Why do some people develop alcohol use disorder (AUD) while others do not? The answer is rarely a simple one. Alcohol problems are shaped by factors including family history, trauma, culture, social norms, and access to care, with AUD affecting an estimated 400 million people worldwide (World Health Organization, 2025). It is a chronic, relapsing disorder and carries more stigma than most other mental health conditions (Kilian C. et al, 2021).
Despite its high prevalence, only one in six people affected by AUD receives treatment (Mekonen T. et al, 2020). Understanding the genetic and cognitive factors that contribute could help develop more targeted, personalised therapies. Also, recognising the disorder’s complexity may reduce stigma and encourage individuals to seek treatment.
Alcohol use disorder has an estimated heritability of approximately 50% (Verhulst B, et al 2015), meaning genetic factors contribute about half of a person’s risk for developing AUD, the other half coming from environmental factors. Many individual genes contribute to this risk, including those involved in impulsivity, alcohol metabolism and the neurobiology of reward. Genetic evidence has linked cognitive performance and AUD risk (Latvala M. et al, 2010), and some studies have reported associations between cognitive performance and later development of AUD (Gronkjaer M. et al, 2019). However, few studies have directly examined these links.
The present study aimed to examine exactly that, exploring the relationship between alcohol use disorder risk and distinct constructs of general intelligence (IQ), cognitive performance and educational attainment across national populations using complementary analytical approaches.
Genes contribute about half of a person’s risk for developing alcohol use disorder .
Methods
This study explores the association between IQ and alcohol use disorder (AUD) risk using three complementary approaches:
1. Epidemiological Analysis
Longitudinal data were used to examine the association between IQ and AUD risk in a Swedish conscription cohort of 573,855 males. Participants were included if they had completed IQ measures from a validated test battery at 18 years and had no prior diagnosed substance use disorder (SUD) including AUD.
IQ was categorised as low, medium or high in relation to the population mean, to enable clinical interpretation. Lifetime AUD status was based on clinical diagnosis or alcohol-related-death data.
2. Genetic Analysis
Recognising that expression and consequences of cognitive traits may vary across sociocultural contexts Mendelian randomisation (MR) was used to examine the causal relationship between genetic liability for cognitive performance and AUD risk, with summary statistics from genome wide association studies (GWAS) across cohorts from the US, UK, Australia and Finland. You can read more about MR in a previous blog by Crick, D (2023).
3. Polygenic Analysis
Data from an independent US sample (n = 5424) were explored for associations between cognitive performance polygenic scores (PGS) – an aggregate measure of genetic liability and AUD risk. You can read more about polygenic scores in a previous blog by Palmer, E (2023).
Results
1. Epidemiological Analysis
Low IQ was associated with a 64% higher lifetime AUD risk and a 43% higher risk after adjusting for diagnosis of depression, anxiety, attention-deficit/hyperactive disorder (ADHD), parental SUD and socio-economic status. Mediation analysis indicated that a small but significant proportion (14%) of IQ’s effect on AUD risk was mediated by educational attainment.
Sibling-pair analyses, which partially control for shared genetic and environmental factors, showed similar results. Within-sibling comparisons showed that compared with siblings with medium IQ, those with low IQ had a greater risk for AUD. Higher IQ did not appear to protect against AUD risk.
Together, these data suggest that lower IQ at 18 years is associated with greater AUD risk in a sample of Swedish males, when adjusting for the contribution of mental health and other socioeconomic factors.
2. Genetic Analysis
In the US and UK samples, MR analyses showed that lower genetic liability for cognitive performance (β = 0.11 [SE = 0.02]; P < 0.001) and lower educational attainment (β = 0.19 [SE = 0.01]; P < 0.001) were both associated with increased AUD. These effects were replicated in the Finnish cohort.
In the US and UK cohorts, when controlling for educational attainment, lower genetic liability for cognitive performance had no direct effect on AUD (β = -0.02 [SE = 0.02]; P = 0.22), but educational attainment did (β = 0.20 [SE = 0.02]; P < 0.001). Around 55% of the effect of cognitive performance on AUD was mediated by educational attainment (SE = 0.50; P = 0.27).
Analyses in the Finnish cohort data showed an opposite pattern, where the effects of lower genetic liability for cognitive performance on AUD were larger after accounting for educational attainment (β = 0.56 [SE = 0.05]; P < 0.001) and there was no direct effect of educational attainment on AUD risk (β = 0.04 [SE = 0.05]; P = 0.40).
Together these data suggest a causal association between genetic liability for cognitive performance and AUD risk. The mediating effect of educational attainment differed by national cohort, suggesting that sociocultural context may influence how educational attainment shapes the cognitive performance-AUD risk.
3. Polygenic Analysis
In a US sample, cognitive performance PGS were associated with 16.7% lower odds of an AUD diagnosis (OR = 0.83 [95% CI = 0.78 to 0.89]). This suggests that individuals with the highest cognitive performance PGS had lower odds of AUD compared to those with average PGS. However, those with lowest PGS did not have increased risk for AUD.
Lower IQ, genetic liability, and polygenic scores all link to higher AUD risk, with education partially explaining the effect.
Conclusions
Data from epidemiological, genetic and polygenic analyses show a robust association between IQ and AUD risk, across different measures of IQ and cognitive traits. A mediating effect of educational attainment on the association between IQ and AUD risk was found, but varied between cohorts, suggesting the influence of national and sociocultural factors.
Together these data suggest that further exploration of the complex interplay between genetic and environmental factors in susceptibility to AUD is warranted.
Educational attainment was a mediating factor on the association between IQ and alcohol use disorder risk.
Strengths and limitations
A major strength of this work is the use of complementary analyses, across several different national populations. Triangulating epidemiological, genetic and polygenic data, across several large samples enhances confidence in the observed association between IQ, cognitive performance and AUD risk.
A limitation of the longitudinal data is the use of an entirely male sample. This is particularly relevant to the study’s focus, given that rates of AUD (Verplaetse T.L. et al 2025) and levels of educational attainment (Hadjar A. et al 2014) often differ between males and females. This gender imbalance potentially limits the generalisability of the epidemiological findings to females. However, the PGS data showed similar findings in a mixed-gender sample, suggesting generalisability of results beyond a male sample.
Conventional measures of IQ have been the focus of a longstanding debate about the nature of human intelligence, including whether IQ scores fully encompass critical cognitive domains (Stanovich K.E. et al 2014). The Swedish cohort study of almost half a million males solely relied on a standardised IQ test battery to assess general intelligence. These tests arguably focus on a narrow component of mental functioning, missing the contribution of emotion and motivational processes to cognition (Ganuthula V.R.R. et al, 2025). IQ measures have also been critiqued for their poor ability to predict real-world decision-making and reasoning (Stanovich K.E, 2009). The present study addresses this limitation by triangulating data across measures of IQ and cognitive performance that vary in validity and rigour.
A common shortcoming of many genetic studies, including this one, is the exclusive use of populations with European ancestry. The authors indicated that this limitation is due to poor availability of data from populations with genetic diversity. This limitation prevents the generalisation of findings to other groups, and the identification of biomarkers for AUD risk in those of non-European ancestry. As recommended by the authors and other researchers alike (Fatumo S. et al 2023), this highlights the importance of conducting genomic studies in diverse global populations.
Triangulated methods increase confidence in the IQ-AUD link, but gaps in gender balance and ancestry limit how widely they apply.
Implications for practice
The present data add to our knowledge of the many individual genes that contribute to AUD risk. Beyond the usual suspects of genes involved in the metabolism of alcohol and neurobiological processes in reward and impulsivity, these data identify cognitive traits as an pathway contributing to AUD risk.
Importantly, these data suggest that the association between IQ and alcohol use disorder risk is not a straightforward causal relationship. The authors state:
The findings suggest that causal pathways from cognitive performance to alcohol use outcomes are not universal and that structural conditions can magnify or mitigate the effects of genetic predispositions.
Indeed, their data show that educational attainment, a sociocultural factor shaped by access to and attendance at educational establishments, parental education, poverty, use of digital technology and one’s experience of school (Department of Education, 2024) appears to influence how cognitive traits impact AUD risk. These findings suggest that efforts to reduce AUD risk should not solely focus on individual differences in IQ and cognitive performance. Further research should explore the sociocultural conditions that influence the broader environment in which genetic liability is expressed.
These data also suggest that the contribution of educational attainment to the genetic liability of IQ and cognitive performance on AUD risk is dependent on national context. In countries such as Finland and Sweden, who provide more equitable access to education, educational attainment may reduce the risk that IQ and cognitive performance have on AUD vulnerability. In contrast, in countries such as the US and UK, who have more stratified societies, educational attainment may increase the vulnerability that cognitive traits confer, potentially via differences in access to and opportunity for education.
Together these data suggest that there is not a uniform causal pathway between cognitive traits and alcohol use disorder risk. That is, individual differences in cognitive traits are not deterministic in conferring genetic susceptibility to AUD. Rather, the findings suggest that the broader societal factor of educational attainment can mitigate or amplify this relationship.
The findings from this study have implications for policies aimed at improving equitable access to education as part of wider strategies to reduce AUD risk. Perhaps most importantly these data suggest that the impact of IQ and cognitive performance on AUD risk is not the result of a predetermined ability but instead a complex interplay between genetic liability and structural conditions.
These findings add to the argument against AUD as a fixed trajectory, where intervention is not possible. By highlighting the complexity of AUD risk, the present work has the potential to help reduce stigma and improve agency for individuals living with AUD.
The impact of IQ and cognitive performance on alcohol use disorder risk is not the result of a predetermined ability, but instead a complex interplay between genetic liability and societal conditions.
Statement of interests
Dr Sally Adams has no conflicting interests in respect of content of this blog post.
Editor
Edited by Éimear Foley. AI tools assisted with language refinement and formatting during the editorial phase.
Links
Primary paper
Capusan, Andrea Johansson, Davis, Christal N., Thern, Emelie, Rehm, Jürgen, Gelernter, Joel, Kranzler, Henry R., & Heilig, Markus (2025) Measures of General Intelligence and Risk for Alcohol Use Disorder. JAMA Psychiatry 2025 82 (12)1195-1202. https://doi.org/10.1001/jamapsychiatry.2025.2689
Other references
Crick, D. Does what you eat affect how you feel? The Mental Elf. 2023
Department of Education, last accessed 09 Feb 2026. https://assets.publishing.service.gov.uk/media/66e4006e3f1299ce5d5c3e11/
Factors_influencing_secondary_school_pupils__educational_outcomes.pdf
Fatumo S. et al (2023) Diversity in Genomic Studies: A Roadmap to Address the Imbalance. Nature Medicine 2023 28 (2) 243–250. https://doi.org/10.1038/s41591-021-01672-4
Ganuthula V.R.R. et al (2025) The Looking Glass for Intelligence Quotient Tests: The Interplay of Motivation, Cognitive Functioning and Affect. Frontiers in Psychology 17(10) 2857. https://doi.org/10.3389/fpsyg.2019.02857
Gronkjaer M. et al (2019) Intelligence Test Scores Before and After Alcohol-Related Disorders- A Longitudinal Study of Danish Male Conscripts. Alcohol Clinical and Experimental Research 2019 43 (10) 2187-2195. https://doi.org/10.1111/acer.14174
Hadjar A. et al (2014) Gender and Educational Achievement. Educational Research 2014 56 117-125. https://doi.org/10.1080/00131881.2014.898908
Kilian C. et al (2021) Stigmatization of People with Alcohol Use Disorders: An Updated Systematic Review of Population Studies. Alcohol Clinical and Experimental Research 2021 45 (5) 893-1160. https://doi.org/10.1111/acer.14598
Latvala A. et al (2010) Genetic Origins of the Association Between Verbal Ability and Alcohol Dependence Symptoms in Young Adulthood. Psychological Medicine 2010 41 (3) 641-651. https://doi.org/10.1017/S0033291710001194
Mekonen T. et al (2020) Treatment Rates for Alcohol Use Disorders: A Systematic Review and Meta-Analysis. Addiction 116 2617-2634. https://doi.org/10.1111/add.15357
Palmer, E. Genetic risk for Tourette Syndrome and related conditions. The Mental Elf, 23 November 2023
Stanovich K.E. (2009) What intelligence tests miss: The psychology of rational thought. New Haven, CT: Yale University Press.
Stanovich K.E. et al (2014) What Intelligence Tests Miss. British Psychological Society, last accessed 09 Feb 2026. https://www.bps.org.uk/psychologist/what-intelligence-tests-miss
Verhulst B. et al (2015) The Heritability of Alcohol Use Disorders: A Meta-Analysis of Twin and Adoption Studies. Psychological Medicine 2015 45 (5) 1061-1072. https://doi.org/10.1017/S0033291714002165
Verplaetse T.L. et al (2025) Gender Difference in Alcohol Use Disorder Trends from 2009-2019: An Intersectional Analysis. Alcohol 2025 123 101-107. https://doi.org/10.1016/j.alcohol.2024.11.003
World Health Organization, last accessed 05 Feb 2026. https://www.who.int/news-room/fact-sheets/detail/alcohol
