Gabor Maté - In the Realm of Hungry Ghosts - Close Encounters with Addiction

Adoption and Twin Study Fallacies

The weighted emphasis on genetic causation in medical literature, particularly when it comes to mental dysfunctions and addictions, is astonishing given the shaky logic on which the supporting studies are based. As one review stated:

A critical analysis of the assumptions of any adoption or twin study, coupled with the succession of retractions of the genetic linkage studies indicates that the evidence for the genetic basis of mental illnesses is far from overwhelming. 1

The two assumptions on which the heavily gene-based estimates in addiction medicine rely are not sustainable if we examine them closely. They are:

1. that studies of adopted children can distinguish genetic from environmental effects

2. that we can separate out genetic from environmental effects by looking at the similarities and differences between identical twins on the one hand, and fraternal twins on the other

A prominent researcher in the field of mental illness, including addiction, sums up this line of reasoning:

Twin and adoption studies provide convincing evidence for significant genetic effects on virtually all major psychiatric disorders. Therefore, genes that affect risk for these disorders must exist somewhere on the human genome. 2

The problem is insidiously circular: for someone to look at these studies and perceive convincing evidence of genetic causality, one already has to have accepted the idea that genes cause.

Why have geneticists chosen adoption studies as testing grounds for genetic effects? To understand this, imagine a regular (nonadoption) family situation, in which a child has been brought up in his family of biological origin. If a parent and child have the same disorder, that condition may, of course, have been passed on through genes. So far so good—but since it’s obvious that children can be influenced by their parents in many other ways, the mere incidence of an ailment “running in the family” does not necessarily point to a genetic cause. For example, if one of my children went to medical school, it wouldn’t necessarily establish that wanting to be a doctor is a hereditary disorder. As a leading behavioural geneticist points out, “because parents share family environment as well as heredity with their offspring, parent-offspring resemblance does not prove the existence of genetic influence.” 3

This is where adoption comes in. If a child is adopted, so the argument goes, he brings with him the genes he received from his parents but is now being raised in an entirely different environment. If he still manifests the same disorder that afflicted his birth father or mother, then that condition must be genetic. If we accept this logic and then look at the findings of adoption studies, an addiction like alcoholism will appear to be induced to a large degree by genetic inheritance—but that, upon inspection, is a rather enormous “if.”

In Chapter 19 we saw how prenatal stresses affect the developing brain. To conclude from adoption studies that a predisposition to alcoholism “runs in the family” and must, therefore, be genetic is to ignore all this evidence of environmental effects before birth.

Then, not all adoptions take place immediately at birth. In the largest, most oft-quoted and perhaps most influential study “proving” a genetic cause for alcoholism, the adopted children stayed with their parent (or parents) of origin for up to three years; the mean age of adoption was eight months. This study, which compared the adopted children of alcoholic biological parents with those of nonalcoholic parents, concluded that the biological father’s alcoholism had the greatest effect on the subsequent alcoholism of the male offspring. 4Even if that is so, it doesn’t necessarily indicate a genetic cause.

Given the long-term effects of prenatal stress and the dominant influence of the environment on brain development following birth, is it surprising that infants of alcoholic biological fathers would also have a greater propensity to drink? We know from the Adverse Childhood Experiences (ACE) Study that alcoholism is associated with many other traumatic circumstances—for example, either parent being alcoholic increases the chance of the mother being battered by a factor of thirteen. 5When we consider what it’s like for a woman to live with an alcoholic male partner—the insecurity she experiences through the pregnancy and beyond, and the abuse she may be subjected to—we can see that the stresses on such a woman, both before and after birth, would have been greater than the stresses on most other pregnant women. Furthermore, if a child spent the first months of his or her life—and possibly the first three years—under such circumstances, it would mean that by the time he was adopted, his attachment-reward, incentive-motivation and self-regulation systems would have been significantly impaired, along with his stress-response mechanisms. Such a study can tell us nothing about genetic effects. Similar objections, and a wide range of others, could be made—and have been made—to the other adoption studies. 6

Twin studies are accepted to be the gold standard of genetic surveys of human populations. Many genetic researchers believe that we can separate the effects of genes from those of the environment by comparing identical with fraternal twin pairs. The underlying belief is that identical and fraternal twin pairs both share the same environment to the same degree. As a geneticist who has done many twin studies admits, “our twin models assume that the exposure to relevant environmental factors was similar in monozygotic and dizygotic *37twins.” 7As we will now see, this is a completely unwarranted assumption.

Identical twins share the same genes; fraternal twins share some genes in common, but no more than any other pair of nonidentical siblings: about 50 per cent. A pair of identical twins, goes the argument, share not only genes but exactly the same environment—unless they are adopted by different families. Fraternal twins, being born at the same time to the same parents, also share the same environment but not the same genes. Therefore, goes the logic, any differences between such kinds of pairs must be genetic. Indeed, in twin studies of addiction the similarities in findings—known in technical language as concordance—between identical twins are consistently high as compared with the concordance for fraternal twins. That is, identical twins are more likely to share an addiction than are fraternal twins. In alcoholism, for example, the concordance for identical twins is about twice the rate for fraternal twins: a result which, according to a review article, “is consistent with addictive genetic factors.” 8

But this finding is at least equally consistent with environmental factors. It’s very obviously untrue that members of fraternal twin pairs share the environment to the same degree as identical twins. Far from it.

First, the fraternal twins are physiologically as different from each other as any pair of siblings. Whatever they experience, they will experience differently. If one, for example, is constitutionally highly sensitive, she will feel and absorb the effects of the same event more acutely than her “tougher” sibling, from early in the uterus and throughout childhood. Differences in temperament may also exist between identical twins, but not nearly to the same degree.

Second, recall that by far the most important aspect of the nurturing environment is the emotional interaction with the parent. Even with the best of love and good will, parents are much more likely to respond in the same way to identical twins than to nonidentical ones. For example, will a father or mother really look in the same way at nonidentical twins with different genders and temperaments? Will the parent use the same tone of voice or play in the same way with, say, a smaller female child than with her larger and more robust male sibling, or vice versa? On a deeper level, will the parents project the same fears, hopes and expectations on the children? Clearly not: each child represents something different to each parent and that means these two children do not grow up under identical conditions. They don’t share the same formative environment—not in the home and not on the playground or in school where nonidentical twins are much more likely to have very different peers and experiences than identical twins. So the assumption that you can tell genetic from environmental effects by comparing identical with nonidentical twin pairs also collapses. Identical twins share the environment much more extensively than any nonidentical pair possible can. *38