For more than 20 years, I have hammered behavioral genetics, which attempts to pinpoint the genetic underpinnings of an enormous variety of human traits and disorders, from homosexuality and religiosity to alcoholism and depression. I’m especially disenchanted with research linking genes to intelligence, which both historically and today is often used to promote racist theories of genetic superiority.
My objections are primarily scientific, not moral. The field of behavioral genetics has a horrendous track record, with a long string of sensational claims that turned out to be erroneous.
As I have repeatedly explained, the methodology of behavioral geneticists is highly susceptible to false positives. Researchers select a group of people who share a trait and then start searching for a gene that occurs not universally and exclusively but simply more often in this group than in a control group. If you look at enough genes, you will almost inevitably find one that meets these criteria simply through chance.
A new report in the Proceedings of the National Academy of Sciences has bolstered my skepticism. Titled “Common genetic variants associated with cognitive performance identified using the proxy-phenotype method,” the paper is authored by an international consortium of 59 researchers, including such stars as Steven Pinker and Robert Plomin.
The authors compiled data on academic performance and cognitive tests from more than 100,000 subjects and found three genetic markers “significantly associated with cognitive performance after correction for multiple hypothesis testing.”
As usual, some journalists are touting the study as a major advance. “‘Smart’ genes associated with human intelligence have been identified for the first time,” proclaims an Australian paper. But not all journalists are impressed. In an article for Nature, “Smart Genes” Prove Elusive,” Ewen Callaway points out that the effects reported in the PNAS paper are “maddeningly small.”
Callaway elaborates: “The three variants the researchers identified were each responsible for an average of 0.3 points on an IQ test. (About two-thirds of the population score between 85 and 115.) That means that a person with two copies of each variant would score 1.8 points higher on an intelligence test than a person with none of them. To put those figures in perspective, those variants have about one-twentieth the influence on intelligence as do gene variants linked to other complex traits such as height, says Daniel Benjamin, a social scientist at Cornell University in Ithaca, New York, who co-led the study.”
Kevin Mitchell, a neurogeneticist at Trinity College, Dublin, tells Callaway: “With effects this small, the chances that they represent false positives are vastly increased. While intelligence — and proxy measures such as cognitive test performance or educational attainment — are quite heritable, the idea that this trait is determined by common [genetic] variants in the population at large is really unproven.”
Some behavioral geneticists are beginning to acknowledge their field’s profound problems. A 2012 editorial in Behavior Genetics stated: “The literature on candidate gene associations is full of reports that have not stood up to rigorous replication. This is the case both for straightforward main effects and for candidate gene-by-environment interactions… As a result, the psychiatric and behavior genetics literature has become confusing and it now seems likely that many of the published findings of the last decade are wrong or misleading and have not contributed to real advances in knowledge.”
If history is any guide, the new PNAS paper will turn out to be yet another “wrong or misleading” report. I applaud Ewen Callaway for treating it with the skepticism it deserves. We need more journalists like him.
John Horgan directs the Center for Science Writings, which is part of the College of Arts & Letters. This column is adapted from one originally published in his ScientificAmerican.com blog, “Cross-check.”
Be First to Comment