Work

Is genetic liability to ADHD and ASD causally linked to educational attainment?

Year: 2021

Type: article

Abstract: ABSTRACT Background Individuals with Attention Deficit Hyperactivity Disorder (ADHD) or Autism Spectrum Disorder (ASD) are at risk of poor educational outcomes. Parental educational attainment has also been associated with risk of ADHD/ASD in the offspring. Despite evidence that ADHD and ASD show genetic links to educational attainment, less is known on the causal nature of the associations and the possible role of IQ. Methods We assessed the total causal effects of genetic liability to ADHD/ASD on educational attainment using two-sample Mendelian randomization (MR). We assessed the possible contribution of IQ to the identified causal effects by estimating the “direct” effects of ADHD/ASD on educational attainment, independent of IQ, using Multivariable MR (MVMR). Reverse direction analyses were performed. The latest GWAS meta-analyses of ADHD, ASD, educational attainment and IQ were used. Causal effect estimates were generated using inverse variance weighted models (IVW). Sensitivity analyses were performed to assess the robustness of the estimates and the presence of pleiotropy. Results Genetic liability to ADHD had a total ( MR IVW:-3.3 months per doubling of liability to ADHD; 95% CI: -4.8 to -1.9; pval= 5*10 −6 ) and direct negative causal effect on educational attainment ( MVMR IVW:-1.6 months per doubling of liability to ADHD; 95% CI: -2.5 to -0.6; pval= 4*10 −4 ). There was little evidence of a total causal effect of genetic liability to ASD on educational attainment ( MR IVW: 4 days, per doubling of liability to ASD; 95% CI: -4.9 months to 5.6 months; pval= 0.9) but some evidence of a direct effect not via IQ ( MVMR IVW:29 days per doubling the genetic liability to ASD; 95% CI: 2 to 48; pval= 0.03). Reverse direction analyses suggested that genetic liability to higher educational attainment was associated with lower risk of ADHD ( MR IVW OR : 0.3 per standard deviation (SD) increase; 95% CI: 0.26 to 0.36; pval= 6*10 −51 ), even after IQ was entered in the models ( MVMR IVW OR: 0.33 per SD increase; 95% CI: 0.26 to 0.43; pval= 6*10 −17 ). On the contrary, there was evidence consistent with a positive causal effect of genetic liability to higher educational attainment on risk of ASD ( MR IVW OR : 1.51 per SD increase; 95% CI: 1.29 to 1.77; pval= 4*10 −7 ), which was found to be largely explained by IQ ( MVMR IVW OR per SD increase: 1.24; 95% CI: 0.96 to 1.60; pval= 0.09). Conclusions Our findings suggest that despite the genetic and phenotypic overlap between ADHD and ASD, they present highly differentiated causal associations with educational attainment. This highlights the necessity for specialized educational interventions for children with ADHD and ASD. Further research is needed in order to decipher whether the identified causal effects reflect parentally transmitted effects, diagnostic masking, or selection bias.

Source: International Journal of Epidemiology

DOI: 10.1093/ije/dyab107