objectives. This study estimated the prevalence of self-rated visual impair
ment among US adults with. diabetes and identified correlates of impairment
.
Methods. Self-reported data from the 1995 Behavioral Risk Factor surveillan
ce System survey of adults 18 years and older with diabetes were analyzed C
orrelates of visual impairment were examined by multiple logistic regressio
n analysis.
Results. The prevalence of self-rated visual impairment was 24.8%; (95% con
fidence interval [CI] = 22.3%, 27.3%). Among insulin users, multivariable-a
djusted odds ratios were 4.9 (95% GI = 2.6, 9.2) for those who had not comp
leted. high school and 1.8 (95% CI = 1.0, 2.8) for those who had completed
high school compared with those with higher levels of education Ea Comparab
le estimates of odds ratios for non-users of insulin were 22 (95% CI:= 1.4,
3.4) and 1.3 (95% CI- 0.9, 2.0) respectively. Among nonusers the adjusted
odds for minority adults were 2.4 (95% CI = 1.0, 3.7) times the odds for no
nHispanic Whites.
Conclusions. By these data, 1.6 million US adults with diabetes reported ha
ving some degree of visual impairment. Future research:li on the specific c
auses of visual impairment may help in estimating the avoidable public heal
th burden.