Beware of the ancient of days!

By Razib Khan | August 23, 2012 1:29 am

By now you have probably read in The New York Times, or on the blogs, about the new paper in Nature which reports on the empirical trend toward the children of older fathers carrying more de novo mutations. Really all you need is this figure:

It’s probably also useful to remember that you expect 2 more de novo mutations per every year of paternal age, at least across this interval.* Some of the stuff on this weblog is abstruse, but this sure isn’t. As a father I’m to the right of the median of this plot. And let’s move beyond just new mutations arising from your father. What about from your forefathers? My maternal grandfather was 55 years old when my mother was born (he lived from 1896 to 1996!). My paternal grandfather was 38.

Humans of a certain station in life imagine a particular life trajectory. We gain education, establish ourselves in careers, etc. Until recently this long apprenticeship was particular stressful for women because of the “biological clock.” But these results suggest that men too might need to be worried about the long term consequences of their delayed reproduction. I don’t see our society making adjustments to biological reality. Therefore, I think fetal screening, and perhaps a huge industry of sperm and egg preservation is going to have to emerge in the near future. In fact, both are actually rather affordable today.

* There is a lot of variance though. This is all the more reason for high coverage genome sequencing (or at least exome sequencing). Some people could actually make life decisions based on their mutational load, as well as preimplantation screening.

CATEGORIZED UNDER: Personal Genomics

Comments (25)

  1. Jason Malloy

    “and perhaps a huge industry of sperm and egg preservation”

    Right, this is the ideal behavior for every population destined for non-early marriage. Freeze your own gametes in high school. This applies to both men and women: Miscarriage, retardation, and disease is due to the age of the egg, not the age of the woman. A 60 year old woman with a nice young egg will grow the same healthy child as her 20 year old donor.

  2. “Freeze your own gametes in high school.”

    Assuming, of course, that the process of freezing and thawing does not introduce its own mutations. (Has anybody even looked into that?)

  3. AG

    The real challenge is for women’s future orientation. She needs to identify young mate before he become successful. Picking winning stock is hard.

    Do high IQ women have better picking when they are young?

  4. Dm

    In the old days the older fathers tended to be more fit, and their reproductive success extended to their children, so it seems to me that the evolutionary risks may be quite well counterbalanced (the true magnitude of fitness risk associated with older paternal age may be difficult to gauge with the fragmentary data available to date, of course, but I would be surprised if it comes even close to the historical benefits of increased reproductive success of their progeny).

    And another natural mitigating factor could be a higher rate of cuckoldry in the old male – young female couples?

  5. ryan

    #3 – Females don’t want mates with good genes, which are in competition with her genes, after all. They want to find mates with substantial resources, which can help HER genes proliferate. It’s actually better genetically if the mate is an idiot with money.

    I’m only partly joking. Obviously a partner with deleterious genes is a problem. But the big goal is for her offspring to survive in order to pass along her genes, not theirs. A mate with reasonable but not superior genetic fitness who has lucked into a fortune seems mathematically more useful to the goal of proliferating her own genetic material. Her genes do best if the DNA of “more fit” individuals, male or female, meets an unlucky end.

  6. 5.

    #This line of reasoning might be specious. It assumes a lot of things too, which need to be clarified like what ‘an idiot with money’ translates to in a formal sense in the animal kingdom. I assume you mean relatively unfit male with a lot of resources for child rearing, since I am not sure what ‘idiot’ would mean in a non-anthropocentric sense. Why would non-idiots with money be maladaptive for the females? It’s not entirely obvious to me. Maybe someone can explain. In the case of Songbirds, it’s been argued through genetic testing of offsprings that the females cheat on the ‘unfit’ partners by mating with the fitter males in the neighborhood but leaving the child rearing duties to the ‘idiot’ male.

  7. Karl Zimmerman

    Didn’t ancient Greek men typically not get married until age 30? One would assume there would be records of deleterious traits being higher in Greece than elsewhere in the ancient world, but AFAIK this isn’t the case.

  8. pconroy


    My paternal grandfather was 57 yo when my father was born, my father was 35 yowhen I was born.
    My maternal grandfather was 35 when my mother was born.

    I was 48 yo when my most recent daughter was born. However when NYU Hospital enrolled me in a 5-year study of advanced parental age, where they took an embryonic sample and tested for 120 known micro-insertions and micro-deletions and 90 other known conditions. They told me to expect 2 or 3 positives, due to my age. When the results came back, they were negative for everything.

    My second youngest does have some Autism or Asperger symptoms, but they are abating – he also has very advanced vocabulary. My youngest is very advanced for her age, ans seems like she is going to be even smarter than her older academically gifted sister (9 yo).

    So, I’ve either dodged a bullet or there are other unknown low-level stuff lurking out there, that will manifest later?!

  9. zach

    @8 Paul – the chart reinforces that the variance of increased mutational load is high, you could have acquired some beneficial mutation that was not seen by the test (not likely) but the most likely explanation is just plain old regular genetics and environmental interactions: smart parents having smart babies. Intelligent people are already prone to have autistic symptoms and, unlike when you (or even I) were young, this diagnosis didn’t really exist.

  10. Patrick

    Should we necessarily treat these mutations as something to be avoided? Do we need an industry that works to maintain the current version of humans out into the future?

  11. @2 Ahcuah: Existing sperm donation practices result in something like 1/5 birth defect rates; so whatever damage the freezing is causing, if anything, seems to be massively outweighed in at least some respects by the better quality sperm.

    @7 Karl: Given that the Greeks practiced infanticide, had extremely high infant mortality, were frequently (usually?) illiterate, kept few/no surviving systematic records, etc., I have no idea why you would expect Greek info to be at all useful on this topic.

    @10 Patrick: Random mutations are rarely worth keeping, especially when compared against narrowly tailored specific engineered ‘mutations’ as we may be able to do in the future. (Here, let me shoot a bullet into a random spot in your brain – after all, it might kill a tumor!)

  12. Is there not a balancing force here vis-a-vis young breeders, specifically the previous genetic load. You speak of your older ancestors, but in the same time period as the Khan’s have bred 3 generations, has not the 15-year old parents bred more generations?

    A little quick and dirty math suggests that 2 generations of 20 year old fathers would gain us 40+40 de novo mutations, and 1 generation of a 40 year old father would gain us 80. A wash.

  13. gcochran

    Purifying selection is at the same time reducing the number of deleterious mutations every generation: it does more in two generations than in one generation. A higher mutation rate eventually results in a higher genetic load, all else equal.

  14. Douglas Knight

    Given an observed correlation between paternal age and a particular condition in the child, say, autism or schizophrenia, how much credence should we put on the hypothesis that de novo mutations are a significant cause?

  15. pconroy

    12 Chris,

    Good point, unless some specific types of mutations are more common in older males offspring

  16. #14, this isn’t just a hypothesis. the hypothesis was formulated decades ago by haldane. rather, we have some genomic evidence now that de novo mutation rates ARE higher for older fathers. and it’s just getting stronger.

  17. Karl Zimmerman

    11 –

    I suppose my only point was there have been at least some cultures in which delayed paternal age was the norm for hundreds of years, and yet there doesn’t appear to be some culture-wide loss of fitness as a result. Obviously we shouldn’t expect detailed birth statistics for a population where literacy itself was probably limited to the upper social strata.

  18. pconroy

    @17, Karl,

    IIRC, delayed paternal age was a feature of Australian Aborigine societies until recently, where societies were organized into male age bands, with the oldest male band getting the pickings of the youngest women.

  19. Douglas Knight

    Should I have said “mechanism” rather than “hypothesis”?

    There are lots of explanations of why paternal age could have a correlation. How much weight should we put on the particular mechanism of de novo mutations?

  20. particular mechanism of de novo mutations?

    well, i think a molecular biologist would use the term ‘mechanism,’ and make explicit recourse to biophysical processes (e.g., DNA mismatch repair, etc.). to me ‘de novo mutation’ is an abstract phenomena, with contextual, not mechanistic relevance (we’re usually talking about germline). in any case, sperm replicates a lot over your life. so DNA repair is imperfect, and therefore…. that’s the mechanism you’re thinking about.

  21. The overall absolute rate of new mutation caused issues at both ages matters too.

    A doubling of risk when the actual risk is 0.2% for the young, and 0.4% for the old, is a considerably different consideration than it is if the actual risk is 20% for the young and 40% for the old.

    If there is an overwhelming likelihood of no problem in both cases (and most new mutations have no observable phenotype effect of any kind), the obsessing over paternal age may not make sense. On the other hand, if the increased likelihood of new mutations translates into a sufficiently high actual risk for a particular father in a given birth in an absolute sense, when maybe it wouldn’t have before, then it is worth caring about.

    To the extent that conditions like psychosis (bipolar and schizophrenia mostly) and autism spectrum disorders are mostly products of high mutational loads in general, as opposed to mutations in particular genes — in other words, if these conditions are to genetic anomalies what fevers are to infections, a common generalized symptom with an infinite number of particular causes — then genetic testing to determine mutational load at around high school age may make a lot of sense.

    The importance of advanced paternal age is much greater for people who already have a high mutational load than it is for people with low mutational loads when it comes to conditions that seems to be associated with generalized mutational load levels. People who want to act on that information need to know it when they can act on it, i.e. not later than the point at which they reach an age when it could make sense in the larger context to have kids if there was a good reason to start early.

    The study also suggests that someone interested in mapping neurodiversity generally would be well advised to focus on samples of children (or better yet, as Razib notes, people who are both children and grandchildren) of advanced paternal age fathers.

    @1 I recall recently reading a story on assisted reproduction that indicated that while sperm keeps quite well when frozen and is easily transported, that long term storage of eggs doesn’t work as well. In practice, assisted reproduction industry participations have concluded that egg extraction for human assisted reproduction works best when it takes place immediately before the egg is fertilized and happens at the same location where the eggs is fertilized. Egg donors are typically flown in to the place where the extraction and fertilization will take place rather than at a location convenient to them, and then the fertilized egg is transported to whereever it is convenient for the gestational mother. Fertilized eggs are apparently much more amenable to storage and transportation prior to implantation in womb, than unfertilized eggs.

    On the other hand, given that genetic defects are much more closely tied to age for sperm (which are produced anew regularly) than for eggs (which are produced early in life and then sit around waiting to be dispensed in a woman’s body), this inbalance may be tolerable.

    Another implication of the difference in when sperm v. eggs are created by the body is that men are much more vulnerable to passing on germline viruses and to having pre-conception environmental effects influence their children than women. Women are mostly vulnerable to environmental effects that impact their children while they are pregnant, not before becoming pregnant.

  22. IIRC, delayed paternal age was a feature of Australian Aborigine societies until recently, where societies were organized into male age bands, with the oldest male band getting the pickings of the youngest women.

    And how’d that work out for them …. ?

  23. Sandgroper

    They survived very successfully in conditions that would kill you in 3 days max.

  24. Miguel Madeira

    14 – In the specifical cases of the correlation of paternal age with autism and schizophrenia, the data can be confused by the point that is expectable some association of “autistic” or “schizophrenic” genes with extreme introversion, meaning that is expectable that a man with a genetic tendency for autism or schizophrenia will marry and have children later (if he has children at all).

  25. Eurologist

    I should like to point out again – although this is really trivial – that the number of mutations scales approximately linearly with age (especially in the most important range, from a little under 20 to a little over 40). This of course means that for a population at large, there is really no difference in mutational load between young fathers or old fathers. A shorter generation time leads to the same number of mutations over time as a longer one, assuming the specific environment/ economic circumstances allow for the same population number dynamics.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar