## I shall tell you about Weight percentile calculation and BMI percentile

By Steven B. Halls, MD

I made this page because I occasionally receive emails from parents of small children, wondering why the BMI calculator would classify their child as Overweight or Obese, and yet the Weight Percentile value didn’t seem overweight. |
Back to the BMI and Weight percentile calculator |

**The bottom line**:

### Weight Percentiles are different from BMI percentiles.

Please don’t think the "Weight percentile" value shown below the "Body Description" is a BMI percentile, because it is not exactly the same.

Before I explain the difference, let me give some background about BMI percentiles. In adults, BMI percentiles are not usually relevant because the classification of people as Overweight or Obese, is based on fixed BMI thresholds^{1} of 25 or 30. And since most of the visitors to my website are adults, young adults especially, and since BMI as a classifier of people’s fatness has many known flaws, I decided to also show something extra; Weight percentiles.

In adults, using a BMI threshold of 25 as the definition of "overweight" will classify over half of the American population as overweight. Thus, it is entirely possible to have a Weight percentile at the 50^{th} percentile, and be considered overweight. You could be average weight and overweight at the same time.

In Children, the Body Description shown on my BMI calculator uses the BMI percentiles 85^{th} and 95^{th} percentiles, as thresholds^{1}. My calculator differs from the CDC-based calculators, only in the wording: CDC says "*at risk of… overweight*" whereas I just come right out and simply say "*overweight/obese*".

For children: | halls.md terminology |
CDC terminology |

BMI > 85th percentile: | "Overweight" |
"At risk of overweight" |

BMI > 95th percentile: | "Obese" |
"overweight" |

Why did I choose to differ from the CDCs wording? Because most of my website visitors are young adults and I didn’t want my calculator to behave inconsistently in the transition years between children and adults. A person of age 17.9 years would not be pleased to be just "at risk of overweight" and then be suddenly labelled "obese" when they become age 18.0 years old. That doesn’t make sense.

But for smaller children and the parents of those children, I agree it is both harsh and overly judgmental to label young children as obese. The word "obese" just isn’t an appropriate descriptor of younger children. Since I couldn’t think of a scientific justification to set a specific age boundary, I instead simply hope parents will use common sense to realize that these Body Description words are not very suitable for young children.

The BMI percentile thresholds for children and the BMI value thresholds for adults^{1}, are based on Chronological Age. But what if a child is taller than average? Think about it.

A tall child could easily be mis-labelled as obese, even on the Body Description given by my own BMI calculator, or on any other BMI calculator or chart based on the CDC thresholds.

But the Weight Percentile calculation takes Height into account, which makes it **special **and different from the BMI thresholds. For both children and adults, the Weight percentile calculation considers both chronological age and median-age-for-height, as it estimates how your weight compares to a population of other people of similar gender, age and height.

If I haven’t confused you yet, I’m about to… There’s another reason why the BMI-derived Body Description and the Weight Percentile, can seem to disagree. They are based on different datasets from different years.

Strange as it may seem, the CDC published new charts^{5} in year 2000, of height, weight and BMI for children, but those charts are based on a broad mix of data spanning many decades^{2}, and the end result are charts that reflect the way kids were in the 1970s^{3}.

On the other hand, the Weight Percentile calculation is based exclusively on the NHANES III dataset from the 1990s^{4}.

You should think of the CDC’s BMI thresholds as idealized criteria from the 1970s – before America started to get really fat, whereas the Weight Percentile calculation lets you compare yourself to current people.

Is the halls.md BMI calculator webpage suitable for evaluating children’s weight status? It’s not perfect, but it could be of help. Hardly any other BMI calculators on the internet bother to compare a child’s BMI to the 85^{th} and 95^{th} BMI thresholds. No other BMI calculators on the internet can show Weight Percentiles – adjusted for Age AND height. As long as you remember that the Weight Percentile is not a BMI percentile, and as long as you don’t apply the labels "overweight" and "obese" to children, then you should be OK.

References

- Body Mass Index evaluation Criteria
- Observatons about the CDC child growth charts
- NHANES I study’s similarity to current CDC’s BMI curves
- NHANES III. Third National Health and Nutrition Examination Survey. 1988-1994 data from USA.
- http://www.cdc.gov/nccdphp/dnpa/bmi/bmi-for-age.htm
- Additional References about Body Mass Index

back to the Body Mass Index calculator