A meta-analysis in the Journal of the American Medical Association says that the "addition of the serotonin transporter genotype did not improve the prediction of risk of depression beyond that associated with exposure to negative life events" — contrary to what a 2003 study found.