Purpose
To report the prevalence of refractive error in an adolescent British cohort, and to examine the influence of cognition over childhood development on myopia risk
Methods
The Twins Early Development Study is a longitudinal cohort of 10,000 British twins enrolled at birth between 1994-96, who have been studied from a neurodevelopmental perspective using multivariate quantitative and molecular genetic techniques. A sample of 2,625 pairs (aged 16-18) was invited to participate (52% response rate) and consent requested to contact their optometrist for subjective refraction data (87% response rate). Myopia was defined as spherical equivalent (SE) ≤-0.75 D, high myopia if ≤-6.0 D and hyperopia as ≥1.0 D. Multivariable regressions and mixed effect models were used to assess demographic features and associations during childhood (adjusted for age, sex, ethnicity, maternal education & family structure).
Results
SE data was obtained on 1992 subjects. Responders reflected the UK population (95% European, 55% female). Mean SE was -0.35 D (SD 1.80). In those aged 16-18, 32.2% (95% CI 30.0-34.9) were myopic, 1.6% (95% CI 1.0-2.4) were highly myopic and 8.4% (95% CI 7.0-9.9) were hyperopic. Significant odds ratios (OR) for myopia were identified for age (1.72), non-white ethnicity (3.06) and maternal education (1.13). In a mixed effect model of composite scores of cognitive ability (g) at 2,3,4,7,10,12,14 & 16, the association with myopia was greatest in late adolescence (age 16: β -0.287 p=0.03). Association between upper quartile cognitive ability and myopia tended to become significant with age: age 4 (OR 1.15 p=0.51), age 7 (OR 1.45 p=0.11), age 10 (OR 1.68 p=0.05) and age 16 (OR 2.11 p=0.003). Verbal cognition (VC) was consistently more associated with myopia than non-verbal cognition (NVC), with greatest disparity and effect at younger ages (VC age 10 OR 1.26 [p=0.03] vs NVC OR 1.02 vs VC OR age 16 1.05). Bivariate twin modelling suggested shared genetic effects underlying SE and g.
Conclusions
Myopia prevalence in a UK-representative population of 16-18 year-olds is 32%. This suggests subsequent prevalence in adulthood may be higher than previous estimates, congruous with a cohort effect. A consistent and significant association of cognition for myopia was identified, with verbal cognition in early childhood and overall cognitive ability in adolescence most strongly associated with myopia risk.
Keywords: 605 myopia •
463 clinical (human) or epidemiologic studies: prevalence/incidence •
464 clinical (human) or epidemiologic studies: risk factor assessment