I spent more than a good few hours at the Leisure Industry Week expo at the NEC in Birmingham. It’s a strange mix of exhibitors – as well as the gym and fitness area that I was planning to see, many other forms of leisure equipment suppliers were present, ranging from leisure centre flooring to gym locker key mechanisms to hot tubs and zip wires.

Machines, machines everywhere!

However, I spent most of my time among the stands that were showing cardio and  resistance training machines, fitness education companies and nutritional supplements. And I also took a look at and took part in some of the latest exercise routines. And, oh yes, I did have a go at the zip wire too! I was amazed at some of the new and ingenious resistance and cardio machines but I did wonder how many more devices can be invented to replicate a basic ab crunch! Technology can be useful but sometimes it  perhaps over-complicates. The bells and whistles that now enhance a stationary bike although designed to make it this year’s latest must-have for the cardio elite are stunning – and stunningly pricey too!


Bio-electrical Impedance machines

But one area held my interest for a different reason. I’m driven a lot by numbers and statistics and I spotted three different companies offering bio-electrical impedance based (BIA) body composition analysis devices, or as I say, bathroom scales. But these have evolved way beyond just reporting weight! As well as selling them at the expo each company was offering a free test on their flagship product – and I was keen to see my results. It also occurred to me that there may be some variance in the figures and I wanted to review the consistency of the analyses from across the suppliers. I’ve no idea whether the electronic circuitry or mathematical algorithms are the same in all three, but I thought that a comparison of the outputs on a test of a willing volunteer (that’s me) would show a decent level of uniformity across all the devices.

How do they work?

BIA machines have been part of the fitness and health landscape on a commonly available basis since the 1980’s and they have become increasingly sophisticated in terms of how they do their job as the years have passed. The machines measure the impedance of the flow of a small electrical current in the body. Equations are then used to predict lean body mass and subsequently estimate body fat percentage. These machines measure impedance not fat – many assumptions and equations have to be applied to provide adequate results. Different types of body tissue react differently to impedance to the flow of the electrical current. Muscles with their very high water content present lower impedance than tissues such as fat and bone. The machines work on the basis that by measuring the resistance of a pathway in the body – a fair idea can be gained of the likely ratio of lean ‘water-rich’ tissue to fatty tissue. Hydration levels are therefore absolutely critical for decent results and equations may not adequately handle ‘unusual’ subjects. Expanding on this: the equations used in estimating BIA can involve a number of assumptions about the amount of water a certain tissue type will contain, the hydration level of the subject and the geometry and volume of the body pathway and have to factor in the racial and gender differences. Now you can see why the formulae are crucial for the machines to provide optimum readings.  My results varied between machines and this could reflect the way they calculated the water content in my body and also whether I’d had a sneaky cup of coffee between tests (caffeine acts as a diuretic).

One test to rate them all

So I devised a simple test of having my body analysis taken at all three stands, supervised by their appropriately-trained individuals, with as little time between the tests as practicable. This would reduce the effects that could be had on my test results due, for example, to my hydration levels. As the stands were some way apart and a fair amount of walking time was required, plus repeated bouts of on, off, emptying, and refilling my shoes, socks and pockets the gap between each test was about was about 45 minutes. I will say that this was a little longer than any scientists among you may like but it’s the best I could do on the day – and I didn’t eat or visit the loo in that time (thought you’d like to know that last fact!). Chris Zaremba Scales of Truth_2

And the results are in!

The three I tested were: – Tanita MC- 980inBody 230 (from Derwent Healthcare) – BCA-2A from Bodivis. Initial results were that I weighed 75.3kg on the Tanita, 75.8kg on the Bodivis and 76.3kg on the InBody. This spread of 1kg was perhaps a little more than I expected. However, there was a bigger disparity on the body fat numbers: 5.9kg on the InBody, 8.7kg on the Tanita, and 10.2kg on the Bodivis, giving percentage bodyfat results of 7.8%, 11.6% and 13.5% respectively. A level of difference that I wasn’t expecting! Many of the machines’ other results showed a similar level of disparity. Here are some examples:The output from each machine goes to the level of analysing muscle and fat on each limb, as well as the trunk. All the numbers differed – one machine had my left at  3.9kg of muscle and the right 0.2kg less than that whereas the machine at the other extreme had the left at 4.4kg and the right to be 0.1kg greater. – Basal Metabolic Rate is another example of a computation made by each device and when it came to me the result varied from 1600 to 1890 calories across the products – again a substantial disagreement. –  And a final example, a couple of the machines calculate a theoretical metabolic age – how old the body stats indicate you are, rather than actual years. I’m 55 and the Bodivis had me at 51 and the Tanita at 40 – thanks Tanita! (the InBody doesn’t perform this calculation).

Are any of them correct?

I was surprised by the variance in the results. I’m now not sure which machine to ‘believe’ (as it were) and if one was 100% accurate I don’t know which one it is. If you are after total accuracy then there are other methods that can be tried to measure body fat levels – these range from hydrostatic (inwater) testing methods (which we believe are the most accurate, if far from accessible method of body composition analysis) and  even good old skin-fold callipers – providing the user is skilled in their usage. However, I’m sure that sticking with one specific machine/method and measuring over time would give an accurate trend on variable change as they affected you and this would probably be fine if you are more interested in your progress over time rather than perhaps the most accurate of numbers.

In conclusion

The machines I put to the test at LIW retail at thousands of pounds with one over £6000. If I was in the market to spend that sort of cash, then I’d like to know that the machine I was purchasing was accurate to 100% (if not at least a high 90% reading). Unfortunately my straw poll test showed, in my opinion, an unacceptable level of disparity between machines and results and this left me still searching for the answer to just what my body fat levels are

Read More from Expert Chris Zaremba.

WatchFit Experts change lives!

And they can do the same for you.


Pollyanna Hale Health and Lifestyle coaches
Lost 13 Kg in Total
Mel, 32y Location: London, United Kingdom Working with Pollyanna changed everything. I lost 13kg, got toned and have more energy than ever! Get same results!


Chriz Zaremba Fitness Consultant
Lost 45 Kg in Total
Chris, 50y Location: London, United Kingdom Lost 45kg after the age of 50 and now competes and wins physique competitions and runs marathons Check our weight loss plans