Gallagher · The American journal of clinical nutrition 2000 · cross-sectional study · n=1626

Healthy percentage body fat ranges: an approach for developing guidelines based on body mass index.

Cited 2143 times in the scientific literature.

Level 4 - case-series / case-control

Cross-sectional observational study developing prediction equations from a convenience sample

PubMed 10966886 · doi:10.1093/ajcn/72.3.694 · record verified 2026-08-26

What was done

Researchers evaluated a convenience sample of 1,626 adults (BMI ≤ 35 kg/m²) across three ethnic groups (white, African American, and Asian) at three universities (Cambridge, Columbia, and Jikei). Body fat was measured using reference methods: a 4-compartment model at two sites and dual-energy X-ray absorptiometry (DXA) at all three sites. Multiple regression models were developed to predict percentage body fat from 1/BMI, sex, age, and ethnic group, which were then linked to established NIH/WHO BMI cutoffs (<18.5, ≥25, and ≥30 kg/m²).

What was found

Independent predictor variables for percentage body fat included 1/BMI, sex, age, and ethnic group, yielding regression models with R² values ranging from 0.74 to 0.92 and standard errors of the estimate (SEEs) from 2.8% to 5.4% body fat. These equations were used to generate provisional healthy percentage body fat ranges aligned with established BMI categories.

Why it matters

This study provides an empirical framework to translate widely used BMI cutoffs into corresponding percentage body fat ranges across different age, sex, and ethnic groups.

Limits

The study used a convenience sample and excluded individuals with BMI > 35 kg/m². Body composition reference methods varied across study sites (4-compartment model at two centers, DXA alone at the third). The resulting percentage body fat ranges are derived secondarily from BMI cutoffs rather than directly linked to clinical or health outcomes.

Cited by