Professional Documents
Culture Documents
The Prevalence and Types of Glaucoma in An Urban Indian Population: The Singapore Indian Eye Study
The Prevalence and Types of Glaucoma in An Urban Indian Population: The Singapore Indian Eye Study
The Prevalence and Types of Glaucoma in An Urban Indian Population: The Singapore Indian Eye Study
Correspondence: Tin Aung, Singa- PURPOSE. To determine the prevalence and types of glaucoma in an urban Singaporean Indian
pore National Eye Centre, 11 Third population.
Hospital Avenue, Singapore 168751;
aung.tin@snec.com.sg. METHODS. The Singapore Indian Eye Study (SINDI) was a population-based, cross-sectional
survey that examined 3400 (75.6% response) persons aged 40 to 80 years. Participants
Submitted: March 1, 2013
Accepted: May 30, 2013
underwent a standardized examination including slit-lamp biomicroscopy, Goldmann
applanation tonometry, and dilated optic disc assessment. Participants suspected to have
Citation: Narayanaswamy A, Baskaran glaucoma also underwent visual field examination (24-2 SITA standard, Humphrey Visual
M, Zheng Y, et al. The prevalence and Field Analyzer II), gonioscopy, and repeat applanation tonometry. Glaucoma was defined
types of glaucoma in an urban Indian
population: The Singapore Indian Eye
according to International Society for Geographical and Epidemiologic Ophthalmology
Study. Invest Ophthalmol Vis Sci. criteria.
2013;54:4621–4627, DOI:10.1167/ RESULTS. Of the 3400 participants, 78 (2.29%) had diagnosed glaucoma, giving an age-
iovs.13-11950 standardized prevalence of 1.95% (95% confidence interval [CI], 1.5%–2.5%). The age-
standardized prevalence of primary open-angle glaucoma (POAG) was 1.25% (95% CI,
0.89%–1.73%), primary angle-closure glaucoma (PACG) 0.12% (95% CI, 0.04%–0.33%),
and secondary glaucoma 0.55% (95% CI, 0.35%–0.86%). The mean IOP among the
participants in the normal group in the study population was 15.6 6 2.6 mm Hg and
17.7 6 6.1 mm Hg in subjects with glaucoma (P ¼ 0.003). The mean central corneal
thickness (CCT) in the normal study population was 540.31 6 33.79; the mean CCT in
subjects with POAG (529.8 6 30.8 lm) was statistically different from the normal study
group (P ¼ 0.003).
CONCLUSIONS. The prevalence of glaucoma among Singaporean Indians 40 years of age and
older in Singapore was 1.95%, approximately half that of the Chinese and the Malay persons
in Singapore. As in other Asian studies, POAG was the main form of glaucoma accounting for
nearly 60% of cases.
Keywords: glaucoma, epidemiology, population, intraocular pressure
Copyright 2013 The Association for Research in Vision and Ophthalmology, Inc.
www.iovs.org j ISSN: 1552-5783 4621
All glaucoma
40–49 y 957 12 (1.25) 465 9 (1.94) 492 3 (0.61)
50–59 y 1077 17 (1.58) 529 10 (1.89) 548 7 (1.28)
60–69 y 887 29 (3.27) 464 16 (3.45) 423 13 (3.07)
70þ y 479 20 (4.18) 248 9 (3.63) 231 11 (4.76)
P for trend 0.006 0.191 0.005
Crude total 3400 78 (2.29) 1706 44 (2.58) 1694 34 (2.01)
Adjusted total* 1.95 (1.51–2.50) 2.29 (1.59–3.24) 1.66 (1.14–2.41)
POAG
40–49 y 957 11 (1.15) 465 8 (1.72) 492 3 (0.61)
50–59 y 1077 9 (0.84) 529 7 (1.32) 548 2 (0.36)
60–69 y 887 15 (1.69) 464 9 (1.94) 423 6 (1.42)
70þ y 479 11 (2.30) 248 4 (1.61) 231 7 (3.03)
P for trend 0.042 0.870 0.003
Crude total 3400 46 (1.35) 1706 28 (1.64) 1694 18 (1.06)
Adjusted total* 1.25 (0.89–1.73) 1.62 (1.02–2.49) 0.93 (0.54–1.55)
PACG
40–49 y 957 0 (0.00) 465 0 (0.00) 492 0 (0.00)
50–59 y 1077 2 (0.19) 529 1 (0.19) 548 1 (0.18)
60–69 y 887 4 (0.45) 464 3 (0.65) 423 1 (0.24)
70þ y 479 0 (0.00) 248 0 (0.00) 231 0 (0.00)
P for trend 0.329 0.367 0.709
Crude total 3400 6 (0.18) 1706 4 (0.23) 1694 2 (0.12)
Adjusted total* 0.12 (0.04–0.33) 0.14 (0.04–0.55) 0.09 (0.01–0.45)
Values in bold are statistically significant.
* Age-standardized rates (95% CI) based on Singapore Census of Population 2010 (Indian).
blood samples were drawn from all participants to determine 0.001), and there were no sex differences (P ¼ 0.28) between
levels of serum glucose and glycosylated hemoglobin (HbA1c). the two groups.
A subject was labeled hypertensive if the systolic BP ‡ 140 mm The mean age of the study population was 57.8 6 10.1
Hg or diastolic BP ‡ 90 mm Hg or physician diagnosis or self- years, and 49.8% were women. Mean IOP (right eye) was 15.75
reported history of hypertension. Diabetes mellitus was mm Hg, with 97.5 and 99.5 percentiles of 21.0 and 24.0 mm
diagnosed in a subject who had a nonfasting glucose level ‡ Hg, respectively. Mean VCDR in the normal study group (right
200 mg/dL (11.1 mmol/L) at examination, or had a physician eye) was 0.41, with 97.5 and 99.5 percentiles of 0.60 and 0.62,
diagnosis of diabetes and was using diabetic medications. respectively. Mean VCDR asymmetry was 0.03 with 97.5 and
99.5 percentiles of 0.13 and 0.20, respectively. Of the 3400
Statistical Analysis participants, 365 (10.7%) were classified as having suspected
glaucoma. Of these, 78 (2.29%) had glaucoma, of which 46
Statistical analysis was performed with commercial software (58.97%) had POAG, 6 (7.69%) had PACG, and 26 (33.3%) had
(SPSS; ver. 20, SPSS Inc., Chicago, IL). Prevalence estimate of secondary glaucoma. Table 1 shows the overall crude and age-
glaucoma was performed for the whole cohort and in age- and standardized prevalence of glaucoma, POAG, and PACG. The
sex-stratified groups. Prevalence rates were standardized to the age-standardized prevalence rate of glaucoma was 1.95% (95%
population distribution from the 2010 Singapore Indian confidence interval [CI], 1.51–2.50) with a POAG prevalence
Census,9 using the direct method of adjustment. Logistic of 1.25% (95% CI, 0.89–1.73) and PACG prevalence of 0.12%
regression was used to assess the odds of glaucoma with (95% CI, 0.04–0.33). The prevalence of glaucoma increased
increasing age. Independent t-test was used for comparison of with age (P for trend ¼ 0.006) and was higher in participants
means and chi-square or Fisher’s exact test was used for aged 60 to 69 years (odds ratio [OR], 2.66; 95% CI, 1.35–5.25; P
comparison of proportions between groups. P-value < 0.05 ¼ 0.005) and 70 to 80þ (3.43; 1.66–7.08; P ¼ 0.001) compared
was considered as statistically significant. We also performed a with those aged 40 to 49 years. The standardized prevalence of
comparative analysis of the age-specific prevalence of POAG
secondary glaucoma was 0.55% (95% CI, 0.35%–0.86%). The
from prior reports among South Indian subjects aged 40 years
mean IOP among the normal subjects in the study population
and above.12–14,22
was 15.6 6 2.6 mm Hg and 17.7 6 6.1 mm Hg in subjects with
glaucoma (P ¼ 0.003). Of the 46 participants with POAG, 38
(82.6%) of subjects with IOP 21.0 mm Hg and three (50.0%)
RESULTS of the six participants with PACG had an IOP 21.0 mm Hg.
There were 3400 participants in SINDI (75.6% participation The mean CCT in the normal study population was 540.31 6
rate). Of the nonparticipants, 1021 (22.7%) refused to 33.79 lm. The mean CCT in subjects with POAG (529.8 6 30.8
participate and 76 (1.7%) were not contactable. Nonpartici- lm) was statistically lower than the normal study group (P ¼
pants on average were slightly older than participants (P ¼ 0.003).
Table 2 shows the characteristics of subjects with specific The overall prevalence of glaucoma among Singaporean
types of glaucoma. Among the attributable causes of secondary Indians was 1.95%, with POAG being the most predominant
glaucoma (0.81%), glaucoma with pseudophakia (0.38%) and form at 1.25%. The prevalence of PACG was noted to be rather
neovascular glaucoma (0.12%) were the most common low at 0.12%. There was a significant increase in glaucoma
diagnoses. ISGEO categorization of the glaucomas (Table 3) prevalence with age, and this was higher among females. The
revealed a predominance of categories 1 (53.8%) and 2 (42.3%) glaucoma prevalence in the Singaporean Indian population
in this population. Of the 78 glaucoma cases, 13 (16.6%) had was significantly lower than that of their Chinese (3.4%) and
low-vision (logMAR, >0.30 to <1.00) and eight (10.2%) were Malay (3.2%) counterparts in the Tanjong Pagar Eye Study23
blind (logMAR, ‡1.00), according to the primary definition. and the Singapore Malay Eye Study,24 respectively, which used
The proportion of subjects with low vision was equal (30.7%) the same definitions and classification criteria.
among POAG and PACG subjects. Blindness due to glaucoma The census from year 2000 recorded that Singaporean
was noted to be associated with POAG in two subjects (2.56%) Indians are predominantly persons with ancestry traced to the
and from secondary causes in the remaining six subjects southern region of India (from the states of Tamil Nadu and
(7.69%). Kerala) and account for 66.6% of the Indian community in
The distribution of males compared with females was Singapore.25 A detailed analysis of our study sample revealed
similar in all categories of glaucoma. The associations of age, that the proportion of subjects with a South Indian ancestry
sex, IOP, CCT, myopia, hypertension, and diabetes mellitus was close to 90%. This aspect has important implications
with POAG were evaluated by logistic regression and the data because our data become directly comparable to recent studies
are presented in Table 4. Subjects with POAG were associated on subjects from this region in India, such as the Aravind
with a higher IOP and lower CCT. Out of 78 subjects with
glaucoma, 56 (71.8%) were not aware that they had glaucoma.
Analysis estimating the pooled prevalence of POAG (Figure) TABLE 4. Multiple Logistic Regressions for Risk Factors for POAG
from 17,258 South Indian subjects aged 40 years and above Among Singaporean Indians
from three major studies12,14,22 revealed that an overall
prevalence of POAG of 2.04% (95% CI, 1.30%–2.78%). Further, N OR (95% CI) P
this estimate was 2.53% (95% CI, 0.97%–4.08%) in urban subset
Age group, y
of the population and 1.41 (95% CI, 1.04%–2.78%) among the
rural subset of the population. 40–49 911 Ref
50–59 1000 0.76 (0.31–1.87) 0.543
60–69 803 1.37 (0.58–3.24) 0.480
DISCUSSION 70þ 404 1.83 (0.71–4.72) 0.213
Sex
This study is the first epidemiological study to evaluate the
prevalence of ocular disease in a migrant Indian population. Male 1581 Ref
Female 1538 0.60 (0.32–1.11) 0.104
TABLE 3. Glaucoma Diagnosis Based on ISGEO Categorization IOP, mm Hg 3119 1.11 (1.03–1.20) 0.006
CCT, lm 3119 0.99 (0.98–0.99) 0.019
Category 1 Category 2 Category 3
Myopia
Glaucoma Subtype (%) (%) (%)
Yes 856 1.14 (0.59–2.20) 0.696
POAG 29 (31.17) 17 (21.79) — No 2263 Ref
PACG 3 (3.84) 3 (3.84) — Hypertension
PXF glaucoma 2 (2.56) — —
Yes 1746 0.91 (0.46–1.77) 0.772
Neovascular glaucoma 1 (1.28) 1 (1.28) 2 (2.56)
No 1373 Ref
Pigment dispersion glaucoma 1 (1.28) — —
Steroid-induced glaucoma — 1 (1.28) — Diabetes
Glaucoma with pseudophakia 5 (6.41) 8 (10.25) — Yes 1064 1.13 (0.59–2.18) 0.705
Unspecified glaucoma 2 (2.56) 2 (2.56) 1 (1.28) No 2055 Ref
Total 43 (55.12) 32 (41.02) 3 (3.84)
Ref, reference group.
FIGURE. Pooled prevalence of POAG in South Indians aged 40 years and older (using crude prevalence figures). APEDS, Andhra Pradesh Eye Disease
Study14; CGS, Chennai Glaucoma Study12; SINDI, Singapore Indian Eye Study; ACES, Aravind Comprehensive Eye Survey.22
Comprehensive Eye Survey (ACES),22 Chennai Glaucoma Study Although the prevalence of POAG in our sample was lower, the
(CGS),12,13 and the Andhra Pradesh Eye Disease Study trend follows that of the rural arms of the CGS and APEDS and
(APEDS).14 A compilation of comparative data from glaucoma of note, there is an overlap of the confidence intervals (95% CI,
prevalence from various parts of southern India is shown in 0.89–1.73 vs. 1.19–1.95 [CGS]; 1.13–2.06 [APEDS]).12,14 The
Table 5. The POAG prevalence in our study has a close match overall prevalence of POAG in our study samples correspond
to the prevalence noted in ACES with both reporting a to the prevalence noted in the rural samples of the ACES, CGS,
prevalence of 1.2%.22 An important aspect of this data is that and APEDS, which is intriguing. Though the Singaporean
the population sample in ACES was derived from a predom- Indians by all means would be considered highly urbanized,
inantly rural setting. In comparison, the CGS reported age- the prevalence of POAG did not match the urban South Indians
standardized POAG prevalence of 1.57% in the rural arm of of the CGS (3.46%) and the APEDS (4.0%). The authors from
their study and APEDS reported a prevalence of 1.6%.12,14 the CGS study attributed the discrepancy of POAG prevalence
TABLE 5. Age-Standardized Rates of POAG and PACG in Comparative Population Subsets from India and Singapore
between their rural and urban populations to significant aware population is likely to be more robust. The current level
socioeconomic and lifestyle differences.12 Interestingly, these of awareness does not seem to translate into this ideal state,
aspects of urbanization did not seem to affect the prevalence of and reasons for this could be due to failure of subjects to report
POAG among the Singaporean Indians, who may be considered for routine eye health screenings, high health care costs, or
to have a higher socioeconomic status compared to the rural failure of strategy used in public awareness programs. This
and urban counterparts in the CGS or the APEDS. The indicates that there is a specific need to look into the strategies
prevalence of PACG was also low at 0.12% and matched that and approaches that could help overcome this deficit.
of the Malay subjects in Singapore but was far lower than that The SINDI study involved a large number of subjects with a
reported in native southern Indian population (0.5%– long history of migration and had a good response rate (75.6%).
1.8%).13,14,22 The urban subjects in the CGS13 and APEDS14 The limitations of our study include the lack of visual field and
reported rates of 0.80% and 1.8%, respectively, whereas the gonioscopy assessments in all subjects, which could have
prevalence in the rural subjects of ACES,22 CGS,13 and APEDS14 contributed to some underestimation of disease, specifically
study was 0.5%, 0.85%, and 0.7%, respectively. We have to angle closure. Furthermore, the single reading approach used
acknowledge the likelihood of the disparities between the for IOP measurements and older age of the nonresponders
urban glaucoma prevalence rates noted between CGS, APEDS, could also have led to underestimation of glaucoma prevalence
and our study could be due to differences in methodologies. among the Singaporean Indians.
The definitions of glaucoma in our study and the CGS were In summary, the prevalence of glaucoma among Indian
based on the ISGEO18; however, in the CGS, gonioscopy and Singaporean persons 40 years of age and older in Singapore
automated perimetry were done for every patient, and this was 1.95%, far lower than the Chinese and the Malay cohorts of
could have contributed to a higher yield of suspected the same population. POAG was the main form of glaucoma,
glaucoma cases. Further, factors that could contribute to this accounting for 58.9% of disease. Our study involved a large
large variation could be subjective differences in gonioscopic number of subjects with a long history of migration and had a
assessment and the proportion and mean age of nonrespond- good response rate (75.6%). The prevalence of POAG matched
ers. We noted that there were 1094 (32.1%) nonresponders for the rural cohorts of the native Southern Indian population and
the study; these subjects had a significantly higher mean age in contrast, the PACG rate was far lower. The effects of urban
than the responders (61.1 vs. 57.8 years; P < 0.001), and this environment on glaucoma prevalence were not obvious among
could have also partially contributed to the low prevalence the Singaporean Indians compared with their South Indian
rate. We also considered the possibility of a higher cataract counterparts.
surgery rate in Singapore as a potential cause for lower
prevalence of PACG but found that the age-adjusted cataract
surgery rates were marginally lower compared to CGS (9.7% vs. Acknowledgments
10.16%)26 and far lower than that reported in APEDS (9.7% vs. Supported by the Biomedical Research Council (08/1/35/19/550)
14.6%).27 and the National Medical Research Council, STaR/0003/2008,
The trend of POAG increasing with age12,14,22,28,29 has been Singapore.
substantiated in our sample as well, and similarly, the risk of Disclosure: A. Narayanaswamy, None; M. Baskaran, None; Y.
glaucoma was twice as much in subjects older than 70 years. Zheng, None; R. Lavanya, None; R. Wu, None; W.-L. Wong,
Our study noted a higher male preponderance for both POAG None; S.-M. Saw, None; C.-Y. Cheng, None; T.-Y. Wong, None; T.
and PACG, but this variation in sex distribution was not Aung, None
significant and was similar to the findings noted in the urban
subset of APEDS. The CGS also reported equal prevalence of
PACG across the sexes in the urban population (17:17) and a References
female preponderance among the rural population (7:27); 1. Mott IVWH. Globalization: People, Perspectives, and Pro-
however, the authors concluded that this did not translate to gress. Praeger Publishers: Westport, CT; 2004.
significantly higher prevalence of PACG among the rural
2. Appadurai A. Modernity at Large: Cultural Dimensions of
women.13 The mean age of our study population was slightly
Globalization. University of Minnesota Press: Minneapolis;
higher at 57.8 6 10.1 years compared with 54.8 6 10.6 years 1996.
in CGS and 53.2 6 10.9 in APEDS. Despite this, the glaucoma
prevalence in our population was lower, and this could be 3. Gushulak BD, MacPherson DW. The basic principles of
migration health: population mobility and gaps in disease
partly attributed to differing study methodologies.
prevalence. Emerg Themes Epidemiol. 2006;3:3.
Glaucoma with pseudophakia was the predominant catego-
ry of secondary glaucoma at 0.38%, closely matching the 4. Swende TZ, Sokpo J, Tamen FI. Globalization and health: a
reports from APEDS (0.21%), ACES (0.30%), and SIMES critical appraisal. Niger J Med. 2008;17:135–138.
(0.42%).22,24,30 The CGS reported a fairly high prevalence rate 5. Gushulak B, Weekers J, Macpherson D. Migrants and emerging
of glaucoma associated with aphakia and pseudophakia (urban public health issues in a globalized world: threats, risks and
arm, 0.99%; rural arm, 1.38%).26,31 The authors of CGS challenges, an evidence-based framework. Emerg Health
attributed this to inadequate preoperative assessment and Threats J. 2009;2:e10. doi:10.3134/ehtj.09.010.
poor quality of surgery especially with these zones of the 6. Salehi R. Intersection of health, immigration, and youth: a
country being served by large-scale rural mass camp surgery systematic literature review. J Immigr Minor Health. 2010;12:
programs.31 788–797.
Glaucoma and eye diseases awareness continues to be a 7. Bundred P, Gibbs T. Medical migration and Africa: an
major public health issue even in the most developed countries unwanted legacy of educational change. Med Teach. 2007;
and health care systems.32,33 More than 70% of subjects in this 29:893–896.
study were not aware of their glaucoma. Though this is better 8. Hargreaves S, Friedland JS, Gothard P, et al. Impact on and use
than the rate of >80% reported in the urban population of of health services by international migrants: questionnaire
CGS,34 it did not translate to a lower rate of POAG-related survey of inner city London A&E attenders. BMC Health Serv
visual morbidity among the Singaporean Indians (SINDI, 2.56% Res. 2006;29:153.
versus CGS, 2.2%). In general higher awareness could lead to 9. Singapore Department of Statistics. Singapore in figures 2010.
improved detection because the health-seeking pattern of an Ethnic distribution. 2009. Available at: http://www.singstat.