The incidence of visual impairment and its determinants in the WHO Multinational Study of Vascular Disease in Diabetes.

Aims/hypothesis: Incidence of severe visual impairment and the ultimate prevalence of all grades of impairment were estimated in the 10 centres of the WHO Multinational Study of Vascular Disease in Diabetes (WHO MSVDD) participating in the follow-up.

Methods: Visual function was ascertained at follow-up in 2994 (77.9 %) of the 3845 eligible participating survivors of the 4709 originally recruited for the WHO MSVDD using the same baseline enquiry method. The associations between incident severe visual impairment, follow-up prevalence of all grades of impairment and baseline risk factors were examined by univariate and stepwise multiple logistic regression analysis.

Results: Overall, 8.4 year incidence of severe visual impairment was 1.94 % and showed statistically significant univariate correlations with age at diagnosis, diabetes duration, systolic blood pressure, fasting blood glucose and cholesterol, insulin treatment and strongly with baseline retinopathy. Baseline retinopathy, systolic pressure and cholesterol were statistically significant in multivariable analysis. Differences between centres (0.3% to 3.45%) were not significant. Ultimate prevalence of all grades of impairment differed between centres and within almost all of them was correlated in multivariable analysis with baseline retinopathy and proteinuria.

Conclusion/interpretation: Comparisons of incident severe visual impairment between centres are restricted by selective mortality, low incidence rates and relatively small numbers in each centre but before retinopathy, baseline systolic pressure and cholesterol predicted severe visual impairment. Follow-up prevalence of all degrees of impairment varied among centres and were associated with prior retinopathy and renal disease at baseline.