Article Text

Download PDFPDF

Glaucoma prevalence in Indigenous Australians
  1. Brian E Chua1,
  2. Jing Xie1,2,
  3. Anna-Lena Arnold1,2,
  4. Ignatios Koukouras1,
  5. Jill E Keeffe1,2,
  6. Hugh R Taylor2,3
  1. 1Centre for Eye Research Australia, Department of Ophthalmology, University of Melbourne, Australia and Royal Victorian Eye & Ear Hospital, Melbourne, Australia
  2. 2Vision CRC, Sydney, Australia
  3. 3Melbourne School of Population Health, University of Melbourne, Melbourne, Australia
  1. Correspondence to Professor Hugh Taylor, Melbourne School of Population Health, University of Melbourne, 207 Bouverie Street, Carlton, Victoria 3053, Australia; h.taylor{at}unimelb.edu.au

Abstract

Aim Define the prevalence of glaucoma within the Indigenous Australian population.

Methods Aboriginal and Torres Strait Islander adults aged ≥40 years were recruited from 30 randomly selected sample areas or communities. Participants were regarded to have glaucoma if they had a cup:disc ratio (CDR) of >0.8, or missed ≥2 points on the Frequency Doubling Test and a CDR >0.7 in at least one eye.

Results 1189 eligible adult participants were examined, representing 72% of the target population. After excluding cases with missing or ungradable information, the authors found a 2.2% (95% CI 1.6% to 3.6%) overall prevalence of glaucoma. In univariate analyses, the odds of glaucoma increased with age (χ2 trend=4.38, p<0.001), male gender (OR 2.43, 95% CI 1.10 to 5.41), less than secondary education (OR 4.74, 95% CI 1.96 to 11.45) and self-reported history of glaucoma (OR 20.8; 95% CI 6.23 to 69.51). After a multivariate analysis, none of these attributes other than history of glaucoma remained significant. No cases of low vision (presenting visual acuity (VA) <6/12 to ≤6/60) or blindness (presenting VA<6/60) were solely attributable to glaucoma. The mean optic disc diameter was 1.93 mm (SD 0.19) for left and right eyes, while the mean CDR for right eyes was 0.44 (SD 0.15) and for left eyes 0.43 (SD 0.16).

Conclusions This population-based study examined the prevalence of glaucoma within the Indigenous Australian population, and although an infrequent cause of vision loss, definable rates of disease were seen. The results may suggest a potential introduction of Caucasian glaucoma-associated genes into this community, differently used diagnostic criteria or sampling bias compared with previous surveys.

  • Glaucoma
  • prevalence
  • Australia
  • Aboriginal
  • angle
  • public health
  • epidemiology

Statistics from Altmetric.com

Introduction

Glaucoma is the second leading cause of vision loss worldwide.1 2 Previous studies of the Australian Aboriginal and Torres Strait Islander population have reported very low to negligible rates of glaucoma.3–5 To date, there remains relatively little information on the status of eye health in the Australian Indigenous population. By comparison, the prevalence of primary open-angle glaucoma of Caucasian Australians has been reported to range between 1.7% and 3.0%, based on two large Australian population studies, the Melbourne Visual Impairment Project6 7 and the Blue Mountains Eye Study,8 respectively. We aim to describe data from the first indigenous population sampled, the National Indigenous Eye Health Survey.

Taylor et al4 were the first to report the prevalence of pseudoexfoliation syndrome in the Australian Aboriginal population. After examining 80% of the estimated Aboriginal population in remote South Australian communities, they found a pseudoexfoliation syndrome prevalence of 1.3% (37/ 2773). Only 0.11% (3/ 2773) of this population was diagnosed as having pseudoexfoliative glaucoma, which they defined as the presence of both raised IOP and abnormal optic disc in the setting of pseudoexfoliation. No cases of primary open-angle glaucoma were identified in that survey.

Increasingly, there is a more reflective understanding of the worldwide prevalence of glaucoma in persons aged 40 years and older arising from recently published population surveys. Indigenous peoples of Mongolia are reported to have an overall glaucoma prevalence at 2.2%, with more angle closure glaucoma compared with open angle glaucoma.9 The principally Chinese-Singaporean population from the Tanjong Pagar Study revealed an overall glaucoma prevalence rate of 3.2%.10 The Aravind Eye Study found a prevalence of 2.6% in India,11 while a Bangeladeshi population prevalence was reported at 2.1%.12 Meanwhile, data from the African continent report prevalence rates ranging from 4% to even 8.5%.13–15 The Barbados Eye Study stratified glaucoma prevalence according to self-reported race, and found a rate of 7% in black compared with 0.8% in white participants.16

This report aims to present the prevalence of glaucoma in Indigenous Australians using an epidemiological definition of glaucoma17 from data collected within the auspices of a National Indigenous Eye Health population-based survey. We also report for the first time the distribution of optic disc diameter and cup:disc ratios in this population.

Methods

Sample selection

The sampling method has been described in detail elsewhere.18 19 Briefly, Indigenous Areas were grouped into six strata based on the Australian Accessibility and Remoteness Index (ARIA): Major City, Inner Regional, Outer Regional, Remote, Very Remote Coastal and Very Remote Inland. Within each ARIA, five sample areas or communities were randomly selected proportional to size, using June 2006 census data, to identify geographic areas containing approximately 300 Indigenous people. This gave a total of 30 sample areas. The sample size of 1507 adults aged 40 years and older was selected based on the ability to detect changes in vision impairment.18 This research was conducted in accordance with the tenets of the Declaration of Helsinki as revised in 2000. Written, informed consent was obtained for all participants prior to examination.19 20

Recruitment

An informal census was undertaken in each site using community data and local informants to establish the size of the eligible population of Indigenous children aged 5–15 years inclusive and Indigenous adults aged 40 and above. As children did not undertake fundus photography or frequency doubling test (FDT), they were excluded from this analysis on glaucoma prevalence.

Examination procedures

A standardised questionnaire was used to collect data on demographics, general health, eye health and health-service utilisation.19 It was adapted from that used in the assessment of The Vision Initiative.21 22 A standardised eye examination included presenting distance and near visual acuity using an E chart.22 23 Pinhole testing was performed if the presenting acuity was less than 6/12. When uncorrected refractive error was suspected, auto refraction (Retromax 3, Righton, Tokyo, Japan) and testing with correction were performed. Visual fields were assessed in adults only with a Humphrey Frequency Doubling Test (Zeiss, Dublin, California).

Fundal photographs of each eye were taken with a non-mydriatic retinal camera (Canon CR—DGi, Tokyo, Japan). If an adequate non-mydriatic retinal photograph could not be obtained, dilating drops (tropicamide (0.5%) and phenylephrine hydrochloride 2.5%) were used. Upon completion of the examination, arrangements were made for the appropriate treatment or referral for those people who required further attention. All retinal photographs were assessed in a masked fashion.

Participants in our study were regarded to have glaucoma if they had a cup:disc ratio (CDR) greater than 0.8 or if they missed 2 or more points on FDT testing and had a CDR greater than 0.7 in at least one eye.22 24 Quigley reported a sensitivity of 91% and specificity of 94% using two or more abnormal points on FDT perimetry for detecting glaucoma,25 while Casson et al reported results of 66% and 93% respectively.26 From our analysis, a CDR of 0.8 was >2 SDs above the mean, consistent with the scheme by Foster and co-authors for diagnosing glaucoma in population surveys.17

Data analysis

Data were entered into an electronic database using Access 2000. Data were checked for missing values, and where possible, missing data were followed up. The χ2 test was used to test categorical outcomes for significant differences in participants' characteristics by group. For continuous outcomes (eg, age, duration of diabetes) significant differences between strata were evaluated by the Mann–Whitney–Wilcoxon test or Student t test. Age-specific rates for adults are given in 10-year age groups. Logistic regression analysis was performed for analyses of risk factors and prevalence of glaucoma. OR and 95% CIs were presented. p Values of <0.05 have been taken to indicate statistical significance. All statistical analyses were carried out using STATA version 10.2.

Results

A total of 1189 eligible adult participants were examined for the study, representing 71.8% of the target population. A total of 128 cases were excluded from analysis owing to missing FDT data or ungradable images, leaving 1061 participants for analysis. In our cohort, 256 right eyes and 242 left eyes tested missed two or more points on FDT testing.

Optic-disc photographs from a 30% random sample of eligible adults from each of the 30 sample areas were formally graded for optic disc size and CDR. Results are shown in figures 1, 2. In summary, the mean optic disc diameter was 1.93 mm (SD 0.19) for left and right eyes, while the mean CDR for right eyes was 0.44 (SD 0.15), and that for left eyes was 0.43 (SD 0.16). The 97.5th percentile (2 SDs above the mean) is calculated to be 0.74–0.75. As grading of CDR was performed to only one decimal place, these results were applied as part of our definition for glaucoma.

Figure 1

Distribution of optic disc diameters for right and left eyes.

Figure 2

Distribution of cup:disc ratio (CDR) in the Australian aboriginal population.

Based on our definition of glaucoma (ie, CDR >0.8, or missing ≥2 points on FDT and a CDR >0.7 in at least one eye), the overall prevalence of glaucoma from this population sample was 26/1061 (2.2%). Persons with glaucoma were older, with a median age of 61 years compared with persons without glaucoma (χ2=3.62, p<0.001; table 1). There was also a greater proportion of males (p<0.03), visual acuity <6/12 (p=0.01) and blindness (p<0.001) in persons with glaucoma. Only 19.3% of persons with glaucoma had a self-reported history of glaucoma, while 2.8% of persons who had no signs of glaucoma also gave a history of glaucoma. Of the 26 persons with glaucoma, three were defined as having low vision (presenting VA <6/12 to ≤6/60), and three were defined as blind (presenting VA<6/60). The principal cause of low vision in these three persons was cataract in two (66%) cases and diabetic retinopathy in the other; and of the three persons with blindness, two (66%) were due to non-glaucomatous optic atrophy of unknown cause, and one (33%) was caused by cataract. These results are compared with persons without glaucoma and summarised in table 2. Interestingly, there were no cases of low vision or blindness solely attributable to glaucoma.

Table 1

Demographic characteristics of persons with glaucoma

Table 2

Causes of low vision and blindness in persons with and without glaucoma

Univariate analysis (table 3) revealed a strong association between glaucoma prevalence and increasing age (trend test: Z=4.38, p<0.001), with prevalence increasing from 1.5% in persons aged 40–49 years to 16.7% in persons aged 80 years or more (figure 3). Our multivariate logistic regression modelling showed males had twice the odds of being diagnosed as having glaucoma compared with women (OR 1.79, 95% CI 0.55 to 5.85). Interestingly, persons with no school or only primary school education also had increased odds of glaucoma 4.55 (95% CI 0.94 to 22.04).

Table 3

Univariate analysis of OR on glaucoma (yes/no)

Figure 3

Prevalence of glaucoma with age. MVIP, Melbourne Visual Impairment Study; NIEHS, National Indigenous Eye Health Survey; POAG, primary open-angle glaucoma.

Discussion

We report results on the optic disc and visual-field characteristics of adult Australian Aborigines from a large population-based survey of Aboriginal Australian eye health. We found a prevalence of glaucoma of 2.2% among Aboriginal Australians, which is higher than previously reported.3 4

Importantly, only 19.3% of persons with glaucoma reported a known history of glaucoma. The corollary is that almost 80% of the group with glaucoma were not aware of their condition. This is significantly higher than the commonly reported 50% prevalence of undiagnosed glaucoma for the general Australian population from the Visual Impairment Project6 and Blue Mountains Eye Study,8 and other international population-based surveys. This higher rate may also reflect the general lay community's confusion and lack of knowledge about glaucoma, and its potential confusion with trachoma. As Indigenous Australians have historically had the disadvantage of poorer access to and delivery of healthcare, our findings of a low rate of glaucoma awareness in combination with the higher-than-previously-thought prevalence of glaucoma among Indigenous Australians will need increased efforts by government and health practitioners to rectify.

Interestingly, more males were diagnosed as having glaucoma in this Aboriginal population. Most population studies find either no association with gender or a higher prevalence of glaucoma among females, probably as a consequence of survivor bias.6 8 27

The strongest risk factor for glaucoma is the presence of a first-degree relative with a history of glaucoma. Although this strongly suggests a genetic basis for glaucoma, evidence to date only shows a low percentage of primary open-angle glaucoma has a genetic link. Previous studies have reported glaucoma to be exceedingly rare in Indigenous Australians, and this could be attributed to the absence of genes predisposing to glaucoma in this population. One explanation for our observed prevalence of glaucoma could be the acquisition of glaucoma genes from intermarriage. We were unable to verify the degree of intermarriage among our population, and were interested to see if differences may become apparent by stratifying these data for analysis. We used education level, English language spoken at home and remote ARIA locations as potential proxies for the acquisition of European genes. No conclusive relationships were identified, other than finding persons with higher education levels being at lower risk of glaucoma. Although difficult to explain, this may reflect a greater likelihood for trauma in less educated persons, although evidence of ocular trauma was lacking.

Age is another important risk factor for glaucoma, and the increased prevalence found in our survey could be explained by sampling a larger proportion of older Indigenous Australians compared with previous studies.3 4 19

There has always been a recognition that optic disc size varies widely within the population, and as a corollary so does CDR. This is the first population survey to report characteristics such as optic disc diameter and CDR in an Indigenous Australian population. We found that the mean optic disc diameter was 1.93 mm, and the mean CDR was 0.44 for right eyes and 0.43 for left eyes. By comparison, Crowston et al reported that the median optic disc diameter of a Caucasian Australian population from the Blue Mountains Eye Study was 1.50 mm, with a median CDR of 0.43.28 Interestingly, Gerry and Johnson29 also found a statistically significant difference in CDR in their assessment of 52 Aboriginal and 52 non-Aboriginal youths, but noted an average CDR of 0.295 in Aboriginals and 0.159 in non-Aboriginals. Many reports have also described differences in the morphological characteristics of optic discs in persons of various ethnicities.30 31

The strengths of our study include a census-based study design and sampling framework, and the use of a standardised approach to identify a large sample of Indigenous Australians living in urban to remote locations of Australia. This is the first study of its kind to document findings from an indigenous Australian population. We acknowledge several limitations in our study, including the limited number of sampled sites and variable participation rate. The absence of a comprehensive ophthalmic examination has resulted in an inability to characterise factors such as central corneal thickness, nerve fibre layer analyses or even classifying glaucoma into its primary and secondary subtypes. Some readers may criticise how we defined glaucoma; however, because we were working within the auspices of an Aboriginal population general eye health survey, we used a well-recognised population-based epidemiological definition proposed by Foster et al to define cases with CDR ≥97.5th population percentile as glaucoma.17 Furthermore, several studies have shown high sensitivity and specificity of FDT perimetry.24–26 Lastly, our clinical data collection was insufficient to determine the genetic background and associations with other physiological risk factors for development of glaucoma in this population.

In conclusion, our study is the first to document a population prevalence of glaucoma, using an epidemiological definition of glaucoma in Indigenous Australians. We once again highlight the population variance in optic disc size and caution against simple reliance on an arbritary percentile cut-off CDR for a diagnosis of glaucoma. We postulate that the higher prevalence of glaucoma found from our survey may be due to the gradual introduction of glaucoma-associated genes through intermarriage, an older population sampled or differing diagnostic criteria for glaucoma.

References

Footnotes

  • Competing interests None.

  • Ethics approval Ethics approval was provided by the Royal Victorian Eye & Ear Hospital, Aboriginal Health and Medical Research Council of NSW, Aboriginal Health Council of South Australia, Menzies School of Health Research, Central Australia Human Research Ethics Committee, Western Australia Aboriginal Health Information and Ethics Committee, ACT Health, Tasmania Scientific Research Advisory Committee, Tasmania Health and Medical Human Research Ethics Committee, Queensland Aboriginal and Islander Health Council.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

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.