Lower IQ in youth predicts higher alcoholism risk in adulthood

A large-scale study published in JAMA Psychiatry suggests that lower intelligence scores in late adolescence are linked to a higher risk of developing alcohol use disorder later in life. The research, which combined data from Sweden, the United States, Finland, and other countries, also provides evidence for a possible genetic pathway connecting cognitive performance to alcohol-related problems. However, the strength and direction of this association appear to vary depending on the social and cultural environment.

Alcohol use disorder, like many mental health conditions, runs in families. Estimates suggest that genetics account for about half the risk of developing the disorder. Despite this, most genetic variants identified so far explain only a small share of that risk. While some genetic research has highlighted the role of impulsivity and alcohol metabolism, researchers remain uncertain about how broader cognitive traits, like intelligence, may factor into alcohol misuse.

Lower intelligence test scores are often found in individuals with alcohol problems, but the direction of this relationship is unclear. It is possible that lower cognitive ability increases vulnerability to alcohol use disorder. But it’s also possible that alcohol itself impairs thinking over time. This study aimed to clarify whether lower cognitive performance predicts the risk of alcohol use disorder and whether this connection is shaped by genetics or by environmental factors like education and socioeconomic background.

“I have been active in the field of addiction for decades, and taken for granted that low socioeconomic status – poverty – contributes to the risk for developing addiction, as a causal factor. I am a clinician and a neuroscientist, not an epidemiologist, and did not pay attention to how SES is measured in the epidemiological studies,” explained study author Markus Heilig, a professor, Wallenberg Clinical Scholar, and director of the Center for Social and Affective Neuroscience at Linköping University.

“I then was invited to collaborate with one of the co-authors on the current paper, Jurgen Rehm, who is one of the leading epidemiologists in this area. In the course of the collaboration, I realized that SES is typically measured as educational attainment. It is well established that educational attainment is highly correlated with IQ.”

Free daily newsletter

“In fact, in many studies, it is used as a proxy for IQ. IQ, in turn, is known to be highly heritable, upwards 80%. So it was clear that when we thought we looked at the relationship between SES and addiction risk, we might in fact indirectly be looking at an influence of IQ.”

To explore this question, researchers used several different methods. First, they examined a cohort of over 570,000 Swedish men who were assessed for IQ at age 18 during mandatory military conscription between 1969 and 1984. These intelligence scores were grouped into low, medium, and high categories. The men were then followed for several decades, with data on alcohol-related diagnoses and deaths gathered from national health and population registers.

The researchers found that men with lower IQ scores at age 18 were significantly more likely to be diagnosed with alcohol use disorder later in life. Specifically, those in the low IQ group had a 64 percent higher risk compared to those with average scores. Even after adjusting for factors such as parental substance use, attention-related or emotional disorders, and childhood living conditions, the elevated risk remained at 43 percent. Those with high IQ scores had a 40 percent lower risk.

“At about 50% elevated risk for someone 15 IQ points down from the population average or less, I guess this is a moderately sized risk factor – lesser than early life traumatization, greater than a lot of other things we worry and talk about,” Heilig told PsyPost.

Sibling comparisons further supported this pattern. In families with two or more brothers, the one with the lower IQ score had a greater chance of developing alcohol use disorder, even when growing up in the same household environment. These findings suggest that individual differences in cognitive ability may shape vulnerability to alcohol-related problems, beyond shared family background.

The study also looked at whether education might explain some of the link between intelligence and alcohol use disorder. Educational attainment did appear to play a role, but only accounted for about 14 percent of the association. Other factors like depression and attention problems had much smaller effects.

To test whether the association was likely to be causal rather than due to confounding, the researchers also performed genetic analyses. Using large datasets from genome-wide studies, they examined whether genetic variants associated with cognitive performance were also related to alcohol use disorder. They used a method called Mendelian randomization, which leverages the random inheritance of genes to estimate causal relationships.

These genetic analyses showed that people with a higher genetic likelihood for strong cognitive performance were less likely to develop alcohol use disorder. The effect was consistent across different analytical models and replicated in separate Finnish data. However, the strength and direction of the effect varied depending on the social context. In a sample that combined data from the United States, the United Kingdom, and Australia, the impact of cognitive performance on alcohol disorder risk was largely explained by educational attainment. In the Finnish data, education appeared to buffer the effects of lower cognitive ability rather than mediate them.

In a final step, the researchers tested whether individuals with a genetic profile linked to better cognitive performance were less likely to be diagnosed with alcohol use disorder in a U.S. sample. They found that people with the highest scores on this genetic profile had about a 35 percent lower chance of receiving an alcohol use disorder diagnosis compared to those with average genetic scores.

Heilig said the consistency of the findings was a surprise: “We’re looking at a very complex web of factors, that operate over decades of life. Yet, every way we looked – a cohort of sibling pairs assessed at age 18, as well as two very different kinds of very complicated genetic analyses in multiple population from the United States and Europe, the same result always came out.”

“We need to identify kids at risk, who are failing in school early, and help them avert the risk,” Heilig said. “They need special ed to develop a level of theoretical skills, but also alternative tracks that do not heavily emphasize those, but rather allow them to develop marketable skills using other strengths they may have. And that these kids need not only these measures, that focus on education, but also specific preventive measures when it comes to the risks of using alcohol and drugs.”

“And, finally, if addiction nevertheless develops, treatment measures need to be adapted to the patient’s capacity – cognitive-behavioral therapy is great if you have the resources to understand and use it, but otherwise sets you up for failure. Which, in its own terrible way, is a teachable experience in itself.”

Although the study used robust methods and multiple large datasets, the authors acknowledge some limitations. The Swedish data included only men, so findings may not generalize to women. Also, the Swedish diagnosis data likely captured only the most severe alcohol problems, possibly underestimating milder forms of the disorder.

In the genetic analyses, the cognitive measures were based on a mix of tests from different studies. This may introduce some inconsistencies in how cognitive ability was defined. Additionally, the genetic studies included mostly individuals of European descent. As a result, the findings may not apply to populations with different genetic backgrounds, and the authors stress the need for more diverse genetic research in the future.

Another challenge in interpreting the results is distinguishing correlation from causation. While Mendelian randomization reduces the risk of certain biases, it cannot rule out all alternative explanations. However, the consistency of the findings across different methods and populations strengthens the case for a genuine connection between cognitive ability and alcohol use disorder risk.

The researchers emphasize that these findings should not be used to stigmatize individuals with lower cognitive performance. Cognitive ability is only one of many traits that can influence risk, and it interacts with a wide range of environmental and social factors.

“Stigma is one of the most destructive phenomena surrounding addiction,” Heilig explained. “There is an obvious risk that our findings, if misinterpreted, could worsen that. I’ve spent a lifetime working to improve the lives of people with addiction, so this concerns me. Remember, people are different in so many various ways. Right now, society happens to emphasize cognitive performance, but that is a rather recent phenomenon. In other times, it has been other things. We need to keep our eyes open for the diversity of human beings, but we also need to have a firm ethical foundation that embraces that diversity. This is no different from other types of diversity of humans.”

Looking ahead, the research team plans to investigate how cognitive traits may influence risk for other life challenges, especially in environments where individuals with lower cognitive performance may face hidden disadvantages. They also hope that these insights will inform more targeted approaches to prevention and treatment of alcohol use disorder, including strategies that account for cognitive strengths and vulnerabilities.

The study, “Measures of General Intelligence and Risk for Alcohol Use Disorder,” was authored by Andrea Johansson Capusan, Christal N. Davis, Emelie Thern, Jürgen Rehm, Joel Gelernter, Henry R Kranzler, and Markus Heilig.