PT on the Net Research

Body Fat Testing - Reliability vs Validity

One of the great dilemmas I have is what to tell the person sitting next to me on a flight when they ask, “So, what do you do?” The right answer is, “I am an exercise physiologist,” which is always answered by, “Great, can you tell me how I can lose weight?” I must admit that to avoid the unbelievably long list of questions that always follow this scenario, I now tell them I teach 5th grade history.

Fortunately, I did break this rule on a recent trip. I happened to be sitting next to a man and his wife that would alternate between English and some sort of unusual slang whenever other passengers would pass us in the aisle. I couldn’t help myself so I had to ask them where they were from. They said they were from London, that this was their first time visiting the US and the slang was some sort of a regional form of communication. They then asked me, “What do you do?” Feeling safe, I gave them the real answer. To my surprise, they had never heard of such a profession and wanted to know more about what I did. I told them I educated people about the benefits of exercise and that I also performed a variety of functional tests. I started off mentioning maximal treadmill tests and tests that measure body fat levels. It was at this point the woman immediately stopped me and said, “You really measure body fat levels on Americans?” I said, “Yes” and she quickly interjected, “You must be very wealthy.”

I continue to find that incident very enlightening for the simple fact that before coming to America this couple had no idea about the concept of exercise, but they quickly realized that America would be a great place to measure body fat. The problem is not that we have an enormous number of overweight Americans, but how do we accurately measure who is “overweight” and how “overweight” they actually are?

As far back as 1950, health professionals have suspected that the risk of developing certain diseases increased in "overweight" individuals. In the past, the exact risk was sometimes difficult to determine because it had always been quite difficult to define "overweight." More recent research has been able to demonstrate that this risk is more associated with being "overfat" than just simply being "overweight." In other words, if you want to know if your weight is unhealthy, you need to measure your relative fatness or your percent of body fat.

Based upon current health information, it appears that the upper limit for a healthy percentage of body fat is 25% for men and 30% for women. Although the data is not as well defined for lower limits, there seems to be no health benefit associated with body fat percentages below 12% for men and 18% for women. With these values in mind, one could make the statement that a healthy percent body fat range for men could be 12-18% and 18-26% for women.

In addition, once you know your percent body fat, the next step is to figure your total number of pounds of fat and muscle. These calculations allow you to determine a healthy weight based on how much fat you have and not just on what you think you should weigh. Unfortunately, height/weight charts fail to determine percent body fat; therefore, they offer little help in determining a healthy weight or more importantly, in monitoring gains or losses in either muscle or fat pounds.

Today there are many methods available for estimating percent body fat. It is important to realize that even the most expensive and "high tech" procedures provide only estimates of percent body fat. Regardless of what procedure you use, all measurements can have inherent error. The degree of error is commonly compared by either validity or reliability. Validity of a given measure is described as how well it measures what it is supposed to measure. On the other hand, reliability is described as how reproducible the measure is. In other words, your % body fat results could be accurate but not reproducible (you are really 15%, but you only get this reading occasionally) or your % body fat results could be inaccurate but very reproducible (you are really 15%, but you always measured at 35%). The following should help you compare the validity and reliability of today's most common methods of measuring percent body fat.


Has always been considered the "gold standard" for determining body density and therefore, body fat. Because it is the gold standard, all other body fat tests are never more accurate than underwater weighing; however, their goal is to come as close as possible to this standard. Unfortunately, even though it sets the standard, underwater weighing relies on the false assumption that both fat and muscle tissue have a unique and a never-changing density. Regardless of that false assumption, the undisputed science is that muscle – because its density is greater than water – will sink and fat – because its density is less than water – will float. Therefore, if you determine your weight under water, you can determine your percent of muscle and fat.


Skinfold sites and their equations are determined by initially underwater weighing a group of individuals and then generating a prediction equation that will, in essence, predict their underwater weighing score. This methodology can have an inherent error if the subject who is being “pinched” was not represented in the original sample. An example of how to avoid this would be to develop a separate prediction equation for a very specific population such as high school wrestlers. As for the general population, deriving a prediction equation using a large sample of ages and body fat percentages could minimize the error. However, even with following these guidelines the equation will still be wrong at the extremes. This could result in telling a 5% fat male that he is actually 9% and telling a 25% fat male that he is only 20%. This scenario will encourage the leaner male to workout more and the heavier male to continue a sedentary lifestyle. Fortunately, in today’s sedentary world there continues to be a larger number of Americans whose skinfold sites are too large to even measure. This point was recently brought out in a very well written New York Times article. Other considerable errors are possible when the tests are performed by unskilled technicians and by the use of inferior skin fold calipers.

3. CIRCUMFERENCE MEASUREMENTS (girth measurements)

Circumference measurements are becoming more and more popular due to the fact that they do not require expensive equipment or a skilled technician. In fact, it is possible to measure yourself. Like skinfolds, the measurement sites and equations are determined by initially underwater weighing a group of individuals and then generating a prediction equation that will, in essence, predict their underwater weighing score. The science behind girth measurements is based on the fact that fat tends to accumulate around the midsection. Consequently, if your circumference measurements increase, you are increasing your body fat. In other words, you can now correlate "inches gained" to fat pounds gained and “inches lost" to fat pounds lost. However, girth measurements like skinfold measurements, have a population that is best suited for the test. This type of test is most successful when used on subjects who have average or above average body fat percentages. This is due to the fact that girth measurements tend to increase significantly when individuals are already quite fat and do not decrease significantly when the individual is already quite lean. The calculations to predict % body fat can be done using a hand held calculator or on the web by entering the measurements in inches.


This form of testing has continued to become very popular in the one-day health fair, athletic club and wellness settings primarily because the device is relatively inexpensive and testing can be performed quickly by relatively unskilled technicians. The science behind the procedure involves the transmittance of a low level current through the body from two surface electrodes placed on the back of the wrist and the front of the ankle. Other versions can be purchased using a bathroom scale with preset settings for “athletes” and “normal”. All are based on the fact that since muscle has a relatively large water content it is also very conductive. The basic theory behind this testing is that the more body water that is present, the higher the amount of muscle per pound. Unfortunately, most exercisers have a tremendous amount of fluctuation in their fluid levels due to sweat loss vs. fluid consumption. It is imperative – with respect to reliability – that pre and post test settings pay particular attention to the subject’s exercise and fluid status prior to testing.


Infrared testing, like the impedance testing, has become more and more accepted in similar settings. In its present form, infrared interactance estimates percent body fat from a single measure taken at the bicep of the individual. The measuring device is capable of recognizing both fat and lean at the bicep and total body fat is then estimated from this measure. This type of test has an inherent reliability error that is related to a consistent placement on the bicep. In an attempt to improve the validity of this device, the potential for using additional measuring sites are currently being researched.

In summary, the reliability and validity of tests that measure % body fat will undoubtedly be a matter of constant debate. As health professionals, you need to choose the best test that fits the specific population that you are testing. If done correctly, periodic body fat testing can provide the subject with data as to the relative health risk associated with their weight, their current pounds of fat and muscle and the net changes that are related to initiating an exercise or nutrition program. Lastly, remember body fat testing was not invented to tell fat people that they are fat.