DXA Analysis

[Guest guest]

Hi folks:

I have just sent the following to an organization that should know

the answers to my two questions. I do not expect to get an answer.

In my experience, when one writes to an organization asking questions

which suggest their methods may be inadequate the natural response

(if my suggestion is correct) is to ignore the issue, and the email.

But if I am mistaken I will get an answer straightening me out for

sure. I will of course keep you good folks updated. Here is the

content of the email:

" Good morning:

I have a question about the reason for the use of a reference group

matched only for age, sex and race for determination of T and Z

scores for BMD.

In several studies I have read it is agreed that BMI is a major

factor that is well correlated with BMD. Indeed, one study, done in

Japan, found that BMI was the *only* factor with a statistically

significant correlation with BMD. Most authorities these days seem

to agree that the ideally healthy BMI is between 20 and 25. A BMI of

between 25 and 30 is considered overweight, while a BMI of 30 is

considered to be the threshold for obesity.

According to the most recent data from the NHANES survey, for the

period 1999 - 2002, the average BMI was well above 25. For example,

for males aged 60 to 74, the average BMI was found to have risen

appreciably since the previous survey and stood at 28.6. If the same

rate of increase has continued since the last survey then the current

average BMI for that group is 29.3.

It appears therefore that the T and Z scores provided by DXA analysis

software compare the BMD of each patient with the average BMD score

for people who are grossly overweight - indeed only marginally short

of being classified as obese.

So I have two questions:

First, since BMD scores are so closely related to BMI does it make

sense to compare the BMD of a patient whose weight places them right

in the middle of the 'ideal' BMI range (say, a BMI of 22.5) with that

of patients very nearly categorized as obese? Why would anyone want

to take as a reference group, a group that is quite clearly

unhealthy? Isn't the usual custom with other medical tests to

compare a patient's score with a **healthy** reference group? Is

there a good reason to make an exception in the case of BMD, and use

an unhealthy group for reference purposes?

Second, could you please tell me where I can find data for North

Americans which show BMD 'norms' broken out not only for age, race

and gender, but also for BMI for each group?

Many thanks for any assistance you can provide with this.

(signed) "