Abstract
Purpose.:
To provide data on prevalence and types of refractive error and the spectacle-wearing rate among adults in Nigeria and the degree to which the need for distance correction could be met by off-the-shelf spectacles.
Methods.:
Multistage, stratified, cluster random sampling with probability proportional to size was used to identify a nationally representative sample of 15,027 persons aged ≥40 years. Distance vision was measured using a reduced logMAR tumbling-E chart. All participants underwent autorefraction, and those with presenting acuity of <6/12 in one or both eyes had their corrected acuity measured and underwent detailed clinical examination to determine the cause.
Results.:
Included in the survey were 13,599 (89.9%) of the 15,122 persons aged ≥40 years who were enumerated. Uncorrected refractive error was responsible for 77.9% of mild visual impairment (<6/12–6/18), 57.1% of moderate visual impairment (<6/18–6/60), 11.3% of severe visual impairment (<6/60–3/60), and 1.4% of blindness (<3/60). The crude prevalence of myopia (≤0.5 D) and high myopia (≤5.0 D) were 16.2% and 2.1%, respectively. Spectacles could improve the vision of 1279 (9.4%) and 882 (6.5%) participants at the 6/12 and 6/18 level, respectively, but only 3.4% and 4.4% of these individuals wore spectacles to the examination site. Approximately 2,140,000 adults in Nigeria would benefit from spectacles that improved their vision from <6/12 to ≥6/12. More than a third of the need could be met by low-cost, off-the-shelf spectacles.
Conclusions.:
Uncorrected refractive errors are an important cause of visual impairment in Nigeria, and services must be dramatically improved to meet the need.
Refractive error (RE) is a priority for the global VISION2020: The Right to Sight initiative.
1 Worldwide, uncorrected RE is the main cause of moderate and severe visual impairment (VI) and the second leading cause of blindness, accounting for an estimated 153 million and 8 million affected persons, respectively, despite the fact that correction of RE with appropriate spectacles is one of the most cost-effective interventions in eye health.
2
RE is a significant cause of low vision in African countries, but available data are limited.
3 Nigeria is the most populous country in Africa.
4 Small, population-based studies in Nigeria have shown uncorrected RE, along with cataract and glaucoma, to be among the leading causes of blindness,
5 –7 VI, and low vision.
8 –12 Hospital- or industry-based Nigerian studies have shown RE to be the commonest ocular condition,
13 –15 a leading cause of VI,
16 –20 and associated with increased absenteeism and reduced productivity.
14 However, these studies were not nationally representative and cannot be extrapolated to the entire country due to their limited geographic scope, small sample sizes, and lack of validated methodology.
RE is a complex and multifactorial condition that varies in prevalence across populations with different genetics, demographics, ocular, and extrinsic factors, such as education.
21 The Nigerian National Blindness and VI Survey
22 –24 indicated that uncorrected RE accounts for 57.1% of moderate VI (visual acuity [VA], >6/18–6/60). Economic consequences are likely to be considerable, as uncorrected RE affects people in the working-age group. To our knowledge, no national survey of RE has been undertaken in Africa, whereas some have been undertaken in Asia.
25 –27 This article reports data on the prevalence and types of RE among adults aged 40 or more years in Nigeria.
Refractive Error.
Improvers.
Need for Spectacles.
The need for spectacles among the improvers could either have been “met” or “unmet.” So as to be comparable with other surveys, “met need” describes the number of participants who wore distance spectacles and had VA <6/12 in the better eye without correction, but who achieved 6/12 or better in the better eye with their present distance spectacles. “Unmet need” was defined as the number of participants who did not wear spectacles and who had VA <6/12 in the better eye without correction, but who could achieve ≥6/12 in the better eye with correction. Met need and unmet needs were also calculated with a cutoff of <6/18 in the better eye.
Some participants presented for VA measurement wearing spectacles, but with an incorrect prescription, defined as a presenting VA of <6/12 (or <6/18), which improved by one or more VA categories with best correction. If best correction improved their VA to ≥6/12 (or ≥6/18) they were defined as having unmet need at the relevant cutoff.
Spectacle Coverage.
The sample size necessary to meet the parameters of the study was calculated as 15,375 persons aged 40 years or above. Multistage, stratified, cluster random sampling, with probability proportional to size was used to identify a nationally representative sample. Stratification was by place of usual residence (urban/rural). A cluster size of 50 eligible adults was used in randomly selecting a total of 310 clusters across the country, of which 226 (72.9%) were rural and 84 (27.1%) were urban. Five clusters had to be abandoned due to civil unrest or refusal to participate.
Enumerated respondents were invited to attend the “temporary clinical station” set up in each cluster. Eligible respondents were registered with a unique identification number, after verifying their age and residency status, and recording information on sex, ethnic group, occupational status, religion, educational attainment, water supply, and household sanitation.
Participants had an initial anterior segment examination using a torch, including grading lens opacities (LO) against the red reflex, using the Mehra-Minassian (MM) system.
29 Information on the location and type of cataract surgery was elicited, as well as on the use of aphakic correction. Participants proceeded to a more detailed examination by a clinical ophthalmologist if they met certain criteria, including presenting VA <6/12 in one or both eyes. For these participants, the ophthalmologists determined the cause(s) of visual loss by using the principles outlined in the WHO Prevention of Blindness Performa (Version III).
30 All participants with VI were referred to the nearest eye facility.
A detailed examination was performed with a slit lamp microscope (SL 115 Classic Slit Lamp; Carl Zeiss Meditec AG, Jena, Germany), 81-D aspheric condensing lens (Volk Optical, Mentor, OH), Goldmann applanation tonometer, a two-mirror lens (Volk) with no flange for gonioscopy, and a digital camera (Visucam Lite Desktop Fundus Camera; Carl Zeiss Meditec AG).
A total of 15,122 eligible adults aged 40 years and older were enumerated, 13,599 of whom were examined (89.9% response rate, which was similar across all geopolitical zones). The age and sex of those enumerated and those examined were similar, but younger men (40–49 years) were underrepresented (Pearson R = −3.94; P < 0.001).
Of the 13,599 participants examined, eight had no VA data, and 890 (6.5%) had no information on RE because of ocular factors including corneal opacity, phthisis, and inability to undergo refraction because of blindness. A further 299 (2.2%) participants who had undergone cataract surgery were excluded, leaving 12,402 participants for analysis. Some analyses also excluded 1,715 participants with significant LO, leaving 10,687 participants for analysis.
Myopia.
Hypermetropia.
Astigmatism.
Only 1.2% (169) of phakic participants (1.2%) wore distance spectacles to the examination site. Another 38 claimed to own distance spectacles, but did not habitually wear them and were classified as nonwearers. Of the 2003 adults identified as having myopia, only 28 (1.4%) were wearing spectacles, and none of those had high myopia (n = 258). Of the 6823 participants identified with hypermetropia, 79 (1.3%) were wearing spectacles. The prevalence of spectacle wear increased with age (0.69% in 40–49-year-olds, 0.97% in 50–59-year-olds, 1.22% in 60–69-year-olds, and 70–79-year-olds, and 1.47% in >80-year-olds). The 299 participants who had undergone cataract surgery were also more likely to be wearing spectacles (14.4% versus 0.9% in phakic participants).
A total of 2248 (16.5%; 95% CI, 15.7–17.4) participants were improvers. We estimate that 3,890,000 (95% CI, 3,700,000–4,100,000) adults over 40 years of age would require optical correction to improve VA status by at least one vision category. Just over half (n = 80, 51.0%) of the spectacle wearers had an incorrect prescription at the 6/12 cutoff; this number was lower at the 6/18 cutoff (n = 65, 41.4%).
Over 90% (OR 96.6%, 95% CI, 95.5–97.7) of participants who needed spectacles did not own them, owned a pair but did not use them routinely, or used an incorrect prescription. Our results show that 9.1% (95% CI: 8.5–9.6) of all Nigerian adults over 40 years (2,140,000 individuals), have an unmet need for spectacles, which would improve their distance vision from <6/12 to ≥6/12.
This survey provides the first population-based data on the magnitude of RE in Nigeria. The two main findings are the relatively low prevalence of myopia and the extremely low spectacle coverage. Data from studies undertaken in Asia, Europe, the Americas, and Australia are shown in
Table 5 for comparison. The table contains only studies of adults conducted since 1985 that measured RE with reproducible methods.
Table 5. Summary of Results from Selected Population-Based Refractive Error Studies
Table 5. Summary of Results from Selected Population-Based Refractive Error Studies
| Study | Country | Sample Size | Age Group (y) | Myopia | High Myopia | Hypermetropia | Astigmatism |
| NBVIS | Pakistan 25 | 14,490 | ≥30 | <−0.5G; 36.5 | <−5.0D; 4.6 | >0.5D; 27.1 | >0.75D; 27.1 |
| NBLVS | Bangladesh 26 | 11,624 | ≥30 | ≤−0.5D; 22.1 | ≤−5.0D; 1.8 | >0.5D; 20.6 | >0.5D; 34.6 |
| APEDS | India 31 | 10,293 | ≥40 | <−0.5D; 34.6 | <−5.0D; 4.5 | >0.5D; 18.4 | >0.5D; 37.6 |
| MES | Myanmar 32 | 1,863 | ≥40 | <−1.0D; 42.7 | <−6.0D; 6.5 | >1.0D; 15.0 | >1.0D; 30.6 |
| SMES | Singapore 33 | 2,974 | 40–80 | <−0.5D; 38.7 | <−5.0D; 3.9 | >0.5D; 27.4 | <−0.5D; 33.3 |
| RES | Netherlands 34 | 6,543 | ≥55 | ≤1.0D; 17.6 | ≤−5.0D; 4.0 | ≥3.0D; 17.6 | Not studied |
| BES | United States 35 | 5,036 | ≥40 | <−0.5D; 19.4 (B) | <−6.0D; 0.0–1.4 (B) | >0.5D | >0.5D |
| | | | | | 41.0 (B) | 15.8–38.3 (B) |
| | | | 28.1 (W) | 1.3–2.5 (W) | 43.9 (W) | 24.4–48.9 (W) |
| BDES | United States 36 | 4,533 | 43–84 | <−0.5D; 26.2 | <−5.0D; 3.8 | >0.5D; 49 | Not studied |
| MVIP | Australia 37 | 4,744 | ≥40 | <−0.5D; 17.0 | <−5.0D; 2.1 | >0.5D; 37.0 | Not studied |
| BMES | Australia 38 | 3,654 | 49–97 | <−0.5D; 15.0 | ≤−4.0D; 3.0 | >0.5D; 57.0 | >1.0D; 37.0 |
| BdES | Barbados 39 | 4,709 | ≥40 | <−0.5D; 21.9 | Not studied | >0.5D; 46.9 | Not studied |
Using autorefraction results and a VA cutoff of <6/12, we estimate that there are 15,765,000 (95% CI, 15,530,000–16,001,000) adults with RE in Nigeria. Optical correction can potentially improve the vision of 4 million adults by one or more VA categories, and more than 2 million to normal levels of vision. Many of the remaining 2 million may require other interventions (i.e., cataract surgery).
The crude prevalence of myopia in Nigeria (16.2%) was lower than that in Asia (Pakistan, Bangladesh, India, Singapore, and Myanmar),
25,26,31 –33 but comparable to Western populations (Netherlands, United States, and Australia)
34,35,37 and similar to the adult black population in Barbados (21.9%) (
Table 5).
39 The prevalence of high myopia (2.1%) was similar to the rate in white Australians aged 40 years and older,
37 higher than findings in the Baltimore Eye Study (1.4% overall)
35 and in a study in Bangladesh (1.8%).
26 It is likely that these population differences are partly accounted for by a genetic mechanism for myopia, as demonstrated in multiple familial, familial aggregation, and twin studies that suggest the involvement of multiple genes rather than a single major gene effect.
40 More data on the genetic basis of RE are needed from African populations.
The prevalence of myopia showed a steady increase with age, similar to reports from Pakistan
25 and Bangladesh.
26 Much of the myopia in older age groups was the result of cataract, and participants with significant LO accounted for 41.5% of all myopes. Effect of age on prevalence of myopia has been observed in other studies.
25,26,31 –39,41 –46 However, this age-related trend was evident, even after participants with visually significant cataract were excluded. In contrast, most studies of Western populations show a decrease in prevalence of myopia with age, followed by an increase at older ages (J-shaped relationship).
39 There may be a U-shaped trend in Nigeria, with the decrease and subsequent increase in prevalence coming at earlier ages than Western populations. It would be instructive for future research in Nigeria to include refractive error in younger age groups as well, to elucidate the longitudinal changes and cohort effects that have been observed in other studies.
42 –45 The men in our survey had a slightly higher prevalence of myopia than did the women, as has been reported in other studies.
26,37,47 –49
The prevalence of hypermetropia (50.7%) was considerably higher than that in most Asian studies, being closer to that in populations of predominantly European or African descent. The relationship between hypermetropia and age showed a pattern similar that in to studies from Asia and Barbados,
25 –27,31,33,39,49 with a rise to maximum levels in the 50- to 59-year-old group, followed by a decline in later years. Hypermetropia prevalence was significantly higher in women (55.6% versus 44.7%), which has been observed elsewhere.
26,31,37,39,50
Astigmatism was prevalent in 63.0% (58.7%, after excluding those with visually significant cataract). Many studies of RE have not examined the prevalence of astigmatism. The prevalence in our survey was similar to findings from South India (60.4% and 59.1% in rural and urban participants in Tamilnadu, respectively).
49
Participants who were illiterate were more likely to be myopic or hypermetropic and to have astigmatism. Participants with manual occupations and those living in rural areas were also more likely to be myopic and have astigmatism. Some of these findings contrast with other population-based studies, which have shown associations between myopia, higher education levels, professional occupations and residence in urban areas
25,26,32,33,40,43,46,48,51 (in support of the use–abuse theory of myopia). There are, however, exceptions to these trends: The Baltimore Eye Study showed that the association between years of education and myopia was stronger for white persons than for black persons,
35 the Chennai Study (India) showed a higher prevalence of myopia in participants living in rural areas,
49 and the Bangladesh study showed a significantly lower risk of myopia with literacy, urban living, and nonmanual occupation.
26 This variability in findings suggests that the risk factors for myopia are context specific, and that time spent outdoors may be the more important environmental variable in myopia in some populations.
52
The leptokurtosis and negative skewness of the distribution of spherical equivalent RE in this population was similar to that in other studies.
25,26,31 –33,40,50,53 After participants with significant LO were excluded, the mean spherical equivalent was +0.67 D, the same as Australians aged ≥49 years,
54 but different from adults aged ≥40 years in Bangladesh (−0.19 D),
26 Pakistan (−0.4 D),
25 and Myanmar (−1.3 D).
32
Spectacle coverage rates were significantly lower than reported among similar age groups in Bangladesh (3.0%),
27 Pakistan (6.2%),
25 and India (17.4%),
31 although it should be appreciated that the definition of “unmet need” for spectacles does not necessarily equate with demand for correction. None of the participants with high myopia were wearing spectacles. Incorrect prescriptions were common among the few wearing spectacles, with just over half improving by ≥1 VA category with best correction. This suggests a need to improve both quality and affordability of optical and refractive services in Nigeria. Over one third of the need for distance correction among individuals who were phakic in one or both eyes could be met by off-the-shelf spectacles.
Limitations of this study include possible overestimation of myopia in younger participants, as autorefraction was not performed after cycloplegia. The analysis used refractive data from the right eye, which is in keeping with several other studies,
26,32,33 but differs from some studies that included the worse eye in their analyses.
25,45,49 Younger males were underrepresented, as they were more likely to be at work at the time of examination, which may have led to a slight overestimation of refractive error. The MM lens-grading system was used to provide some data on lens opacities in all participants, regardless of their visual acuity. Individuals undergoing full ophthalmic examination had their lenses graded using LOCS III.
55,56 finally, presbyopia, and anisometropia were not addressed.
This is the first population-based, national RE survey in Africa, to the authors' knowledge. The distribution of RE in Nigeria appears closer to that of white and black populations in Europe and America, and differs from Asian populations. Findings indicate a low prevalence of myopia in Nigeria, exceedingly low spectacle coverage, a large unmet need for spectacles, and a need to improve the quality, access and affordability of optical and refractive services—a VISION2020 priority.
Supported by Sightsavers International, Christian Blind Mission (CBM), and Vellux Stiftung.
Disclosure:
C. Ezelum, None;
H. Razavi, None;
S. Sivasubramaniam, None;
C.E. Gilbert, None;
G.V.S. Murthy, None;
G. Entekume, None;
T. Abubakar, None
The authors thank Oye Quaye for managing the finances for the study; Auwal Shehu and Dania Charles for data entry; and the teams of ophthalmic nurses, enumerators, and interviewers in the six geopolitical zones who assisted in data collection.
Additional Members of the Nigeria National Blindness and Visual Impairment Study Group
Abdull Mahdi, Adenike Abiose, Olufunmilayo Bankole, Fatima Kyari, Hannah Faal, Abudallahi Imam, Pak Sang Lee, and Mansur Rabiu.