What You'll Learn
Table of Contents
Why Weight Alone Misleads
Weight is one number covering several different changes at once. Fat, lean tissue, water and gut contents all move, and the scale reports their sum.
That matters most in exactly the situation where people watch it hardest. During substantial weight loss the question is not how much came off, it is what it was made of β the argument our muscle-loss guide makes at length.
Two people can lose the same eleven kilograms and end up in completely different places. The scale gives both the same number.
Waist: The Measure With A Consensus Behind It
If you only ever measure one thing beyond weight, measure this. It is the cheapest measurement in this article and the best evidenced.
The statement is blunt: βBMI alone is not sufficient to properly assess or manage the cardiometabolic risk associated with increased adiposity.β Within any BMI category, higher waist circumference carries substantially higher risk, and waist remains predictive after adjusting for BMI while BMIβs own association weakens. Crucially for anyone losing weight: waist reductions can occur with or without weight loss.
Full statement on PMC βBottom line: Read that last line again. Waist reductions can occur with or without weight loss. If the scale has stalled and the tape is still moving, something is working that the scale cannot see.
Our Picks
For the trend, if you want a device. An eight-electrode design with handles measures both an arm and a leg path rather than inferring the upper body from the legs alone.

FITINDEX Smart Body Composition Scale with Handle
8-electrode design with handles, app sync, 50+ reported metrics. Use it for the trend under controlled conditions, not for the absolute number. Confirm current price on Amazon.
Check price on Amazon βFor the measure that actually has a consensus statement behind it. A retracting tape with a fixed tension mechanism removes the largest source of error in home measurement β pulling it tighter on the days you want a better number.

FITINDEX Bluetooth Body Measuring Tape
Retracting tape with app logging, for waist, hips, arms and thighs. Costs a fraction of the scale and measures the variable with better evidence behind it.
Check price on Amazon βIf you buy only one, buy the tape.
How To Measure It So The Number Means Something
- At the navel, or midway between the lowest rib and the top of the hip bone. Pick one and never change it.
- Standing, at the end of a normal breath out. Not sucked in, not pushed out.
- Tape snug, not tight. It should sit against the skin without compressing it.
- Same time of day, ideally first thing, before eating.
- Every four weeks, not weekly. It moves slowly, and watching it daily produces noise rather than information.
Consistency of method matters more than precision. A measurement taken the same wrong way every time still tracks the trend correctly.
What A Smart Scale Is Actually Doing
Body-composition scales use bioelectrical impedance. A small current passes through the body, and because fat and lean tissue conduct differently, the resistance is used to estimate the proportions.
Two things follow from that mechanism, and both matter more than the marketing.
It is an estimate built on an equation, not a measurement of fat. The device measures resistance and then infers body fat using a formula fitted to a reference population.
Hydration changes the answer. Water conducts. Being dehydrated, having just trained, or having just eaten all shift the reading β which is a real problem on a medication where fluid intake falls without you noticing.
Foot-to-foot scales pass current up one leg and down the other, so the upper body is inferred rather than measured. Devices with handles add an arm-to-arm path, which is why eight-electrode models are a different class from the cheapest ones.
The Honest Limits
This is the part product pages leave out, and it decides how you should use the device.
Four consumer impedance devices measured against DXA and whole-body MRI. The best correlated well (rΒ² 0.92β0.96) with a mean bias under one percentage point β but its 95% limits of agreement ran from β6.59 to +4.61 percentage points of body fat. The weakest device spanned β14.54 to +8.58. Good on average; wide on any individual reading.
Full study on PMC βBottom line: A mean bias under one percentage point sounds excellent. Limits of agreement from β6.59 to +4.61 mean your individual reading could be six points out in either direction. Do not treat the number as your body-fat percentage.
So What Is It Good For?
The direction, measured under identical conditions.
The bias in these devices is largely systematic β it errs in the same direction for the same person each time. That makes the absolute figure unreliable and the change considerably more useful. If the scale said 31% in March and says 27% in September, the four-point move is meaningful even though neither figure may be correct.
To get that, control the conditions:
- Same time of day, first thing, after the toilet and before eating or drinking
- Same hydration state β not after training, not after a heavy meal
- Bare feet, same surface, hard floor rather than carpet
- Weekly at most, and read the four-week trend rather than the day
One caveat on the study above: it dates from 2008 and tested that era's devices, three of them foot-to-foot. Eight-electrode designs are a newer and better class. The direction of the finding still holds β consumer impedance tracks trends far better than it pins absolutes β but treat the specific limits as an illustration rather than as this device's specification.
What To Actually Track
Four numbers, four different frequencies. Together they catch what any one of them misses.
- Weight β weekly average, not a single day. Bodyweight swings one to two kilograms on water alone.
- Waist β every four weeks. The measure with the consensus behind it, and the one that moves when the scale will not.
- Body fat percentage β monthly, trend only. Never the absolute figure.
- Chair stands in thirty seconds β monthly. Free, needs no device, and it is the only one of the four that distinguishes losing weight from losing capability.
Bottom line: If weight is falling and chair stands are falling with it, none of the other numbers matter yet. Fix protein and training first.
What You Can Ignore
- Metric counts. β50+ body metricsβ mostly means the same impedance reading run through fifty formulas. Visceral fat rating, metabolic age and bone mass are derived from the identical measurement and inherit all of its uncertainty.
- Metabolic age. Not a clinical concept. It is a marketing number.
- Daily weighing anxiety. Weigh often if you like, but read the weekly average. The daily figure is mostly water.
Important: Bioelectrical impedance devices pass a small current through the body. Manufacturers generally advise against use with a pacemaker or other implanted electronic device, and during pregnancy. Check the manual and ask a clinician if either applies to you.