Abstract

Aims: To evaluate the relationship between best corrected visual acuity (BCVA), age, type of diabetes, sight-threatening diabetic retinopathy (STDR) and ocular co-morbidity.

Methods: 1549 randomly selected people with diabetes mellitus (DM) from a countywide digital photographic screening programme had standardised logarithm of minimum angle of resolution (logMAR) BCVA measurement, followed by slit-lamp biomicroscopy examination by an experienced ophthalmologist.

Results: Subnormal vision (logMAR ⩾0.3, Snellen ⩽6/12) and blindness (logMAR >1.3, Snellen <3/60) in the better-seeing eye were found in 9.0% and 0.45%. The sensitivity, specificity and positive and negative predictive values of using subnormal vision to screen for STDR in an individual eye were 33.4%, 85.9%, 18.6% and 93.0%, respectively. Important contributory causes of moderate visual loss (logMAR 0.50 to 0.98, Snellen 6/18 or worse but better than 6/60) and of Acuity Blindness (logMAR ⩾1.0, Snellen 6/60 or worse) in an individual eye were lenticular opacity (including capsular opacification) 49%, macular degeneration (including myopic degeneration) 29%, diabetic maculopathy 15%, other media causes (including corneal opacity) 13% and amblyopia 10%.

Conclusion: The majority of visual loss in a population with diabetes is due to causes other than diabetic retinopathy. BCVA alone is not a reliable criterion in predicting STDR.