Sunday, March 18, 2012

Too Good to be True - The "98% Turnover" Dilemma

"If it sounds too good to be true, it probably is" - Anonymous Skeptic

On a gorgeous Spring day, I awoke to find birds singing and the sun shining. After rolling up a window shade and brewing a fresh pot of coffee, I was ready to tackle overnight emails and visit the great Twitter machine.

The first thing I saw?

("Your atoms?" Do I own my atoms? Does anyone? And doesn't that figure seem, well, a little high?)

OK, wait, I'm a scientist, so it's just in my nature to be skeptical. But with 50+ re-tweets, and 18 'favorites' already, I felt this 'fact' needed a bit of perspective. Surely a little digging (assisted by @doctorchemed and @Chemjobber, thanks!) will turn up the truth, right? 

First, let's stop at the likely source for the factoid: this segment from a 2007 All Things Considered episode. It mentions the mysterious "98%" figure, and goes on to give a perfectly rational-sounding explanation, backed up by a scientist (Ph.D. = credibility). The reporters' comments tend towards the usual suspects - oxidative stress, dead skin cells, DNA copying errors. It's well known that several body tissues do indeed cycle through, so they leave another breadcrumb trail for us to follow: "...a study published in the Annual Report for Smithsonian Institution (huh?) in 1953..."

"Mr. Isotope" Seal of Approval?
Source: Fourmilab
Well, thanks to, we can find the "study" fairly quickly. It seems that, in 1953, Paul C. Aebersold (dubbed "Mr. Isotope" by the U.S. Atomic Energy Commission), wrote to the Annual Report of the Board of Regents of the Smithsonian Institution. Given the appearance of annual budgets, secretarial minutes, and auction items, this has to be the AEC's equivalent of a progress report. Consider Aebersold's title: Director, Isotopes Division, Atomic Energy Commission (now part of the NRC). While it's tempting to believe this is a "scientific publication," there's little in the way of hard data, and the text reads more like an historical recap, or perhaps a biography, than an experimental procedure. Read the following quotes, from p.239 of the document:
"We need more people trained in the use of isotopes — people who can apply this new tool to tomorrow's problems in medicine, science, and technology - more 'isotopologists'...We have hardly scratched the possibilities of scientific achievement."

Make sense yet? It's a funding appeal! The speech recounts the glory of the atomic age, from the Curies to the end of WWII, and the line about the "approximately 98 percent of the atoms in us..." (p. 232) serves as another brick in the staircase, built towards the temple of program renewal.

Since Aebersold discloses no study or reference data, my skepticism mounts. If we really recycle almost our whole body, how could we detect blood contaminants, such as PFOA or BPA? How could teeth be stained, if you were always churning out new enamel? (you're not) How would "belly-button microflora" from different places still be detectable on you, if you were always throwing them away?

George de Hevesy
Source: Nobel Foundation
Let's go one more notch back in the literature. George de Hevesy, 1943 Nobelist in Chemistry, developed many of the earliest radiotracers, radioactive compounds used to monitor human, plant, and animal metabolism. Read his Nobel lecture; de Hevesy indicates, with hard data, that work on rabbit skeletal exchange (that is, watching radioactive phosphorus diffuse in and out) resulted in 6.7-29.7% change in various bones over 50 days. So, for certain tissues (the higher values), we could say that, over the course of a year, the 98% value sounds correct. But for the 6.7%? Well, a year is not 746 days long, so that can't possibly apply. Further de Hevesy work utilized labeled calcium ions to show that only 1/3 of the mouse skeleton is replaced during its entire life.

Source: Edupics
Well, old data is exactly that...old. So, you have to flash forward in time to verify conclusions. Here's an NC State Extension primer on osteoporosis, which mentions that approximately 1/5 (20%) of the adult human skeleton is renewed each year. 

So, why all this focus on bones, bones, bones? Turns out, the skeleton in an average human being weighs somewhere around 12-18% of total weight. If 20% is renewed, that means 80% is not. Thus, without considering any other vital body system (neurons? scar tissue? cartilage?) the 98% exchange can't be correct - somewhere around 10% of your body's atoms stay unchanged every year!

(Readers, if you see factual or logical errors here, I welcome future impassioned discussions in the comments!


  1. I don't know if the 98% is bogus or not, but even if it is true, it is very misleading. Most of our bodies are water (75% by weight, I've heard), which certainly has a high turnover. And water has a disproportionately high number of atoms per weight when compared to the dry body tissues. Also, atoms that may be recycled for years and years, such as iron, are pretty heavy (they have a low # atoms per weight. Durable hard tissues contain significant mass fractions of calcium, which is also a relatively heavy atom. Softer tissues with higher turnover are composed of primarily lighter atoms.
    All this said, however, I still don't know whether the 98% figure is an exaggeration or even made up entirely. What's more important, however, is that it is meaningless, whether true or not.

  2. @Bend - Let's just assume, for the moment, that the 75% H2O is correct. If so, that means (roughly) 25% "solid" ingredients, of which 2% would remain constant under Aebersold's calculations. That's an ~11:1 ratio switching out every year (or about 92% turnover). For certain tissues (blood, lymph, epithelium) that certainly fits, but I still don't see that being right for structural tissues.

    1. See Arr Oh, my point is not merely that bodies are made primarily of water that turns over very quickly, but also that the durable tissues that you mention (e.g. teeth, bones) have a disproportionate fraction of heavy atoms relative to soft tissues. Proteins, fats, DNA and carbohydrates (primarily hydrogen, carbon, oxygen, nitrogen) are relatively easy to replace. Skin, the largest organ of the human body, turns over its structural collagen and elastin quickly. 1 year? I don’t know. Bone? Less quickly. Again, I’m not saying that the 98% is correct. As you point out with your historical perspective, it seems like the figure was pulled out of someone’s backside. But facially I don’t think it sounds absolutely unbelievable. Generally, I’m on your side, though. Correct (technically) or incorrect, the statistic is wrong. It’s wrong because it’s misleading and pointless.

  3. I'm glad you tried to figure out the truth. With your data at hand, let's do some fermi calculations. Assuming the human body is made out a theoretical molecule with three atoms with molecular weight 21 (slightly heavier than water). A person weighing 63 kilograms will have 3000 x 6 x 10^23 molecules = 3 x 3000 x 6 x 10^23 atoms.

    Now let's assume, for simplicity, that 20% of that weight is bone mass and also assume that bone is nothing but calcium (atomic weight 40). So about 12 kilograms of calcium which is about 300 x 6 x 10^23 atoms of calcium. If 80% of it remains unchanged every year that os 240 x 6 x 10^23.

    So dividing 240 by 3 x 3000 already gives me 2.7%. In other words, the 98% figure is wrong.

    PS: I am assuming that the argument is well understood here. I've assumed that bone is nothing but calcium atoms. But that's not true, they are made out of cells which have a large number of atoms. So if assuming single heavy atom still means 2.7% of the body atoms remains unchanged, in the real world a lot of the body atoms must remain unchanged per year.

    1. You assume that bone turnover is zero, which it is not. Just because it turns over less than skin doesn’t mean it is constant (if this were so, broken bones would stay broken). Also consider that bone has two primary constituents: 70% hydroxylapatite (a calcium phosphate) and 30% collagen. For every atom of calcium in hydroxylapatite there is approximately one atom of phosphorus, and four atoms of oxygen. The collagen portion is, as are proteins generally, heavy in hydrogen, carbon nitrogen and oxygen. When bone tissue does turn over, the atoms most likely to be recycled to new bone are the calcium but there are actually relatively few calcium atoms in bone. The atoms most likely to be replace by recent comers to the body via digestion are the lighter atoms.

  4. This comment has been removed by a blog administrator.

  5. This comment has been removed by a blog administrator.