Educational attainment, health outcomes and mortality: a within-sibship Mendelian randomization study

Abstract

Background: Previous Mendelian randomization (MR) studies using population samples (population MR) have provided evidence for beneficial effects of educational attainment on health outcomes in adulthood. However, estimates from these studies may have been susceptible to bias from population stratification, assortative mating and indirect genetic effects due to unadjusted parental genotypes. MR using genetic association estimates derived from within-sibship models (within-sibship MR) can avoid these potential biases because genetic differences between siblings are due to random segregation at meiosis.

Methods: Applying both population and within-sibship MR, we estimated the effects of genetic liability to educational attainment on body mass index (BMI), cigarette smoking, systolic blood pressure (SBP) and all-cause mortality. MR analyses used individual-level data on 72 932 siblings from UK Biobank and the Norwegian HUNT study, and summary-level data from a within-sibship Genome-wide Association Study including >140 000 individuals.

Results: Both population and within-sibship MR estimates provided evidence that educational attainment decreased BMI, cigarette smoking and SBP. Genetic variant-outcome associations attenuated in the within-sibship model, but genetic variant-educational attainment associations also attenuated to a similar extent. Thus, within-sibship and population MR estimates were largely consistent. The within-sibship MR estimate of education on mortality was imprecise but consistent with a putative effect.

Conclusions: These results provide evidence of beneficial individual-level effects of education (or liability to education) on adulthood health, independently of potential demographic and family-level confounders.

Keywords: Mendelian randomization; Within-sibship; educational attainment; mortality.

Figure 1

Population stratification, assortative mating and indirect genetic effects. ‘Population stratification’ occurs when ancestry is associated with the allele frequency of the genetic variant (G) and the phenotype of interest, distorting the association between G and the phenotype. In the context of MR, population stratification could distort the association between G and educational attainment (EA) and/or the association between G and the outcome, both of which could lead to bias in MR. ‘Assortative mating’ occurs when a heritable phenotype influences mate choice, e.g. if individuals are more likely to select a partner with a similar EA. Assortative mating leads to correlations for parental genotypes related to assorted phenotypes, which in turn leads to correlations between otherwise independent genotypes in the offspring. For example, if two genetic variants G₁ and G₂ influence EA then assortative mating on EA will lead to correlations in offspring for the EA-increasing alleles of G₁ and G₂ even if the two alleles are unlinked (linkage disequilibrium = 0). ‘Indirect genetic effects’ occur when the genotypes of relatives (e.g. parents, siblings) influence the phenotypes of the index individual. For example, parents with a higher EA polygenic score may produce an environment for their offspring that is more conducive to learning than parents with a lower EA polygenic score. This has been previously illustrated by evidence that non-transmitted parental EA polygenic scores also associate with offspring phenotypes

Figure 2

Phenotypic educational attainment and health outcomes. Figure 2 shows associations between phenotypic educational attainment (years in full-time education derived using qualifications) and body mass index, smoking (cigarettes, measured in pack years), systolic blood pressure and mortality in the population and within-sibship models in UK Biobank and HUNT. Estimates for mortality are presented as hazard ratios with the rest of the estimates presented in standard deviation units. BMI, body mass index; SBP, systolic blood pressure

Figure 3

Educational attainment polygenic score (PGS), educational attainment and health outcomes. Figure 3 shows associations between an educational attainment PGS and education (measured educational attainment), body mass index, cigarette smoking (pack years), systolic blood pressure and mortality in the population and within-sibship models in UK Biobank and HUNT. Estimates for mortality are presented as hazard ratios per standard deviation increase in the polygenic score with the rest of the outcome estimates presented in standard deviation units

Figure 4

Mendelian randomization (MR) estimates of educational attainment on health outcomes from UK Biobank and HUNT. Figure 4 shows population and within-sibship MR estimates of the effect of educational attainment on body mass index, smoking (pack years of cigarette smoking), systolic blood pressure and mortality. These estimates were derived using the polygenic score association estimates in Figure 3 from the UK Biobank and HUNT studies. Estimates are presented in standard deviation units for body mass index, smoking and systolic blood pressure, and as hazard ratios for mortality

Figure 5

Mendelian randomization estimates of educational attainment on health outcomes using summary data from the within-sibship Genome-wide Association Study (GWAS). Figure 5 shows population and within-sibship Mendelian randomization estimates (inverse-variance weighted) of the effect of educational attainment on body mass index, cigarettes per day measured in ever smokers only, ever smoking and systolic blood pressure. These estimates were derived using GWAS summary statistics from a within-sibship meta-analysis GWAS of ≤18 studies. Estimates are presented in standard deviation units except for ever smoking (binary) where estimates are in terms of risk difference %. BMI, body mass index; CPD, cigarettes per day; SBP, systolic blood pressure