I have been waiting to see how long it would take the mainstream press to make the obvious joke about news the New York Times recently reported, that
a protein called osteocalcin, which is produced by bone-forming cells called osteoblasts, binds to a specific receptor on cells of the testes. Male mice that were unable to make osteocalcin (as a result of genetic manipulation) produced less testosterone and were less fertile. When they mated, they had fewer and smaller offspring.
My metaphorical ears perked up when I read the brief description of the research, by Professor Gerard Karsenty of Columbia University’s Medical Center. The full article appeared in Cell in March of this year.
The original NY Times blog post was titled “Examining the Mystery of Skeleton, Sugar and Sex” and then was retitled “Examining Bone’s Role in Fertility” (although my links, regardless of title, call up a post with the original title). Karsenty’s research, as the more complex title suggests, is about how bones interact with other skeletal systems, not just specifically about sexuality.
Bone biology is, for reasons idiosyncratic and motivated, a major example I use in teaching about the flexibility of human sex. One of the first things I do in the semester is elicit from students a list of what they understand to be “markers” of sex: everything from hair style to hormones to skeletal morphology. Then we walk through the list, examining whether they can actually rely on any specific signal to do what in most naive groups they assume it does: divide human beings into two, and only two, mutually exclusive categories.
Students are generally untroubled by arguments that clearly cultural characteristics don’t work particularly well (dress, hairstyle, and the like they conceive of as performative without much urging on my part). Most of my work goes into making them understand that there is not a simple pathway from embryonic chromosomal variability to adult sexual identity. This is where the resistance comes in, and bones play an important role for many of my students: they think they know that the human skeleton comes in two completely distinct variants.
Which is great for me, because I am lucky enough to be in a department with a biological anthropologist who specializes in bone biology, Associate Professor Sabrina Agarwal. Her research looks at variability in bone biology by examining populations from different historical periods. It calls into question some things so widely accepted as facts that the NY Times states them without reservation:
It is well known that the hormones estrogen and testosterone, produced in the ovaries and testes, help to regulate bone growth. When women reach menopause, estrogen levels decrease along with bone mass, putting them at increased risk for osteoporosis. As men age, their testosterone and estrogen levels decline, as well. Men lose bone, but much more slowly than women do.
So, I am in a privileged position: I can bring into my classroom an expert who actually knows how development of the apparently solid and, for many students, determinative skeletal biology is actually dynamic and systemic. As she and Patricia Stuart Macadam wrote in a review of evolutionary biology of osteoporosis
a women’s risk of developing osteoporosis is greatly mediated by factors that are independent of the menopause-induced drop in estrogen levels, such as genetics, nutrition, and physical activity. Pregnancy and lactation can also play potentially important roles in female bone maintenance.
Agarwal’s analyses of medieval British populations are illustrative. Examining a rural population, she concluded in the American Journal of Physical Anthropology that in fact,
Significant age-related changes… were observed to occur primarily by middle age with significant differences between the youngest and two older age groups. Neither sex showed continuing change in trabecular structure between the middle and old age groups. …females showed no statistical differences among the age groups in bone connectivity. … We speculate that while nutritional factors may have initiated some bone loss in both sexes, physical activity could have conserved bone architecture in old age in both sexes, and reproductive factors such as high parity and extended periods of lactation could have played a key role in female bone maintenance in this historic population.
A more recent article, also in the AJPA, adds that in the same population
Females appear to suffer greater bone loss at an earlier age with no change in [bone mineral density] between middle and old age, whereas males show a more steady loss of [bone mineral density] across the age groups.
To rephrase the New York Times generalization to account for these results: in this medieval British rural village,
When women reached menopause, estrogen levels decreased but bone mass did not, having already suffered bone loss at younger ages. As men age, their testosterone and estrogen levels decline, as well. Men lost bone, much more steadily than women did.
Placing these results in context, Agarwal also conducted studies of city residents living at the same time. In an article in press in American Anthropologist, she adds discussion of two populations from medieval London, where she found more familiar patterns that adhere more closely to what today is treated as normal, and thus unavoidable: post-menopausal women suffered the highest levels of osteoporosis. Subsequent studies she and her grad students and postdocs have carried out demonstrate repeatedly that at different times and places, skeletal development may or may not mirror what we have come to take as natural today.
This serendipitous exposure to a body of research that I might not have become familiar with as early or as well if the author were not my colleague shapes my reception of the report on Karsenty’s research. The NY Times blog post presents it in the deterministic mode that has become the dominant way biological findings are represented today. Biology has become, in the mind of the public, fed and reinforced by media, a search for trans-historically true and relatively simple facts that together will account for all the human variability we see today, and can propose based on the material traces of past populations.
This desire for a simple model in which humans are mechanistic products of uniform forces may be comforting. It fulfills a need to be able to control the world, if not in reality (where we know that, in the famous image, a butterfly in the Amazon may cause storms far away, in systems too complex to be reduced to a few variables) then at least in theory. But it is fundamentally untrue.
I want to be clear: I don’t presume that Professor Karsenty or his colleagues are under an illusion that things are simple. Indeed, we can read his work as an attempt to make what was a simple model more appropriately complex; he is quoted as saying
“We thought that if the sex organs talk to the skeleton, then the skeleton should talk back to the sex organs”.
Asking questions like this can be a step along the path to developmental systems theory, which “replaces unhelpful genetic metaphors with a molecular understanding of gene action during the course of development” and “incorporates a concept of experience that goes beyond the traditional equation of experience with learning” so that we understand that “every pattern of behavior has multiple determinants…and the task of developmental analysis is to specify the ways in which the determinants act together in particular cases”.
That Karsenty is thinking in these more interesting ways is suggested by his characterization of the possible implications of the research:
“One idea is that bone might not just be a victim of aging,” he said. “It might also be a contributor.”
But getting readers to see the developmental implications continues to be impeded by the metaphors we resort to that reduce the body to the status of a machine. Dr. William Crowley of Harvard Medical School, while perceptively suggesting the full story of skeletal biology and fertility might be more “complicated”, falls into the trap:
Luteinizing hormone is “the on-off switch” for testosterone, said Dr. Crowley. Osteocalcin, on the other hand, looks more like a “dimmer switch” that modulates the process.
This introduces an unwarranted hierarchy into what actually is beginning to be understood as a complex system. Osteocalcin isn’t subordinate to luteinizing hormone; both act together, and the way they are taken up needs to be thought of as an interaction that takes place simultaneously, not in serial, as the on/off and dimmer switch metaphor suggests.
And in fact, the best demonstration of the failure of that simple, mechanistic metaphor is contained in the description of how osteocalcin works– a description that explains the original, complicated title of the NY Times blog post.
Why was the blog post originally about the mystery of “Skeleton, Sugar and Sex”?
Osteocalcin doesn’t just bind to receptors on male testes and influence testosterone production. Karsenty was actually originally pursuing its role in another bodily process:
osteocalcin boosts insulin production in the pancreas and also increases insulin sensitivity (making the body more responsive to the hormone). Insulin, in turn, acts to lower blood sugar.
No mechanical metaphor can cover these two diverse effects that osteocalcin has. The “dimmer switch” would need to simultaneously raise the thermostat in the house, or maybe initiate the internet connection. This is real human biology, really complicated. Not unknowable, but messy, with lots of particularities that we are only just beginning to understand.
Our understanding starts, of course, by recognizing that things formerly thought to be inert products of hierarchically more important systems are actually actively involved in a network of processes that are remarkably plastic, that give rise to emergent properties, where “biology” cannot be walled off from “experience”.
Let’s give Professor Karsenty the last word here; in a characterization that captures the essence of developmental systems approaches, he added
“The body is not an assembly of silos that don’t speak to each other, but is full of surprising examples of crosstalk.”