By Tim Bates
Tim Bate refutes a criticism of twin studies in an article in Slate here and provides a handy debunking of common assertions made about twin studies below:
Facts and Findings
1. Assertion: “How do we know [that some traits are heritable]? Twin studies.”
Fact-check: Twin studies are not the only source of information about the influence of genes: Adoption studies, pedigree studies, molecular marker studies, to name just three, also provide critical information about the heritable basis of traits. Importantly these designs differ widely in their approach, and yet converge on the same effects of genes and environments.
2. Assertion: “In theory, the siblings in each pair have been raised in the same way.”
Fact-check: In most twin studies, the siblings are raised in the same home: That is what allows researchers to examine the effect of factors such as housing and neighbourhood. However, twin studies also allow us to look at the effect of experiences unique to each twin. These combine to have an effect usually far bigger effect on scores than that of parenting.
3. Claim: Twin studies are “potentially misleading, and based on an antiquated view of genetics.”
Fact-check: The methods of modern twin studies are just that: modern. The advent of structural equation modelling in the 1970s and 80s opened up an era of research with unprecedented capability to test assumptions such as the equal environment, to examine links between developing traits such as reading and mathematics, directly model the interactions of genes with social class, teaching quality, and many dozens of other innovations.
4. Claim: “As a result of mutations during development, about one in 10 human brain cells has more or less than the typical two copies of a chromosome”.
Fact-check: Differentiated neurons are very unusual, in that they have just one copy of each chromosome, rather than the usual two. But this has nothing to do with mutation. If cells gain an entire chromosome in ways that are not programmed, then the effects on the cell of an extra dose of the hundreds of additional gene copies are typically devastating. Down’s Syndrome (an extra copy of chromosome 21) is one of few examples “mild” enough to allow survival.
5. Claim: “Identical twins also have different mitochondrial DNA, the genetic information stored in the cellar organelle responsible for processing glucose”.
Fact-check: Identical and fraternal twins share their similarity on mitochondrial DNA: It is all maternally derived. When mitochondrial DNA plays a role in a trait, we should expect low heritability in twin studies. Again, it works in opposite ways to Palmer’s suggestion. Nick Martin’s team of researchers in Queensland, Australia recently examined the mitochondrial DNA to see if it played a role in IQ. It didn’t.
6. Assertion: “The equal environments assumption is similarly questionable“.
Fact-check: They are not “questionable”, they are testable, and continue to be routinely tested. Some of the best work has been done by Kenneth Kendler, Virginia Commonwealth University Professor of Psychiatric Genetics, and colleagues. These studies compare twins who are more or less often confused with each by their parents and others, who are dressed alike or not, or even who are raised as identical when molecular testing shows they are fraternal twins, with null results for the effects of these confusions. Importantly, the last few years have seen the advent of direct measures of heritability without using any twins: Just direct measures of different gene sharing. One more just came out. Over 3500 people, specifically chosen to be only distantly related, estimating the heritability of IQ at 52%, providing independent validation for twin studies.
7. Assertion: “genetics and environment aren’t separable elements,” and attempting to “discern [the] independent contributions [of genetics and environment] is foolish”.
Fact-check: Often, we are told that this question is like asking ‘which contributes more to the area of a field, its length or its width?’ This apparent inseparability of genes and environment misrepresents what twin researchers are doing.
The field analogy is wrong in two basic ways: Firstly, both length and width measure the same thing. Genes and environments are different kinds of variable. A more appropriate agricultural analogy would be whether the fertiliser or water matter more for a plant. Secondly, we can immediately see the practical importance of studies of the effect of fertiliser or water on agricultural productivity. Imagine if a farmer found out that nearly all the differences in whether his crops grew or not were due to water, and not fertiliser, and that with a low-cost change to watering, he could increase his yields tenfold. Would he laugh, say ‘fertiliser and water are inseparable’, and continue doing nothing? Would you? Finding that events have more than one influence doesn’t render the situation unanalyzable, any more than we have to conclude that one and only of the influences is the one that “really” matters.
8. Claim: “Twin studies also rely on the false assumption that genetics are constant throughout one’s lifetime” and “[b]y the time a pair of twins reaches middle age, it’s very difficult to make any assumptions whatsoever about the similarity of their genes.”
Fact-check: We appear to moving from farce to fantasy. Here’s a reference to a lovely study looking at whether or not genes involved in personality are the same at ages 11, 12, and 16. Fascinating recent work studying twins across the lifecourse suggests that the genes involved in adolescent depression differ greatly from those affecting depression in 50 year olds. That’s an incredibly interesting finding: It means that depression in an 18-year old and 50 year old’s depression are, biologically, different diseases. But according to Palmer, researchers cannot ever have even asked such questions… It is perhaps worth stating too that these new effects of genes throughout development are not “mutations” as Palmer claimed. How can we tell? Because just as an MZ twin’s level of mood at age 15 is an indicator of their co-twin’s likelihood of a mood disorder, that same relationship holds in their 30s and 50s. So the DNA is largely constant, but its effects may (for some traits) vary across the lifespan.
9. Claim: “exercise, which can accelerate the formation of new neurons and potentially increase genetic variation among individual brain cells”.
Fact-check: In the adult human brain there are just two tiny areas where new neurons have ever been observed growing: the olfactory bulb, and, (possibly) one thin layer of the hippocampus. Critically, cortical neurons don’t appear to grow in the adult brain – as demonstrated in a wonderfully lucid study by Bhardwaj in PNAS. In a psychological sense the mind is massively plastic: You’re changing its state right now. And biologically, as you change your mind, the membranes, synapses, transmitters of the brain are similarly plastic. But at the level of new neurons, we are sadly running on a supply largely fixed shortly after birth.