Genetic Study of One Million People Reveals Genetic Patterns Linked to Length of Schooling

The largest ever genetic study of human cognition has found more than 1,000 links between people’s genes and their educational path.

The work, which involved the DNA of 1.1 million people and researchers from 40 institutions, led to a scoring system that can roughly predict a person’s level of education by examining the DNA of that person.

Those with the lowest genetic scores had only a 10% chance of graduating from college. In contrast, those in the highest quintile of genetic promise did so 50% of the time.

It is not surprising that the path a person takes in school is partly determined by genes. Studies of identical twins raised separately, for example, show that they are surprisingly similar. Until recently, however, scientists lacked the tools to locate the genes that influence human behaviors.

What has changed is that researchers can now study much larger groups of people. This allows them to focus on tiny differences in the genome that, acting together, help explain how tall a person is, or the likelihood of developing a common disease like diabetes, or even how smart they are.

“This document will [be] a landmark in this new kind of social science,” says Eric Turkheimer, a psychologist at the University of Virginia, who was not involved in the study. “As a very successful application of new genetic technology, this is extraordinary.”

Specifically, the large transport of education-related genes will allow scientists to “begin to ask questions about how individual genes contribute to the biological pathways that ultimately lead to the brain and learning,” he says.

The new effort to link DNA to education, described today in Natural genetics, is among the first to simultaneously assess the genes of more than one million people. It used more than 400,000 DNA profiles collected in Britain as part of the national UK Biobank project, and another 365,536 were provided by 23andMe, the consumer genetic testing company in the San Francisco area.

Some researchers say the findings will allow children’s learning potential to be assessed from their DNA in the form of a genetic intelligence test, giving parents or school systems a way to identify those who are more promising or to explain why others have problems.

The authors of the present study strongly contest this idea. In an FAQ document they distributed to reporters, they said their rating system was just a scientific tool. “Any practical response – individual or at the policy level – to this or similar research would be grossly premature and unsupported by science,” they wrote.

According to Daniel Benjamin, a behavioral economist at the University of Southern California and one of the study’s lead authors, the predictions are still too unreliable to apply to individuals. The genetic variants he and his colleagues measured can only explain about 11% of the variability between people in educational attainment.

“Until the score is better and we understand the causal factors behind it, I’m pretty uncomfortable using it to predict individual outcomes,” Benjamin said. “There’s still a lot of work to do before we even have a conversation about using it this way.”

Still, Benjamin acknowledged that DNA is now a better predictor of how long people have been in school growing up in a rich or poor household, and almost as good a predictor as their parents’ level of education.

Researchers say this new type of genetic assessment, called a polygenic risk score, can also provide insight into a person’s risk of developing heart disease, mental illness or other conditions.

Exactly how genes create a tendency towards more or less education remains fundamentally uncertain. This could result from the action of other traits, such as conscientiousness, intelligence, or even body mass. The effect of genes is also highly dependent on social context. In a society without formal education, for example, people’s DNA would say nothing about the level of education they complete.

“These are not genes that have the same effect everywhere,” says Turkheimer. “Instead, they influence results in subtle, context-sensitive and hard-to-trace ways, with effects that can only be detected in huge samples.”

Comments are closed.