Relationship of 10-year change in refraction to nuclear cataract and axial length

Purpose: To examine 10-year changes and other influences on spherical equivalent refraction (SER) in older persons. Design: Prospective population-based study. Participants: Three thousand six hundred fifty-four Blue Mountains Eye Study participants 49 or older at baseline (1992-1994) were observed after 5 years (2335; 75% of survivors) and 10 years (1952; 76% of survivors). Methods: At each visit, subjective refraction was performed using modified Early Treatment Diabetic Retinopathy Study protocols. Spherical equivalent refraction was calculated as sphere + half cylinder power. Axial length was measured only at the 10-year examinations, using an IOL Master. Right phakic eyes with best-corrected visual acuity > 20/40 (n = 1340) at baseline and 10-year examinations were included. Main Outcome Measures: Temporal refractive change. Results: Over the decade, a hyperopic shift was observed among persons < 65 years old, and a myopic shift was associated with increasing age among older subjects, with gender-adjusted mean SER changes of 0.40, 0.33, -0.02, and -0.65 diopters (D) in persons with baseline ages 49 to 54, 55 to 64, 65 to 74 and >= 75, respectively. Apart from age, myopic refractive change was strongly associated with baseline nuclear cataract; mean changes of -0.96 and 0.26 D were observed in eyes with and without nuclear cataract, respectively (P < 0.001). A birth cohort effect on refraction was also observed. After adjusting for age and nuclear cataract, baseline refractive status (P = 0.58), education (P = 0.34), and diabetes (P = 0.16) were not associated with changing SER. A 10% increase in against-the-rule astigmatism was observed over the 10-year period. Axial length showed an age-related reduction in 10-year cross-sectional data, from a mean of 23.61 mm (95% confidence interval [CI], 23.50-23.73) in 59- to 64-year-olds to a mean of 23.15 mm (95% CI, 22.83-23.47) in >= 85-year-olds. After adjusting for age, education, and nuclear cataract, axial length measured 10 years later was not associated with change in SER (P = 0.34). Conclusions: This longitudinal study confirms a hyperopic shift in persons younger than 65 and a myopic shift for older ages. Although underlying causes for this age-related hyperopic shift are unknown, it does not appear related to axial length. The myopic shift, however, is most likely caused by increasing nuclear cataract.