Genetic amplification versus G-E correlations

One curious aspect of IQ is that the heritability of it increases with age [3]. Given the nature of the increase, there are three plausible explanations. The increasing heritability is either a function of: 1) active GE correlations (rGE), 2) genetic amplification, or 3) novel gene expression through the developmental period. As it is, the current evidence suggests that the third possibility does not play a prominent role. Longitudinal studies show that genes contribute to the continuity of IQ across the development period [6,7]. Figure 3.4, graphically depicts the possible relations between genes and IQ continuity; model C corresponds to the data. Basically, the genes that lead IQ differences in infancy are the same genes that lead to differences in adulthood. Were novel genes to come into play, this would not be the case.

Environmentalists, naturally, have seized on the active rGE explanation. Accordingly, genetic differences in intelligence result from cognitive shaping environments, environments which individuals select on the basis of their genetically predisposed predilections [5]. By this model, the heritability of intelligence increases with age because with age individuals become freer to follow their dispositions and select their environments. Quasi-environmentalists are not the only ones that find this model appealing. For example, in the discussion section of their ground breaking study of 11,000 twin pairs across four counties, Haworth et al (2009) state [14]:

Why, despite life’s ‘slings and arrows of outrageous fortune’, do genetically driven differences increasingly account for differences in general cognitive ability during the school years? It is possible that heritability increases as more genes come into play as the brain undergoes its major transitions from infancy to childhood and again during adolescence. However, longitudinal genetic research indicates that genes largely contribute to continuity rather than change in g during the school years. We suggest that the developmental increase in the heritability of g lies with genotype–environment correlation: as children grow up, they increasingly select, modify and even create their own experiences in part on the basis of their genetic propensities. This leads to an active view of experiences relevant to cognitive development, including educational experiences, in which children make their own environments that not only reflect but also accentuate their genetic differences.

Oddly, they fail to mention the well known third alternative.

The reason that (the now gene-)environmentalists have embraced this model is fairly obvious: causally, it’s an environmental account of IQ differences; it just has a genetic veneer. For behavior geneticists, it offers more politically palatable interpretation to their findings.

Initially, the rGE explanation seems reasonable; when it comes to individuals differences, it doesn’t seem outlandish to suppose that naturally born intellectuals might increase their verbal IQ through bookish behavior. On the subpopulation level, likewise, it’s doesn’t seem implausible that a propensity for studiousness might lead to cognitive enhancement. Yet, there are a few bumps which preclude a simple rGE model. For one, it’s not merely this or that measure of intelligence that increases but, rather, the central factor and its numerous correlates [6]. In effect, rGE theorists are forced to maintain that g is created from the outside in. (See theoretical diagram below). Since g is structurally the same across individuals, cultures, sexes, and subpopulations, not only would the patterns of one’s environment have to construct g, the patterns of everyone’s environment would have to construct the same g.

Additionally, IQ g has numerous endophenotypic correlates, such as the volume of white and grey matter, the mass of the prefrontal lobe, and total brain size [1, 11, 13] and the overlap between IQ (g) and many of these endophenotypes is entirely due to genetic influences [1, 13]. To explain this genetic covariation, rGE theorists must maintain that genetics sets the parameters for environmental selection, which leads to the development of different cognitive phenotypes, which, in turn, molds the endophenotypic differences, thus creating the three way correlation. Since the Phenotypic/endophenotypic correlations have been found to be a function of differential rates of change during the development process [9] (see figure 2), rGE theorists must maintain that this environmentally induced endophenotypic molding occurs primarily during the developmental process and starts early on. If we kept in mind what we said above, that the genes that lead to slight genetic IQ differences in infancy are the same genes that lead to large genetic based differences in adulthood, and note that the heritability of many dispositions also increases with age [4], we can readily identify the problem with this conception. Somehow, dispositional differences, which are under heavy environmental influence early on, must set the phenotypic/endophenotypic molding (environmental) parameters in a way that happens to correspond to the genetic driven phenotype that the individual will later express.

None of the above logically precludes a rGE explanation; the explanation would just have to be exceedingly complex. That said, there is a growing body of evidence against the active rGE model. According to it, as environmentally conditioned phenotypic differences cause endophenotypic differences, environments will correlate with endophenotyes [13]. This was found to not be the case by Posthuma et al. (2003), van Leeuwen et al (2009), and Betjeman (2009), disconfirming the model.

In addition to the above, Shikishima et al. (2009) found correlational evidence of a causal genetic g. This effectively rules out the possibility of a purely active rGE created g:

Accordingly, our findings could furnish an argument against the typical criticisms offered by those who are opposed to the concept of g; in other words, g is an “artifact” (Simon, 1969) of the statistical methods that psychologists apply to the data. Gould (1981) argued that g, as a factor extracted from the factor analysis, is neither a “thing with physical reality” nor a “causal entity”, but is a “mathematical abstraction”, maintaining that “we cannot reify g as a ‘thing’ unless we have convincing, independent information beyond the fact of correlation itself.” Although the present study also draws information from correlations, we were able to depict the structure of human intelligence beyond the fact of phenotypic and genetic correlations with an explicit comparison between the independent pathway and the common pathway model; and as a “causal entity”, as a highly genetically driven entity…

…Several recent reports have shown that g is also correlated with a variety of neural mechanisms, such as glucose metabolism (Haier, 2003), cortical development (Shaw et al., 2006), and biochemical activity (Jung et al., 2005), along with the identification of promising endophenotypes for intelligence such as working memory and processing speed (van Leeuwen, van den Berg, Hoekstra, & Boomsma, 2007). These studies allow us to assume that it is now reasonable to consider g to be a physiological or biological, genetic entity.

The alternative to the rGE model is the readily falsifiable genetic amplification model. Plomin (1987) summarizes it thusly:

Genes that affect IQ make only a small contribution to phenotypic variance at first, but their effects are amplified throughout development. Suppose, for example, that genetic differences among infants are responsible for differences among them in the formation of dendritic spines during the first few years of life and that the complexity of dendritic spines is related to information processing capabilities. At first, these structural differences do not have a chance to cause function differences because so little information has been processed at this point. Gradually, the functional differences are amplified as more and more information is processed by children. If we were to measure differences in categorizing ability early in childhood, the genetic differences among children due to the complexity of dendritic spines would contribute a negligible amount of variance to observed variability among children in categorization ability. The differences snowball as development proceeds, so that a study of the children when they are older will show more genetic variance. Yet the genetic correlation between the two ages is near unity because the genetic portion of observed variability at both ages originates with the same set of genes whose effect become amplified during development.

The model predicts that the genetic correlations from childhood to adulthood, as inferred by parental-offspring correlations and seen in longitude twin data, will continue to be high even as heritability increases [7]. As mentioned above, this has turned out to be the case. The model also predicts that genetic differences which cause small differences early on will cause larger differences latter on. Boomsma (1998) found confirmation of this: differences in intelligence between 5 and 7 year olds were due to the same genes; while at 5 those genetic differences resulted in minor IQ differences, at 7 they resulted in a much larger differences. Other studies have found that genetic amplification explains increased heritability in the samples studied [16].

In light of the evidence to date, a substantial gene-environmental explanation for the increasing heritability of general intelligence and with it individual differences is improbable; alternatively, a genetic amplification model is probable. When it comes to that oft discussed subpopulation difference, from a non-hereditarian standpoint, the improbability of the rGE explanation leaves one with an equally improbable environmental alternative.

References

[1] Betjemann, et al., 2009. Genetic Covariation Between Brain Volumes and IQ
[2] Boomsma, et al., 1998. Genetic influences on childhood IQ in 5- and 7-year-old Dutch twins
[3] Bouchard, 2009. Genetic influence on human intelligence (Spearman’s g): How much?
[4] Gardner, 2007. A meta-analysis of age-related changes in heritability of behavioral phenotypes over adolescence and young adulthood.
[5] Flynn, 2007. What is intelligence? Beyond the Flynn effect.
[6] Haworth, 2009. The heritability of general cognitive ability increases linearly from childhood to young adulthood
[7] Plomin, 1987. Development, genetics, and psychology.
[8] Posthuma, et al., 2003. Brain volumes and the WAIS-III dimensions of verbal comprehension, working memory, perceptual organization, and processing speed.
[9] Shaw et al., 2006. Intellectual ability and cortical development in children and adolescent
[10] Shikishima, et al., 2009. Is g an entity? A Japanese twin study using syllogisms and intelligence tests.
[11] Smit et al., 2010. Endophenotypes in a Dynamically Connected Brain.
[12] van Leeuwen, 2008. A twin-family study of general IQ
[13] van Leeuwen et al., 2009. A genetic analysis of brain volumes and IQ in children.
[14] Jensen, 1973, Educatability and group differences.
[15] Haworth, et al., 2009. The heritability of general cognitive ability increases linearly from childhood to young adulthood
[16] Soelen, 2011. Heritability of Verbal and Performance Intelligence in a Pediatric Longitudinal Sample

Advertisements
This entry was posted in Uncategorized. Bookmark the permalink.

4 Responses to Genetic amplification versus G-E correlations

  1. statsquatch says:

    Thanks for this post. Why is this the case, “The GE model predicts that the correlation between the average phenotypic difference between twins |t1-t2| and the average of the phenotype (t1 + t2)/2 will be significantly different from zero [14]. ” I think it is going to be hard to banish the GXE interaction. You have indentified the politics around it.

    • Chuck says:

      This article has a nice summary of the method:

      “Detecting Genotype–Environment Interaction in Monozygotic Twin Data: Comparing the Jinks and Fulker Test and a New Test Based on Marginal Maximum Likelihood Estimation”
      http://www.vipbg.vcu.edu/vipbg/Articles/trhg-2006-gxe.pdf

      As for GE interactions, given the finding to date, it’s unlikely that they explain much of the genetic component of intelligence. Maybe a more sensitive model of detection will show that GE interactions account for 10% of the genetic component of IQ, but that’s a far cry from what environmentalists need. (Incidentally, all the studies listed above did show a consistently negative (though statistically non-significant) correlation — so there likely is some GE, just not a very potent one.

      The point of this post — and the “is the sociologist’s constant still constant” one was to follow up on my assessment of the between-within argument.

    • Chuck says:

      Speaking of which, Jensen made two within-between arguments. The second was from twin data. You wouldn’t know the current absolute twin difference?
      ………………………..
      Ok, I got it. Here’s the twin version of the between within argument:

      To say that the B and W subpopulations are genetically identical, is to say that these subpopulations are equivalent to being sets of MZ twins. As we said, the total average difference between twins is ~6.5 points (as compared to ~17 for any random pair of individuals). Given a total average difference of 6.5 points, the standard deviation between twins is about 5 points (1). Since the B-W difference is 15 points (2), there must be 3 SD (15/5) of environmental effect between them, which is to say that Bs must live in a cognitive environment equivalent to the Ws 0.14 percentile. Which they don’t.

      (1) Jensen (1973) gives the twin standard deviation formula:

      The MZ Twin SD (4.74) is equal to sqrt of the environmental variance [sqrt (22.5)]. The environmental variance (22.5) is equal to the total variance (15^2) minus the genetic variance (.85 x 15^2) and the error variance (-1.95(15^2). The genetic variance is the twin correlation X the population SD (15). The twin correlation (.85) is r = 1-(|dk|/|dp|)^2, where
      |dk| is the mean absolute difference between kin
      |dp| is the mean absolute difference between all random people
      and
      |dp|=2pi/sqrt(pi) = 1.13SD

      (2) My .9SD across the age spectrum would give us 2.7SD of needed environmental effect circa 2008.

  2. statsquatch says:

    Thanks, the source is from Amsterdam again. Jensen’s indirect effect brought us this far but what will bring us over the finish line?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s