Key Points
Question
How do sociodemographic, economic, and health-related characteristics associated with vision impairment and receipt of eye care vary across low- and middle-income countries?
Findings
In this study of cross-sectional survey data from 34 159 adults in the Study on Global Aging and Adult Health, in each low- and middle-income country studied, a unique set of characteristics associated with vision impairment and receipt of eye care was found. However, certain traits, such as educational attainment, medical comorbidities, and memory, were more consistently associated with these outcomes across most countries.
Meaning
These findings suggest that cross-national comparisons to identify shared and unique characteristics associated with vision impairment and receipt of eye care may be important for targeting those most in need and optimally allocating health care resources.
This study of cross-sectional survey data from the World Health Organization Study on Global Aging and Adult Health identifies the characteristics associated with vision impairment and receipt of eye care in a sample of low- and middle-income countries.
Abstract
Importance
Vision impairment (VI), including blindness, affects hundreds of millions globally, and 90% of those with VI live in low- and middle-income countries. Cross-national comparisons are important to elucidate the unique and shared factors associated with VI and receipt of eye care in different countries and to target those most in need.
Objective
To identify the characteristics associated with VI and receipt of eye care in a sample of low- and middle-income countries.
Design, Setting, and Participants
In this study of cross-sectional survey data from wave 1 of the World Health Organization Study on Global Aging and Adult Health, data on sociodemographic characteristics and health were collected from nationally representative samples in China, Ghana, India, Mexico, Russia, and South Africa from 2007 to 2010. Probability sampling with multistage, stratified, random-cluster samples was used to identify households and participants. The survey was completed by 34 159 adults 50 years and older. Data were analyzed from December 2017 to February 2018.
Main Outcomes and Measures
We analyzed associations of individual-level and household-level covariates with 3 primary outcomes: distance VI (visual acuity worse than 6/18 in the better-seeing eye), near VI (visual acuity worse than 6/18 in the better-seeing eye), and receipt of an eye examination within the previous 2 years.
Results
The study sample in China consisted of 13 350 participants (50.2% female; mean [SD] age, 62.6 [9.0] years); in Ghana, 4725 participants (50.4% female; mean [SD] age, 64.2 [10.8] years); in India, 7150 participants (48.9% female; mean [SD] age, 61.5 [9.0] years); in Mexico, 2103 participants (52.3% female; mean [SD] age, 69.2 [9.2] years); in Russia, 3763 participants (61.1% female; mean [SD] age, 63.9 [10.4] years); and in South Africa, 3838 participants (55.9% female; mean [SD] age 61.6 [9.5]) (all demographic characteristics weighted to reflect respective populations). The weighted proportion of the study sample with distance VI ranged from 9.9% (95% CI, 9.3-10.5) in China to 25.4% (95% CI, 22.0-29.2) in Russia; near VI, from 28.5% (95% CI, 26.9-30.1) in Ghana to 43.1% (95% CI, 41.1-45.1) in India; and receipt of a recent eye examination, from 15.0% (95% CI, 13.8-16.2) in Ghana to 53.1% (95% CI, 49.3-56.8) in Russia. Educational attainment, medical comorbidities, and memory were significantly associated with all outcomes across most low- and middle-income countries. Female sex, low household wealth, food insecurity, no health insurance, rurality, disability, being unmarried, and low social participation were significantly associated with adverse vision-related outcomes, though less consistently.
Conclusions and Relevance
There are both common and unique characteristics associated with VI and receipt of eye care across low- and middle-income countries. Our findings suggest that recognizing these factors is important to identify those most at risk and allocate resources optimally. Additional local epidemiological studies are needed.
Introduction
Vision impairment (VI), including blindness, affects more than 250 million people worldwide,1 with 81% of those affected 50 years and older and 90% living in low- and middle-income countries (LMICs).2 Although the prevalence of VI and blindness has decreased in recent years because of improvements in socioeconomic development and the accessibility of eye care, the number of older adults affected continues to increase owing to aging of the population and improved life expectancy.1 In fact, age-related vision disorders are projected to overtake diabetes by 2030 as the eighth largest cause of disability in middle-income countries.3
Age-related eye conditions, rather than infectious and neonatal diseases, are now the predominant causes of VI throughout the world. In 2015, uncorrected refractive error and cataract together accounted for 55% of blindness and 77% of VI in adults 50 years and older worldwide.1 While the prevalence of other causes varied by region, chronic age-related diseases, like glaucoma, macular degeneration, and diabetic retinopathy, are universally important causes of VI and blindness. Age-related eye diseases will likely continue to affect an increasing number of people globally, as the number of individuals 60 years and older is expected to double over the next 30 years and will account for more than one-fifth of the world’s population.4
In a 2017 study, Wang et al5 found that gross domestic product, human development index, and health expenditures explained 69.4% of between-country variation in the prevalence of moderate to severe VI. This work provided economically attainable targets for countries seeking to combat VI and blindness. However, it was beyond the scope of that study to identify the within-country socioeconomic factors that explain who is least likely to receive eye care and most likely to be visually impaired.
The same LMICs that are subject to the most extreme demographic and epidemiologic transitions are also the least prepared to deal with chronic conditions and their accompanying disability.6 Accordingly, national and local public health officials could benefit from a granular understanding of the associations of local socioeconomic and demographic characteristics with VI and receipt of eye care in their countries.7 Comparing the individual and household traits associated with VI and receipt of eye care across a sample of LMICs could elucidate common patterns while also reinforcing the need for more local epidemiologic studies. These types of data are crucial to attaining the World Health Organization VISION 2020 goal of reducing avoidable VI through universal access and equity.2
To address this knowledge gap, we used data from the World Health Organization Study on Global Aging and Adult Health (SAGE), which includes nationally representative samples of older adults from 6 LMICs. Using these data, we examined the associations of demographic, socioeconomic, and health-related traits of individuals and their households with VI and receipt of eye care in China, Ghana, India, Mexico, Russia, and South Africa. The purpose of this study was to provide insights into the characteristics of older adults and households in these countries that are associated with an increased or decreased likelihood of VI and receipt of eye care.
Methods
Data Source
The Study on Global Aging and Adult Health is a nationally representative survey on health and well-being in individuals from 6 LMICs: China, Ghana, India, Mexico, Russia, and South Africa. We used cross-sectional data from wave 1 (collected between 2007 and 2010), since wave 2 data were not yet available. Adults 50 years and older were oversampled, and data from these respondents were used for this study.
Sampling methods were based on the methodology of the World Health Survey,8 using probability sampling with multistage, stratified, random-cluster samples.9 The sample weighting used a sample selection and poststratification factor to ensure each sample was nationally representative. The survey was conducted using an interviewer-administered questionnaire in the participant’s native language.10 Proxy respondents were interviewed if individuals were unable to complete the survey. The University of Michigan Institutional Review Board deemed this study exempt because it was a secondary analysis of publicly available data.
Variable Definitions
Outcome Variables
The outcomes of interest were distance VI, near VI, and receipt of eye care. Visual acuity was measured using the tumbling E logMAR chart for distance and near acuity separately for each eye. A string was used to measure 40 cm as the test distance for near acuity. We defined VI (at distance and near) according to the World Health Organization definition for moderate VI, which is presenting visual acuity worse than 6/18 (0.48 logMAR) in the better-seeing eye.11
Receipt of eye care was determined from participants’ response to the question, “When was the last time you had your eyes examined by a medical professional?” which was recorded as never or the number of years ago they were last examined. Based on the distribution of responses, we categorized respondents as having had a recent eye examination if their last one was within 2 years.
Covariates
We analyzed covariates that corresponded to the health, well-being, demographic characteristics, and socioeconomic circumstances of respondents and their households. Covariates were chosen based on their conceptual relevance to the outcome variables in the study.
Health and Well-being
Physical functioning was measured using the World Health Organization Disability Assessment Schedule 2.0. Responses to the 12 items were recorded on a 5-point scale, and a score ranging from 0% to 100% was derived using the recommended algorithm.12 Based on the distribution of scores, we defined disability as none, mild, moderate, or severe for scores less than 5%, 5% to 13%, greater than 13% to 30%, and greater than 30%, respectively. We also constructed a variable indicating if an individual had 0, 1, or more than 1 comorbid conditions based on self-reported diagnosis or symptoms suggestive of arthritis, stroke, angina, diabetes, chronic lung disease, asthma, depression, or hypertension. Memory, a marker of cognition,13 was assessed by a series of verbal recall trials. The total number of words recalled in 4 memory trials was summed. Based on the distribution of memory scores, we defined poor, fair, good, and very good memory as scores ranging from 0 to 18, 19 to 22, 23 to 26, and 27 to 40, respectively.
Social Participation and Support
Low levels of social participation14,15 and support16,17,18 are associated with poor health outcomes. We calculated a social support score from 0 to 4 using previously described methods19 and dichotomized the variable to compare those with low (0-1) and high (2-4) levels of support. We also derived a measure of social participation from 4 questions related to personal relationships, making and maintaining friendships, joining community activities, and attending social activities or visits with friends or relatives. Responses to each item were recorded on a 5-point scale and dichotomized, with poor participation represented by the lower 3 levels (moderate, severe, or extreme difficulty) for items 1 to 3 and the lower 2 levels (0-2 times per year) for item 4. A participation score from 0 to 4 was calculated and dichotomized to compare those with low (0-2) and high (3-4) levels of participation.
Socioeconomic Status
Measurement error and bias is common in reporting income from surveys in LMICs.20 Therefore, a wealth index using a validated method that measures consumption, which is an accurate measure of permanent income in LMICs, was used.21 The wealth index accounts for whether households possess assets or attributes, such as durable goods (eg, TV and livestock), services (eg, water source and internet access), and traits (eg, cooking fuel and toilet). A random-effects model was used to estimate assets per household, and a Bayesian postestimation method was used to generate income estimates corresponding to each household’s location on their country’s asset ladder and to categorize households into wealth quintiles.21 In addition, educational attainment was used as a separate indicator of socioeconomic status.
Other Covariates
Age, sex, and marital status were recorded. Household locations were described as urban (including suburban and metropolitan) or rural based on local legal definitions. We also determined whether each participant had health insurance from any source (data on health insurance were not available for Mexico), received health care the last time they needed it (termed health access), and was ever hungry because they could not afford food during the prior year (termed food insecurity).
Data Analysis
Based on the SAGE design, we calculated the weighted proportions of participants for each covariate group stratified by VI status and timing of last eye examination. We report unadjusted P values from Pearson χ2 tests. We performed univariable logistic regression with distance VI, near VI, and receipt of eye care as outcome variables and each of the covariates as predictors. The unadjusted odds ratios (ORs) from univariable models were used to generate heat maps with color-coding to depict the relative effect sizes of associations. We also constructed multivariable logistic models adjusted for age, educational attainment, and wealth for each of the 3 outcome variables. All analyses were conducted using R version 3.4.3 (The R Foundation) and accounted for the complex and unique design features of SAGE in each country. All statistical tests were 2-tailed, and a P value less than or equal to .05 was considered statistically significant.
Results
The study sample in China consisted of 13 350 participants (50.2% female; mean [SD] age, 62.6 [9.0] years); in Ghana, 4725 participants (50.4% female; mean [SD] age, 64.2 [10.8] years); in India, 7150 participants (48.9% female; mean [SD] age, 61.5 [9.0] years); in Mexico, 2103 participants (52.3% female; mean [SD] age, 69.2 [9.2] years); in Russia, 3763 participants (61.1% female; mean [SD] age, 63.9 [10.4] years); and in South Africa, 3838 participants (55.9% female; mean [SD] age 61.6 [9.5]). All characteristics of the study samples are weighted to reflect the respective populations from which they were drawn.
Distance Vision Impairment
Complete distance visual acuity data were available for 12 283 of 13 350 participants (92.0%) from China, 4273 of 4725 (90.4%) from Ghana, 6289 of 7150 (88.0%) from India, 1983 of 2103 (94.3%) from Mexico, 2927 of 3763 (77.8%) from Russia, and 3692 of 3838 (96.2%) from South Africa. The heat map illustrates the magnitudes of the associations of distance VI with each covariate (Figure, A). The weighted proportions of participants for each covariate group stratified by distance VI are shown in Table 1. Unadjusted and adjusted ORs are displayed in eTable 1 in the Supplement.
Figure. Heat Maps of Unadjusted Associations of Demographic, Socioeconomic, and Health-Related Variables With Distance Vision Impairment (VI), Near VI, and Receipt of a Recent Eye Examination.
The legend applies to all 3 heat maps and shows the effect size (based on odds ratios) depicted by various colors. Variables with a strong positive and negative association with the outcome are illustrated by progressively darker shades of hot and cold colors, respectively. Unadjusted associations with distance VI, near VI, and receipt of a recent eye examination are depicted. Odds ratios were calculated using logistic models, and reference groups are noted last in the variable descriptions. Mild VI indicates visual acuity between worse than 6/12 and 6/18 or better; moderate VI, between worse than 6/18 and 6/60 or better; severe VI, worse than 6/60 and 6/120 or better; and blindness, worse than 6/120.
Table 1. Summary of Associations of Demographic, Socioeconomic, and Health-Related Variables With Distance Vision Impairment (VI)a.
| Characteristic | % | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| China | Ghana | India | Mexico | Russia | South Africa | |||||||||||||
| VI (n = 1362) | No VI (n = 10 921) | P Value | VI (n = 553) | No VI (n = 3720) | P Value | VI (n = 1071) | No VI (n = 5218) | P Value | VI (n = 342) | No VI (n = 1641) | P Value | VI (n = 579) | No VI (n = 2348) | P Value | VI (n = 451) | No VI (n = 3241) | P Value | |
| Overall prevalence, % (95% CI) | 9.9 (9.3-10.5) | 90.1 (89.5-90.7) | NA | 12.2 (11.1-13.4) | 87.8 (86.6-88.9) | NA | 18.5 (17.0-20.1) | 81.5 (79.9-83.0) | NA | 15.5 (13.4-17.9) | 84.5 (82.1-86.6) | NA | 25.4 (22.0-29.2) | 74.6 (70.8-78.0) | NA | 10.9 (9.4-12.6) | 89.1 (87.4-90.6) | NA |
| Age, y | <.001 | <.001 | <.001 | <.001 | <.001 | .006 | ||||||||||||
| 50-59 | 4.2 | 95.8 | 7.5 | 92.5 | 11.0 | 89.0 | 9.5 | 90.5 | 19.2 | 80.8 | 8.5 | 91.5 | ||||||
| 60-69 | 7.9 | 92.1 | 12.2 | 87.8 | 21.0 | 79.0 | 11.5 | 88.5 | 23.5 | 76.5 | 12.2 | 87.8 | ||||||
| 70-79 | 20.4 | 79.6 | 15.8 | 84.2 | 31.1 | 68.9 | 18.6 | 81.4 | 32.4 | 67.6 | 15.4 | 84.6 | ||||||
| ≥80 | 42.0 | 58.0 | 23.3 | 76.7 | 42.1 | 57.9 | 28.1 | 71.9 | 45.9 | 54.1 | 14.1 | 85.9 | ||||||
| Sex | <.001 | <.001 | <.001 | .55 | .32 | .02 | ||||||||||||
| Male | 7.8 | 92.2 | 10.4 | 89.6 | 16.0 | 84.0 | 16.2 | 83.8 | 23.7 | 76.3 | 9.1 | 90.9 | ||||||
| Female | 12.2 | 87.8 | 14.3 | 85.7 | 21.1 | 78.9 | 14.9 | 85.1 | 26.5 | 73.5 | 12.4 | 87.6 | ||||||
| Education | <.001 | <.001 | <.001 | .06 | .001 | .04 | ||||||||||||
| None | 21.1 | 78.9 | 16.5 | 83.5 | 22.6 | 77.4 | 18.4 | 81.6 | 34.8 | 65.2 | 13.4 | 86.6 | ||||||
| <Primary school | 10.0 | 90.0 | 8.2 | 91.8 | 18.3 | 81.7 | 16.9 | 83.1 | 27.7 | 72.3 | 11.1 | 88.9 | ||||||
| Primary school | 7.1 | 92.9 | 10.4 | 89.6 | 18.1 | 81.9 | 15.3 | 84.7 | 48.7 | 51.3 | 13.6 | 86.4 | ||||||
| Secondary/high school | 4.1 | 95.9 | 5.2 | 94.8 | 11.4 | 88.6 | 5.9 | 94.1 | 25.8 | 74.2 | 8.4 | 91.6 | ||||||
| University or postgraduate | 7.0 | 93.0 | 7.0 | 93.0 | 5.4 | 94.6 | 12.8 | 87.2 | 16.0 | 84.0 | 4.9 | 95.1 | ||||||
| Wealth indexb | <.001 | <.001 | <.001 | .11 | <.001 | .04 | ||||||||||||
| 1 (Lowest quintile) | 18.9 | 81.1 | 17.1 | 82.9 | 25.2 | 74.8 | 17.2 | 82.8 | 27.0 | 73.0 | 15.2 | 84.8 | ||||||
| 2 | 11.2 | 88.8 | 14.7 | 85.3 | 18.5 | 81.5 | 18.0 | 82.0 | 31.2 | 68.8 | 10.2 | 89.8 | ||||||
| 3 | 9.4 | 90.6 | 12.8 | 87.2 | 20.4 | 79.6 | 14.9 | 85.1 | 39.3 | 60.7 | 8.1 | 91.9 | ||||||
| 4 | 6.0 | 94.0 | 9.5 | 90.5 | 15.9 | 84.1 | 17.0 | 83.0 | 20.1 | 79.9 | 12.2 | 87.8 | ||||||
| 5 (Highest quintile) | 6.5 | 93.5 | 7.9 | 92.1 | 14.1 | 85.9 | 9.2 | 90.8 | 12.3 | 87.7 | 8.9 | 91.1 | ||||||
| Food insecurity | .008 | <.001 | .02 | .96 | .04 | .08 | ||||||||||||
| Ever | 20.0 | 80.0 | 16.2 | 83.8 | 22.8 | 77.2 | 15.5 | 84.5 | 37.1 | 62.9 | 13.2 | 86.8 | ||||||
| Never | 9.8 | 90.2 | 10.5 | 89.5 | 17.8 | 82.2 | 15.4 | 84.6 | 24.4 | 75.6 | 10.1 | 89.9 | ||||||
| Health insurance | .08 | .11 | .92 | NA | .34 | .001 | ||||||||||||
| Yes | 9.7 | 90.3 | 13.4 | 86.6 | 18.1 | 81.9 | NA | NA | 25.4 | 74.6 | 6.1 | 93.9 | ||||||
| No | 11.6 | 88.4 | 11.5 | 88.5 | 18.5 | 81.5 | NA | NA | 10.6 | 89.4 | 12.2 | 87.8 | ||||||
| Health accessc | .12 | .63 | .47 | .21 | .77 | .13 | ||||||||||||
| Yes | 10.4 | 89.6 | 12.3 | 87.7 | 18.3 | 81.7 | 15.0 | 85.0 | 25.7 | 74.3 | 10.9 | 89.1 | ||||||
| No | 8.4 | 91.6 | 11.5 | 88.5 | 14.9 | 85.1 | 25.6 | 74.4 | 28.4 | 71.6 | 21.4 | 78.6 | ||||||
| Location | <.001 | <.001 | .39 | .96 | .79 | .93 | ||||||||||||
| Urban | 8.5 | 91.5 | 9.4 | 90.6 | 17.0 | 83.0 | 15.5 | 84.5 | 25.7 | 74.3 | 11.0 | 89.0 | ||||||
| Rural | 11.0 | 89.0 | 14.2 | 85.8 | 19.0 | 81.0 | 15.6 | 84.4 | 24.4 | 75.6 | 10.8 | 89.2 | ||||||
| Disabilityd | <.001 | <.001 | <.001 | <.001 | .01 | <.001 | ||||||||||||
| None | 5.9 | 94.1 | 5.8 | 94.2 | 9.0 | 91.0 | 10.9 | 89.1 | 20.0 | 80.0 | 8.3 | 91.7 | ||||||
| Mild | 12.0 | 88.0 | 6.5 | 93.5 | 14.6 | 85.4 | 12.6 | 87.4 | 23.5 | 76.5 | 7.8 | 92.2 | ||||||
| Moderate | 16.8 | 83.2 | 12.7 | 87.3 | 17.3 | 82.7 | 17.5 | 82.5 | 26.4 | 73.6 | 10.1 | 89.9 | ||||||
| Severe | 31.1 | 68.9 | 21.3 | 78.7 | 26.5 | 73.5 | 25.0 | 75.0 | 36.3 | 63.7 | 17.9 | 82.1 | ||||||
| Comorbidities, No.e | <.001 | <.001 | .003 | .53 | .03 | .13 | ||||||||||||
| 0 | 6.5 | 93.5 | 10.2 | 89.8 | 15.7 | 84.3 | 15.8 | 84.2 | 16.5 | 83.5 | 7.7 | 92.3 | ||||||
| 1 | 9.3 | 90.7 | 10.8 | 89.2 | 17.5 | 82.5 | 13.7 | 86.3 | 23.5 | 76.5 | 11.1 | 88.9 | ||||||
| 2-8 | 13.0 | 87.0 | 15.3 | 84.7 | 22.0 | 78.0 | 16.7 | 83.3 | 29.0 | 71.0 | 12.2 | 87.8 | ||||||
| Memoryf | <.001 | <.001 | <.001 | <.001 | <.001 | .07 | ||||||||||||
| Poor | 17.2 | 82.8 | 18.4 | 81.6 | 25.2 | 74.8 | 19.6 | 80.4 | 38.2 | 61.8 | 14.8 | 85.2 | ||||||
| Fair | 8.4 | 91.6 | 11.6 | 88.4 | 17.5 | 82.5 | 13.0 | 87.0 | 22.9 | 77.1 | 9.2 | 90.8 | ||||||
| Good | 7.0 | 93.0 | 9.2 | 90.8 | 13.2 | 86.8 | 14.3 | 85.7 | 18.2 | 81.8 | 10.5 | 89.5 | ||||||
| Very good | 5.3 | 94.7 | 8.7 | 91.3 | 10.6 | 89.4 | 6.9 | 93.1 | 19.0 | 81.0 | 9.8 | 90.2 | ||||||
| Marriedg | <.001 | .04 | <.001 | .31 | .77 | .03 | ||||||||||||
| Yes | 8.1 | 91.9 | 11.2 | 88.8 | 16.1 | 83.9 | 14.6 | 85.4 | 25.8 | 74.2 | 9.2 | 90.8 | ||||||
| No | 20.0 | 80.0 | 13.6 | 86.4 | 26.6 | 73.4 | 17.0 | 83.0 | 24.8 | 75.2 | 12.7 | 87.3 | ||||||
| Social supporth | .12 | .005 | .25 | .38 | .11 | .78 | ||||||||||||
| Low | 10.0 | 90.0 | 17.8 | 82.2 | 19.4 | 80.6 | 16.5 | 83.5 | 24.2 | 75.8 | 11.0 | 89.0 | ||||||
| High | 7.9 | 92.1 | 12.1 | 87.9 | 17.2 | 82.8 | 13.9 | 86.1 | 31.7 | 68.3 | 10.4 | 89.6 | ||||||
| Social participationi | <.001 | <.001 | <.001 | .02 | .24 | .002 | ||||||||||||
| Low | 25.8 | 74.2 | 18.4 | 81.6 | 24.5 | 75.5 | 21.3 | 78.7 | 30.0 | 70.0 | 15.5 | 84.5 | ||||||
| High | 8.8 | 91.2 | 9.7 | 90.3 | 15.1 | 84.9 | 14.0 | 86.0 | 24.8 | 75.2 | 9.6 | 90.4 | ||||||
| Recent eye examinationj | .25 | .02 | .13 | .79 | .27 | .94 | ||||||||||||
| Yes | 7.4 | 92.6 | 15.5 | 84.5 | 16.2 | 83.8 | 15.1 | 84.9 | 23.5 | 76.5 | 10.8 | 89.2 | ||||||
| No | 8.8 | 91.2 | 11.6 | 88.4 | 19.1 | 80.9 | 15.7 | 84.3 | 27.4 | 72.6 | 11.0 | 89.0 | ||||||
Abbreviation: NA, not applicable.
Moderate VI (or worse) was defined as presenting visual acuity worse than 6/18 in the better-seeing eye.
Wealth index was a modeled estimate of household assets and characteristics in each country divided into quintiles.
Health access was defined as ability to access health care when needed.
12-Item World Health Organization Disability Assessment Schedule scores: none, <5%; mild, 5%-13%; moderate, >13%-30%; severe, ≥30%.
Number of comorbidities based on self-report and strongly suggestive symptoms.
Memory score based on number of words correctly recalled: poor, 0-18; fair, 19-22; good, 23-26; very good, 27-40.
Married or cohabitating.
Social support scores: low, 0-1; high, 2-4.
Social participation scores: low, 0-2; high, 3-4.
Recent eye examination defined as within the last 2 years.
The overall proportion of the study sample with distance VI ranged from 9.9% (95% CI, 9.3-10.5) in China to 25.4% (95% CI, 22.0-29.2) in Russia. Only higher age, less wealth, and self-reported disability were significantly associated with distance VI in all 6 countries. In most countries, distance VI was significantly associated with female sex, being unmarried, lower educational attainment, more medical comorbidities, poorer memory, and lower social participation. Distance VI was less consistently associated with living in a rural location, having no health insurance, lower levels of social support, food insecurity, and receipt of a recent eye examination.
Near Vision Impairment
Complete near visual acuity data were available for 12 263 of 13 350 participants (91.9%) from China, 4278 of 4725 (90.5%) from Ghana, 6291 of 7150 (88.0%) from India, 1983 of 2103 (94.3%) from Mexico, 3003 of 3763 (79.8%) from Russia, and 3682 of 3838 (95.9%) from South Africa. The heat map illustrates the magnitudes of the associations of near VI with each covariate (Figure, B). The weighted proportions of participants for each covariate group stratified by near VI are shown in Table 2. Unadjusted and adjusted ORs are displayed in eTable2 in the Supplement.
Table 2. Summary of Associations of Demographic, Socioeconomic, and Health-Related Variables With Near Vision Impairment (VI)a.
| Characteristic | % | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| China | Ghana | India | Mexico | Russia | South Africa | |||||||||||||
| VI (n = 4315) | No VI (n = 7948) | P Value | VI (n = 1179) | No VI (n = 3099) | P Value | VI (n = 2662) | No VI (n = 3629) | P Value | VI (n = 770) | No VI (n = 1213) | P Value | VI (n = 1177) | No VI (n = 1826) | P Value | VI (n = 1278) | No VI (n = 2404) | P Value | |
| Overall prevalence, % (95% CI) | 36.1 (35.0-37.1) | 63.9 (62.9-65.0) | NA | 28.5 (26.9-30.1) | 71.5 (69.9-73.1) | NA | 43.1 (41.1-45.1) | 56.9 (54.9-58.9) | NA | 40.4 (37.1-43.7) | 59.6 (56.3-62.9) | NA | 39.8 (36.0-43.8) | 60.2 (56.2-64.0) | NA | 35.5 (33.0-38.2) | 64.5 (61.8-67.0) | NA |
| Age, y | <.001 | <.001 | .008 | .67 | <.001 | .05 | ||||||||||||
| 50-59 | 26.0 | 74.0 | 22.3 | 77.7 | 40.1 | 59.9 | 41.0 | 59.0 | 34.2 | 65.8 | 32.1 | 67.9 | ||||||
| 60-69 | 39.4 | 60.6 | 27.7 | 72.3 | 44.6 | 55.4 | 40.2 | 59.8 | 35.5 | 64.5 | 39.7 | 60.3 | ||||||
| 70-79 | 49.6 | 50.4 | 32.2 | 67.8 | 48.4 | 51.6 | 43.6 | 56.4 | 46.4 | 53.6 | 36.8 | 63.2 | ||||||
| ≥80 | 62.1 | 37.9 | 47.3 | 52.7 | 48.3 | 51.7 | 37.0 | 63.0 | 66.0 | 34.0 | 40.3 | 59.7 | ||||||
| Sex | <.001 | <.001 | <.001 | .15 | .40 | .31 | ||||||||||||
| Male | 32.2 | 67.8 | 24.7 | 75.3 | 38.1 | 61.9 | 37.8 | 62.2 | 41.3 | 58.7 | 34.1 | 65.9 | ||||||
| Female | 39.9 | 60.1 | 32.6 | 67.4 | 48.3 | 51.7 | 42.7 | 57.3 | 38.8 | 61.2 | 36.7 | 63.3 | ||||||
| Education | <.001 | .004 | .01 | .56 | .001 | .29 | ||||||||||||
| None | 51.3 | 48.7 | 31.5 | 68.5 | 45.0 | 55.0 | 43.8 | 56.2 | 36.2 | 63.8 | 38.0 | 62.0 | ||||||
| <Primary school | 41.7 | 58.3 | 27.8 | 72.2 | 47.2 | 52.8 | 42.1 | 57.9 | 52.0 | 48.0 | 38.6 | 61.4 | ||||||
| Primary school | 34.3 | 65.7 | 25.7 | 74.3 | 43.7 | 56.3 | 37.1 | 62.9 | 53.0 | 47.0 | 34.2 | 65.8 | ||||||
| Secondary/high school | 24.3 | 75.7 | 23.2 | 76.8 | 37.6 | 62.4 | 35.4 | 64.6 | 41.8 | 58.2 | 31.2 | 68.8 | ||||||
| University or postgraduate | 27.2 | 72.8 | 27.8 | 72.2 | 35.1 | 64.9 | 43.5 | 56.5 | 26.9 | 73.1 | 30.8 | 69.2 | ||||||
| Wealth indexb | <.001 | .30 | .56 | .07 | .02 | .79 | ||||||||||||
| 1 (Lowest quintile) | 46.6 | 53.4 | 31.0 | 69.0 | 42.5 | 57.5 | 41.2 | 58.8 | 38.0 | 62.0 | 34.4 | 65.6 | ||||||
| 2 | 39.8 | 60.2 | 28.8 | 71.2 | 44.6 | 55.4 | 47.6 | 52.4 | 45.2 | 54.8 | 37.5 | 62.5 | ||||||
| 3 | 35.8 | 64.2 | 29.8 | 70.2 | 43.0 | 57.0 | 41.9 | 58.1 | 43.3 | 56.7 | 36.7 | 63.3 | ||||||
| 4 | 28.5 | 71.5 | 27.4 | 72.6 | 45.4 | 54.6 | 36.4 | 63.6 | 45.3 | 54.7 | 32.8 | 67.2 | ||||||
| 5 (Highest quintile) | 33.3 | 66.7 | 25.8 | 74.2 | 40.5 | 59.5 | 33.2 | 66.8 | 28.6 | 71.4 | 36.6 | 63.4 | ||||||
| Food insecurity | .19 | .52 | .04 | .01 | .89 | .02 | ||||||||||||
| Ever | 44.0 | 56.0 | 29.4 | 70.6 | 38.4 | 61.6 | 49.2 | 50.8 | 40.6 | 59.4 | 40.7 | 59.3 | ||||||
| Never | 36.0 | 64.0 | 28.2 | 71.8 | 43.8 | 56.2 | 38.0 | 62.0 | 39.8 | 60.2 | 33.6 | 66.4 | ||||||
| Health insurance | .53 | .21 | .44 | NA | .004 | .52 | ||||||||||||
| Yes | 36.2 | 63.8 | 29.8 | 70.2 | 40.0 | 60.0 | NA | NA | 39.7 | 60.3 | 33.9 | 66.1 | ||||||
| No | 35.1 | 64.9 | 27.7 | 72.3 | 43.2 | 56.8 | NA | NA | 74.8 | 25.2 | 36.0 | 64.0 | ||||||
| Health accessc | .14 | .58 | .97 | .13 | .51 | .90 | ||||||||||||
| Yes | 37.5 | 62.5 | 28.1 | 71.9 | 43.8 | 56.2 | 40.2 | 59.8 | 40.1 | 59.9 | 35.3 | 64.7 | ||||||
| No | 33.9 | 66.1 | 29.4 | 70.6 | 44.1 | 55.9 | 22.7 | 77.3 | 46.8 | 53.2 | 33.9 | 66.1 | ||||||
| Location | <.001 | .11 | .57 | .71 | .19 | .66 | ||||||||||||
| Urban | 30.4 | 69.6 | 26.9 | 73.1 | 44.2 | 55.8 | 40.0 | 60.0 | 41.5 | 58.5 | 35.1 | 64.9 | ||||||
| Rural | 40.9 | 59.1 | 29.6 | 70.4 | 42.6 | 57.4 | 41.3 | 58.7 | 34.2 | 65.8 | 36.3 | 63.7 | ||||||
| Disabilityd | <.001 | <.001 | .001 | .77 | .005 | .01 | ||||||||||||
| None | 30.1 | 69.9 | 19.5 | 80.5 | 38.2 | 61.8 | 41.1 | 58.9 | 33.4 | 66.6 | 29.9 | 70.1 | ||||||
| Mild | 40.4 | 59.6 | 24.5 | 75.5 | 38.0 | 62.0 | 43.6 | 56.4 | 37.1 | 62.9 | 34.4 | 65.6 | ||||||
| Moderate | 47.2 | 52.8 | 29.5 | 70.5 | 42.9 | 57.1 | 39.1 | 60.9 | 41.7 | 58.3 | 39.8 | 60.2 | ||||||
| Severe | 58.8 | 41.2 | 39.4 | 60.6 | 48.5 | 51.5 | 39.4 | 60.6 | 51.7 | 48.3 | 39.2 | 60.8 | ||||||
| Comorbidities, No.e | <.001 | <.001 | .71 | .17 | <.001 | .05 | ||||||||||||
| 0 | 28.6 | 71.4 | 22.9 | 77.1 | 42.7 | 57.3 | 43.5 | 56.5 | 25.9 | 74.1 | 31.4 | 68.6 | ||||||
| 1 | 35.2 | 64.8 | 26.6 | 73.4 | 42.3 | 57.7 | 38.7 | 61.3 | 36.4 | 63.6 | 33.6 | 66.4 | ||||||
| 2-8 | 42.5 | 57.5 | 34.8 | 65.2 | 44.1 | 55.9 | 38.8 | 61.2 | 45.5 | 54.5 | 39.5 | 60.5 | ||||||
| Memoryf | <.001 | <.001 | .02 | .22 | <.001 | .005 | ||||||||||||
| Poor | 47.4 | 52.6 | 38.0 | 62.0 | 44.2 | 55.8 | 41.7 | 58.3 | 57.5 | 42.5 | 43.5 | 56.5 | ||||||
| Fair | 35.4 | 64.6 | 31.6 | 68.4 | 46.1 | 53.9 | 38.3 | 61.7 | 38.0 | 62.0 | 29.6 | 70.4 | ||||||
| Good | 30.6 | 69.4 | 23.4 | 76.6 | 40.4 | 59.6 | 35.3 | 64.7 | 36.5 | 63.5 | 33.9 | 66.1 | ||||||
| Very good | 28.6 | 71.4 | 19.1 | 80.9 | 37.2 | 62.8 | 47.7 | 52.3 | 27.5 | 72.5 | 35.5 | 64.5 | ||||||
| Marriedg | <.001 | <.001 | .16 | .28 | .93 | .40 | ||||||||||||
| Yes | 34.1 | 65.9 | 25.7 | 74.3 | 42.4 | 57.6 | 42.3 | 57.7 | 39.7 | 60.3 | 34.6 | 65.4 | ||||||
| No | 47.5 | 52.5 | 32.8 | 67.2 | 45.6 | 54.4 | 38.3 | 61.7 | 40.1 | 59.9 | 36.8 | 63.2 | ||||||
| Social supporth | .98 | .03 | .13 | .41 | .78 | .26 | ||||||||||||
| Low | 36.1 | 63.9 | 34.7 | 65.3 | 44.1 | 55.9 | 38.5 | 61.5 | 40.8 | 59.2 | 35.8 | 64.2 | ||||||
| High | 36.2 | 63.8 | 28.4 | 71.6 | 40.9 | 59.1 | 42.5 | 57.5 | 39.5 | 60.5 | 31.5 | 68.5 | ||||||
| Social participationi | <.001 | <.001 | <.001 | .37 | .36 | .07 | ||||||||||||
| Low | 53.9 | 46.1 | 33.5 | 66.5 | 48.8 | 51.2 | 44.8 | 55.2 | 43.8 | 56.2 | 40.5 | 59.5 | ||||||
| High | 34.9 | 65.1 | 26.5 | 73.5 | 40.1 | 59.9 | 40.3 | 59.7 | 39.3 | 60.7 | 34.4 | 65.6 | ||||||
| Recent eye examinationj | .30 | .38 | .50 | <.001 | .16 | .64 | ||||||||||||
| Yes | 31.8 | 68.2 | 30.1 | 69.9 | 44.4 | 55.6 | 33.5 | 66.5 | 37.4 | 62.6 | 34.5 | 65.5 | ||||||
| No | 34.1 | 65.9 | 28.2 | 71.8 | 42.7 | 57.3 | 46.1 | 53.9 | 42.5 | 57.5 | 35.9 | 64.1 | ||||||
Abbreviation: NA, not applicable.
Moderate VI (or worse) was defined as presenting visual acuity worse than 6/18 in the better-seeing eye.
Wealth index was a modeled estimate of household assets and characteristics in each country divided into quintiles.
Health access was defined as ability to access health care when needed.
12-Item World Health Organization Disability Assessment Schedule scores: none, <5%; mild, 5%-13%; moderate, >13%-30%; severe, ≥30%.
Number of comorbidities based on self-report and strongly suggestive symptoms.
Memory score based on number of words correctly recalled: poor, 0-18; fair, 19-22; good, 23-26; very good, 27-40.
Married or cohabitating.
Social support scores: low, 0-1; high, 2-4.
Social participation scores: low, 0-2; high, 3-4.
Recent eye examination defined as within the last 2 years.
The overall proportion of the study sample with near VI ranged from 28.5% (95% CI, 26.9-30.1) in Ghana to 43.1% (95% CI, 41.1-45.1) in India. We did not identify any covariate that was associated with near VI across all 6 countries. In most countries, near VI was significantly associated with higher age, lower educational attainment, greater self-reported disability, more medical comorbidities, and poorer memory. Near VI was less consistently associated with female sex, being unmarried, living in a rural location, receipt of a recent eye examination, having no health insurance, lower levels of social support and social participation, food insecurity, and less wealth.
Receipt of Eye Care
Complete data on receipt of eye examinations were available for 5341 of 13 350 participants (40.0%) from China, 4302 of 4725 (91.0%) from Ghana, 6559 of 7150 (91.7%) from India, 2007 of 2103 (95.4%) from Mexico, 3747 of 3763 (99.6%) from Russia, and 3771 of 3838 (98.3%) from South Africa. The heat map illustrates the magnitudes of the associations of receipt of a recent eye examination with each covariate (Figure, C). The weighted proportions of participants for each covariate group stratified by eye examinations are shown in Table 3. Unadjusted and adjusted ORs are displayed in eTable 3 in the Supplement.
Table 3. Summary of Associations of Demographic, Socioeconomic, and Health-Related Variables With Receipt of a Recent Eye Examinationa.
| Characteristic | % | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| China | Ghana | India | Mexico | Russia | South Africa | |||||||||||||
| Recent Eye Exam (n = 742) | No Recent Exam (n = 4599) | P Value | Recent Eye Exam (n = 630) | No Recent Exam (n = 3672) | P Value | Recent Eye Exam (n = 1359) | No Recent Exam (n = 5200) | P Value | Recent Eye Exam (n = 892) | No Recent Exam (n = 1115) | P Value | Recent Eye Exam (n = 2109) | No Recent Exam (n = 1638) | P Value | Recent Eye Exam (n = 1018) | No Recent Exam (n = 2753) | P Value | |
| Overall prevalence, % (95% CI) | 15.9 (14.7-17.2) | 84.1 (82.8-85.3) | NA | 15.0 (13.8-16.2) | 85.0 (83.8-86.2) | NA | 21.8 (20.2-23.4) | 78.2 (76.6-79.8) | NA | 41.5 (38.3-44.8) | 58.5 (55.2-61.7) | NA | 53.1 (49.3-56.8) | 46.9 (43.2-50.7) | NA | 27.7 (25.4-30.1) | 72.3 (69.9-74.6) | NA |
| Age, y | .20 | .10 | .47 | .39 | .001 | .11 | ||||||||||||
| 50-59 | 15.4 | 84.6 | 13.6 | 86.4 | 21.2 | 78.8 | 43.7 | 56.3 | 60.5 | 39.5 | 26.5 | 73.5 | ||||||
| 60-69 | 16.7 | 83.3 | 14.3 | 85.7 | 22.8 | 77.2 | 44.1 | 55.9 | 50.3 | 49.7 | 27.4 | 72.6 | ||||||
| 70-79 | 17.0 | 83.0 | 17.3 | 82.7 | 22.7 | 77.3 | 40.2 | 59.8 | 46.7 | 53.3 | 34.4 | 65.6 | ||||||
| ≥80 | 10.8 | 89.2 | 17.0 | 83.0 | 17.3 | 82.7 | 35.8 | 64.2 | 38.1 | 61.9 | 23.3 | 76.7 | ||||||
| Sex | .28 | .26 | .18 | .02 | .46 | .001 | ||||||||||||
| Male | 15.3 | 84.7 | 15.6 | 84.4 | 22.8 | 77.2 | 37.3 | 62.7 | 51.1 | 48.9 | 32.0 | 68.0 | ||||||
| Female | 16.6 | 83.4 | 14.2 | 85.8 | 20.7 | 79.3 | 45.2 | 54.8 | 54.4 | 45.6 | 24.3 | 75.7 | ||||||
| Education | <.001 | <.001 | <.001 | .004 | .17 | <.001 | ||||||||||||
| None | 9.6 | 90.4 | 11.6 | 88.4 | 16.1 | 83.9 | 35.1 | 64.9 | 52.5 | 47.5 | 14.9 | 85.1 | ||||||
| <Primary school | 12.2 | 87.8 | 9.7 | 90.3 | 20.5 | 79.5 | 39.9 | 60.1 | 42.6 | 57.4 | 23.4 | 76.6 | ||||||
| Primary school | 16.5 | 83.5 | 14.3 | 85.7 | 22.9 | 77.1 | 45.2 | 54.8 | 39.1 | 60.9 | 23.1 | 76.9 | ||||||
| Secondary/high school | 21.4 | 78.6 | 21.9 | 78.1 | 29.6 | 70.4 | 37.9 | 62.1 | 53.6 | 46.4 | 38.8 | 61.2 | ||||||
| University or postgraduate | 37.5 | 62.5 | 42.7 | 57.3 | 48.1 | 51.9 | 62.6 | 37.4 | 56.0 | 44.0 | 58.7 | 41.3 | ||||||
| Wealth indexb | <.001 | <.001 | <.001 | .009 | .06 | <.001 | ||||||||||||
| 1 (Lowest quintile) | 7.5 | 92.5 | 6.4 | 93.6 | 14.4 | 85.6 | 34.5 | 65.5 | 44.0 | 56.0 | 11.5 | 88.5 | ||||||
| 2 | 12.1 | 87.9 | 10.9 | 89.1 | 18.1 | 81.9 | 37.1 | 62.9 | 48.9 | 51.1 | 24.9 | 75.1 | ||||||
| 3 | 17.5 | 82.5 | 12.7 | 87.3 | 19.2 | 80.8 | 40.7 | 59.3 | 51.8 | 48.2 | 18.8 | 81.2 | ||||||
| 4 | 18.2 | 81.8 | 15.1 | 84.9 | 22.9 | 77.1 | 49.2 | 50.8 | 58.6 | 41.4 | 33.1 | 66.9 | ||||||
| 5 (Highest quintile) | 23.5 | 76.5 | 28.0 | 72.0 | 31.5 | 68.5 | 49.0 | 51.0 | 58.9 | 41.1 | 48.6 | 51.4 | ||||||
| Food insecurity | .93 | .004 | <.001 | <.001 | .62 | <.001 | ||||||||||||
| Ever | 15.3 | 84.7 | 12.2 | 87.8 | 12.5 | 87.5 | 31.0 | 69.0 | 57.7 | 42.3 | 16.3 | 83.7 | ||||||
| Never | 15.9 | 84.1 | 16.2 | 83.8 | 23.2 | 76.8 | 45.0 | 55.0 | 52.7 | 47.3 | 32.5 | 67.5 | ||||||
| Health insurance | .05 | <.001 | <.001 | NA | .44 | <.001 | ||||||||||||
| Yes | 16.5 | 83.5 | 21.0 | 79.0 | 43.2 | 56.8 | NA | NA | 53.2 | 46.8 | 42.9 | 57.1 | ||||||
| No | 13.4 | 86.6 | 11.3 | 88.7 | 20.9 | 79.1 | NA | NA | 36.9 | 63.1 | 23.9 | 76.1 | ||||||
| Health accessc | .79 | <.001 | .93 | .78 | .20 | .66 | ||||||||||||
| Yes | 16.0 | 84.0 | 16.3 | 83.7 | 22.8 | 77.2 | 44.2 | 55.8 | 55.5 | 44.5 | 31.0 | 69.0 | ||||||
| No | 16.7 | 83.3 | 10.1 | 89.9 | 23.4 | 76.6 | 47.5 | 52.5 | 77.7 | 22.3 | 37.0 | 63.0 | ||||||
| Location | <.001 | <.001 | <.001 | <.001 | .22 | <.001 | ||||||||||||
| Urban | 21.1 | 78.9 | 21.1 | 78.9 | 29.0 | 71.0 | 44.9 | 55.1 | 55.0 | 45.0 | 33.0 | 67.0 | ||||||
| Rural | 7.5 | 92.5 | 10.7 | 89.3 | 18.8 | 81.2 | 32.2 | 67.8 | 47.9 | 52.1 | 17.9 | 82.1 | ||||||
| Disabilityd | .04 | .49 | .32 | .52 | <.001 | <.001 | ||||||||||||
| None | 15.4 | 84.6 | 14.0 | 86.0 | 25.1 | 74.9 | 38.6 | 61.4 | 46.5 | 53.5 | 35.3 | 64.7 | ||||||
| Mild | 17.6 | 82.4 | 13.7 | 86.3 | 22.4 | 77.6 | 41.9 | 58.1 | 63.5 | 36.5 | 25.8 | 74.2 | ||||||
| Moderate | 17.9 | 82.1 | 15.9 | 84.1 | 20.6 | 79.4 | 45.0 | 55.0 | 55.9 | 44.1 | 26.5 | 73.5 | ||||||
| Severe | 9.3 | 90.7 | 15.8 | 84.2 | 21.2 | 78.8 | 41.5 | 58.5 | 42.6 | 57.4 | 20.4 | 79.6 | ||||||
| Comorbidities, No.e | <.001 | .04 | .05 | <.001 | .004 | .45 | ||||||||||||
| 0 | 13.6 | 86.4 | 12.7 | 87.3 | 19.2 | 80.8 | 29.8 | 70.2 | 44.8 | 55.2 | 29.0 | 71.0 | ||||||
| 1 | 14.8 | 85.2 | 14.9 | 85.1 | 22.1 | 77.9 | 44.2 | 55.8 | 46.3 | 53.7 | 25.9 | 74.1 | ||||||
| 2-8 | 19.7 | 80.3 | 16.8 | 83.2 | 24.0 | 76.0 | 49.2 | 50.8 | 58.3 | 41.7 | 29.2 | 70.8 | ||||||
| Memoryf | <.001 | .77 | <.001 | .003 | <.001 | <.001 | ||||||||||||
| Poor | 12.1 | 87.9 | 14.5 | 85.5 | 17.8 | 82.2 | 36.1 | 63.9 | 44.5 | 55.5 | 24.8 | 75.2 | ||||||
| Fair | 16.2 | 83.8 | 14.6 | 85.4 | 21.4 | 78.6 | 47.2 | 52.8 | 49.5 | 50.5 | 25.2 | 74.8 | ||||||
| Good | 16.6 | 83.4 | 16.1 | 83.9 | 22.4 | 77.6 | 49.5 | 50.5 | 56.0 | 44.0 | 22.3 | 77.7 | ||||||
| Very good | 20.2 | 79.8 | 14.6 | 85.4 | 32.2 | 67.8 | 34.8 | 65.2 | 63.1 | 36.9 | 35.2 | 64.8 | ||||||
| Marriedg | .04 | .19 | .10 | .39 | .50 | <.001 | ||||||||||||
| Yes | 16.5 | 83.5 | 15.7 | 84.3 | 22.4 | 77.6 | 40.5 | 59.5 | 54.1 | 45.9 | 32.7 | 67.3 | ||||||
| No | 13.3 | 86.7 | 14.0 | 86.0 | 19.4 | 80.6 | 43.4 | 56.6 | 51.6 | 48.4 | 21.4 | 78.6 | ||||||
| Social supporth | .008 | .29 | .01 | .22 | .28 | .54 | ||||||||||||
| Low | 15.6 | 84.4 | 12.6 | 87.4 | 19.6 | 80.4 | 40.2 | 59.8 | 51.9 | 48.1 | 26.6 | 73.4 | ||||||
| High | 22.7 | 77.3 | 14.7 | 85.3 | 23.9 | 76.1 | 35.2 | 64.8 | 56.6 | 43.4 | 29.0 | 71.0 | ||||||
| Social participationi | .04 | .03 | .12 | .10 | <.001 | .14 | ||||||||||||
| Low | 11.0 | 89.0 | 17.1 | 82.9 | 19.9 | 80.1 | 35.6 | 64.4 | 40.7 | 59.3 | 24.1 | 75.9 | ||||||
| High | 16.1 | 83.9 | 13.9 | 86.1 | 22.4 | 77.6 | 42.6 | 57.4 | 56.9 | 43.1 | 28.8 | 71.2 | ||||||
| Vision impairmentj | .05 | .05 | .81 | .35 | .34 | .37 | ||||||||||||
| None | 17.5 | 82.5 | 13.6 | 86.4 | 22.3 | 77.7 | 45.2 | 54.8 | 54.9 | 45.1 | 29.6 | 70.4 | ||||||
| Mild | 14.7 | 85.3 | 16.0 | 84.0 | 21.8 | 78.2 | 39.1 | 60.9 | 49.7 | 50.3 | 25.2 | 74.8 | ||||||
| Moderate | 13.6 | 86.4 | 19.3 | 80.7 | 21.6 | 78.4 | 40.0 | 60.0 | 49.1 | 50.9 | 27.6 | 72.4 | ||||||
| Severe or blindness | 11.8 | 88.2 | 17.3 | 82.7 | 18.7 | 81.3 | 35.0 | 65.0 | 42.7 | 57.3 | 33.5 | 66.5 | ||||||
Abbreviation: NA, not applicable.
Recent eye examination defined as within the last 2 years.
Wealth index was a modeled estimate of household assets and characteristics in each country divided into quintiles.
Health access was defined as ability to access health care when needed.
12-Item World Health Organization Disability Assessment Schedule scores: none, <5%; mild, 5%-13%; moderate, >13%-30%; severe, ≥30%.
Number of comorbidities based on self-report and strongly suggestive symptoms.
Memory score based on number of words correctly recalled: poor, 0-18; fair, 19-22; good, 23-26; very good, 27-40.
Married or cohabitating.
Social support scores: low, 0-1; high, 2-4.
Social participation scores: low, 0-2; high, 3-4.
Mild vision impairment indicates visual acuity between worse than 6/12 and 6/18 or better; moderate vision impairment, between worse than 6/18 and 6/60 or better; severe vision impairment, worse than 6/60 and 6/120 or better; and blindness, worse than 6/120.
The overall proportion of the study sample that had received a recent eye examination ranged from 15.0% (95% CI, 13.8-16.2) in Ghana to 53.1% (95% CI, 49.3-56.8) in Russia. The only covariate associated with receipt of a recent eye examination in all 6 countries was greater wealth. In most countries, receipt of a recent eye examination was significantly associated with higher educational attainment, urban location, more medical comorbidities, better memory, and not experiencing food insecurity. Receipt of a recent eye examination was less consistently associated with age, sex, being married, vision impairment, self-reported disability, access to health care, having health insurance, and levels of social support and participation.
Discussion
In this study, using data from SAGE in 6 LMICs, we found that VI and receipt of eye care among older adults were associated with a distinct group of sociodemographic features in each country. Our study makes several important and novel contributions. First, because each of the 6 SAGE countries used the same survey, we are able to make cross-national comparisons that suggest how different environmental, social, and economic factors are associated with vision in different places. Second, the SAGE data set contains variables that are not routinely collected in epidemiologic field studies, which permitted us to examine constructs like social participation, food insecurity, and household consumption that have not been widely considered in prior investigations. While we have provided results adjusted for age and socioeconomics, bivariate associations may provide the most pertinent information for health policy planners seeking to identify and target those most likely to be affected by VI and a lack of eye care.
Few other studies have used SAGE data to investigate associations with VI.22,23,24 Vellakkal et al23 showed that a 5% reduction of VI among the poor and an 11% reduction among the less educated would eradicate socioeconomic disparities in the prevalence of VI in India. Also using SAGE data, Akuamoah-Boateng et al24 found that more than half of the population of Kassena-Nankana District, Ghana, reported difficulty seeing, but less than 3% used corrective lenses. We have built on these findings by identifying context-specific sociodemographic factors associated with VI and eye care, which our study suggests could help guide targeted public health efforts in these countries.
Our study found that older age, lower educational attainment, greater disability, having more medical comorbidities, and poorer memory were significantly associated with both distance and near VI in most SAGE countries. Other characteristics, including female sex, being unmarried, low social participation, and having less wealth, were commonly associated with distance VI but were less consistently associated with near VI. Some of these traits may represent common criteria that could be used to identify those most likely to have specific forms of VI across LMICs, although additional work should be done to determine the generalizability of these findings.
Importantly, we also identified traits that were significantly associated with VI only in specific countries, which may reflect unique cultural or environmental circumstances. For example, those without health insurance had approximately twice the odds of distance VI in South Africa and 4 times the odds of near VI in Russia, although health insurance was not significantly associated with either form of VI in other SAGE countries. Russia was also 1 of 3 SAGE countries (along with China and the lowest wealth quintile in Ghana) in which near VI was associated with less wealth. Conversely, distance VI was associated with less wealth in all SAGE countries. These patterns could reflect differences between countries in their health insurance systems and the out-of-pocket cost of eye care. While there may be relative ease of access to presbyopia correction in most SAGE countries, financial barriers could play a larger role for those in need of higher level refractive, medical, or surgical services to prevent or treat distance VI.
In at least half of SAGE countries, adults were less likely to have received a recent eye examination if they had lower educational attainment, lived in a rural location, had fewer comorbidities, experienced food insecurity, or lacked health insurance. Other investigations have also shown that access and use of eye care services are comparatively low in rural locations in LMICs.25,26,27 In a population-based study in Andhra Pradesh, India,27 only 34.1% of those with unilateral VI who had noticed a change in vision had sought eye care. Those with more education were more likely to have received care, and having less financial resources was the most common reason for not getting care. In a separate study,28 the same investigators found that 41.0% of adults with VI in urban areas of Andhra Pradesh had sought eye care. Our results were similar, with older adults from rural areas in all countries except Russia being less likely to have had a recent eye examination, even independent of age and socioeconomic status.
Prior studies have also explored the association of sex with VI and eye care.29,30,31 Globally, women are 30% more likely than men to be blind.30 In a systematic review, Ramke et al32 found that women have higher rates of blindness than men in all regions of the world, a higher prevalence of trachoma, and, in some places, are less likely to receive cataract surgery. Our results also revealed sex-based disparities, although this was not consistent across all countries. Women were significantly more likely than men to have distance VI in China, Ghana, India, and South Africa but not in Mexico or Russia. The most extreme sex difference was found in China, where the proportion of women with distance VI was 56% greater than for men. In China, Ghana, and India, women were also significantly more likely than men to have near VI. Moreover, the fact that women did not report greater receipt of eye care in countries where they were found to have excess VI is concerning and should be addressed to achieve the United Nations’ Sustainable Development Goals aimed at promoting good health and well-being, achieving gender equality, and reducing inequalities.33
Limitations
There are several limitations to this study. Responses to survey questions may be affected by recall and desirability biases. There were considerable missing data (60.0%) on receipt of eye examinations in China, although data were available for more than 90% of respondents in all other countries. The reason for this is not clear, but results related to this outcome for China should be interpreted with caution. Finally, additional work should be done in other LMICs to elucidate associations of common and unique factors with VI and receipt of eye care, as these may help guide efforts to combat avoidable and preventable VI and blindness where they are needed most.
Conclusions
Cross-national comparisons suggest that there are factors commonly associated with VI and receipt of eye care across LMICs as well as factors that are context-specific. Additional local epidemiological studies are needed to better understand who is most likely to be affected by VI and least likely to receive eye care as well as to design contextually relevant interventions to reduce the burden of preventable VI.
eTable 1. Odds of distance vision impairment in low- and middle-income countries.
eTable 2. Odds of near vision impairment in low- and middle-income countries.
eTable 3. Odds of having had a recent eye exam in low- and middle-income countries.
References
- 1.Flaxman SR, Bourne RRA, Resnikoff S, et al. ; Vision Loss Expert Group of the Global Burden of Disease Study . Global causes of blindness and distance vision impairment 1990-2020: a systematic review and meta-analysis. Lancet Glob Health. 2017;5(12):e1221-e1234. doi: 10.1016/S2214-109X(17)30393-5 [DOI] [PubMed] [Google Scholar]
- 2.World Health Organization Universal Eye Health: A Global Action Plan, 2014-2019. Geneva, Switzerland: World Health Organization; 2013. [Google Scholar]
- 3.Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006;3(11):e442. doi: 10.1371/journal.pmed.0030442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.United Nations Department of Economic and Social Affairs World Population Prospects: The 2017 Revision. New York, NY: United Nations; 2017. [Google Scholar]
- 5.Wang W, Yan W, Müller A, Keel S, He M. Association of socioeconomics with prevalence of visual impairment and blindness. JAMA Ophthalmol. 2017;135(12):1295-1302. doi: 10.1001/jamaophthalmol.2017.3449 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Bollyky TJ, Templin T, Cohen M, Dieleman JL. Lower-income countries that face the most rapid shift in noncommunicable disease burden are also the least prepared. Health Aff (Millwood). 2017;36(11):1866-1875. doi: 10.1377/hlthaff.2017.0708 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.India State-level Disease Burden Initiative Collaborators Nations within a nation: variations in epidemiological transition across the states of India, 1990-2016 in the Global Burden of Disease Study [published online November 13, 2017]. Lancet Lond Engl. doi: 10.1016/S0140-6736(17)32804-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.World Health Organization WHO World Health Survey. http://www.who.int/healthinfo/survey/en/. Accessed December 6, 2017.
- 9.Kowal P, Chatterji S, Naidoo N, et al. ; SAGE Collaborators . Data resource profile: the World Health Organization Study on Global Ageing and Adult Health (SAGE). Int J Epidemiol. 2012;41(6):1639-1649. doi: 10.1093/ije/dys210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.World Health Organization WHO SAGE Survey Manual: The WHO Study on Global AGEing and Adult Health (SAGE). Geneva, Switzerland: World Health Organization; 2006. [Google Scholar]
- 11.World Health Organization International Classification of Diseases (ICD). http://www.who.int/classifications/icd/en/. Accessed December 7, 2017.
- 12.World Health Organization WHO Disability Assessment Schedule 2.0 (WHODAS 2.0). http://www.who.int/classifications/icf/whodasii/en/. Accessed December 22, 2017.
- 13.Tulsky DS, Carlozzi N, Chiaravalloti ND, et al. . NIH Toolbox Cognition Battery (NIHTB-CB): list sorting test to measure working memory. J Int Neuropsychol Soc. 2014;20(6):599-610. doi: 10.1017/S135561771400040X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.World Health Organization International Classification of Functioning, Disability and Health. Geneva, Switzerland: World Health Organization; 2001. [Google Scholar]
- 15.Rejeski WJ, Ip EH, Marsh AP, Miller ME, Farmer DF. Measuring disability in older adults: the International Classification System of Functioning, Disability and Health (ICF) framework. Geriatr Gerontol Int. 2008;8(1):48-54. doi: 10.1111/j.1447-0594.2008.00446.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Lee S, Seo DH, Kim KM, et al. . Contingent association between the size of the social support network and osteoporosis among Korean elderly women. PLoS One. 2017;12(7):e0180017. doi: 10.1371/journal.pone.0180017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.de Brito TRP, Nunes DP, Corona LP, da Silva Alexandre T, de Oliveira Duarte YA. Low supply of social support as risk factor for mortality in the older adults. Arch Gerontol Geriatr. 2017;73:77-81. doi: 10.1016/j.archger.2017.07.016 [DOI] [PubMed] [Google Scholar]
- 18.Stagg BC, Choi H, Woodward MA, Ehrlich JR. Association of social support network size with receipt of cataract surgery in older adults. JAMA Ophthalmol. 2018;136(4):423-427. doi: 10.1001/jamaophthalmol.2018.0244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zamora-Macorra M, de Castro EFA, Ávila-Funes JA, et al. . The association between social support and cognitive function in Mexican adults aged 50 and older. Arch Gerontol Geriatr. 2017;68:113-118. doi: 10.1016/j.archger.2016.10.005 [DOI] [PubMed] [Google Scholar]
- 20.Moore JC, Stinson LL, Welniak EJ. Income measurement error in surveys: a review. J Off Stat. 2000;16(4):331-361. [Google Scholar]
- 21.Ferguson BD, Tandon A, Gakidou E, Murray CJ. Estimating Permanent Income Using Indicator Variables. Geneva, Switzerland: World Health Organization; 2003. [Google Scholar]
- 22.Arokiasamy P, Uttamacharya U, Jain K, et al. . The impact of multimorbidity on adult physical and mental health in low- and middle-income countries: what does the study on global ageing and adult health (SAGE) reveal? BMC Med. 2015;13:178. doi: 10.1186/s12916-015-0402-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Vellakkal S, Subramanian SV, Millett C, Basu S, Stuckler D, Ebrahim S. Socioeconomic inequalities in non-communicable diseases prevalence in India: disparities between self-reported diagnoses and standardized measures. PLoS One. 2013;8(7):e68219. doi: 10.1371/journal.pone.0068219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Akuamoah-Boateng H. Self-reported vision health status among older people in the Kassena-Nankana District, Ghana. Glob Health Action. 2013;6:19012. doi: 10.3402/gha.v6i0.19012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Salomão SR, Cinoto RW, Berezovsky A, et al. . Prevalence and causes of visual impairment in low-middle income school children in Sao Paulo, Brazil. Invest Ophthalmol Vis Sci. 2008;49(10):4308-4313. doi: 10.1167/iovs.08-2073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhao J, Ellwein LB, Cui H, et al. . Prevalence of vision impairment in older adults in rural China: the China Nine-Province Survey. Ophthalmology. 2010;117(3):409-416, 416.e1. doi: 10.1016/j.ophtha.2009.11.023 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Marmamula S, Giridhar P, Khanna RC. Utilization of eye care services among those with unilateral visual impairment in rural South India: Andhra Pradesh Eye Disease Study (APEDS). Int J Ophthalmol. 2017;10(3):473-479. doi: 10.18240/ijo.2017.03.23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Dandona R, Dandona L, Naduvilath TJ, McCarty CA, Rao GN. Utilisation of eyecare services in an urban population in southern India: the Andhra Pradesh Eye Disease Study. Br J Ophthalmol. 2000;84(1):22-27. doi: 10.1136/bjo.84.1.22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lou L, Ye X, Xu P, et al. . Association of sex with the global burden of cataract. JAMA Ophthalmol. 2018;136(2):116-121. doi: 10.1001/jamaophthalmol.2017.5668 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Bourne RRA, Flaxman SR, Braithwaite T, et al. ; Vision Loss Expert Group . Magnitude, temporal trends, and projections of the global prevalence of blindness and distance and near vision impairment: a systematic review and meta-analysis. Lancet Glob Health. 2017;5(9):e888-e897. doi: 10.1016/S2214-109X(17)30293-0 [DOI] [PubMed] [Google Scholar]
- 31.Abou-Gareeb I, Lewallen S, Bassett K, Courtright P. Gender and blindness: a meta-analysis of population-based prevalence surveys. Ophthalmic Epidemiol. 2001;8(1):39-56. doi: 10.1076/opep.8.1.39.1540 [DOI] [PubMed] [Google Scholar]
- 32.Ramke J, Zwi AB, Palagyi A, Blignault I, Gilbert CE. Equity and blindness: closing evidence gaps to support universal eye health. Ophthalmic Epidemiol. 2015;22(5):297-307. doi: 10.3109/09286586.2015.1077977 [DOI] [PubMed] [Google Scholar]
- 33.United Nations Transforming Our World: The 2030 Agenda for Sustainable Development. New York, NY: United Nations; 2015. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Odds of distance vision impairment in low- and middle-income countries.
eTable 2. Odds of near vision impairment in low- and middle-income countries.
eTable 3. Odds of having had a recent eye exam in low- and middle-income countries.