DEXA vs smart scales: what each really measures

The short version: a DEXA scan measures your body composition with low-dose X-rays and is the reference method used in clinical trials. A smart scale estimates it from an electrical signal and assumptions about bodies like yours. Both are useful; they are not the same kind of number, and they shouldn't be blended.

How each one works

DEXA passes two very low-dose X-ray energies through the body and measures how each tissue attenuates them, producing fat mass, lean mass and bone mineral separately — including where the fat sits (trunk versus limbs, and estimated visceral fat). It's the method the big GLP-1 trials used in their composition sub-studies, which is why we call it the reference method.

Smart scales use bioelectrical impedance (BIA): a tiny current travels through your body, and because fat conducts differently from lean tissue, the scale estimates composition from the resistance — then runs it through equations built on population data. Hydration, food timing, recent exercise, warm feet and the equation itself all move the estimate. Reviews of BIA report individual errors of several percentage points of body fat compared with reference methods.

So is a smart scale useless?

No — it's a *trend* instrument. A single reading can be off, but the same scale, used at the same time of day under similar conditions, tracks direction reasonably well over weeks. The errors that make a single number unreliable are fairly consistent for the same person, so the slope means more than the level. Treat any single-day jump as noise first.

When a DEXA is worth it

If you want to know your actual fat and lean mass — especially to see what months of weight change were made of — a scan before-and-after a journey is the informative pair. In New Zealand, DEXA body-composition scans are available privately in the main centres; prices are commonly in the low hundreds of dollars per scan, so a baseline plus a follow-up every few months is the typical pattern. Scan on the same machine where possible: cross-machine differences are real.

Reading the two together

Keep them as separate lines, not one blended number: the scan is your calibrated reference point, the scale is your daily trend between scans. (That separation is exactly how the Recomp Health app treats them — scan figures and scale estimates are never merged.) And remember DEXA's "lean mass" includes water and organs, not just muscle — so day-to-day hydration shifts belong to the scale's noise, and real change is measured scan to scan.

What we still don't know

BIA equations keep improving and vary by brand, so your scale's absolute accuracy is hard to state in general; and visceral-fat estimates from DEXA are themselves modelled, not directly measured. Neither limitation changes the practical rule: trends from the scale, truth from the scan.