You are here

Article Text

Article menu

Download PDFPDF

Clinical science
Six-year incidence and progression of diabetic retinopathy in Indian adults: the Singapore Indian Eye study
  1. Neelam Kumari1,2,3,4,
  2. Mayuri Bhargava2,3,5,
  3. Duc Quang Nguyen2,
  4. Alfred Tau Liang Gan2,
  5. Gavin Tan2,3,5,
  6. Ning Cheung2,3,
  7. Nicholas Tan5,
  8. Charlene Wong5,
  9. Jie Jin Wang3,
  10. Paul Mitchell6,
  11. Ecosse L Lamoureux2,3,
  12. Ching Yu Cheng2,3,5,
  13. Tien Yin Wong2,3,5,
  14. Charumathi Sabanayagam2,3,5
  1. 1 Department of Ophthalmology and Visual Sciences, Khoo Teck Puat Hospital, Singapore, Singapore
  2. 2 Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, Singapore
  3. 3 Ophthalmology and Visual Sciences Academic Clinical Program, Duke-NUS Medical School, National University of Singapore, Singapore, Singapore
  4. 4 Department of ophthalmology, National University Health System, Singapore, Singapore
  5. 5 Department of Ophthalmology, Yong Loo Lin School of Medicine, National University of Singapore and National University Health System, Singapore, Singapore
  6. 6 Center for Vision Research, Westmead Institute for Medical Research, University of Sydney, Sydney, New South Wales, Australia
  1. Correspondence to Dr Charumathi Sabanayagam, Singapore Eye Research Institute, The academia, Singapore, 169856, Singapore; charumathi.sabanayagam{at}seri.com.sg

Abstract

Aims Diabetes is a major public health problem in migrants and ethnic minorities worldwide. We determined the incidence and risk factors of diabetic retinopathy (DR) in migrant Indians living in Singapore.

Methods We included data from 759 Indian adults with diabetes, who participated in the baseline (aged 40–80 years, 2007–2009) and 6-year follow-up 2012–2015 of the Singapore Indian Eye Study. Retinal photographs were graded for the presence and severity of DR using modified Airlie House Classification. Incidence was assessed in participants who were free of DR at baseline visit (n=501), while progression in those with DR but free of proliferative DR at baseline visit (n=189). Risk factors included demographic, lifestyle, socioeconomic, family history, genes, duration of diabetes, glycaemic control, insulin use, ocular and clinical factors.

Results The 6-year age-standardised DR incidence and progression were 21.89% and 33.45%, respectively. HbA1c (risk ratio (RR) 1.41, 95% CI 1.28 to 1.55 per unit increase), current smoking (RR 1.63, 95% CI 1.02 to 2.62) and insulin use (RR 2.63, 95% CI 1.44 to 4.82) were associated with higher incidence, whereas estimated cerebrospinal fluid pressure (RR 0.90, 95% CI 0.82 to 0.98) and body mass index (BMI) (RR 0.74, 95% CI 0.60 to 0.93) were associated with lower incidence of DR. Higher HbA1c (RR 1.26, 95% CI 1.13 to 1.42), BMI (RR 1.26, 95% CI 1.02 to 1.56) and estimated cerebrospinal fluid pressure (RR 1.11, 95% CI 1.02 to 1.21) were associated with DR progression. The population attributable risk of HbA1c >8% was 41.29% and 49.63% for DR incidence and progression.

Conclusion DR incidence and progression in migrant Indians living in Singapore was more than double that reported in Indians living in urban India. Consistent with past studies, poor glycaemic control was an important predictor for incidence and progression of DR.

  • epidemiology
  • public health
  • retina

http://dx.doi.org/10.1136/bjophthalmol-2018-313282

Statistics from Altmetric.com

    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.

    Introduction

    Diabetes is a major public health problem worldwide, in particular in ethnic minorities and migrant populations. Diabetic retinopathy (DR) is a relatively specific microvascular complication of diabetes and remains a leading cause of avoidable blindness among individuals of working age worldwide.1 It has been estimated that total number of diabetic people will double from 171 million in 2000 to 366 million by 2030 with greatest increases expected in Asia (India and China).2 With increasing prevalence of diabetes and increasing life expectancy among those with diabetes, the number of persons with DR and vision-threatening DR has been projected to rise to 191.0 million and 56.3 million, respectively, by 2030.3

    Several population-based studies have reported prevalence and risk factors of DR in both developed and low- and middle-income world, including regional and ethnic differences.4–6 However, up to date incidence estimates and risk factors for DR progression from population-based cohort studies are limited. Most of the incidence estimates and risk factor information described in Western populations were conducted before year 2000, an era of poor diabetic control and areas with no population-based screening for diabetes or ineffective public health systems.7–9 Moreover, classic long-term incidence studies have been conducted in type 1 diabetes while studies targeting type 2 diabetic patients are relatively fewer.10

    Asian Indians are one of the fastest growing ethnic minorities around the world. We have earlier shown that prevalence of diabetes, hypertension and DR in migrant Indians living in Singapore were significantly higher than Indians living in India.11 Three previous studies conducted in China and India that examined the incidence and risk factors of DR have shown vastly different incidence estimates ranging from 10-year cumulative incidence of 4.2% in the Beijing Eye study,12 to 4-year incidence of 9.2% in the Sankara Nethralaya-Diabetic Retinopathy Epidemiology and Molecular Genetics Study (SN-DREAMS),13 to 5-year incidence of 46.9% in the Beixinjing community in China.14 It is unclear if these variations reflect differences in study population characteristics, access to care or other methodological differences. Moreover, these studies being conducted in low- and middle-income countries where coordinated public health efforts to tackle the burden of DR are lacking, they may not be reflective of the contemporary estimates in the post public health era. To address the above limitations, we conducted a population-based cohort study in Singapore involving Asian Indians who are at higher risk of developing diabetes and live in a country with well-designed public health system and improved access to care.

    Methods

    Study design and population

    The Singapore Indian Eye Study (SINDI) is a population-based study of Indian adults aged ≥40 years living in Singapore. Baseline examinations were conducted between years 2007 and 2009 (SINDI-1, n=3400) and 6-year follow-up examinations between years 2012 and 2015 (SINDI-2, n=2200). Both studies were conducted in accordance with the tenets of the Declaration of Helsinki. All participants understood the study protocol and provided written informed consent to take part in the study.

    The methodology for the baseline and follow-up examination has been described in detail previously.11 15 A total of 1289 subjects (both previously detected and newly diagnosed) had diabetes at baseline examination, and of these, 759 participated in the 6-year follow-up study. Diabetes was defined as self-report of a previous diagnosis of the disease by a doctor, use of diabetes medication or haemoglobin A1c (HbA1c) ≥6.5% as recommended by the American Diabetes Association. Only participants with type 2 diabetes (defined as age of onset after 30 years) were included. The details of the 759 diabetic subjects who participated in the follow-up study are shown in the flow chart (figure 1). All participants underwent standardised clinical and ocular assessment, questionnaire interview and blood biochemical analyses.

    Figure 1
    Figure 1

    Patient selection flow chart.DR, diabetic retinopathy; SINDI, Singapore Indian Eye Study.

    Ophthalmic examination and DR grading

    A comprehensive eye examination following baseline examination protocol was performed to capture the following data: distance presenting visual acuity (VA), subjective refraction, best corrected distance and near VA, autorefraction, keratometry, ocular biometry, slit-lamp biomicroscopy of anterior segment and tonometry. Colour photographs of ETDRS standard fundus fields 1 (centred on the optic disc) and 2 (centred on the macula) were obtained for both eyes following pupil dilation using a digital retinal camera (Canon CR-DGi with 10-D SLR back, Canon, Tokyo, Japan). Photographs were graded by trained graders in a masked manner using an adaptation of the modified Airlie House Classification.16

    DR was considered present if any characteristic lesion as defined by ETDRS was visualised. Questionable retinopathy lesions were adjudicated and other definite lesions reviewed before the final retinopathy grades were determined. In grouping participants according to retinopathy level, we followed the ETDRS summary, with an additional severity level (14 haemorrhages without microaneurysms, and 15, hard exudates, soft exudates or intraretinal microvascular abnormalities in the absence of any microaneurysms) to define the presence of ‘questionable’ DR. As in the Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR),17 the retinopathy level of each participant was derived by concatenating the levels for the two eyes, giving the eye with the higher level greater weight, as shown in the online supplementary table 1. If the DR severity could not be graded in an eye, it was considered to have a score equivalent to that in the other eye.

    Incidence of DR was defined as development of any DR (defined as score level of ≥20 in at least one eye) at follow-up examination with no DR at baseline (with a score level of 10/10 in both eyes). An increase of at least two steps on the 15-step scale over the 6-year period from the baseline examination in eyes with any level of existing DR (including questionable DR with a score level of 14/15) was considered to indicate progression of retinopathy. Vision-threatening DR (VTDR) was defined as the presence of severe non-proliferative DR (NPDR) and PDR, using the Eye Diseases Prevalence Research Group definition. Diabetic macular oedema (DME) was defined by hard exudates in the presence of microaneurysms and blot haemorrhage within one disc diameter from the foveal centre or the presence of focal photocoagulation scars in the macular area.

    Assessment of systemic risk factors

    All participants underwent a detailed interview: information on socioeconomic status (eg, education, income), lifestyle risk factors (eg, smoking), medication use and self-reported history of cardiovascular diseases (eg, myocardial infarction, stroke) were collected. Information on clinical factors such as anthropometric measurements (height, weight, body mass index (BMI)) and blood pressure (BP) was also collected. Non-fasting venous blood samples were drawn and sent for analysis of serum lipid levels (total cholesterol, high-density lipoprotein cholesterol and low density lipoprotein cholesterol), HbA1c, creatinine and glucose at the National University Hospital Reference Laboratory on the same day. Cerebrospinal fluid pressure (CSFP) was estimated based on the associations between higher CSFP and younger age, higher BMI and higher diastolic BP. The formula used was as follows: CSFP (mm Hg)=0.44×BMI+0.16×diastolic BP−0.18×age−1.91.

    Statistical analysis

    R language (R V.3.4.2, R Foundation for statistical computing 2016, Vienna, Austria) was used to perform all statistical analyses. The baseline characteristics are presented as mean±SD for continuous variables and as rates (proportions) for categorical variables. Characteristics between groups were compared using t-test for normally distributed continuous variables or Cochran-Mantel-Haenszel test for categorical variables. Age-standardised 6-year incidence and progression of DR were estimated based on the prevalence of diabetes in Singapore and the population distribution according to Singapore Census 2010.18

    The impact of different risk factors on incidence and progression of DR was examined first in a model adjusting for age and gender. Subsequently, multivariable-adjusted regression model was performed with incidence (and progression) of DR as outcome and potential risk factors as exploratory variables. Risk factors for inclusion were either identified previously as risk factors in the literature or on the basis of their significance in the univariate analysis. Relative risk (RR) ratio and their 95% CIs were estimated using modified Poisson regression analysis. We regarded p value of <0.05 from two-sided tests to indicate statistical significance. We calculated the population attributable risk (PAR) of significant risk factors for incidence and progression of DR using Levin’s formula: PAR=Pe*(RR-1)/(Pe*(RR−1)+1) where Pe is the prevalence of risk factor in the study population.19

    Results

    Table 1 compares the characteristics of diabetic subjects who participated (n=759) at the baseline examination and those who did not participate in follow-up examination (n=530). Participants lost to follow-up examination were more likely to be older, had primary/below education, higher prevalence of hypertension and CKD, have longer duration of diabetes, higher use of insulin and higher levels of HbA1c, total and HDL cholesterol and albuminuria at baseline.

    View this table:
    Table 1

    Characteristics of patients with diabetes at baseline examination for participants and non-participants

    Ninety-three of the 501 participants developed any DR over 6-year period giving a crude incidence of 18.56% and age-standardised incidence 21.89% (95% CI 17.09 to 27.78). Among patients with incident DR, minimal DR was the most common stage seen in 52 patients (55.91%), followed by mild DR in 23 (24.73%) and moderate DR in 15 (16.13%), whereas VTDR was identified in three patients (3.23%) only. Thirty-eight patients (40.86%) developed incident DR in both eyes. Higher HbA1c levels, current smoking and higher insulin use were associated with higher incidence, whereas higher BMI and higher estimated CSFP were associated with lower incidence (table 2). The severity of incident DR increased with increase in mean HbA1c levels, from 7.17% to 9.67% (p-trend <0.001) and from 7.49% to 9.34% (p-trend <0.001), at baseline and follow-up visits, respectively. As shown in figure 2, increasing categories of HbA1c at baseline were associated with increased risk within categories of diabetes duration and antidiabetic medication use. Compared with controls (HbA1c<7%), RR estimates of HbA1c ≥8% were higher in duration ≥5 years as well as <5 years and no antidiabetic medication as well as use of antidiabetic medication. The PAR of HbA1c>8% for incidence DR was 41.29%.

    Figure 2
    Figure 2

    Baseline glycated haemoglobin levels and incidence of diabetic retinopathy within subgroups. RR, risk ratio.

    View this table:
    Table 2

    Risk factors associated with incident DR in univariate and multivariate regression analysis (n=501)

    The incidence of CSME and VTDR were 2.82% and 2.47%, respectively, among those who showed some evidence of DR but no CSME and VTDR at baseline. Patients within younger age were associated with significantly higher incident DR when compared with relatively older age (27.6% vs 19.4%, p-trend 0.005) but no association of age was observed with either CSME (p-trend 0.091) or VTDR (p-trend 0.505), after adjusting for glycaemic control. There was significant positive relationship between duration of diabetes and incident VTDR (p-trend 0.001) but not any DR (p-trend 0.096) and CSME (p-trend 0.971) after adjusting for glycaemic control.

    Of the 189 patients included for progression analysis, 56 demonstrated progression (29.6%) over 6-year period. When stratified by severity, 6.9% (n=13) progressed to mild (from questionable at baseline), 15.3% (n=29) to moderate and 7.45 (n=14) progressed to VTDR. The age-standardised progression was 33.45% (95% CI 22.69 to 48.5). Higher HbA1c, higher BMI and higher estimated CSFP were associated with increased DR progression (table 3). Compared with those with no change in DR status (n=30, mean HbA1c 7.92%), baseline mean HbA1c was higher in those who progressed two steps or more (n=56, mean HbA1c 8.94%), followed by those where DR progressed by one step (n=13, mean HbA1c 8.51%). Mean HbA1c was lowest in those whose DR regressed (n=90, mean HbA1c 7.28%) (p-trend <0.001). The PAR of HbA1c >8% for DR progression was 49.22%.

    View this table:
    Table 3

    Risk factors associated with progression of DR in regression models (n=189)

    Discussion

    This is the first population-based cohort study to assess the incidence, progression and risk factors of DR in Asian Indians, a high-risk population for diabetes. The 6-year crude and age-standardised incidence were 18.56% and 21.89% and of progression were 29.6% and 33.45%, respectively. The DR incidence in our population was similar to that reported in Caucasian populations by Blue Mountains Eye Study (BMES) (5-year cumulative incidence 22.2%)8 and UK Prospective Diabetes Study (UKPDS) (6-year incidence 22%)9 as well as in Asian populations (4-year incidence in Hong Kong 20.3%, 5-year incidence in Korea 22%, 5-year incidence in Japan 22%, 4-year incidence in Taiwan 19.2%).20–23 However, DR progression was slightly higher than that reported in Caucasian population, 25.9% in BMES and 29% in UKPDS,8 9 but comparable to those reported in Asian populations (4-year progression in Hong Kong 34.71%, 4-year progression in Taiwan 30%).20 21

    When compared with Indians living in Urban India, DR incidence and progression were found to be higher in Singapore Indians (incidence: 21.89% vs 9.25%, progression 33.45% vs 12.61%).13 This discordance may be due to either shorter follow-up period (4 year vs 6 year) or use of simpler DR grading scale (International Clinical Diabetic Retinopathy Severity Scale) in the latter study and/or identification of more patients with DR through systematic screening programmes available in Singapore. Alternatively, migrant Indians in Singapore experience new lifestyle patterns and dietary habits with calorie-dense/low-fibre foods that make tight diabetes control difficult to attain leading to increased development and progression of DR. However, DR incidence in Singaporean Indians was found to be lower than Beixinjing Chinese community (incidence: 21.89% vs 46.9%) and this difference could probably be attributable to younger age (minimum age 19 years), longer duration of diabetes (50.93% of patient >10 years with longest duration being 30 years) and/or use of fluorescein angiography to confirm DR diagnosis in the latter study.14 Beijing Eye Study reported a very lower DR incidence because the denominator used for estimation of DR incidence was the whole population and not only the diabetic subjects.12

    In this study, risk factors for DR incidence were HbA1c, insulin use, current smoking, lower BMI and lower CSFP, whereas for DR progression, the risk factors were higher HbA1c, BMI and CSFP. Similar to our study, HbA1c was associated with incident DR in Indians living in Urban India but we did not find any significant associations with other risk factors (BP, duration of diabetes, the presence of anaemia and increased cholesterol levels) identified in that cohort (SN-DREAMS II).13 Higher estimated CSFP was associated with higher incident DR in Beijing Eye Study whereas in our study, higher estimated CSFP was associated with higher DR progression (weaker correlation compared with HbA1c and BMI).12 Additional risk factors identified in past studies (baseline BP, duration of diabetes, male gender, the presence of albuminuria and cholesterol levels) were not found to be associated with incidence and/or progression of DR in our study.7 9 13 21

    In the current study, baseline mean HbA1c levels of ≥8% had nearly fourfold higher risk for incident DR independent of diabetes duration and use of antidiabetic medication. The severity of incident DR also increased with increase in mean HbA1c levels at both baseline and follow-up visits. Similarly, baseline HbA1c levels of ≥8% were associated with nearly 3.5-folds higher risk of DR progression. In addition, mild and moderate levels of DR progressed despite mean HbA1c levels showing slight reduction at follow-up visit from baseline visit. This could be attributable to the ‘legacy’ effect or a ‘metabolic memory’ phenomenon, where DR continues to progress even after hyperglycaemic insult is replaced by a normal glyacemic phase.24 Our findings are consistent with UKPDS,25 The Diabetes Control and Complications Trial26 and several other studies conducted among diabetics in Asia.14 20 21 These consistent observations reiterate the notion that it is crucial to optimise the HbA1c level early in the disease process as a long-term management goal to reduce the risk of DR development, incident severity and progression.

    Insulin use was associated with higher incidence of DR and this is in agreement with RETINODIAB (Study Group for Diabetic Retinopathy Screening) programme where insulin use was found to be a significant risk factor for DR development.27 Moreover, a meta-analysis of seven cohort studies reported significant association between insulin use and risk of DR despite major heterogeneity existing in the study.28 The underlying mechanisms by which insulin use increases DR risk remains unknown; however, according to a recently proposed synergistic hypothesis, exogenous insulin could act synergistically with the vascular endothelial growth factor expressed by ischaemic retina so as to trigger vascular proliferation and worsening of retinopathy.

    A positive history of smoking was associated with higher incident DR in this study. The relationship between smoking and DR is not clear and past studies reveal conflicting results. A history of smoking or current smoking status was positively associated with increased risk of DR in few studies.21 29 On the contrary, in WESDR, smoking was not found to be a contributing factor to DR development, and in UKPDS, smoking was even protective.7 9 Moreover, according to a recent met

    a-analysis, smoking was significantly associated with an increased risk of DR in type 1 diabetics but a decreased risk of DR in type 2 diabetics.30 It is important to note that these inconsistent associations between smoking and DR should not change the importance of smoking cessation for public health.

    We also found that patients with diabetes with higher BMI were less likely to have incident DR but more likely to have progression of DR than subjects with lower BMI. While past studies have shown protective effect of greater BMI on incident DR akin to our observation,31 few epidemiological studies have shown an adverse effect of BMI on DR,32 and more recently, a recent meta-analysis failed to show any association between obesity and DR.33 Although the association between obesity and DR remains inconclusive, several reasons may attribute to an inverse association between BMI and DR, such as ethnic differences in the relationship between BMI and DR (positive association in Caucasian and inverse in Asians), obesity paradox (higher BMI have lower risk of DR probably due to better beta cell function) and/or inability of BMI to distinguish between fat and muscle.

    The risk factors identified in our models are unable to fully explain the risk of incidence and progression of DR. Several biomarkers are currently being explored that may provide additional risk information. Ongoing studies on non-invasive biomarkers (such as retinal vessel calibre, retinal vascular fractal dimension and retinal oximetry) with respect to incidence and progression of DR may provide useful information.34 Furthermore, optical coherence tomography angiography (a novel potential biomarker) may provide valuable information if cross-sectional associations with DR can be confirmed in prospective studies.35 In addition, future studies examining potential markers for neuronal degeneration with early DR might enable future novel therapies to prevent onset or delay the progression of DR.36

    The major strength of this study is the population-based cohort that provides an unbiased estimation of incidence and progression estimation of DR. The other strengths are standardised grading of fundus photographs for DR by trained graders in a masked manner along with low number of unreadable photographs and standardised assessment of risk factors. Following are the limitations of this study: assessment of diabetes based on random blood glucose at baseline and follow-up visits might have misclassified diabetes and consequently DR status; DR was graded on two-field non-stereoscopic fundus photography that may miss peripheral lesions (compared with standard seven-field ETDRS photography) and DME (compared with stereoscopic photographs); optical coherence tomography of the macula was not used for assessing DME, and therefore, may underestimate the DR incidence.

    In conclusion, DR incidence and progression in migrant Indians living in Singapore was more than double than that reported in Indians living in urban India. Poor glycaemic control remains the most important predictor for incidence and progression of DR.

    Acknowledgments

    The authors would like to thank Miss Riswana Banu Binte Mohamed Abdul Sokor, Singapore Eye Research Institute, Singapore, for her help with the formatting of the manuscript.

    References

    1. ↵
      2. Cheung N ,
      3. Mitchell P ,
      4. Wong TY
      . Diabetic retinopathy. The Lancet 2010;376:124–36.doi:10.1016/S0140-6736(09)62124-3
      OpenUrlCrossRef
    2. ↵
      2. Wild S ,
      3. Roglic G ,
      4. Green A , et al
      . Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 2004;27:1047–53.doi:10.2337/diacare.27.5.1047
      OpenUrlAbstract/FREE Full Text
    3. ↵
      1. International Diabetes Federation
      . Diabetes atlas. 7th edition. Brussels, Belgium: International Diabetes Federation, 2015.
    4. ↵
      2. Yau JWY ,
      3. Rogers SL ,
      4. Kawasaki R
      . Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care 2012;35:556–64.doi:10.2337/dc11-1909
      OpenUrlAbstract/FREE Full Text
    5. ↵
      2. Lee R ,
      3. Wong TY ,
      4. Sabanayagam C
      . Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss. Eye and Vision 2015;2:17.doi:10.1186/s40662-015-0026-2
      OpenUrl
    6. ↵
      2. Sivaprasad S ,
      3. Gupta B ,
      4. Gulliford MC , et al
      . Ethnic variations in the prevalence of diabetic retinopathy in people with diabetes attending screening in the United Kingdom (drive UK). PLoS ONE 2012;7:e32182.doi:10.1371/journal.pone.0032182
    7. ↵
      2. Klein R1 ,
      3. Klein BE ,
      4. Moss SE
      . The wisconsin epidemiologic study of diabetic retinopathy. X. Four-year incidence and progression of diabetic retinopathy when age at diagnosis is 30 years or more. Arch Ophthalmol 1989;2:244–9.
      OpenUrl
    8. ↵
      2. Cikamatana L ,
      3. Mitchell P ,
      4. Rochtchina E , et al
      . Five-year incidence and progression of diabetic retinopathy in a defined older population: the Blue Mountains Eye study. Eye 2007;21:465–71.doi:10.1038/sj.eye.6702771
      OpenUrlPubMedWeb of Science
    9. ↵
      2. Stratton IM ,
      3. Kohner EM ,
      4. Aldington SJ , et al
      . UKPDS 50: risk factors for incidence and progression of retinopathy in type II diabetes over 6 years from diagnosis. Diabetologia 2001;44:156–63.doi:10.1007/s001250051594
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      2. Broe R ,
      3. Rasmussen ML ,
      4. Frydkjaer-Olsen U , et al
      . The 16-year incidence, progression and regression of diabetic retinopathy in a young population-based Danish cohort with type 1 diabetes mellitus: the Danish cohort of pediatric diabetes 1987 (DCPD1987). Acta Diabetol 2014;51:413–20.doi:10.1007/s00592-013-0527-1
      OpenUrlCrossRefPubMed
    11. ↵
      2. Sabanayagam C ,
      3. Yip W ,
      4. Gupta P , et al
      . Singapore Indian eye Study-2: methodology and impact of migration on systemic and eye outcomes. Clin Exp Ophthalmol 2017;45:779–89.doi:10.1111/ceo.12974
      OpenUrl
    12. ↵
      2. Xu J ,
      3. Xu L ,
      4. Wang YX , et al
      . Ten-year cumulative incidence of diabetic retinopathy. the Beijing eye study 2001/2011. PLoS One 2014;9:e111320.doi:10.1371/journal.pone.0111320
    13. ↵
      2. Raman R ,
      3. Ganesan S ,
      4. Pal SS , et al
      . Incidence and progression of diabetic retinopathy in urban India: Sankara Nethralaya-Diabetic Retinopathy Epidemiology and molecular Genetics Study (SN-DREAMS II), Report 1. Ophthalmic Epidemiol 2017;24:294–302.doi:10.1080/09286586.2017.1290257
      OpenUrl
    14. ↵
      2. Jin P ,
      3. Peng J ,
      4. Zou H , et al
      . The 5-year onset and regression of diabetic retinopathy in Chinese type 2 diabetes patients. PLoS One 2014;9:e113359.doi:10.1371/journal.pone.0113359
    15. ↵
      2. Lavanya R ,
      3. Jeganathan VSE ,
      4. Zheng Y , et al
      . Methodology of the Singapore Indian Chinese cohort (SICC) eye study: quantifying ethnic variations in the epidemiology of eye diseases in Asians. Ophthalmic Epidemiology 2009;16:325–36.doi:10.3109/09286580903144738
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      Grading diabetic retinopathy from stereoscopic color fundus photographs--an extension of the modified airlie house classification. Etdrs report number 10. Early treatment diabetic retinopathy study research group. Ophthalmology 1991;98(5 Suppl):786–806.
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      2. Klein R ,
      3. Klein BE ,
      4. Moss SE , et al
      . The Wisconsin Epidemiologic study of diabetic retinopathy. XIV. Ten-year incidence and progression of diabetic retinopathy. Arch Ophthalmol 1994;112:1217–28.
      OpenUrlCrossRefPubMedWeb of Science
    18. ↵
      1. Census of Population
      . Statistical Release 1. Demographic characteristics, Education, Language and Religion. Department of Statistics, Ministry of Trade and Industry. In: Republic of Singpaore, 2010ISBN 978-981-08-7808-5.
    19. ↵
      2. Rockhill B ,
      3. Newman B ,
      4. Weinberg C
      . Use and misuse of population attributable fractions. Am J Public Health 1998; 88(1): 15-9. Erratum in. Am J Public Health 2008;98:2119.
      OpenUrlCrossRef
    20. ↵
      2. Chen MS ,
      3. Kao CS ,
      4. Fu CC , et al
      . Incidence and progression of diabetic retinopathy among non-insulin-dependent diabetic subjects: a 4-year follow-up. Int J Epidemiol 1995;24:787–95.doi:10.1093/ije/24.4.787
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      2. Tam VH ,
      3. Lam EP ,
      4. Chu BC , et al
      . Incidence and progression of diabetic retinopathy in Hong Kong Chinese with type 2 diabetes mellitus. J Diabetes Complications 2009;23:185–93.doi:10.1016/j.jdiacomp.2008.03.001
      OpenUrlCrossRefPubMed
    22. ↵
      2. Morisaki N ,
      3. Watanabe S ,
      4. Kobayashi J , et al
      . Diabetic control and progression of retinopathy in elderly patients: five-year follow-up study. J Am Geriatric Soc 1994;42:142–5.doi:10.1111/j.1532-5415.1994.tb04941.x
      OpenUrlPubMedWeb of Science
    23. ↵
      2. Kim YJ ,
      3. Kim JG ,
      4. Lee JY , et al
      . Development and progression of diabetic retinopathy and associated risk factors in Korean patients with type 2 diabetes: the experience of a tertiary center. J Korean Med Sci 2014;29:1699–705.doi:10.3346/jkms.2014.29.12.1699
      OpenUrl
    24. ↵
      2. Zhang E ,
      3. Wu Y
      . Metabolic memory: mechanisms and implications for diabetic vasculopathies. Sci China Life Sci 2014;57:845–51.doi:10.1007/s11427-014-4710-6
      OpenUrl
    25. ↵
      2. Stratton IM ,
      3. Adler AI ,
      4. Neil HA , et al
      . Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ 2000;321:405–12.doi:10.1136/bmj.321.7258.405
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. The DCCT Research Group
      . The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin- dependent diabetes mellitus. The Diabetes control and complications trial Research Group. N Engl J Med 1993;329:977–86.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      2. Dutra Medeiros M ,
      3. Mesquita E ,
      4. Gardete-Correia L , et al
      . First incidence and progression study for diabetic retinopathy in Portugal, the RETINODIAB study: evaluation of the screening program for Lisbon region. Ophthalmology 2015;122:2473–81.doi:10.1016/j.ophtha.2015.08.004
      OpenUrlPubMed
    28. ↵
      2. Zhao C ,
      3. Wang W ,
      4. Xu D , et al
      . Insulin and risk of diabetic retinopathy in patients with type 2 diabetes mellitus: data from a meta-analysis of seven cohort studies. Diagnostic Pathology 2014;9:130.doi:10.1186/1746-1596-9-130
      OpenUrl
    29. ↵
      2. Tseng ST ,
      3. Chou ST ,
      4. Low BH , et al
      . Risk factors associated with diabetic retinopathy onset and progression in diabetes patients: a Taiwanese cohort study. Int J Clin Exp Med 2015;8:21507–15.
      OpenUrl
    30. ↵
      2. Cai X ,
      3. Chen Y ,
      4. Yang W , et al
      . The association of smoking and risk of diabetic retinopathy in patients with type 1 and type 2 diabetes: a meta-analysis. Endocrine 2018;62:299–306 https://doi.org/.doi:10.1007/s12020-018-1697-y
      OpenUrl
    31. ↵
      2. Lim LS ,
      3. Tai ES ,
      4. Mitchell P , et al
      . C-reactive protein, body mass index, and diabetic retinopathy. Invest Ophthalmol Vis Sci 2010;51:4458–63.doi:10.1167/iovs.09-4939
      OpenUrlAbstract/FREE Full Text
    32. ↵
      2. van Leiden HA ,
      3. Dekker JM ,
      4. Moll AC , et al
      . Blood pressure, lipids, and obesity are associated with retinopathy: the Hoorn study. Diabetes Care 2002;25:1320–5.doi:10.2337/diacare.25.8.1320
      OpenUrlAbstract/FREE Full Text
    33. ↵
      2. Zhou Y ,
      3. Zhang Y ,
      4. Shi K
      . Body mass index and risk of diabetic retinopathy: a meta-analysis and systematic review. Medicine 2017;96:e6754.
    34. ↵
      2. Blindbaek SL ,
      3. Torp LE ,
      4. Lundberg K
      . Noninvasive retinal markers in diabetic retinopathy: advancing from bench towards bedside. J Diab Res 2017.
    35. ↵
      2. Ting DSW ,
      3. Tan GSW ,
      4. Agrawal R , et al
      . Optical coherence tomographic angiography in type 2 diabetes and diabetic retinopathy. JAMA Ophthalmol 2017;135:306.doi:10.1001/jamaophthalmol.2016.5877
      OpenUrl
    36. ↵
      2. Srinivasan S ,
      3. Dehghani C ,
      4. Pritchard N , et al
      . Corneal and retinal neuronal degeneration in early stages of diabetic retinopathy. Invest Ophthalmol Vis Sci 2017;58:6365–73.doi:10.1167/iovs.17-22736
      OpenUrlCrossRef

    Footnotes

    • NK and MB are joint first authors.

    • Contributors All authors contributed to the intellectual development of this paper. CS designed the study. NGQ and AG performed the statistical analyses. KN wrote the initial and revised draft. MB assisted initial draft. CS supervised data analysis. TYW obtained funding. GT, NC, NT, CW, JJW, PM, EL, CYC, TYW and CS provided critical corrections to the manuscript. Final version of the paper has been seen and approved by all the authors.

    • Funding The study is funded by Singapore Ministry of Health’s National Medical Research Council (NMRC), NMRC/STaR/0003/2008, NMRC/STaR/0016/2013, NMRC/CIRG/1371/2013 and Biomedical Research Council (BMRC), 08/1/35/19/550. The funding organization had no role in the design or conduct of this research.

    • Competing interests None declared.

    • Patient consent for publication Not required.

    • Ethics approval Both studies were approved by the Institutional Review Board of the Singapore National Eye Center.

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

    Linked Articles

    • At a glance
      Highlights from this issue
      Keith Barton Jost B Jonas James Chodosh
      British Journal of Ophthalmology 2019; 103 i-ii Published Online First: 21 Nov 2019. doi: 10.1136/bjophthalmol-2019-315509