Monday, August 29, 2011

Men who are skinny-fat: There are quite a few of them

The graph below (from Wikipedia) plots body fat percentage (BF) against body mass index (BMI) for men. The data is a bit old: 1994. The top-left quadrant refers to men with BF greater than 25 percent and BMI lower than 25. A man with a BF greater than 25 has crossed into obese territory, even though a BMI lower than 25 would suggest that he is not even overweight. These folks are what we could call skinny-fat men.


The data is from the National Health and Nutrition Examination Survey (NHANES), so it is from the USA only. Interesting that even though this data is from 1994, we already could find quite a few men with more than 25 percent BF and a BMI of around 20. One example of this would be a man who is 5’11’’, weighing 145 lbs, and who would be technically obese!

About 8 percent of the entire sample of men used as a basis for the plot fell into the area defined by the top-left quadrant – the skinny-fat men. (That quadrant is one in which the BMI measure is quite deceiving; another is the bottom-right quadrant.) Most of us would be tempted to conclude that all of these men were sick or on the path to becoming so. But we do not know this for sure. On the standard American diet, I think it is a reasonably good guess that these skinny-fat men would not fare very well.

What is most interesting for me regarding this data, which definitely has some measurement error built in (e.g., zero BF), is that it suggests that the percentage of skinny-fat men in the general population is surprisingly high. (And this seems to be the case for women as well.) Almost too high to characterize being skinny-fat as a disease per se, much less a genetic disease. Genetic diseases tend to be rarer.

In populations under significant natural selection pressure, which does not include modern humans living in developed countries, genetic diseases tend to be wiped out by evolution. (The unfortunate reality is that modern medicine helps these diseases spread, although quite slowly.)  Moreover, the prevalence of diabetes in the population was not as high as 8 percent in 1994, and is not that high today either; although it tends to be concentrated in some areas and cluster with obesity as defined based on both BF and BMI.

And again, who knows, maybe these folks (the skinny-fat men) were not even the least healthy in the whole sample, as one may be tempted to conclude.

Maybe being skinny-fat is a trait, passed on across generations, not a disease. Maybe such a trait was useful at some point in the not so distant past to some of our ancestors, but leads to degenerative diseases in the context of a typical Western diet. Long-living Asians with low BMI tend to gravitate more toward the skinny-fat quadrant than many of their non-Asian counterparts. That is, long-living Asians generally tend have higher BF percentage at the same BMI (see a discussion about the Okinawans on this post).

Evolution is a deceptively simple process, which can lead to very odd results.

This “trait-not-disease” idea may sound like semantics, but it has major implications. It would mean that many of the folks who are currently seen as diseased or disease-prone, are in fact simply “different”. At a point in time in our past, under a unique set of circumstances, they might have been the ones who would have survived. The ones who would have been perceived as healthier than average.

25 comments:

Mike said...

It is weird that there would be some people with 0% body fat. That can't be right.

Richard A. said...

Weight lifting will tend to push down %BF even if BMI stays the same.

Ned Kock said...

A few of the outliers are definitely due to measurement error, including the ones where BF is zero.

Not good, but always present in datasets. This is why in advanced stats multiple measures are aggregated into one – it contributes to minimizing measurement error.

Ned Kock said...
This comment has been removed by the author.
Ned Kock said...

Oops, the comment above (deleted) was meant for the previous post.

Anonymous said...

Not to nit-pick a good post, but BMI of 30 is obese, not 25.

Ned Kock said...

Hi Anon. Thanks, but did I imply anywhere that a BMI > 25 equals obesity? I meant to say only that a BF > 25 percent equals obesity (and only for men), and that a BMI < 25 suggests otherwise (not even overweight).

gwarm said...

Perhaps skinny fat asian men live longer due to lower testosterone, lower PSA (higher carbohydrate, soy).

http://www.youtube.com/watch?playnext=1&index=4&feature=PlayList&v=oqVbDWE94RU&list=PL53AA35449C7DD652#t=5m23s @8min20sec "Japan has just as much prostate cancer as we do, but the amount dying is 1/10th of American men. #1 life expectancy of any nation, dying with their cancer instead of from their cancer."

Ned Kock said...

Hi Gwarm. I think that the prostate cancer connection is with obesity, and perhaps other factors associated with the Westernized diet/lifestyle (e.g., stress), but not with higher levels of testosterone.

Testosterone levels have been steadily going down over the years in American men, and prostate cancer rates do not seem to be going down in a correlated way.

Moreover, free testosterone is positively correlated with various health markers in men, and negatively with various diseases. For example, see this graph of free testosterone and intima-media thickness, which suggests a negative association with heart disease:

http://www.nature.com/ijo/journal/v27/n7/fig_tab/0802292f1.html

Lerner said...

Hi, Ned. One thing I'll say for you, you certainly do have your finger on the sources of good visuals :) In fact, I'd first found your site when I came across that great image showing relative sizes of lipoproteins. Seeing images, for me, really adds to merely seeing figures from tables.

I would second gwarm's idea on testosterone being related not to incidence but to progression of cancers. In fact, anything that an exercising person considers as beneficial to building muscle would likely tend to increase cancer growth rates as well.

WRT specifically to prostate cancer, "chemical castration" is a Tx.

Heart disease is entirely different from neoplasm.

Ned Kock said...

Thanks Lerner. It looks like a second comment by you did not make it through. I’ll add it myself below.

Ned Kock said...
This comment has been removed by the author.
Ned Kock said...

Mistaken comment above from another post -- sorry.

Lerner said...

HGH has also been talked about as promoting progression of cancers. That's especially true for the typically very slow growing prostrate cancer. There is a move on in oncology to not test for prostrate cancer in PTs after a certain age - because a diagnosis would likely only lead to treatment with the side-effects and suffering, while not affecting lifespan.

I'm just throwing out some thoughts, not implying that you don't already know these things.

But... a really good graph for this topic would indeed be about exercise affecting skinny-fatness. Anybody can refashion themselves, though it is not as easy for some as for others.

Ned Kock said...

I still have my doubts about the ideas that naturally occurring muscle growth promoters, such as endogenous HGH and T, have anything to do with cancer growth.

At post-developmental stages (i.e., in adults) cancer is much more common; but muscle growth is very difficult to achieve, and normally does not occur without some form of muscle damage happening regularly (e.g., weight training).

Moreover, there is a lot of evidence pointing at other factors, of which the main suspects are industrial foods. The mediating effect of the metabolic syndrome is also very strongly suggested by empirical data.

Ned Kock said...

In any case, these ideas are worth exploring, and have been added to my ever growing to-do list.

David Isaak said...

Contrary to popular belief--even among many oncologists--testosterone is not associated with more rapid progression of prostate cancer across the course of the disease.

Castration--chemical or physical--initially slows prostate cancer growth, but it then picks up speed. Higher testosterone results in more rapid initial growth, but this effect then slows.

Castration is popular because it seems to be doing something and slows--for a while--increases in PSA. It is not clear that it alters the overall outcome, and its net effect may even be harmful.

David Isaak said...

BTW, one of the interesting features of your graph is the relative size of the upper left and lower right quadrants.

As you note, there are a lot of people with high body fat who look fine on a BMI basis. But the total number of people who are wrongly classified as overweight by BMI (lower right quad) is about twice as large.

We've always known BMI wasn't a good guide, but these N values give some idea of percentages. Together the upper left and lower right make up about 25% of the US male population.

BMI tells about 17% of guys they are fat when they aren't. It tells about 8% of them they are slim when they are fat.

Ned Kock said...

Very good points David.

Lerner said...

Hi, David. My reflexive thought on seeing your post was to think as usual that there's almost nothing in medicine that's without controversy these days. Since I just now took a quick look around on this topic, I'll post what I ran across.

From 2006 at webmd: "Testosterone Replacement Therapy Doesn't Raise Prostate Cancer Risk, Study Shows"

"...effects on the prostate of testosterone replacement therapy in 40 men. The men were aged 44 to 78 and had low testosterone levels" and "No treatment-related change in the number of cancer cases or cancer severity was found."

http://www.webmd.com/prostate-cancer/news/20061114/testosterone-may-be-safe-for-prostate

[Note that even back then there were ~2 million prescriptions for T Replacement Therapy.]


From April 2011, "The long-standing prohibition against testosterone therapy in men with untreated or low-risk prostate cancer merits reevaluation, according to a new study published in The Journal of Urology."

"This study, involving 13 symptomatic testosterone deficient men who also had untreated prostate cancer, suggests this traditional view is incorrect, and that testosterone treatment in men does not cause rapid growth of prostate cancer. It is the first to directly and rigorously assess changes in the prostate among men with prostate cancer who received testosterone therapy." and "received testosterone therapy ... for a median of 2.5 years" and "none developed cancer progression".

However, I'd say that it's premature to describe the T/prostate connection as a myth. History is full of things like this that didn't pan out. It's not unusual for patients to complain that doctors are slow to adopt the latest great thing. The common reason assumed these days is that if Pharma doesn't make money on something, then it gets ignored. But I'd think that the real reason is that docs have seen many theories come and go through the years. These are only small studies. It might turn out, e.g., that they are only dealing with a subset and a large study would end up differently.

http://www.medicalnewstoday.com/releases/222946.php

[Note FWIW that the study authors have a long list of Pharma disclosures. Also, the study lead has a book out (and wrote an article at LEF) and I'd guess offhand his practice centers on TRT.]


So, from "How Do You Treat Prostate Cancer That Has Progressed On Primary Androgen Deprivation Therapy?" 1998

I see 'In fact, there is significant heterogeneity in the "hormone-refractory" patient population such that some men retain some degree of hormonal sensitivity.' So heterogeneity tends to maybe supprt the idea of patient subsets behaving differently. In fact, next it says, "We now know that an androgen receptor mutation can cause an antiandrogen to paradoxically stimulate tumor growth. Withdrawal of antiandrogen therapy has been shown to result in tumor regression, on average, in approximately 20% of patients."


With all that, I'd guess that new outlook seems possible but is far from proven. No medical body would change practice guidelines on the evidence so far.


Speaking of ideas that didn't pan out, perhaps the biggest was Dr McCully's 'Homocysteine Revolution" (homocysteine causes CVD so reduce it with folate etc). There was a lot to that, from the initial discovery of CAD in very young children, to the drama of McCulley being the victim of oppression by the powers that be and his eventual vindication ~20 years later. The propsed mechanism seemed very sound and there were studies showing association. But then there were actual intervention RCTs done, and oops --- no benefit whatsoever. So homocyteine was dead for years... but oops, now a possible comeback with another Hcy analysis from NHANES and MESA studies. Controversy, as usual, is everywhere.

David Isaak said...

@Lerner--

Yep, controversy is everywhere, but most oncologists seem to automatically assume Test is bad for prostate cancer.

Here's an interesting counterpoint:

"What I believe caused most doctors to believe that TRT is
always harmful is that initially, withdrawing T causes metastatic prostate cancer to go into
remission for well over 90% of all patients. These responses are rapid and dramatic, and since
removing T helps men with metastatic prostate cancer, everyone assumed that adding T would logically make them worse.

However, for men with metastatic disease, the average duration of response to hormone blockade
is only 18-20 months; after that, hormone resistant prostate cancer develops. Therefore, if
patients are treated with continuous hormone blockade, as was first recommended in the early
1940's, the average duration of remission is only 18-20 months. That is not much to brag about.
Treat men with metastatic prostate cancer with the so-called community standard of practice, and
your patient will develop hormone resistance in about one and a half years, and die of metastatic
prostate cancer, on average, one to two years later."

(from Robert Leibowitz, MD, http://compassionateoncology.org/pdfs/TRTHandout.pdf ).

As you note, the story is much more comlex and nuanced than the "testosterone fuels prostate cancer growth" story we have heard so often.

learningat said...

I'll take a moment to thank David for correcting my own initial mis-impression that T is always bad for prostate cancer. My best wishes go out to anybody who is dealing with it.

WRT some things that are good for you but then *might* be bad for cancer:

The first time I saw it was years ago I was reading a CRON group (for the ON, not the CR). Up popped a prostate study that copper was bad.

That same CRON group was all decided to avoid the inflammatory PUFA, Arachadonic Acid, as relates to CVD. I did a websearch and quickly unearthed a bodybuilding site selling AA in tablet form. You do want inflammation after a workout, to be sure. IIRC?, steak has more AA than regular beef and lots more than chicken.

Nowadays, science oriented weightlifter sites will talk about promoting mTOR signaling, yet oncology has developed mTor inhibitor drugs like temsorilimus.

Probably all muscle builders take O-3s, yet here is DHA: "men with the highest blood percentages ... have two-and-a-half-times the risk of developing aggressive, high-grade prostate cancer compared to men with the lowest DHA levels."

http://www.sciencedaily.com/releases/2011/04/110425135643.htm

That's different than:
"Omega-3 Fatty Acids Reduce Risk Of Advanced Prostate Cancer"
http://www.sciencedaily.com/releases/2009/03/090324131444.htm

but maybe it's only the DHA and not EPA that's causing harm.

Lerner said...

oops, that above is me, Lerner. I started a wordpress account because OpenID was acting flakey.

Pearly Home Remedy said...

I think it is a reasonably good guess that these skinny-fat men would not fare very well.

jordan moore said...

skinny fat seems like a low T male who is mostly sedentary.

just do some high intensity cardio, fools.

row, bike, burpee, wallball, pull ups.

just get active and lose the bf.