Let's pick up our in-depth exploration of body fat measurement with bioimpedance, also known as bioelectrical impedance, impedance, or BIA
Have you ever seen a home bathroom scale that estimates your body fat? Or seen a handheld device with conductive grips that can somehow provide you with your fat mass? More than likely, these devices are using bioimpedance (or BIA).
BIA is a quick and dirty method for predicting body fat. Most of these devices allow you to wear clothing, step up or hold onto a device for a few seconds, possibly enter some information about yourself – and voila! - You’re rewarded with a body fat number.
These devices are certainly convenient, and for the most part affordable, but this often comes at the price of both accuracy and consistency. A vast peer-reviewed literature has discussed what bioimpedance can and can’t do when it comes to predicting body composition. The conclusion? If you care about accuracy or precision in body composition, BIA is a risky choice.
BIA is a quick and easy ‘at home’ way to get a rough estimate of your body composition. It is low cost, low barrier to entry and doesn’t even require you to take off your clothing!
BIA is riddled with inaccuracies arising from a large variety of sources. There are high end BIA devices (such as InBody) that address some of these concerns, but at its core, BIA is a flawed technique when it comes to estimating and tracking body composition.
You may be asking, are there any affordable devices that allow me to track my composition with higher accuracy at home? Up next, we’ll discuss both commercial and in-home 3D scanning devices and our view on the future of body composition!
See below for the long form answer…
What is it and how does it work?
Bioimpedance, or bioelectrical impedance, or BIA, is scientifically described as the ability of biological tissue to impede electric current.
What on earth does that mean?
In plain speak, this means that different types of tissue and substances in the body respond differently to a low voltage electrical current. Fat tissue, for example, does not conduct electric charge very well. Fat free mass – which includes water, muscle, and bone – conducts electricity quite well. It’s largely the high volume of water (and electrolytes dissolved in your body’s water) in fat free mass that makes this component strongly conductive.
Sensors in these BIA devices measure the impedance – the voltage that returns to the sensors after traveling through the body – and use the measured value in equations validated from higher compartment body composition models. This logic relies on the assumption that there is an inverse relationship between body water and presence of fat (meaning the more body water, the less body fat and vice versa); and that impedance is the key that links these two phenomena.
There are a number of different types of BIA devices. There are handheld devices, BIA scales (that look much like your traditional bathroom scale), and some more complex ‘medical grade’ BIA devices. Some of these devices require additional inputs into their equations, possibly including gender, height, race, age, and exercise habits. Others have derived their equations for body fat using only impedance and a reference method, thereby predicting body fat from bioimpedance alone while eliminating other regressors (gender, race, etc).
BIA devices can run you as little as $30 and as much as thousands. Generally speaking, the more sophisticated and expensive the device, the more accurate. That said, they are still subject to the same sources of error and that error is quite large.
What are the primary sources of error?
BIA often infers composition of large segments of the body
Although this is not the case for every bioimpedance device, the majority of those used by non-medical consumers use a great deal of inference to measure body composition. The BIA foot scale (like that made by Tanita), for example, sends an electrical current up one leg and receives the current down the other leg. While this is useful for regional impedance, the trunk (which makes up nearly 50% of the body’s lean mass!) is largely inferred, rather than directly measured.
Individuals with unevenly distributed adiposity – for example someone with high levels of abdominal obesity – will experience a sizeable under-estimation of fat mass using leg-to-leg bioimpedance devices, such as BIA foot scales. Hand-to-hand impedance devices experience the converse; they often measure arm and upper trunk bioimpedance, but miss the lower trunk and leg regions. In recent years, more expensive and complex devices have been created that primarily mitigate this particular problem, but they are still subject to the numerous other limitations of BIA.
Photo courtesy of InBody
BIA is a prediction based on a prediction
In order to develop a new BIA technology, a manufacturer uses or develops an equation for body composition based off the impedance results. This equation may use a gold standard 4-compartment model as the reference method, but more often than not, the equation is based off a 2-compartment criterion method (like hydrostatic weighing).
This means that the error of the criterion method is compounded in the BIA prediction. Looking back at prior articles in this series, the gold-standard method of DXA, a 3-compartment model, sees approximately a 2% margin of error for the average user, with some individuals deviating as much at 7% from a higher compartment model. Thus, if hydrostatic weighing (a method already biased by assumptions) is used as a reference, BIA estimates will only amplify that error.
BIA is subject to fluctuations in the hydration of fat free mass
Accurate measurement of body water is central to the prediction of body fat from BIA. Fat free mass contains two types of water, intracellular and extracellular water. Intracellular water is that found in the body’s cells; thus, intracellular water is mostly observed in the muscles, bones, and organs. Extracellular water on the other hand is found outside of the cells, mostly in the blood and in the solution surrounding the body’s cells (interstitial fluids).
Almost all commercially available BIA devices use a frequency that is able to measure the body’s extracellular water, but not its intracellular water. Instead, the intracellular water is predicted from the extracellular water by assuming a constant ratio. This works reasonably well in normally hydrated and healthy individuals, but the assumption of this ratio is invalidated with abnormal hydration, disease, old age and obesity.
Current will always follow the path of least resistance
A tried and true physics principle teaches that an electrical current will follow the path of least resistance. This means that in extreme cases, the current won’t even come in contact with subcutaneous fat – which will likely result in a drastic under-estimation of percent fat.
How well does BIA do for different individuals?
This method is quite different from many of the other methods we’ve discussed thus far in that impedance does not measure any biological quantity. DXA, for example, measures bone AND tissue directly. Hydrostatic weighing measures fat mass and indirectly infers lean mass from this measurement. BIA, however, uses the impedance index (reminder: this is the value that permits the inference of total body water volume) as an independent predictor in a regression formula to predict body composition.
This regression formula is derived using a reference method (such as hydrostatic weighing) for a specific population – most commonly white, middle age, non-obese men and women. This means that any population measured using bioimpedance should be similar to the reference population (the population by which the formula was derived) in order for the results to be valid. Unfortunately, this is often not the case and can result in some pretty skewed results.
The takeaway here? Given absolutely perfect testing conditions, BIA will always be more valid for individuals closely replicating the sample with which the BIA formula was derived. For outliers – such as very lean, athletic or obese individuals – BIA is poorly suited and often will deviate as much as 8-12% from a criterion method like Hydrostatic Weighing or DXA.
How does BIA perform within the same individual?
Remember how BIA relies on an accurate representation of total body water (both intracellular and extracellular fluid) to predict body composition? Herein lies the problem! As the body changes, often so does the relative and absolute hydration of fat free mass.
In one study, people who engaged in a cardio regimen for weight loss saw an underestimation in fat loss and overestimation in fat-free mass loss. The authors propose that this is due to an increase in plasma volume, a common adaptation to cardiovascular training, which in turn increases extracellular water and inflates body fat estimates.
In obese individuals who lost weight, hydration of fat-free mass also decreased alongside decreasing weight, resulting in underestimated fat loss by BIA. The converse is also true! With increased weight, hydration of fat free mass increases, so gains in body fat may be obscured.
So, if you hear “BIA may be inaccurate, but at least I can track my progress”, you’re now armed with the evidence suggesting otherwise!