I wrote in my last blog about how I have long wanted a way to assess my body fat in detail inexpensively and at home. I found measuring tapes too imprecise for capturing most of the body fat changes that I see in the mirror in maintenance, and have not taken the time to become an expert caliper user. But I believe I have finally found a great tool for this purpose-the Skulpt Aim. I’ve been using this handheld device daily for about three weeks now. I use the Quick Test daily, which measures muscle quality (MQ) and bodyfat percentage (% fat) at 4 sites. The Skulpt app uses this data to generate an overall estimate of my body fat percentage. Less frequently, I also measure MQ and % fat for other individual muscles.
What are these measures, and what do they tell me?
You might be wondering (as I was initially) how about the device works and how the two things that it measures (% fat and muscle quality) are even different. I’ll start with % fat because I think this is by far the most familiar of the two, commonly called body fat percentage.
The % fat metric captured by the Aim estimates the percentage of a body’s weight that is made up by fat. The Aim assesses body fat by measuring the resistance of a muscle area. The more fat associated with the muscle, the more resistance it has to a current passed through it. The remainder of your total weight is collectively referred to as lean body mass, and is made up of various tissues and body fluids like bone, organs, muscle, blood, water, etc.
While it’s true that having more muscle does contribute to a low body fat percentage, skeletal muscle is just one component of your total lean body mass. I guess since it happens to be the component that becomes prominently visible when body fat is low, skeletal muscle has become strongly associated with low body fat in most people’s minds. But the two are actually quite distinct concepts. You can have low body fat and low amounts of skeletal muscle, and high body fat with high amounts of skeletal muscle and in both cases have little to no visible muscles. We usually see someone as having an athletic look when they have the combination of visible skeletal muscle and body fat that is low enough to let the muscle show through. I enjoyed this post by Brad Pilon about how these two variables contribute to the shape of an athletic woman. And not just because he featured my childhood idol and favorite superhero in it!
The fact that the Aim measures any attribute of the actual muscle fibers is really unique, in my experience. The web site says, “Muscle Quality (MQ) is a rating of your muscles’ fitness. A high MQ score means a lean, strong, and fit muscle.” MQ is measured using the time delay between a current being passed through the muscle and the voltage measurement that is detected afterwards. Apparently fitter and larger muscle fibers retain current longer. (I knew some day I’d really wish I had studied harder in physics class…sigh!) You can read a lot more details here on the Skulpt site. They have the explanations covered better than I ever could and there are some helpful diagrams. In any case, you’d expect higher MQ scores in people who are naturally gifted with more muscle, work out, and don’t store much fat inside their muscles. On the other hand, the lowest scores would be in people who have low skeletal muscle to start with, don’t work out, and tend to store fat intramuscularly.
A few comments first about why the information provided by the Aim is so unique and why I believe the data I am getting. I had a DXA scan done in January 2014 and it provided me the highest resolution information I’ve ever had about my lean body mass and body fat distribution. All my previous body fat assessments provided an overall number, where the DEXA gave me compartmental information on my left and right arms, legs, and trunk. The Aim improves on this by providing MQ and % fat assessments for various individual muscle areas. For example, instead of getting an arm measurement that is an average over my whole arm and shoulder area like I had with my DXA, I can now see details for bicep, tricep, shoulders, and forearms broken out individually. This new information reveals that these areas have % fat and MQ profiles that are somewhat different from one another.
In reviewing my results, I can see immediately that my highest MQ scores are found in areas that I train specifically and regularly. For example, I never train my calves directly, and my lowest MQ scores are found there. On the other hand, I train my back, chest, and shoulders regularly, and those areas have higher muscle quality scores. Imagine how useful this kind of information would be for assessing my progress if I decided to begin training my calves or wanted to selectively train another lagging body part!
The Aim also inspired my confidence by pinpointing a known issue with my abdominal musculature. I know that I have an abdominal diastasis on my right side, and the Aim consistently gives lower MQ and higher % fat scores on my right side versus my higher quality (and unaffected) left side. I’m not certain whether this has to do with additional body fat collecting in the right side area or the noticeable shifting of the muscle to the side, but I find it fascinating that this pattern is apparent in my data. Also, when I had my DXA, it became apparent that my spine is slightly curved, and I did see a slight imbalance in my lower back muscles. Whether this is a result of the spinal curvature or compensation for the abdominal imbalance, I don’t know, but it hasn’t been of clinical significance to me.
Soon after I began using the Aim, I took nine days off lifting to see if I could improve a neck strain issue that I have been experiencing off and on for a few years. I have been seeing a physical therapist for this issue and he said since taking time off did not improve it and the PT was, I could return to lifting. I collected data throughout this rest period, and interestingly I saw only small changes in my MQ and % fat scores during that time. This was somewhat reassuring as I had never had any previous assessment as to how an extended break like this impacted my muscles (other than visual cues).
When I introduced the Aim, I promised to share the following:
- Which areas are the easiest and hardest to measure
- What areas of my own body are leanest and which need more work
- How to get the most reproducible data out of your Aim
I’m going to address the first and last items first, and leave #2 to the end.
Which areas are the easiest and hardest to measure
Without a doubt, body areas you can’t see well are the most difficult to measure. These include lower and upper back and triceps. I have been able to get consistent tricep measurements, but I can’t easily reach the intended upper back measurement area without flexing and I often wait several seconds, then look at the device, only to see the error message to try again. Since I can’t see whether I’ve had a successful measurement, it requires a lot of trial and error and it’s very hard to tell if I’m positioning the device consistently each time. There isn’t really anything that can be done about this, except to have someone else do the back measurements for you, as the how-to video suggests, so I’ve enlisted my husband’s help for those.
Speaking of the how-to videos, after a few weeks of regular use I spent some time reviewing them to make sure I was still positioning the Aim correctly. When I saw the positioning of the back measurement, I realized that some of my earlier measurements were taken in a different area, so they’re not really comparable to my most recent set, but I’m OK with that. When taken at the correct location by my husband, the MQ scores were much higher than before. The upper back measurement is located on the trapezius, which I never work in isolation. Apparently it receives plenty of stimulation during other exercises, since it had a very high MQ score and low % fat relative to other areas.
How to get the most reproducible data out of your Aim
Like all such devices, the Aim shows some error on repeated measurements. The Skulpt website quotes the test-retest error to be about 5%. So for example, if I measured a body part at 20%, the next day’s measurement would likely be somewhere between 19-21%. My goal since I began using the Aim has been to determine the variability of my own measurement patterns and then seek to reduce that variation by improving my technique. Although I would like to tweak my training and observe the impact on my muscle quality and body fat percentages, if I am not taking reproducible measurements to start with, then it will be harder to assess the impact of any gains or losses due to training patterns.
There are a few things you too can do to make sure that you get the most consistent data when you measure with the Aim. I recommend experimenting to convince yourself that these factors are important:
- Spray the sensors, every single time, if you remove the device from its location between measurements. When I first started using the Aim, I did not appreciate the importance of the sensors being sprayed before each reading. Once I began to realize that this was affecting my consistency, I adjusted my technique to spray the contacts each time and my data got better. Later on I found that I could skip this step if I leave the wet device pressed on my skin while I do several replicate measurements in a row. This can be a little tricky to accomplish, as it requires you to remember which side the select and up/down buttons are on and sometimes press them without seeing them, but readings are very consistent this way.
- Spray the area you’re about to measure before you start. After tweaking my technique above, I started to notice that my first replicate measurement for a given body part seemed to be the most variable, often showing a MQ score that was 2-4 points different from later reps. Clearly there was something I was doing on the first rep that was different than the other reps, or the measurement conditions were slightly different. I examined my data closely to confirm my impression, and realized that although the sensor was wet before rep 1, the body area was not. I experimented with spraying the area first, which helped bring rep 1 back into line.
- Choose a position at which to measure each muscle and stick with it. I’ve noticed that the measurements for certain body parts-especially small muscles-appear to be more variable than others. Once I fixed my other measurement issues, I noticed a disturbing trend over several days-the standard deviation of my right bicep measurement was steadily increasing! I looked at my data and watched carefully where I was putting the device and came to believe this indicated a device positioning issue. I started experimenting to confirm that suspicion and found if I positioned the device half an inch to an inch closer to the top of the muscle near the shoulder, the MQ scores were indeed 3-4 points lower than when they were taken in areas slightly closer to the elbow. I didn’t go so far as to put a dot on my arm to match up to the top of the device, but I’m now much more aware of this impact and try to choose the position consistently.
Where do I store fat and which areas have the best and worse muscle quality?
I took screen shots of my MQ and % fat results with the most recent results for each area as of 8/4/2015. Below are the results for front and back % fat measures. Most of my body areas are within a percent or two of one another when comparing left to right. Clearly I see major body part to body part differences, though. I also have the traditional female fat storage pattern-higher body fat in my lower half. Fortunately for my heart disease risks, my abdominal fat is lower.
I used JMP’s Fit Y by X platform to show how mean values for muscle quality and body fat percentage differed across body parts. The graph below shows the results for fat % for a simple model using a concatenated variable that treats each unique combination of body part and side as a separate group. The green diamonds represent the means for each body part-side combination. You can see that the parts I measure each day in the Quick Test have the most observations and hence the widest mean diamonds.
This graph uses the full data table and does not exclude any outlier points. But it is really useful not only for identifying outlier body part areas like upper back, abs, and calves, but for pinpointing outlier measurements which I may want to exclude from my analysis.
Not surprisingly, my MQ numbers below reflect the inverse relationship between % fat and MQ, which I’ll talk more about in a later blog. Areas where I had the lowest % fat above tend to have the highest MQ and vice versa.. As I mentioned above, my well-trained back and abs have the highest MQ scores. I don’t work my abs directly but they are a stabilizer muscle for many exercises. My calves, which I never train directly, are my lowest quality area (110-115). The rest of my MQ scores are clustered in the 120-135 range.
Looking again at a simple model using a single variable to represent each side-part combination, the same outliers are apparent as showed up in my % fat data. Again, outlier points are also obvious in the data.
The screenshots below show my Total Body (overall body fat percentage) measurements for MQ and % fat from 8/4/15. As you can see on the progress graph at the bottom, my MQ scores have been trending slightly up and % fat scores have been trending slightly down over the past week or so (ignoring days in the middle of the week that were impacted by some bicep measurement variability issues that I’ll discuss at a later date). My daily average body fat percentage was around 20.5% on July 19th when my weight was 135.8 and measured 18.6% today at a weight of 135.8.
I have been actively working on cutting my calories slightly each day for several weeks now, but I can’t yet verify whether this 2% difference is due to water weight changes from to my monthly cycle, measurement consistency improvements, true changes in body composition, or a combination of those factors. I’ll be tracking it over time and see what happens going forward. I’ll also be looking at my data more with outlier points excluded, likely in a future JMP blog post in my Fitness and Food series.
In summary, I’m not all that concerned about the 100% accuracy of the numbers I am getting from my Skulpt and I can’t validate them easily with DXA anytime soon. But just going off past information, I’m pretty certain my numbers are in the ballpark of “the truth,” so I’ll keep collecting my data and I think it will be very useful going forward as I start to tweak my training approach and workout timing. By the way, to order your own Aim, you can click on my referral link here. Full disclosure: you’ll get a better price on the Aim and I’ll get a $20 gift card if you purchase (and don’t return it). You’ll get your own referral link and opportunity to get $20 when you purchase. (On a side note, it’s nice to be able to benefit from sharing a better price for a device I think is awesome!)