11.5 Body Composition Assessment in Children
Key Takeaways
- BMI is a proxy for adiposity, not a measure of it, and misclassifies children with cerebral palsy, Down syndrome, short stature, edema, or steroid exposure
- Bioelectrical impedance estimates fat-free mass from total body water, so it is invalid in dehydration, edema, ascites, and rapid fluid shifts
- Air displacement plethysmography uses the PEA POD for infants up to about 8 kg and the BOD POD for cooperative older children
- Fat-free mass index (fat-free mass divided by height squared) tracks lung function in cystic fibrosis better than BMI does
- Whole-body dual-energy X-ray absorptiometry is the practical clinical reference for fat, lean, and bone mass, while the four-compartment model is the research gold standard
Why BMI Is Not Enough
Body mass index divides weight by height squared and cannot tell fat from muscle from fluid. In population screening that limitation is tolerable; in specialty pediatric practice it is a source of real clinical error. Four scenarios show why the CSP blueprint pushes beyond anthropometry:
- Non-ambulatory cerebral palsy. Muscle mass is markedly reduced, so a child at the 25th BMI percentile may carry a high fat fraction with almost no lean tissue. Feeding to a "normal" BMI produces overfeeding, excess adiposity, and harder transfers for caregivers.
- Cystic fibrosis. BMI can be preserved by fat while lean mass erodes. Fat-free mass index (FFMI = fat-free mass in kg / height in m²) correlates with forced expiratory volume in 1 second more tightly than BMI does, which is why CF centers increasingly track composition rather than weight alone.
- Oncology and inflammatory bowel disease on corticosteroids. Steroids add fat and fluid while catabolizing muscle, so weight rises as function falls - the classic "weight is stable, the child is wasting" trap.
- Anorexia nervosa in recovery. Early weight restoration is disproportionately central adiposity and fluid; composition data reassures patients and clinicians that lean tissue recovery follows and helps interpret plateaus.
Below age 2, BMI is not used at all - weight-for-length is the index (see 11.2) - and even above age 2, short stature mathematically inflates BMI, so syndromic and skeletal-dysplasia populations need composition or condition-specific references.
The Measurement Toolkit
| Method | What it actually measures | Pediatric strengths | Key limitations |
|---|---|---|---|
| Dual-energy X-ray absorptiometry (DXA) | Three compartments: fat mass, lean soft tissue, bone mineral | Practical clinical reference; also yields bone mineral density z-scores; whole-body-less-head is the preferred pediatric readout | Small radiation dose; motion artifact in young children; results vary by manufacturer and software; scan bed size limits very large patients |
| Bioelectrical impedance analysis (BIA) | Total body water via impedance, converted to fat-free mass by prediction equations | Cheap, portable, fast, repeatable; useful for tracking trends in one child on one device | Hydration-dependent - invalid with edema, ascites, dehydration, dialysis shifts; requires pediatric-validated, population-specific equations |
| Air displacement plethysmography | Body volume, then density, then a two-compartment fat/fat-free split | PEA POD for infants to about 8 kg; BOD POD for cooperative older children and adolescents; no radiation | Requires stillness and minimal clothing; assumes a fixed fat-free mass density that shifts with age and hydration |
| Isotope (deuterium) dilution | Total body water directly | Reference method for fat-free mass; usable in infants | Expensive, slow, research settings |
| Skinfold thickness | Subcutaneous fat at defined sites, converted with pediatric equations such as Slaughter (triceps + subscapular, validated roughly ages 8-18) | Bedside, inexpensive, serial tracking of one child by one measurer | Large intra- and inter-observer error; calipers and training required; ignores visceral fat |
| Mid-upper arm circumference and arm muscle area | Combined muscle-plus-fat, or muscle alone when paired with triceps skinfold | Bedside surrogate that is relatively independent of fluid status and usable when height cannot be measured | Coarse; reference data thinner than for weight and height |
| Four-compartment model | Fat, water, protein, mineral combined from several methods | Research gold standard against which the others are validated | Not a clinical tool |
Choosing the Right Tool for the Question
Match the method to the clinical question rather than reaching for whatever the unit owns:
- "Is this child's low weight muscle loss or fat loss?" - DXA if available; otherwise triceps skinfold plus mid-upper arm muscle area at the bedside.
- "Is this fluid or tissue?" - Do not use BIA in an edematous or fluid-overloaded child; use MUAC trends and physical exam, and interpret weight against dry weight.
- "Is catch-up growth proportional?" - Track length/height velocity alongside weight; disproportionate weight gain with flat linear growth is adiposity, not recovery, and predicts later insulin resistance.
- "Is bone at risk?" - DXA bone mineral density with height-for-age adjustment, because raw areal density under-reads in short children. Indications include long-term parenteral nutrition, chronic steroids, cystic fibrosis from about age 8, anorexia nervosa, non-ambulatory cerebral palsy, and inflammatory bowel disease.
- "Is the infant's growth quality good?" - PEA POD-derived fat mass in preterm graduates distinguishes healthy lean accretion from the excess fat deposition that follows aggressive caloric catch-up.
Interpreting and Documenting Results
Report composition as age- and sex-specific z-scores or percentiles, never as raw kilograms in isolation, and always name the method and device - values are not interchangeable between DXA and BIA, or even between DXA manufacturers. Because the reference databases for pediatric fat mass index and fat-free mass index are narrower than for BMI, the within-child trend on a single device is far more trustworthy than a single cross-sectional value.
Finally, keep the ethical frame in view. Composition data should sharpen a nutrition prescription, not become another number a child with obesity or an eating disorder is measured against. Share results in the context of function and growth, use person-first language, and skip routine composition testing in children where it will not change management.
A 9-year-old with nephrotic syndrome has significant peripheral edema and ascites. The team requests body composition analysis. Which method is least appropriate in this situation?
Which body composition index correlates most closely with lung function (FEV1) in children with cystic fibrosis and is therefore increasingly tracked alongside BMI?
Which device is designed for measuring body composition by air displacement plethysmography in infants weighing up to about 8 kg?