Is a DEXA or InBody scan worth it on a GLP-1?

Once you know that some of the weight you lose on a GLP-1 is muscle, not fat,⁵ the obvious next question is how to check. Gyms, med spas, and longevity clinics have made two tools widely available for exactly this: the DEXA scan and the InBody (or similar bioelectrical impedance) scan. Both spit out a number for lean mass. Neither number means what most marketing implies it means.

This article walks through what each device actually measures, how accurate each one is against the other, what a scan costs versus what a free calculator gets you, and how often re-scanning is actually useful rather than just expensive noise.

What each one actually measures

A DEXA (dual-energy X-ray absorptiometry) scan passes two low-dose X-ray beams through your body. Because fat, lean tissue, and bone absorb the two energies differently, the machine can separate them directly, producing a full-body and regional breakdown of fat mass, lean mass, and bone mineral density.¹ It is the method researchers use as a reference standard in the GLP-1 trials that measured body composition, including the SURMOUNT-1 substudy.⁴

An InBody or similar BIA device works differently. It sends a small, imperceptible electrical current through your body and measures resistance. Because fat, muscle, and water conduct electricity differently, the device estimates composition from that resistance using population-based prediction equations, not a direct measurement of tissue.² That distinction, estimate versus direct measurement, is the source of most of the accuracy gap between the two.

How accurate is each one, really

A 2022 comparison of high-frequency BIA devices, including an InBody model, against DEXA in Korean adults found strong statistical agreement: standardized coefficients of 0.95 or higher for appendicular lean mass and 0.98 or higher for fat-free mass, with the two methods tracking each other closely across individuals.² That sounds like a green light for BIA. But the same study reported a systematic bias underneath that correlation: BIA consistently underestimated appendicular lean mass and body-fat percentage while overestimating fat-free mass relative to DEXA.² Strong correlation and accurate agreement are not the same thing, and this is the gap that matters if you are trying to read a single number off a report.

BIA also reacts to things that have nothing to do with your actual muscle: hydration status, whether you ate or exercised recently, and even room temperature can shift the reading. That does not make it useless, a consistent bias is still useful for tracking a trend if conditions are held constant, but it means a single InBody printout is a rough estimate, not a lab-grade measurement.

DEXA is not immune to its own version of this problem. A methodology review focused on athletic populations found that DEXA precision depends heavily on standardized scanning conditions, consistent positioning, a controlled fasting and hydration state, and, ideally, the same scanner and analysis settings on every visit.³ A follow-up scan run on a different machine, or without the same prep, can introduce enough noise to make a small real change look bigger or smaller than it is, or to manufacture a change that is not really there.⁶

Cost and access: the tradeoff that actually decides this

DEXA machines are medical-grade X-ray equipment, so scans are generally offered through specialty body-composition clinics, some gyms, or imaging centers, and they tend to cost more per session and be less widely available than a BIA scan.⁷ InBody and similar BIA devices are common in commercial gyms, sometimes included free with membership, and produce a result in under a minute with no radiation exposure at all.

That access gap is worth being honest about. A slightly-less-precise number you can get for free every week is often more useful in practice than a more precise number you have to pay for and drive to a clinic to get every few months. The right tool depends on what you are actually trying to answer, and how often you are willing and able to check.

How often is a scan actually useful

More frequent is not automatically better. DEXA's own precision literature is explicit that a change has to be large enough to exceed the scan's measurement noise before it means anything, which is why researchers standardize fasting state and, where possible, use the same densitometer for every follow-up scan on the same person.³ Scanning weekly on a device with a few percent of built-in variability will mostly show you noise, not progress.

There is no dedicated clinical trial establishing an ideal scan interval for people on a GLP-1 specifically. As a practical, non-clinical reference point, DEXA-focused providers commonly suggest roughly every 8 to 12 weeks as an interval long enough for a real body-composition change to plausibly clear the noise floor for most people.⁸ That is a provider recommendation, not a trial finding, and it should be treated as a reasonable starting rhythm rather than a rule.

What a scan can and cannot tell you

A DEXA-defined drop in "lean mass" includes water and organ tissue, not only contractile muscle, so even a well-run scan is an imperfect proxy for how strong or functional you are.⁵ Clinicians use DEXA-measured appendicular lean mass as part of the formal criteria for diagnosing sarcopenia,⁹ which tells you the measurement is clinically meaningful, but a single number below or above a population cutoff still is not the same thing as knowing how you actually perform in daily life.

This is where a scan pairs best with something free: our protein and lean-mass calculator gives you a starting daily protein target and an estimated composition baseline in a couple of minutes, no appointment or radiation required, and you can pair that with tracking what you can actually lift or carry over time. A periodic scan then becomes a confirming data point on top of that baseline, rather than the only source of information you have.

What the evidence does not say

No published trial has compared DEXA and InBody head-to-head specifically in people taking a GLP-1 medication; the accuracy data above comes from general adult populations, and rapid fluid shifts from a new medication or a diuretic-like effect in early treatment could plausibly affect a BIA reading in ways not captured by validation studies done outside that context. The 8-to-12-week scanning interval mentioned above is industry guidance from DEXA providers, not a result from a randomized trial, and no study has established an optimal re-scan schedule for tracking lean mass during active GLP-1 treatment specifically. Cost figures for scans vary widely by provider, region, and whether a package or single session is purchased, so no specific price is quoted here; check current pricing locally before booking.

The bottom line

DEXA is the more accurate tool for a single snapshot, and it is why researchers use it in the trials this site cites elsewhere. InBody and other BIA devices are close enough, correlated at 0.95 or better with DEXA for lean mass, to be genuinely useful for tracking a trend, as long as you accept a known bias and control for hydration and timing between checks.² Neither replaces the free, always-available signal of your own protein intake and the weight on the bar. A scan is worth it if you want a periodic confirming data point and can afford the cost or access; it is not a requirement for doing the two things, protein and resistance training, that the evidence says actually protect your muscle.¹⁰

Frequently asked