Why do some people sail happily through life whereas others are brought low by its slings and arrows? A 2003 study offered an answer: a gene variant that made some people more susceptible to depression. The work was hailed as a prime example of how someone's life experiences could activate their genes. But a new analysis of the 2003 study and many that came after it calls the link into question.
The 2003 study (Science, 18 July 2003, p. 291) dealt with serotonin, a brain chemical that plays a major role in depression and is a key target of antidepressant drugs. Researchers led by psychologist Avshalom Caspi, now at Duke University in Durham, North Carolina, found from a long-term study of 847 people in New Zealand that those with a short version, or allele, of the serotonin transporter gene were more likely to become depressed by adverse life events than were those with only long alleles. The researchers thought that adverse experiences somehow caused the gene to make less serotonin available in brain cells. "Everybody was excited about this at the time," says Kathleen Merikangas, a genetic epidemiologist at the National Institute of Mental Health in Bethesda, Maryland. "It was very widely embraced."
Now a multidisciplinary team headed by Merikangas and geneticist Neil Risch of the University of California, San Francisco, has concluded that the finding doesn't hold up. The group pored over the data from the original 2003 study, plus 13 later studies. In all, the meta-analysis covered some 12,500 individuals. Only three of the smaller studies replicated the original finding, the authors report today in the Journal of the American Medical Association. Although depression clearly is associated with stressful life events, Risch and his colleagues say that no matter which way they chopped up the data, they were unable to find a connection between the suspect allele and depression induced by adverse life events.
Caspi and co-author Terrie Moffitt, also now at Duke, are sticking to their guns. "The article ignores the complete body of scientific evidence," says Caspi. He and Moffitt argue that Risch's team excluded laboratory studies showing that humans with the short allele show greater biological stress responses in the lab, and other research showing people with the short allele are more vulnerable to disorders such as post-traumatic stress disorder and anxiety, which are related to depression. They also claim that the meta-analysis accords too much weight to large studies "not [of] the best quality"--where subjects were probed with phone calls or questionnaires, instead of in person.
Experts on both sides of the issue say that the Risch study was well-done. But psychiatric geneticist Joel Gelernter of Yale University isn't ready to discount the 2003 paper. "I still think the Caspi study is a very good study," he says, noting that the serotonin alleles may have a more nuanced role in depression than previously thought.
Psychiatric geneticist Kenneth Kendler of Virginia Commonwealth University in Richmond, says he's not surprised by the new study. Even though he published a paper that supports the Caspi findings, he points out that many studies have failed to link the serotonin transporter gene to clinical depression. "Gene discovery in psychiatric illness has been very hard--the hardest kind of science," says Kendler. Reliable clues will require large studies with thousands of possible genes, he says; individual ones can only be expected to have very small effects.