BMI Poor Measure of Body Fat in Heart Failure
Tuesday, April 5, 2011
Body mass index (BMI) is a widely used anthropmetric surrogate for body composition as it correlates well with body fat content across a wide range of body ages, shapes, and sizes.
However it does have its limitations in special populations (e.g. infants, competitive athletes, pregnancy, etc.). It may now be time to add patients with heart failure to this list.
In a study just published by Antigone Oreopoulos and colleagues from the University of Alberta, Edmonton, Canada, in Congestive Heart Failure, we examined the relationship between various anthropometric indices and body composition in men and women with heart failure.
Body composition was assessed by dual-energy x-ray absorptiometry in 140 patients with congestive heart failure and compared to BMI, waist circumference, waist-stature ratio, and waist index. Diagnostic accuracy of detecting obesity or high central fat was also examined.
While in men, all of the anthropometric indices except waist index were just as well (albeit moderately) correlated with lean body mass as with the actual percentage of body fat, in women, all 4 anthropometric measures were unable to significantly differentiate between body fat and lean body mass.
Thus, none of the anthropometric indices accurately reflected body composition in women with congestive heart failure and were also of limited use in men.
These finding may be due to the remarkable variability in lean body mass and fluid retention seen in these patients – both of which can strongly determine body weight irrespective of changes in body fat.
The implications for clinicians are to interpret anthropometric indices including BMI with caution in patients with heart failure – both with regard to predicting excess fat as well as in regard to any changes in body weight.
AMS
Edmonton, Alberta
p.s. The Canadian Obesity Network is seeking a new Finance Manager. For more information click here.
Oreopoulos A, Fonarow GC, Ezekowitz JA, McAlister FA, Sharma AM, Kalantar-Zadeh K, Norris CM, Johnson JA, & Padwal RS (2011). Do Anthropometric Indices Accurately Reflect Directly Measured Body Composition in Men and Women With Chronic Heart Failure? Congestive heart failure (Greenwich, Conn.), 17 (2), 90-92 PMID: 21449998
Tuesday, April 5, 2011
First sentence, paragraph 5, did you mean to say “While in men . . .” or were you saying that “on average.” If you were saying what I think you were saying, you were comparing men and women, and then the sentence ends as follows . . .
“in women, all 4 anthropometric measures were unable to significantly differentiate between body fat and lean body mass.”
Okay. And yet the conclusion is, simply:
“. . . interpret anthropometric indices including BMI with caution in patients with heart failure – both with regard to predicting excess fat as well as in regard to any changes in body weight.”
Am I smelling sexism? Shouldn’t we also regard BMI cautiously with regard to, er, women in general? Or, in medicine, a measure still doesn’t matter unless it applies to the true generic people . . . men. What year are we living in?
Tuesday, April 5, 2011
“Am I smelling sexism?”
Not sure what to make of your comment – in medicine (and biology) there are (thankfully?) some very ‘real’ differences between men and women – body composition is one such parameter with important sex differences. BMI works fine in populations of non-pregnant women but apparently not in women with heart failure.
Regarding the clinical value of using BMI to define obesity in men and women – that’s a very different issue that I’ve addressed in many previous postings.
Thanks for spotting the typo in para 5 – fixed 🙂
Tuesday, April 5, 2011
Oh, I get that there are very real differences in men and women. The sexism comes in precisely because we don’t always start from that assumption and have, instead, started from a male-centric view and just generalized. (And I wish I could say it’s stopped, but . . . another day.)
So, you’re saying that BMI “works fine in populations of non-pregnant women, but not in women with heart failure.” Hmmmm. Okay. “works fine”? I don’t know what you mean by that. The clinical value (or lack of value) is exactly the problem, and you have talked about that, if I’m remembering correctly, in other posts. BMI, taken alone, as a measure or indicator of anything, is troublesome. This study, which showed that it doesn’t predict fat composition in women with heart disease seems to underscore that, or at least add more evidence to that notion, especially for women. Hmmm.
Maybe I’m just not at the right depot for your train of thought today. That’s okay. I’ll come back tomorrow. Maybe my head will be clearer. Don’t give it another thought.
Oh, and thanks for the one-week experiment with the plural. It looked good to me. Someday soon, the world will be ready.
Tuesday, April 5, 2011
There’s a difference between using a measure at the population level and the individual level. BMI is a relatively good measure of risk at the population level in the general population, not so good at the individual level (this is called ecological fallacy – the assumption that the associations are the same at both levels). In the context of HF, it is especially important to learn that it is a poor measure of risk at the population level because we know that a higher BMI is paradoxically associated with better survival in this population. But BMI is assumed to be a measure of fat – and we can see from this study that it is not doing a good job of measuring fat in this population, especially in women. So although we don’t yet fully understand the separate effects of muscle and fat on mortality in this population, we can at least say that clinicians should probably be measuring body composition rather than BMI in heart failure as a better measure of risk.