mass. A child with a BMI at the 95th centile who is muscular rather than overfat is not uncommon.

BMI misclassification also affects short and tall children. The BMI-for-age percentile adjusts for age and sex, but not for height independently. Very tall children can have an upward bias in BMI z-score, while very short children can have a downward bias. This is not a flaw in the arithmetic; it is a feature of the ratio. It means you cannot read a centile as a simple health verdict.

The BMI centile and the BMI z-score are related but not identical. The centile is a rank; the z-score is the number of standard deviations from the median. At extremes, the z-score is more informative because centiles compress. A child at the 99.9th centile and a child at the 99.99th centile both appear at the extreme end of the centile scale, but their z-scores differ substantially. This is why the CDC extended growth charts matter for clinical management of severe obesity.

Puberty changes body composition in ways that can make a teenager's BMI look like it has moved off a cliff or shot up in a month. BMI increases with pubertal stage independent of age, and the CDC and WHO growth charts do not incorporate Tanner stage. A boy gaining lean mass during growth spurts will see his BMI rise without accumulating fat. A girl gaining body fat as part of normal pubertal development will also see her BMI rise, but the composition is different. The growth chart does not distinguish between these scenarios.

The BMI Rebound And What It Signals

The BMI rebound, the point where the BMI curve reaches its nadir and starts rising again, typically between age 4 and 7, is a separate but related phenomenon. An earlier rebound, before age 5, is associated with increased adult obesity risk, according to Whitaker et al. (1998) following 105 children to age 21-29. But again, this is a population-level association, not a prediction for an individual teenager.

Why Self-Reported Numbers Mislead

For teenagers, self-reported height and weight introduce systematic bias into any BMI calculation. Adolescents under-report weight by 0.5-2.0 kg and over-report height by 0.5-2.0 cm, resulting in a BMI underestimation of 0.5-1.5 kg/m² on average, based on NHANES III data for 2,965 adolescents aged 12-16. If your teenager's BMI is based on what they told the school nurse rather than on measured height and weight, know that the true value is almost certainly higher.

This guide routes; it does not prescribe. If your child's BMI percentile has crossed a threshold and you are worried, the next step is a consultation with a paediatrician, not a search for a diet plan. Bring two things: the growth chart showing the child's centile history, and a list of questions about what the centile means for your child. A paediatric endocrinologist can assess growth velocity, pubertal stage, and body composition more accurately than any screening number can.

What The National Child Measurement Programme Delivers

The National Child Measurement Programme in England measures children in Reception (age 4-5) and Year 6 (age 10-11). In the 2023/24 school year, 9.6% of Reception children and 22.1% of Year 6 children had obesity. The NCMP uses standardised protocols: height measured to the nearest 0.1 cm with a Leicester Height Measure, weight to the nearest 0.1 kg on Class III electronic scales, children in light clothing and no shoes. If you receive a letter with your child's BMI category, use it as a conversation starter with your GP, not as a verdict.

Do not apply adult BMI cut-offs to your child. Do not use a waist-to-height ratio threshold of 0.5 for your child; NICE CG189 advises that threshold for adults only and has not validated paediatric cut-offs. Do not compare your child's body fat percentage from a BIA scale to a DXA measurement. The measurements are not interchangeable, and no universal healthy-range standard exists for paediatric body fat percentage across all methods and populations.