Why did my InBody and DEXA results not match?

Because they are measuring differently, and one of them is more sensitive to what you did that morning. InBody estimates body composition by running a small current through your body and calculating resistance, which shifts with hydration, recent meals, and recent exercise. DEXA uses low-dose X-ray and does not move around with those daily factors the same way. A gap between the two scans is expected, not a sign either machine is broken.

Which number on my body composition scan should I actually trust?

The direction of change across repeat scans taken under the same conditions, more than any single absolute number. Comparability research on multi-frequency bioelectrical devices found accuracy holds up much better when conditions are standardized (same time of day, same hydration, same pre-scan routine) than in typical real-world use. If your fat mass or lean mass is trending down or up scan over scan, believe that trend. Treat one isolated body fat percentage as an estimate, not a verdict.

Is my visceral fat number from an InBody scan reliable?

Treat it as the least reliable number on the printout. The same 1,000-person comparison found visceral adipose tissue estimates had the weakest agreement between InBody and DEXA of any measurement on the scan, notably weaker than total fat mass or lean mass. It is fine to glance at as a rough directional flag, but it is not precise enough to make a decision on by itself.

Back to Blog

How to Read Your InBody or DEXA Scan Results (Without Panicking Over One Number)

Q: Is InBody or DEXA more accurate for body fat percentage?

DEXA is the more accurate method. A real-world comparison of 1,000 healthy adults found InBody (a multi-frequency bioelectrical impedance device) underestimated total fat mass by an average of 3.7 kilograms in men and 1.9 kilograms in women compared with DEXA. InBody still correlates reasonably well with DEXA across a population, but for any one person's exact number, DEXA is the closer read.

InBody and DEXA do not agree with each other, and that is normal. Here is which numbers on your scan to trust, which to treat as directional only, and what to actually change this week.

Heather Bauer, RD · September 5, 2026

Here is the direct answer: no two body composition scans agree perfectly, and that is not a malfunction. If your InBody printout and a DEXA scan gave you different numbers, or two InBody scans a month apart swung more than you expected, the fix is not to distrust body composition data altogether. It is knowing which numbers on the page to actually act on, and which ones to read as a rough direction rather than a fact.

Why don’t InBody and DEXA agree?

They are not measuring the same way. InBody uses bioelectrical impedance: it sends a small, safe current through your body and estimates fat, muscle, and water based on how much resistance that current meets. Resistance shifts with how hydrated you are, what you ate a few hours ago, and whether you exercised that day. DEXA uses low-dose X-ray to directly distinguish fat, lean tissue, and bone, and it is far less sensitive to those same-day factors.

A real-world comparison of 1,000 healthy adults measured on both machines found InBody underestimated total fat mass by an average of 3.7 kilograms in men and 1.9 kilograms in women relative to DEXA (Potter, Ward, Chapman et al., European Journal of Clinical Nutrition, 2025). The two devices still moved together reasonably well across the group as a whole. The gap shows up most when you zoom in on any one person’s single reading.

Which number should you actually trust?

Trend over single reading, every time. A separate comparison of several BIA devices against DEXA in a clinical setting found strong overall correlation (r of 0.97 or higher) but wide individual-level disagreement on any given scan, the kind of spread that shows up in a Bland-Altman plot as good-on-average, unreliable-on-one-reading (Lee, Sun, Bae, Son & Yoo, Korean Journal of Health Promotion, 2025). Practically: if your fat mass, muscle mass, or body fat percentage is moving in a consistent direction across two or three scans taken the same way, that trend is real information. One number in isolation is closer to a snapshot with some blur around the edges.

Consistency in how you take repeat scans matters more than most people realize. Lab research tracking the same 14 people across five visits over three weeks found bioelectrical impedance measurements were meaningfully more reliable when time of day, hydration, and pre-scan routine were held constant (Looney, Schafer, Chapman, Pryor, Potter, Roberts & Friedl, Frontiers in Nutrition, 2024). If you get repeat InBody scans, do them fasted, at a similar hydration level, and around the same time of day each time. That single habit removes most of the noise you would otherwise be chasing.

What about visceral fat and muscle mass numbers?

Treat visceral fat as the least trustworthy line on the page. In that same 1,000-person comparison, visceral adipose tissue had the weakest agreement between InBody and DEXA of anything measured, clearly behind total fat mass and lean mass. It is fine as a rough directional flag over time, not as a precise figure to react to on its own.

Muscle (lean mass) is worth the same trend-not-single-reading read, especially if you are also watching this number while on a GLP-1, where a real share of weight lost can come from lean tissue rather than fat alone. The protein target that protects muscle while appetite is suppressed matters here regardless of which machine gave you the number.

What to actually do with your results this week

Do not overhaul your plan off one scan. If a number surprised you, either direction, hold off on a reaction until you have a second reading taken the same way. If a decision genuinely hinges on a precise number (a medical threshold, a program requirement), a DEXA scan is the closer read; an InBody number is a good-enough estimate for tracking your own trend over time, not for pinning down an exact figure. Either way, the printout is a data point for a conversation, not a grade.

The bottom line

InBody and DEXA disagreeing is normal, not a red flag on either machine. Read the trend across scans taken the same way, treat visceral fat as the least precise number on the page, and bring the result into a plan rather than reacting to it alone. A coaching layer built to track that trend with you, not just hand you a number, is what turns a body composition printout into an actual next step.