Abstract
Purpose.:
To compare the prevalence and risk factors of diabetes (DM) and diabetic retinopathy (DR) in a multi-ethnic Asian population of Chinese, Malays, and Indians in Singapore.
Methods.:
A total of 2919 individuals participated in a population-based, cross-sectional study in Singapore of Chinese (n = 1633), Malays (n = 658), and Indians (n = 628) aged 40 to 95 years, with retinal photographs, graded using the Early Treatment Diabetic Retinopathy Study (ETDRS) severity scale. DM was defined as fasting plasma glucose ≥7 mmol/L, self-reported physician-diagnosed diabetes, and use of glucose-lowering medication.
Results.:
The overall age-standardized prevalence of diabetes was 13.8% (Chinese, 11.5%; Malays, 17.1%; and Indians, 21.6%; P < 0.0001). Among persons with diabetes (n = 401), the overall age-standardized prevalence of DR was 25.4% (20.1%, 24.8%, and 28.9% in Chinese, Malays, and Indians, respectively; P = 0.290). In multivariate analysis, longer diabetes duration (odds ratio [OR], 1.05; 95% confidence interval [CI], 1.01–1.09, per year increase), higher glycated hemoglobin (OR 1.25; 95% CI, 1.01–1.54, per 1% increase), and serum creatinine levels (OR, 1.01; 95% CI, 1.00–1.03, per mg/dL increase) were the independent risk factors of DR in the whole population. Race was not found to be associated with DR (OR, 1.35; CI, 1.00–1.83). The associations of major risk factors with DR were similar among the three ethnic groups.
Conclusions.:
There was a significant difference in the prevalence of diabetes between Chinese, Malays, and Indians. The main risk factors of DR, similar among the three ethnic groups, are longer diabetes duration, higher hbA1c, and higher creatinine levels. No significant racial differences were found in the prevalence of DR among persons with diabetes.
Diabetes mellitus (DM) affects over 246 million people worldwide. About a third have diabetic retinopathy (DR)
1 and of those, a third will have vision threatening retinopathy (defined as severe retinopathy or macular edema).
2 The World Health Organization (WHO) estimates that by 2025, an estimated 300 million people will have diabetes, with many expected to develop some level of retinopathy.
1 DR is the major cause of vision impairment and blindness in people with diabetes.
3 –6 The prevalence of DR is expected to grow exponentially as diabetes continues to increase globally.
7,8
The prevalence of diabetes appears to vary between racial and ethnic groups.
9,10 For example, McBean et al.
11 found the increase in prevalence rates to be the greatest among Hispanics and Asian Americans. The authors attributed the differences to Medicare entitlements in the Hispanic groups and the westernization of lifestyle, particular in diet among Asians.
11 Similar findings have been reported by Harris et al.
12 There have also been suggestions that environmental and socioeconomic factors may contribute to racial disparities in the prevalence of diabetes.
13 For instance, LaVeist et al.
13 found, in integrated communities with similar health risk environments and sociodemographic factors, that the prevalence estimates of diabetes were similar in African Americans and whites.
Like diabetes, it has also been suggested that the frequency of DR may vary by race/ethnicity. African Americans have a fourfold risk of DR-induced vision impairment compared to non-Hispanic whites.
14 In the LALES study (Los Angeles Latino Eye Study), Latinos were found to have a higher rate of more severe threatening DR than whites (Varma R, et al.
IOVS 2005;46:ARVO E-Abstract 1164).
15 Dowse et al.
16 did not find any variation in the prevalence of DR observed in all major ethnic groups in Mauritius. However, although there have been large studies conducted on the epidemiology of DR in Asia,
17 –19 none has investigated interracial/ethnic variations in the prevalence and risk factors of DR among the three major Asian groups (Chinese, Malays, and Indians). This information is important, as disparities in health for racial and ethnic groups can represent unequal access to care.
20,21
In this study we examined the prevalence and risk factors of diabetes and DR in a population-based sample of Chinese, Malays, and Indians living in Singapore. These three ethnic groups represent two thirds of the world's population. Furthermore, Singapore is the ideal location in which to conduct this research because all three groups live in one geographic area, with largely similar environmental and sociodemographic factors.
All serum biochemistry tests were performed at the National University Hospital Reference Laboratory. The biochemistry tests included total cholesterol, high-density lipoprotein cholesterol (HDL), low-density lipoprotein (LDL) cholesterol, triglycerides, and fasting plasma glucose. Fasting blood samples measured serum lipids and plasma glucose. During the interview, cardiovascular disease (CVD) was defined as self-reported myocardial infarction, angina, or stroke. Main risk factors of CVD were also assessed: cigarette smoking (currently smoking everyday or on some days) and alcohol consumption (currently consuming alcoholic beverages daily or on some days). Systolic and diastolic blood pressures were measured with a digital automatic blood pressure monitor (Dinamap model Pro100V2; Criticon GmbH, Norderstedt, Germany), after the participant had been seated for at least 5 minutes. Participants classified as having hypertension had a systolic blood pressure of 140 mm Hg or more or a diastolic blood pressure of 90 mm Hg or more at examination, a reported history or a self-reported history of physician-diagnosed hypertension, or use of prescribed antihypertensive medications, or both. Mean arterial blood pressure was calculated as two thirds of the diastolic value plus one third of the systolic value. Body mass index (BMI) was calculated as weight divided by the square of height in meters (kg/m2).
Characteristics of the study population and risk factors were examined using proportions, means, medians, percentiles, and standard deviations. The χ2 test and t-test were used for univariate associations. Age-standardized prevalence estimates were calculated via the direct method using the 2010 Singapore population census. The prevalences of diabetes and DR were estimated and compared between the three ethnic groups: Chinese, Malay, and Indians. Binary logistic regressions were performed to examine the association of DR with various risk factors in the ethnic groups and were adjusted for age and sex. Interactions between the presence of any DR and each of the potential confounders were explored by including appropriate interaction terms in the model. The Mantel-Haenszel method was used to test for differences between the prevalence estimates of the three ethnic groups (all statistical analyses: Stata, ver. 11.0; StataCorp, College Station, TX).
This is the first population-based study comparing interracial/ethnic variations in the prevalence and risk factors of diabetes and DR in three major Asian groups aged 40 to 95 years. There was a significant difference in the prevalence of diabetes between the Chinese (20.1%), Malays (24.8%), and Indians (28.9%). However, overall, race was not found to be a risk factor for diabetes and when further stratified by the three ethnic groups, the risk factors were similar across all groups. The prevalence of DR among persons with diabetes was also largely similar among the three ethnic groups. Race again was not found to be an independent risk factor for DR, although creatinine level, longer duration of diabetes, and higher HbA1c level were found to be associated with the outcome.
Our race-specific DR prevalence rates differed from other population-based studies conducted in Asians. For instance, the prevalence rates among Chinese people aged 45 years and above in the Beijing
31 and Handan eye studies were 27.9% and 43.1%, respectively.
19 Similar studies in India have reported DR prevalence rates of 22.4%
32 and 26.8%.
33 The Singapore Malay Eye Study (SiMES)
17 found the overall prevalence of any retinopathy to be 35.0%.
17 Several reasons could explain these underlying variations. First, the differences could be related to the duration of diabetes, because it is one of the strongest risk factors for DR. Because of the delay in diagnosis of diabetes, the actual diabetes duration is often imprecise. Socioeconomic issues or biases related to care referrals or access (i.e., patterns of access to and utilization of medical care between races and across states or urban and rural areas) may be some of the factors causing the variations.
34 –36 For example, in the United Kingdom, the coverage rates of screening and uptake of eye care among ethnic minority groups in inner city areas can be much lower than those for white Europeans.
37
Many Indian studies and the Handan Eye Study were conducted in predominantly rural regions, where the diagnosis of diabetes is usually delayed for some years, most probably because of a lack of awareness and poor access to health care services. The prevalence of DR in Handan is high and differs approximately 5 years in duration compared to that in some Western countries.
19 This difference could signify that the diagnosis of diabetes is delayed on average about 5 years, so the DR prevalence for 10 years' duration in Handan probably equates to that for 15 years' duration in Western countries. Many Indians develop type 2 diabetes at an earlier age.
38 In fact, the Indians in our study were reported to have a longer duration of diabetes (
Table 1), which further suggests that other mechanisms or different pathogeneses are at play.
39
The prevalence rates for Malays differed from those in the SiMES study, perhaps because of detection biases and methodology differences. Differences between the studies include the graders, the number of retinal photograph fields per eye, and the definitions used to define diabetes. The SiMES study collected nonfasting venous blood samples, whereas in the SP2 study fasting venous blood samples were analyzed.
Few studies have assessed variations in the prevalence of DR among different races and most have used non-Asian populations. For instance, Raymond et al.
39 found a significantly higher prevalence of DR and higher levels of HbA1c among patients of South Asian ethnicity than in white Europeans in the community. Eberhardt et al.
34 and Harries et al.
40 found that black people had higher levels of HbA1c in two South Carolina communities. When social factors such as low income or geography between the groups were equalized the effect of race was neutralized, suggesting predictors other than race. For example, Lim et al.
41 did not find race to be a predictor of DR within an urban, underserved population of whites, blacks, Hispanics, and Asians. This could also partially explain the underlying reason that our study did not find race to be a risk factor, as all Singaporeans live in a highly contained urbanized geographic area in proximity to one another. Thus, one can postulate that access to care and socioeconomic statuses for instance could be similar.
Age, lipids, BMI, and glucose are all documented risk factors of diabetes.
42 –44 The development of PAD has also been found to be associated with diabetes.
45 Research has shown that LDL and cholesterol can regulate the function and survival of β cells.
46 Thus, dyslipidemias may not only be consequences of but also contributors to the pathogenesis of type 2 diabetes and hence are targets for prevention.
46
HbA1c and duration of diabetes also remain the classic predictors for the onset and progression of DR regardless of ethnicity and this was confirmed by our present study findings.
16,17,31,36,47,48 Wong et al.
36 showed that the risk factors for DR were similar across four racial/ethnic populations in six U.S. communities. The role of diabetes duration and HbA1c is well described by numerous clinical trials.
49 –52 As for creatinine, it is an indicator of chronic kidney disease,
53,54 and there has been evidence showing that diabetic nephropathy may be used as a tell-tale sign of DR, necessitating more intensive ophthalmic care, especially in patients with a longer duration of diabetes.
55 Thus, our findings of creatinine as a possible risk factor are also consistent with those in the existing literature.
The major strength of our study is that it was a large, population-based sample of three racial and ethnic groups, and information was available for a range of possible confounders. However, this study also has its limitations. First, the number of people with DM and DR was small (
n = 107), even though the overall sample size was large. Insufficient statistical power may be the reason underpinning our inability to detect a difference in the prevalence and risk factors of DR among the three ethnic groups.
56 Second, we could not explore the race associations between the different types of DR (i.e., severity, macular edema, and vision-threatening retinopathy) as the numbers were too small. Finally, the analyses were cross-sectional, which limited us in making causational inferences between the variables and the outcome.
In summary, our study demonstrates significant differences in the prevalence of diabetes between Chinese, Malays, and Indians. We found no race effect, although this may be related to our small sample of people with DM or DR. Future research in this area is warranted. Major risk factors of diabetes are age, lipids, glucose, and PAD. The major risk factors of DR are higher creatinine level, longer diabetes duration, and higher hbA1c level. The risk factors for both diabetes and DR are similar among the three ethnic groups aged 40 years and older. Our study findings imply the importance of optimizing diabetes control and addressing the key risk factors that lead to DR through early prevention, detection, and maintaining a high quality of clinical care.
Supported by Biomedical Research Council (BMRC) Grant 08/1/35/19/550 and National Medical Research Council (NMRC) Grant Star/0003/2008, Singapore.
Disclosure:
P.P.C. Chiang, None;
E.L. Lamoureux, None;
C.Y. Cheung, None;
C. Sabanayagam, None;
W. Wong, None;
E.S. Tai, None;
J. Lee, None;
T.Y. Wong, None