Educational attainment is strongly linked to positive life outcomes, including higher income, better health, and increased marital stability. However, its impact on wellbeing, a broad concept encompassing happiness and life satisfaction, remains less clear, particularly when considering its high correlation with intelligence. Previous observational studies yielded mixed results, sometimes demonstrating both positive and negative associations depending on the inclusion of mediating factors like income and employment. These studies primarily focused on Australian samples, with limited research in other populations. This study addresses these gaps by utilizing both observational data and a novel Mendelian randomization (MR) approach to disentangle the independent causal effects of educational attainment and intelligence on wellbeing, while also examining potential sex differences.

The literature on the relationship between education and wellbeing shows inconsistent findings. Some studies highlight a positive association, mediated through increased income and improved employment prospects, particularly for men. Other studies suggest a negative association, primarily for women, when mediating factors aren't considered, suggesting that the benefits of education are realized through diverse mechanisms. Furthermore, the role of intelligence in this complex interplay has not been sufficiently explored, leaving the independent contribution of educational attainment unclear. The existing research lacks causal inference, relying on correlations that are vulnerable to confounding variables.

This study employed a two-pronged approach: Mendelian randomization (MR) and observational analyses. For the MR analysis, summary-level genetic data from genome-wide association studies (GWAS) were used for educational attainment, intelligence, and a composite wellbeing phenotype (combining life satisfaction, positive affect, neuroticism, and depressive symptoms). Univariable MR assessed the total causal effects, while multivariable MR examined the independent effects of educational attainment and intelligence on wellbeing, accounting for their correlation. Several MR methods (inverse variance weighted, MR-Egger, weighted median, weighted mode) were used to address potential biases. The observational analysis used data from the Avon Longitudinal Study of Parents and Children (ALSPAC), a large UK-based birth cohort study, analyzing the association between educational attainment (university degree completion), intelligence (WISC-III scores at age 8), and wellbeing (subjective happiness and life satisfaction at age 26). Linear and interaction models accounting for sex were used. Multiple imputation addressed missing data.

Univariable MR analysis showed a small positive causal effect of educational attainment on wellbeing and a positive effect of wellbeing on educational attainment. Multivariable MR revealed independent effects, with a positive causal impact of educational attainment on wellbeing (controlling for intelligence) and a negative causal impact of intelligence on wellbeing (controlling for educational attainment). The observational analysis largely supported the MR findings, showing that higher education was associated with increased life satisfaction, with a stronger effect for females. For subjective happiness, the association with education showed a sex interaction. Among females, higher education was associated with greater happiness, while the opposite was true for males. For intelligence, higher scores were negatively correlated with happiness in males, but not in females. The observational analyses showed a small positive correlation between intelligence and life satisfaction.

This study's findings provide novel causal evidence supporting the beneficial effect of educational attainment on wellbeing, independent of intelligence. The negative association between intelligence and wellbeing, observed after accounting for education, suggests that high intelligence may not always translate to higher wellbeing, especially for individuals who do not pursue further education. Sex differences highlighted the importance of considering gender-specific factors when evaluating the impact of education and intelligence on wellbeing. The stronger association between education and wellbeing in women might be explained by the greater impact of socialization opportunities offered by education on female happiness. These findings have significant implications for policymakers, suggesting that interventions focusing on keeping students engaged in education are crucial for maximizing wellbeing outcomes, particularly for males.

This study's combination of MR and observational data provides strong evidence for the independent and positive causal effect of educational attainment on wellbeing, which is partially sex-specific. It also reveals a potential negative impact of intelligence on wellbeing when educational attainment is low. Future research should explore the mechanisms underlying these effects, particularly sex differences and the role of mediating factors like social support and health behaviors. Furthermore, replication in diverse populations is necessary to ensure generalizability.

The MR analysis relies on GWAS data, which may not be fully representative of the general population. Assortative mating and dynastic effects could potentially inflate MR estimates. Observational analyses were limited by the binary measure of educational attainment (university degree vs. no degree) and the assessment of wellbeing at a single time point. The study's focus on a UK sample limits its generalizability to other populations.