Simple, cheap, and objective

I've got an armload of questions to ask about it, but I have to applaud ideas like mid-upper arm circumference is a simple, objective method of assessing nutritional status.

BMI is a controversial way to judge health - but it is also a fairly easily manipulated one. Patients game weights all the time in all sorts of ways. Having a reliable and simple measure like this is an interesting angle. I look forward to hearing more about this.


  1. This is actually one of the key signs they use to determine how malnourished kids are in Third World countires because you don't always have a scale, nor do you have the time to weigh everyone. And the arm circumference is pretty accurate. They also look at pitting edema and whether you can walk.

    Very interesting.

  2. True - very interesting, although to be totally accurate would one not have to have some kind of guage of lifetime arm circumference - mine and, not surprisingly given our genetic links, my daughters' is quite small and although one daughter is slender and the other in danger of being underweight, I am most certainly NOT, I just have puny arms!

  3. Also, some people have larger arms and everyone loses weight differently -- I think its a bit too much to make blanket statements such as: anyone who has an arm circumference of 20cm is not malnourished. . .


  4. A:) et all - good points. D has tremendous natural upper body strength, reflected in rather thick arms. Even when she weighed less than 75% of her w/r weight, her arms would have easily exceeded 20cm. She also never got below a BMI of 18 really, even when she was >50lbs below w/r.

    One should always be careful in interpreting abstracts as the whole study. Word limits can realy affect interpreting the quality of the work! But, for those who may be interested, the r value reported (.79) translates to about 63% co-variance (known as R-squared) between the two measures. So, on average over this group of 55 people, if you know the arm circumference, you can predict BMI with about 63% accuracy. The flip-side is that there is a remaining 37% of variance. So if you are using this measure on a given individual, you could be correspondingly VERY wrong. It is also unclear whether the sample patients were followed from presentation as outpatients or inpatients. If the latter, the data might be problematic, as you'd be trying to make generalizations to a broader population based on the characteristics of kids who were already so sick they had to be admitted.

    While I agree that this is a simpler tool than BMI, and less prone to manipulation by the patient, I think it still is vulnerable to the same kinds of problems as BMI if it is used as the whole or biggest tool in the box, rather than just one of a larger kit.

  5. Interesting Irishup -- do you have training in statistics?

    I know underweight as I am now, my arms are small, but nowhere near as small as most people's arms at my BMI I like Marcella's idea of using lifetime arm circumference statistics -- that would make more sense -- perhaps looking also at percent change?

    Interesting post though Laura!


  6. I don't think arm circumference would work. What if someone has naturally thicker arms and she starves or purges herself into having "medium" sized arms? If you just went by circumference measurements then she would be considered "healthy".

    I like the idea of lifetime arm circumference statistics but then again, who gets their arms measure on a regular basis?


Post a Comment

Popular Posts