[Skip to Content]
[Skip to Content Landing]
Figure 1.
Annual Mortality Rate vs Age
Annual Mortality Rate vs Age

Solid circles indicate the Centers for Disease Control and Prevention estimates of annual mortality rate in the US population as a function of age. The curve fit to the data is generated by the function in m(t) = 6 × 10−5 exp (0.082t), where m refers to annual mortality rate and t refers to age.

Figure 2.
Prevalence Rate vs Age for 3 Best-Corrected Visual Acuity (BCVA) Levels
Prevalence Rate vs Age for 3 Best-Corrected Visual Acuity (BCVA) Levels

The solid line is loge prevalence rate vs age with the slope fixed to 0.087 for all BCVA levels. This figure illustrates how well the model fits the 3 data sets after the data are adjusted to equate the intercepts.

Figure 3.
Prevalence Rate vs Age for 3 Best-Corrected Visual Acuity (BCVA) Criteria
Prevalence Rate vs Age for 3 Best-Corrected Visual Acuity (BCVA) Criteria

Solid circles indicate the fit of the exponential model to the National Health and Nutrition Examination Survey prevalence rate vs age for the BCVA criteria of less than 20/40 (A), less than 20/60 (B), and 20/200 or less (C), with the rate constant constrained to 0.087. Dotted lines bound the 95% CI for the estimated function (indicated by solid lines). Open circles in panels A and C are estimates from the Congdon et al5 study and are plotted for comparison.

Table 1.  
Prevalence and Annual Incidence Estimates of Low Vision in the United States for Older Adults and All Agesa
Prevalence and Annual Incidence Estimates of Low Vision in the United States for Older Adults and All Agesa
Table 2.  
Prevalence Rate of Vision Impairment by Agea
Prevalence Rate of Vision Impairment by Agea
1.
Jette  AM, Branch  LG.  Impairment and disability in the aged.  J Chronic Dis. 1985;38(1):59-65.PubMedGoogle ScholarCrossref
2.
West  SK, Munoz  B, Rubin  GS,  et al.  Function and visual impairment in a population-based study of older adults. The SEE Project. Salisbury Eye Evaluation.  Invest Ophthalmol Vis Sci. 1997;38(1):72-82.PubMedGoogle Scholar
3.
Varma  R, Vajaranant  TS, Burkemper  B,  et al.  Visual impairment and blindness in adults in the United States: demographic and geographic variations from 2015 to 2050.  JAMA Ophthalmol. 2016;134(7):802-809.PubMedGoogle ScholarCrossref
4.
Massof  RW.  A model of the prevalence and incidence of low vision and blindness among adults in the U.S.  Optom Vis Sci. 2002;79(1):31-38.PubMedGoogle ScholarCrossref
5.
Congdon  N, O’Colmain  B, Klaver  CC,  et al; Eye Diseases Prevalence Research Group.  Causes and prevalence of visual impairment among adults in the United States.  Arch Ophthalmol. 2004;122(4):477-485.PubMedGoogle ScholarCrossref
6.
Vitale  S, Cotch  MF, Sperduto  RD.  Prevalence of visual impairment in the United States.  JAMA. 2006;295(18):2158-2163.PubMedGoogle ScholarCrossref
7.
World Health Organization. International classification of impairments, disabilities, and handicaps: a manual of classification relating to the consequences of disease. Geneva, Switzerland: World Health Organization; 1980. http://apps.who.int/iris/bitstream/10665/41003/1/9241541261_eng.pdf. Published 1980. Accessed October 2, 2017.
8.
Centers for Medicare and Medicaid Services, US Department of Health and Human Services. Provider education article: Medicare coverage of rehabilitation services for beneficiaries with vision impairment. Program Memorandum Intermediaries/Carriers. Transmittal AB-02-078. https://www.cms.gov/Regulations-and-Guidance/Guidance/Transmittals/downloads/AB02078.pdf. Published May 29, 2002. Accessed October 2, 2017.
9.
American Academy of Ophthalmology. The Academy’s initiative in vision rehabilitation. https://www.aao.org/low-vision-and-vision-rehab. Accessed October 2, 2017.
10.
West  SK, Rubin  GS, Broman  AT, Muñoz  B, Bandeen-Roche  K, Turano  K.  How does visual impairment affect performance on tasks of everyday life? the SEE Project.  Arch Ophthalmol. 2002;120(6):774-780.PubMedGoogle ScholarCrossref
11.
Ko  F, Vitale  S, Chou  CF, Cotch  MF, Saaddine  J, Friedman  DS.  Prevalence of nonrefractive visual impairment in US adults and associated risk factors, 1999-2002 and 2005-2008.  JAMA. 2012;308(22):2361-2368.PubMedGoogle ScholarCrossref
12.
Goldstein  JE, Massof  RW, Deremeik  JT,  et al; Low Vision Research Network Study Group.  Baseline traits of low vision patients served by private outpatient clinical centers in the United States.  Arch Ophthalmol. 2012;130(8):1028-1037.PubMedGoogle ScholarCrossref
13.
US Social Security Administration. Disability evaluation under Social Security. 2.02 Loss of central visual acuity.https://www.ssa.gov/disability/professionals/bluebook/2.00-SpecialSensesandSpeech-Adult.htm#2_02. Accessed October 2, 2017.
14.
National Health and Nutrition Examination Survey. Public-use data files and documentation. https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/Default.aspx?BeginYear=2007. Accessed October 2, 2017.
15.
Centers for Disease Control and Prevention. Deaths and mortality. http://www.cdc.gov/nchs/fastats/deaths.htm. Updated May 3, 2017. Accessed October 2, 2017.
16.
US Census Bureau. Population projections. https://www.census.gov/programs-surveys/popproj.html. Accessed October 2, 2017.
17.
Lee  DJ, Gómez-Marín  O, Lam  BL, Zheng  DD.  Visual acuity impairment and mortality in US adults.  Arch Ophthalmol. 2002;120(11):1544-1550.PubMedGoogle ScholarCrossref
18.
Lamoureux  EL, Chong  E, Wang  JJ,  et al.  Visual impairment, causes of vision loss, and falls: the Singapore Malay Eye Study  [published correction appears in Invest Ophthalmol Vis Sci. 2008;49(4):1298].  Invest Ophthalmol Vis Sci. 2008;49(2):528-533.PubMedGoogle ScholarCrossref
19.
Lord  SR, Dayhew  J.  Visual risk factors for falls in older people.  J Am Geriatr Soc. 2001;49(5):508-515.PubMedGoogle ScholarCrossref
20.
Ivers  R, Cumming  R, Mitchell  P.  Poor vision and risk of falls and fractures in older Australians: the Blue Mountains Eye Study.  N S W Public Health Bull. 2002;13(1-2):8-10.PubMedGoogle ScholarCrossref
21.
Klein  BE, Klein  R, Lee  KE, Cruickshanks  KJ.  Performance-based and self-assessed measures of visual function as related to history of falls, hip fractures, and measured gait time: the Beaver Dam Eye Study.  Ophthalmology. 1998;105(1):160-164.PubMedGoogle ScholarCrossref
22.
Dhital  A, Pey  T, Stanford  MR.  Visual loss and falls: a review.  Eye (Lond). 2010;24(9):1437-1446.PubMedGoogle ScholarCrossref
23.
Smith  M, Bailey  T.  Identifying solutions to medication adherence in the visually impaired elderly.  Consult Pharm. 2014;29(2):131-134.PubMedGoogle ScholarCrossref
24.
US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. National Health Interview Survey, 1994. Second Supplement on Aging ICPSR 02563-v3. Ann Arbor, MI: Inter-university Consortium for Political and Social Research. 10.3886/ICPSR02563.v3. Accessed October 2, 2017.
25.
Cimarolli  VR, Jopp  DS.  Sensory impairments and their associations with functional disability in a sample of the oldest-old.  Qual Life Res. 2014;23(7):1977-1984.PubMedGoogle ScholarCrossref
26.
Keller  BK, Morton  JL, Thomas  VS, Potter  JF.  The effect of visual and hearing impairments on functional status.  J Am Geriatr Soc. 1999;47(11):1319-1325.PubMedGoogle ScholarCrossref
27.
Rubin  GS, Roche  KB, Prasada-Rao  P, Fried  LP.  Visual impairment and disability in older adults.  Optom Vis Sci. 1994;71(12):750-760.PubMedGoogle ScholarCrossref
28.
Owsley  C.  Aging and vision.  Vision Res. 2011;51(13):1610-1622.PubMedGoogle ScholarCrossref
29.
Chou  CF, Cotch  MF, Vitale  S,  et al.  Age-related eye diseases and visual impairment among U.S. adults.  Am J Prev Med. 2013;45(1):29-35.PubMedGoogle ScholarCrossref
Original Investigation
January 2018

Estimates of Incidence and Prevalence of Visual Impairment, Low Vision, and Blindness in the United States

Author Affiliations
  • 1The Lions Vision Research and Rehabilitation Center at the Johns Hopkins University School of Medicine, Baltimore, Maryland
  • 2Dana Center for Preventive Ophthalmology, Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, Maryland
JAMA Ophthalmol. 2018;136(1):12-19. doi:10.1001/jamaophthalmol.2017.4655
Key Points

Question  What is the projected prevalence and incidence of low vision and blindness in the United States from 2017 to 2050?

Findings  This study of data from 6016 participants in the 2007-2008 National Health and Nutrition Examination Survey found that the number of new cases of low vision and blindness each year is estimated to more than double in 30 years.

Meaning  Prevalence and incidence rates show a substantial increase in the need for low vision rehabilitation services with the aging of the US population.

Abstract

Importance  Updated estimates of the prevalence and incidence rates of low vision and blindness are needed to inform policy makers and develop plans to meet the future demands for low vision rehabilitation services.

Objective  To provide updated estimates of the incidence and prevalence of low vision and blindness in the United States.

Design, Setting, and Participants  Visual acuity measurements as a function of age from the 2007-2008 National Health and Nutrition Examination Survey, with representation of racial and ethnic groups, were used to estimate the prevalence and incidence of visual impairments. Data from 6016 survey participants, ranging in age from younger than 18 years to older than 45 years, were obtained to estimate prevalence rates for different age groups. Incidence and prevalence rates of low vision (best-corrected visual acuity [BCVA] in the better-seeing eye of <20/40 and <20/60) and blindness (BCVA of ≤20/200) in older adults were estimated from exponential models, fit to prevalence rates as a function of age (specified in 5-year age bins). The prevalence and annual incidence of low vision and blindness in the United States were estimated, using the 2010 US census data by age, from the rate models applied to the census projections for 2017, 2030, and 2050. Data were collected from November 1, 2007, to October 31, 2008. Data analysis took place from March 31, 2016, to March 19, 2017.

Main Outcomes and Measures  Prevalence and incidence rates of low vision and blindness in the United States.

Results  Of the 6016 people in the study, 1714 (28.4%) were younger than 18 years of age, 2358 (39.1%) were 18 to 44 years of age, and 1944 (32.3%) were 45 years of age or older. There were 2888 male (48%) and 3128 female (52%) participants. The prevalence of low vision and blindness for older adults (≥45 years) in the United States in 2017 is estimated to be 3 894 406 persons (95% CI, 3 034 442-4 862 549 persons) with a BCVA less than 20/40, 1 483 703 persons (95% CI, 968 656-2 370 513 persons) with a BCVA less than 20/60, and 1 082 790 persons (95% CI, 637 771-1 741 864 persons) with a BCVA of 20/200 or less. The estimated 2017 annual incidence (projected from 2010 census data) of low vision and blindness among older adults (≥45 years) in the United States is 481 970 persons (95% CI, 375 541-601 787 persons) with a BCVA less than 20/40, 183 618 persons (95% CI, 119 878-293 367 persons) with a BCVA less than 20/60, and 134 002 persons (95% CI, 83 383-215 567 persons) with a BCVA of 20/200 or less. The total annual incidence for each BCVA criterion is 12.4% of the total prevalence.

Conclusions and Relevance  Low vision and blindness affect a substantial portion of the older population in the United States. Estimates of the prevalence and annual incidence of visual impairment assist policy planners in allocating and developing resources for this life-changing loss of function.

Introduction

Low vision (LV) and blindness are leading causes of disability among US residents.1,2 Updated estimates of the prevalence of blindness and LV were recently reported using population-based studies applied to current US census data.3 However, the incidence of blindness and LV in the United States was formally estimated more than a decade ago using 2000 census data.4-6 More current and detailed estimates are needed to inform policy makers to adequately meet present and growing demands for LV rehabilitation services.

The World Health Organization defines LV as a best-corrected visual acuity (BCVA) in the better-seeing eye of less than 20/60.7 Many programs, like Medicare,8 use this standard, but considerable evidence (plus the American Academy of Ophthalmology Preferred Practice Patterns9) supports the consensus that people experience functional limitations when the BCVA in the better-seeing eye is less than 20/40.3,5,10-12Legal blindness is defined as a BCVA of 20/200 or less by the US Social Security Administration with a modified definition of less than 20/100 when measured with modern visual acuity charts.13 The World Health Organization defines blindness as a visual acuity of less than 20/400.7 The term visual impairment applies to any individual presenting with vision loss, which may encompass uncorrected refractive error. Because most LV cases are caused by age-related eye diseases, the number of individuals with LV will increase along with the aging of the US population. For the purposes of policy planning and identifying LV care demand, however, it is more important to know the number of new cases each year. Setting up programs on the basis of prevalence alone would address only the backlog of individuals with LV. Therefore, continued service provision relies on the incidence.

Quiz Ref IDThe most recent estimates of the prevalence and incidence of LV and blindness as a function of age used meta-analyses.3,5 The weakness of this approach is that the results of multiple epidemiological studies (some more than 30 years old) are merged. The present study takes advantage of presenting visual acuity (PVA) and BCVA raw data, which were collected as part of the most recent (2007-2008) National Health and Nutrition Examination Survey (NHANES). The NHANES data on the prevalence of correctable and uncorrectable visual impairment are a function of age, sex, race and ethnicity, and a variety of other personal and demographic traits. Vitale et al6 used data from the 1999-2002 NHANES combined with the 2000 US census data to estimate the total prevalence of correctable and uncorrectable visual impairment in the United States. The purpose of the present study is to provide updated estimates of the prevalence of LV and blindness for different age groups in the United States and develop models of LV prevalence and annual incidence rates as a function of age for the different visual acuity criteria commonly used to define LV and legal blindness.

Methods

The NHANES is a cross-sectional sample of the noninstitutionalized US population and is a program of ongoing surveys of health status performed in 2-year cycles by the National Center for Health Statistics of the Centers for Disease Control and Prevention.14 The 1999-2008 NHANES protocol was reviewed and approved by the National Center for Health Statistics research ethics review board. The 2007-2008 NHANES data (2008 was the last year visual acuity measurements were obtained) were used to construct new models and provided updated estimates of incidence and prevalence rates of adults with visual impairment, LV, and legal blindness. This present study was exempt from institutional review board review because, according to the US Department of Health and Human Services, it included the analysis of existing, publicly available data, which were recorded in a way that participants cannot be identified. The study is not classified as human subjects research by the Johns Hopkins Institutional Review Board. Data were collected from November 1, 2007, to October 31, 2008. Data analysis took place from March 31, 2016, to March 19, 2017.

In the NHANES, PVA was measured for each eye while participants used their usual distance vision correction, if any. Visual acuity was measured using an autorefractor (ARK-760; Nidek Co Ltd) 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. If the individual’s PVA measured 20/30 or worse, automated refraction was performed, and the BCVA based on autorefraction was recorded.

The NHANES measured PVA and BCVA in a sample of 6016 people. Many epidemiological LV studies use an age of 50 years or older as the criterion for adults, but this study included individuals 45 years of age or older. The rationale was that a data point close to the asymptote improves the estimate of the model parameters. Five-year intervals were used, and values were assigned to the midpoint of the data. Including the age category of 45 years allowed the use of the midpoint age of 47.5 years rather than 52.5 years, the midpoint for starting with age 50 years. Those without complete BCVA data were excluded. Age was reported to the nearest year. For the purpose of the present analysis, LV was defined in 2 ways: a BCVA in the better-seeing eye of less than 20/40 and a BCVA of less than 20/60. Legal blindness was defined as a BCVA in the better-seeing eye of 20/200 or less. The NHANES did not report visual acuity measurements between 20/80 and 20/200 or of less than 20/200, precluding the analyses for other definitions of LV and blindness.

Prevalence of visual impairment in the NHANES sample was estimated by racial/ethnic group: non-Hispanic white (NHW); Hispanic white (HW); and total black (TB), which combined Hispanic and non-Hispanic black. All other racial/ethnic groups had too few individuals in the sample to analyze separately. The number of persons as a function of age was counted for each of the visual acuity categories in the 3 racial/ethnic groups. The TB and HW groups are overrepresented in the NHANES sample, and the NHW group is underrepresented relative to their distribution in the US population. Therefore, the LV prevalence for the NHANES subsamples in each age group was weighted to correct the sample biases in racial/ethnic groups.10 The weighted observed prevalence for each group was combined in each age bin for each visual acuity category and then divided by the corresponding weighted total sample size to obtain an estimated prevalence rate as a function of age representative of the total population. For all visual acuity categories, the prevalence rate appeared to be an exponential function of age; that is, the loge prevalence rate should be linear with age for all visual acuity categories.

To determine the incidence of LV and blindness in the US population, the exponential model was linearized by estimating the loge prevalence rate vs age for older adults (≥45 years). These estimates used each visual acuity data set. (Data were omitted for ages at which the empirical prevalence rate is 0 because the corresponding loge is undefined.) The rate of change in loge prevalence was assumed to be the same for each visual acuity range; therefore, the model rate constant was calculated from linear regression of the loge prevalence rate (corrected for differences between visual acuity category intercepts) vs age for each visual acuity data set. The exponential model with a fixed rate constant (ie, slope of the loge regression model) was then used to refine the estimates of the visual acuity–specific coefficient (ie, intercept of the loge regression model) because empirical prevalence rates of 0 could be used in the nonlinear regression analyses.

The use of the exponential model presumes that (1) once an individual has LV, visual acuity does not improve over time, and (2) the rate of change in prevalence rate is a constant percentage of the prevalence rate (ie, prevalence grows by adding new cases over time to the existing pool of cases). However, we were interested not only in the growth of the prevalence rate with time but also in the percentage of new cases annually for each age group (ie, the incidence rate). To estimate the incidence rate, the growth of the prevalence rate vs age must be corrected by accounting for the percentage of individuals in the pool who die (death rate vs age). The following equations illustrate how these assumptions were incorporated in the models.

The exponential LV prevalence rate model is defined by the equation p(t) = exp (rt + bVA), where t is the person’s age in years; r is the rate constant (slope of the linear regression on loge prevalence rate vs age), which is constrained to be the same for all 3 visual acuity criteria for LV; and bVA is a visual acuity–specific coefficient (intercept of the linear regression on loge prevalence rate vs age).

These models of prevalence rate were differentiated by age to estimate the annual growth rates in LV prevalence for each of the 3 BCVA criteria as a function of age:

p′(t) = [dp(t)/dt] = r exp (rt + bVA).

The rate of growth in prevalence rate, p′(t), includes the rate of new LV cases less the mortality rate of existing LV cases. Thus, to calculate the annual LV incidence rate (ie, the number of new cases of LV each year), i(t), adding back in the LV mortality rate, p(t)m(t), is necessary:

i(t) = r exp (rt + bVA) + p(t)m(t).

Figure 1 illustrates estimates of annual mortality rates for the US population as a function of age in 2010 (solid circles).15 The curve fit to the mortality rate estimates is the exponential function m(t) = 6 × 10−5 exp (0.082t), in which t refers to age.

The mean of the slopes of the regression lines fit to loge prevalence rate vs age data for 3 different BCVA categories: (<20/40 to 20/60, <20/60 to >20/200, and ≤20/200) is 0.087 (Figure 2). To display the data together with the regression line having the average slope, the intercept for the regression on the data set of less than 20/40 to 20/60 BCVA was used as the reference, and the loge prevalence rates for the 2 other BCVA categories were adjusted to ensure that their respective regression line intercepts matched the reference. The R2 values for the 2-parameter regression line fit to each data set were 0.44 for less than 20/40 to 20/60 BCVA, 0.92 for less than 20/60 to greater than 20/200 BCVA, and 0.44 for 20/200 or less BCVA. The R2 values for the regression line fit with a fixed slope of 0.087 were 0.40 for less than 20/40 to 20/60 BCVA, 0.90 for less than 20/60 to greater than 20/200 BCVA, and 0.43 for 20/200 or less BCVA. The excellent agreement between models in R2 values supports the use of the same rate constant in the prevalence rate models for different BCVA criteria for LV and blindness. The fit of exponential prevalence rate vs age function was determined for each BCVA criterion, with the slope (rate constant) constrained to 0.087. The estimated model coefficients for each BCVA criterion for defining LV and blindness, bVA, were −9.79 for less than 20/40 to 20/60, −12.11 for less than 20/60 to greater than 20/200, and −10.75 for 20/200 or less.

The prevalence and annual incidence of LV among older adults (≥45 years) in the United States were estimated from the rate models applied to the 2010 US census data by age. The US Administration on Aging population growth estimates16 in each age group were used to project the 2010 census data to 2017, 2030, and 2050.

Using methods similar to those used by Vitale et al,6 we calculated the prevalence of presenting correctable and uncorrectable visual impairment over the entire age range from the NHANES samples for children (<18 years), young adults (18-44 years), and older adults (≥45 years). Prevalence rates of PVA and BCVA were estimated from the 2007-2008 NHANES data sets and were broken down and weighted according to racial/ethnic representation in each age group.

Results

Quiz Ref IDThe 2017 estimated prevalence rates of LV for older adults (≥45 years) in the United States were 3% (3 894 406) for less than 20/40 BCVA, 1.1% (1 483 703) using a cutoff of less than 20/60 BCVA, and 0.8% (1 082 790) for 20/200 or less BCVA (legal blindness) (Table 1). The 2017 total annual incidence rate for each BCVA criterion was 12.4% of the total prevalence, translating to 481 970 incident cases of less than 20/40 BCVA, 183 618 incident cases of less than 20/60 BCVA, and 134 002 incident cases of legal blindness (≤20/200 BCVA). Of the 6016 people in the study, 1714 (28.4%) were younger than 18 years of age, 2358 (39.1%) were 18 to 44 years of age, and 1944 (32.3%) were 45 years of age or older. There were 2888 male (48%) and 3128 female (52%) participants.

The results of the nonlinear regression analyses of LV prevalence rate vs age are illustrated for the different BCVA criteria in Figure 3 along with the estimated prevalence rate data, plotted at the midpoint of each 5-year age category. The dashed curves in Figure 3 define the model boundaries from the 95% CI, calculated from ±1.96 SEs for each exponential function representing an LV or legal blindness category.

The overall prevalence rates of LV and legal blindness were estimated for 3 broad age categories (<18 years, 18-44 years, and ≥45 years) for each of the 3 BCVA categories (<20/40, <20/60, and ≤20/200). Table 2 lists the estimated prevalence rates in each BCVA category, racial/ethnic category, and age category, organized by participants’ PVA in the better-seeing eye and BCVA in the better-seeing eye. Table 2 also lists the percentage of people in each category whose PVA could be improved to 20/40 or greater by correcting refractive error. The racial/ethnic categories, organized by age groups and by BCVA criterion, give us 27 comparisons to assess for significance. A Bonferroni correction for α = .05 gives the criterion of a statistically significant P = .002. Only 3 comparisons met the criterion for significance: NHW vs HW individuals with a visual acuity of less than 20/40 who were (1) 18 to 44 years of age compared with those who were (2) 45 years of age or older, as well as (3) NHW vs TB individuals with a visual acuity of less than 20/40 who were 18 to 44 years of age. Although the comparison of NHW vs TB individual with a visual acuity of less than 20/40 who were 45 years of age or older showed a trend, it did not meet statistical significance after correction.

Table 1 lists the estimated prevalence and annual incidence of low vision for adults 45 years of age or older in the US population for each visual criterion projected for 2017, 2030, and 2050 from the 2010 US census data by age. Table 1 also lists the estimated prevalence of low vision in the total population (including younger individuals) for each visual criterion projected for 2017, 2030, and 2050 from the 2010 US census data by age. Incidence estimates were not calculated in the younger age group owing to the small number of individuals in some categories.

Discussion

Quiz Ref IDOverall, the youngest group (<18 years) had the highest rate of presenting visual impairment in the better-seeing eye for all 3 BCVA categories. Similar to the results obtained by Vitale et al,6 findings from our study indicate that 85% to 96% of children with presenting visual impairments could have their BCVA corrected to greater than 20/40. Because most causes of permanent impairment increase in prevalence with increasing age, the highest rates of nonrefractive LV and blindness were seen in older individuals.

Presenting visual impairment was greatest in the HW and TB racial/ethnic categories; however, those groups also demonstrated a higher rate of visual impairment that could be corrected with spectacles to 20/40 or greater BCVA. No significant differences were detected between racial/ethnic categories (after Bonferroni correction), with the exception of the 3 comparisons mentioned earlier. These findings are consistent with previous reports that NHW individuals represent the largest number of people with LV and blindness.3 Ko et al11 reported that the factors associated with nonrefractive visual impairment included older age, poverty, lower educational level, and diabetes diagnosed 10 or more years ago. However, increasing rates of visual impairment in younger individuals may be attributed to an increase in systemic illness and a lack of health care, which may identify the need for myopic correction.

The current prevalence rate estimates are similar to previously reported rates.4,5Figure 3 compares the estimates of prevalence rate vs age by Congdon et al5 with those of the present study, along with the exponential model we propose. Earlier estimates of prevalence rates as a function of age drew summary data from different studies that used differing BCVA criteria and age category definitions. The present study used updated census data, with age in the database specified in 1-year increments, and therefore may provide more precise estimates. In addition, visual acuity was measured the same way for all participants in the NHANES data set and specified to the nearest line. Numerous assumptions and approximations had to be made in the previous meta-analyses, but there is remarkable quantitative agreement between the studies.

To build our models, we assumed that LV does not increase the risk of death and applied the mortality rate as a function of age from the general population to the LV population. However, an increased risk of mortality has been shown in cases of patient-reported severe bilateral visual impairment.17 In addition, it is possible, if not probable, that LV increases death rates because LV increases the risk of falls,18-22 medication mismanagement,23,24 and reduced activity levels.24-28 If the death rate has been underestimated for the LV population, then the annual LV incidence rate also has been underestimated. The risk of age-related eye disease increases with increasing age,5,28,29 which explains the acceleration of the LV prevalence rate with age (Figure 3).

Limitations

Quiz Ref IDThis study had several limitations. The NHANES vision screening data did not provide visual acuity measurements between 20/80 and 20/200. Therefore, the legal blindness definition was limited to 20/200 or less visual acuity in the better-seeing eye. Individuals with profound visual impairment (hand motions, light perception, or no light perception) were excluded from visual acuity testing but were included in the BCVA category of less than 20/200; with refraction, some of these people may have had BCVA improvement. Because autorefraction was used to determine corrected visual acuity, it is possible that refractive error was not fully corrected for all participants, potentially contributing to an overestimation of uncorrectable visual acuity loss.

The sample contained some racial/ethnic groups with too few individuals to include in the analyses. By design, the NHANES does not include people who are institutionalized (eg, nursing home population); therefore, the prevalence and annual incidence of visual acuity loss in the population were likely underestimated. In addition, NHANES vision screening data did not provide data on visual fields. Patients with visual field loss, with relatively intact visual acuity, would not be included in the visual impairment estimates. Estimates of LV and blindness were based on projections from the census data, which did not specify ocular conditions (such as the rate of myopia) or identify visual impairment associated with systemic diseases (such as diabetes). Therefore, the incidence and prevalence of LV and blindness estimated in this study are likely to be too low.

Conclusions

The exponential growth in prevalence rate vs age reflects the increased risk of vision loss from age-related eye disease with each passing year. As demonstrated in Tables 1 and 2, assuming the prevalence and mortality rates stay constant, we expect a greater need for services for those patients with LV as the aging population increases over the next several decades. Regarding health care delivery, prevalence rates can be used to estimate the backlog of people with visual impairment needing LV services. However, of likely greater relevance for planning purposes and provision of health care services is the number of new cases of visual impairment and blindness each year, which is estimated by the annual incidence for older adults: more than 480 000 cases of mild LV or worse (<20/40), more than 180 000 cases of moderate LV or worse (<20/60), and more than 134 000 legally blind cases (≤20/200). In addition, consistent with prior studies, data suggest that vision screenings for uncorrected refractive error could substantially reduce the prevalence of visual impairment.

Back to top
Article Information

Corresponding Author: Robert Massof, PhD, The Lions Vision Research and Rehabilitation Center at the Johns Hopkins School of Medicine, 600 N Wolfe St, Baltimore, MD 21287 (bmassof@jhmi.edu).

Accepted for Publication: September 18, 2017.

Published Online: November 2, 2017. doi:10.1001/jamaophthalmol.2017.4655

Author Contributions: Drs Chan and Massof had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Chan, Friedman, Massof.

Acquisition, analysis, or interpretation of data: Chan, Bradley, Massof.

Drafting of the manuscript: Chan, Friedman.

Critical revision of the manuscript for important intellectual content: Chan, Bradley, Massof.

Statistical analysis: Bradley, Massof.

Administrative, technical, or material support: Friedman.

Study supervision: Friedman, Massof.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Massof reported being a consultant for Research and Development at Janssen. No other disclosures were reported.

Funding/Support: This study was funded by a grant from Reader’s Digest Partners for Sight Foundation and by grant EY018696 from the National Eye Institute of the National Institutes of Health.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Jette  AM, Branch  LG.  Impairment and disability in the aged.  J Chronic Dis. 1985;38(1):59-65.PubMedGoogle ScholarCrossref
2.
West  SK, Munoz  B, Rubin  GS,  et al.  Function and visual impairment in a population-based study of older adults. The SEE Project. Salisbury Eye Evaluation.  Invest Ophthalmol Vis Sci. 1997;38(1):72-82.PubMedGoogle Scholar
3.
Varma  R, Vajaranant  TS, Burkemper  B,  et al.  Visual impairment and blindness in adults in the United States: demographic and geographic variations from 2015 to 2050.  JAMA Ophthalmol. 2016;134(7):802-809.PubMedGoogle ScholarCrossref
4.
Massof  RW.  A model of the prevalence and incidence of low vision and blindness among adults in the U.S.  Optom Vis Sci. 2002;79(1):31-38.PubMedGoogle ScholarCrossref
5.
Congdon  N, O’Colmain  B, Klaver  CC,  et al; Eye Diseases Prevalence Research Group.  Causes and prevalence of visual impairment among adults in the United States.  Arch Ophthalmol. 2004;122(4):477-485.PubMedGoogle ScholarCrossref
6.
Vitale  S, Cotch  MF, Sperduto  RD.  Prevalence of visual impairment in the United States.  JAMA. 2006;295(18):2158-2163.PubMedGoogle ScholarCrossref
7.
World Health Organization. International classification of impairments, disabilities, and handicaps: a manual of classification relating to the consequences of disease. Geneva, Switzerland: World Health Organization; 1980. http://apps.who.int/iris/bitstream/10665/41003/1/9241541261_eng.pdf. Published 1980. Accessed October 2, 2017.
8.
Centers for Medicare and Medicaid Services, US Department of Health and Human Services. Provider education article: Medicare coverage of rehabilitation services for beneficiaries with vision impairment. Program Memorandum Intermediaries/Carriers. Transmittal AB-02-078. https://www.cms.gov/Regulations-and-Guidance/Guidance/Transmittals/downloads/AB02078.pdf. Published May 29, 2002. Accessed October 2, 2017.
9.
American Academy of Ophthalmology. The Academy’s initiative in vision rehabilitation. https://www.aao.org/low-vision-and-vision-rehab. Accessed October 2, 2017.
10.
West  SK, Rubin  GS, Broman  AT, Muñoz  B, Bandeen-Roche  K, Turano  K.  How does visual impairment affect performance on tasks of everyday life? the SEE Project.  Arch Ophthalmol. 2002;120(6):774-780.PubMedGoogle ScholarCrossref
11.
Ko  F, Vitale  S, Chou  CF, Cotch  MF, Saaddine  J, Friedman  DS.  Prevalence of nonrefractive visual impairment in US adults and associated risk factors, 1999-2002 and 2005-2008.  JAMA. 2012;308(22):2361-2368.PubMedGoogle ScholarCrossref
12.
Goldstein  JE, Massof  RW, Deremeik  JT,  et al; Low Vision Research Network Study Group.  Baseline traits of low vision patients served by private outpatient clinical centers in the United States.  Arch Ophthalmol. 2012;130(8):1028-1037.PubMedGoogle ScholarCrossref
13.
US Social Security Administration. Disability evaluation under Social Security. 2.02 Loss of central visual acuity.https://www.ssa.gov/disability/professionals/bluebook/2.00-SpecialSensesandSpeech-Adult.htm#2_02. Accessed October 2, 2017.
14.
National Health and Nutrition Examination Survey. Public-use data files and documentation. https://wwwn.cdc.gov/nchs/nhanes/ContinuousNhanes/Default.aspx?BeginYear=2007. Accessed October 2, 2017.
15.
Centers for Disease Control and Prevention. Deaths and mortality. http://www.cdc.gov/nchs/fastats/deaths.htm. Updated May 3, 2017. Accessed October 2, 2017.
16.
US Census Bureau. Population projections. https://www.census.gov/programs-surveys/popproj.html. Accessed October 2, 2017.
17.
Lee  DJ, Gómez-Marín  O, Lam  BL, Zheng  DD.  Visual acuity impairment and mortality in US adults.  Arch Ophthalmol. 2002;120(11):1544-1550.PubMedGoogle ScholarCrossref
18.
Lamoureux  EL, Chong  E, Wang  JJ,  et al.  Visual impairment, causes of vision loss, and falls: the Singapore Malay Eye Study  [published correction appears in Invest Ophthalmol Vis Sci. 2008;49(4):1298].  Invest Ophthalmol Vis Sci. 2008;49(2):528-533.PubMedGoogle ScholarCrossref
19.
Lord  SR, Dayhew  J.  Visual risk factors for falls in older people.  J Am Geriatr Soc. 2001;49(5):508-515.PubMedGoogle ScholarCrossref
20.
Ivers  R, Cumming  R, Mitchell  P.  Poor vision and risk of falls and fractures in older Australians: the Blue Mountains Eye Study.  N S W Public Health Bull. 2002;13(1-2):8-10.PubMedGoogle ScholarCrossref
21.
Klein  BE, Klein  R, Lee  KE, Cruickshanks  KJ.  Performance-based and self-assessed measures of visual function as related to history of falls, hip fractures, and measured gait time: the Beaver Dam Eye Study.  Ophthalmology. 1998;105(1):160-164.PubMedGoogle ScholarCrossref
22.
Dhital  A, Pey  T, Stanford  MR.  Visual loss and falls: a review.  Eye (Lond). 2010;24(9):1437-1446.PubMedGoogle ScholarCrossref
23.
Smith  M, Bailey  T.  Identifying solutions to medication adherence in the visually impaired elderly.  Consult Pharm. 2014;29(2):131-134.PubMedGoogle ScholarCrossref
24.
US Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Health Statistics. National Health Interview Survey, 1994. Second Supplement on Aging ICPSR 02563-v3. Ann Arbor, MI: Inter-university Consortium for Political and Social Research. 10.3886/ICPSR02563.v3. Accessed October 2, 2017.
25.
Cimarolli  VR, Jopp  DS.  Sensory impairments and their associations with functional disability in a sample of the oldest-old.  Qual Life Res. 2014;23(7):1977-1984.PubMedGoogle ScholarCrossref
26.
Keller  BK, Morton  JL, Thomas  VS, Potter  JF.  The effect of visual and hearing impairments on functional status.  J Am Geriatr Soc. 1999;47(11):1319-1325.PubMedGoogle ScholarCrossref
27.
Rubin  GS, Roche  KB, Prasada-Rao  P, Fried  LP.  Visual impairment and disability in older adults.  Optom Vis Sci. 1994;71(12):750-760.PubMedGoogle ScholarCrossref
28.
Owsley  C.  Aging and vision.  Vision Res. 2011;51(13):1610-1622.PubMedGoogle ScholarCrossref
29.
Chou  CF, Cotch  MF, Vitale  S,  et al.  Age-related eye diseases and visual impairment among U.S. adults.  Am J Prev Med. 2013;45(1):29-35.PubMedGoogle ScholarCrossref
×