Purpose:
Estimate visual impairment (VI) prevalence among US adults NHANES2001-2006 participants aged >=12 y. To assess associationbetween weight gain and VI.
Methods:
A study cohort of 29489 aged >=12 y in NHANES 2001-2006 survey.Visualimpairment defined as presenting visual acuity of 20/50 or worsein better seeing-eye. Persons with missing presenting or correctedvisual acuity were excluded from analysis unless 1 eye had presentingvisual acuity of 20/40 or better in which case participantswere not visually impaired.SAS version 9.1 was used to computeunbiased weighted prevalence estimates for subpopulations ofinterest with minimum sample size requirements taking into accountcomplex sample design and non-response.
Results:
Logistic regression modeling VI and body mass index showed diabetesmellitus, 1.5 (1.07-2.01), less than high school, 1.6 (1.35-1.93),below poverty line, 1.34 (1.03-1.76) race/ethnicity [Blacks1.5 (1.16-1.88); Mexican-Americans 1.5 (1.16-1.87)], ≥60 years,1.8 (1.38-2.21) were significant risk factors for VI while gender,high blood pressure, hyperlipidemia and health insurance werenot. Weight gain was not significantly associated with visualimpairment. About 13 million persons in the US population [6.3%(5.8-6.9)] are visually impaired; 10% (8.5-11.3) Mexican Americans,8.5% (7.2-9.8) non-Hispanic Blacks and 5.3% (4.8-5.8) non-HispanicWhites.
Conclusions:
From analysis, we did not find direct effect on visual impairmentby BMI. Weight gain has an indirect effect on visual impairmentthrough other outcomes that involve diabetes mellitus, highblood pressure and hyperlipidemia. We will conduct further studyto determine if there is significant indirect effect.
Keywords: low vision • diabetes • refractive error development