Understanding the InBody 970 and Y-Scope: A Comprehensive Guide
Updated: Sep 25
An InBody scan takes under a minute and prints a page of numbers: skeletal muscle mass, body fat percentage, segmental lean mass, an extracellular water ratio, phase angle, visceral fat and an estimated basal metabolic rate. Ohana Performance Institute runs Norway's only InBody 970, and the most common question after a scan is "what does this number mean". This guide explains each metric, how the device arrives at it, and how much trust the published research says it deserves.
The evidence
Juliana and colleagues (2026, Frontiers in Analytical Science) compared the InBody 970S with DXA in 148 healthy Malaysian adults aged 20 to 45. Looney and colleagues (2024, Frontiers in Nutrition) tested the reliability, day-to-day variability and DXA accuracy of the InBody 770 in 14 military-age adults over five visits. Rosa, Lukaski and Sardinha (2025, Reviews in Endocrine and Metabolic Disorders) review what phase angle represents. Chan and colleagues (2023, Frontiers in Endocrinology) compared bioimpedance and DXA fat estimates with three-dimensional MRI visceral fat in 4,558 UK Biobank participants, using a Tanita rather than an InBody device.
What happened
Agreement with DXA is strong at group level. The 970S gave R² of 0.96 for fat mass, 0.97 for fat-free mass and 0.92 for body fat percentage, with mean bias of +0.91 kg, −2.61 kg and +2.54 percentage points respectively, and limits of agreement of about −2.7 to +7.7 points for body fat.
The direction of error varies by population. The InBody 770 read body fat 4.0 points lower than DXA (standard error 2.6 per cent) and fat-free mass 2.8 kg higher: a different sign from the 2026 study.
Repeatability is excellent. Test-retest reliability for whole-body water and mass was 0.999 or higher; same-day repeat scans differed by 0.0 to 0.2 kg and between days by 0.1 to 0.7 kg under standardised conditions.
Screening use is supported. The 970S identified sarcopenia risk with 95.1 per cent sensitivity and classified obesity with 100 per cent specificity and 75.2 per cent sensitivity.
Visceral fat from impedance is an estimate. In UK Biobank, bioimpedance trunk fat explained 49 per cent of the variance in MRI visceral fat (R² 0.49) against 0.94 for DXA, although that device lacked a dedicated visceral algorithm.
The current answer
Bioelectrical impedance passes small currents at several frequencies through the body and measures resistance and reactance. Water conducts, fat does not, and cell membranes act as capacitors, so the device estimates total body water, then fat-free mass, then fat mass, and derives the rest.
Skeletal muscle mass and segmental lean mass are estimates of water-rich tissue in the whole body and in each limb and the trunk.
Body fat percentage is what remains after fat-free mass.
The extracellular-to-total water ratio drifts upward with inflammation, oedema and muscle loss.
Phase angle is the delay between current and voltage at the cell membrane; it tracks cell mass and membrane integrity, rising with muscle and falling with illness and age.
Visceral fat is modelled from trunk impedance and body dimensions.
Basal metabolic rate is an equation applied to fat-free mass, not a measurement.
Certainty is moderate: the device is highly repeatable and agrees with DXA at group level, with individual errors of a few kilograms or percentage points.
How certain is this?
Moderate. Reliability is established. Accuracy against DXA is consistent in strength but not in direction of bias, and the samples were young, healthy, mostly normal-weight adults, mostly men in the 2024 study, so transfer to older, heavier or very lean athletic bodies is uncertain. DXA is itself a model rather than ground truth. A 970-specific comparison against a four-compartment reference in trained men and women would firm this up.
An InBody scan is a precise instrument for tracking change in the same person under the same conditions, and an approximate one for stating any single number as fact.
What this means in practice
Read trends, not snapshots. Two scans under the same conditions tell you far more than one.
Standardise the scan: same time of day, fasted, empty bladder, no training in the previous 24 hours, no alcohol.
Expect any single body fat reading to sit within a few percentage points of a DXA value, in either direction.
Treat visceral fat and basal metabolic rate as modelled estimates for tracking, not diagnosis.
Use phase angle and the water ratio as flags that prompt a look at recovery, illness or muscle loss, not as diagnoses.
Remember who was studied: young, healthy adults. Very lean or very heavy bodies may sit outside the validated range.
Ohana Performance Institute offers InBody 970 scans with a standardised protocol and a walk-through of every number in the context of your goals. Book a scan.
References
Last reviewed: 4 September 2026. Re-review if a new validation study of the InBody 970 against a four-compartment reference is published, or by September 2028.
Comments