Identification of loci involved in childhood visual acuity and associations with cognitive skills and educational attainment

Abstract

Visual acuity significantly contributes to quality of life. Deficits in childhood are associated with reading difficulties, which can have detrimental effects on education outcomes. In adults, it has been observed that vision defects such as myopia are associated with higher educational attainment (EA). Understanding genetic factors contributing to visual acuity could help to dissect its links with cognitive skills, neurodevelopmental conditions, and education. We examined associations between distance visual acuity, cognitive measures including school grades, and neurodevelopmental conditions in a longitudinal cohort of British children (ALSPAC, n = 6807, M age = 11.8). We performed a genome-wide association study (GWAS, n = 5571) on visual acuity and tested for genetic associations with relevant phenotypes using polygenic scores (PGS) and genetic correlation analyses. Visual acuity was associated with better cognitive performance and school grades, and reduced in individuals with reading difficulties compared to controls. GWAS revealed genetic associations at the NPLOC4 locus and highlighted other genes involved in sensory function. In line with positive genetic correlations between visual acuity and cognitive measures, EA PGS were positively associated with visual acuity, while there was a less robust negative association with myopia PGS. In conclusion, increased visual acuity is associated with a range of positive outcomes, including better school grades. Our results suggest an association between a higher EA PGS and slightly increased visual acuity in childhood. This could indicate gene-environment correlation, in which environmental exposures linked to higher EA might have detrimental effects on vision offsetting the initial positive effect.

Fig. 1. Associations between visual acuity, cognitive skills, and adult myopia.

There is a negative link between visual acuity and myopia risk in childhood (not shown). Therefore, we expect that genetic factors known to increase myopia risk (myopia PGS) are negatively associated with visual acuity (black dotted arrow). However, visual acuity has been associated with increased cognitive skills, contributing to more years spent in education, which have been shown to increase myopia risk (red solid arrows). Therefore, we expect a positive association between the PGS for educational attainment (EA) and childhood visual acuity (red dotted arrow). The discrepancy of EA PGS being associated positively with childhood visual function and EA being associated negatively with adult visual function (in that it increases myopia risk) suggests gene-environment correlation. Specifically, the same factors associated with better vision earlier (EA PGS) are also associated with environmental exposures with detrimental effects on vision. Arrows represent a positive association unless otherwise indicated with a minus symbol (-).

Fig. 2. Distribution of visual acuity.

a Distribution of visual acuity (better eye) in the overall sample (n = 6807), and b as a function of sex (n_females = 3444; n_males = 3356).

Fig. 3. Correlation coefficients for visual acuity and cognitive measures.

a Behavioural correlations. b Genetic correlations in unrelated individuals. Correlations are shown if passing the Bonferroni-corrected significance level (0.05/8 = 0.00625).

Fig. 4. Visual acuity subgroup analyses.

a grouped by SES and b grouped by neurodevelopmental conditions. Error bars of violin plots represent standard deviations (s.d.). ^***p < 0 .001, ^**p < 0.01, ^*p < 0.05 in Tukey post-hoc tests after adjustment for multiple comparisons.

Fig. 5. Manhattan plots for visual acuity.

a SNP-based and b gene-based association p values are plotted against chromosome and position. SNP-based p values origin from linear mixed models in BOLT-LMM. Gene-based p values origin from re-weighting of SNP-based p values in MAGMA. The solid line represents the genome-wide significance level (a: p = 5 × 10⁻⁸, b: p = 2.7 × 10⁻⁶), the dotted line represents a suggestive significance level (a: p = 1 × 10⁻⁵, b: p = 1 × 10⁻⁴), respectively.