LBM, FFMI, and Body Fat: A Decision Tree for Body Composition

Lean body mass, FFMI, and body fat percentage are the three numbers that show up most often in body-composition conversations. They are related but not interchangeable, and the LBM engine produces different lean-mass numbers depending on which formula you pick. This article walks the three engines for an 80 kg male and produces a decision tree for which to anchor on under which condition.

Engine outputs

The same 80 kg / 178 cm male through the FFMI Calculator (16% body fat input) and the Body Fat Percentage Calculator (Navy, 38 cm neck / 84 cm waist). The three-formula Lean Body Mass Calculator takes only sex/weight/height, and an 80 kg / 178 cm male is its built-in demo case, so its outputs are quoted below rather than embedded:

# ffmi-calculator (computed live from /engines/ffmi-calculator.js)
Engine input
  weight_kg             = 80
  height_cm             = 178
  body_fat_pct          = 16

Engine output
  ffmi                  = 21.20944325211463
  adjustedFfmi          = 21.33144325211463
  fatFreeMassKg         = 67.2
  interpretation        = Above average — consistent training

# body-fat-percentage-calculator (computed live from /engines/body-fat-percentage-calculator.js)
Engine input
  sex                   = male
  waist_cm              = 84
  neck_cm               = 38
  height_cm             = 178

Engine output
  bodyFatPercent        = 15.654191587762455
  methods[0].name       = U.S. Navy
  methods[0].bodyFatPercent= 15.654191587762455
  methods[0].note       = DoD circumference method. Standard for military fitness assessments.
  average               = 15.654191587762455
  fatMassKg             = null
  leanMassKg            = null
  category              = Fitness
  coachSummary          = Your estimated body fat is 15.7% (Fitness range for males) using the Navy method.

The Lean Body Mass Calculator returns Boer 60.9 kg, James 62.1 kg, and Hume 57.1 kg lean-mass estimates that cluster tightly within about 5 kg. The three-formula average is 60.05 kg lean, which implies a body fat of 24.9% for this 80 kg male. Because all three are adult-validated regressions, the average is a trustworthy single number and the realistic adult-male LBM range here is 57.1 to 62.1 kg.^[4]

FFMI uses 16% body fat (the explicit input) to derive lean mass: 80 × 0.84 = 67.2 kg, divided by 1.78² = 21.21. The output sits clearly above the LBM engine's anthropometric estimates because the body-fat input was lower (16%) than what the LBM regression formulas implicitly assume.^[3]

Navy body fat returns 15.65% — agreeing closely with the FFMI input. Applied to the 80 kg bodyweight, that 15.65% implies roughly 67.5 kg of lean mass (80 × 0.8435), far closer to the FFMI engine's 67.2 kg than to the LBM engine's Boer/James estimates of 61–62 kg.^[1]

Reading the disagreement

Three engines for the same lifter, three different lean-mass estimates:

• LBM engine (Boer + James average): ~61.5 kg lean.
• FFMI engine (16% body fat input): 67.2 kg lean.
• Body Fat engine (Navy circumference): 67.48 kg lean.

The ~6 kg lean-mass spread is the entire diagnostic value. The LBM regression formulas (Boer, James) are anthropometric — they predict lean mass from height and weight alone, implicitly assuming an average body-fat percentage for the input population. The FFMI and Navy engines use measured or estimated body fat directly. For lean lifters (body fat under 17%), the anthropometric formulas read low, because a regression fitted on an average-body-fat population carries that average into every prediction. The gap is a property of the formulas rather than a measurement error.

When to anchor on each number

Anchor on body fat: during a cut

Body fat percentage tracked monthly is the right anchor during a deficit phase. Against DXA in 609 fit US Marines, the circumference method under-read men by about 2.5 points and over-read women by 1 to 2 points, with individual scatter of roughly ±4 points.^[1] That bias runs in a consistent direction for a given person, so successive measurements still track real change: a 2-percentage-point drop over 6 weeks is a real fat-mass change even though the absolute number is not research-grade.

Anchor on FFMI: for trained-population context

FFMI is the right number when the question is "where am I in the natural-trained distribution?" or "how close am I to the natural muscular ceiling?" The published natural-cap near 25 for men provides a meaningful endpoint to the metric; LBM and body fat don't.^[3]

Anchor on LBM: for absolute lean mass over time

LBM is the right number when the question is "did the lean mass actually grow?" Tracked LBM over a 6–12 month period with consistent measurement technique reveals lean-mass trajectory more cleanly than FFMI (which mixes the height divisor) or body fat (which moves with both fat-mass and lean-mass changes).

How they disagree

The three engines disagree primarily because they handle body fat differently:

1. LBM engine (Boer/James): doesn't ask for body fat. Returns lean mass implicit from anthropometry alone.
2. FFMI engine: requires body fat as an input. Lean mass is body-fat-derived.
3. Body fat engine: derives body fat from circumferences, then implies lean mass from there.

Two of the three (FFMI, Navy body fat) depend on the body-fat estimate. The LBM engine is independent of it. For lifters with body fat that differs meaningfully from population average (say below 17% or above 27% for men), the LBM and FFMI engines will systematically disagree by 3–8 kg of lean mass.

A decision tree

1. Is this a cut? → Anchor on body fat. Track monthly with Navy circumference.
2. Is this a bulk? → Anchor on LBM, secondary watch on body fat to confirm fat mass isn't outpacing lean gains.
3. Is the question "am I close to my natural cap?" → Anchor on FFMI.
4. Is the question "are my training adaptations comparable to a previous training block?" → Anchor on LBM (independent of body-fat-measurement noise).
5. Are you a contest-level lean lifter? → Use DXA. All three engines degrade at extreme body fat.

The hydration noise floor

All three engines share a common noise source that often gets overlooked: hydration state. A 1.5 kg swing in body water (typical day-to-day variation for an 80 kg lifter) moves the Navy body-fat reading by 0.5–1.0 percentage points, the LBM output by 1.5 kg, and the FFMI by ~0.5 points. For tracking purposes the published recommendation is to measure on the same morning of the week, fasted: in the first systematic reliability study of DXA in trained people, a day of ordinary activity and even a simple breakfast each substantially increased both the error and the mean of the lean-mass estimate.^[2] Inconsistent measurement conditions explain most of the "did I really gain muscle?" anxiety in week-to-week tracking.

The muscle-gain-potential cross-check

The Muscle Gain Potential Calculator sits one step downstream from FFMI: it takes current lean mass and trajectory and projects the realistic next 6–24 months of muscular development. Useful as a sanity check on whether the FFMI gap to the natural cap is reachable on a normal training timeline (typically yes if the gap is 1–2 FFMI points; no if it's 4+).

Related reading

Body Composition For Athletes covers the trained-population framing in depth, How To Measure Body Fat At Home compares the field methods that feed these engines, and FFMI Natty Boundaries examines where the FFMI natural-cap argument holds and breaks.

FAQ

Should I take an average of all three LBM formulas?

Yes. The engine returns Boer, James, and Hume, all adult-validated regressions that cluster within about 5 kg, so their average (60.05 kg for the 80 kg demo male) is a reliable single number. If you want one formula, Boer and James track best for trained adult males; treat Hume as a sanity check.^[4]

Which engine should I track over a 12-month training block?

LBM from the Boer formula tracked at consistent measurement times (same scale, same time of day, same hydration state). Treat the absolute number as an estimate and read the change over time instead — but that signal only holds if conditions are standardised, since ordinary daily activity and a single meal are enough to move a lean-mass measurement.^[2]

Is Navy body fat accurate for very lean lifters?

The circumference method also drifts with adiposity rather than failing evenly: in the Marine Corps survey the leanest individuals were over-estimated and the fattest under-estimated.^[1] Contest-prep lifters, who sit at the lean end of that curve, should anchor on callipers or DXA rather than Navy tape.