David Rettew :
Go through the Powerpoint presentations of faculty who teach students about depression, and very likely you will you find a slide with a simple-looking graph that displays the roles that both early maltreatment and a particular gene have in the condition. The figure, which comes from a 2003 paper published in the prestigious journal Science, shows that the risk of depression among people who have experienced stressful life events becomes much greater if they also possess a certain version of a certain gene. This landmark study, in the midst of an endless nature-nurture debate, was evidence that not only were both genetic and environmental factors important, the two worlds actually depended on each other. This demonstration of what was called a “gene-environment interaction” became a cornerstone principle in our understanding of how behavior, both typical and atypical, develops.
One of the remarkable things about this study, in addition to the novel way that genetic and environmental factors were considered together, was the finding that such strong results could be seen when considering the status of a single gene out of the thousands that encode for proteins involved in brain function and structure. While at the time there was appreciation that complex disorders like depression were not solely determined by the status of a single gene, it did seem plausible that some individual genes-including the one described in the Science study, which codes for a receptor in the brain that binds the important neurotransmitter serotonin-could be important in themselves, especially if considered along with particular environmental factors like trauma, early attachment, or poverty.
What followed over the next decade and a half was an explosion of studies documenting the apparent importance of a small number of what were termed “candidate” genes in the development of many kinds of psychiatric disorders-each with a very plausible story about how it might operate in brain development to create particular symptoms. This particular serotonin gene (SLC6A4) often was referred to as “the depression gene.” Rapidly, the interest in candidate genes spawned thousands of journal articles and cost many millions of dollars in research budgets, not to mention countless hours of effort. Soon thereafter came private industry, willing to test your own DNA to see how many of these “risk alleles” you carry in order to help you determine your level of risk and to recommend which medications you should take if you do develop symptoms.
A new study, now, throws all of this into question.
To be fair, skepticism about candidate gene studies were there from the start, and some earlier work had already thrown some very cold water on some of the more famous candidate gene hypotheses. But a study published last month in the American Journal of Psychiatry has dealt a major blow to the entire candidate gene world. Using samples of up to nearly half a million people, the researchers decided to try and replicate the results of dozens of previously published links (which were often based on much smaller samples) between 18 particular candidate genes and depression, either on their own or in interaction with environmental factors such as early abuse or poverty. What they found was essentially nothing, and the authors argued that these particular genes are “no more associated with depression… than genes chosen at random.” In the end, the authors call researchers to “abandon” candidate gene research in depression. And since the release of the paper, few people seem to be arguing otherwise.
Science, of course, is based on hypotheses and knowledge that turn out to be incorrect. That said, it is hard to spin this saga into a positive story for psychiatric research. How could so many people (including myself; I dabbled in candidate gene research) get this wrong? Most people think that what happened was publication bias. With so many interesting genes to test, some of them would inevitably show links to depression simply due to chance. Those false positives would get written up, published, and sometimes covered in the media while negative associations were simply left unpublished and forgotten, what has been called the “file drawer effect.” To be clear, it is also important to point out what this study does not mean. First and foremost, it does not mean that there isn’t a genetic basis to depression, just that it more likely involves a much larger number of genes, each of which alone has a very small effect. Scientific evidence from many other sources, like twin studies, still point quite convincingly to the important-although certainly not overpowering-role of genetics in the development of depression and most other psychiatric conditions as well. The study also doesn’t shoot down the gene-environment interaction hypothesis either, it just again makes it a bit more complicated. Finally, it is important to note that the study was focused on depression and there are examples in other conditions, from cystic fibrosis to breast cancer, where single genes can play large roles in the development of the illness.
One area that will certainly get some additional scrutiny is companies charging hundreds of dollars to tell you about your risk for depression based on your status of many of the very same genes that were debunked in this study. Whether some insurance companies will still be willing to cover these tests is also very much in question.
Medical research in general, and psychiatric research in particular, will go on as it should, but hopefully with some lessons learned. Just like in other areas, things that sound too good (or simple) to be true often are. And studies that don’t replicate the flashy finding of last month need to be given the same air time as the ones that do. This won’t be easy as ever more journals compete for readers and clicks, but some additional discipline all around might help keep the next cool-sounding hypothesis from getting too big for its britches.
Go through the Powerpoint presentations of faculty who teach students about depression, and very likely you will you find a slide with a simple-looking graph that displays the roles that both early maltreatment and a particular gene have in the condition. The figure, which comes from a 2003 paper published in the prestigious journal Science, shows that the risk of depression among people who have experienced stressful life events becomes much greater if they also possess a certain version of a certain gene. This landmark study, in the midst of an endless nature-nurture debate, was evidence that not only were both genetic and environmental factors important, the two worlds actually depended on each other. This demonstration of what was called a “gene-environment interaction” became a cornerstone principle in our understanding of how behavior, both typical and atypical, develops.
One of the remarkable things about this study, in addition to the novel way that genetic and environmental factors were considered together, was the finding that such strong results could be seen when considering the status of a single gene out of the thousands that encode for proteins involved in brain function and structure. While at the time there was appreciation that complex disorders like depression were not solely determined by the status of a single gene, it did seem plausible that some individual genes-including the one described in the Science study, which codes for a receptor in the brain that binds the important neurotransmitter serotonin-could be important in themselves, especially if considered along with particular environmental factors like trauma, early attachment, or poverty.
What followed over the next decade and a half was an explosion of studies documenting the apparent importance of a small number of what were termed “candidate” genes in the development of many kinds of psychiatric disorders-each with a very plausible story about how it might operate in brain development to create particular symptoms. This particular serotonin gene (SLC6A4) often was referred to as “the depression gene.” Rapidly, the interest in candidate genes spawned thousands of journal articles and cost many millions of dollars in research budgets, not to mention countless hours of effort. Soon thereafter came private industry, willing to test your own DNA to see how many of these “risk alleles” you carry in order to help you determine your level of risk and to recommend which medications you should take if you do develop symptoms.
A new study, now, throws all of this into question.
To be fair, skepticism about candidate gene studies were there from the start, and some earlier work had already thrown some very cold water on some of the more famous candidate gene hypotheses. But a study published last month in the American Journal of Psychiatry has dealt a major blow to the entire candidate gene world. Using samples of up to nearly half a million people, the researchers decided to try and replicate the results of dozens of previously published links (which were often based on much smaller samples) between 18 particular candidate genes and depression, either on their own or in interaction with environmental factors such as early abuse or poverty. What they found was essentially nothing, and the authors argued that these particular genes are “no more associated with depression… than genes chosen at random.” In the end, the authors call researchers to “abandon” candidate gene research in depression. And since the release of the paper, few people seem to be arguing otherwise.
Science, of course, is based on hypotheses and knowledge that turn out to be incorrect. That said, it is hard to spin this saga into a positive story for psychiatric research. How could so many people (including myself; I dabbled in candidate gene research) get this wrong? Most people think that what happened was publication bias. With so many interesting genes to test, some of them would inevitably show links to depression simply due to chance. Those false positives would get written up, published, and sometimes covered in the media while negative associations were simply left unpublished and forgotten, what has been called the “file drawer effect.” To be clear, it is also important to point out what this study does not mean. First and foremost, it does not mean that there isn’t a genetic basis to depression, just that it more likely involves a much larger number of genes, each of which alone has a very small effect. Scientific evidence from many other sources, like twin studies, still point quite convincingly to the important-although certainly not overpowering-role of genetics in the development of depression and most other psychiatric conditions as well. The study also doesn’t shoot down the gene-environment interaction hypothesis either, it just again makes it a bit more complicated. Finally, it is important to note that the study was focused on depression and there are examples in other conditions, from cystic fibrosis to breast cancer, where single genes can play large roles in the development of the illness.
One area that will certainly get some additional scrutiny is companies charging hundreds of dollars to tell you about your risk for depression based on your status of many of the very same genes that were debunked in this study. Whether some insurance companies will still be willing to cover these tests is also very much in question.
Medical research in general, and psychiatric research in particular, will go on as it should, but hopefully with some lessons learned. Just like in other areas, things that sound too good (or simple) to be true often are. And studies that don’t replicate the flashy finding of last month need to be given the same air time as the ones that do. This won’t be easy as ever more journals compete for readers and clicks, but some additional discipline all around might help keep the next cool-sounding hypothesis from getting too big for its britches.
(David Rettew, M.D., is a child psychiatrist at the University of Vermont and author of Child Temperament: New Thinking about the Boundary between Traits and Illness).
