Background/Aim To study the association between exposure to biomass smoke from cooking fuels andi cataract, visual acuity and ocular symptoms in women.
Methods We conducted a community-based cross-sectional study among women (≥20 years and without a previous diagnosis of cataract, ocular trauma or diabetes or those taking steroids) from hilly and plain regions of Nepal. Eligible participants received an interview and a comprehensive eye assessment (cataract development, visual acuity test and ocular symptoms). Participants’ data on demographics, cooking fuel type and duration of use, and cooking habits were collected. We addressed potential confounders using the propensity score and other risk factors for ocular diseases through regression analysis.
Results Of 784 participants, 30.6% used clean fuel (liquefied petroleum gas, methane, electricity) as their primary current fuel, and the remaining 69.4% used biomass fuels. Thirty-nine per cent of the total participants had cataracts—about twofold higher in those who currently used biomass fuel compared with those who used clean fuel (OR=2.27; 95% CI 1.09 to 4.77) and over threefold higher in those who always used biomass. Similarly, the nuclear cataract was twofold higher in the current biomass user group compared with the clean fuel user group (OR=2.53; 95% CI 1.18–5.42) and over threefold higher among those who always used biomass. A higher proportion of women using biomass had impaired vision, reported more ocular symptoms compared with those using clean fuel. Severe impaired vision and blindness were only present in biomass fuel users. However, the differences were only statistically significant for symptoms such as redness, burning sensation, a complaint of pain in the eye and tear in the eyes.
Conclusions Cataract was more prevalent in women using biomass for cooking compared with those using clean fuel.
- Eye (Globe)
Statistics from Altmetric.com
Approximately 41% of the world’s population uses solid fuels (biomass and coal) as their primary source of energy for household activities including cooking and heating, mostly in low- and middle-income countries (LMICs).1 2 Households using biomass fuels for cooking in LMICs are often poorly ventilated, trapping airborne particulate matter in high levels of household air pollution (HAP).3 As women perform the majority of the cooking indoors, often for long hours, they may have disproportionately high exposure to HAP and a possibly higher risk of adverse health outcomes.1 4–6
Cigarette and biomass fuel smoke contain similar constituents, and this supports the argument that inhaled substances can directly or indirectly affect the lens tissue. Animal studies have shown that the condensate from firewood smoke permeates the lens capsule, imparts colour and opacifies the lens when continuously exposed over some time.3 Previous studies have reported that exposure to smoke from burning biomass fuels depletes antioxidants such as plasma ascorbate, carotenoids and glutathione, and enhances the formation of superoxide free radicals inducing oxidative stress.3 Additionally, exposure to naphthalene and formaldehyde released during the combustion of many biomass fuels are known risk factors for cataract development.3
Previous studies have reported that HAP is associated with ocular diseases such as ocular irritation, ocular infection including trachoma, and cataract.7 Cataract, an opacity of the lens in the eye, is associated with around 48% of the world’s blindness.8 While cataract is common with increasing age, it is also associated with exposure to cigarette smoke and ultraviolet radiation, trauma, diabetes mellitus and genetic factors.9–11 12 The true impact of long-term exposure to HAP upon cataract may be masked by age as it is highly collinear with cumulative HAP exposure. Recent publications reported that risk factors including exposure to HAP from the solid fuel burning play a substantial role in the development of cataract.5 13 In LMICs (when compared with high-income countries (HICs)), cataracts occur at a younger age and are more common in women than men.14 15 Poverty is associated with risk factors such as poor access to healthcare (which, eg, would lead to poor control of diabetes, a risk factor for cataract), long hours of outdoor work without eye protection (ie, high ultraviolet radiation exposure), malnutrition and eye infections. Thus, there is a significant risk of confounding within observational studies evaluating the relationship between the use of biomass fuel and cataract development.
As it is not ethical or feasible to randomise fuel type and, at the same time, study a chronic outcome such as cataract, we conducted a propensity score analysis to address these potential sources of bias. Propensity score analysis allows researchers to design and analyse an observational study so that it mimics some of the characteristics of a randomised controlled trial by adjusting for treatment bias in observational cohorts16 The main aim of this study was to understand the effect of HAP on women ocular morbidity in Nepal and to know the differencing impacts of each cooking fuels. In this community-based study, we report on the prevalence of visual acuity and association between ocular symptoms and cataract with exposure to HAP, as well as ocular morbidity among those who have switched to clean fuel from biomass and vice versa.
We conducted a community-based cross-sectional study among women from two village development committees (VDCs): one from a hilly region and the other from a plain region (Madhesh) of Nepal. The VDCs were purposively selected to ensure a diversity of fuel use, with one (hilly area) predominantly using biomass and the other (plain area) using, in addition to biomass fuel, biogas fuels. Within the VDCs, one woman from each household (aged ≥20 years) who cooked at least four to five times a week was invited to take part in the study. Only women were included as in Nepal; women are mostly responsible for daily cooking. Those who had the previous diagnosis of cataract were excluded, as they were likely to have the cataract removed. We also excluded women with ocular trauma or diabetes, and those who were taking steroids (either eye-drops or tablets), which are known risk factors for cataract. For safety reasons, we did not include pregnant women in the study.
Consented participants received an interview and a comprehensive eye assessment at temporary clinics in their villages. The interview was conducted using an interview guide addressing socio-demographics, housing and kitchen characteristics (layout and space), cooking fuel type used, cooking type and duration, lifestyle factors such as smoking habits of study participants, environmental tobacco smoke (ETS) exposure, medical history and ocular symptoms. The interview guide was translated into Nepali and back-translated into English by an independent translator to ensure accuracy. After the interview, each participant underwent visual acuity assessment by an optometrist/ophthalmic technician and then eye examination by an ophthalmologist.
Assessment of HAP exposure
Participants were asked if they have lived in more than one residence. For the current and previous homes, the participants were asked to report, separately, the primary and secondary types of fuel used for cooking and the duration of use. We combined electricity, liquefied petroleum gas and biogas into ‘clean fuel’, and charcoal, wood, dung and mixed biomass as ‘solid fuel’. We also captured those participants who had consistently used clean or solid fuels or switched between the two categories. We acquired information on the frequency of cooking (number of days of food cooking per week) and the age when they started cooking, where the participants cooked (open-air or indoors), and whether the kitchen was ventilated (presence of window or extraction fan in the kitchen).
Ocular outcomes assessment
During the interview, we asked about the presence of a variety of ocular symptoms, that is, blurring of vision, redness, burning sensation, grittiness, photophobia, eye pain and tears. The study participants underwent a visual acuity test conducted by an ophthalmic technician from Tilganga Tertiary Eye Hospital and eye examination by an ophthalmologist using a portable slit lamp and direct ophthalmoscopy. Those presenting with visual acuity of 6/6–6/12 were classified as having normal vision, those worse than 6/12 classified as mild vision impairment, worse than 6/18 as moderate visual impairment, worse than 6/60 as severe visual impairment and worse than 3/60 as blind as per the WHO International Classification of Diseases-11 criteria.17 An ophthalmologist used Lens Opacities Classification System to grade cataract among participants those who had it.
We used interview questionnaires to obtain information on socio-demographics of the women, that is, family size and income, education and occupation, ownership of assets, and housing. We asked about other environmental airborne exposure including detailed histories of tobacco smoking (categorised as never, former or current smokers), age started smoking, exposure to ETS (never, <1 hour/day, ≥1 hour/day) and use of mosquito coils (none, <5 months/year, ≥5 months/year). Other potential confounders such as exposure to sunlight (none, 1–4 hours/day, >4 hours/day working under the sun) and dust (no, yes) in their primary occupation, and regular consumption of fruit (no, yes) and dairy products (no, yes) were assessed.
Propensity score model
We created a propensity score model to address factors that might have created systemic variation between women (study participants) who currently used relatively clean fuel for cooking as compared with those who cooked with solid fuels. Considering women who cooked food using relatively clean fuel as a reference group, women using solid fuel were compared across many variables (literacy, annual household income, ownership of land, animals, and motorised vehicles, housing materials, family size, number of rooms in their residence, whether they had engaged in manual work and whether they had switched fuel) that were potentially associated with solid fuel use but might have been confounded by the socioeconomic status. Univariate logistic regression was performed for each covariate that might have been possibly associated with socio-demographic characteristics. Each covariate that was found to be significantly associated with biomass use was selected for inclusion in the propensity score model. A propensity score was calculated for each participant based on the final chosen covariate series in the model. We did not include variables such as kitchen ventilation type, age and smoking in the model to calculate the propensity score. Still, these variables were adjusted for when carrying out the multivariable logistic regression.
Demographic characteristics were compared between clean fuel and solid fuel users using t-tests and ᵡ² tests where appropriate.
We carried out logistic regression to estimate the univariate effect size and associated 95% CIs for eye symptoms and the presence of cataract. The propensity scores derived earlier were categorised into quintiles. Due to very few participants in our study, we chose not to match our participants by propensity score as this would make the sample size too small to derive any inferences.
To address the collinearity between poverty and choice of fuel, and the strong association between the former and the presence of adverse health outcomes, for each participant, we calculated the propensity score (as described earlier) for solid fuel use, which is the predicted probability that the participant used solid fuel for cooking. The propensity score quintiles were then included in a logistic regression model, in which the relationship between the presence of cataract and solid fuel use was assessed, controlling for other potential confounders (smoking, age, education and ventilation, conjunctivitis, visual impairment and duration of exposure in hours/day). We report the OR and the associated 95% CIs. The analysis of the presence of cataracts was repeated for those with nuclear cataracts only as a sensitivity analysis, as 85% had a nuclear cataract. All statistical analyses were performed using Stata (version 16.0, College Station, TX, USA).
Of 784 participants who met the inclusion criteria, 30.6% (n=240) used clean fuel (liquefied petroleum gas (LPG) or methane or electricity) as their main current fuel (fuel type used at the time of questionnaire). In contrast, the remaining 69.4% (n=544) used solid fuels. The overall participants cooked for an average (±SD) of 2.41±2.35 hours/day. Out of the total participants, only 9.0% (n=71) had always used clean fuel, 58.8% (n=461) had always used solid fuel, 21.6% (n=169) had switched from solid fuel to clean fuel and 10.6% (n=83) had switched to solid from clean fuel (table 1).
Factors associated with clean versus solid fuel use
Solid fuel is overwhelmingly the most common cooking fuel used by Nepalese women and is more frequently used compared to clean fuel in both hilly (75.6% vs 24.4%) and plain terrains (60.2% vs 39.8%) by women of all age groups (table 1).
A larger proportion of women using biomass fuel for cooking were smokers (p<0.001), and among the smokers, biomass fuel users had higher pack-years compared with those using clean fuel (p<0.001). Exposure to ETS was also higher in women using biomass fuel compared with those using clean fuel (p=0.007) (table 2).
The occupation was correlated to cooking fuel type, with farmers, labourers and homemakers being more likely to use biomass fuel (94.9%) while government employees, those involved in business/commerce and teachers or students relatively being more likely to use clean fuel (13.9%). This difference was statistically significant and perhaps tied to occupation as women earning less than 100 000 Nepalese rupees (NPR) (~US$900) per year were more likely to use biomass fuel. In comparison, those with an income higher than 100 000 NPR were more likely to use clean fuel (p<0.001).
Reasons for fuel use varied, from cooking to lighting and heating. Those participants who used biomass fuel to cook also used biomass fuel more frequently as a heating source (86.4% vs 68.8%, p<0.001), unlike the clean fuel user group who used mainly electricity. The proportion of women in the biomass fuel user group that used electricity (93.4% vs 99.6%) for lighting purposes was lower, and kerosene (6.6% vs 0.4%) was higher compared with those using clean fuel.
Women using biomass fuel were more likely to have kitchens that were open/semi-enclosed/in a separate building and kitchens that are not separate from the living area. In contrast, the majority of the women in the clean fuel user group had a separate kitchen within the house (table 1). Women using biomass fuel were more likely to drink alcohol (20.8% vs 6.7%) and less likely to have seasonal fruit (86% vs 97.1%) compared with those using clean fuel. A lower proportion of women in the biomass user group owned land compared with the clean fuel user group (87.8% vs 86.6%).
Comparison of eye symptoms by primary cooking fuel
Women using solid fuel as their current primary fuel reported more eye-related symptoms (redness, burning sensation, grittiness, photophobia, pain and tearing) as compared with those who used clean fuel (table 2). However, the differences were statistically significant only for symptoms such as burning sensation (p=0.007), ocular pain (p<0.001) and tearing of the eyes (p=0.011). A higher prevalence of eye symptoms was reported in women who switched from biomass to clean fuel and women who switched from clean to biomass fuel when compared with those who had only used clean fuel. Regardless of primary fuel type, participants reported eye symptoms were more common experiences while cooking, and the symptoms worsened while cooking (table 3).
The odds of reporting eye symptoms (redness, burning sensation, grittiness, photophobia, pain and tears) by women using biomass fuel were higher compared with the clean fuel user group during unadjusted analysis (significant for grittiness and tear only); however, after adjusting for covariates (both the models with different adjustment), the risk was statistically significant in those with redness, burning sensation, pain and tears in the eye (table 4).
Visual acuity by primary cooking fuel use
Of 784 participants, 82.6% (n=648) had normal vision, 5.2% (n=41) mild, 9.4% (n=74) moderate, 1.1% (n=9) severe visual impairment and 1.5% (n=12) were blind (table 5). Similarly, 82.3% (n=645) had normal vision, 7.0% (n=55) mild, 7.1% (n=56) moderate, 1.3% (n=10) severe visual impairment and 2.3% (n=18) were blind. A higher proportion of participants currently exposed to biomass smoke (6.2%, 10.5%; 1.5% and 1.6%) compared to using relatively cleaner fuel (2.9%, 7.1%, 0.4%, 1.2%) had mild, moderate, severe vision impairment and blindness respectively in their right eye. Similarly, the proportion of participants with mild, moderate, severe vision impairment and blindness was higher among those currently exposed to biomass smoke (8.1%, 7.5%; 1.8% and 2.2%) compared to using relatively cleaner fuel (4.6%, 6.2%, 0% and 2.5%) respectively (table 5). On eye examination, participants who always used biomass fuel were found to have 1.9% blindness in the right and 2.6% in the left eye compared to no blindness in those who reported always using cleaner fuel.
Solid fuel use and cataract
Overall, 39% of the total participants (n=307/784) had cataracts, and these were more common in those currently using biomass fuel compared with clean fuel users (42% vs 32%, p=0.01) (figure 1 and online supplemental table S1). Women who had always used clean fuel had the lowest prevalence of cataracts, and women who switched to clean fuel from biomass fuel had the second-lowest prevalence of cataracts. There was over twofold (OR=2.27; 95% CI 1.09 to 4.77) increased risk of cataract among those who were currently using biomass fuel for cooking.
Using the unadjusted model, the OR of having cataracts in those who always used biomass fuel, switched to biomass fuel from clean fuel and switched to clean fuel from biomass compared with those who always used clean fuels were 2.74 (95% CI 1.50 to 4.98), 2.59 (1.26 to 5.31) and 2.22 (1.56 to 4.25) respectively. After adjustment, there was still an increased odds of having a cataract in those who always used biomass fuel only compared with those who used clean fuel (OR=3.58, 95% CI 1.13 to 11.37). Directionally, a trend was observed from ever clean fuel users, switched to clean fuel from biomass, switched to biomass fuel from clean fuel and always biomass fuel user (p=0.006). The risk of cataract among those who switched to clean fuel from previous biomass fuel use was lower compared with those who always used biomass fuel, although statistically not significant in the adjusted model. Similarly, the risk of cataract was higher among those who switched to biomass fuel from clean fuel (reference always clean fuel user). However, the results were statistically not significant in the adjusted model.
Of 307 cataracts, the majority (85%) had nuclear cataracts. Upon sensitivity analysis, limiting the data to the presence of nuclear cataracts (reference group: no cataract) only, participants with current use of biomass fuel had over twofold (OR 2.53; 95% CI 1.18 to 5.42) increased risk of nuclear cataract (figure 1 and online supplemental table S2). There was a dose–response relationship between the exposure to biomass fuel emissions and nuclear cataracts. Those who always used clean fuel had the lowest followed by switched to clean fuel from biomass fuel use, followed by switching to biomass fuel from clean fuel to those who always used biomass fuel (p=0.016). The risk was statistically significant only for those who always used biomass fuel (OR=3.30; 95% CI 1.04 to 10.48).
This cross-sectional study from two different regions of Nepal (hilly and plain) reports on the association between cooking fuel airborne exposure and eye disorder (visual acuity, ocular symptoms and cataract types) among women who switched from lifelong use of biomass to relatively cleaner fuel, and those who have switched from lifelong use of clean fuel to biomass fuel for cooking. The findings suggest that those who switched to clean fuel from biomass fuel had a comparatively lower risk of experiencing visual acuity problems, worsening of eye symptoms during cooking and cataract compared with those who continue using biomass. Similarly, those who switched to biomass fuel from previous use of clean fuel experienced a higher risk of visual acuity problems, worsening of various eye symptoms during cooking and cataract compared with those who continued to use cleaner fuel.
Two previous studies reported on the association between exposure to biomass stove and tear in eyes. A study from rural Pakistan found women who used improved cookstoves compared with those who used traditional stoves reported 46% lower risk of teary eyes.18 Another randomised controlled trial from Guatemala reported women using open fires reported around 5.5 times increased risk of self-reported eye symptoms compared with those with improved cookstoves (Plancha).19 These findings are in agreement with this study, where women who used biomass fuel as their primary cooking fuel reported a greater presence of each of the assessed eye symptoms. None of the previous studies on HAP and eye problems contained information about the presence of specific eye symptoms based on the primary fuel source.
Only two other studies have reported an association between exposure to biomass fuel and visual acuity.5 20 A study in children aged 7–8 years from Barcelona reported association both in cross-sectional and longitudinal analyses between different components of ambient air pollution (NOx and particulate matter) with 13– 32% higher use of spectacles.21 In our study, we observed that those currently exposed to biomass had the highest proportion of visual impairment ranging from mild to moderate among those who used fuel biomass (right and left eye: 18.2% and 17.4%, respectively) compared with those who currently used clean fuel (right and left eye: 10.4% and 10.8%, respectively). In this study, women who always used biomass fuel for cooking had severe vision impairment and blindness in both the right and left eye, but none of the participants who always used clean fuel had severe vision impairment. Although the study cannot confirm the causality given the design, this is a plausible finding worth investigating in future studies.
A report suggesting a possible association between HAP and cataract was published by Mohan et al as early as in 1989 when they reported the use of biomass fuels instead of gas as a cooking fuel resulted in an increased risk of cataracts.13 Since then, more research has been undertaken in South Asia in the hope of providing more definitive evidence, including four case–control studies and four cross-sectional studies (online supplemental table S3).4 5 12 16 22–24 Pokhrel et al performed a hospital-based case–control study on the Nepal–India border of 206 women with confirmed cataracts and compared them to controls matched by age.12 They found, compared with using a clean-burning-fuel stove, the OR for using a flued solid fuel stove was 1.23, while the OR for using an unflued solid fuel stove was 1.9. This study was hospital-based and is highly unlikely to be representative of the community-based. Haq et al, who performed a cross-sectional study of 645 subjects in Aligarh over 1 year, found the prevalence of cataract was 21.7%, with a bilateral cataract prevalence of 16.9% of the population, and cataract was significantly associated with age, education and fuel use.16 Other studies have also suggested kerosene as a possible cause of cortical cataract, but more research is needed to confirm this association.5 While a consistent increased cataract risk with biomass fuel use was found in these studies, causality could not be ascertained, as it was difficult to remove selection bias and to confound. Furthermore, there is a lack of research into other aspects of ocular morbidity in these studies, including ocular symptomatology and visual acuity. However, findings from many studies seem to confirm the plausibility of a causal relationship between exposure to HAP and cataract.
Many previous studies have reported a positive association between exposure to biomass and cataract but with varying effect sizes.2–5 13 22 23 25–28 Although many of these studies had clinical confirmation of cataracts, only one previous study from Nepal differentiated types of cataracts.5 In our study, 42.3% of women who use biomass fuel exclusively developed cataracts compared to 21.1% of women who use clean fuel exclusively. Among women who switched from biomass fuel to clean fuel (perhaps due to an increase in income), 41% had cataracts, probably due to chronic exposure of biomass smoke. Of the women that switched from clean fuel to biomass fuel, 37.3% had of cataract. A study from India reported that 27.2% of participants who used biomass fuels exclusively had a cataract and 10.3% in the clean fuels exclusively experienced cataract.27
Not all the results from previous studies are directly comparable to the present study. Uniquely, this study examines the presence of cataracts in women who exclusively used clean or biomass fuel as well as those who switched between biomass fuel and clean fuel. These factors may contribute to the differences in results.
The aforementioned cross-sectional studies consistently reported a higher risk of cataract in the biomass fuel user; however, the nature of the studies does not remove the selection bias and confounding. A study by Saha et al reported that 27.1% of participants who used solid fuels exclusively developed cataracts, and 10.3% of participants who used clean fuels exclusively experienced cataracts.4 The number of subjects in our study who had cataract was higher; however, we examined a different population to Saha et al. We studied women aged 20 years or over, as they typically are more involved in-home cooking for long periods. Saha et al included both males and female patients, and 18% of their population were aged <20 years. Thus, while the prevalence of cataracts in our study is higher, it is not surprising given previous studies have shown cataracts are more common in female populations in LMICs compared with that of HICs.8 15 Their reasoning for these differences supports our study methods, as they felt men were less likely to develop eye irritation as they were not actively involved in cooking.4
The principal recommendation of published studies is to switch from biomass fuel to liquid petroleum gas, biogas or solar energy as these sources. It should not be surprising that biomass fuel users tend to be of lower socioeconomic status compared with gas or electricity users. Poverty itself is associated with (or even as the root of) risk factors such as poor access to healthcare (and hence poor control of diabetes), long hours of outdoor work without eye protection (and thus high ultraviolet radiation exposure and high risk of eye injury), malnutrition and eye infections. On the other hand, while ageing is seen as the most important risk factor for senile cataract, it is also possible that the true effect of long-term exposure to HAP, which is highly collinear with age, might have been masked by the latter and could, therefore, be underestimated.
This study, while larger than many previous studies addressing the topic, does have some limitations, including having an unequal number of participants in each fuel type group. The self-reported exposure data collected from the interview format may have resulted in reporting bias, despite efforts to reduce the residual bias by adjusting for important confounders. The strengths of this study include its large sample size, and the mitigation measurement by using a set of predetermined eye symptoms to ensure consistency in reporting. We also adjusted for propensity score to reduce the bias mainly due to socioeconomic status and other potential confounders such as smoking. The study is community-based, and a mobile clinic was set up with ophthalmologists and ophthalmic technicians, reducing the detection bias, but due to the nature of design, causality could not be ascertaine. This would need a prospective cohort study but more appropriately randomised controlled trials to understand the causality of the relationship between exposure to biomass smoke on lenticular changes and/or formation of cataract. However, the other important strengths of the study are detailed eye examination with visual acuity, self-reported symptoms examined by ophthalmologist and detection of types of cataract, and have excluded participants with ocular trauma or diabetes, and those who were taking steroids (either eye-drops or tablets), which are known risk factors for cataract.
In conclusion, this study provides further evidence confirming the association of HAP and ocular morbidity through the burning of biomass fuel as a primary cooking source. The use of the propensity score model addresses systemic variation and aids in reducing the impact of socioeconomic confounders. Further, logistic regression based on this model aided in delineating poverty as a co-variable. Poverty appears to be a key driver in fuel choice and therefore restricts self-directed resolution due to economic restrictions of the participant communities. Symptoms such as a burning sensation and eye pain are noted at roughly equal levels, suggesting that these factors pertain to the process of cooking, more so than fuel. In sharp contrast, cataract was found to be prevalent at approximately twice the rate in participants who currently used solid fuel over those who used clean fuel, and the risk was over threefold higher among those who always used solid fuel. This study adds credibility to previous studies, as the propensity score method used produced somewhat similar results within a range of likely outcomes and factors. Our study suggests that switching to cleaner fuel may help in reducing the burden of ocular health problems in LMICs. These results may be useful to inform fuel choices for cooking or in developing national policy when it comes to fuel options.
MP, MKS, AM, JA and OPK contributed equally.
Twitter Nagendra Chaudhary @nagendraaiims.
Acknowledgements We would like to thank all the research assistants from the Tilganga Institute of Ophthalmology who helped in the data collection and all the participants for taking part in this study. The authors attest that they meet the current ICMJE criteria for authorship.
Contributors OPK and JA conceived the study, participated in its design, data acquisition and/or research execution, data analysis and/or interpretation, and manuscript preparation and/or revision. MKS and AM were involved in data acquisition and/or research execution, data interpretation and manuscript revision. MP was involved in data analysis and/or interpretation and wrote the first draft of the manuscript. SS, NC and RC were involved in the data interpretation and manuscript revision.
Funding This was supported by the University of Birmingham, UK, as internal seed funding (no reference number). The funder had no role in data collection, analysis and interpretation of the results.
Competing interests None declared.
Patient consent for publication Not required.
Ethics approval The research followed the tenets of the Declaration of Helsinki and was approved by the institutional ethical committee of the Tilganga Institute of Ophthalmology (reference number: 1/2011–2012). Informed consent was obtained from the subjects and their guardians after an explanation of the nature and possible consequences of the study. For those who were unable to read or write, the purpose of the study and consent form was read out, and if they agreed to participate, thumb impressions were taken instead of their signatures.
Provenance and peer review Not commissioned, externally peer-reviewed.
Data availability statement Data are available upon reasonable request. The data used to support the findings of this study are available from the corresponding author upon request.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.