All visualizations

Body composition percentiles

Explore body fat and lean tissue by age and sex.

Body fat across adulthood

Fat mass as a percentage of total body mass.

Body fat across adulthood — female, Non-Hispanic WhiteAge 20 to 85 on the horizontal axis; Body fat in % on the vertical axis. Shaded bands show population percentiles, not confidence intervals. The middle curve is the median. A vertical line marks age 40. A numerical table follows the chart.%Percentile20304050603102550 · median7590972030405060708085Age (years; 85 is the top-coded group)

Shading: middle 50%, 80% and 94% of the reference group. The blue line is the median.

Compare a DXA measurement

Enter DXA body fat percentage.

Add a value to see its position on the chart.

Values stay on this page and are cleared when you leave it. They are not saved or added to the URL.

Percentile values at this age
Body fat at age 40 · female · Non-Hispanic White
Percentile3102550759097
%25.329.734.138.943.547.651.6
What these comparisons can tell you

The regional lean definition needs care. Hinton’s paper labels the measurements lean mass but does not identify whether its NHANES field includes bone mineral. We retain the published values without making a bone correction. Measurement ranks are exploratory until the scan’s field convention is matched.

Lean tissue is not a direct muscle measurement. DXA lean soft tissue includes water and other non-fat, non-bone tissue. Arm and leg lean tissue is a useful muscle proxy; it does not measure strength or diagnose sarcopenia.

Height scaling is not matching by height. Lean and fat mass indices divide kilograms by height in meters squared. The kg/lb view multiplies those indices by the height you choose. It does not estimate a distribution conditional on that exact height.

A population snapshot, not an aging forecast. Curves compare different people at different ages, not the same people followed over time. A percentile describes relative position, not an ideal body composition or a treatment target.

Study coverage is limited. The scan protocol excluded people taller than 196 cm or heavier than 136 kg, and pregnant participants. Age 85 includes people aged 85 and older. The height control can illustrate other heights, but cannot supply evidence missing from the study.

Measurements must be comparable. These references come from Hologic DXA with NHANES calibration; other scanners, software and methods can differ. Smart-scale or caliper estimates are not interchangeable. These historical U.S. data exclude some people with very large bodies, and do not represent every ancestry or the present-day population.

Sources & calculation

Kelly et al. · 2009 · Published reference table S2

Whole-body fat references from weighted, multiply imputed NHANES data. Published tables report σ; this dataset converts to S = σ / M.

The published groups are Non-Hispanic White, Non-Hispanic Black, Mexican American. No pooled curve is available in these tables; the selected group is a reference choice, not an inferred personal characteristic. The Mexican American group does not represent all Hispanic populations.

Curves use the published LMS parameters: L describes skewness, M the median, and S relative spread. Parameters are linearly interpolated between published ages; values are not extrapolated beyond ages 20–85. Percentile estimates outside 3–97 are reported as a bound rather than an exact tail rank. Shaded bands show population spread, not confidence intervals.

Download the reference parameters