Visual impairment and blindness in rural central India: the Central India Eye and Medical Study

Recent population-based studies have revealed that age-related macular degeneration, glaucoma and retinal vascular diseases including diabetic retinopathy and retinal vein occlusions have been the most common causes for visual impairment in elderly populations (Buch et al. 2001; Ramrattan et al. 2001; Jonas et al. 2009; Gunnlaugsdottir et al. 2010; Song et al. 2010). We examined the prevalence of visual impairment and its associated factors in a rural central Indian population relatively untouched by modern civilization.

The Central India Eye and Medical Study (CIEMS) is a population-based cross-sectional study in central India (Jonas et al. 2010; Nangia et al. 2010a). The Medical Ethics Committee of the Medical Faculty Mannheim of the Ruprecht-Karls-University Heidelberg and the ethical committee of Suraj Eye Institute/Nagpur approved the study, and all participants gave informed consent, according to the Declaration of Helsinki. Inclusion criterion was an age of 30+ years. Of the 5885 eligible subjects, 4711 people participated (response rate, 80.1%). On the basis of presenting visual acuity (PRVA) and using the World Health Organization (WHO) standard and the United States (US) standard, blindness (PRVA-blindness) was defined as PRVA in the better-seeing eye of <20/400 and of ≤20/200, respectively. Visual impairment (PRVA-visual impairment) was defined as presenting vision of <20/60 and ≥20/400, and of <20/40 and >20/200, respectively. In a similar manner, On the basis of best-corrected visual acuity (BCVA) and using the WHO standard and the US, blindness (BCVA-blindness) was defined as BCVA in the better-seeing eye of <20/400 and of ≤20/200, respectively, and visual impairment (BCVA-visual impairment) was defined as best-corrected vision of <20/60 and ≥20/400, and of <20/40 and >20/200, respectively. Undercorrection of refractive error was defined as an improvement in PRVA of the examined eye by at least two lines with best refractive correction. An undercorrection of refractive error as primary cause for visual impairment was defined as PRVA of <6/18, which could be improved by optimizing the correction of the refractive error to at least 6/18. Only the measurements of BCVA or of PRVA were taken as the basis for the assessment of visual impairment or blindness. Statistical analysis was performed using a commercially available statistical software package (SPSS for Windows, version 19.0; IBM-SPSS Inc., Chicago, IL, USA). The frequency of visual impairment and blindness were given as proportions and 95% confidence intervals (CI). The mean values of normally distributed parameters were given as mean ± standard deviation. Analyses were stratified by age and gender. For the comparison of means, we used the parametric student’s t-test if the parameters showed a Gaussian distribution, tested by the Kolgorov–Smirnov test. If the parameters were not normally distributed, we applied the nonparametric Mann–Whitney U-test. Multiple logistic regression analysis was used to analyse the factors related to presence or absence of visual impairment and blindness. All p-values were two-sided and were considered statistically significant when the values were less than 0.05. Owing to the differences in the demographic structures between the CIEMS and the general rural population of India, all prevalence figures for visual impairment and blindness were additionally corrected and adjusted for age, using the national Indian census 2001 (National Indian Census 2001).

Visual acuity measurements were available for 9411 (99.9%) eyes of 4706 (99.9%) individuals of the 4711 subjects participating in the study. The mean age was 49.5 ± 13.4 years (range, 30–100 years). On the basis of PRVA and using the WHO definition, 1049 (22%; 95% CI: 21.1, 23.5) subjects fulfilled the criteria for PRVA-visual impairment, and 35 (0.7%; 95% CI: 0.5, 1.0) subjects were blind. Based on the US standards, 1290 (27.4%; 95% CI: 26.1, 28.7) subjects had PRVA-visual impairment, and 116 (2.5%; 95% CI: 2.0, 2.9) subjects were blind. The prevalence of combined PRVA-visual impairment/blindness (WHO definition) increased to 77% in the subjects aged 80+ years. In 729 subjects of the 1084 subjects with PRVA-visual impairment or PRVA-blindness (67% of the visually impaired or blind population; 16% of the total study population), visual acuity could be improved by using best-correcting glasses, so that the definition of visual impairment or blindness was no longer fulfilled. Using the data of the national Indian census from 2001, the age-standardized prevalence of PRVA-visual impairment (WHO definition) was 17%, and for PRVA-blindness (WHO definition) 0.5%. In a similar manner, the age-standardized prevalence of PRVA-visual impairment (US definition) was 21%, and for PRVA-blindness (US definition) 1.9%.

On the basis of BCVA and using the WHO definition, 333 (7.1%; 95% CI: 6.3, 7.8) subjects fulfilled the criteria for visual impairment, and 22 (0.5%; 95% CI: 0.3, 0.7) subjects were blind based on the BCVA in the better eye (Fig. 1). The age-standardized prevalence of visual impairment was 5% for the population aged 30+ years, and the age-standardized prevalence of blindness was 0.3%. The prevalence of BCVA-visual impairment and blindness (WHO definition) increased to 39% and 5.4%, respectively, in the subjects aged 80+ years. Using the US definition, 473 (10%; 95% CI: 9.2, 10.9) subjects had visual impairment, and 31 (0.7%; 95% CI: 0.4, 0.9) subjects were blind. The corresponding age-standardized prevalence of visual impairment was 8% for the population aged 30+ years, and the age-standardized prevalence of blindness was 0.5%. The prevalence of BCVA-visual impairment and blindness (US definition) increased to 49% and 5%, respectively, in the subjects aged 80+ years.

Using the WHO criteria, univariate analysis showed that the prevalence of BCVA-visual impairment was significantly correlated with higher age (p < 0.001), lower education level (p < 0.001) and more myopic refractive error (p < 0.001). It was not significantly associated with gender (p = 0.11). In a similar manner, the prevalence of BCVA-blindness was related to higher age (p < 0.001) and lower education level (p = 0.001). It was independent of gender (p = 0.40) and myopic refraction error (p = 0.93). Combining BCVA-visual impairment and BCVA-blindness showed significant associations with higher age (p < 0.001), lower education level (p < 0.001) and more myopic refractive error (p < 0.001). It was not significantly associated with gender (p = 0.19). In multiple logistic regression analysis including the parameters that were significant for visual impairment/blindness in univariate analysis, prevalence of combined BCVA-low vision/blindness remained significantly (p < 0.001) associated with higher age, more myopic refractive error and lower level of educational background. Again, gender was not significantly associated (p = 0.30). In a subgroup of the study population including only subjects with an age of 70+ years, the percentage of subjects with visual impairment or blindness was significantly higher for the women than that for the men (42% versus 31%; p = 0.01). In a multiple logistic regression analysis in that age-defined subgroup, the prevalence of visual impairment/blindness was related to higher age (p = 0.04), more myopic refractive error (p = 0.003), lower level of education (p = 0.01) and female gender (p = 0.04).

The most frequent cause for BCVA-low vision/blindness was cataract (75%), followed by postoperative posterior capsular opacification (4%), surgical complications (2%), corneal opacities (2%), age-related macular degeneration (2%), other macular diseases (1%), glaucoma (1%) and other optic nerve or retinal diseases (4%). Among subjects with bilateral cataract surgery (n = 84), 19 (23%) subjects were visually impaired (WHO standard) and 1 (1%) subjects was blind. The reasons for visual impairment and blindness (WHO definition) in the subjects after bilateral cataract surgery were surgical complications after cataract surgery [5 (26%) subjects], secondary cataract [3 (16%) subjects], age-related macular degeneration [3 (16%) subjects], glaucoma [1 (5%) subject], nonglaucomatous optic nerve damage [1 (5%) subject], retinal detachment [1 (5%) subject] and others. Applying the US definitions, 31 (37%) of 84 subjects with bilateral cataract surgery were visually impaired, and 9 (11%) subjects were blind. The reasons for visual impairment (US definition) were surgical complications [12 (39%) subjects], secondary cataract [9 (29%) subjects], age-related macular degeneration [3 (10%) subjects], glaucoma [1 (3%) subject], nonglaucomatous optic nerve damage [1 (3%) subject], retinal detachment [1 (3%) subject] and others. The reasons for blindness (US definition) were surgical complications [2 (22%) subjects], age-related macular degeneration [2 (22%) subjects], glaucoma [1 (11%) subject], retinal detachment [1 (11%) subject] and others. With increasing age, the proportion of cataract and posterior capsular opacity as reasons for visual impairment increased (cataract: increase from 72% for the subjects aged less than 70 years to 79% for subjects aged ≥70 years), and the proportion of glaucomatous optic neuropathy, corneal opacities and myopic degenerative retinopathy slightly decreased. Overall, cataract was by far the outstandingly most common cause for BCVA-visual impairment and BCVA-blindness.

The data indicate that cataract was by far the main contributor to the exponential increase in the prevalence of BCVA-blindness and BCVA-visual impairment in persons aged 30 years or older (Fig. 1). There was no marked qualitative difference in the various age groups in the ranking of cataract as the primary cause of visual impairment and blindness. In all age groups, cataract was by far the most frequent cause. Age-related macular degeneration and vascular retinopathies including diabetic retinopathy played an almost negligible role as frequent cause for visual impairment. The results of our study with respect to the causes for visual impairment agree with two previous studies from south India, that is, the Aravind Comprehensive Eye Survey and the Andhra Pradesh Eye Study (Dandona et al. 1999, 2001; Thulasiraj et al. 2003), in the frequency of cataract as major reason for low vision/blindness. In contrast to our study, major causes for visual impairment were glaucoma and diabetic retinopathy in the Los Angeles Latino Eye Study on a Hispanic population group (Varma et al. 2004). As compared with our study population, the rural Chinese study population of the Handan Study showed a considerably lower prevalence of PRVA-visual impairment (4.7%), PRVA-blindness (0.6%), BCVA-visual impairment (1%) and BCVA-blindness (0.5%) (Liang et al. 2008). One of the most striking findings of our study was that for two-thirds (67%) of the visually impaired or blind subjects, visual impairment could be treated by simply supplying correct glasses. It is in agreement with the Handan Study (Liang et al. 2008). Potential limitations of our study are, first, that optic nerve- and particularly glaucoma-related blindness may have been underestimated in our and other population-based studies as the visual field was usually not taken for the definition of blindness, and central visual acuity is affected only in the late stage of glaucoma. Second, the majority of visually impaired subjects had advanced cataract, so that additional retinal or optic nerve disease leading to the visual impairment may have been overlooked. Third, for few subjects, a morphologic cause for the visual impairment was not detected. As these subjects made out less 8% of the whole group of subjects with visual impairment, this weakness of the study may not have markedly influenced the study results and conclusions.

In conclusion, the prevalence of PRVA-visual impairment/blindness (WHO definition) in the adult population of rural central India was 22% with a steep increase to 77% in the subjects aged 80+ years. The outstandingly most frequent cause was undercorrected refractive error. Supply of correct glasses may be the most efficient and cheapest way to improve vision in the rural central India. Prevalence of BCVA-visual impairment and BCVA-blindness (WHO definition) was 7% and 0.5%, respectively, with a steep increase to 39% and 5%, respectively, in the subjects aged 80+ years. These figures are considerably higher than those reported for other populations. The most frequent cause for BCVA-low vision was nontreated cataract, accounting for ¾ of all visually disabled subjects. The by far leading causes for visual impairment in rural central India, that is, undercorrection of refractive error and cataract, are treatable. In contrast to studies in urbanized societies in East Asia and Western countries, age-related macular degeneration, diabetic retinopathy and degenerative myopia played minor roles as causes for visual impairment in the rural central India.