What Counts as a Genetic Trait

What does it mean to call something a genetic trait or disease? Clearly, at least part of that judgment rests on some kind of causal assessment. If a disease is genetic, then it is caused by one or more of an organism's genes. Indeed, this seems to fit a more general concept of disease, in which the causal basis of disease is incorporated into our nosologies. As Richard Hull has explained:

In its efforts to understand, control, and avoid disease, modern medicine has incorporated into the very identification of a disease the notion of the cause of the syndrome. This permits the individua-tion of similar syndromes with distinct causes into different diseases. (p. 61)

There is a fairly obvious problem with this as a way of distinguishing between genetic and epigenetic diseases. That is because there are genetic and nongenetic factors which are causally relevant to every trait, a fact recognized by virtually all commentators on the concept of genetic disease (see Gifford; Hull, 1979). So the real issue in deciding that something is a genetic disease, is whether the causal factors which are genetic are the most important causes. How do we decide whether genetic factors or environmental factors are more important in the production of various diseases? In response to the selection problem, a number of solutions have been proposed. These can be grouped into a few major categories.

One approach is to try to tease out a notion of genes as direct causes of disease. In 1990 Fred Gifford tried to capture this notion in one of his two definitions:

...the trait must be the specific effect of some genetic cause, that the trait must be described or individuated in such a way that it is properly matched to what the gene causes specifically. (p. 329)

However, this approach seems hopeless in the face of the actual complexity of development. Quite simply, this definition probably does not identify any diseases or traits as genetic. As Kelly Smith argued in 1990, "genes do not directly cause anything of immediate phenotypic significance" (p. 338).

Perhaps the most obvious and promising approach to the selection problems is to try a statistical approach. A number of variants on this have been attempted.

The first and central sense of genetic is this: a trait is genetic if genetic differences in a given population account for the phenotypic differences in the trait-variable amongst members of that population. (Gifford, p. 334)

This seems to exactly capture at least something important about society's concept of genetic disease. It can be put perhaps more precisely in terms of covariance. When some trait is identified as genetic, it can be argued that (in that population) the covariance of the trait with some genetic factor(s) is greater than the covariance of the trait with other (nongenetic) factors. This solves the selection problem neatly by allowing us to pick out which causal factors are irrelevant (the ones which are fixed) and highlight the important ones (the ones that make the difference). In one of the canonical examples of causality, one is inclined to say that the lighting of a match (under normal circumstances) was the cause of the fire, while the presence of oxygen (while a contributing causal factor) was not. In contrast, in an environment where fire was normally present and oxygen was not, one might well pick out the (unusual) presence of oxygen as the cause of a fire.

There are several advantages to this approach to the selection problem. First, it corresponds to the use of analysis of variance that is used by biologists to measure the causal contribution of hereditary and environmental factors in a population. Second, it is capable of clear explication. Third, it has at least some intuitive support. However, this account seems to conflict with common usage in cases where pathogens typically identified as the cause of disease are nearly ubiquitous (so that, for example, genetic factors may make the difference between which people exposed to the pathogen become ill).

In spite of its advantages, the statistical approach fails to capture all of the myriad uses of the concept of genetic disease. Another approach has been developed from the way the most important causal factor in an explanation is picked out.

Philosophers have claimed on quite general grounds that the most important cause is chosen in terms of the manipulability of the various factors. Whatever the general virtues of this approach, it is promising when it comes to medicine. In the natural sciences, it could be argued that there is a strong interest in prediction and explanation. In contrast it has been argued that the medical realm is more concerned with the prevention and treatment of disease than with explanation (Wulff; Engelhardt). Instrumentalist interests play a much more central role in medical practice than in science. Hence, the appropriate solution to the selection problem can be formulated in terms of manipulability. The most important cause is the one that is identified as the most easily manipulated to prevent or treat disease. A disease is genetic if it is genes that play this role and epigenetic if it is non-genetic factors that are most easily manipulated.

Like the statistical definition, the manipulability definition captures something important about our usage of the term. In addition it is often an implicit aspect of the justification for the extension of the concept of genetic disease to new cases. However there are some problems with this approach as well. The obvious problem seems to be that on this analysis, no disease could be classified as genetic. Many of the paradigm genetic diseases (phenylketonuria [PKU], cystic fibrosis [CF]) involve treatments that are not molecular. Indeed, in the case of PKU, the standard treatment involves a change in diet. At the same time the tests for PKU were developed before the actual mutation responsible for the disease had been identified. It is impossible to adhere to the manipulability definition and accept that PKU is a genetic disease. This seems to be a fatal flaw in the manipulability definition. In addition, it is not true that biomedical science is always instrumentally oriented. A great deal of effort is aimed not just at treating and preventing disease, but at understanding it. This may lead to a conflict over which causal factor is most important (the factor most easily manipulated for treating or preventing a disease may not be the most revealing for the purposes of understanding a disease).

It is worth noting that both the statistical approaches and the manipulability approaches seem to imply a relativity in the concept of genetic disease. In the case of the statistical notion, something will count as a genetic disease or not, depending on the population it is a part of. The manipulability definition implies that technological advances will affect what counts as a genetic disease as the reach of our technology is extended. Yet, this result seems to be incompatible with an ontological conception of disease. If diseases are real entities (and independent of values) then the solution to the selection problem should not depend on factors outside of the organism (Boorse). Thus the normativist or constructivist position on disease seems to be supported by these analyses (however inadequate they are as a general account).

Anxiety and Depression 101

Anxiety and Depression 101

Everything you ever wanted to know about. We have been discussing depression and anxiety and how different information that is out on the market only seems to target one particular cure for these two common conditions that seem to walk hand in hand.

Get My Free Ebook

Post a comment