A new study by Roberto Colom and his coauthors (published in the ICA Journal) examines the stability and change in IQ in children with above-average intelligence at age 7. What it finds is revealing.
The major finding is that IQ changes in childhood are common. In early childhood, large IQ fluctuations are common. These changes get smaller in adolescence, but they still happen. Moreover, the changes tend to be larger for children with IQs of 115+ at age 7 (right panel) than those with IQs of 99-114 (left panel). This is not terribly surprising because regression towards the mean should be larger in the higher-IQ group.
Documenting these changes is important, but the authors also investigated whether IQ changes could be predicted by DNA-based polygenic scores, background variables, home environment, and behavioral problems.
The results showed that increasing IQ through childhood and into early adulthood was positively associated with higher polygenic scores and higher socioeconomic status. The most consistent predictors of increasing IQ was the DNA-based polygenic scores and socioeconomic status. The most consistent predictor of decreasing IQ was behavioral problems, though adverse life events were pretty consistent in the 99-114 IQ group.
These results match prior studies on cognitive development and confirm the importance of genes in determining the adult IQ of a person. They also show the importance of seeing children’s intelligence as a trait that is still in the process of developing. Practices like giving IQ tests to very young children and labeling the as “gifted” for the rest of their education are not justified. In this study, only 16% of children with IQs of 115+ still had a score that high at age 21. Regularly reassessing children’s cognitive development is best practice.
The instability of early high IQ is striking. Only 16% of kids with 115+ IQs at age 7 maintained that level by age 21. This destroys the idea that gifted identification in early elementary school is reliable. The finding that polygenic scores and SES consistently predict upward trajectories while behavioral problems predict downward ones shows that IQ development reflects both genetic potential and environmental support. Regression to the mean explains some decline in the high group, but the fact that behavior problems consistently predict drops suggests that non-cognitive factors matter for realizing cognitive potential. Schools need to reassess regularly rather than locking kids into tracks based on age 6-7 testing.
What stands out is that most high-scoring young kids are “cognitively mobile” rather than stable geniuses. Early high scores often don’t persist, which has huge implications for gifted programs that identify kids once and never reassess. The polygenic score findings are important because they show genes influence developmental trajectory, not just starting point. Kids with higher genetic loading maintain or increase scores more often. The behavioral problems finding suggests that things like ADHD, emotional dysregulation, or conduct issues can derail cognitive development even in kids with high potential. This supports early intervention for behavioral issues being crucial for long-term outcomes.
