By Daniel W. Belsky, PhD
Large-scale data mining of more than one hundred thousand human genomes recently discovered genetic variants related to differences in educational attainment. We followed-up results from that data mining study to uncover the developmental and behavioral paths that connected DNA sequences with life outcomes.
We studied a cohort of 1,037 individuals all born in 1972-3 and followed-up at regular intervals through their 38th year of life: The Dunedin Study. We started at the end. We asked whether children born with a higher complement of education-associated genetic variants were better off four decades later as compared to their peers who carried fewer of these genetic variants. They were. At age 38 years, Study members who carried more of the education-associated variants had more prestigious jobs, higher incomes, better credit scores, fewer financial problems, and so on. In fact, even among Study members who completed the same level of education, those who carried more of the genetic variants we studied achieved better socioeconomic outcomes. In other words, the genetics we were studying were not the genetics of education only. Instead, these genetics predicted a broad pattern of socioeconomic success.
Our next step was to go back to the beginning. Children inherit genes and social position from their parents. And we know that where kids start in life influences where they end up. So we tested if the reason children who carried more of the education-associated genetic variants did better in life was because they started off that way. As it turned out, children’s social origins were related to their genetics. Kids born into better-off families did tend to carry slightly more of the education-associated genetics. But starting life with a bit of social advantage didn’t explain why kids with these genetics did better as adults. In fact, the genetics predicted children’s success in life whether they were born rich or poor. Regardless of where they started off in life, having more education-associated genetic variants predicted more success later on.
To understand why children with these particular genetics did better in life, we dug deep into the Dunedin Study archives. We compiled records of children’s developmental milestones, behavioral observations made of the children with their mothers, reading tests, psychometric testing, interviews, and physical exams. Our analysis of this trove of data revealed three important findings.
First, children with more education-associated genetic variants were distinguished from their peers early on by more rapid development of language and reading skill. In contrast, they did not show an accelerated pace of physical development.
Second, children with more education-associated genetic variants scored higher on intelligence tests. But they also demonstrated better self-control and were more interpersonally skillful (friendly, cooperative, communicative, etc.). Collectively, these characteristics, measured when Study members were children, accounted for about half of the genetic association with midlife socioeconomic success. In other words, both cognitive and “non-cognitive” skills linked children’s DNA with their life success.
Third, children with more education-associated genetic variants were no healthier than their peers. One possible explanation for our findings was that children with more education-associated genetic variants were simply healthier. But they weren’t. We examined children’s lung function, blood pressure, body fat, height, weight, motor skills, and histories of illness, all measured at repeated intervals during their first decade of life. We found no relationship between education-associated genetics and children’s physical health. Thus, the link between the genetics we studied and life success seems to be specific to the brain, at least in childhood.
As adults, although they were better-off economically, Dunedin study members who carried more of the genetics of success were not more satisfied with their lives. (For more details on adult life courses, see the paper linked at the end of this post)
Complex traits (like educational success) are influenced by large numbers of genetic variants, each with small effects. We study this sort of genetic influence using a technique called “polygenic scoring.” Polygenic scoring uses results from genome-wide association studies to build a simple algorithm that can be applied to a person’s genetic data. The output of this algorithm is a summary of the influence of tens of thousands of genetic variants on a trait. We computed polygenic scores for Dunedin Study members based on results from a previously published genome-wide association study of educational attainment. Polygenic scores are normally distributed (bell-shaped). In the case of the education polygenic score, that means some of us carry very few education-associated variants, some carry a lot, and most of us are in the middle.
The genetics of success are not ready for prime time but are an opportunity for social and behavioral science
The magnitudes of genetic influences identified in our study were small. The sensitivity and specificity of genetic prediction was too low for “precision education.” It is possible that with newer and more powerful genetic studies, we could devise more accurate polygenic scores. Policy makers, scientists, and the general public need to start the conversation about how such algorithms should be used beyond the realm of scientific inquiry – if at all.
The genetics of success are an opportunity for social and behavioral science. We studied as single birth cohort in a single country. Will we see the same general patterns at different points in history or under different policy contexts? Do more generous social welfare states limit or amplify influence of the genetics of success? Do changing social norms shape genetic influences? What role do families, schools, and neighborhoods play as contexts within which genetic influence are expressed? Critically, can interventions that target pathways linking the genetics of success with life outcomes promote upward social mobility?
These questions and many more beg answers. Social and behavioral scientists are getting into the game, e.g. using genetics to study economic inequality, assortative mating, fertility, and more. Interested scientists might check out the Integrating Genetics and Social Science conference as a great way to learn more.
Read the article
Belsky DW, Moffitt TE, Corcoran DL, Domingue B, Harrington HL, Hogan S, Houts R, Ramrakha S, Sugden K, Williams B, Poulton R, Caspi A. The genetics of success: How SNPs associated with educational attainment relate to life course development. Psychological Science, 27(7)957-972. PMC4946990.
This work was supported by the National Institute on Aging (AG032282, AG048895, AG049789, AG028716) The National Institute of Child Health and Human Development (HD078031), and by the Jacobs Foundation.
About the author
Daniel W. Belsky, Ph.D. is Assistant Professor of Medicine at Duke University School of Medicine, Assistant Research Professor at the Duke Social Science Research Institute, and Early Career Fellow of the Jacobs Foundation. Dan works at the intersection of population sciences and public health, biomedicine, and genomics. His research seeks to understand causes of socioeconomic disparities in health in the aging process with the aim of devising novel approaches to intervention, in particular efforts to promote successful development early in life and slow deterioration in aging. Find out more here.