Prevalence of Visual Impairment in the United States FREE

Sight is an important indicator of health and quality of life.^1- 2 Individuals with impaired vision may be at increased risk of injuries^3- 4 and are more likely to curtail or stop driving.^5- 6 Visually impaired elderly persons are at increased risk of falls and fractures^7- 10 and depression.¹¹ The most common reason for impaired vision may be refractive error,^12- 13 a condition usually correctable with eyeglasses or contact lenses.

Vision was assessed in the First National Health and Nutrition Examination Survey (NHANES I) (1974-1975)¹⁴ and in the Hispanic Health and Nutrition Examination Survey (1982-1984).¹⁵ Since then, the US population has aged and become more racially and ethnically diverse¹⁶ and rates of myopia (nearsightedness) have increased worldwide.¹⁷ Accordingly, visual acuity assessment was reinstituted as an NHANES component in 1999. We estimated the number of US individuals with impaired vision aged 12 years or older and the proportion of individuals who can achieve good distance vision with refractive correction.

Data from NHANES 1999-2002 were used.¹⁸ NHANES is a nationally representative survey of the US civilian population, conducted on an ongoing basis by the National Center for Health Statistics, Centers for Disease Control and Prevention. Each year, approximately 7000 individuals are interviewed in their homes and are invited to undergo a comprehensive health examination in a mobile examination center (MEC), comprising functional assessments, collection of biological samples, and laboratory test results.¹⁸ Ethnicity is assessed in NHANES to provide estimates of health conditions within specific ethnic groups. The NHANES 1999-2002 protocol was reviewed and approved by the National Center for Health Statistics research ethics review board. Written informed consent was obtained from all participants.

Survey participants provided data on age, sex, ethnicity, years of schooling, income, and health insurance coverage,¹⁶ using response options defined by the NHANES investigators. Visual acuity and automated refraction were measured for each eye using the ARK-760 (Nidek Co Ltd, Tokyo, Japan), an autorefractor containing built-in visual acuity charts with 20/20, 20/25, 20/30, 20/40, 20/50, 20/60, 20/80, and 20/200 lines.

Presenting visual acuity was assessed using the participant's usual distance vision correction, if any. The 20/50 line was presented first. At least 4 of the 5 characters (numbers, letters, or both) had to be read correctly to advance to the next line (with smaller characters). If the participant was unable to read the 20/50 line, the 20/200 line was presented. Participants were asked to read each line until they missed more than 1 character per line, for 2 lines in a row. Visual acuity was recorded as the last line for which 4 or more characters were read correctly. For eyes with presenting visual acuity of 20/30 or worse, visual acuity was also measured aided by the automated refraction result (corrected visual acuity). Simple (character recognition) literacy was required for visual acuity testing. Visual acuity was not tested in participants who were too cognitively or physically impaired to undertake the vision examination.¹⁹ The autorefractor was calibrated weekly.

Visual impairment was defined as presenting distance visual acuity of 20/50 or worse in the better-seeing eye. We defined visual impairment due to uncorrected refractive error as (presenting) visual impairment that improved, aided by the automated refraction results, to 20/40 or better in the better-seeing eye. Persons with missing presenting or corrected visual acuity for either eye were excluded from analyses (unless 1 eye had presenting visual acuity of 20/40 or better, in which case the participant was classified as not having visual impairment). Participants who did not bring their distance corrective lenses (eyeglasses or contact lenses) to the MEC were classified as not using distance correction.

SUDAAN version 9.0.0 (Research Triangle Institute, Research Triangle Park, NC) was used to compute weighted prevalence estimates accounting for the multistage probability sampling design and nonresponse to the MEC examination.^20- 22 Estimates were age-standardized to the US 2000 Census population.²³

A total of 15 193 individuals aged 12 years or older participated in NHANES 1999-2002. Of these participants, 14 203 (93.5%) reported to the MEC and 13 265 (93.4%) had complete visual acuity data (Figure). Those participants with incomplete visual acuity data were more likely to be older, black, poor, and have fewer years of education, and were less likely to have private health insurance compared with those with complete visual acuity data (Table 1).

Overall, 1190 NHANES participants presented with visual impairment (weighted prevalence, 6.4%; 95% confidence interval [CI], 6.0%-6.8%) (Table 2). Rates of visual impairment were highest among persons who were Hispanic or other (includes Asian or mixed race) ethnicity, were poor, had diabetes mellitus, lacked private health insurance, or had fewer years of education. The majority (83.3%; 95% CI, 80.9%-85.8%) of the 1190 persons with visual impairment could achieve good corrected visual acuity (20/40 or better in the better-seeing eye) using autorefractor correction. The prevalence of visual impairment due to uncorrected refractive error in the total NHANES population (n = 13 265) was 5.3% (95% CI, 4.9%-5.7%).

Extrapolating the NHANES findings to the general US population, approximately 14 million persons aged 12 years or older would be projected to have visual impairment (Table 3). Of these individuals, more than 11 million could achieve good visual acuity with refractive correction. Of the estimated 2 million persons with presenting visual acuity of 20/200 or worse in the better-seeing eye, 1.4 million could achieve good vision with refractive correction.

National surveys, such as the NHANES, provide data to track the health status of the US population for public health planning and policy purposes.²⁴ We estimated that 93.6% of individuals aged 12 years or older have good presenting visual acuity (20/40 or better in the better-seeing eye). Of the remaining 6.4% of persons with visual impairment, the vast majority, 83%, could achieve good visual acuity with refractive correction.

Our results are consistent with those of population-based studies conducted in specific US communities.^13,25 Other countries with demographic and socioeconomic profiles similar to the United States have also reported a high proportion of visual impairment due to uncorrected refractive error.^26- 28 Consistent with previous studies,^{13,25,27- 28} we defined visual impairment as visual acuity of 20/50 or worse, and good vision as visual acuity of 20/40 or better. Although 20/40 is the visual acuity criterion used to determine eligibility for a driver's license in most states, we are reluctant to call this normal vision.

We found that prevalence of visual impairment was higher in persons who were of black, Hispanic, or other ethnicity, or who were poor, less educated, or lacked private health insurance. Although these findings are not unexpected, they suggest that health care access and resources may be important barriers to consider in addressing the need for refractive correction of visual impairment.

Our analyses are subject to several limitations. The MEC vision examination was conducted by trained operators using standardized protocols under stringent quality control procedures.¹⁹ Nevertheless, visual acuity measured by an autorefractor is not equivalent²⁹ to visual acuity assessed in a clinical ophthalmic setting (precluded by logistical constraints in NHANES). We are unaware of studies validating the visual acuity measured using autorefractor charts and are therefore unable to address the accuracy of the NHANES measurements. Although similar autorefractors have been shown to be valid and reliable,^30- 31 the automated refraction obtained in NHANES is not equivalent to a subjectively refined refraction. Our corrected visual acuity is therefore not comparable with best-corrected visual acuity used in other studies. Despite these methodological limitations, our estimates are similar to results from previous epidemiological studies of visual impairment due to refractive error.

Our results may be influenced by nonparticipation in the vision examination. We followed the NHANES analytic guidelines to adjust for nonresponse (to the MEC examination) by using sampling weights.²⁰ However, our results indicate that persons without visual acuity data were more likely to be poor, less educated, and of nonwhite ethnicity. These characteristics were also associated with higher rates of visual impairment in our sample; therefore, prevalence of visual impairment may be underestimated in our study.

We elected to include in our analysis the 765 persons (5.8% of the total sample) who informed NHANES personnel they had refractive correction for distance but forgot to bring their lenses to the MEC; 29% of these persons were found to need refractive correction to have visual acuity of 20/40 or better. If these individuals were excluded, prevalence of visual impairment would decrease to 5.1% (vs 6.4%) and prevalence of visual impairment due to refractive error would decrease to 4.1% (vs 5.3%). Although we think it is appropriate to include these individuals in estimates of visual impairment based on presenting visual acuity, their inclusion may result in an overestimate of the prevalence of visual impairment due to refractive error.

In conclusion, the results of this study suggest that approximately 14 million individuals aged 12 years or older were estimated to be visually impaired. More than 11 million of these individuals could attain visual acuity good enough to qualify them for a driver's license in most states with refractive correction. The provision of corrective lenses to those individuals in need will be an important public health endeavor with implications for safety and quality of life.