Friday, August 21, 2015

A New Kind of Twin Study and the Heritability Fraud

In my post on Psychology Today, "Scientific Fraud in the Nature versus Nurture Debate," I discussed the disturbing tendency of psychiatric researchers to use the term heritability as a synonym for genetic, which it certainly is not. The heritability statistic is a measure of phenotype, not genotype, meaning it is a measure of the final outcome of the influence of the interactions between genes and the environment on such things as certain personality characteristics or psychiatric symptoms.

The statistic is derived from twin studies in which fraternal and identical twins who were raised together are compared to each other and to those raised apart on various traits. It is not a measure of purely genetic influences but instead a measure of a mix of purely genetic influences plus gene-environment interactional influences. 

There is no way to tell how much of each is present in the statistic. The determination of heritability can also be manipulated in a number of ways, such as by setting the bar for saying that a symptom is present or absent at different levels.

Interestingly, a recent study employing a very different type of twin study has been getting a fair amount of press (Thalia C. Eley, Tom A. McAdams, Fruhling V. Rijsdijk, et. al., "The Intergenerational Transmission of Anxiety: A Children-of-Twins Study," American Journal of Psychiatry, 172 [7], pp. 630-637, 2015).

Rather than comparing twins with each other, the authors compared the children of twins with one another. The subjects were anxiety and a dimension of normal personality known as neuroticism - a measure of emotional reactivity. People with higher neuroticism scores tend to get more anxious and/or depressed in reaction to negative environmental stimuli, and remain dysregulated longer, than those with lower scores.

By comparing the extent to which correlations between children and their twin uncle/aunt (avuncular correlations) differ for monozygotic (identical) and dizygotic (fraternal) twin families, the authors were able to infer the extent to which genetic and environmental factors influence transmission from one generation to another. Children share a greater level of genetic influence with their uncle/aunt when in monozygotic families than when in dizygotic families.

Thus, if children resemble their uncle/aunt to a greater extent in monozygotic families than in dizygotic families, this implies a genetic influence on transmission of the trait of interest. In contrast, if these two sets of correlations are similar, and are significantly lower than the parent-child correlations, this is indicative of an environmental mode of transmission.

The results of the new study showed almost the opposite of the usual results of heritability studies on neuroticism: environmental factors came out very much more important than genetic ones! Living with one's parents was found to be far more influential than merely inheriting 50% of their genes.

It appeared that children and adolescents learned anxious behavior from their parents rather than inheriting a tendency towards it from their parents genetically.

Now, I must say that the authors used a statistical technique to come to their conclusion called "structural equation modeling"—of which I know absolutely nothing. So I am not able to say if the methodological techniques used in traditional twin studies yield more accurate results than those found in this type of study. This may, in fact, be a case of scientists being able to get the results they want to get through statistical manipulation of their study data.

And surely neuroticism must have some significant genetic component. Clearly, some people are naturally more high strung than others.

Nonetheless, I do know from the observation of blatantly obvious behavioral patterns within families and other social groups that anxiety can be highly contagious. Since as of now mental health professionals can't fix your genes but we can fix your relationships, I know on which factors therapists should focus the majority of their attention.


  1. From your statement of principles:

    ...and desperate parents who want to believe that their children have a brain disease to avoid an overwhelming sense of guilt.

    Arguing in favor of genetics over heritable behaviors also helps confused and immobile parents avoid that guilt. In my family, every problem behavior ever performed has been put down to genes, "I guess we're just genetically disposed to rage, depression, horrible self images, guilt, workaholism, drug abuse..."

  2. I think you'd appreciate this article:

    It even lets you have a tool for hacking the "p-value" until you get a "significant result" (p < 0.05) to bias the results anyway you want, whether for the Republicans or Democrats.

    It's pretty telling, showing just how tough science can be.