Skip to main content

Using DNA Tests to Predict IQ: How Personal Genomics May Influence the Way We Think about Intelligence

By Celeste Fong

Image courtesy of Pixabay.

Thanks to the increasing popularity of direct-to-consumer genetic testing, it is now cool to talk about your DNA. Companies like 23andMe market personal genomics to analyze and interpret a customer’s DNA to provide them with information about their ancestry or health. As of now, your genes can tell you about your predisposition for arthritis or baldness, but what if it could tell you about your intelligence? Recent scientific research has identified specific genes correlated with intelligence. It’s possible that such predictors could accompany the next generation of DNA test results. But how will this new insight change the way we think about genetics and intelligence?

Intelligence, which is partially heritable, has long been associated with a number of social, economic, and health-related outcomes (Conti et al., 2010; Cutler & Lleras-Muney, 2008). Until recently, however, there have been no confirmed genetic associations to intelligence; there simply wasn’t a sample size. Then in 2017, a Dutch led genome-wide association study used the DNA of nearly 80,000 individuals to implicate over 50 genes related to intelligence. The data was gathered across 13 online databases and measured Spearman’s g (expressed as IQ), which most closely captures a factor of cognitive ability and general intelligence. The results had the power to predict 3% of the variance in intelligence (Sniekers et al., 2017).

While 80,000 is a large sample size, a prediction of 3% is not enough to have much of a real-world effect. A subsequent study used over 1 million samples collected from 23andMe and was able to find more than 1,200 genetic variants associated with educational attainment. This showed the value that commercial genomics could have in understanding intelligence. The increased data set was able to predict 11% of the variance in educational attainment and up to 10% in cognitive performance (Lee et al., 2018).

Image courtesy of Pixabay.

At this point you may be asking: how can DNA be used to predict my intelligence? The answer lies in polygenic scores. Essentially, once the genes that influence intelligence are located, a predictive likelihood that a certain genetic makeup will result in a particular phenotype is calculated according to the “weight” associated with certain genes. This method of analysis is similar to that of DNA health predictors, some of which are already included in test results.

It’s a controversial thought. 23andMe allows its customers to consent to have their data used in research, but the company itself refuses to report on their customers’ intelligence. According to an article by MIT Technological Review, this is due to the risk of “potential misinterpretation,” among other reasons (Regalado, 2018). However, predictive intelligence is already being quantified and provided to customers though other services like DNA Land and GenePlaza. The results tend to be accompanied by a number of disclaimers, but the number of users that these services cater to indicate that it’s time to take more seriously the impact that personal genomics has on intelligence.

Robert Plomin, for example, is keen on the idea that intelligence predictors will soon be routinely available in direct-to-consumer DNA tests. He outlines the value of the genetic components of intelligence in his paper, “The New Genetics of Intelligence,” seeing predictive IQ as a guide for personalized education – what he calls “precision education” (Plomin and Stumm, 2017). Similar to precision medicine, predictive intelligence will allow individuals to receive a customized education recommendation based on the intellectual strengths and weaknesses indicated by their predicted IQ. Perhaps one’s DNA predicts that he/she will be likely to excel academically and ought to be placed in a more rigorous classroom. Maybe another’s will predict that he/she is expected to have a harder time in the classroom and would be better suited to other activities. Plomin says using predictive IQ in such a manner might “help parents understand why their children differ in school achievement” (Plomin and Stumm, 2017).

Image courtesy of  Pixabay .

Relying on genetic components as the basis of intelligence, like Plomin does, poses its own risks as well. For some, it might perpetuate the concept of social categorization, potentially leading to the labeling of others based on their genetic predictors. It could be that those with high predictive IQs will face pressure to cultivate their academic potential to become high achievers, while those with low predictive IQs will face pressure to improve academically or risk being rejected. This thought is based on biological determinism, the belief that behavior or ability is controlled by genetics. This type of determinism ultimately misappropriates the nature of predictivity as deterministic rather than merely probabilistic. It also underscores the role that an individual’s environment has in affecting their intelligence, although the interplay between environment and genetics with regards to intelligence is not fully understood yet.

Other potential worries include issues related to discrimination and stigmatization. With predictive IQ, a line of demarcation is inherently being drawn when we decide to characterize individuals using what could be a nearly arbitrary figure. This is true when we create a threshold for whether or not children should be placed in overachieving versus underachieving academic settings, and the same could be true if predictive intelligence were to significantly impact an individual’s future opportunities. Perhaps getting accepted into university or being offered a coveted position at a company will be dependent on whether or not your predictive IQ is above your otherwise equally qualified competitors. Similar to height in sports like basketball, your height, while not equivalent to your talent in the sport, may make you more attractive to recruiters. With intelligence, your predictive IQ won’t necessarily equate to your ability to perform a certain job, but it could make you seem more appealing to employers.

At its worst, one could even make a eugenics argument by claiming that either higher or lower predictive IQs are more valuable than the other. It has already been suggested that parents using in-vitro fertilization might choose to undergo genetic testing for predictive intelligence in order to choose the best sperm donors, implant the smartest embryos, or avoid those that display a potential for cognitive disorders and difficulties (Regalado, 2018). The topic of conversation in bioethics recently has been the successful creation of genetically modified CRISPR babies in China (Marchione, 2018). Does this mean we could be nearing a reality where we design our babies right down to their intellectual abilities?

Lastly, it is also worth mentioning the ethical quandary of privacy when an entire genome of individuals can be passed between genomics companies or secondary vendors. With the understanding that one’s body belongs to oneself, how is ownership of and access to one’s genetic sequence managed with the increasing value and use of big data?

I suppose you’re now wondering if we should be worried about a world where we can use direct-to-consumer DNA tests to predict intelligence. My answer to that is no – at least, not yet.

Image courtesy of Pixabay.

While advancements are being made in locating a genetic basis for intelligence, the data do not provide much support for the ability to accurately predict real-world intelligence. After all, IQ scores are merely indicators of different factors related to intelligence. These numbers represent ability in areas like memory, mathematics, or spatial visualization, but the results do not equate to our overall concept of intelligence. It could be that there will always remain some level of abstraction in intelligence.

The accuracy of polygenetic scores is also up for debate, especially when it comes to predicting characteristics like intelligence. Despite what it may look like, the data sets used in meta-analyses are not large enough to provide a statistically accurate estimation of intelligence because it is naturally a highly variable trait among populations. There might never be a static measure of intelligence among the same population sizes, and as a result, it would be difficult to provide an accurate prediction of intelligence based on the constantly changing data sets.

Finally, it is always important to remember that correlation does not imply causation. Simply because there is a relationship between IQ score, DNA, and life outcomes in certain populations, it does not guarantee that the next individuals in those same groups are going to follow similar trajectories.

As genetics intelligence research continues, it will ultimately be important for the implications of genetic testing and intelligence to be thought about and discussed at all stages.


Celeste Fong is a current student in the Masters in Bioethics program at Emory University. Having received her Bachelor’s degree in Neuroscience from the University of Alabama at Birmingham, she is interested in exploring how advances in neuroscience research and technology can challenge society’s perception of itself and its values.

  1. Conti, G., Heckman, J. & Urzua, S. (2010). The education-health gradient. Am. Econ. Rev. 100, 234-238
  2. Cutler, D. M. & Lleras-Muney, A. (2008). In House, J. et al. (Ed.), Making Americans Healthier: Social and Economic Policy as Health Policy. New York, NY: Russell Sage Foundation.
  3. Deary, I. J., Pattie, A. & Starr, J. M. (2013). The stability of intelligence from age 11 to age 90 years: the Lothian birth cohort of 1921. Psychol. Sci. 24, 2361-2368.
  4. Lee, J. J., Wedow, R., Okbay, A., Kong, E., Maghzian, O., Zacher, M., . . . (2018). Social Science Genetic Association, C. Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nature Genetics, 50(8), 1112-1121. doi:10.1038/s41588-018-0147-3
  5. Marchione, M. (2018, November 26). Chinese researcher claims first gene-edited babies. Retrieved from
  6. Plomin, R., & von Stumm, S. (2018). The new genetics of intelligence. Nature Reviews Genetics, 19, 148. doi:10.1038/nrg.2017.104
  7. Regalado, A. (2018, August 02). DNA tests for IQ are coming, but it might not be smart to take one. Retrieved from
  8. Regalado, A. (2018, August 13). Eugenics 2.0: We’re at the dawn of choosing embryos by health, height, and more. Retrieved from
  9. Sniekers, S., Stringer, S., Watanabe, K., Jansen, P. R., Coleman, J. R. I., Krapohl, E., . . . Posthuma, D. (2017). Genome-wide association meta-analysis of 78,308 individuals identifies new loci and genes influencing human intelligence. Nature Genetics, 49, 1107. doi:10.1038/ng.3869

Want to cite this post?

Fong, C. (2019). Using DNA Tests to Predict IQ: How Personal Genomics May Influence the Way We Think about Intelligence. The Neuroethics Blog. Retrieved on , from


Emory Neuroethics on Facebook