Monday, January 23, 2017

Why does the human population carry an allele that increases the risk of Alzheimer's?

The human apolipoprotein E gene (ApoE) has several alleles segregating in the human population. One of them, E4, is associated with increased risk of Alzheimer's. Ed Yong, writing for The Atlantic, asks "Why Do Humans Still Have a Gene That Increases the Risk of Alzheimer's?

I can think of several answers off the top of my head. The most important one is that Alzheimer's has very little effect on your ability to have children. The disease may not even have developed in most of our ancestors who tended to die younger. In order to be subject to negative selection the allele has to affect adults before they reproduce.

The second reason is that the slight deleterious effect, if there is one from an evolution perspective, may not have been significant enough in small populations. I know, and I hope my students know, that neutral and deleterious alleles can reach significant frequency in a population by chance. The general public doesn't know this.

Check out Ed Yong's article to see his explanation.
“It doesn’t make sense,” says Ben Trumble, from Arizona State University. “You’d have thought that natural selection would have weeded out ApoE4 a long time ago. The fact that we have it at all is a little bizarre.”


  1. There is a http too many in the link to the aricle you gave.

  2. Correct me if I'm wrong, but it seems as if Dr. Trumble mistakenly believes that phenotypes, like Alzheimer's, are ALWAYS governed by natural selection. Consequently, his "hammer" is always looking for the "selective nail." I, too, think this is a common mistake.

    As Dr. Moran has pointed out, population genetics is based on the premise that alleles often appear within a population, via random mutations. The resulting phenotypes:

    1) may be selected for or against - via natural selection (the famous and cool stuff)


    2) phenotypes (traits), in general, may not always confer an advantage or disadvantage to the this case, any reproductive success is random, and completely NEUTRAL, and due to non-selective forces (bottleneck effect or improbabilities associated with small populations). This is sometimes referred to as genetic drift.

    In addition, as Dr. Moran has pointed out, the age that most people succumb to Alzheimer's is well beyond our reproductive years. Based on what we know about the disease, it has no effect on biological fitness (the ability to survive and reproduce).

    I don't have a complete grasp on all of the mathematical complexities of population genetics, but it seems as if this study can be understood within one of the more basic, and demonstrable, principles of modern genetics and evolutionary theory.

  3. A little off thread. I question that this disease comes from genes like presented. I see that old age is the dominant point. if we never became older then 45 we never would have this disease in any important numbers.
    So e4 might rather be just affected by our decaying bodies. nOt that its a trigger in any way to this problem.
    Further i strongly suspect this disease is a memory problem completely. So e4 would only be affecting the memory as such.
    Its the aging issue that to me says its not a pure malfunction independent of aging.
    Everyone would get this if we lived long enough surely.
    So evolutionary selection would have no effect.
    Evolution can't get rid of us who decay faster. Reproduction happens early.

  4. Of course it's always possible that the allele provides some advantage or other that makes up for any deleterious effects. No idea.

    1. Yes, as Gould and Lewontin point out, that's the fall-back position of the adaptationist program. Rather than consider the null hypothesis there's always a way to rationalize selection.

    2. I agree with Larry that selection to avoid Alzheimer's disease will be weak, especially since basically all reproduction occurs before there is much risk of Alzheimer's disease. And yes, effective population sizes in the human lineage were small. Note, though, that E4 is not fixed in humans.

      In fact, studies of polymorphism of Apoliliprotein loci in apes show (McIntosh et al., PLoS One, 2012) that apes are fixed for sequences related to the E4 allele. That at least raises the possibility that it is the other alleles that have been favored in humans. Larry should be interested in this issue.

    3. Human societies/populations and their survival during evolutionary time spans is not a subject we know everything about quite yet. I can fault the article for not mentioning the default explanation (as long as we can tell the allele involved isn't subjected to strong negative or positive selection - anyone know?), but not for following scientists and their curiosity regarding more or less plausible adaptationist explanations. I thought the research was interesting.

    4. Thanks, Joe re fixation - I hadn't seen your comment while I was writing mine.

    5. The observation that E4 is fixed in apes, but not in humans, implies (in the adaptationist perspective at any rate,) that E4 is undergoing negative selection is directly opposed to the Yong article. The thing is, the question of whether this negative selection pressure actually exists at all is still left unexamined. Both positions, E4 is adaptive and E4 isn't adaptive but a mere vestige, are equally unsupported.

    6. In fact, studies of polymorphism of Apoliliprotein loci in apes show (McIntosh et al., PLoS One, 2012) that apes are fixed for sequences related to the E4 allele.

      When you say "apes" are you referring to Pan troglodytes only, or is there some other paper I haven't found that assays lots of species?

    7. @John: I overstated. The McIntosh et al. (2012) paper shows all chimps in their sample having an E4-like allele. For other apes it shows an allele like that but the samples are too small to assess whether these are fixed.

  5. Interesting. Perhaps parasite infections could ameliorate some of the symptoms seen in Alzheimer's patients?

    Just wondering out loud ...

  6. If you click on the link, you can go down the page to find a more rewarding article on M. Night Shyamalan's new movie, Split. The science isn't much worse.

    Adaptationism has the great advantage of adhering to the common observation that "God don't make no junk." That includes DNA as well as children apparently.

    Yong conspicuously omits to ask about the evidence showing sharper wits in old people is beneficial to their offspring's reproductive success in a technologically primitive society.

    1. Adaptationism has the great advantage of adhering to the common observation that "God don't make no junk."

      Then how do we account for creationist arguments like yours?

    2. You do not quote a creationist argument. Perfection of design by natural selection is not really distinguishable from perfection of design from a Creator, which means it's those who presume perfection who are making the creationist arguments. You have everything upside down.

    3. Evolutionary biology does not predict perfect adaptation. It is the imperfections that provide one source of evidence that evolution produced the adaptations. On your side the argument is obviously for perfection since "God don't make no junk".

    4. Your positition seems to be either E4 is an adaptation to brain parasites, caused by natural selection for the reproductive advantage of neurologically healthy parents and grand-parents to the offspring *or* for its opposite, natural selection in humans for the alleles that are replacing this vestige of ape neurology. Perhaps this is not literally perfect, but your position is that particular DNA is not junk. This seems to me to be wholly unfounded. Natural selection is not powerful enough to optimize traits with marginal or variable effects on reproductive success in the face of mutation, drift, etc.

      Therefore I am unrepentant about equating adaptationists of Yong's sort with creationists, which is the topic at hand after all. Your version where natural selection hasn't finished perfecting apolipoprotein in the human brain, sorry, no, it's a possibility. But there's no evidence for it that I know of.

  7. In order to be subject to negative selection the allele has to affect adults before they reproduce.

    Surely you mean before they cease to reproduce?

  8. On the other hand...

    Human-specific derived alleles of CD33 and other genes protect against postreproductive cognitive decline, Schwarz et al, 2015,


    The individuals of most vertebrate species die when they can no longer reproduce. Humans are a rare exception, having evolved a prolonged postreproductive lifespan. Elders contribute to cooperative offspring care, assist in foraging, and communicate important ecological and cultural knowledge, increasing the survival of younger individuals. Age-related deterioration of cognitive capacity in humans compromises these benefits and also burdens the group with socially costly members. We investigated the contribution of the immunoregulatory receptor CD33 to a uniquely human postreproductive disease, Alzheimer’s dementia. Surprisingly, even though selection at advanced age is expected to be weak, a CD33 allele protective against Alzheimer’s disease is derived and unique to humans and favors a functional molecular state of CD33 resembling that of the chimpanzee. Thus, derived alleles may be compensatory and restore interactions altered as a consequence of human-specific brain evolution. We found several other examples of derived alleles at other human loci that protect against age-related cognitive deterioration arising from neurodegenerative disease or cerebrovascular insufficiency. Selection by inclusive fitness may be strong enough to favor alleles protecting specifically against cognitive decline in postreproductive humans. Such selection would operate by maximizing the contributions of postreproductive individuals to the fitness of younger kin.

    1. This is called a just-so story. The idea that there's been selection for elders in a society because they contribute to overall fitness seems superficially attractive but there's no evidence that it's true. Even in historical times the number of individuals over the age of 65 was a small percentage of the population.

      If such a selective advantage had been a real force in human evolution then why hasn't it eliminated cancer, cardiovascular disease, arthritis, Parkinson's, and dementia?

      It seems just as reasonable to me to assume that the CD33 allele arose by chance and it's effects don't make a significant difference to fitness. There's no reason to make up a just-so story, especially one that makes so little sense in light of all the imperfections in modern humans. We certainly don't look to me like a species that has been honed to optimal fitness by subtle effects such as the ones described in the abstract.

  9. The first author of the study published 16 papers (Pubmed) with corelational data on the 'Amazonian forager-horticulturalists' (published in between 2012 and 2016) with 8 as a first author). Am I jealous?

    The correlations tested involve (taken from the titles):

    -parasite burden, ApoE alleles and cognition
    -parasite infection with cardiovascular disease and type 2 diabetes
    -parasite infection with Depression as sickness behavior
    -immune function (largely parasite related) and age-related differences forager-horticulturalists, US and Europeans
    -parasite burden and metabolism rates

    -testosterone & parasite infection with immune function
    -testosterone levels with physical competition
    -testosterone levels (and cortisol) with successful hunting
    -testosterone levels (age dependent) with horticultural activity
    -testosterone levels (male) with cognitive performance
    -testosterone and oxitocin levels when hunters come home

    -Cortisol with political influence

    -fertility and mineral density among adult women
    -fertility and parasite infection

    -No Sex or Age Difference in Dead-Reckoning Ability among Tsimane Forager-Horticulturalists.
    -An epigenetic clock analysis of race/ethnicity, sex, and coronary heart disease.

    I had a look in the paper and I don't understand the method they used to compare the cognitive capabilities of the groups tested. Maybe someone with access & more statistical knowledge can help me out? It might turn out to be an interesting lesson in statistics & correlations.

    In Figure 2 'z Scan' they plot the z-scores for the cognitive tests for the different genotypes against the parasitic load (measured by Eosinophil count;) no data point are given.
    I don't understand how the z-scores (?) are calculated and how the performance in the tests is actually calculated.

    The naive eye would interpret the graph as follows: having the E4 allele your cognition gets better and better the more parasites you have.

    1. A z-score is a measure of how many standard deviations away from the arithmetic mean (average) an observation is. (Observed value minus mean value) divided by the standard deviation is the z-score.


    2. Thanks jaxkayaker.
      Actually I know what a z-score is.
      But I have no Idea, how they actually calculated their mean(s). For each Eosinophil count a new one or a global one?

      Best Michael
      Thx again,

  10. “ “You’d have thought that natural selection would have weeded out ApoE4 a long time ago. The fact that we have it at all is a little bizarre.” “

    Why does he say we all have it?

    Common Alzheimer’s gene APOE4 raises risk for women but not men
    “APOE4 is the riskiest known Alzheimer’s gene variant…It’s riskiest because it is so common. As many as 1 in 5 people possess APOE4……….Researchers don’t yet know exactly how the allele does its work, but a 2012 review declared that a person with one copy of APOE4 has three times the risk of LOAD as a noncarrier. Two copies increase the risk by about 12 times.”

    1. It doesn't say that we all have it. Check the sentence: "The fact that we have it at all" does not mean "the fact that we all have it."

    2. He doesn't, at least not in that sentence. He is referring to the species, not every individual.

    3. Ah, after this elementary reading comprehension fail, some things begin to become clearer.

    4. Scan error. I beg your pardon. Thanks for the correction.

    5. I used to work in a research project on cardiovascular disease using ApoE3 knock out mice. If you fed the mice normal rodent chow, cholesterol and fat contents of blood remained at normal levels. If you started to feed them 'western type foods' cholesterol and fat in blood would sky rocket. In the long run mice most mice would get cardiovascular issues related to the high fat/ cholesterol.

      This sounds to me to be a combination of environment and prevalence of the gene(s) which lead to Alzheimer. Until about 30 years ago, most people rarely ate modern 'western style diets' ie high fat, sugar. And people are getting older due to better health care.

      But I have to agree, these diseases occur years, even decades, after people reproduce. And don't forget ApoE has an important function in normal catabolism.

      It seems ApoE4 is most common in Nigeria, but since few people there have access to western style food, Alzheimers is rare.

    6. @Ed says,

      Until about thirty years ago, most people rarely ate modern 'western style diets.'"

      Thirty years ago is 1987. I don't know what you mean by "Western type" foods but my eating habits haven't changed very much since I was a teenager in the 1960s. I don't see any big change that occurred in the late 1980s.

  11. No one mentions possible pleiotropy or linkage?

    1. Pleiotropy is a good one, was thinking about that too. If it deleterious but shows up after reproduction, it wouldn't be subjected to negative selection hence was not weeded out.

  12. There are several genetic diseases (syndromes) that are not acted on to any great extent by selection. Marfans runs in my family and I have the genetic markers. It is a connective tissue disorder who's most lethal symptoms are heart valve problems and aneurisms. But, since they don't usually present themselves until the person is in their 50s, it doesn't prevent me from reproducing.

  13. Here is access to an oldie but goodie by George C. Williams on the general situation of the evolution of senescence. Several of the issues raised seem relevant.