|Author, Co-Author||Karla Zadnik, Donald Mutti, The Longitudinal Evaluation Of Ethnicity And Refractiv|
Northern Hemisphere E 1-2
|Abstract|| PURPOSE. Many different criteria are used by investigators to define myopia, both clinically and for research purposes. We sought to determine the effect of different criteria on the estimated proportion of myopia.
METHODS. We examined the effect of measurement method, no cycloplegia vs. cycloplegia, and the numeric cutoff for myopia in a sample of eighth graders from the CLEERE Study. We also tabulated the varying criteria for myopia across clinical research laboratories.
RESULTS. There is a marked effect of measurement method and the numeric cutoff for myopia on the proportion of myopia. The proportion of myopes in the CLEERE dataset (8th graders in 1991-1998) varies from 18.2% (cycloplegic autorefraction: at least -0.75 DS of myopia) to 36.9% (non-cycloplegic autorefraction measures: at least -0.50 DS of myopia). The May/June 1999 issues of Optometry and Vision Science include 13 patient-based studies on myopia. There is very little consensus on such important study design issues as entry criteria, method of measurement, or categorization of myopia. Only 5 of the 13 studies agree on an entry criteria of at least -0.50 D of myopia. Among the 10 papers, most express outcome data as spherical equivalent, but the wisdom of excluding astigmatism may be questioned. Nine studies use no cycloplegia, but, again, this may not be the ideal methodological decision.
CONCLUSIONS. The choice of measurement methods can alter the proportion of myopes in a given study. Investigators use a wide range of myopia definitions in studies of human myopia. Clearly, the worldwide myopia community needs to enter into a discussion of the most appropriate measurement methods and definition of myopia in an attempt to unify research in this area and to strengthen the interpretability of results. Supported by: NEI & Office of Minority Research/NIH, U10-EY08893, R21-EY12273, Ohio Lions Eye Research Foundation, & the EF Wildermuth Foundation.
|Affiliation of Co-Authors||Ohio State University|