This article, published in the Journal of Anatomy four years ago, was also highlighted in ScienceAlert this January 18, which is how Matthew Cobb found it. And although the results aren’t new, I find them interesting from an evolutionary point of view and sure didn’t know about them before. (I’m not sure why ScienceAlert chose to highlight them this week.)
The paper (and the shorter popular summary) describes an Australian study of a variable trait: an extra artery in the forearm and hands of humans called “the median artery”. It is present in fetuses, where it feeds the growing arm and hand, but regresses during development so that it’s not usually present in newborns. However, in a substantial number of cases—now about 30%—it remains as a functioning artery in adults. The paper describes a present study of the incidence of this “vestigial artery” in modern adult Australians, and compares this incidence with that seen in adults going back to the late 19th century. There has been a marked increase in persistence—threefold!—over that period. What we don’t know is why this is happening. It could be strong natural selection, an environmental change we don’t understand, or both.
You can see the paper by clicking on the title below, or download a pdf here.
First, here’s what the artery looks like in an adult (caption from the paper). I’ve put a red oval around the artery:
Now although the artery feeds the arm and hand, we don’t know whether it actually benefits those who have it. The authors and ScienceAlert appear to favor natural selection as the reason for the increase over time, but we don’t know that. To know for sure, we’d have to do long-term studies of the reproductive output of individuals having the artery versus those lacking it, or perhaps genetic studies (see below). We don’t have that data and therefore cannot say anything about natural selection.
Further, perhaps its increased persistence into adulthood is due to some environmental effect. We have no data on that, either. All we can say, and we can’t even say that with a high degree of confidence, is that the percentage of adults having the artery seems to have increased drastically over time.
But I’m getting ahead of myself. The authors dissected 78 arms of Australians aged from 51 to 101 years who died between 2015 and 2016, determining how many of them had the persisting median artery. Individuals were excluded who might have skewed the studies, including individuals with only the hands and not arms examined, people who had carpal tunnel syndrome (possibly caused by persistence of the artery), and examinations using angiography, which has a greater ability to detect arteries. Exactly a third of adults (33.3%) showed the artery.
The authors then went back and scoured the literature, using data on adults from 47 published papers going back to 1897. Using data from that arms in individuals who died at a known age, we have a dataset of individuals born from about 1846 to 1997—a span of roughly 150 years, or about 5 human generations. That’s a remarkably short span of time from an evolutionary viewpoint.
Nevertheless, they found a significant increase over this period of the proportion of individuals having a median artery nearly tripled—from about 10% to 30%. Here’s the most relevant graph plotting the percentage of individuals showing the artery as adults born between 1880 and 2000. (There’s considerable scatter because sample sizes at each date are small.). The authors gives a probability of less than 0.0001 that this temporal trend would be due to chance, so it’s highly statistically significant (they don’t specify whether they’re testing the regression coefficient or the correlation coefficient, but it doesn’t really matter with p values that low.
They also extrapolate this trend and say that one “could predict that the median artery will be present in 100% of individuals born in the year 2100 or later.” It will then no longer be a persisting fetal trait, but a trait that persists throughout life, and the persisting adult trait could no longer be seen as “vestigial”, like persisting wisdom teeth in some people.
The authors do suggest that environmental factors could play a role in this increase, but also that it could be due to natural selection. Such selection, to cause such a strong change in just a few generations, would have to be strong! The ScienceAlert article plays up the selection part, saying this:
“This increase could have resulted from mutations of genes involved in median artery development or health problems in mothers during pregnancy, or both actually,” said Lucas.
We might imagine having a persistent median artery could give dexterous fingers or strong forearms a dependable boost of blood long after we’re born. Yet having one also puts us at a greater risk of carpal tunnel syndrome, an uncomfortable condition that makes us less able to use our hands.
Nailing down the kinds of factors that play a major role in the processes selecting for a persistent median artery will require a lot more sleuthing.
Indeed, a TON of more sleuthing. What would be required to show selection would be either or both of two things:
1.) Show that, over a long period of time, individuals with median arteries as adults leave more offspring than individuals lacking these arteries. This is how the Framingham Heart Study, which began in 1948, showed that there appeared to be natural selection in women for reduced height, increased stoutness, reduced total cholesterol levels, and lower systolic blood pressure. Further, there appears to have been selection for women to produce their first child earlier and to reach menopause later. This is what I tell people who ask me, as they inevitably do when I lecture on human evolution, where our spercies is going. Not that exciting, is it? But of course the time span of such studies are necessarily limited.
2.) Find the genes responsible for the persistence of the artery and show, by population-genetic analysis, that those genes leading to persistence have been undergoing positive selection. This would be even harder because we have no idea what those genes are.
Absent those two types of studies, all we can say is that we have a putative case of evolution occurring over a short period of human evolution.
Caveats: The authors offer these caveats, and I have one more:
Limitations of the present study include the fact that the number of whole cadavers that were available for the study was not adequate. In addition, our search of the literature may have missed some publications not listed in Google Scholar. Finally, the definitions of ‘persistent median artery’ may have differed somewhat among the various published studies included in the present study.
Finally, as far as I can determine from looking at a few of the papers they cite in the older literature, the samples of arms came not just from Australia, but from other countries like Brazil and South Africa. Given that we know that at present populations from different places differ in the persistence of the artery, this could also throw some bias into the data. However, to create a time course this significant, I don’t think that using arms from different places could be the explanation, for it would require that arms from older people tended to come from places which had a lower incidence of the artery in general.
h/t: Matthew Cobb
So great to read and know about, thanks!
Extra blood to the hand – I’m thinking musicians like piano, string, or instruments requiring lots of finger / hand strength…
Is this specifically left? Or does that just happen to be the dissection that they took a photo of?
In living children, the persistent median artery can be detected with Doppler ultrasound without risk of iatrogenic injury…”noninvasively” as we say.
https://pmc.ncbi.nlm.nih.gov/articles/PMC6109622/
(The anomaly occurred in either or both limbs in one fourth of the children. All the children studied were going to have orthopaedic surgery for a variety of upper limb problems, where knowledge of this well-known anomaly would assist surgical planning. In some cases the anomalous artery may be a cause of the condition. So a sample of convenience, not a population-level sample.)
Here you go, a practical way to realize your idea for a nice study if you could get professional virtuoso musicians (or sculptors? or champion video-game competitors?) to lend a few minutes of their time to science. The subjects would have to be reassured that the diagnostic ultrasound technique used in the study carried no risk of injuring their preciously valuable hands! Figure out an appropriate control group, people similarly accomplished in careers not requiring great dexterity such as philosophers or mathematicians maybe, and the paper almost writes itself.
Intellectually, the challenge of defining controls is the most stimulating part of a study like this. A matched control should be as much like the pianist as possible except that he doesn’t play piano. But not her sibling because if the condition is familial then the sib would have a higher than chance likelihood of having it also, obscuring any association with piano virtuosity. Vladimir Horowitz said there are only three kinds of pianists: Jewish pianists, homosexual pianists, and bad pianists. Pick your controls carefully.
Huh, there it is – Thanks!
That graph does not look very convincing. Not far off a swarm of bees.
You are using the eyeball test rather than the statistics that were used, apparently properly. You may not be convinced, but the statistics convince me that there is a trend.
The R^2 value of 0.1173 indicates that ~12% percent of the variance in the prevalence of the median artery is accounted for by the variance in the year of birth. This is a stronger explainer than many similar studies that try to find signal in noise, such as heritability of IQ, etc.
Another caveat in addition to those mentioned elsewhere is that autopsy rates have been declining for many decades and are now done in many locations only if ordered by a coroner or in bodies donated to medical schools, a tiny fraction of all deaths. This introduces bias if for some reason the deceased in a “coroner’s case” was more likely to have a persisting median artery than someone who died of obvious natural causes. This doesn’t seem obviously likely but bias creeps in in the strangest of ways and you won’t look for bias unless you think of it.
Isn’t heritability of IQ (as shown by twin studies, etc) up at about 0.7?
I was thinking more about studies comparing populations of not closely related individuals where heritability is postulated to work despite intragroup variance being large. Even if there is a temporal factor causing median arteries to persist in modern times, you wouldn’t be surprised that this factor alone doesn’t explain more than 12% of the variance in a character that can vary widely for multiple random and unknown reasons. It could still be real but on the scale of a nudge, not a shove.
Comment by Greg Mayer
A few things about the graph: I can’t see why they fitted an x^2 term; it seems superfluous. Second, the last point on the left (late 1870s) is far from the other x-values– it has high leverage, in statistical parlance. It is thus likely to be influential in the estimation of the regression equation. The 4 points to the right may also have high influence. (The increase in the variance of prevalence over time is more striking to me than the increase in mean prevalence over time.)
The regression is the regression, but the R^2 is pretty low, and the estimated regression may depend quite a bit on a just a few points. In their own studies (as opposed to comparison to other studies from the literature), the authors did not find a significant increase in prevalence, so the overall results here are mixed.
GCM
Mortality (especially of infants and kids) was so much higher 100 years ago than now, that it makes me wonder if the relaxation of a past selection pressure against persistence of the median artery is likelier than present-day selection in favor of it. That explanation seems like it might be even harder to test, though.
Thank you for this very clear summary and introduction to this topic. I had no idea there was such an artery in the fetus and that it was usually gone in new borns. This is itself interesting.
One should be scientifically skeptical that something could drive the evolution of the median artery this quickly. Curious — is this something you can self-diagnose or ask nurse/doctor about (or is it really obscure).
Infants show traits that are thought to be from our ape cousins. You should look up lanugo, and also the palmer grasp reflex.
Reminded me of lanugo as well.
Great minds think alike Mark!
D.A.
NYC
On a quick scout of the paper, I do not believe this result. The scatter in the reported prevalence rates is huge, including, to take the example of births around 1960, rates of both ~ 8% and ~60% at the same time. Both of those numbers cannot be sensible.
When the authors analyse only the results of their own observations the trend with time is not significant (ok, small number of statistics perhaps), and it only becomes significant when adding in lots of previous studies.
But, the list of those previous studies (Table 2) does not give error bars on the prevalence values, indeed it doesn’t even give sample sizes (so maybe one estimate came from a sample of 3 people and another from a sample of 300).
The authors make no comment on this, they don’t say whether they have given each of the previous estimates equal weight in the analysis, or, if not, how they have weighted them. They don’t say whether they have evaluated each of those previous studies to assess how reliable the reported prevalence is. In short, there is not enough information presented in the paper to have confidence that this is a real effect.
I don’t have a lot to say about statistics, but I have questions and doubts about this being about natural selection.
I was wondering too if the artery is present in our ape cousins.
Good question about apes.
I always thought evolution and natural selection worked exceedingly slow with randomness as always a factor. Dennett talks, I think persuasively about ”saltations” or adaptions that seem to expedite the process, but this median artery study only dates back to 1846 which in evolutionary terms would be virtually invisible. For my two cents, this random artery supplement analysis is interesting, but statistically insignicant. But what do I know, I’m a visual artist who is inspired by the intersection of science and art. Excuse me now while I renew myself with re reading my beloved Robert Sapolsky books.
No, if selection is strong, as it was in the medium ground finch in the Galapagos, there can be very fast change. Read “The Beak of the Finch” by Jon Weiner to see a 10% change in beak shape in one generation.
I will do that. I appreciate the suggestion.
Thanks.
I presume that these older results are not photographic evidence, which could lead toward investigator bias. If an investigator is not expecting an additional artery, they may overlook it.
There were only a few populations that deviated greatly in the last century and half or so. Anything appearing to change this fast is more likely due to selection bias or is some epigenetic effect.
It can be genetically based and still change fast. Let’s say there is moderate selection – perhaps in utero – against the persistent-artery trait or against something it’s correlated with. Under modern nutrition and medical care, nearly all babies survive, so the trait becomes more prevalent in the adult population. This enhanced survival and relaxed selection should eventually cause changes in gene frequencies, but it can cause observable change in the adult population much faster than that. This is probably true of a lot of human traits; people are surviving much longer with genetically-based conditions than they did a few decades ago (e.g. people with trisomy-21 now have a mean life expectancy of 60 years, compared with 25 years in the 1980s).
That’s a good point.
It seems possible that the ‘reduced height, increased stoutness, reduced total cholesterol levels, and lower systolic blood pressure’ seen in the Framingham Study could be associated with anatomical changes such as the persistence of this fetal artery into adulthood.
Is there any correlation between premature birth and retention of the artery into adulthood?
Count me (a mathematician who has done his share of data analysis) among the skeptics where that graph is concerned.
But I look forward to reading the follow-up paper which compares the DNA of the cohort with this artery against the cohort who lack it.
A couple of clarifications:
The Methods section does specify that the P values are for correlation coefficients (r). They analyze three sets of data. The very low P of 0.0001 is for the all-inclusive Set 1. The figure shown here is for a (more reliable?) subset, Set 2, but the P value for those data is 0.018.
For reasons not specified, quadratic polynomials were fitted to the data instead of illustrating the simple (linear) increase. Higher R^2 values result, but the correlation coefficients that were tested are nowhere reported.
Birthdates were not reported for most of the literature-derived prevalences but had to be “reconstructed”.
The extrapolation to 100% is not based on the polynomials illustrated but just by squinting at the ends of the curves and claiming a tripling from 10 to 30%. Variation aside.
I don’t know. The pattern itself seems pretty weak. As for natural selection, could human reproductive success really depend so strongly on the presence of a superfluous hand artery?
Fascinating. I would definitely be skeptical of the quality of the dissections and data collection—particularly the old dissections. And, the trend seems to be strongly influenced by a few outliers. I call this an interesting result, but I want to see more data.
Call THAT an artery?
THAT’S an artery
— says Australian born me.
Great article.
D.A.
NYC
Very interesting. Thanks for posting.
Perhaps the median artery isn’t directly selected for. Perhaps some other trait/gene is selected for and persistence of the median artery is just a consequence. Does evolutionary theory allow for such a possibility? I really don’t know either way.