Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss

Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss Diabetic retinopathy (DR) is a leading cause of vision-loss globally. Of an estimated 285 million people with diabetes mellitus worldwide, approximately one third have signs of DR and of these, a further one third of DR is vision-threatening DR, including diabetic macular edema (DME). The identification of established modifiable risk factors for DR such as hyperglycemia and hypertension has provided the basis for risk factor control in preventing onset and progression of DR. Additional research investigating novel risk factors has improved our understanding of multiple biological pathways involved in the pathogenesis of DR and DME, especially those involved in inflammation and oxidative stress. Variations in DR prevalence between populations have also sparked interest in genetic studies to identify loci associated with disease susceptibility. In this review, major trends in the prevalence, incidence, progression and regression of DR and DME are explored, and gaps in literature identified. Established and novel risk factors are also extensively reviewed with a focus on landmark studies and updates from the recent literature. Keywords: Diabetic retinopathy, Diabetic macular edema, Epidemiology, Risk factors Introduction vision loss in the highly prevalent type 2 diabetes [4] and Diabetic Retinopathy (DR) is the leading cause of vision is invariably present in patients with type 2 diabetes with loss in adults aged 20–74 years [1]. From 1990–2010, DR PDR [5]. In addition to vision loss, DR and DME have also ranked as the fifth most common cause of preventable been shown to contribute to the development of other blindness and fifth most common cause of moderate to diabetes-related complications including nephropathy, severe visual impairment [2]. In 2010, of an estimated 285 peripheral neuropathy and cardiovascular events [6–9]. million people worldwide with diabetes, over one-third The most clinically important risk factors for progres- have signs of DR, and a third of these are afflicted with sion to vision loss include duration of diabetes, hypergly- vision-threatening diabetic retinopathy (VTDR), defined cemia and hypertension. Control of serum glucose and as severe non-proliferative DR or proliferative DR (PDR) blood pressure have been shown to be effective in pre- or the presence of diabetic macular edema (DME) [3]. venting vision loss due to DR. Prevalence and risk factors These estimates are expected to rise further due to the of DR have been studied widely in previous studies includ- increasing prevalence of diabetes, ageing of the population ing regional and ethnic differences, but epidemiological and increasing of life expectancy of those with diabetes. data on DME are relatively scarce. A review conducted in PDR is the most common vision-threatening lesion 2012 suggested that up to 7 % of people with diabetes may particularly among patients with type 1 diabetes. However, have DME and risk factors of DME are largely similar to DME is responsible for most of the visual loss experienced DR. Recently, new information on the epidemiology of DR by patients with diabetes as it remains the major cause of and DME has been published from both developed and developing countries. In this review, we summarize the prevalence of DR and highlight regional differences in the * Correspondence: charumathi.sabanayagam@seri.com.sg Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, epidemiology of DR from recent studies. We also review Singapore the incidence, progression and regression of DR and Yong Loo Lin School of Medicine, National University of Singapore, DME, as well as factors contributing to the progression or Singapore, Singapore Full list of author information is available at the end of the article regression of DR and DME. © 2015 Lee et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lee et al. Eye and Vision (2015) 2:17 Page 2 of 25 29.6 %, respectively). Of concern is that a large proportion Review of diagnosed DR is vision threatening, with VTDR preva- Prevalence of DR lence estimated to be higher (10.6–17.5 %) than that A pooled individual participant meta-analysis involving 35 observed in the Western world. These observations imply studies conducted worldwide from 1980 to 2008, esti- that most of these cases of DR have been detected late, mated global prevalence of any DR and PDR among when it has already progressed to a vision-threatening patients with diabetes to be 35.4 and 7.5 % respectively stage, or that these populations are particularly susceptible [3]. Prevalence of any DR and PDR was higher in those to severe DR due to ethnic predisposition. Other devel- with type 1 diabetes, compared to those with type 2 oped Asian countries such as Hong Kong [19] and South diabetes (77.3 vs. 25.2 % for any DR, 32.4 vs. 3.0 % for Korea [20] report DR prevalence that is much lower than PDR). Table 1 summarizes the findings of various preva- the global average (12.1 and 15.8 %, respectively). lence studies, organized by region, in comparison to the Apart from the east–west divide, rapidly developing global estimate. Estimates on DR prevalence in type 1 dia- economies in Asia such as China and India are observing betes in Europe and the USA range between 36.5–93.6 %, urban–rural divides in terms of DR disease burden. In with VTDR prevalence estimated between 6.7–34.9 % China, prevalence of DR was reported to be higher among [10–16]. The wide range of prevalence observed may be adults with type 2 diabetes living in rural regions (29.1– due to differences in healthcare systems and socioeco- 43.1 %) [22, 28], compared to their urban counterparts nomic factors between the studied populations, but con- (18.1 %) [22]. Conversely, in a study conducted in Chennai, clusions cannot be made as key characteristics such as India, DR prevalence was reported to be higher in urban known duration of diabetes vary greatly between the sam- (18.0 %) [21] compared to rural areas (10.8 %) [29], pled populations. In the East (Asia and the Middle East), possibly due to the increasing affluence accompanied prevalence studies focused on DR in type 2 diabetes alone, by changes in diet in the urban regions and selective due to the low prevalence of type 1 diabetes in these pop- mortality of those with diabetes-related complications ulations. Hence, comparison of DR prevalence between in rural regions because of poor access to healthcare. the East and West is restricted only to type 2 diabetes. The reason why this urban–rural relationship is re- In general, patients with type 2 diabetes in Western versed in China may represent a case of ethnic pre- communities have a higher prevalence of DR than their disposition, but this is an area that requires further Asian counterparts. In the USA, studies estimate that study. In the past two years, reports on DR prevalence 28.5–40.3 % of patients with type 2 diabetes had DR, and 4.4–8.2 % of them had VTDR [17, 18]. In con- from many developing countries in Asia and Africa have been published [30–35]. Prevalence of DR in Sri Lanka, trast, most Asian countries report DR prevalence to Bangladesh, Nepal, Tunisia, Kenya and Ethiopia ranged be between 12.1–23.0 %, and VTDR prevalence to be be- tween 4.3–4.6 % [19–22]. from 21.6–41.4 %. While the sample sizes of these studies tend to be smaller, they still provide insight into the bur- Singapore is a notable exception to this trend. Despite den of DR in these communities. being an Asian country, paralleling the rapid urbanization, industrialization and internal migration that took place Although duration of diabetes is a major risk factor for DR, a few studies reported DR prevalence in newly over the past five decades in Singapore, DR prevalence in diagnosed diabetes. Prevalence found in these studies Singapore is reported to be higher (33.9 %) than other Asian countries but comparable to that seen in the West- ranged from 2.8 % in South Korea to 28.6 % in Singapore ern world [23]. Within the three major ethnic groups in [20, 27, 32, 36–39]. Surprisingly, a large percentage (19.2 %) of newly diagnosed patients with diabetes have Singapore, the Malays and Indians were reported to have higher a prevalence of DR (33.4 % in Malays, 33.0 % DR in Scotland, UK, where there is universal healthcare. in Indians) compared to the Chinese (25.4 %) [23]. In T his prevalence is even higher than in Nepal (13.0 %) [32], addition to ethnic differences, a study conducted in where access to healthcare is presumably more limited. Singapore also highlighted geographic heterogeneity in However, the prevalence of advanced stages of DR or the prevalence of DR within ethnic Indian groups living in DME was found to be lower among those with newly Singapore (30.4 %) [24] and in urban India (18 %) [21, 25]. diagnosed diabetes suggesting diagnosis of DR early in the It has been speculated that increased acculturation to a course of the disease [40]. westernized lifestyle associated with increased prevalence of obesity and diabetes, and increased awareness among Incidence of DR Indians living in Singapore has led to a higher prevalence, There are few population-based cohort studies, outside of while selective mortality of those with DR in the urban the USA or UK, which have investigated DR incidence. Indian cohorts led to a lower prevalence. In the Middle- Various cohort studies investigating DR incidence over East, Saudi Arabia [26] and Iran [27] both report preva- the past two decades are listed in Table 2. Comparisons lence that are similar to Western communities (36.8 and between the East and West, urban and rural populations, Lee et al. Eye and Vision (2015) 2:17 Page 3 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects Author (Year) Type of study Location Sample size, Diabetes type Prevalence of DR (%) Prevalence of VTDR (%) Age in years Yau (2012) [3] Meta-analysis Global 12,620 Overall 35.36 11.72 (PDR and/or DME) Mean 58.1 Type 1 77.31 38.48 (PDR and/or DME) Range 3–97 Type 2 25.16 6.92 (PDR and/or DME) Asia Liu (2012) [22] Meta-analysis China 11,996 Unspecified 23.0 2.8 (PDR) Range 15–87 Kung (2014) [19] Hospital Hong Kong 15,856 Type 2 12.1 0.3 (PDR) Range ≥ 20 Jee (2013) [202] Population South Korea 1678 Type 2 15.8 4.6 Mean 58.0 ± 11.6 Wong (2008) [39] Population Singapore 757 Type 2 35.0 9.0 Mean 58.7 ± 11.0 Range 40–80 Chiang (2011) [203] Population Singapore 401 Type 2 25.4 Not investigated Mean 53.0 ± 9.0 Range 40–95 Zheng (2012) [204] Population Singapore 1295 Type 2 30.4 7.1 Range ≥ 40 Raman (2009) [21] Population India 1,414 Unspecified 18.0 4.3 Mean 56.1 ± 10.1 Range ≥ 40 Katulanda (2014) [205] Hospital Sri Lanka 536 Unspecified 27.4 Not investigated Mean 56.4 ± 10.9 Range ≥ 18 Akhter (2013) [31] Population Bangladesh 60 Unspecified 21.6 Not investigated Mean 46.0 ± 12 Thapa (2014) [32] Hospital Nepal 277 Unspecified 38.26 14.44 Mean 62.3 ± 13.3 Range ≥ 20 Middle East Al-Rubeaan (2015) [64] Population Saudi Arabia 50,464 Type 2 19.7 10.6 (PDR) 5.7 (DME) (Registry) Mean 59.7 ± 12.8 Range ≥ 25 Al Ghamdi (2012) [26] Population Saudi Arabia 612 Unspecified 36.8 17.5 (Scottish DR Grading R4 and/or M2) Mean 63.3 Range ≥ 50 Papakonstantinou Population Iran 529 Type 2 29.6 11.1 (2015) [27] Range 40–80 Europe Thomas (2015) [10] Population United Kingdom 91,393 Overall 32.4 3.4 (PDR and/or DME) Mean 36.5 ± 16.4 Type 1 56.0 11.2 (PDR and/or DME) Mean 65.3 ± 11.7 Type 2 30.3 2.9 (PDR and/or DME) Lee et al. Eye and Vision (2015) 2:17 Page 4 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects (Continued) Pugliese (2012) [65] Hospital Italy 15,773 Type 2 22.2 9.8 Range 59–75 Pedro (2010) [11] Population Spain 8675 Overall 26.7 0.59 (PDR) Mean 34.9 ± 10.5 Type 1 36.5 1.0 (PDR) 5.73 (DME) Mean 64.6 ± 10.8 Type 2 26.1 0.56 (PDR) 6.44 (DME) Dutra Medeiros Population Portugal 52,739 Type 2 16.3 3.1 (2015) [66] Mean 69.1 ± 11.1 Range ≥ 45 Hautala (2014) [12] Population Finland 172 Type 1 93.6 34.9 (PDR) Mean 30 ± 3 Range 22–35 Bertelsen (2013) [13] Population Norway 514 Overall 26.8 1.2 (PDR) 3.9 (DME) Mean 66.4 Type 1 78.0 Range 46–87 Type 2 25.0 Knudsen (2006) [14] Population Denmark 984 Overall 48.8 Median 37.3 Type 1 53.8 2.9 (PPDR) 5.6 (PDR) 7.9 (CSME) IQR 19.0–48.5 Median 58.1 Type 2 38.7 3.6 (PPDR) 0.9 (PDR) 12.8 (CSME) IQR 15.0–65.0 Dedov (2009) [15] Population Russia 7186 Overall 45.9 8.1 (PPDR) 6.7 (PDR) Median 38.0 Type 1 54.6 9.1 (PPDR) 11.1 (PDR) IQR 27.0–49.0 Median 59.0 Type 2 34.2 7.2 (PPDR) 2.7 (PDR) IQR 54.0–66.0 North America Zhang (2010) [18] Population United States 1006 Unspecified 28.5 4.4 of America Range ≥ 40 Kempen (2004) [17] Pooled population United States 4440 Type 2 40.3 8.2 from 8 studies of America Range ≥ 40 Roy (2004) [16] Pooled population United States 1384 Type 1 79.1 31.2 from 2 studies of America Range ≥ 18 Nathoo (2010) [67] Population Canada 394 Unspecified 27.2 2.3 (PDR) 2.0 (CSME) Mean 58.8 Range 10–100 South America Schellini (2014) [206] Population Brazil 407 Type 2 7.62 Not investigated Mean 51.8 ± 13.6 Range ≥ 30 Esteves (2009) [68] Hospital Brazil 437 Type 1 44.4 3.0 (PPDR) 22.2 (PDR) 9.4 (CSME) Mean 26.8 ± 7.8 Range ≥ 18 Villena (2011) [69] Hospital Peru 1222 Type 2 23.1 1.6 (PPDR) 2.8 (PDR) 2.3 (CSME) Median 59.0 IQR 52.0–67.0 Lee et al. Eye and Vision (2015) 2:17 Page 5 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects (Continued) Africa Thomas (2013) [70] Hospital South Africa 5565 Overall 25.8 7.5 Mean 35.4 ± 15.4 Type 1 36.9 9.7 Mean 56.8 ± 11.8 Type 2 21.4 6.6 Kahloun (2014) [33] Hospital Tunisia 2320 Type 1 and 2 26.3 5.4 (PPDR) 3.4 (PDR) 4.2 (CSME) Mean 54.5 Range 10–92 Mathenge (2014) [34] Population Kenya 195 Unspecified 35.9 13.9 (PPDR + PDR) 4.1 (CSME) Range ≥ 50 Sharew (2013) [35] Hospital Ethiopia 324 Unspecified 41.4 7.3 Oceania Kaidonis (2014) [71] Pooled population Australia 12,666 Type 1 and 2 30.4 7.5 (PDR and/or DME) from 11 studies Range ≥ 15 Papali’i-Curtin Population New Zealand 5647 Unspecified 19.0 0.4 (PDR) (2013) [207] Win Tin (2014) [208] Population Pacific Islands 459 Type 2 47.1 Not investigated (Vanuatu, Nauru, Mean 54 Solomon Islands) DR diabetic retinopathy, VTDR vision-threatening diabetic retinopathy, PDR proliferative diabetic retinopathy, DME diabetic macular edema, PPDR preproliferative diabetic retinopathy, CSME clinically significant macular edema, IQR interquartile range gender and age of diabetes onset. The reduction in risk was and developed versus developing countries are not pos- sible due to the lack of population-based cohort studies in even greater in the cohort diagnosed from 1985 onwards, Asia and many developing countries. In the USA, the at 64 %. Overall, these studies indicate that while almost all patients with type 1 diabetes may eventually develop DR Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) found that among patients with insulin- over time, the incidence of DR and VTDR among patients dependent diabetes with onset before the age of 30, who with type 1 diabetes is probably on the decline. In the UK, population studies involving patients with are presumed to have type 1 diabetes, the 4-year cumula- tive incidence of DR was 59.0 % [41]. At 10, 14 and type 2 diabetes estimated cumulative incidence of DR to 25 years, cumulative incidence of DR in the same cohort be 26.0 % at 4 years [49] 38.1–41.0 % at 6 years [50, 51], and 66 % at 10 years [52]. These findings seem compar- rose to 89.3 % [42], 95.9 % [43], and 97 % [44], respect- ively. Similar observations were made in the Danish able to that found in US population studies, which esti- Cohort of Pediatric Diabetes 1987 (DCPD1987), which mated cumulative incidence of DR to be 22.5–34.0 % at reported a 16-year cumulative incidence of 95.1 % [45]. 4 years [53, 54] and 72.3 % at 14 years [55], despite differ- While these cohorts have long follow-up times, it should ences in ethnicity and age of the cohorts at the time of be noted that the participants were recruited between diabetes diagnosis. Cohorts in Australia [56], Barbados 1979 and 1989. The incidence reported in these studies [57] and Mauritius [58] report cumulative incidence that may not reflect actual DR incidence today, owing to sig- is similar to the UK and US studies. In contrast, the 4-year nificant advancements in retinopathy diagnosis techniques cumulative incidence of DR in a Spanish cohort is much and risk factor management in the past three decades. For lower, estimated at 8.1 % [59]. Age and duration of diabetes example, in a UK cohort recruited between 1991 and are comparable between the US, UK and Spanish studies, 1999, 6-year cumulative incidence of DR in patients with and this significant difference in incidence is attributed to type 1 diabetes was estimated to be only 45.3 % [46]. A unusually good glycemic control within the Spanish cohort, separate UK study, involving only newly diagnosed cases of with mean HbA1c at 7 %, with 55 % of the cohort achieving type 1 diabetes recruited between 2000 and 2007, found 9- HbA1c of less than 7 %. In contrast, patients in one of the year cumulative incidence of DR to be only 23.9 % [47]. In US cohorts [53] had HbA1c of 9.9 % on average. Finland, the incidence of VTDR was reported to be de- As with prevalence, incidence data from Asia is re- creasing in patients with type 1 diabetes [48]. In this study, stricted only to that of type 2 diabetes. A population- patients who were diagnosed with diabetes from 1980 to based study in urban Shanghai, China, found the 5-year 1984 had 47 % reduced risk of VTDR as compared to pa- cumulative incidence to be much higher than in the US tients diagnosed from 1975 to 1979, after adjusting for and UK, at 46.9 %, of which more than a third of it is Lee et al. Eye and Vision (2015) 2:17 Page 6 of 25 Table 2 Incidence of diabetic retinopathy among diabetic subjects Author (Year) Type of study Location Sample size, Diabetes Follow-up Cumulative incidence Incidence of Age in years type time in years of DR, % (95 % CI) VTDR (%) Asia Xu (2014) [75] Population China 2602 Unspecified 10 4.2 (3.45–5.03)** Not investigated Mean 64.6 ± 9.7 Jin (2014) [98] Population China 322 Type 2 5 46.9 13.9 (Severe NPDR) Mean 66.1 ± 13.2 4.6 (PDR) Tam (2009) [60] Hospital Hong Kong 212 Type 2 4 20.3 0.47 (PDR) Mean 55.2 ± 9.5 Song (2011) [61] Hospital Hong Kong 3647 Type 2 4 15.2 0.03 Mean 62.60 ± 9.58 Kawasaki Hospital Japan 1221 Type 2 8 26.6 Not investigated (2011) [62] Mean 58.2 ± 6.9 Kajiwara Hospital Japan 383 Type 2 5.8 ± 2.5 58.5/1000 person Not investigated (2014) [76] (retrospective) years Mean 59.4 ± 11.0 Tsugawa Hospital Japan 1083 Unspecified 3 15.7 Not investigated (2012) [209] Mean 51.0 ± 11.7 Ahmed Hospital Bangladesh 977 Type 2 15 50.6 (47.5–53.8) Not investigated (2012) [210] Mean 41 ± 8 Middle East Manaviat Hospital Iran 120 Type 2 4 47.5 (38.6–56.4) Not investigated (2008) [211] Mean 55.2 ± 9.6 Janghorbani Hospital Iran 549 Type 2 5.1 ± 2.1 45.4 Not investigated (2003) [212] (retrospective) Mean 45.7 ± 9.3 Europe Stratton Population United 1216 Type 2 6 41 Not investigated (2001) [51] Kingdom Mean 52.2 ± 8.5 Younis Population United 305 Type 1 6 45.3 (36.9–53.7) 5.4 (2.3–8.5) (2003) [46] Kingdom Median 30.2 IQR 21.5–39.8 Younis Population United 3743 Type 2 6 38.1 (35.1–41.2) 6.1 (4.4–7.8) (2003) [50] Kingdom Median 63.4 IQR 56.1–69.8 Jones Population United 16,444 Type 2 10 66 18.7 (2012) [52] Kingdom Median 66.7 IQR 58.0–74.5 Martin-Merino Population United 1757 Type 1 9 23.9 4.4 (DME) (2012) [47] Kingdom Mean 19.1 63,226 Type 2 27.8 3.6 (DME) Mean 61.3 Thomas Population United 49,763 Type 2 4 26.0 0.7 (2012) [49] Kingdom Mean 60.2 ± 11.3 Perol Hospital France 236 Unspecified 3 14.0 (9.5–18.4) 0 (2012) [213] Mean 54.0 ± 12.8 Lee et al. Eye and Vision (2015) 2:17 Page 7 of 25 Table 2 Incidence of diabetic retinopathy among diabetic subjects (Continued) Romero-Aroca Hospital Spain 334 Type 1 10 35.9 11.07 (DME) (2011) [77] Mean 25.7 ± 11.7 Salinero-Fort Population Spain 2405 Type 2 4 8.07 (7.04–9.22) 2.8 (PDR) (2013) [59] 1.2 (DME) Mean 67.5 ± 10.6 Henricsson Population Sweden 627 Types 1 and 10 39 1.8 (PDR) (2003) [111] 2 Mean 35.3 ± 5.8 Broe (2014) Population Denmark 185 Type 1 16 95.1 Not investigated [45] Mean 7.5 ± 3.7 North America Lee (1992) [55] Population United States 380 Type 2 14 72.3 15.4 (PDR) (Oklahoma of America Mean 48.2 ± 8.4 12.5 improved Indians) Klein (1998) Population United States 634 Type 1 14 95.9 (93.2–98.6) 26.1 (22.6–29.6) [43] of America (DME) Mean 14.2 ± 7.4 Tudor (1998) Population United States 169 Type 2 4 22.5 Not investigated [53] of America Mean 58.1 Varma (2010) [54] Population United States 775 Unspecified 4 34.0 (30.0–38.0) 5.4 (3.8–7.1) (DME) (Latino) of America Mean 58 ± 9.7 Oceania Cikamatana Population Australia 150 Unspecified 5 22.2 (14.1–32.2) Not investigated (2007) [56] Mean 66.2 ± 8.4 Others Leske Population Barbados 436 Types 1 and 9 39.6 (33.6–45.5) 8.3 (2006) [57] 2 Mean 57.6 ± 9.4 Tapp Population Mauritius 227 Unspecified 6 23.8 0.4 (PDR) (2006) [58] Mean 50 ± 11 DR diabetic retinopathy, VTDR vision-threatening diabetic retinopathy, CI confidence interval, NPDR nonproliferative diabetic retinopathy, PDR proliferative diabetic retinopathy, DME diabetic macular edema **Cumulative incidence of DR among total sample, incidence among participants with diabetes not reported VTDR. This may just be due to differences in known scale, and progression was defined as increase in diabetes duration of the cohorts; the Chinese cohort has severity of 2 steps or more. Some other studies diabetes duration of 11 years on average at baseline assess- assigned DR severity based on the severity grade in ment, while studies in the US and UK report diabetes the worse eye alone. The findings on DR progression duration to be 4 to 7 years on average. More prospective and regression from the various cohort studies are studies are warranted to compare the incidence of DR in summarized in Table 3. Four to six-year cumulative Asia with that observed in Europe or the US. incidence of 2-step progression among the studies ranged from 24.1 to 38.9 %, which increased to 64.1 and 83.1 % in studies with 16-year or 25-year follow-up. Progression and regression of DR In general, progression was much more common than A large number of cohort studies have investigated progres- regression. Two Asian cohort studies, both hospital-based sion and regression of DR [44, 45, 52–54, 56–58, 60–62]. and carried out in Hong Kong, investigated the regression Disease severity was most often classified by the Early of DR. One of the studies found 4-year progression of DR Treatment Diabetic Retinopathy Study (ETDRS) classifica- to be 34.7 % and 4-year regression to be 13.2 % [60], tion for DR severity [63]. The cohort with the longest which is similar to that seen in the population-based US follow-up time was the WESDR cohort, which reported cohorts. However, the other study found 4-year regression 25-year progression of DR in patients with type 1 diabetes to be substantially higher (45.8 %) and progression to be [44]. In this study, DR severity was assigned a level by lower (6.6 %) [61]. This study defined progression or re- concatenating the severity grade in both eyes, with the gression by 1-step change in severity, while most of the worse eye given greater weight. This created a 15-step Lee et al. Eye and Vision (2015) 2:17 Page 8 of 25 Table 3 Progression and regression of diabetic retinopathy Author (Year) Method of severity grading Progression intervals Criteria for progression Progression of Progression from Regression of DR (%) or regression DR (%) NPDR to PDR (%) Asia Tam (2009) [60] Concatenated ETDRS severity Cumulative at 4 years 2-step 34.7 9.9 13.2 of both eyes, with 11 levels Song (2011) [61] Eye with worse ETDRS severity Cumulative at 4 years 1-step 6.6 Not investigated 45.8 Kawasaki (2011) [62] Mild DR to at least severe NPDR according Cumulative at 8 years N/A N/A 15.9 Not investigated to ETDRS Europe Jones (2012) [52] Eye with worse ETDRS severity Cumulative at 5 years Not stated Data irretrievable 6.1 Not investigated Cumulative at 10 years 9.6 Broe (2014) [45] Eye with worse ETDRS severity Cumulative at 16 years 2-step 64.1 31.0 0 North America Tudor (1998) [53] Eye with worse ETDRS severity Cumulative at 4 years 2-step 24.1 Not investigated 13.3 Varma (2010) [54] Concatenated ETDRS severity of both eyes, Cumulative at 4 years 2-step 38.9 5.3 14.0 with 15 levels Klein (2008) [44] Concatenated ETDRS severity of both eyes, Between 0 to 4 years 2-step 13.5 annually 2.5 annually 3.0 annually with 15 levels Between 4 to 10 years 13.0 annually 4.0 annually 0.8 annually Between 10 to 14 years 12.0 annually 2.5 annually 0.4 annually Between 14 to 25 years 2.4 annually 1.5 annually 0.4 annually Cumulative at 25 years 83.1 42.0 17.8 Oceania Cikamatana (2007) [56] Concatenated ETDRS severity of both eyes, Cumulative at 5 years 1-step 25.9 4.1 Not reported with 15 levels Others Leske (2006) [57] Mild or moderate DR to at least severe NPDR Cumulative at 9 years N/A N/A 8.2 Not investigated according to ETDRS Tapp (2006) [58] Mild or moderate DR to at least severe NPDR Cumulative at 6 years N/A 27.7 5.2 Not investigated according to ETDRS DR diabetic retinopathy, NPDR nonproliferative diabetic retinopathy, PDR proliferative diabetic retinopathy, ETDRS Early Treatment for Diabetic Retinopathy Study, N/A not available Lee et al. Eye and Vision (2015) 2:17 Page 9 of 25 other studies defined progression or regression by 2-step populations or a difference in methodology. Of note, clin- ical stereoscopic fundus examination by an ophthalmolo- changes in severity. Moreover, this study was based in a community optometry clinic. Hence, the population sam- gist was carried out in both of these studies and factored ple may be biased towards patients with mild baseline in the diagnosis of DME whereas most studies relied on non-stereoscopic fundus photographs alone, thus raising severity of DR, as patients with more severe disease may the question if prevalence studies using non-stereoscopic have been referred to tertiary hospitals for follow-up. Indeed, 91.7 % of patients with DR at baseline in this study fundus photographs may be severely underdiagnosing DME. In patients with newly diagnosed diabetes, observed had only mild NPDR, and the 1-step regression of prevalence of DME was almost non-existent, with studies mild NPDR to no DR accounted for the majority of the regression observed in this study. The results of reporting it to be within 0 to 0.8 % [21, 39]. A Cochrane review of prevalence of DME assessed by optical coher- this study are hence not directly comparable with that ence tomography (OCT) has found a large range of preva- of the other cohorts, but it highlights the high prob- ability of disease regression in patients with only mild lence rates (19–65 %) [72]. Of note, none of the studies included in the review were population-based studies. NPDR. The absence of data on population-based co- OCT-detected DME was found to have a great degree horts in Asia also precludes direct comparison of pro- of disagreement with the clinical definition of CSME, gression and regression rates between Asian and Western and not all patients who had macular thickening de- populations. tected on OCT progressed to have clinical DME, hence its validity as a diagnostic tool in epidemiologic studies Prevalence of DME is questionable. In most studies, DME was defined by hard exudates in the presence of microaneurysms and blot hemorrhages within one disc diameter of the foveal center. Clinically signifi- Incidence of DME cant macular edema (CSME) is the more severe spectrum Cohort studies that investigated DME incidence are sum- of DME, and was defined by the presence of edema within marized in Table 5. Only studies conducted in the US and 500 μm of the foveal center, or focal photocoagulation Europe investigated DME incidence. The WESDR cohort scars present in the macular area. The prevalence of DME of patients with type 1 diabetes had the longest follow-up among recent cross-sectional studies is summarized in time of 25 years [73]. Interestingly, cumulative incidence Table 4. Among the population-based studies, prevalence of DME and CSME in this cohort seemed to plateau at of DME among patients with type 1 diabetes was between the 14-year mark (DME 26.1 %, CSME 17.0 %), with the 4.2 and 7.9 %. In patients with type 2 diabetes, it was latter 11 years adding minimally to the 25-year cumulative between 1.4 and 12.8 %. Non-stereoscopic fundus photog- incidence (DME 29 %, CSME 17 %). Data available on raphy was used in most studies, which affects the accuracy DME incidence in type 2 diabetes is limited and inconsist- of DME assessment. About half of the studies defined ent [50, 52, 59]. macular edema using the CSME criteria, and hence only the more severe spectrum of DME was captured in these Risk factors for DR and DME studies. Overall, the heterogeneity in methodology causes DR and DME share many common risk factors. comparison of prevalence between these studies to be a Incidence-derived risk factors for DR and DME reported challenge. The prevalence of DME among patients with in the various cohort studies are summarized in Table 6. diabetes is generally much lower than that of DR [11, 13, The major and established risk factors have been reviewed 14, 16–18, 20, 21, 24, 26, 27, 32–35, 39, 64–71]. There extensively before [74]. The most pertinent observations was no observable difference between prevalence of DME will be highlighted again in this review, with updates from between Western or Eastern populations. the latest literature. Novel risk factors were also reviewed. In the Diabetic Retinopathy Screening Service for Wales, a high prevalence of DR (56.0 % in type 1 diabetes, Non-modifiable risk factors 30.3 % in type 2 diabetes) was reported, but the prevalence Duration of diabetes of DME was not found to be higher than other studies Cohort studies with the longest follow-up times found (4.2 % in type 1 diabetes, 1.4 % in type 2 diabetes) [10]. that almost all patients with type 1 diabetes develop some There were a few outliers among the studies that degree of retinopathy if duration of disease exposure is reported exceptionally high prevalence of DME. In Kenya, long enough [44, 45]. This relationship is not as clear in a population-based study found a prevalence of DME of cohort studies on type 2 diabetes, probably due to the 33.3 % among participants with diabetes [34], while a competing risk of mortality in patients with type 2 Canadian study found DME prevalence to be 15.7 %. It is diabetes, who are older and may have more age-related difficult to ascertain if this abnormally - high observed comorbidities. Nevertheless, many studies, both in type 1 prevalence is due to genuinely high prevalence in these and type 2 diabetes [49, 52, 59, 75–77], found disease Lee et al. Eye and Vision (2015) 2:17 Page 10 of 25 Table 4 Prevalence of diabetic macular edema among diabetic subjects Author (Year) Type of study Location Type of Prevalence Definition of macular diabetes (%) edema Yau (2012) [3] Meta-analysis Global Overall 7.48 DME and/or CSME Type 1 14.25 DME and/or CSME Type 2 5.57 DME and/or CSME Xie (2008) [214] Population China Unspecified 4 CSME Jee (2013) [20] Population South Korea Type 2 2.8 DME 1.4 CSME Wong (2008) [39] Population Singapore Type 2 5.7 DME 3.0 CSME Zheng (2012) [24] Population Singapore Type 2 7.2 DME 4.5 CSME Raman (2009) [21] Population India Unspecified 1.4 CSME Thapa (2014) [32] Hospital Nepal Unspecified 5.78 CSME Al-Rubeaan (2015) [64] Population (Registry) Saudi Arabia Type 2 5.7 DME Al Ghamdi (2012) [26] Population Saudi Arabia Unspecified 20.3 Scottish DR Grading M1 15.9 Scottish DR Grading M2 (M2 equivalent to DME) Papakonstantinou (2015) Population Iran Type 2 4.7 CSME [27] Thomas (2015) [10] Population United Kingdom Type 1 4.2 DME Type 2 1.4 DME Pugliese (2012) [65] Hospital Italy Type 2 1.3 DME Pedro (2010) [11] Population Spain Type 1 5.73 CSME Type 2 6.44 CSME Dutra Medeiros (2015) [66] Population Portugal Type 2 1.4 DME Bertelsen (2013) [13] Population Norway Types 1 and 2 3.9 DME Knudsen (2006) [14] Population Denmark Type 1 7.9 CSME Type 2 12.8 CSME Zhang (2010) [18] Population United States of Unspecified 2.7 CSME America Varma (2014) [215] Population United States of Unspecified 3.8 DME America Petrella (2012) [216] Population (registry) Canada Type 1 and 2 15.7 DME Nathoo (2010) [67] Population Canada Unspecified 2.0 CSME Esteves (2009) [68] Hospital Brazil Type 1 9.4 CSME Villena (2011) [69] Hospital Peru Type 2 2.3 CSME Thomas (2013) [70] Hospital South Africa Type 1 and 2 3.2 DME Kahloun (2014) [33] Hospital Tunisia Type 1 and 2 8.7 DME 4.2 CSME Mathenge (2014) [34] Population Kenya Unspecified 33.3 DME 4.1 CSME Sharew (2013) [35] Hospital Ethiopia Unspecified 6.0 CSME Kaidonis (2014) [71] Pooled population from 11 Australia Types 1 and 2 7.6 DME studies DR diabetic retinopathy, DME diabetic macular edema, CSME clinically significant macular edema Lee et al. Eye and Vision (2015) 2:17 Page 11 of 25 Table 5 Incidence of diabetic macular edema among diabetic subjects Author (Year) Type of study Location Type of diabetes Follow-up time Cumulative incidence, Definition of macular in years % (95 % CI) edema Younis (2003) [46] Population United Kingdom Type 1 6 3.2 (0.8–5.7) DME Younis (2003) [50] Population United Kingdom Type 2 6 6.1 (4.4–7.8) DME Jones (2012) [52] Population United Kingdom Type 2 10 1.5 DME Martin-Merino Population United Kingdom Type 1 9 4.4 DME (2012) [47] Type 2 3.6 DME Thomas (2012) [49] Population United Kingdom Type 2 4 1.4 DME Perol (2012) [213] Hospital France Unspecified 3 0 DME Romero-Aroca Hospital Spain Type 1 10 11.07 DME (2011) [77] Salinero-Fort Population Spain Type 2 4 0.01 DME (2013) [59] Klein (1998) [43] Population United States of America Type 1 14 26.1 (22.6–29.6) DME 17.0 (14.1–19.9) CSME Klein (2009) [73] Population United States of America Type 1 25 29 DME 17 CSME Varma (2010) [54] Population (Latino) United States of America Unspecified 4 5.4 (3.8–7.1) DME exclusive of CSME 7.2 (5.2–9.1) CSME Leske (2006) [57] Population Barbados Types 1 and 2 9 8.7 (5.4–12.0) CSME DR diabetic retinopathy, DME diabetic macular edema, CSME clinically significant macular edema, CI confidence interval duration to be a significant risk factor for DR, and this is times as likely to occur in mothers with type 1 diabetes as independent of adequacy of glycemic control. mothers with type 2 diabetes (31.3 vs. 11.7 %, p = 0.001) [82]. This progression is often transient and accompanied Puberty and pregnancy by rapid regression of DR in the postpartum period. At Puberty is a well-known risk factor for DR in type 1 dia- the end of 6.5 years of follow-up on average, prevalence betes. Pre-pubertal years of diabetes exposure contributes and severity of retinopathy was comparable between women with pregnancies and women without pregnancies to added risk of DR [78, 79], but it seems that it is disease exposure during puberty itself, when the body is undergo- [83]. Possible mechanisms behind the progression of DR ing rapid development and maturation, that has the in pregnancy include both hormonal and immune theories [84, 85]. greater impact on the risk of DR. In Finland, the FinnDiane Study Group found that onset of diabetes dur- ing pubertal or post-pubertal age increases risk of devel- Modifiable risk factors oping severe retinopathy requiring laser treatment when Hyperglycemia compared to patients with pre-pubertal onset of diabetes Hyperglycemia is one of the most important risk factors [80]. This was particularly significant among the male for DR and DME. A meta-analysis of three large participants. Biological pathways that may contribute to population-based studies found a graded relationship this phenomenon include the transforming growth factor between the level of glycemia and frequency of retinop- beta (TGF-β) signaling pathway, which is an important athy signs [86]. The United Kingdom Prospective Diabetes mediator of renal microvascular damage [81]. Androgens Study (UKPDS) and the Diabetes Control and Complica- promote and accelerate TGF-β transcriptional activity, tions Trial (DCCT) provided strong evidence that tight which can explain the male preponderance. However, control of glycemia (HbA1c <7 %) reduces the risk of evidence of activation of similar pathways in retinal vessels development and progression of DR in both type 1 and is lacking. type 2 diabetes [87]. The DCCT showed that intensive DR and DME can progress rapidly during pregnancy, glycemic control reduced the incidence of retinopathy by especially in patients with type 1 diabetes. A recent study 76 % and progression from early to advanced retinopathy found progression of DR in pregnancy to be almost 3 by 54 % [88]. This highlights that strict glycemic control is Lee et al. Eye and Vision (2015) 2:17 Page 12 of 25 Table 6 Incidence-derived risk factors for the development of diabetic retinopathy in cohort studies Risk factor Author (Year) Strength of association Age Xu (2014) [75] OR (95 % CI) = 1.00 (0.98–1.02) per year increase Ahmed (2012) [210] HR (95 % CI) = 1.29 (1.07–1.58) per year Janghorbani (2003) [212] HR (95 % CI) = 1.03 (1.006–1.04) per year increase Jones (2012) [52]Comparedto40–70 years, HR (95 % CI) = 1.49 (1.09–2.05) for < 40 years; 1.26 (1.00–1.27) for > 70 years Gender Xu (2014) [75] OR (95 % CI) = 1.32 (0.88–1.96) *reference gender not reported Kajiwara (2014) [76] HR (95 % CI) = 1.85 (1.06–3.24) for female Ahmed (2012) [210] HR (95 % CI) = 1.08 (0.91–1.29) *reference gender not specified Smoking Stratton (2001) [51] OR (95 % CI) = 0.63 (0.48–0.82) if current smoker Duration of diabetes Xu (2014) [75] OR (95 % CI) = 1.16 (1.10–1.22) per year increase Kajiwara (2014) [76] OR (95 % CI) = 1.13 (1.09–1.17) per year increase Romero-Aroca (2011) [77] OR (95 % CI) = 8.90 (4.83–17.4) for ≤ 15 years vs. > 15 years Jones (2012) [52] Compared to < 10 years, HR (95 % CI) = 1.21 (1.01–1.44) for 10–20 years; 0.93 (0.68–1.26) if ≥ 20 years Thomas (2012) [49] Compared to < 5 years, HR (95 % CI) = 1.29 (1.23–1.34) for 5–9 years; 1.68 (1.59–1.77) for 10 years Salinero-Fort (2013) [59] Compared to < 6 years, HR (95 % CI) = 1.22 (0.88–1.70) for 7–14 years; 1.64 (1.05–2.57) for 15–22 years; 2.00 (1.18–3.39) for 22 years HbA1C Xu (2014) [75] OR (95 % CI) = 1.73 (1.35–2.21) per 1 % increase Jin (2014) [98] OR (95 % CI) = 1.12 (1.01–1.24) per 1 % increase Tam (2009) [60] OR (95 % CI) = 1.57 (1.23–2.00) per 1 % increase Kajiwara (2014) [76] OR (95 % CI) = 1.21 (1.08–1.36) per 1 % increase Stratton (2001) [51] Compared to HbA1C < 6.2 %, OR (95 % CI) = 1.4 (1.1–1.8) for 6.2–7.4 %; 2.5 (2.0–3.2) for > 7.4 % Romero-Aroca (2011) [77] OR (95 % CI) = 4.01 (1.91–8.39) if > 7.0 % vs. ≤ 7.0 % Tudor (1998) [53] OR (95 % CI) = 1.50 (0.96–2.36) per 2 % increase Kajiwara (2014) [76] HR (95 % CI) = 1.33 (1.18–1.51) per 1 % increase Janghorbani (2003) [212] HR (95 % CI) = 1.08 (1.007–1.15) per 1 % increase Salinero-Fort (2013) [59] Compared to HbA1C < 7 % HR (95 % CI) = 1.39 (1.01–1.92) for 7–8 %; 1.90 (1.30–2.77) for > 8 % Henricsson (2003) [111] HR (95 % CI) = 1.7 (1.43–1.93) per 1 % increase Use of insulin/diabetes treatment Tudor (1998) [53] OR (95 % CI) = 2.00 (0.75–5.35) if on oral treatment vs. no medications OR (95 % CI) = 9.30 (2.69–32.16) if on insulin vs. no medications Jones (2012) [52] Compared to diet control only, HR (95 % CI) = 1.77 (1.44–2.17) if oral hypoglycemics only HR (95 % CI) = 2.17 (1.68–2.81) if using insulin Thomas (2012) [49] Compared to diet control only, HR (95 % CI) = 1.41 (1.36–1.47) if oral hypoglycemics only HR (95 % CI) = 2.03 (1.89–2.18) if using insulin Blood pressure Jin (2014) [98] OR (95 % CI) = 1.80 (1.14–2.86) if SBP > 140 mmHg and/or DBP > 90 mmHg Kajiwara (2014) [76] OR (95 % CI) = 1.02 (1.01–1.03) per mmHg increase in SBP Stratton (2001) [51] Compared to < 125 mmHg, OR (95 % CI) = 1.5 (1.2–2.6) for SBP was 125–139 mmHg; 2.8 (2.2–3.5) if SBP was ≥ 140 mmHg Romero-Aroca (2011) [77] OR (95 % CI) = 3.31 (1.62–6.75) if SBP > 140 mmHg and/or DBP > 90 mmHg Tudor (1998) [53] OR (95 % CI) = 1.81 (1.02–3.20) per 20 mmHg increase in SBP Lee et al. Eye and Vision (2015) 2:17 Page 13 of 25 Table 6 Incidence-derived risk factors for the development of diabetic retinopathy in cohort studies (Continued) Kajiwara (2014) [76] HR (95 % CI) = 1.01 (0.99–1.03) per mmHg increase in SBP Jones (2012) [52] HR (95 % CI) = 0.72 (0.64–0.81) if on anti-hypertensive medications Obesity Kajiwara (2014) [76] OR (95 % CI) = 1.07 (1.01–1.13) per kg/m increase in BMI Kajiwara (2014) [76] HR (95 % CI) = 1.16 (1.06–1.26) per kg/m increase in BMI Henricsson (2003) [111] HR (95 % CI) = 1.11 (1.04–1.18) per kg/m increase in BMI Nephropathy Xu (2014) [75] OR (95 % CI) = 1.01 (1.002–1.022) per mmol/L increase in serum creatinine concentration Axial Length Xu (2014) [75] OR (95 % CI) = 0.48 (0.33–0.71) per mm increase Cerebrospinal fluid pressure Xu (2014) [75] OR (95 % CI) = 1.10 (1.01–1.21) per mmHg increase Fasting blood glucose Janghorbani (2003) [212] HR (95 % CI) = 1.003 (1.0003–1.005) per mg/dL increase in fasting blood glucose Cholesterol Salinero-Fort (2013) [59] Compared to < 100 mg/dL, HR (95 % CI) = 0.87 (0.65–1.16) for LDL 100–190 mg/dL; 7.91 (3.39–18.47) for LDL > 190 mg/dL Aspirin use Salinero-Fort (2013) [59] HR (95 % CI) = 1.65 (1.22–2.24) if patient takes aspirin Risk factors for DME Incidence Risk factor Author (Year) Strength of association Duration of diabetes Romero-Aroca (2011) [77] OR (95 % CI) = 8.921 (4.321–26.773) if > 15 years of diabetes duration HbA1c Romero-Aroca (2011) [77] OR (95 % CI) = 3.121 (1.823–10.332) if HbA1c is > 7.0 % Blood pressure Klein (2009) [73] HR (95 % CI) = 1.17 (1.10–1.25) per 1 % increase Romero-Aroca (2011) [77] OR (95 % CI) = 3.115 (0.907–10.70) if SBP > 140 mmHg and/or DBP > 90 mmHg Klein (2009) [73] HR (95 % CI) = 1.15 (1.04–1.26) for every 10 mmHg increase in SBP Nephropathy Romero-Aroca (2011) [77] OR (95 % CI) = 6.774 (3.442–18.236) if protein excretion > 200 μg/min or > 300 μg/mg of albumin: creatinine ratio Klein (2009) [73] HR (95 % CI) = 1.43 (0.99–2.08) if urine protein concentration ≥ 30 mg/dL Cholesterol Romero-Aroca (2011) [77] OR (95 % CI) = 4.125 (1.125–15.857) if Total cholesterol/HDL-cholesterol ratio is > 3.5 in men and > 3.0 in women OR odds ratio, HR hazard ratios, CI confidence interval, LDL low density lipoprotein, HDL high density lipoprotein, SBP systolic blood pressure, DBP diastolic blood pressure much more effective in preventing or delaying the onset Glycemic control should be achieved early in the disease of DR in patients with diabetes without DR, rather than course and maintained for as long as possible, since its limiting the severity of DR after it has occurred. In the protective effect is sustained even if tight glycemic control case of DME, intensive glycemic control was associated is lost. This is the metabolic memory effect observed after with 46 % reduction in the incidence of DME at the end the DCCT. Within a year after the end of DCCT, the gly- of the trial and a 58 % reduction 4 years later compared cemic control in the conventional group and intensive with those in the conventional group [89]. The burden of control group had converged, but the participants in the primary prevention of DR and DME hence falls heavily on intensive control group still had lower prevalence of DR primary care physicians, who are in the best position to and DME than the participants in the conventional con- achieve good glycemic control in patients who have not trol group at 10 years after DCCT [91]. Risk reduction in developed complications. In everyday clinical care how- the intensive control group was 52 % between years 1 to ever, it is difficult to replicate the intensity of glycemic 10 after DCCT, but dwindled to 12 % between years 11 to control seen in these studies that were achieved under 18 [92]. This implies that the metabolic memory effect trial conditions. From the findings reported by the DCCT, fades with time, but this is confounded by improved gly- intensive glycemic control actually increases risk of pro- cemic control and risk reduction in the conventional con- gression of existing DR in the first year of treatment [90]. trol group since the end of DCCT. Besides implications for However, this should not deter achieving tight glycemic clinical treatment, metabolic memory also has implications control in patients with existing DR, as the long-term pro- on methodology of diabetes research, seeing that obtaining gression risk reduction outweighs that of the increased mean HbA1c of the entire course of diabetes may be risk in the first year alone. needed to control for the effect of metabolic memory [93]. Lee et al. Eye and Vision (2015) 2:17 Page 14 of 25 Apart from the absolute value of glycemia alone, the or more steps by 35 % in type 1 diabetes, and increased short-term variability of glycemia, such as spikes in post- regression of retinopathy by 34 % in type 2 diabetes prandial glucose, is found to be associated with increased [101, 102]. However, regression only occurred in mild risk of microvascular complications [94]. However, there DR, and candesartan had no effect on incidence or pro- is insufficient data at this point to conclude that fluctua- gression of DME. In the RASS, enalapril and losartan re- tions in blood sugar levels is a causative factor in micro- duced the risk of retinopathy progression by 65 and 70 %, vascular complications considering increased glycemic respectively. Since it was observed that this effect was in- fluctuation can be due to a multitude of correlated factors dependent of blood pressure changes across the period of that may all contribute to microvascular injury, such as the trial, it was proposed that DR risk reduction was not severity of disease or poor compliance. mediated by an effect on hypertension. A recently published Cochrane review concluded that The benefits of achieving euglycemia should be bal- anced with the risk of hypoglycemia, especially in the eld- intensive blood pressure control had a modest effect in erly. In both the Action in Diabetes and Vascular Disease reducing incidence of DR, but does not reduce risk of pro- gression [103]. Insufficient evidence on adverse effects of (ADVANCE) [95] and Action to Control Cardiovascular Risk in Diabetes (ACCORD) [96] trials, aggressive gly- strict blood pressure control in patients with diabetes cemic control (HbA1c <6.5 %) did not significantly reduce made a cost-benefit analysis impossible in the review, and both clinicians and researchers should be aware of this risk of retinopathy development or progression in type 2 diabetes. In ACCORD, it was found that such an aggres- gap in literature. Hence, the overall recommendation is to sive manner of glycemic control may in fact be associated avoid intensive blood pressure control for the sole purpose of slowing DR progression. Instead, control of hyperten- with increased mortality, but it was not ascertained sion in a patient with diabetes should be focused on whether this was directly due to metabolic complications of treatment, such as hypoglycemia. Current institution preventing or limiting progression of other vascular com- plications, particularly nephropathy, as well as lowering guidelines state that treatment goals of hyperglycemia are mortality. There is insufficient evidence for the use of to be anywhere between <6.5 to <7.5 % of HbA1c. Accord- ing to a recently published Cochrane review [97] however, RAS targeting anti-hypertensive medication specifically for preventing or treating retinopathy. there is no concrete evidence on any specific treatment target. Instead, the authors recommend that clinicians set individualized treatment goals based on age, disease pro- Dyslipidemia gression, risk of hypoglycemic episodes, and psychological As outlined in a previous review, the evidence for dyslipid- factors of the patient. emia as a risk factor for DR are inconsistent, and no single lipid measure had been consistently found to be associated Hypertension with DR or DME [74]. In recent cohort studies, only the Multiple epidemiologic studies have identified hyperten- Madrid Diabetes Study found an association between low sion as a risk factor for DR and DME [51, 53, 76, 77, 98]. density lipoprotein (LDL) cholesterol and incidence of DR In the UKPDS, tight blood pressure control (defined as [59]. Moreover, a meta-analysis found that there was a target blood pressure <150/85 mmHg) in patients with dose-dependent relationship of statin use with increasing type 2 diabetes reduced the risk of microvascular disease risk of diabetes [104]. It was then believed that statins by 37 %, the rate of progression of DR by 34 %, and the might have effects on glucose homeostasis, such as risk of deterioration of visual acuity by 47 % [99]. Unlike decreasing insulin production or increasing insulin resist- in the case of hyperglycemia, the protective effect of blood ance, or both [105]. Therefore, while the use of statins is pressure control waned quickly upon stopping intensive first-line treatment for dyslipidemia in the prevention of control [100]. Anti-hypertensive medications targeting the cardiovascular events in patients with diabetes, the evi- renin-angiotensin-aldosterone system (RAAS) are now the dence for intensive control by statins for the purposes of first line treatment for control of hypertension in patients treating DR and DME are lacking. with nephropathy as it was found that they had additional Fenofibrate, a peroxisome proliferator-activated receptor beneficial effects independent of their absolute hypotensive alpha (PPARα) agonist, has gathered interest on its effects action. Since retinopathy and diabetic nephropathy are re- on DR and DME. In an ancillary study of the Fenofibrate lated microvascular complications, clinical trials such as Intervention and Event-lowering in Diabetes (FIELD) the Diabetic Retinopathy Candesartan Trials (DIRECT) cohort, participants treated with fenofibrate had a 31 % re- and Renin-Angiotensin System Study (RASS) measured duced risk of requiring laser treatment for PDR or DME, the beneficial effects these classes of anti-hypertensive compared to placebo [106]. However, 2-step progression medications had on DR and DME. Candesartan was found of retinopathy did not differ significantly between the fenofibrate and placebo group, except for the subgroup to reduce the incidence of retinopathy by two or more steps in severity on the ETDRS scale by 18 % or by three with pre-existing DR. In this subgroup, risk of 2-step Lee et al. Eye and Vision (2015) 2:17 Page 15 of 25 progression was almost a fifth of that compared to pla- Asians with type 2 diabetesis yet to be confirmed by a cebo. Moreover, in a more recent trial by the ACCORD cohort study. group, adjunct fenofibrate with simvastatin compared Closely related to obesity is the study of obstructive to simvastatin alone reduced the rate of progression of sleep apnea (OSA) as a potential risk factor for DR and DME. A cross-sectional study in patients with type 2 DR (6.5 vs. 10.2 %, respectively) by at least 3 steps at 4 years [107]. Fenofibrate treatment may also have benefi- diabetes found that OSA was associated with DR severity, cial effects on DME, as it was found to have a moderate but not DME [115]. A separate study on patients with CSME found high prevalence of sleep-disordered breath- effect in decreasing macular volume in patients with DME [108]. The sample size of this study however, was relatively ing in these patients, but severity of sleep-disordered small, and more studies are required to study this associ- breathing was not correlated with severity of DR or DME in this study [116]. However, the sample sizes of these ation. Given the current evidence, it is found that patients with DR benefit most from fibrate therapy if they have studies were too small to draw any concrete conclusions. hypertriglyceridemia and low serum high density lipopro- Bariatric surgery is a highly effective treatment for mor- bid obesity that achieves glycemic control of diabetes tein (HDL)-cholesterol, and hence treatment can be justi- rapidly. However, much like how intensive glucose control fied in this subset of patients, with the hopes of slowing progression to PDR. However, generalization of fibrate with medications or insulin increases risk of DR progres- sion in the short-run, this rapid improvement in glycemic treatment to all patients with diabetes at risk of DR is not control post-bariatric surgery has been associated with recommended without stronger evidence [109]. progression of DR. Most studies presented in this area are case series, and a recent meta-analysis of these studies Obesity found that patients with pre-existing DR are 2.77 times The effect obesity has on DR has been relatively well- studied but with inconclusive and conflicting findings (95 % CI 1.10–6.99) more likely to have adverse outcomes in DR post-operatively than patients without pre-existing [110]. It may be possible that obesity has differing im- DR [117]. As mentioned earlier, increased risk of progres- pacts on DR in type 1 diabetes as compared to type 2 sion with intensive glycemic control occurred only in the diabetes. In the Diabetic Incidence Study in Sweden in- first year of follow-up, with subsequent risk reduction with volving predominantly participants with type 1 dia- longer-term control [90]. It remains to be seen if this is betes, it was found that risk of developing DR the case with bariatric surgery as well, as no studies had increased by 1.11 (95 % confidence interval (CI) 1.04– sufficient follow-up time to determine if bariatric surgery 1.18) times per kg/m increase in Body Mass Index (BMI) has long-term benefits on DR. after 10 years of follow-up [111]. In the EURODIAB Prospective Complications Study, also involving patients with type 1 diabetes, larger waist to hip ratio was asso- Novel risk factors ciated with incidence of DR after more than 7 years of Inflammation follow-up [112]. Retinal and vitreous inflammation was observed in sub- In contrast, many studies in type 2 diabetes, performed jects with diabetes, both in animal models and human primarily in Asia, found an inverse relationship between studies. The role of inflammation in DR and DME is obesity and DR. In a cross-sectional study of the Shanghai therefore an area of extensive study, and has been Diabetes Registry Database, participants who were over- reviewed previously [118]. As pointed out in the review weight had reduced risk of DR and VTDR [113]. A similar however, current data suggests systemic inflammation study on the multi-ethnic population in Singapore found cannot account for the characteristic lesions seen in DR the same risk reduction in obese patients for DR, VTDR and DME. Many conditions can lead to systemic inflam- and CSME [114]. mation (e.g. sepsis, autoimmune disease), but DR-like The exact mechanisms underlying this discrepancy be- lesions and DME are not seen in these diseases. Hence, it tween type 1 and type 2 diabetes are not well understood. seems that the local retinal inflammation seen in subjects It was postulated that unintentional weight loss is a sign with diabetes is not related to systemic inflammation. This of advanced and severe type 2 diabetes, hence the obser- challenges the validity of investigating systemic inflamma- vation of non-obese patients with type 2 diabetes being at tory markers such as serum C-reactive protein (CRP), higher risk of DR. In contrast, obesity and metabolic syn- interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) drome do not contribute to the etiology of type 1 diabetes, as risk factors for DR or DME. Indeed, inconsistencies in which is autoimmune in nature, and obese patients with the association between systemic inflammatory markers type 1 diabetes may simply have more difficulties achiev- and risk of DR and DME exist in the current literature. ing good glycemic control. It should be noted that there The EURODIAB Prospective Complications Study found are no prospective population-based studies in Asia on an association between CRP, IL-6, TNF-α and presence of DR incidence, and the protective effect of obesity in DR in subjects with type 1 diabetes via a cross-sectional Lee et al. Eye and Vision (2015) 2:17 Page 16 of 25 study [119]. Other cross-sectional studies found no such Metabolic hormones association. The Multi-ethnic Study of Atherosclerosis did Hormones involved in metabolism have been hypothe- not find an association between CRP and DR or VTDR sized to play key roles in the pathogenesis of microvascu- (which includes DME), but found an association between lar complications in diabetes, due to their roles in both fibrinogen, an acute-phase reactant in systemic inflamma- metabolic and inflammatory pathways [132]. In particular, tion, and DR and VTDR [120]. The Singapore Malay Eye leptin and adiponectin, which are actively secreted by Study even found that raised CRP was associated with a adipocytes to regulate energy balance in the body, have lower prevalence of DR [121]. None of the studies found been implicated as potential risk factors. an association between systemic inflammatory markers Leptin may play a role in inciting inflammation. Leptin and DME specifically. was found to cause upregulation of VEGF in retinal Local retinal inflammation forms the basis of intra- pericytes [133], hence stimulating angiogenesis in the is- venous administration of corticosteroids. The Diabetic chemic retina [134], and possibly contributing to the neo- Retinopathy Clinical Research Network (DRCR.net) vascularization seen in PDR. Elevated serum and vitreous compared intravitreal triamcinolone versus focal/grid leptin was observed in patients with diabetes, and vitreous laser photocoagulation in patients with DME. The leptin was especially elevated in patients with PDR [135]. findings showed that the triamcinolone group had However, cross-sectional studies could not find an associ- better visual acuity at the 4-month interval, but ation between elevated serum leptin and DR [136, 137], equivalent visual acuity at the 1-year interval. At the though it should be noted that the sample sizes of 2-year [122] and 3-year interval [123], mean visual these studies were relatively small and they may be acuity was better in the photocoagulation than the underpowered. triamcinolone groups. Hence, corticosteroid treatment Adiponectin has been found to induce dilation of retinal for DME is effective, but the effect is transient. Clini- arterioles via upregulation of endothelial cell nitric oxide cians also have to be cautious with adverse effects production, in animal studies [138]. Studies by the same such as elevated intraocular pressure and cataract group in human subjects with mild DR found that serum formation. adiponectin was positively correlated with retinal blood flow velocity and negatively correlated with retinal arterial Vascular endothelial growth factor (VEGF) is a key modulator of angiogenesis and vascular permeability, and resistance [139]. Hence, adiponectin may have a role in is upregulated by inflammatory cytokines [124]. Anti- countering ischemia by promoting reperfusion in the VEGF agents have been used successfully for the treat- ischemic retina. In vitro studies also found that it downre- ment of both PDR and DME [125, 126]. Ranibizumab, an gulates VEGF and thus may have anti-angiogenic proper- anti-VEGF agent, was more effective than laser therapy in ties [140]. Large population-based cross-sectional studies restoring vision for DME [127], although just like with found that elevated serum adiponectin in patients with corticosteroids, ranibizumab is associated with elevations DR correlated with severity of DR when compared to in intraocular pressure [128]. In recent reports, the patients without DR [141, 142]. However, there are incon- DRCR.net compared outcomes in DME treated by afliber- sistencies in literature, with one study finding decreased cept, bevacizumab or ranibizumab, and found that afliber- serum adiponectin in participants with PDR [143]. Given cept provided superior visual recovery if baseline visual that basic science suggests adiponectin as mainly protect- acuity was poorer than 69 ETDRS letters (approximately ive against the development of microvascular complica- 6/15 Snellen) when compared to the other anti-VEGF tions, the observation that serum adiponectin is elevated agents, but there was no significant difference between in patients with severe DR appears contradictory. It may aflibercept and the other anti-VEGF agents if baseline be that upregulation of adiponectin secretion can be at- visual acuity was better than 69 letters [129]. tributed to a natural response that ameliorates the effects Anti-VEGF agents appear superior to corticosteroids in of severe microvascular disease, but prospective cohort terms of efficacy. DRCR.net compared ranibizumab and studies are needed to establish the temporal link between concurrent photocoagulation against triamcinolone with adiponectin levels and the development and progression photocoagulation in patients with DME, and found that of DR. Overall, it appears research in adiponectin has ranibizumab achieved better visual outcome at 1-year produced more promising and consistent results than follow-up than triamcinolone, except in a subset of leptin. The association between these hormones and DME patients with pseudophakic eyes [130]. In this subset of has yet to be studied. participants, triamcinolone achieved comparable visual outcome when compared with ranibizumab, possibly Oxidative stress because of the removed effect of steroid-induced cataract Oxidative stress is the accumulation of free radicals in the formation in pseudophakic eyes. Consistent results were form of reactive oxygen species (ROS). Highly efficient obtained at 2-year follow-up [131]. physiological mechanisms consisting of endogenous free Lee et al. Eye and Vision (2015) 2:17 Page 17 of 25 suggesting that serum LPO may be a suitable proxy meas- radical scavengers usually keep oxidative stress low. How- ure of DR severity [154]. More studies will be needed to ever, under pathological conditions, ROS production may confirm this association. be increased such that the defensive mechanisms are over- whelmed, or the protective mechanisms themselves may be impaired, or both [144]. Oxidative stress has been Vitamin D linked to the histopathological changes of DR, such as On top of its well-known effects on calcium metabolism, retinal basement membrane thickening [145] and capillary Vitamin D has anti-angiogenic and anti-inflammatory cell loss [146]. Increased ROS and decreased antioxidant effects that have implicated Vitamin D deficiency in the potential has also been found in patients with diabetes, es- pathogenesis of various types of pathology, such as malig- pecially if they have DR [147]. The effects of oxidative nancy, autoimmune disease, cardiovascular disease and stress are observed early in the course of diabetes, and its diabetes [156]. effects on microvasculature persist even if hyperglycemia It is thus intuitive that Vitamin D has a protective ef- is subsequently corrected. Hence, oxidative stress is likely fect on DR and DME, since anti-angiogenesis may slow to be the mechanism behind the “metabolic memory” progression to PDR and anti-inflammatory properties effect mentioned earlier, where sustained periods of hyper- may counteract development of both DR and DME. glycemia early in the disease course has long-lasting Calcitriol, or 1,25-dihydroxycholecalciferol, is the meta- effects on future microvascular complications [148]. bolically active form of Vitamin D, and has been found Multiple biochemical pathways involved in DR patho- to be a potent inhibitor of retinal neovascularization in genesis are linked to oxidative stress. The accumulation vitro [157], possibly through suppressing TGF-β and of advanced glycation end products (AGE) in retinal VEGF levels [158]. Epidemiologic studies have found pericytes upregulates cellular expression of its receptor vitamin D deficiency to be associated with increased (RAGE). AGE-RAGE overexpression produces ROS, prevalence and severity of diabetic retinopathy, in both activating apoptotic pathways to cause pericyte loss, seen type 1 [159, 160] and type 2 diabetes [161–163]. How- in early DR [149] The polyol pathway is augmented in ever, all these studies are cross-sectional. No data is hyperglycemic conditions, resulting in overconsumption available on how Vitamin D influences prevalence of of NADPH, reducing its availability for formation of the DME. key endogenous antioxidant glutathione [150]. ROS has also been found to increase the activity of protein kinase Genetic factors C (PKC), a family of serine-threonine kinases that cause As highlighted earlier in this review, certain trends in DR vascular dysfunction by increasing permeability, altering prevalence and incidence cannot be explained by environ- blood flow, and stimulating neovascularization. Vascular mental or socioeconomic factors, such as the abnormally dysfunction and neovascularization is potentiated further high prevalence of DR in rural China, or the large propor- as PKC induces VEGF [144]. Due to how multiple path- tion of VTDR in the Middle East. Some patients appear ways activate and can be activated by oxidative stress, predisposed to severe DR even with adequate risk factor therapeutic strategies targeting any single pathway is control, while others avoided DR despite poor control and unlikely to be effective, as shown in the multiple long diabetes duration [164]. Familial aggregation studies randomized-controlled trials [151–153]. Research has and clinical trials including the DCCT have demon- since focused on mitochondrial dysfunction as the main strated a heritable tendency for severe retinopathy in upstream source of oxidative stress, but whether research type 1 and type 2 diabetes, independent of shared risk in this area will yield novel treatment strategies remains to factors [165–168]. Hence, the hypothesis of differential be seen [148]. genetic susceptibility to DR has drawn interest. The list From an epidemiologic standpoint, given the import- of polymorphisms reviewed here is not exhaustive, but ance of oxidative stress in the pathogenesis of DR, reliable focuses on genes affecting the biological pathways men- and accessible markers of oxidative stress are valuable tioned earlier in the review. measures of disease severity and prognosis. To date, most Polymorphisms in the adipose most abundant gene studies relating oxidative stress to DR involve in vitro and transcript-1 (apM-1) gene on chromosome 1q21.3-q23 animal studies, and oxidative stress markers have not been that codes for adiponectin have been detected to influence investigated in large epidemiologic studies. Small cross- serum adiponectin levels and risk of DR [142]. Partici- sectional studies have consistently found elevated markers pants with type 2 diabetes heterozygous for the Tyr111His of oxidative stress such as lipid peroxide (LPO) and polymorphism at exon 3 (Tyr/His) had significantly higher malondialdehyde in both vitreous and serum of human serum adiponectin levels than participants who were subjects with DR [154, 155]. In particular, serum LPO was homozygous for Tyr111His (Tyr/Tyr), but this had no sta- found to correlate highly with vitreous LPO, and that LPO tistically significant effect on the risk of DR. Participants correlated well with key disease mediators such as VEGF, with type 2 diabetes who had the mutant +45TG allele at Lee et al. Eye and Vision (2015) 2:17 Page 18 of 25 the Gly15Gly polymorphism had no observable differ- effective treatment. Vision loss from DR or DME is hence ences in serum adiponectin levels when compared to a significant healthcare burden [1]. participants with the wild type +45TT allele, but they had A recent systematic review estimated that in 2010, a significantly lower risk of DR. It was unclear why the 3.63 million people worldwide suffer from moderate and reduced risk of DR in this study appeared independent of severe vision loss due to DR and its related sequelae, serum adiponectin levels. Multiple VEGF polymorphisms defined as visual acuity in the better eye being worse have been investigated for their link to DR. The -2578C/ than Snellen 6/18 but at least 3/60. An estimated 850 A, +936C/T and -460 T/C polymorphisms of VEGF have thousand more people suffer from DR-related blindness, been associated with DR in Asians by meta-analysis of defined as visual acuity worse than 3/60 in the better eye cross-sectional studies [169, 170]; The C-634G poly- [2]. Prevalence of vision impairment and blindness due morphism was linked to risk of DME. The CC geno- to DR was found to be on the uptrend, even though total type of this polymorphism was associated with the prevalence of vision impairment and blindness was presence of DME, but was also associated with better decreasing. Findings from reviews of cross-sectional treatment response to bevacizumab when compared studies in Europe [182], South-East Asia and Oceania to the CG and GG genotypes [171]. Recently, single [183], consistently found DR to be the fifth most com- nucleotide polymorphisms in the VEGF-C gene have mon cause of moderate and severe vision loss and blind- been associated with DR and DME in both type 1 ness, behind causes such as uncorrected refractive error, and type 2 diabetes [172]. cataracts, macular degeneration and glaucoma. In Africa, Aldose reductase is the rate-limiting enzyme in the DR is the sixth most common cause of visual impair- polyol pathway that contributes to oxidative stress in ment and blindness, behind the above-listed conditions patients with diabetes. The C(−106)T polymorphism was and trachoma [184]. In the USA, the WESDR investi- found on meta-analysis to be associated with risk of DR in gated visual impairment in patients with type 1 diabetes, type 1, but not type 2 diabetes [173]. Genes coding for and found that 25-year cumulative incidence of visual enzymes in antioxidant pathways such as catalase, impairment (defined as poorer than 6/12 best-corrected superoxide dismutase and glutathione peroxidase are visual acuity in the better eye) and severe visual impair- ment (defined as poorer than 6/60 best-corrected visual downregulated in patients with DR compared to pa- tients with diabetes but without DR, but it is unknown acuity in the better eye) to be 13 and 3 %, respectively if certain polymorphisms predispose to this observation [185]. [174]. Vitamin D receptor gene polymorphisms may Recent data in Leeds, UK, found that in 2008 to 2010, also predispose to DR. T to C substitution at the Taq I DR accounted for 6.1–8.3 % of visual impairment certi- site of the Vitamin D receptor gene [175], and T to C fication. Extrapolated to the total population of the substitution at the start codon FokI [176], was associ- metropolitan area in Leeds, this estimates that 30.0 to ated with severe DR in patients with type 1 diabetes. 43.2 people per million per year will become severely A few genome-wide studies have identified novel gene visually impaired due to DR and its sequelae [186]. In loci associated with DR [177–180]. Association of novel Fife, Scotland, between 2000 and 2009, the mean inci- genes related to vascular endothelium proliferation and dence of blindness (defined as above) was 13.8 per capillary permeability, such as PLXDC2 and ARHGAP22, million per year for the total population of the county imply that our understanding of angiogenic and inflam- [187]. In the Sankara Nethralaya Diabetic Retinopathy matory pathways is still incomplete [178]. Interestingly, Epidemiology and Molecular Genetics Study (SN- polymorphism of RP1-90 L14.1, a long intergenic non- DREAMS) in type 2 diabetes, the prevalence of visual coding RNA gene adjacent to CEP162 was found to be impairment and blindness was 4 and 0.1 %, respectively associated with susceptibility to DR [180]. Since CEP162 [188]. is a key protein in cell ciliogenesis [181], it raises the ques- tion if dysregulation of ciliary assembly plays a role in DR Other eye complications of diabetes pathogenesis. While DR and DME are the most important and well- studied diabetes-related eye complication, many patients Epidemiology of diabetes-related vision loss with diabetes are at risk of vision loss from other While treatment options such as pan-retinal laser photo- diabetes-related eye conditions that range from mild vi- coagulation can largely control neovascularization and sion impairment to blindness. Diabetes is associated with prevent blindness, these treatments cannot restore vision, early and rapid development of cataracts, and is hence a and in fact have vision-impairing effects of their own. In- major cause of visual impairment among patients with travitreal agents such as anti-vascular endothelial growth diabetes. The Singapore Malay Eye Study (SiMES) found factor (VEGF) agents do not fully restore vision in all patients with diabetes to be more likely to have cortical patients, and require frequent and costly doses for and posterior subcapsular cataracts [189]. In the WESDR Lee et al. Eye and Vision (2015) 2:17 Page 19 of 25 study and SN-DREAMS study, presence of cataracts were strong as in nephropathy [8]. In the Chennai Urban Rural significant factors contributing to visual impairment and Epidemiology Study, prevalence of coronary heart disease blindness in patients with diabetes [185, 188]. Many pa- was higher among patients with DR as compared to those tients with diabetes require cataract surgery at a relatively without DR [198]. An eight-year cohort study in Japan younger age. In the WESDR, 10-year cumulative incidence found that patients who developed signs of mild DR were of cataract surgery was 8 % in patients with type 1 diabetes already at higher risk of coronary heart disease or stroke and 25 % in patients with type 2 diabetes [190]. While [9]. Factoring presence of DR in the assessment of patients usually a surgical procedure with good outcomes, cataract with diabetes also improved risk assessment of silent myo- surgery is complicated in patients with diabetes as they cardial infarcts [199]. Presence of DR was also associated may develop DME after surgery [191]. with mortality from cardiovascular disease, especially if Although findings have been inconsistent, diabetes has there is concomitant nephropathy [200]. Literature relat- been found to be a risk factor for developing primary ing DR with peripheral vascular disease is sparse, but a glaucoma in some population-based studies [192]. For recent cross-sectional study in China found an association instance, SiMES found an association between ocular between presence of PDR with lower ankle-brachial index hypertension and diabetes, but not glaucoma [189]. and lower toe-brachial index [201]. Neovascular glaucoma, which is both a blinding and painful condition, can also arise from PDR. A recent Conclusions report found that 7.1 % of patients with PDR requiring As this review shows, the epidemiology of DR has been vitrectomy developed neovascular glaucoma 1 year after extensively studied. The use of a common grading system, surgery [193]. Epiretinal membranes, which can cause the ETDRS severity scale and its modifications, has facili- significant visual impairment, were also found to be tated standardized diagnosis and severity classification of more common among patients with diabetes that have DR in multiple epidemiologic studies, allowing compari- undergone cataract surgery [189]. sons of prevalence, incidence, progression and regression of DR. Review of literature published within the past five Relationship of DR and DME with diabetes related years consistently found higher DR prevalence in Western systemic complications countries compared to Middle-East and Asian countries. Microvascular complications Notable exceptions include Saudi Arabia and Singapore, Diabetic nephropathy is closely associated to DR and two of the most affluent countries in Asia, where DR preva- DME, as many of the pathologic processes affecting mi- lence is comparable to that observed in the US and UK. crovasculature in DR are likely to be causative of diabetic Given the increasing affluence of developing economies nephropathy as well. In a cross-sectional study in Korea, such as China and India, the healthcare burden of DR can compared to patients without DR, patients with DR had be expected to be on the uptrend in the decades ahead. 2.11 the odds (95 % CI 1.04–4.26) of having overt diabetic More recently, cross-sectional studies from developing nephropathy, defined as protein excretion of more than countries are being published. Understandably, the sample 300 mg per 24 h or albumin/creatinine ratio greater than sizes of these studies tend to be small, and few are 300 μg/mg [194]. Ischemic diabetic retinopathy, as evi- population-based. However, it is clear that while people in denced by capillary non-perfusion found on fundal fluor- developing countries are at lower risk of developing dia- escein angiogram, was found to be associated with betes, they have an equivalent if not higher risk of devel- progression of diabetic nephropathy. Patients with more oping DR upon onset of diabetes. While traditional causes than or equal to 10 optic disc areas of capillary non- of visual impairment and blindness in developing coun- perfusion had 6.64 times the risk of progression of tries such as cataracts and trachoma are declining, the nephropathy [195]. Increasing severity of DR was associ- prevalence of DR is growing. Gaps in the literature on ated with increasing severity of chronic kidney disease and theepidemiologyofDRincludethelackof population- based cohort studies investigating the incidence, pro- decreased estimated glomerular filtration rate [196]. In a 15-year follow-up study, development of overt nephropa- gression, and regression in Asian and developing-world thy (defined as above) was found to be associated with the populations. In contrast to DR, the epidemiology of DME is much development of DME [197]. Few studies related the devel- opment of neuropathy with DR. However, the SN- less well studied. Existing studies are split between the use DREAMS found an association between neuropathy and of two diagnostic criteria, one for DME and the other for CSME. Since the CSME criteria are substantially stricter visual-impairment in patients with diabetes [188]. than the DME criteria, direct comparisons between these Macrovascular complications studies cannot be made. The lack of a severity scale also precludes the study of progression and regression of The strength of association between DR and macrovascu- DME. The diagnosis of DME itself is more challenging lar complications, such as cardiovascular disease is just as Lee et al. Eye and Vision (2015) 2:17 Page 20 of 25 than DR. While DR can be diagnosed and classified ad- species; SiMES: Singapore Malay Eye Study; SN-DREAMS: Sankara Nethralaya Diabetic Retinopathy Epidemiology and Molecular Genetics Study; TGF- equately with the assessment of non-stereoscopic fundus β: Transforming growth factor beta; TNF-α: Tumor necrosis factor-α; photos, the diagnosis of DME using this same modality is UK: United Kingdom; UKPDS: United Kingdom prospective diabetes study; challenging as macular thickening is difficult to assess in USA: United States of America; VEGF: Vascular endothelial growth factor; VTDR: Vision-threatening diabetic retinopathy; WESDR: Wisconsin non-stereoscopic photographs. There is no consensus on epidemiologic study of diabetic retinopathy. OCT-based severity classification for DME. More research will have to be carried out to overcome these hurdles in Competing interests diagnosis and classification of DME. The authors declare that they have no competing interests. The investigation of risk factors has also revealed in- teresting considerations both in clinical practice and re- Authors’ contributions RL performed the literature review and drafted the manuscript. CS was search. Hyperglycemia remains the most important involved in drafting the manuscript and critically revising it. WTY critically modifiable risk factor for DR, and intensive glycemic revised the manuscript and gave final approval of the version to be control has been proven to have potent and long-lasting published. All authors read and approved the final manuscript. protective effects against development and progression Author details of DR and DME. As the evidence behind hypertension 1 Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, and dyslipidemia as risk factors is weaker than in hyper- Singapore. Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore. Ophthalmology and Visual Sciences glycemia, intensive control of hypertension and dyslipid- Academic Clinical Program, Duke-NUS Graduate Medical School, Singapore, emia should not be sought solely on the basis to prevent Singapore. onset or progression of DR and DME, but taken in con- Received: 11 August 2015 Accepted: 1 September 2015 sideration of other complications (e.g. reduction in ne- phropathy and cardiovascular diseases). Among novel risk factors, increased serum adiponectin References and LPO were found to be associated with greater preva- 1. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. lence of DR. Vitamin D deficiency has also been found to 2010;376(9735):124–36. 2. Bourne RR, Stevens GA, White RA, Smith JL, Flaxman SR, Price H, et al. be associated with DR, but more evidence is needed to Causes of vision loss worldwide, 1990–2010: a systematic analysis. Lancet ascertain efficacy of Vitamin D supplementation in the Glob Health. 2013;1(6):e339–49. prevention of DR. These novel risk factors are promising, 3. Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes but the findings that have been made in cross-sectional Care. 2012;35(3):556–64. studies have to be supported by consistent findings in pro- 4. Lightman S, Towler HM. Diabetic retinopathy. Clin Cornerstone. spective cohort studies. The relationship between these 2003;5(2):12–21. 5. Tong L, Vernon SA, Kiel W, Sung V, Orr GM. Association of macular factors and DME is unknown and is worth exploring. involvement with proliferative retinopathy in type 2 diabetes. Diabet Med. The association between DR and other vascular diseases 2001;18(5):388–94. are important areas of study. DR is strongly associated 6. He F, Xia X, Wu XF, Yu XQ, Huang FX. Diabetic retinopathy in predicting diabetic nephropathy in patients with type 2 diabetes and renal disease: a with nephropathy, which has significant burden on health- meta-analysis. Diabetologia. 2013;56(3):457–66. care systems due to the need for renal replacement ther- 7. Hagg S, Thorn LM, Putaala J, Liebkind R, Harjutsalo V, Forsblom CM, et al. apy. The presence of DR is also associated with vascular Incidence of stroke according to presence of diabetic nephropathy and severe diabetic retinopathy in patients with type 1 diabetes. Diabetes Care. diseases that are disabling, such as stroke and peripheral 2013;36(12):4140–6. vascular disease, or life threatening, such as myocardial 8. Mottl AK, Pajewski N, Fonseca V, Ismail-Beigi F, Chew E, Ambrosius WT, et al. infarction. Physicians and ophthalmologists should there- The degree of retinopathy is equally predictive for renal and macrovascular outcomes in the ACCORD Trial. J Diabetes Complications. 2014;28(6):874–9. fore be aware that patients with DR and DME are receiv- 9. Kawasaki R, Tanaka S, Tanaka S, Abe S, Sone H, Yokote K, et al. Risk of ing appropriate assessment and treatment for these cardiovascular diseases is increased even with mild diabetic retinopathy: the comorbidities. Japan Diabetes Complications Study. Ophthalmology. 2013;120(3):574–82. 10. Thomas RL, Dunstan FD, Luzio SD, Chowdhury SR, North RV, Hale SL, et al. Abbreviations Prevalence of diabetic retinopathy within a national diabetic retinopathy ACCORD: Action to control cardiovascular risk in diabetes; ADVANCE: Action screening service. Br J Ophthalmol. 2015;99(1):64–8. in diabetes and vascular disease; AGE: Advanced glycation end products; 11. Pedro RA, Ramon SA, Marc BB, Juan FB, Isabel MM. Prevalence and apM-1: Adipose most abundant gene transcript-1; CRP: C-reactive protein; relationship between diabetic retinopathy and nephropathy, and its risk CSME: Clinically significant macular edema; DCCT: Diabetes control and factors in the North-East of Spain, a population-based study. Ophthalmic complications trial; DCPD1987: Danish cohort of pediatric diabetes 1987; Epidemiol. 2010;17(4):251–65. DIRECT: Diabetic retinopathy candesartan trials; DME: Diabetic macular 12. Hautala N, Hannula V, Palosaari T, Ebeling T, Falck A. Prevalence of diabetic edema; DR: Diabetic retinopathy; DRCR.net: Diabetic retinopathy clinical retinopathy in young adults with type 1 diabetes since childhood: the Oulu research network; ETDRS: Early Treatment for diabetic retinopathy study; cohort study of diabetic retinopathy. Acta Ophthalmol. 2014;92(8):749–52. FIELD: Fenofibrate intervention and event-lowering in diabetes; 13. Bertelsen G, Peto T, Lindekleiv H, Schirmer H, Solbu MD, Toft I, et al. Tromso IL-6: Interleukin-6; OCT: Optical coherence tomography; OSA: Obstructive eye study: prevalence and risk factors of diabetic retinopathy. Acta sleep apnea; PDR: Proliferative diabetic retinopathy; PPARα: Peroxisome Ophthalmol. 2013;91(8):716–21. proliferator-activated receptor alpha; PKC: Protein kinase C; RAAS: Renin- 14. Knudsen LL, Lervang HH, Lundbye-Christensen S, Gorst-Rasmussen A. The angiotensin-aldosterone system; RAGE: Receptor of advanced glycation end north Jutland county diabetic retinopathy study: population characteristics. products; RASS: Renin-Angiotensin System Study; ROS: Reactive oxygen Br J Ophthalmol. 2006;90(11):1404–9. Lee et al. Eye and Vision (2015) 2:17 Page 21 of 25 15. Dedov I, Maslova O, Suntsov Y, Bolotskaia L, Milenkaia T, Besmertnaia L. 35. Sharew G, Ilako DR, Kimani K, Gelaw Y. Prevalence of diabetic retinopathy in Prevalence of diabetic retinopathy and cataract in adult patients with type Jimma University Hospital, Southwest Ethiopia. Ethiop Med J. 1 and type 2 diabetes in Russia. Rev Diabet Stud. 2009;6(2):124–9. 2013;51(2):105–13. 16. Roy MS, Klein R, O’Colmain BJ, Klein BE, Moss SE, Kempen JH. The 36. Raman R, Gupta A, Krishna S, Kulothungan V, Sharma T. Prevalence and risk prevalence of diabetic retinopathy among adult type 1 diabetic persons in factors for diabetic microvascular complications in newly diagnosed type II the United States. Arch Ophthalmol. 2004;122(4):546–51. diabetes mellitus. Sankara Nethralaya Diabetic Retinopathy Epidemiology 17. Kempen JH, O’Colmain BJ, Leske MC, Haffner SM, Klein R, Moss SE, et al. The and Molecular Genetic Study (SN-DREAMS, report 27). J Diabetes prevalence of diabetic retinopathy among adults in the United States. Arch Complications. 2012;26(2):123–8. Ophthalmol. 2004;122(4):552–63. 37. Kostev K, Rathmann W. Diabetic retinopathy at diagnosis of type 2 diabetes 18. Zhang X, Saaddine JB, Chou CF, Cotch MF, Cheng YJ, Geiss LS, et al. in the UK: a database analysis. Diabetologia. 2013;56(1):109–11. Prevalence of diabetic retinopathy in the United States, 2005–2008. JAMA. 38. Tapp RJ, Shaw JE, Harper CA, de Courten MP, Balkau B, McCarty DJ, et al. 2010;304(6):649–56. The prevalence of and factors associated with diabetic retinopathy in the Australian population. Diabetes Care. 2003;26(6):1731–7. 19. Kung K, Chow KM, Hui EM, Leung M, Leung SY, Szeto CC, et al. Prevalence of complications among Chinese diabetic patients in urban primary care 39. Wong TY, Cheung N, Tay WT, Wang JJ, Aung T, Saw SM, et al. Prevalence clinics: a cross-sectional study. BMC Fam Pract. 2014;15:8. and risk factors for diabetic retinopathy: the Singapore Malay Eye Study. 20. Jee D, Lee WK, Kang S. Prevalence and risk factors for diabetic retinopathy: Ophthalmology. 2008;115(11):1869–75. the Korea National Health and Nutrition Examination Survey 2008–2011. 40. Looker HC, Nyangoma SO, Cromie D, Olson JA, Leese GP, Black M, et al. Invest Ophthalmol Vis Sci. 2013;54(10):6827–33. Diabetic retinopathy at diagnosis of type 2 diabetes in Scotland. 21. Raman R, Rani PK, Reddi Rachepalle S, Gnanamoorthy P, Uthra S, Diabetologia. 2012;55(9):2335–42. Kumaramanickavel G, et al. Prevalence of diabetic retinopathy in India: 41. Klein R, Klein BK, Moss SE, Davis MD, DeMets DL. The wisconsin Sankara Nethralaya diabetic retinopathy epidemiology and molecular epidemiologic study of diabetic retinopathy: Ix. four-year incidence and genetics study report 2. Ophthalmology. 2009;116(2):311–8. progression of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol. 1989;107(2):237–43. 22. Liu L, Wu X, Liu L, Geng J, Yuan Z, Shan Z, et al. Prevalence of diabetic retinopathy in mainland China: a meta-analysis. PLoS One. 42. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic 2012;7(9):e45264. Study of diabetic retinopathy. XIV. Ten-year incidence and progression of 23. Huang OS, Tay WT, Ong PG, Sabanayagam C, Cheng CY, Tan GS, et al. diabetic retinopathy. Arch Ophthalmol. 1994;112(9):1217–28. Prevalence and determinants of undiagnosed diabetic retinopathy and 43. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic vision-threatening retinopathy in a multiethnic Asian cohort: the Singapore Study of Diabetic Retinopathy: XVII. The 14-year incidence and progression Epidemiology of Eye Diseases (SEED) study. Br J Ophthalmol. 2015. of diabetic retinopathy and associated risk factors in type 1 diabetes. 24. Zheng Y, Lamoureux EL, Lavanya R, Wu R, Ikram MK, Wang JJ, et al. Ophthalmology. 1998;105(10):1801–15. Prevalence and risk factors of diabetic retinopathy in migrant Indians in an 44. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin urbanized society in Asia: the Singapore Indian eye study. Ophthalmology. Epidemiologic Study of Diabetic Retinopathy: XXII the twenty-five-year 2012;119(10):2119–24. progression of retinopathy in persons with type 1 diabetes. Ophthalmology. 2008;115(11):1859–68. 25. Rema M, Premkumar S, Anitha B, Deepa R, Pradeepa R, Mohan V. Prevalence of diabetic retinopathy in urban India: the Chennai Urban Rural 45. Broe R, Rasmussen ML, Frydkjaer-Olsen U, Olsen BS, Mortensen HB, Peto T, Epidemiology Study (CURES) eye study, I. Invest Ophthalmol Vis Sci. et al. The 16-year incidence, progression and regression of diabetic 2005;46(7):2328–33. retinopathy in a young population-based Danish cohort with type 1 26. Al Ghamdi AH, Rabiu M, Hajar S, Yorston D, Kuper H, Polack S. Rapid diabetes mellitus: The Danish cohort of pediatric diabetes 1987 (DCPD1987). assessment of avoidable blindness and diabetic retinopathy in Taif, Saudi Acta Diabetol. 2014;51(3):413–20. Arabia. Br J Ophthalmol. 2012;96(9):1168–72. 46. Younis N, Broadbent DM, Harding SP, Vora JP. Incidence of sight- 27. Papakonstantinou E, Tsinopoulos I, Dimitrakos S, Dehghan MH, Katibeh M, threatening retinopathy in type 1 diabetes in a systematic screening Ahmadieh H, et al. Prevalence and risk factors for diabetic retinopathy in programme. Diabet Med. 2003;20(9):758–65. the 40 to 80 year-old population in Yazd, Iran: the Yazd Eye Study. J 47. Martin-Merino E, Fortuny J, Rivero-Ferrer E, Garcia-Rodriguez LA. Incidence Diabetes. 2015;7(1):139–41. of retinal complications in a cohort of newly diagnosed diabetic patients. 28. Wang FH, Liang YB, Zhang F, Wang JJ, Wei WB, Tao QS, et al. Prevalence of PLoS One. 2014;9(6):e100283. diabetic retinopathy in rural China: the Handan Eye Study. Ophthalmology. 48. Kyto JP, Harjutsalo V, Forsblom C, Hietala K, Summanen PA, Groop PH. 2009;116(3):461–7. Decline in the cumulative incidence of severe diabetic retinopathy in patients with type 1 diabetes. Diabetes Care. 2011;34(9):2005–7. 29. Raman R,Ganesan S,PalSS,KulothunganV,SharmaT.Prevalenceand risk factors for diabetic retinopathy in rural India. Sankara Nethralaya 49. Thomas RL, Dunstan F, Luzio SD, Roy Chowdury S, Hale SL, North RV, et al. Diabetic Retinopathy Epidemiology and Molecular Genetic Study III Incidence of diabetic retinopathy in people with type 2 diabetes mellitus (SN-DREAMS III), report no 2. BMJ Open Diabetes Res Care. attending the diabetic retinopathy screening service for wales: retrospective 2014;2(1):e000005. analysis. BMJ. 2012;344:e874. 30. Katulanda P, Waniganayake YC, Ranasinghe P, Wijetunga WU, Jayaweera M, 50. Younis N, Broadbent DM, Vora JP, Harding SP. Incidence of sight- Wijesinghe NP, et al. Retinopathy among young adults with diabetes threatening retinopathy in patients with type 2 diabetes in the mellitus from a tertiary care setting in Sri Lanka. BMC Endocr Disord. Liverpool Diabetic Eye Study: a cohort study. Lancet. 2014;14:20. 2003;361(9353):195–200. 31. Akhter A, Fatema K, Ahmed SF, Afroz A, Ali L, Hussain A. Prevalence and 51. Stratton IM, Kohner EM, Aldington SJ, Turner RC, Holman RR, Manley SE, et associated risk indicators of retinopathy in a rural Bangladeshi al. UKPDS 50: risk factors for incidence and progression of retinopathy in population with and without diabetes. Ophthalmic Epidemiol. Type II diabetes over 6 years from diagnosis. Diabetologia. 2013;20(4):220–7. 2001;44(2):156–63. 32. Thapa R, Joshi DM, Rizyal A, Maharjan N, Joshi RD. Prevalence, risk factors 52. Jones CD, Greenwood RH, Misra A, Bachmann MO. Incidence and and awareness of diabetic retinopathy among admitted diabetic patients at progression of diabetic retinopathy during 17 years of a population-based a tertiary level hospital in Kathmandu. Nepal J Ophthalmol. 2014;6(11) screening program in England. Diabetes Care. 2012;35(3):592–6. doi:10.3126/nepjoph.v6i1.10760. 53. Tudor SM, Hamman RF, Baron A, Johnson DW, Shetterly SM. Incidence and 33. Kahloun R, Jelliti B, Zaouali S, Attia S, Ben Yahia S, Resnikoff S, et al. progression of diabetic retinopathy in Hispanics and non-Hispanic whites Prevalence and causes of visual impairment in diabetic patients in Tunisia, with type 2 diabetes. San Luis Valley Diabetes Study, Colorado. Diabetes North Africa. Eye (Lond). 2014;28(8):986–91. Care. 1998;21(1):53–61. 34. Mathenge W, Bastawrous A, Peto T, Leung I, Yorston D, Foster A, et al. 54. Varma R, Choudhury F, Klein R, Chung J, Torres M, Azen SP. Four-year Prevalence and correlates of diabetic retinopathy in a population-based incidence and progression of diabetic retinopathy and macular survey of older people in Nakuru, Kenya. Ophthalmic Epidemiol. edema: the Los Angeles Latino Eye Study. Am J Ophthalmol. 2014;21(3):169–77. 2010;149(5):752–61. e1-3. Lee et al. Eye and Vision (2015) 2:17 Page 22 of 25 55. Lee ET, Lee VS, Kingsley RM, Lu M, Russell D, Asal NR, et al. Diabetic 76. Kajiwara A, Miyagawa H, Saruwatari J, Kita A, Sakata M, Kawata Y, et al. retinopathy in Oklahoma Indians with NIDDM. Incidence and risk factors. Gender differences in the incidence and progression of diabetic Diabetes Care. 1992;15(11):1620–7. retinopathy among Japanese patients with type 2 diabetes mellitus: a 56. Cikamatana L, Mitchell P, Rochtchina E, Foran S, Wang JJ. Five-year clinic-based retrospective longitudinal study. Diabetes Res Clin Pract. incidence and progression of diabetic retinopathy in a defined older 2014;103(3):e7–10. population: the Blue Mountains Eye Study. Eye (Lond). 2007;21(4):465–71. 77. Romero-Aroca P, Baget-Bernaldiz M, Fernandez-Ballart J, Plana-Gil N, Soler- 57. Leske MC, Wu SY, Hennis A, Nemesure B, Schachat AP, Hyman L, et al. Nine- Lluis N, Mendez-Marin I, et al. Ten-year incidence of diabetic retinopathy year incidence of diabetic retinopathy in the Barbados Eye Studies. Arch and macular edema. Risk factors in a sample of people with type 1 Ophthalmol. 2006;124(2):250–5. diabetes. Diabetes Res Clin Pract. 2011;94(1):126–32. 58. Tapp RJ, Zimmet PZ, Harper CA, McCarty DJ, Chitson P, Tonkin AM, et al. Six 78. Olsen BS, Sjolie AK, Hougaard P, Johannesen J, Marinelli K, Jacobsen BB, et al. year incidence and progression of diabetic retinopathy: results from the The significance of the prepubertal diabetes duration for the development of Mauritius diabetes complication study. Diabetes Res Clin Pract. retinopathy and nephropathy in patients with type 1 diabetes. J Diabetes 2006;73(3):298–303. Complications. 2004;18(3):160–4. 79. Donaghue KC, Fairchild JM, Craig ME, Chan AK, Hing S, Cutler LR, et al. Do 59. Salinero-Fort MA, San Andres-Rebollo FJ, de Burgos-Lunar C, Arrieta-Blanco all prepubertal years of diabetes duration contribute equally to diabetes FJ, Gomez-Campelo P. Four-year incidence of diabetic retinopathy in a complications? Diabetes Care. 2003;26(4):1224–9. Spanish cohort: the MADIABETES study. PLoS One. 2013;8(10):e76417. 60. Tam VH, Lam EP, Chu BC, Tse KK, Fung LM. Incidence and progression of 80. Harjutsalo V, Maric C, Forsblom C, Thorn L, Waden J, Groop PH. Sex-related diabetic retinopathy in Hong Kong Chinese with type 2 diabetes mellitus. differences in the long-term risk of microvascular complications by age at J Diabetes Complications. 2009;23(3):185–93. onset of type 1 diabetes. Diabetologia. 2011;54(8):1992–9. 61. Song H, Liu L, Sum R, Fung M, Yap MK. Incidence of diabetic retinopathy in 81. Chen S, Jim B, Ziyadeh FN. Diabetic nephropathy and transforming growth a Hong Kong Chinese population. Clin Exp Optom. 2011;94(6):563–7. factor-beta: transforming our view of glomerulosclerosis and fibrosis build- up. Semin Nephrol. 2003;23(6):532–43. 62. Kawasaki R, Tanaka S, Tanaka S, Yamamoto T, Sone H, Ohashi Y, et al. 82. Egan AM, McVicker L, Heerey A, Carmody L, Harney F, Dunne FP. Diabetic Incidence and progression of diabetic retinopathy in Japanese adults with retinopathy in pregnancy: a population-based study of women with type 2 diabetes: 8 year follow-up study of the Japan Diabetes pregestational diabetes. J Diabetes Res. 2015;2015:310239. Complications Study (JDCS). Diabetologia. 2011;54(9):2288–94. doi:10.1155/2015/310239. 63. Fundus photographic risk factors for progression of diabetic retinopathy. ETDRS report number 12. Early Treatment Diabetic Retinopathy Study 83. Diabetes Control and Complications Trial Research Group. Effect of Research Group. Ophthalmology. 1991;98(5 Suppl):823–33. pregnancy on microvascular complications in the diabetes control and 64. Al-Rubeaan K, Abu El-Asrar AM, Youssef AM, Subhani SN, Ahmad NA, Al- complications trial. The Diabetes Control and Complications Trial Research Sharqawi AH, et al. Diabetic retinopathy and its risk factors in a society with Group. Diabetes Care. 2000;23(8):1084–91. a type 2 diabetes epidemic: a Saudi National Diabetes Registry-based study. 84. Ringholm L, Vestgaard M, Laugesen CS, Juul A, Damm P, Mathiesen ER. Acta Ophthalmol. 2015;93(2):e140–7. Pregnancy-induced increase in circulating IGF-I is associated with progression of diabetic retinopathy in women with type 1 diabetes. Growth 65. Pugliese G, Solini A, Zoppini G, Fondelli C, Zerbini G, Vedovato M, et al. Horm IGF Res. 2011;21(1):25–30. High prevalence of advanced retinopathy in patients with type 2 85. Kastelan S, Tomic M, Pavan J, Oreskovic S. Maternal immune system diabetes from the Renal Insufficiency and Cardiovascular adaptation to pregnancy–a potential influence on the course of diabetic Events (RIACE) Italian Multicenter Study. Diabetes Res Clin Pract. retinopathy. Reprod Biol Endocrinol. 2010;8:124. 2012;98(2):329–37. 66. Dutra Medeiros M, Mesquita E, Papoila AL, Genro V, Raposo JF. First diabetic 86. Wong TY, Liew G, Tapp RJ, Schmidt MI, Wang JJ, Mitchell P, et al. Relation retinopathy prevalence study in Portugal: RETINODIAB Study-Evaluation of between fasting glucose and retinopathy for diagnosis of diabetes: the screening programme for Lisbon and Tagus Valley region. Br J three population-based cross-sectional studies. Lancet. Ophthalmol. 2015 doi:10.1136/bjophthalmol-2015–306727 2008;371(9614):736–43. [Epub ahead of print]. 87. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a 67. Nathoo N, Ng M, Rudnisky CJ, Tennant MT. The prevalence of diabetic systematic review. JAMA. 2007;298(8):902–16. retinopathy as identified by teleophthalmology in rural Alberta. Can J 88. The effect of intensive treatment of diabetes on the development and Ophthalmol. 2010;45(1):28–32. progression of long-term complications in insulin-dependent diabetes 68. Esteves JF, Kramer CK, Azevedo MJ, Stolz AP, Roggia MF, Larangeira A, et al. mellitus. The Diabetes Control and Complications Trial Research Group. Prevalence of diabetic retinopathy in patients with type 1 diabetes mellitus. N Engl J Med. 1993;329(14):977–86. Rev Assoc Med Bras. 2009;55(3):268–73. 89. Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of intensive therapy. The Diabetes Control and Complications 69. Villena JE, Yoshiyama CA, Sanchez JE, Hilario NL, Merin LM. Prevalence of Trial/Epidemiology of Diabetes Interventions and Complications Research diabetic retinopathy in Peruvian patients with type 2 diabetes: results of a Group. N Engl J Med. 2000;342(6):381–9. hospital-based retinal telescreening program. Rev Panam Salud Publica. 2011;30(5):408–14. 90. Early worsening of diabetic retinopathy in the Diabetes Control and 70. Thomas RL, Distiller L, Luzio SD, Chowdhury SR, Melville VJ, Kramer B, et al. Complications Trial. Arch Ophthalmol. 1998;116(7):874–86. Ethnic differences in the prevalence of diabetic retinopathy in persons with 91. White NH, Sun W, Cleary PA, Danis RP, Davis MD, Hainsworth DP, et al. diabetes when first presenting at a diabetes clinic in South Africa. Diabetes Prolonged effect of intensive therapy on the risk of retinopathy Care. 2013;36(2):336–41. complications in patients with type 1 diabetes mellitus: 10 years after the 71. Kaidonis G, Mills RA, Landers J, Lake SR, Burdon KP, Craig JE. Review of the Diabetes Control and Complications Trial. Arch Ophthalmol. prevalence of diabetic retinopathy in Indigenous Australians. Clin 2008;126(12):1707–15. Experiment Ophthalmol. 2014;42(9):875–82. 92. Lachin JM, White NH, Hainsworth DP, Sun W, Cleary PA, Nathan DM. Effect of intensive diabetes therapy on the progression of diabetic retinopathy in 72. Virgili G, Menchini F, Murro V, Peluso E, Rosa F, Casazza G. Optical patients with type 1 diabetes: 18 years of follow-up in the DCCT/EDIC. coherence tomography (OCT) for detection of macular oedema in patients Diabetes. 2015;64(2):631–42. with diabetic retinopathy. Cochrane Database Syst Rev. 2011;7:CD008081. 73. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin 93. Hirose A, Furushima D, Yamaguchi N, Kitano S, Uchigata Y. Prediction of Epidemiologic Study of Diabetic Retinopathy XXIII: the twenty-five-year retinopathy at 20 years after onset in younger-onset type 1 diabetes using incidence of macular edema in persons with type 1 diabetes. mean metabolic memory-free HbA1c values: the importance of using Ophthalmology. 2009;116(3):497–503. HbA1c data of total, not partial, diabetes duration. Diabetes Care. 74. Ding J, Wong TY. Current epidemiology of diabetic retinopathy and 2013;36(11):3812–4. diabetic macular edema. Curr Diab Rep. 2012;12(4):346–54. 94. Smith-Palmer J, Brandle M, Trevisan R, Orsini Federici M, Liabat S, Valentine W. Assessment of the association between glycemic variability and 75. Xu J, Xu L, Wang YX, You QS, Jonas JB, Wei WB. Ten-year cumulative diabetes-related complications in type 1 and type 2 diabetes. Diabetes Res incidence of diabetic retinopathy. The Beijing Eye Study 2001/2011. PLoS Clin Pract. 2014;105(3):273–84. One. 2014;9(10):e111320. Lee et al. Eye and Vision (2015) 2:17 Page 23 of 25 95. Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, et al. 117. Cheung D, Switzer NJ, Ehmann D, Rudnisky C, Shi X, Karmali S. The impact Intensive blood glucose control and vascular outcomes in patients with of bariatric surgery on diabetic retinopathy: a systematic review and meta- type 2 diabetes. N Engl J Med. 2008;358(24):2560–72. analysis. Obes Surg. 2015;25(9):1604–9. 96. Gerstein HC, Miller ME, Byington RP, Goff Jr DC, Bigger JT, Buse JB, et al. 118. Tang J, Kern TS. Inflammation in diabetic retinopathy. Prog Retin Eye Res. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2011;30(5):343–58. 2008;358(24):2545–59. 119. Schram MT, Chaturvedi N, Schalkwijk CG, Fuller JH, Stehouwer CD. Markers 97. Fullerton B, Jeitler K, Seitz M, Horvath K, Berghold A, Siebenhofer A. of inflammation are cross-sectionally associated with microvascular Intensive glucose control versus conventional glucose control for type 1 complications and cardiovascular disease in type 1 diabetes–the EURODIAB diabetes mellitus. Cochrane Database Syst Rev. 2014;2:CD009122. Prospective Complications Study. Diabetologia. 2005;48(2):370–8. 98. Jin P, Peng J, Zou H, Wang W, Fu J, Shen B, et al. The 5-year onset and 120. Nguyen TT, Alibrahim E, Islam FM, Klein R, Klein BE, Cotch MF, et al. regression of diabetic retinopathy in Chinese type 2 diabetes patients. PLoS Inflammatory, hemostatic, and other novel biomarkers for diabetic One. 2014;9(11):e113359. retinopathy: the multi-ethnic study of atherosclerosis. Diabetes Care. 2009;32(9):1704–9. 99. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. UK Prospective Diabetes Study 121. Lim LS, Tai ES, Mitchell P, Wang JJ, Tay WT, Lamoureux E, et al. C-reactive Group. BMJ. 1998;317(7160):703–13. protein, body mass index, and diabetic retinopathy. Invest Ophthalmol Vis 100. Holman RR, Paul SK, Bethel MA, Neil HA, Matthews DR. Long-term follow-up Sci. 2010;51(9):4458–63. after tight control of blood pressure in type 2 diabetes. N Engl J Med. 122. A randomized trial comparing intravitreal triamcinolone acetonide and 2008;359(15):1565–76. focal/grid photocoagulation for diabetic macular edema. Ophthalmology. 101. Chaturvedi N, Porta M, Klein R, Orchard T, Fuller J, Parving HH, et al. Effect of 2008;115(9):1447–9, 1149.e1-10. doi:10.1016/j.ophtha.2008.06.015. candesartan on prevention (DIRECT-Prevent 1) and progression (DIRECT- 123. Beck RW, Edwards AR, Aiello LP, Bressler NM, Ferris F, Glassman AR, et al. Protect 1) of retinopathy in type 1 diabetes: randomised, placebo-controlled Three-year follow-up of a randomized trial comparing focal/grid trials. Lancet. 2008;372(9647):1394–402. photocoagulation and intravitreal triamcinolone for diabetic macular edema. Arch Ophthalmol. 2009;127(3):245–51. 102. Sjolie AK, Klein R, Porta M, Orchard T, Fuller J, Parving HH, et al. Effect of candesartan on progression and regression of retinopathy in type 2 124. Angelo LS, Kurzrock R. Vascular endothelial growth factor and its diabetes (DIRECT-Protect 2): a randomised placebo-controlled trial. Lancet. relationship to inflammatory mediators. Clin Cancer Res. 2008;372(9647):1385–93. 2007;13(10):2825–30. 103. Do DV, Wang X, Vedula SS, Marrone M, Sleilati G, Hawkins BS, et al. Blood 125. Jardeleza MS, Miller JW. Review of anti-VEGF therapy in proliferative diabetic pressure control for diabetic retinopathy. Cochrane Database Syst Rev. retinopathy. Semin Ophthalmol. 2009;24(2):87–92. 2015;1:CD006127. 126. Colucciello M. Current intravitreal pharmacologic therapies for diabetic 104. Preiss D, Seshasai SR, Welsh P, Murphy SA, Ho JE, Waters DD, et al. Risk of macular edema. Postgrad Med. 2015;127(6):640–53. incident diabetes with intensive-dose compared with moderate-dose statin 127. Mitchell P, Bandello F, Schmidt-Erfurth U, Lang GE, Massin P, Schlingemann therapy: a meta-analysis. JAMA. 2011;305(24):2556–64. RO, et al. The RESTORE study: ranibizumab monotherapy or combined with 105. Sheely D, Jialal I. Strategies to lower low-density lipoprotein cholesterol in laser versus laser monotherapy for diabetic macular edema. Ophthalmology. metabolic syndrome: averting the diabetes risk. Metab Syndr Relat Disord. 2011;118(4):615–25. 2013;11(3):149–51. 128. Bressler SB, Almukhtar T, Bhorade A, Bressler NM, Glassman AR, Huang SS, et al. 106. Keech AC, Mitchell P, Summanen PA, O’Day J, Davis TM, Moffitt MS, et al. Repeated intravitreous ranibizumab injections for diabetic macular edema and Effect of fenofibrate on the need for laser treatment for diabetic the risk of sustained elevation of intraocular pressure or the need for ocular retinopathy (FIELD study): a randomised controlled trial. Lancet. hypotensive treatment. JAMA Ophthalmol. 2007;370(9600):1687–97. 2015;133(5):589–97. 107. Chew EY, Ambrosius WT, Davis MD, Danis RP, Gangaputra S, Greven CM, et al. 129. Wells JA, Glassman AR, Ayala AR, Jampol LM, Aiello LP, Antoszyk AN, et al. Effects of medical therapies on retinopathy progression in type 2 diabetes. N Aflibercept, bevacizumab, or ranibizumab for diabetic macular edema. Engl J Med. 2010;363(3):233–44. N Engl J Med. 2015;372(13):1193–203. 108. Massin P, Peto T, Ansquer JC, Aubonnet P, Macu FENSIFT. Effects of 130. Elman MJ, Aiello LP, Beck RW, Bressler NM, Bressler SB, Edwards AR, et al. fenofibric acid on diabetic macular edema: the MacuFen study. Ophthalmic Randomized trial evaluating ranibizumab plus prompt or deferred laser or Epidemiol. 2014;21(5):307–17. triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2010;117(6):1064–77. e35. 109. Elam M, Lovato LC, Ginsberg H. Role of fibrates in cardiovascular disease prevention, the ACCORD-Lipid perspective. Curr Opin Lipidol. 131. Elman MJ, Bressler NM, Qin H, Beck RW, Ferris 3rd FL, Friedman SM, et al. 2011;22(1):55–61. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or 110. Cheung N, Wong TY. Obesity and eye diseases. Surv Ophthalmol. triamcinolone plus prompt laser for diabetic macular edema. 2007;52(2):180–95. Ophthalmology. 2011;118(4):609–14. 111. Henricsson M, Nystrom L, Blohme G, Ostman J, Kullberg C, Svensson M, et al. 132. Wozniak SE, Gee LL, Wachtel MS, Frezza EE. Adipose tissue: the new The incidence of retinopathy 10 years after diagnosis in young adult people endocrine organ? A review article. Dig Dis Sci. 2009;54(9):1847–56. with diabetes: results from the nationwide population-based Diabetes 133. Yamagishi S, Inagaki Y, Amano S, Okamoto T, Takeuchi M. Up-regulation of Incidence Study in Sweden (DISS). Diabetes Care. vascular endothelial growth factor and down-regulation of pigment 2003;26(2):349–54. epithelium-derived factor messenger ribonucleic acid levels in 112. Chaturvedi N, Sjoelie AK, Porta M, Aldington SJ, Fuller JH, Songini M, et al. leptin-exposed cultured retinal pericytes. Int J Tissue React. Markers of insulin resistance are strong risk factors for retinopathy incidence 2002;24(4):137–42. in type 1 diabetes. Diabetes Care. 2001;24(2):284–9. 134. Suganami E, Takagi H, Ohashi H, Suzuma K, Suzuma I, Oh H, et al. Leptin 113. Lu J, Hou X, Zhang L, Jiang F, Hu C, Bao Y, et al. Association between body stimulates ischemia-induced retinal neovascularization: possible role of mass index and diabetic retinopathy in Chinese patients with type 2 vascular endothelial growth factor expressed in retinal endothelial cells. diabetes. Acta Diabetol. 2015;52(4):701–8. Diabetes. 2004;53(9):2443–8. 114. Rooney D, Lye WK, Tan G, Lamoureux EL, Ikram MK, Cheng CY, et al. Body 135. Gariano RF, Nath AK, D’Amico DJ, Lee T, Sierra-Honigmann MR. Elevation of mass index and retinopathy in Asian populations with diabetes mellitus. vitreous leptin in diabetic retinopathy and retinal detachment. Invest Acta Diabetol. 2015;52(1):73–80. Ophthalmol Vis Sci. 2000;41(11):3576–81. 115. Rudrappa S, Warren G, Idris I. Obstructive sleep apnoea is associated with 136. Dossarps D, Petit JM, Guiu B, Cercueil JP, Duvillard L, Bron AM, et al. Body the development and progression of diabetic retinopathy, independent of fat distribution and adipokine secretion are not associated with diabetic conventional risk factors and novel biomarkers for diabetic retinopathy. Br J retinopathy in patients with type 2 diabetes mellitus. Ophthalmic Res. Ophthalmol. 2012;96(12):1535. 2014;51(1):42–5. 116. Mason RH, West SD, Kiire CA, Groves DC, Lipinski HJ, Jaycock A, et al. High 137. Sari R, Balci MK, Apaydin C. The relationship between plasma leptin levels prevalence of sleep disordered breathing in patients with diabetic macular and chronic complication in patients with type 2 diabetes mellitus. Metab edema. Retina. 2012;32(9):1791–8. Syndr Relat Disord. 2010;8(6):499–503. Lee et al. Eye and Vision (2015) 2:17 Page 24 of 25 138. Omae T, Nagaoka T, Tanano I, Yoshida A. Adiponectin-induced dilation of 160. Shimo N, Yasuda T, Kaneto H, Katakami N, Kuroda A, Sakamoto F, et al. isolated porcine retinal arterioles via production of nitric oxide from Vitamin D deficiency is significantly associated with retinopathy in young endothelial cells. Invest Ophthalmol Vis Sci. 2013;54(7):4586–94. Japanese type 1 diabetic patients. Diabetes Res Clin Pract. 139. Omae T, Nagaoka T, Yoshida A. Relationship between retinal blood flow 2014;106(2):e41–3. and serum adiponectin concentrations in patients with type 2 diabetes 161. He R, Shen J, Liu F, Zeng H, Li L, Yu H, et al. Vitamin D deficiency increases mellitus. Invest Ophthalmol Vis Sci. 2015;56(6):4143–9. the risk of retinopathy in Chinese patients with type 2 diabetes. Diabet 140. Srinivasan V, Sulochana KN. Effect of adiponectin on expression of vascular Med. 2014;31(12):1657–64. endothelial growth factor and pigment epithelium-derived factor: an in vitro 162. Jee D, Han K, Kim EC. Inverse association between high blood 25- study. Indian J Pharmacol. 2015;47(1):117–20. hydroxyvitamin D levels and diabetic retinopathy in a representative Korean 141. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V. population. PLoS One. 2014;9(12):e115199. Association of serum adiponectin with diabetic microvascular complications 163. Alcubierre N, Valls J, Rubinat E, Cao G, Esquerda A, Traveset A, et al. Vitamin among south Indian type 2 diabetic subjects - (CURES-133). Clin Biochem. D deficiency is associated with the presence and severity of diabetic 2015;48(1–2):33–8. retinopathy in type 2 diabetes mellitus. J Diabetes Res. 2015;2015:374178. 142. Zietz B, Buechler C, Kobuch K, Neumeier M, Scholmerich J, Schaffler A. 164. Patel S, Chen H, Tinkham NH, Zhang K. Genetic susceptibility of diabetic Serum levels of adiponectin are associated with diabetic retinopathy and retinopathy. Curr Diab Rep. 2008;8(4):257–62. with adiponectin gene mutations in Caucasian patients with diabetes 165. Hallman DM, Huber Jr JC, Gonzalez VH, Klein BE, Klein R, Hanis CL. Familial mellitus type 2. Exp Clin Endocrinol Diabetes. 2008;116(9):532–6. aggregation of severity of diabetic retinopathy in Mexican Americans from 143. Yilmaz MI, Sonmez A, Acikel C, Celik T, Bingol N, Pinar M, et al. Adiponectin Starr County, Texas. Diabetes Care. 2005;28(5):1163–8. may play a part in the pathogenesis of diabetic retinopathy. Eur J 166. Hietala K, Forsblom C, Summanen P, Groop PH. Heritability of proliferative Endocrinol. 2004;151(1):135–40. diabetic retinopathy. Diabetes. 2008;57(8):2176–80. 144. Madsen-Bouterse SA, Kowluru RA. Oxidative stress and diabetic retinopathy: 167. Arar NH, Freedman BI, Adler SG, Iyengar SK, Chew EY, Davis MD, et al. pathophysiological mechanisms and treatment perspectives. Rev Endocr Heritability of the severity of diabetic retinopathy: the FIND-Eye study. Invest Metab Disord. 2008;9(4):315–27. Ophthalmol Vis Sci. 2008;49(9):3839–45. 145. Robison Jr WG, Jacot JL, Katz ML, Glover JP. Retinal vascular changes 168. Clustering of long-term complications in families with diabetes in the induced by the oxidative stress of alpha-tocopherol deficiency contrasted diabetes control and complications trial. The Diabetes Control and with diabetic microangiopathy. J Ocul Pharmacol Ther. Complications Trial Research Group. Diabetes. 1997;46(11):1829–39. 2000;16(2):109–20. 169. Wang H, Cheng JW, Zhu LS, Wei RL, Cai JP, Li Y, et al. Meta-analysis of 146. Kowluru RA. Diabetic retinopathy: mitochondrial dysfunction and retinal association between the -2578C/A polymorphism of the vascular capillary cell death. Antioxid Redox Signal. 2005;7(11–12):1581–87. endothelial growth factor and retinopathy in type 2 diabetes in Asians and 147. Naruse R, Suetsugu M, Terasawa T, Ito K, Hara K, Takebayashi K, et al. Caucasians. Ophthalmic Res. 2014;52(1):1–8. Oxidative stress and antioxidative potency are closely associated with 170. Han L, Zhang L, Xing W, Zhuo R, Lin X, Hao Y, et al. The associations diabetic retinopathy and nephropathy in patients with type 2 diabetes. between VEGF gene polymorphisms and diabetic retinopathy susceptibility: Saudi Med J. 2013;34(2):135–41. a meta-analysis of 11 case–control studies. J Diabetes Res. 148. Wu Y, Tang L, Chen B. Oxidative stress: implications for the development of 2014;2014:805801. diabetic retinopathy and antioxidant therapeutic perspectives. Oxid Med 171. El-Shazly SF, El-Bradey MH, Tameesh MK. Vascular endothelial growth factor Cell Longev. 2014;2014:752387. gene polymorphism prevalence in patients with diabetic macular oedema 149. Yamagishi S, Matsui T. Advanced glycation end products (AGEs), oxidative and its correlation with anti-vascular endothelial growth factor treatment stress and diabetic retinopathy. Curr Pharm Biotechnol. 2011;12(3):362–8. outcomes. Clin Experiment Ophthalmol. 2014;42(4):369–78. 150. Oates PJ, Mylari BL. Aldose reductase inhibitors: therapeutic implications for 172. Kaidonis G, Burdon KP, Gillies MC, Abhary S, Essex RW, Chang JH, et al. diabetic complications. Expert Opin Investig Drugs. 1999;8(12):2095–119. Common Sequence Variation in the VEGFC Gene Is Associated with 151. PKC-DRS Study Group. The effect of ruboxistaurin on visual loss in patients Diabetic Retinopathy and Diabetic Macular Edema. Ophthalmology. 2015. with moderately severe to very severe nonproliferative diabetic retinopathy: doi:10.1016/j.ophtha.2015.05.004. initial results of the Protein Kinase C beta Inhibitor Diabetic Retinopathy 173. Zhou M, Zhang P, Xu X, Sun X. The relationship between aldose reductase Study (PKC-DRS) multicenter randomized clinical trial. Diabetes. C106T polymorphism and diabetic retinopathy: an updated meta-analysis. 2005;54(7):2188–97. Invest Ophthalmol Vis Sci. 2015;56(4):2279–89. 152. Aiello LP, Davis MD, Girach A, Kles KA, Milton RC, Sheetz MJ, et al. Effect of 174. El-Bab MF, Zaki NS, Mojaddidi MA, Al-Barry M, El-Beshbishy HA. Diabetic ruboxistaurin on visual loss in patients with diabetic retinopathy. retinopathy is associated with oxidative stress and mitigation of gene Ophthalmology. 2006;113(12):2221–30. expression of antioxidant enzymes. Int J Gen Med. 2013;6:799–806. 153. Bolton WK, Cattran DC, Williams ME, Adler SG, Appel GB, Cartwright K, et al. 175. Taverna MJ, Sola A, Guyot-Argenton C, Pacher N, Bruzzo F, Slama G, et al. Randomized trial of an inhibitor of formation of advanced glycation end Taq I polymorphism of the vitamin D receptor and risk of severe diabetic products in diabetic nephropathy. Am J Nephrol. 2004;24(1):32–40. retinopathy. Diabetologia. 2002;45(3):436–42. 154. Brzovic-Saric V, Landeka I, Saric B, Barberic M, Andrijasevic L, Cerovski B, et al. 176. Taverna MJ, Selam JL, Slama G. Association between a protein Levels of selected oxidative stress markers in the vitreous and serum of polymorphism in the start codon of the vitamin D receptor gene and diabetic retinopathy patients. Mol Vis. 2015;21:649–64. severe diabetic retinopathy in C-peptide-negative type 1 diabetes. J Clin 155. Pan HZ, Zhang H, Chang D, Li H, Sui H. The change of oxidative stress Endocrinol Metab. 2005;90(8):4803–8. products in diabetes mellitus and diabetic retinopathy. Br J Ophthalmol. 177. Grassi MA, Tikhomirov A, Ramalingam S, Below JE, Cox NJ, Nicolae DL. 2008;92(4):548–51. Genome-wide meta-analysis for severe diabetic retinopathy. Hum Mol Genet. 2011;20(12):2472–81. 156. Pludowski P, Holick MF, Pilz S, Wagner CL, Hollis BW, Grant WB, et al. Vitamin D effects on musculoskeletal health, immunity, autoimmunity, 178. Huang YC, Lin JM, Lin HJ, Chen CC, Chen SY, Tsai CH, et al. Genome-wide cardiovascular disease, cancer, fertility, pregnancy, dementia and mortality-a association study of diabetic retinopathy in a Taiwanese population. review of recent evidence. Autoimmun Rev. 2013;12(10):976–89. Ophthalmology. 2011;118(4):642–8. 157. Albert DM, Scheef EA, Wang S, Mehraein F, Darjatmoko SR, Sorenson CM, et al. 179. Sheu WH, Kuo JZ, Lee IT, Hung YJ, Lee WJ, Tsai HY, et al. Genome-wide Calcitriol is a potent inhibitor of retinal neovascularization. Invest Ophthalmol association study in a Chinese population with diabetic retinopathy. Hum Vis Sci. 2007;48(5):2327–34. Mol Genet. 2013;22(15):3165–73. 158. Ren Z, Li W, Zhao Q, Ma L, Zhu J. The impact of 1,25-dihydroxy vitamin D3 180. Awata T, Yamashita H, Kurihara S, Morita-Ohkubo T, Miyashita Y, Katayama S, on the expressions of vascular endothelial growth factor and transforming et al. A genome-wide association study for diabetic retinopathy in a growth factor-beta(1) in the retinas of rats with diabetes. Diabetes Res Clin Japanese population: potential association with a long intergenic non- Pract. 2012;98(3):474–80. coding RNA. PLoS One. 2014;9(11):e111715. 159. Kaur H, Donaghue KC, Chan AK, Benitez-Aguirre P, Hing S, Lloyd M, et al. 181. Wang WJ, Tay HG, Soni R, Perumal GS, Goll MG, Macaluso FP, et al. CEP162 Vitamin D deficiency is associated with retinopathy in children and is an axoneme-recognition protein promoting ciliary transition zone adolescents with type 1 diabetes. Diabetes Care. 2011;34(6):1400–2. assembly at the cilia base. Nat Cell Biol. 2013;15(6):591–601. Lee et al. Eye and Vision (2015) 2:17 Page 25 of 25 182. Bourne RR, Jonas JB, Flaxman SR, Keeffe J, Leasher J, Naidoo K, et al. 203. Chiang PP, Lamoureux EL, Cheung CY, Sabanayagam C, Wong W, Tai ES, et al. Prevalence and causes of vision loss in high-income countries and in Racial differences in the prevalence of diabetes but not diabetic retinopathy in a Eastern and Central Europe: 1990–2010. Br J Ophthalmol. 2014;98(5):629–38. multi-ethnic Asian population. Invest Ophthalmol Vis Sci. 2011;52(10):7586–92. 183. Keeffe J, Taylor HR, Fotis K, Pesudovs K, Flaxman SR, Jonas JB, et al. 204. Zheng Y, Lamoureux EL, Chiang PC, Anuar AR, Ding J, Wang JJ, et al. Prevalence and causes of vision loss in Southeast Asia and Oceania: Language barrier and its relationship to diabetes and diabetic retinopathy. 1990–2010. Br J Ophthalmol. 2014;98(5):586–91. BMC Public Health. 2012;12:781. 184. Naidoo K, Gichuhi S, Basanez MG, Flaxman SR, Jonas JB, Keeffe J, et al. 205. Katulanda P, Ranasinghe P, Jayawardena R. Prevalence of retinopathy Prevalence and causes of vision loss in sub-Saharan Africa: 1990–2010. Br J among adults with self-reported diabetes mellitus: the Sri Lanka diabetes Ophthalmol. 2014;98(5):612–8. and Cardiovascular Study. BMC Ophthalmol. 2014;14:100. 185. Klein R, Lee KE, Gangnon RE, Klein BE. The 25-year incidence of visual 206. Schellini SA, Carvalho GM, Rendeiro FS, Padovani CR, Hirai FE. Prevalence of impairment in type 1 diabetes mellitus the wisconsin epidemiologic study diabetes and diabetic retinopathy in a Brazilian population. Ophthalmic of diabetic retinopathy. Ophthalmology. 2010;117(1):63–70. Epidemiol. 2014;21(1):33–8. 207. Papali’i-Curtin AT, Dalziel DM. Prevalence of diabetic retinopathy and 186. Arora S, Kolb S, Goyder E, McKibbin M. Trends in the incidence of visual maculopathy in Northland, New Zealand: 2011–2012. N Z Med J. impairment certification secondary to diabetic retinopathy in the Leeds 2013;126(1383):20–8. metropolitan area, 2005–2010. Diabet Med. 2012;29(7):e112–6. 208. Win Tin ST, Kenilorea G, Gadabu E, Tasserei J, Colagiuri R. The prevalence of 187. Hall HN, Chinn DJ, Sinclair A, Styles CJ. Epidemiology of blindness diabetes complications and associated risk factors in Pacific Islands attributable to diabetes in Scotland: change over 20 years in a defined countries. Diabetes Res Clin Pract. 2014;103(1):114–8. population. Diabet Med. 2013;30(11):1349–54. 209. Tsugawa Y, Takahashi O, Meigs JB, Davis RB, Imamura F, Fukui T, et al. New 188. Rani PK, Raman R, Gella L, Kulothungan V, Sharma T. Prevalence of visual diabetes diagnostic threshold of hemoglobin A(1c) and the 3-year impairment and associated risk factors in subjects with type II diabetes incidence of retinopathy. Diabetes. 2012;61(12):3280–4. mellitus: sankara nethralaya diabetic retinopathy epidemiology and 210. Ahmed KR, Karim MN, Bhowmik B, Habib SH, Bukht MS, Ali L, et al. molecular genetics study (SN-DREAMS, report 16). Middle East Afr J Incidence of diabetic retinopathy in Bangladesh: a 15-year follow-up study. Ophthalmol. 2012;19(1):129–34. J Diabetes. 2012;4(4):386–91. 189. Chiang PP, Lamoureux EL, Zheng Y, Tay WT, Mitchell P, Wang JJ, et al. 211. Manaviat MR, Rashidi M, Afkhami-Ardekani M. Four years incidence of Frequency and risk factors of non-retinopathy ocular conditions in people diabetic retinopathy and effective factors on its progression in type II with diabetes: the Singapore Malay Eye Study. Diabet Med. diabetes. Eur J Ophthalmol. 2008;18(4):572–7. 2013;30(2):e32–40. 212. Janghorbani M, Amini M, Ghanbari H, Safaiee H. Incidence of and risk 190. Klein BE, Klein R, Moss SE. Incidence of cataract surgery in the Wisconsin factors for diabetic retinopathy in Isfahan, Iran. Ophthalmic Epidemiol. Epidemiologic Study of Diabetic Retinopathy. Am J Ophthalmol. 2003;10(2):81–95. 1995;119(3):295–300. 213. Perol J, Balkau B, Guillausseau PJ, Massin P. A study of the 3-year incidence 191. Baker CW, Almukhtar T, Bressler NM, Glassman AR, Grover S, Kim SJ, et al. of diabetic retinopathy in a French diabetic population seen at Lariboisiere Macular edema after cataract surgery in eyes without preoperative Hospital, Paris. Diabetes Metab. 2012;38(3):225–9. central-involved diabetic macular edema. JAMA Ophthalmol. 214. Xie XW, Xu L, Wang YX, Jonas JB. Prevalence and associated factors of 2013;131(7):870–9. diabetic retinopathy. The Beijing Eye Study 2006. Graefes Arch Clin Exp 192. Jeganathan VS, Wang JJ, Wong TY. Ocular associations of diabetes other Ophthalmol. 2008;246(11):1519–26. than diabetic retinopathy. Diabetes Care. 2008;31(9):1905–12. 215. Varma R, Bressler NM, Doan QV, Gleeson M, Danese M, Bower JK, et al. 193. Goto A, Inatani M, Inoue T, Awai-Kasaoka N, Takihara Y, Ito Y, et al. Prevalence of and risk factors for diabetic macular edema in the United Frequency and risk factors for neovascular glaucoma after vitrectomy in States. JAMA Ophthalmol. 2014;132(11):1334–40. eyes with proliferative diabetic retinopathy. J Glaucoma. 2013;22(7):572–6. 216. Petrella RJ, Blouin J, Davies B, Barbeau M. Prevalence, demographics, and 194. Lee WJ, Sobrin L, Lee MJ, Kang MH, Seong M, Cho H. The relationship treatment characteristics of visual impairment due to diabetic macular between diabetic retinopathy and diabetic nephropathy in a population- edema in a representative Canadian cohort. J Ophthalmol. based study in Korea (KNHANES V-2, 3). Invest Ophthalmol Vis Sci. 2012;2012:159167. 2014;55(10):6547–53. 195. Lee WJ, Sobrin L, Kang MH, Seong M, Kim YJ, Yi JH, et al. Ischemic diabetic retinopathy as a possible prognostic factor for chronic kidney disease progression. Eye (Lond). 2014;28(9):1119–25. 196. Grunwald JE, Alexander J, Ying GS, Maguire M, Daniel E, Whittock-Martin R, et al. Retinopathy and chronic kidney disease in the Chronic Renal Insufficiency Cohort (CRIC) study. Arch Ophthalmol. 2012;130(9):1136–44. 197. Romero P, Baget M, Mendez I, Fernandez J, Salvat M, Martinez I. Diabetic macular edema and its relationship to renal microangiopathy: a sample of Type I diabetes mellitus patients in a 15-year follow-up study. J Diabetes Complications. 2007;21(3):172–80. 198. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V. Relationship of diabetic retinopathy with coronary artery disease in Asian Indians with type 2 diabetes: the Chennai Urban Rural Epidemiology Study (CURES) Eye Study–3. Diabetes Technol Ther. 2015;17(2):112–8. 199. Cosson E, Nguyen MT, Chanu B, Balta S, Takbou K, Valensi P. The report of Submit your next manuscript to BioMed Central male gender and retinopathy status improves the current consensus and take full advantage of: guidelines for the screening of myocardial ischemia in asymptomatic type 2 diabetic patients. Nutr Metab Cardiovasc Dis. 2013;23(6):557–65. • Convenient online submission 200. Ricardo AC, Grunwald JE, Parvathaneni S, Goodin S, Ching A, Lash JP. • Thorough peer review Retinopathy and CKD as predictors of all-cause and cardiovascular mortality: National Health and Nutrition Examination Survey (NHANES) 1988–1994. Am • No space constraints or color figure charges J Kidney Dis. 2014;64(2):198–203. • Immediate publication on acceptance 201. Chen YW, Wang YY, Zhao D, Yu CG, Xin Z, Cao X, et al. High prevalence of lower extremity peripheral artery disease in type 2 diabetes patients with • Inclusion in PubMed, CAS, Scopus and Google Scholar proliferative diabetic retinopathy. PLoS One. 2015;10(3):e0122022. • Research which is freely available for redistribution 202. Abougalambou SS, Abougalambou AS. Risk factors associated with diabetic retinopathy among type 2 diabetes patients at teaching hospital in Submit your manuscript at Malaysia. Diabetes Metab Syndr. 2015;9(2):98–103. www.biomedcentral.com/submit http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Eye and Vision Springer Journals

Epidemiology of diabetic retinopathy, diabetic macular edema and related vision loss

Eye and Vision , Volume 2 (1) – Sep 30, 2015

Loading next page...
 
/lp/springer-journals/epidemiology-of-diabetic-retinopathy-diabetic-macular-edema-and-DpbKpdDz3b

References (226)

Publisher
Springer Journals
Copyright
Copyright © 2015 by Lee et al.
Subject
Medicine & Public Health; Ophthalmology
eISSN
2326-0254
DOI
10.1186/s40662-015-0026-2
pmid
26605370
Publisher site
See Article on Publisher Site

Abstract

Diabetic retinopathy (DR) is a leading cause of vision-loss globally. Of an estimated 285 million people with diabetes mellitus worldwide, approximately one third have signs of DR and of these, a further one third of DR is vision-threatening DR, including diabetic macular edema (DME). The identification of established modifiable risk factors for DR such as hyperglycemia and hypertension has provided the basis for risk factor control in preventing onset and progression of DR. Additional research investigating novel risk factors has improved our understanding of multiple biological pathways involved in the pathogenesis of DR and DME, especially those involved in inflammation and oxidative stress. Variations in DR prevalence between populations have also sparked interest in genetic studies to identify loci associated with disease susceptibility. In this review, major trends in the prevalence, incidence, progression and regression of DR and DME are explored, and gaps in literature identified. Established and novel risk factors are also extensively reviewed with a focus on landmark studies and updates from the recent literature. Keywords: Diabetic retinopathy, Diabetic macular edema, Epidemiology, Risk factors Introduction vision loss in the highly prevalent type 2 diabetes [4] and Diabetic Retinopathy (DR) is the leading cause of vision is invariably present in patients with type 2 diabetes with loss in adults aged 20–74 years [1]. From 1990–2010, DR PDR [5]. In addition to vision loss, DR and DME have also ranked as the fifth most common cause of preventable been shown to contribute to the development of other blindness and fifth most common cause of moderate to diabetes-related complications including nephropathy, severe visual impairment [2]. In 2010, of an estimated 285 peripheral neuropathy and cardiovascular events [6–9]. million people worldwide with diabetes, over one-third The most clinically important risk factors for progres- have signs of DR, and a third of these are afflicted with sion to vision loss include duration of diabetes, hypergly- vision-threatening diabetic retinopathy (VTDR), defined cemia and hypertension. Control of serum glucose and as severe non-proliferative DR or proliferative DR (PDR) blood pressure have been shown to be effective in pre- or the presence of diabetic macular edema (DME) [3]. venting vision loss due to DR. Prevalence and risk factors These estimates are expected to rise further due to the of DR have been studied widely in previous studies includ- increasing prevalence of diabetes, ageing of the population ing regional and ethnic differences, but epidemiological and increasing of life expectancy of those with diabetes. data on DME are relatively scarce. A review conducted in PDR is the most common vision-threatening lesion 2012 suggested that up to 7 % of people with diabetes may particularly among patients with type 1 diabetes. However, have DME and risk factors of DME are largely similar to DME is responsible for most of the visual loss experienced DR. Recently, new information on the epidemiology of DR by patients with diabetes as it remains the major cause of and DME has been published from both developed and developing countries. In this review, we summarize the prevalence of DR and highlight regional differences in the * Correspondence: charumathi.sabanayagam@seri.com.sg Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, epidemiology of DR from recent studies. We also review Singapore the incidence, progression and regression of DR and Yong Loo Lin School of Medicine, National University of Singapore, DME, as well as factors contributing to the progression or Singapore, Singapore Full list of author information is available at the end of the article regression of DR and DME. © 2015 Lee et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Lee et al. Eye and Vision (2015) 2:17 Page 2 of 25 29.6 %, respectively). Of concern is that a large proportion Review of diagnosed DR is vision threatening, with VTDR preva- Prevalence of DR lence estimated to be higher (10.6–17.5 %) than that A pooled individual participant meta-analysis involving 35 observed in the Western world. These observations imply studies conducted worldwide from 1980 to 2008, esti- that most of these cases of DR have been detected late, mated global prevalence of any DR and PDR among when it has already progressed to a vision-threatening patients with diabetes to be 35.4 and 7.5 % respectively stage, or that these populations are particularly susceptible [3]. Prevalence of any DR and PDR was higher in those to severe DR due to ethnic predisposition. Other devel- with type 1 diabetes, compared to those with type 2 oped Asian countries such as Hong Kong [19] and South diabetes (77.3 vs. 25.2 % for any DR, 32.4 vs. 3.0 % for Korea [20] report DR prevalence that is much lower than PDR). Table 1 summarizes the findings of various preva- the global average (12.1 and 15.8 %, respectively). lence studies, organized by region, in comparison to the Apart from the east–west divide, rapidly developing global estimate. Estimates on DR prevalence in type 1 dia- economies in Asia such as China and India are observing betes in Europe and the USA range between 36.5–93.6 %, urban–rural divides in terms of DR disease burden. In with VTDR prevalence estimated between 6.7–34.9 % China, prevalence of DR was reported to be higher among [10–16]. The wide range of prevalence observed may be adults with type 2 diabetes living in rural regions (29.1– due to differences in healthcare systems and socioeco- 43.1 %) [22, 28], compared to their urban counterparts nomic factors between the studied populations, but con- (18.1 %) [22]. Conversely, in a study conducted in Chennai, clusions cannot be made as key characteristics such as India, DR prevalence was reported to be higher in urban known duration of diabetes vary greatly between the sam- (18.0 %) [21] compared to rural areas (10.8 %) [29], pled populations. In the East (Asia and the Middle East), possibly due to the increasing affluence accompanied prevalence studies focused on DR in type 2 diabetes alone, by changes in diet in the urban regions and selective due to the low prevalence of type 1 diabetes in these pop- mortality of those with diabetes-related complications ulations. Hence, comparison of DR prevalence between in rural regions because of poor access to healthcare. the East and West is restricted only to type 2 diabetes. The reason why this urban–rural relationship is re- In general, patients with type 2 diabetes in Western versed in China may represent a case of ethnic pre- communities have a higher prevalence of DR than their disposition, but this is an area that requires further Asian counterparts. In the USA, studies estimate that study. In the past two years, reports on DR prevalence 28.5–40.3 % of patients with type 2 diabetes had DR, and 4.4–8.2 % of them had VTDR [17, 18]. In con- from many developing countries in Asia and Africa have been published [30–35]. Prevalence of DR in Sri Lanka, trast, most Asian countries report DR prevalence to Bangladesh, Nepal, Tunisia, Kenya and Ethiopia ranged be between 12.1–23.0 %, and VTDR prevalence to be be- tween 4.3–4.6 % [19–22]. from 21.6–41.4 %. While the sample sizes of these studies tend to be smaller, they still provide insight into the bur- Singapore is a notable exception to this trend. Despite den of DR in these communities. being an Asian country, paralleling the rapid urbanization, industrialization and internal migration that took place Although duration of diabetes is a major risk factor for DR, a few studies reported DR prevalence in newly over the past five decades in Singapore, DR prevalence in diagnosed diabetes. Prevalence found in these studies Singapore is reported to be higher (33.9 %) than other Asian countries but comparable to that seen in the West- ranged from 2.8 % in South Korea to 28.6 % in Singapore ern world [23]. Within the three major ethnic groups in [20, 27, 32, 36–39]. Surprisingly, a large percentage (19.2 %) of newly diagnosed patients with diabetes have Singapore, the Malays and Indians were reported to have higher a prevalence of DR (33.4 % in Malays, 33.0 % DR in Scotland, UK, where there is universal healthcare. in Indians) compared to the Chinese (25.4 %) [23]. In T his prevalence is even higher than in Nepal (13.0 %) [32], addition to ethnic differences, a study conducted in where access to healthcare is presumably more limited. Singapore also highlighted geographic heterogeneity in However, the prevalence of advanced stages of DR or the prevalence of DR within ethnic Indian groups living in DME was found to be lower among those with newly Singapore (30.4 %) [24] and in urban India (18 %) [21, 25]. diagnosed diabetes suggesting diagnosis of DR early in the It has been speculated that increased acculturation to a course of the disease [40]. westernized lifestyle associated with increased prevalence of obesity and diabetes, and increased awareness among Incidence of DR Indians living in Singapore has led to a higher prevalence, There are few population-based cohort studies, outside of while selective mortality of those with DR in the urban the USA or UK, which have investigated DR incidence. Indian cohorts led to a lower prevalence. In the Middle- Various cohort studies investigating DR incidence over East, Saudi Arabia [26] and Iran [27] both report preva- the past two decades are listed in Table 2. Comparisons lence that are similar to Western communities (36.8 and between the East and West, urban and rural populations, Lee et al. Eye and Vision (2015) 2:17 Page 3 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects Author (Year) Type of study Location Sample size, Diabetes type Prevalence of DR (%) Prevalence of VTDR (%) Age in years Yau (2012) [3] Meta-analysis Global 12,620 Overall 35.36 11.72 (PDR and/or DME) Mean 58.1 Type 1 77.31 38.48 (PDR and/or DME) Range 3–97 Type 2 25.16 6.92 (PDR and/or DME) Asia Liu (2012) [22] Meta-analysis China 11,996 Unspecified 23.0 2.8 (PDR) Range 15–87 Kung (2014) [19] Hospital Hong Kong 15,856 Type 2 12.1 0.3 (PDR) Range ≥ 20 Jee (2013) [202] Population South Korea 1678 Type 2 15.8 4.6 Mean 58.0 ± 11.6 Wong (2008) [39] Population Singapore 757 Type 2 35.0 9.0 Mean 58.7 ± 11.0 Range 40–80 Chiang (2011) [203] Population Singapore 401 Type 2 25.4 Not investigated Mean 53.0 ± 9.0 Range 40–95 Zheng (2012) [204] Population Singapore 1295 Type 2 30.4 7.1 Range ≥ 40 Raman (2009) [21] Population India 1,414 Unspecified 18.0 4.3 Mean 56.1 ± 10.1 Range ≥ 40 Katulanda (2014) [205] Hospital Sri Lanka 536 Unspecified 27.4 Not investigated Mean 56.4 ± 10.9 Range ≥ 18 Akhter (2013) [31] Population Bangladesh 60 Unspecified 21.6 Not investigated Mean 46.0 ± 12 Thapa (2014) [32] Hospital Nepal 277 Unspecified 38.26 14.44 Mean 62.3 ± 13.3 Range ≥ 20 Middle East Al-Rubeaan (2015) [64] Population Saudi Arabia 50,464 Type 2 19.7 10.6 (PDR) 5.7 (DME) (Registry) Mean 59.7 ± 12.8 Range ≥ 25 Al Ghamdi (2012) [26] Population Saudi Arabia 612 Unspecified 36.8 17.5 (Scottish DR Grading R4 and/or M2) Mean 63.3 Range ≥ 50 Papakonstantinou Population Iran 529 Type 2 29.6 11.1 (2015) [27] Range 40–80 Europe Thomas (2015) [10] Population United Kingdom 91,393 Overall 32.4 3.4 (PDR and/or DME) Mean 36.5 ± 16.4 Type 1 56.0 11.2 (PDR and/or DME) Mean 65.3 ± 11.7 Type 2 30.3 2.9 (PDR and/or DME) Lee et al. Eye and Vision (2015) 2:17 Page 4 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects (Continued) Pugliese (2012) [65] Hospital Italy 15,773 Type 2 22.2 9.8 Range 59–75 Pedro (2010) [11] Population Spain 8675 Overall 26.7 0.59 (PDR) Mean 34.9 ± 10.5 Type 1 36.5 1.0 (PDR) 5.73 (DME) Mean 64.6 ± 10.8 Type 2 26.1 0.56 (PDR) 6.44 (DME) Dutra Medeiros Population Portugal 52,739 Type 2 16.3 3.1 (2015) [66] Mean 69.1 ± 11.1 Range ≥ 45 Hautala (2014) [12] Population Finland 172 Type 1 93.6 34.9 (PDR) Mean 30 ± 3 Range 22–35 Bertelsen (2013) [13] Population Norway 514 Overall 26.8 1.2 (PDR) 3.9 (DME) Mean 66.4 Type 1 78.0 Range 46–87 Type 2 25.0 Knudsen (2006) [14] Population Denmark 984 Overall 48.8 Median 37.3 Type 1 53.8 2.9 (PPDR) 5.6 (PDR) 7.9 (CSME) IQR 19.0–48.5 Median 58.1 Type 2 38.7 3.6 (PPDR) 0.9 (PDR) 12.8 (CSME) IQR 15.0–65.0 Dedov (2009) [15] Population Russia 7186 Overall 45.9 8.1 (PPDR) 6.7 (PDR) Median 38.0 Type 1 54.6 9.1 (PPDR) 11.1 (PDR) IQR 27.0–49.0 Median 59.0 Type 2 34.2 7.2 (PPDR) 2.7 (PDR) IQR 54.0–66.0 North America Zhang (2010) [18] Population United States 1006 Unspecified 28.5 4.4 of America Range ≥ 40 Kempen (2004) [17] Pooled population United States 4440 Type 2 40.3 8.2 from 8 studies of America Range ≥ 40 Roy (2004) [16] Pooled population United States 1384 Type 1 79.1 31.2 from 2 studies of America Range ≥ 18 Nathoo (2010) [67] Population Canada 394 Unspecified 27.2 2.3 (PDR) 2.0 (CSME) Mean 58.8 Range 10–100 South America Schellini (2014) [206] Population Brazil 407 Type 2 7.62 Not investigated Mean 51.8 ± 13.6 Range ≥ 30 Esteves (2009) [68] Hospital Brazil 437 Type 1 44.4 3.0 (PPDR) 22.2 (PDR) 9.4 (CSME) Mean 26.8 ± 7.8 Range ≥ 18 Villena (2011) [69] Hospital Peru 1222 Type 2 23.1 1.6 (PPDR) 2.8 (PDR) 2.3 (CSME) Median 59.0 IQR 52.0–67.0 Lee et al. Eye and Vision (2015) 2:17 Page 5 of 25 Table 1 Prevalence of diabetic retinopathy among diabetic subjects (Continued) Africa Thomas (2013) [70] Hospital South Africa 5565 Overall 25.8 7.5 Mean 35.4 ± 15.4 Type 1 36.9 9.7 Mean 56.8 ± 11.8 Type 2 21.4 6.6 Kahloun (2014) [33] Hospital Tunisia 2320 Type 1 and 2 26.3 5.4 (PPDR) 3.4 (PDR) 4.2 (CSME) Mean 54.5 Range 10–92 Mathenge (2014) [34] Population Kenya 195 Unspecified 35.9 13.9 (PPDR + PDR) 4.1 (CSME) Range ≥ 50 Sharew (2013) [35] Hospital Ethiopia 324 Unspecified 41.4 7.3 Oceania Kaidonis (2014) [71] Pooled population Australia 12,666 Type 1 and 2 30.4 7.5 (PDR and/or DME) from 11 studies Range ≥ 15 Papali’i-Curtin Population New Zealand 5647 Unspecified 19.0 0.4 (PDR) (2013) [207] Win Tin (2014) [208] Population Pacific Islands 459 Type 2 47.1 Not investigated (Vanuatu, Nauru, Mean 54 Solomon Islands) DR diabetic retinopathy, VTDR vision-threatening diabetic retinopathy, PDR proliferative diabetic retinopathy, DME diabetic macular edema, PPDR preproliferative diabetic retinopathy, CSME clinically significant macular edema, IQR interquartile range gender and age of diabetes onset. The reduction in risk was and developed versus developing countries are not pos- sible due to the lack of population-based cohort studies in even greater in the cohort diagnosed from 1985 onwards, Asia and many developing countries. In the USA, the at 64 %. Overall, these studies indicate that while almost all patients with type 1 diabetes may eventually develop DR Wisconsin Epidemiologic Study of Diabetic Retinopathy (WESDR) found that among patients with insulin- over time, the incidence of DR and VTDR among patients dependent diabetes with onset before the age of 30, who with type 1 diabetes is probably on the decline. In the UK, population studies involving patients with are presumed to have type 1 diabetes, the 4-year cumula- tive incidence of DR was 59.0 % [41]. At 10, 14 and type 2 diabetes estimated cumulative incidence of DR to 25 years, cumulative incidence of DR in the same cohort be 26.0 % at 4 years [49] 38.1–41.0 % at 6 years [50, 51], and 66 % at 10 years [52]. These findings seem compar- rose to 89.3 % [42], 95.9 % [43], and 97 % [44], respect- ively. Similar observations were made in the Danish able to that found in US population studies, which esti- Cohort of Pediatric Diabetes 1987 (DCPD1987), which mated cumulative incidence of DR to be 22.5–34.0 % at reported a 16-year cumulative incidence of 95.1 % [45]. 4 years [53, 54] and 72.3 % at 14 years [55], despite differ- While these cohorts have long follow-up times, it should ences in ethnicity and age of the cohorts at the time of be noted that the participants were recruited between diabetes diagnosis. Cohorts in Australia [56], Barbados 1979 and 1989. The incidence reported in these studies [57] and Mauritius [58] report cumulative incidence that may not reflect actual DR incidence today, owing to sig- is similar to the UK and US studies. In contrast, the 4-year nificant advancements in retinopathy diagnosis techniques cumulative incidence of DR in a Spanish cohort is much and risk factor management in the past three decades. For lower, estimated at 8.1 % [59]. Age and duration of diabetes example, in a UK cohort recruited between 1991 and are comparable between the US, UK and Spanish studies, 1999, 6-year cumulative incidence of DR in patients with and this significant difference in incidence is attributed to type 1 diabetes was estimated to be only 45.3 % [46]. A unusually good glycemic control within the Spanish cohort, separate UK study, involving only newly diagnosed cases of with mean HbA1c at 7 %, with 55 % of the cohort achieving type 1 diabetes recruited between 2000 and 2007, found 9- HbA1c of less than 7 %. In contrast, patients in one of the year cumulative incidence of DR to be only 23.9 % [47]. In US cohorts [53] had HbA1c of 9.9 % on average. Finland, the incidence of VTDR was reported to be de- As with prevalence, incidence data from Asia is re- creasing in patients with type 1 diabetes [48]. In this study, stricted only to that of type 2 diabetes. A population- patients who were diagnosed with diabetes from 1980 to based study in urban Shanghai, China, found the 5-year 1984 had 47 % reduced risk of VTDR as compared to pa- cumulative incidence to be much higher than in the US tients diagnosed from 1975 to 1979, after adjusting for and UK, at 46.9 %, of which more than a third of it is Lee et al. Eye and Vision (2015) 2:17 Page 6 of 25 Table 2 Incidence of diabetic retinopathy among diabetic subjects Author (Year) Type of study Location Sample size, Diabetes Follow-up Cumulative incidence Incidence of Age in years type time in years of DR, % (95 % CI) VTDR (%) Asia Xu (2014) [75] Population China 2602 Unspecified 10 4.2 (3.45–5.03)** Not investigated Mean 64.6 ± 9.7 Jin (2014) [98] Population China 322 Type 2 5 46.9 13.9 (Severe NPDR) Mean 66.1 ± 13.2 4.6 (PDR) Tam (2009) [60] Hospital Hong Kong 212 Type 2 4 20.3 0.47 (PDR) Mean 55.2 ± 9.5 Song (2011) [61] Hospital Hong Kong 3647 Type 2 4 15.2 0.03 Mean 62.60 ± 9.58 Kawasaki Hospital Japan 1221 Type 2 8 26.6 Not investigated (2011) [62] Mean 58.2 ± 6.9 Kajiwara Hospital Japan 383 Type 2 5.8 ± 2.5 58.5/1000 person Not investigated (2014) [76] (retrospective) years Mean 59.4 ± 11.0 Tsugawa Hospital Japan 1083 Unspecified 3 15.7 Not investigated (2012) [209] Mean 51.0 ± 11.7 Ahmed Hospital Bangladesh 977 Type 2 15 50.6 (47.5–53.8) Not investigated (2012) [210] Mean 41 ± 8 Middle East Manaviat Hospital Iran 120 Type 2 4 47.5 (38.6–56.4) Not investigated (2008) [211] Mean 55.2 ± 9.6 Janghorbani Hospital Iran 549 Type 2 5.1 ± 2.1 45.4 Not investigated (2003) [212] (retrospective) Mean 45.7 ± 9.3 Europe Stratton Population United 1216 Type 2 6 41 Not investigated (2001) [51] Kingdom Mean 52.2 ± 8.5 Younis Population United 305 Type 1 6 45.3 (36.9–53.7) 5.4 (2.3–8.5) (2003) [46] Kingdom Median 30.2 IQR 21.5–39.8 Younis Population United 3743 Type 2 6 38.1 (35.1–41.2) 6.1 (4.4–7.8) (2003) [50] Kingdom Median 63.4 IQR 56.1–69.8 Jones Population United 16,444 Type 2 10 66 18.7 (2012) [52] Kingdom Median 66.7 IQR 58.0–74.5 Martin-Merino Population United 1757 Type 1 9 23.9 4.4 (DME) (2012) [47] Kingdom Mean 19.1 63,226 Type 2 27.8 3.6 (DME) Mean 61.3 Thomas Population United 49,763 Type 2 4 26.0 0.7 (2012) [49] Kingdom Mean 60.2 ± 11.3 Perol Hospital France 236 Unspecified 3 14.0 (9.5–18.4) 0 (2012) [213] Mean 54.0 ± 12.8 Lee et al. Eye and Vision (2015) 2:17 Page 7 of 25 Table 2 Incidence of diabetic retinopathy among diabetic subjects (Continued) Romero-Aroca Hospital Spain 334 Type 1 10 35.9 11.07 (DME) (2011) [77] Mean 25.7 ± 11.7 Salinero-Fort Population Spain 2405 Type 2 4 8.07 (7.04–9.22) 2.8 (PDR) (2013) [59] 1.2 (DME) Mean 67.5 ± 10.6 Henricsson Population Sweden 627 Types 1 and 10 39 1.8 (PDR) (2003) [111] 2 Mean 35.3 ± 5.8 Broe (2014) Population Denmark 185 Type 1 16 95.1 Not investigated [45] Mean 7.5 ± 3.7 North America Lee (1992) [55] Population United States 380 Type 2 14 72.3 15.4 (PDR) (Oklahoma of America Mean 48.2 ± 8.4 12.5 improved Indians) Klein (1998) Population United States 634 Type 1 14 95.9 (93.2–98.6) 26.1 (22.6–29.6) [43] of America (DME) Mean 14.2 ± 7.4 Tudor (1998) Population United States 169 Type 2 4 22.5 Not investigated [53] of America Mean 58.1 Varma (2010) [54] Population United States 775 Unspecified 4 34.0 (30.0–38.0) 5.4 (3.8–7.1) (DME) (Latino) of America Mean 58 ± 9.7 Oceania Cikamatana Population Australia 150 Unspecified 5 22.2 (14.1–32.2) Not investigated (2007) [56] Mean 66.2 ± 8.4 Others Leske Population Barbados 436 Types 1 and 9 39.6 (33.6–45.5) 8.3 (2006) [57] 2 Mean 57.6 ± 9.4 Tapp Population Mauritius 227 Unspecified 6 23.8 0.4 (PDR) (2006) [58] Mean 50 ± 11 DR diabetic retinopathy, VTDR vision-threatening diabetic retinopathy, CI confidence interval, NPDR nonproliferative diabetic retinopathy, PDR proliferative diabetic retinopathy, DME diabetic macular edema **Cumulative incidence of DR among total sample, incidence among participants with diabetes not reported VTDR. This may just be due to differences in known scale, and progression was defined as increase in diabetes duration of the cohorts; the Chinese cohort has severity of 2 steps or more. Some other studies diabetes duration of 11 years on average at baseline assess- assigned DR severity based on the severity grade in ment, while studies in the US and UK report diabetes the worse eye alone. The findings on DR progression duration to be 4 to 7 years on average. More prospective and regression from the various cohort studies are studies are warranted to compare the incidence of DR in summarized in Table 3. Four to six-year cumulative Asia with that observed in Europe or the US. incidence of 2-step progression among the studies ranged from 24.1 to 38.9 %, which increased to 64.1 and 83.1 % in studies with 16-year or 25-year follow-up. Progression and regression of DR In general, progression was much more common than A large number of cohort studies have investigated progres- regression. Two Asian cohort studies, both hospital-based sion and regression of DR [44, 45, 52–54, 56–58, 60–62]. and carried out in Hong Kong, investigated the regression Disease severity was most often classified by the Early of DR. One of the studies found 4-year progression of DR Treatment Diabetic Retinopathy Study (ETDRS) classifica- to be 34.7 % and 4-year regression to be 13.2 % [60], tion for DR severity [63]. The cohort with the longest which is similar to that seen in the population-based US follow-up time was the WESDR cohort, which reported cohorts. However, the other study found 4-year regression 25-year progression of DR in patients with type 1 diabetes to be substantially higher (45.8 %) and progression to be [44]. In this study, DR severity was assigned a level by lower (6.6 %) [61]. This study defined progression or re- concatenating the severity grade in both eyes, with the gression by 1-step change in severity, while most of the worse eye given greater weight. This created a 15-step Lee et al. Eye and Vision (2015) 2:17 Page 8 of 25 Table 3 Progression and regression of diabetic retinopathy Author (Year) Method of severity grading Progression intervals Criteria for progression Progression of Progression from Regression of DR (%) or regression DR (%) NPDR to PDR (%) Asia Tam (2009) [60] Concatenated ETDRS severity Cumulative at 4 years 2-step 34.7 9.9 13.2 of both eyes, with 11 levels Song (2011) [61] Eye with worse ETDRS severity Cumulative at 4 years 1-step 6.6 Not investigated 45.8 Kawasaki (2011) [62] Mild DR to at least severe NPDR according Cumulative at 8 years N/A N/A 15.9 Not investigated to ETDRS Europe Jones (2012) [52] Eye with worse ETDRS severity Cumulative at 5 years Not stated Data irretrievable 6.1 Not investigated Cumulative at 10 years 9.6 Broe (2014) [45] Eye with worse ETDRS severity Cumulative at 16 years 2-step 64.1 31.0 0 North America Tudor (1998) [53] Eye with worse ETDRS severity Cumulative at 4 years 2-step 24.1 Not investigated 13.3 Varma (2010) [54] Concatenated ETDRS severity of both eyes, Cumulative at 4 years 2-step 38.9 5.3 14.0 with 15 levels Klein (2008) [44] Concatenated ETDRS severity of both eyes, Between 0 to 4 years 2-step 13.5 annually 2.5 annually 3.0 annually with 15 levels Between 4 to 10 years 13.0 annually 4.0 annually 0.8 annually Between 10 to 14 years 12.0 annually 2.5 annually 0.4 annually Between 14 to 25 years 2.4 annually 1.5 annually 0.4 annually Cumulative at 25 years 83.1 42.0 17.8 Oceania Cikamatana (2007) [56] Concatenated ETDRS severity of both eyes, Cumulative at 5 years 1-step 25.9 4.1 Not reported with 15 levels Others Leske (2006) [57] Mild or moderate DR to at least severe NPDR Cumulative at 9 years N/A N/A 8.2 Not investigated according to ETDRS Tapp (2006) [58] Mild or moderate DR to at least severe NPDR Cumulative at 6 years N/A 27.7 5.2 Not investigated according to ETDRS DR diabetic retinopathy, NPDR nonproliferative diabetic retinopathy, PDR proliferative diabetic retinopathy, ETDRS Early Treatment for Diabetic Retinopathy Study, N/A not available Lee et al. Eye and Vision (2015) 2:17 Page 9 of 25 other studies defined progression or regression by 2-step populations or a difference in methodology. Of note, clin- ical stereoscopic fundus examination by an ophthalmolo- changes in severity. Moreover, this study was based in a community optometry clinic. Hence, the population sam- gist was carried out in both of these studies and factored ple may be biased towards patients with mild baseline in the diagnosis of DME whereas most studies relied on non-stereoscopic fundus photographs alone, thus raising severity of DR, as patients with more severe disease may the question if prevalence studies using non-stereoscopic have been referred to tertiary hospitals for follow-up. Indeed, 91.7 % of patients with DR at baseline in this study fundus photographs may be severely underdiagnosing DME. In patients with newly diagnosed diabetes, observed had only mild NPDR, and the 1-step regression of prevalence of DME was almost non-existent, with studies mild NPDR to no DR accounted for the majority of the regression observed in this study. The results of reporting it to be within 0 to 0.8 % [21, 39]. A Cochrane review of prevalence of DME assessed by optical coher- this study are hence not directly comparable with that ence tomography (OCT) has found a large range of preva- of the other cohorts, but it highlights the high prob- ability of disease regression in patients with only mild lence rates (19–65 %) [72]. Of note, none of the studies included in the review were population-based studies. NPDR. The absence of data on population-based co- OCT-detected DME was found to have a great degree horts in Asia also precludes direct comparison of pro- of disagreement with the clinical definition of CSME, gression and regression rates between Asian and Western and not all patients who had macular thickening de- populations. tected on OCT progressed to have clinical DME, hence its validity as a diagnostic tool in epidemiologic studies Prevalence of DME is questionable. In most studies, DME was defined by hard exudates in the presence of microaneurysms and blot hemorrhages within one disc diameter of the foveal center. Clinically signifi- Incidence of DME cant macular edema (CSME) is the more severe spectrum Cohort studies that investigated DME incidence are sum- of DME, and was defined by the presence of edema within marized in Table 5. Only studies conducted in the US and 500 μm of the foveal center, or focal photocoagulation Europe investigated DME incidence. The WESDR cohort scars present in the macular area. The prevalence of DME of patients with type 1 diabetes had the longest follow-up among recent cross-sectional studies is summarized in time of 25 years [73]. Interestingly, cumulative incidence Table 4. Among the population-based studies, prevalence of DME and CSME in this cohort seemed to plateau at of DME among patients with type 1 diabetes was between the 14-year mark (DME 26.1 %, CSME 17.0 %), with the 4.2 and 7.9 %. In patients with type 2 diabetes, it was latter 11 years adding minimally to the 25-year cumulative between 1.4 and 12.8 %. Non-stereoscopic fundus photog- incidence (DME 29 %, CSME 17 %). Data available on raphy was used in most studies, which affects the accuracy DME incidence in type 2 diabetes is limited and inconsist- of DME assessment. About half of the studies defined ent [50, 52, 59]. macular edema using the CSME criteria, and hence only the more severe spectrum of DME was captured in these Risk factors for DR and DME studies. Overall, the heterogeneity in methodology causes DR and DME share many common risk factors. comparison of prevalence between these studies to be a Incidence-derived risk factors for DR and DME reported challenge. The prevalence of DME among patients with in the various cohort studies are summarized in Table 6. diabetes is generally much lower than that of DR [11, 13, The major and established risk factors have been reviewed 14, 16–18, 20, 21, 24, 26, 27, 32–35, 39, 64–71]. There extensively before [74]. The most pertinent observations was no observable difference between prevalence of DME will be highlighted again in this review, with updates from between Western or Eastern populations. the latest literature. Novel risk factors were also reviewed. In the Diabetic Retinopathy Screening Service for Wales, a high prevalence of DR (56.0 % in type 1 diabetes, Non-modifiable risk factors 30.3 % in type 2 diabetes) was reported, but the prevalence Duration of diabetes of DME was not found to be higher than other studies Cohort studies with the longest follow-up times found (4.2 % in type 1 diabetes, 1.4 % in type 2 diabetes) [10]. that almost all patients with type 1 diabetes develop some There were a few outliers among the studies that degree of retinopathy if duration of disease exposure is reported exceptionally high prevalence of DME. In Kenya, long enough [44, 45]. This relationship is not as clear in a population-based study found a prevalence of DME of cohort studies on type 2 diabetes, probably due to the 33.3 % among participants with diabetes [34], while a competing risk of mortality in patients with type 2 Canadian study found DME prevalence to be 15.7 %. It is diabetes, who are older and may have more age-related difficult to ascertain if this abnormally - high observed comorbidities. Nevertheless, many studies, both in type 1 prevalence is due to genuinely high prevalence in these and type 2 diabetes [49, 52, 59, 75–77], found disease Lee et al. Eye and Vision (2015) 2:17 Page 10 of 25 Table 4 Prevalence of diabetic macular edema among diabetic subjects Author (Year) Type of study Location Type of Prevalence Definition of macular diabetes (%) edema Yau (2012) [3] Meta-analysis Global Overall 7.48 DME and/or CSME Type 1 14.25 DME and/or CSME Type 2 5.57 DME and/or CSME Xie (2008) [214] Population China Unspecified 4 CSME Jee (2013) [20] Population South Korea Type 2 2.8 DME 1.4 CSME Wong (2008) [39] Population Singapore Type 2 5.7 DME 3.0 CSME Zheng (2012) [24] Population Singapore Type 2 7.2 DME 4.5 CSME Raman (2009) [21] Population India Unspecified 1.4 CSME Thapa (2014) [32] Hospital Nepal Unspecified 5.78 CSME Al-Rubeaan (2015) [64] Population (Registry) Saudi Arabia Type 2 5.7 DME Al Ghamdi (2012) [26] Population Saudi Arabia Unspecified 20.3 Scottish DR Grading M1 15.9 Scottish DR Grading M2 (M2 equivalent to DME) Papakonstantinou (2015) Population Iran Type 2 4.7 CSME [27] Thomas (2015) [10] Population United Kingdom Type 1 4.2 DME Type 2 1.4 DME Pugliese (2012) [65] Hospital Italy Type 2 1.3 DME Pedro (2010) [11] Population Spain Type 1 5.73 CSME Type 2 6.44 CSME Dutra Medeiros (2015) [66] Population Portugal Type 2 1.4 DME Bertelsen (2013) [13] Population Norway Types 1 and 2 3.9 DME Knudsen (2006) [14] Population Denmark Type 1 7.9 CSME Type 2 12.8 CSME Zhang (2010) [18] Population United States of Unspecified 2.7 CSME America Varma (2014) [215] Population United States of Unspecified 3.8 DME America Petrella (2012) [216] Population (registry) Canada Type 1 and 2 15.7 DME Nathoo (2010) [67] Population Canada Unspecified 2.0 CSME Esteves (2009) [68] Hospital Brazil Type 1 9.4 CSME Villena (2011) [69] Hospital Peru Type 2 2.3 CSME Thomas (2013) [70] Hospital South Africa Type 1 and 2 3.2 DME Kahloun (2014) [33] Hospital Tunisia Type 1 and 2 8.7 DME 4.2 CSME Mathenge (2014) [34] Population Kenya Unspecified 33.3 DME 4.1 CSME Sharew (2013) [35] Hospital Ethiopia Unspecified 6.0 CSME Kaidonis (2014) [71] Pooled population from 11 Australia Types 1 and 2 7.6 DME studies DR diabetic retinopathy, DME diabetic macular edema, CSME clinically significant macular edema Lee et al. Eye and Vision (2015) 2:17 Page 11 of 25 Table 5 Incidence of diabetic macular edema among diabetic subjects Author (Year) Type of study Location Type of diabetes Follow-up time Cumulative incidence, Definition of macular in years % (95 % CI) edema Younis (2003) [46] Population United Kingdom Type 1 6 3.2 (0.8–5.7) DME Younis (2003) [50] Population United Kingdom Type 2 6 6.1 (4.4–7.8) DME Jones (2012) [52] Population United Kingdom Type 2 10 1.5 DME Martin-Merino Population United Kingdom Type 1 9 4.4 DME (2012) [47] Type 2 3.6 DME Thomas (2012) [49] Population United Kingdom Type 2 4 1.4 DME Perol (2012) [213] Hospital France Unspecified 3 0 DME Romero-Aroca Hospital Spain Type 1 10 11.07 DME (2011) [77] Salinero-Fort Population Spain Type 2 4 0.01 DME (2013) [59] Klein (1998) [43] Population United States of America Type 1 14 26.1 (22.6–29.6) DME 17.0 (14.1–19.9) CSME Klein (2009) [73] Population United States of America Type 1 25 29 DME 17 CSME Varma (2010) [54] Population (Latino) United States of America Unspecified 4 5.4 (3.8–7.1) DME exclusive of CSME 7.2 (5.2–9.1) CSME Leske (2006) [57] Population Barbados Types 1 and 2 9 8.7 (5.4–12.0) CSME DR diabetic retinopathy, DME diabetic macular edema, CSME clinically significant macular edema, CI confidence interval duration to be a significant risk factor for DR, and this is times as likely to occur in mothers with type 1 diabetes as independent of adequacy of glycemic control. mothers with type 2 diabetes (31.3 vs. 11.7 %, p = 0.001) [82]. This progression is often transient and accompanied Puberty and pregnancy by rapid regression of DR in the postpartum period. At Puberty is a well-known risk factor for DR in type 1 dia- the end of 6.5 years of follow-up on average, prevalence betes. Pre-pubertal years of diabetes exposure contributes and severity of retinopathy was comparable between women with pregnancies and women without pregnancies to added risk of DR [78, 79], but it seems that it is disease exposure during puberty itself, when the body is undergo- [83]. Possible mechanisms behind the progression of DR ing rapid development and maturation, that has the in pregnancy include both hormonal and immune theories [84, 85]. greater impact on the risk of DR. In Finland, the FinnDiane Study Group found that onset of diabetes dur- ing pubertal or post-pubertal age increases risk of devel- Modifiable risk factors oping severe retinopathy requiring laser treatment when Hyperglycemia compared to patients with pre-pubertal onset of diabetes Hyperglycemia is one of the most important risk factors [80]. This was particularly significant among the male for DR and DME. A meta-analysis of three large participants. Biological pathways that may contribute to population-based studies found a graded relationship this phenomenon include the transforming growth factor between the level of glycemia and frequency of retinop- beta (TGF-β) signaling pathway, which is an important athy signs [86]. The United Kingdom Prospective Diabetes mediator of renal microvascular damage [81]. Androgens Study (UKPDS) and the Diabetes Control and Complica- promote and accelerate TGF-β transcriptional activity, tions Trial (DCCT) provided strong evidence that tight which can explain the male preponderance. However, control of glycemia (HbA1c <7 %) reduces the risk of evidence of activation of similar pathways in retinal vessels development and progression of DR in both type 1 and is lacking. type 2 diabetes [87]. The DCCT showed that intensive DR and DME can progress rapidly during pregnancy, glycemic control reduced the incidence of retinopathy by especially in patients with type 1 diabetes. A recent study 76 % and progression from early to advanced retinopathy found progression of DR in pregnancy to be almost 3 by 54 % [88]. This highlights that strict glycemic control is Lee et al. Eye and Vision (2015) 2:17 Page 12 of 25 Table 6 Incidence-derived risk factors for the development of diabetic retinopathy in cohort studies Risk factor Author (Year) Strength of association Age Xu (2014) [75] OR (95 % CI) = 1.00 (0.98–1.02) per year increase Ahmed (2012) [210] HR (95 % CI) = 1.29 (1.07–1.58) per year Janghorbani (2003) [212] HR (95 % CI) = 1.03 (1.006–1.04) per year increase Jones (2012) [52]Comparedto40–70 years, HR (95 % CI) = 1.49 (1.09–2.05) for < 40 years; 1.26 (1.00–1.27) for > 70 years Gender Xu (2014) [75] OR (95 % CI) = 1.32 (0.88–1.96) *reference gender not reported Kajiwara (2014) [76] HR (95 % CI) = 1.85 (1.06–3.24) for female Ahmed (2012) [210] HR (95 % CI) = 1.08 (0.91–1.29) *reference gender not specified Smoking Stratton (2001) [51] OR (95 % CI) = 0.63 (0.48–0.82) if current smoker Duration of diabetes Xu (2014) [75] OR (95 % CI) = 1.16 (1.10–1.22) per year increase Kajiwara (2014) [76] OR (95 % CI) = 1.13 (1.09–1.17) per year increase Romero-Aroca (2011) [77] OR (95 % CI) = 8.90 (4.83–17.4) for ≤ 15 years vs. > 15 years Jones (2012) [52] Compared to < 10 years, HR (95 % CI) = 1.21 (1.01–1.44) for 10–20 years; 0.93 (0.68–1.26) if ≥ 20 years Thomas (2012) [49] Compared to < 5 years, HR (95 % CI) = 1.29 (1.23–1.34) for 5–9 years; 1.68 (1.59–1.77) for 10 years Salinero-Fort (2013) [59] Compared to < 6 years, HR (95 % CI) = 1.22 (0.88–1.70) for 7–14 years; 1.64 (1.05–2.57) for 15–22 years; 2.00 (1.18–3.39) for 22 years HbA1C Xu (2014) [75] OR (95 % CI) = 1.73 (1.35–2.21) per 1 % increase Jin (2014) [98] OR (95 % CI) = 1.12 (1.01–1.24) per 1 % increase Tam (2009) [60] OR (95 % CI) = 1.57 (1.23–2.00) per 1 % increase Kajiwara (2014) [76] OR (95 % CI) = 1.21 (1.08–1.36) per 1 % increase Stratton (2001) [51] Compared to HbA1C < 6.2 %, OR (95 % CI) = 1.4 (1.1–1.8) for 6.2–7.4 %; 2.5 (2.0–3.2) for > 7.4 % Romero-Aroca (2011) [77] OR (95 % CI) = 4.01 (1.91–8.39) if > 7.0 % vs. ≤ 7.0 % Tudor (1998) [53] OR (95 % CI) = 1.50 (0.96–2.36) per 2 % increase Kajiwara (2014) [76] HR (95 % CI) = 1.33 (1.18–1.51) per 1 % increase Janghorbani (2003) [212] HR (95 % CI) = 1.08 (1.007–1.15) per 1 % increase Salinero-Fort (2013) [59] Compared to HbA1C < 7 % HR (95 % CI) = 1.39 (1.01–1.92) for 7–8 %; 1.90 (1.30–2.77) for > 8 % Henricsson (2003) [111] HR (95 % CI) = 1.7 (1.43–1.93) per 1 % increase Use of insulin/diabetes treatment Tudor (1998) [53] OR (95 % CI) = 2.00 (0.75–5.35) if on oral treatment vs. no medications OR (95 % CI) = 9.30 (2.69–32.16) if on insulin vs. no medications Jones (2012) [52] Compared to diet control only, HR (95 % CI) = 1.77 (1.44–2.17) if oral hypoglycemics only HR (95 % CI) = 2.17 (1.68–2.81) if using insulin Thomas (2012) [49] Compared to diet control only, HR (95 % CI) = 1.41 (1.36–1.47) if oral hypoglycemics only HR (95 % CI) = 2.03 (1.89–2.18) if using insulin Blood pressure Jin (2014) [98] OR (95 % CI) = 1.80 (1.14–2.86) if SBP > 140 mmHg and/or DBP > 90 mmHg Kajiwara (2014) [76] OR (95 % CI) = 1.02 (1.01–1.03) per mmHg increase in SBP Stratton (2001) [51] Compared to < 125 mmHg, OR (95 % CI) = 1.5 (1.2–2.6) for SBP was 125–139 mmHg; 2.8 (2.2–3.5) if SBP was ≥ 140 mmHg Romero-Aroca (2011) [77] OR (95 % CI) = 3.31 (1.62–6.75) if SBP > 140 mmHg and/or DBP > 90 mmHg Tudor (1998) [53] OR (95 % CI) = 1.81 (1.02–3.20) per 20 mmHg increase in SBP Lee et al. Eye and Vision (2015) 2:17 Page 13 of 25 Table 6 Incidence-derived risk factors for the development of diabetic retinopathy in cohort studies (Continued) Kajiwara (2014) [76] HR (95 % CI) = 1.01 (0.99–1.03) per mmHg increase in SBP Jones (2012) [52] HR (95 % CI) = 0.72 (0.64–0.81) if on anti-hypertensive medications Obesity Kajiwara (2014) [76] OR (95 % CI) = 1.07 (1.01–1.13) per kg/m increase in BMI Kajiwara (2014) [76] HR (95 % CI) = 1.16 (1.06–1.26) per kg/m increase in BMI Henricsson (2003) [111] HR (95 % CI) = 1.11 (1.04–1.18) per kg/m increase in BMI Nephropathy Xu (2014) [75] OR (95 % CI) = 1.01 (1.002–1.022) per mmol/L increase in serum creatinine concentration Axial Length Xu (2014) [75] OR (95 % CI) = 0.48 (0.33–0.71) per mm increase Cerebrospinal fluid pressure Xu (2014) [75] OR (95 % CI) = 1.10 (1.01–1.21) per mmHg increase Fasting blood glucose Janghorbani (2003) [212] HR (95 % CI) = 1.003 (1.0003–1.005) per mg/dL increase in fasting blood glucose Cholesterol Salinero-Fort (2013) [59] Compared to < 100 mg/dL, HR (95 % CI) = 0.87 (0.65–1.16) for LDL 100–190 mg/dL; 7.91 (3.39–18.47) for LDL > 190 mg/dL Aspirin use Salinero-Fort (2013) [59] HR (95 % CI) = 1.65 (1.22–2.24) if patient takes aspirin Risk factors for DME Incidence Risk factor Author (Year) Strength of association Duration of diabetes Romero-Aroca (2011) [77] OR (95 % CI) = 8.921 (4.321–26.773) if > 15 years of diabetes duration HbA1c Romero-Aroca (2011) [77] OR (95 % CI) = 3.121 (1.823–10.332) if HbA1c is > 7.0 % Blood pressure Klein (2009) [73] HR (95 % CI) = 1.17 (1.10–1.25) per 1 % increase Romero-Aroca (2011) [77] OR (95 % CI) = 3.115 (0.907–10.70) if SBP > 140 mmHg and/or DBP > 90 mmHg Klein (2009) [73] HR (95 % CI) = 1.15 (1.04–1.26) for every 10 mmHg increase in SBP Nephropathy Romero-Aroca (2011) [77] OR (95 % CI) = 6.774 (3.442–18.236) if protein excretion > 200 μg/min or > 300 μg/mg of albumin: creatinine ratio Klein (2009) [73] HR (95 % CI) = 1.43 (0.99–2.08) if urine protein concentration ≥ 30 mg/dL Cholesterol Romero-Aroca (2011) [77] OR (95 % CI) = 4.125 (1.125–15.857) if Total cholesterol/HDL-cholesterol ratio is > 3.5 in men and > 3.0 in women OR odds ratio, HR hazard ratios, CI confidence interval, LDL low density lipoprotein, HDL high density lipoprotein, SBP systolic blood pressure, DBP diastolic blood pressure much more effective in preventing or delaying the onset Glycemic control should be achieved early in the disease of DR in patients with diabetes without DR, rather than course and maintained for as long as possible, since its limiting the severity of DR after it has occurred. In the protective effect is sustained even if tight glycemic control case of DME, intensive glycemic control was associated is lost. This is the metabolic memory effect observed after with 46 % reduction in the incidence of DME at the end the DCCT. Within a year after the end of DCCT, the gly- of the trial and a 58 % reduction 4 years later compared cemic control in the conventional group and intensive with those in the conventional group [89]. The burden of control group had converged, but the participants in the primary prevention of DR and DME hence falls heavily on intensive control group still had lower prevalence of DR primary care physicians, who are in the best position to and DME than the participants in the conventional con- achieve good glycemic control in patients who have not trol group at 10 years after DCCT [91]. Risk reduction in developed complications. In everyday clinical care how- the intensive control group was 52 % between years 1 to ever, it is difficult to replicate the intensity of glycemic 10 after DCCT, but dwindled to 12 % between years 11 to control seen in these studies that were achieved under 18 [92]. This implies that the metabolic memory effect trial conditions. From the findings reported by the DCCT, fades with time, but this is confounded by improved gly- intensive glycemic control actually increases risk of pro- cemic control and risk reduction in the conventional con- gression of existing DR in the first year of treatment [90]. trol group since the end of DCCT. Besides implications for However, this should not deter achieving tight glycemic clinical treatment, metabolic memory also has implications control in patients with existing DR, as the long-term pro- on methodology of diabetes research, seeing that obtaining gression risk reduction outweighs that of the increased mean HbA1c of the entire course of diabetes may be risk in the first year alone. needed to control for the effect of metabolic memory [93]. Lee et al. Eye and Vision (2015) 2:17 Page 14 of 25 Apart from the absolute value of glycemia alone, the or more steps by 35 % in type 1 diabetes, and increased short-term variability of glycemia, such as spikes in post- regression of retinopathy by 34 % in type 2 diabetes prandial glucose, is found to be associated with increased [101, 102]. However, regression only occurred in mild risk of microvascular complications [94]. However, there DR, and candesartan had no effect on incidence or pro- is insufficient data at this point to conclude that fluctua- gression of DME. In the RASS, enalapril and losartan re- tions in blood sugar levels is a causative factor in micro- duced the risk of retinopathy progression by 65 and 70 %, vascular complications considering increased glycemic respectively. Since it was observed that this effect was in- fluctuation can be due to a multitude of correlated factors dependent of blood pressure changes across the period of that may all contribute to microvascular injury, such as the trial, it was proposed that DR risk reduction was not severity of disease or poor compliance. mediated by an effect on hypertension. A recently published Cochrane review concluded that The benefits of achieving euglycemia should be bal- anced with the risk of hypoglycemia, especially in the eld- intensive blood pressure control had a modest effect in erly. In both the Action in Diabetes and Vascular Disease reducing incidence of DR, but does not reduce risk of pro- gression [103]. Insufficient evidence on adverse effects of (ADVANCE) [95] and Action to Control Cardiovascular Risk in Diabetes (ACCORD) [96] trials, aggressive gly- strict blood pressure control in patients with diabetes cemic control (HbA1c <6.5 %) did not significantly reduce made a cost-benefit analysis impossible in the review, and both clinicians and researchers should be aware of this risk of retinopathy development or progression in type 2 diabetes. In ACCORD, it was found that such an aggres- gap in literature. Hence, the overall recommendation is to sive manner of glycemic control may in fact be associated avoid intensive blood pressure control for the sole purpose of slowing DR progression. Instead, control of hyperten- with increased mortality, but it was not ascertained sion in a patient with diabetes should be focused on whether this was directly due to metabolic complications of treatment, such as hypoglycemia. Current institution preventing or limiting progression of other vascular com- plications, particularly nephropathy, as well as lowering guidelines state that treatment goals of hyperglycemia are mortality. There is insufficient evidence for the use of to be anywhere between <6.5 to <7.5 % of HbA1c. Accord- ing to a recently published Cochrane review [97] however, RAS targeting anti-hypertensive medication specifically for preventing or treating retinopathy. there is no concrete evidence on any specific treatment target. Instead, the authors recommend that clinicians set individualized treatment goals based on age, disease pro- Dyslipidemia gression, risk of hypoglycemic episodes, and psychological As outlined in a previous review, the evidence for dyslipid- factors of the patient. emia as a risk factor for DR are inconsistent, and no single lipid measure had been consistently found to be associated Hypertension with DR or DME [74]. In recent cohort studies, only the Multiple epidemiologic studies have identified hyperten- Madrid Diabetes Study found an association between low sion as a risk factor for DR and DME [51, 53, 76, 77, 98]. density lipoprotein (LDL) cholesterol and incidence of DR In the UKPDS, tight blood pressure control (defined as [59]. Moreover, a meta-analysis found that there was a target blood pressure <150/85 mmHg) in patients with dose-dependent relationship of statin use with increasing type 2 diabetes reduced the risk of microvascular disease risk of diabetes [104]. It was then believed that statins by 37 %, the rate of progression of DR by 34 %, and the might have effects on glucose homeostasis, such as risk of deterioration of visual acuity by 47 % [99]. Unlike decreasing insulin production or increasing insulin resist- in the case of hyperglycemia, the protective effect of blood ance, or both [105]. Therefore, while the use of statins is pressure control waned quickly upon stopping intensive first-line treatment for dyslipidemia in the prevention of control [100]. Anti-hypertensive medications targeting the cardiovascular events in patients with diabetes, the evi- renin-angiotensin-aldosterone system (RAAS) are now the dence for intensive control by statins for the purposes of first line treatment for control of hypertension in patients treating DR and DME are lacking. with nephropathy as it was found that they had additional Fenofibrate, a peroxisome proliferator-activated receptor beneficial effects independent of their absolute hypotensive alpha (PPARα) agonist, has gathered interest on its effects action. Since retinopathy and diabetic nephropathy are re- on DR and DME. In an ancillary study of the Fenofibrate lated microvascular complications, clinical trials such as Intervention and Event-lowering in Diabetes (FIELD) the Diabetic Retinopathy Candesartan Trials (DIRECT) cohort, participants treated with fenofibrate had a 31 % re- and Renin-Angiotensin System Study (RASS) measured duced risk of requiring laser treatment for PDR or DME, the beneficial effects these classes of anti-hypertensive compared to placebo [106]. However, 2-step progression medications had on DR and DME. Candesartan was found of retinopathy did not differ significantly between the fenofibrate and placebo group, except for the subgroup to reduce the incidence of retinopathy by two or more steps in severity on the ETDRS scale by 18 % or by three with pre-existing DR. In this subgroup, risk of 2-step Lee et al. Eye and Vision (2015) 2:17 Page 15 of 25 progression was almost a fifth of that compared to pla- Asians with type 2 diabetesis yet to be confirmed by a cebo. Moreover, in a more recent trial by the ACCORD cohort study. group, adjunct fenofibrate with simvastatin compared Closely related to obesity is the study of obstructive to simvastatin alone reduced the rate of progression of sleep apnea (OSA) as a potential risk factor for DR and DME. A cross-sectional study in patients with type 2 DR (6.5 vs. 10.2 %, respectively) by at least 3 steps at 4 years [107]. Fenofibrate treatment may also have benefi- diabetes found that OSA was associated with DR severity, cial effects on DME, as it was found to have a moderate but not DME [115]. A separate study on patients with CSME found high prevalence of sleep-disordered breath- effect in decreasing macular volume in patients with DME [108]. The sample size of this study however, was relatively ing in these patients, but severity of sleep-disordered small, and more studies are required to study this associ- breathing was not correlated with severity of DR or DME in this study [116]. However, the sample sizes of these ation. Given the current evidence, it is found that patients with DR benefit most from fibrate therapy if they have studies were too small to draw any concrete conclusions. hypertriglyceridemia and low serum high density lipopro- Bariatric surgery is a highly effective treatment for mor- bid obesity that achieves glycemic control of diabetes tein (HDL)-cholesterol, and hence treatment can be justi- rapidly. However, much like how intensive glucose control fied in this subset of patients, with the hopes of slowing progression to PDR. However, generalization of fibrate with medications or insulin increases risk of DR progres- sion in the short-run, this rapid improvement in glycemic treatment to all patients with diabetes at risk of DR is not control post-bariatric surgery has been associated with recommended without stronger evidence [109]. progression of DR. Most studies presented in this area are case series, and a recent meta-analysis of these studies Obesity found that patients with pre-existing DR are 2.77 times The effect obesity has on DR has been relatively well- studied but with inconclusive and conflicting findings (95 % CI 1.10–6.99) more likely to have adverse outcomes in DR post-operatively than patients without pre-existing [110]. It may be possible that obesity has differing im- DR [117]. As mentioned earlier, increased risk of progres- pacts on DR in type 1 diabetes as compared to type 2 sion with intensive glycemic control occurred only in the diabetes. In the Diabetic Incidence Study in Sweden in- first year of follow-up, with subsequent risk reduction with volving predominantly participants with type 1 dia- longer-term control [90]. It remains to be seen if this is betes, it was found that risk of developing DR the case with bariatric surgery as well, as no studies had increased by 1.11 (95 % confidence interval (CI) 1.04– sufficient follow-up time to determine if bariatric surgery 1.18) times per kg/m increase in Body Mass Index (BMI) has long-term benefits on DR. after 10 years of follow-up [111]. In the EURODIAB Prospective Complications Study, also involving patients with type 1 diabetes, larger waist to hip ratio was asso- Novel risk factors ciated with incidence of DR after more than 7 years of Inflammation follow-up [112]. Retinal and vitreous inflammation was observed in sub- In contrast, many studies in type 2 diabetes, performed jects with diabetes, both in animal models and human primarily in Asia, found an inverse relationship between studies. The role of inflammation in DR and DME is obesity and DR. In a cross-sectional study of the Shanghai therefore an area of extensive study, and has been Diabetes Registry Database, participants who were over- reviewed previously [118]. As pointed out in the review weight had reduced risk of DR and VTDR [113]. A similar however, current data suggests systemic inflammation study on the multi-ethnic population in Singapore found cannot account for the characteristic lesions seen in DR the same risk reduction in obese patients for DR, VTDR and DME. Many conditions can lead to systemic inflam- and CSME [114]. mation (e.g. sepsis, autoimmune disease), but DR-like The exact mechanisms underlying this discrepancy be- lesions and DME are not seen in these diseases. Hence, it tween type 1 and type 2 diabetes are not well understood. seems that the local retinal inflammation seen in subjects It was postulated that unintentional weight loss is a sign with diabetes is not related to systemic inflammation. This of advanced and severe type 2 diabetes, hence the obser- challenges the validity of investigating systemic inflamma- vation of non-obese patients with type 2 diabetes being at tory markers such as serum C-reactive protein (CRP), higher risk of DR. In contrast, obesity and metabolic syn- interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) drome do not contribute to the etiology of type 1 diabetes, as risk factors for DR or DME. Indeed, inconsistencies in which is autoimmune in nature, and obese patients with the association between systemic inflammatory markers type 1 diabetes may simply have more difficulties achiev- and risk of DR and DME exist in the current literature. ing good glycemic control. It should be noted that there The EURODIAB Prospective Complications Study found are no prospective population-based studies in Asia on an association between CRP, IL-6, TNF-α and presence of DR incidence, and the protective effect of obesity in DR in subjects with type 1 diabetes via a cross-sectional Lee et al. Eye and Vision (2015) 2:17 Page 16 of 25 study [119]. Other cross-sectional studies found no such Metabolic hormones association. The Multi-ethnic Study of Atherosclerosis did Hormones involved in metabolism have been hypothe- not find an association between CRP and DR or VTDR sized to play key roles in the pathogenesis of microvascu- (which includes DME), but found an association between lar complications in diabetes, due to their roles in both fibrinogen, an acute-phase reactant in systemic inflamma- metabolic and inflammatory pathways [132]. In particular, tion, and DR and VTDR [120]. The Singapore Malay Eye leptin and adiponectin, which are actively secreted by Study even found that raised CRP was associated with a adipocytes to regulate energy balance in the body, have lower prevalence of DR [121]. None of the studies found been implicated as potential risk factors. an association between systemic inflammatory markers Leptin may play a role in inciting inflammation. Leptin and DME specifically. was found to cause upregulation of VEGF in retinal Local retinal inflammation forms the basis of intra- pericytes [133], hence stimulating angiogenesis in the is- venous administration of corticosteroids. The Diabetic chemic retina [134], and possibly contributing to the neo- Retinopathy Clinical Research Network (DRCR.net) vascularization seen in PDR. Elevated serum and vitreous compared intravitreal triamcinolone versus focal/grid leptin was observed in patients with diabetes, and vitreous laser photocoagulation in patients with DME. The leptin was especially elevated in patients with PDR [135]. findings showed that the triamcinolone group had However, cross-sectional studies could not find an associ- better visual acuity at the 4-month interval, but ation between elevated serum leptin and DR [136, 137], equivalent visual acuity at the 1-year interval. At the though it should be noted that the sample sizes of 2-year [122] and 3-year interval [123], mean visual these studies were relatively small and they may be acuity was better in the photocoagulation than the underpowered. triamcinolone groups. Hence, corticosteroid treatment Adiponectin has been found to induce dilation of retinal for DME is effective, but the effect is transient. Clini- arterioles via upregulation of endothelial cell nitric oxide cians also have to be cautious with adverse effects production, in animal studies [138]. Studies by the same such as elevated intraocular pressure and cataract group in human subjects with mild DR found that serum formation. adiponectin was positively correlated with retinal blood flow velocity and negatively correlated with retinal arterial Vascular endothelial growth factor (VEGF) is a key modulator of angiogenesis and vascular permeability, and resistance [139]. Hence, adiponectin may have a role in is upregulated by inflammatory cytokines [124]. Anti- countering ischemia by promoting reperfusion in the VEGF agents have been used successfully for the treat- ischemic retina. In vitro studies also found that it downre- ment of both PDR and DME [125, 126]. Ranibizumab, an gulates VEGF and thus may have anti-angiogenic proper- anti-VEGF agent, was more effective than laser therapy in ties [140]. Large population-based cross-sectional studies restoring vision for DME [127], although just like with found that elevated serum adiponectin in patients with corticosteroids, ranibizumab is associated with elevations DR correlated with severity of DR when compared to in intraocular pressure [128]. In recent reports, the patients without DR [141, 142]. However, there are incon- DRCR.net compared outcomes in DME treated by afliber- sistencies in literature, with one study finding decreased cept, bevacizumab or ranibizumab, and found that afliber- serum adiponectin in participants with PDR [143]. Given cept provided superior visual recovery if baseline visual that basic science suggests adiponectin as mainly protect- acuity was poorer than 69 ETDRS letters (approximately ive against the development of microvascular complica- 6/15 Snellen) when compared to the other anti-VEGF tions, the observation that serum adiponectin is elevated agents, but there was no significant difference between in patients with severe DR appears contradictory. It may aflibercept and the other anti-VEGF agents if baseline be that upregulation of adiponectin secretion can be at- visual acuity was better than 69 letters [129]. tributed to a natural response that ameliorates the effects Anti-VEGF agents appear superior to corticosteroids in of severe microvascular disease, but prospective cohort terms of efficacy. DRCR.net compared ranibizumab and studies are needed to establish the temporal link between concurrent photocoagulation against triamcinolone with adiponectin levels and the development and progression photocoagulation in patients with DME, and found that of DR. Overall, it appears research in adiponectin has ranibizumab achieved better visual outcome at 1-year produced more promising and consistent results than follow-up than triamcinolone, except in a subset of leptin. The association between these hormones and DME patients with pseudophakic eyes [130]. In this subset of has yet to be studied. participants, triamcinolone achieved comparable visual outcome when compared with ranibizumab, possibly Oxidative stress because of the removed effect of steroid-induced cataract Oxidative stress is the accumulation of free radicals in the formation in pseudophakic eyes. Consistent results were form of reactive oxygen species (ROS). Highly efficient obtained at 2-year follow-up [131]. physiological mechanisms consisting of endogenous free Lee et al. Eye and Vision (2015) 2:17 Page 17 of 25 suggesting that serum LPO may be a suitable proxy meas- radical scavengers usually keep oxidative stress low. How- ure of DR severity [154]. More studies will be needed to ever, under pathological conditions, ROS production may confirm this association. be increased such that the defensive mechanisms are over- whelmed, or the protective mechanisms themselves may be impaired, or both [144]. Oxidative stress has been Vitamin D linked to the histopathological changes of DR, such as On top of its well-known effects on calcium metabolism, retinal basement membrane thickening [145] and capillary Vitamin D has anti-angiogenic and anti-inflammatory cell loss [146]. Increased ROS and decreased antioxidant effects that have implicated Vitamin D deficiency in the potential has also been found in patients with diabetes, es- pathogenesis of various types of pathology, such as malig- pecially if they have DR [147]. The effects of oxidative nancy, autoimmune disease, cardiovascular disease and stress are observed early in the course of diabetes, and its diabetes [156]. effects on microvasculature persist even if hyperglycemia It is thus intuitive that Vitamin D has a protective ef- is subsequently corrected. Hence, oxidative stress is likely fect on DR and DME, since anti-angiogenesis may slow to be the mechanism behind the “metabolic memory” progression to PDR and anti-inflammatory properties effect mentioned earlier, where sustained periods of hyper- may counteract development of both DR and DME. glycemia early in the disease course has long-lasting Calcitriol, or 1,25-dihydroxycholecalciferol, is the meta- effects on future microvascular complications [148]. bolically active form of Vitamin D, and has been found Multiple biochemical pathways involved in DR patho- to be a potent inhibitor of retinal neovascularization in genesis are linked to oxidative stress. The accumulation vitro [157], possibly through suppressing TGF-β and of advanced glycation end products (AGE) in retinal VEGF levels [158]. Epidemiologic studies have found pericytes upregulates cellular expression of its receptor vitamin D deficiency to be associated with increased (RAGE). AGE-RAGE overexpression produces ROS, prevalence and severity of diabetic retinopathy, in both activating apoptotic pathways to cause pericyte loss, seen type 1 [159, 160] and type 2 diabetes [161–163]. How- in early DR [149] The polyol pathway is augmented in ever, all these studies are cross-sectional. No data is hyperglycemic conditions, resulting in overconsumption available on how Vitamin D influences prevalence of of NADPH, reducing its availability for formation of the DME. key endogenous antioxidant glutathione [150]. ROS has also been found to increase the activity of protein kinase Genetic factors C (PKC), a family of serine-threonine kinases that cause As highlighted earlier in this review, certain trends in DR vascular dysfunction by increasing permeability, altering prevalence and incidence cannot be explained by environ- blood flow, and stimulating neovascularization. Vascular mental or socioeconomic factors, such as the abnormally dysfunction and neovascularization is potentiated further high prevalence of DR in rural China, or the large propor- as PKC induces VEGF [144]. Due to how multiple path- tion of VTDR in the Middle East. Some patients appear ways activate and can be activated by oxidative stress, predisposed to severe DR even with adequate risk factor therapeutic strategies targeting any single pathway is control, while others avoided DR despite poor control and unlikely to be effective, as shown in the multiple long diabetes duration [164]. Familial aggregation studies randomized-controlled trials [151–153]. Research has and clinical trials including the DCCT have demon- since focused on mitochondrial dysfunction as the main strated a heritable tendency for severe retinopathy in upstream source of oxidative stress, but whether research type 1 and type 2 diabetes, independent of shared risk in this area will yield novel treatment strategies remains to factors [165–168]. Hence, the hypothesis of differential be seen [148]. genetic susceptibility to DR has drawn interest. The list From an epidemiologic standpoint, given the import- of polymorphisms reviewed here is not exhaustive, but ance of oxidative stress in the pathogenesis of DR, reliable focuses on genes affecting the biological pathways men- and accessible markers of oxidative stress are valuable tioned earlier in the review. measures of disease severity and prognosis. To date, most Polymorphisms in the adipose most abundant gene studies relating oxidative stress to DR involve in vitro and transcript-1 (apM-1) gene on chromosome 1q21.3-q23 animal studies, and oxidative stress markers have not been that codes for adiponectin have been detected to influence investigated in large epidemiologic studies. Small cross- serum adiponectin levels and risk of DR [142]. Partici- sectional studies have consistently found elevated markers pants with type 2 diabetes heterozygous for the Tyr111His of oxidative stress such as lipid peroxide (LPO) and polymorphism at exon 3 (Tyr/His) had significantly higher malondialdehyde in both vitreous and serum of human serum adiponectin levels than participants who were subjects with DR [154, 155]. In particular, serum LPO was homozygous for Tyr111His (Tyr/Tyr), but this had no sta- found to correlate highly with vitreous LPO, and that LPO tistically significant effect on the risk of DR. Participants correlated well with key disease mediators such as VEGF, with type 2 diabetes who had the mutant +45TG allele at Lee et al. Eye and Vision (2015) 2:17 Page 18 of 25 the Gly15Gly polymorphism had no observable differ- effective treatment. Vision loss from DR or DME is hence ences in serum adiponectin levels when compared to a significant healthcare burden [1]. participants with the wild type +45TT allele, but they had A recent systematic review estimated that in 2010, a significantly lower risk of DR. It was unclear why the 3.63 million people worldwide suffer from moderate and reduced risk of DR in this study appeared independent of severe vision loss due to DR and its related sequelae, serum adiponectin levels. Multiple VEGF polymorphisms defined as visual acuity in the better eye being worse have been investigated for their link to DR. The -2578C/ than Snellen 6/18 but at least 3/60. An estimated 850 A, +936C/T and -460 T/C polymorphisms of VEGF have thousand more people suffer from DR-related blindness, been associated with DR in Asians by meta-analysis of defined as visual acuity worse than 3/60 in the better eye cross-sectional studies [169, 170]; The C-634G poly- [2]. Prevalence of vision impairment and blindness due morphism was linked to risk of DME. The CC geno- to DR was found to be on the uptrend, even though total type of this polymorphism was associated with the prevalence of vision impairment and blindness was presence of DME, but was also associated with better decreasing. Findings from reviews of cross-sectional treatment response to bevacizumab when compared studies in Europe [182], South-East Asia and Oceania to the CG and GG genotypes [171]. Recently, single [183], consistently found DR to be the fifth most com- nucleotide polymorphisms in the VEGF-C gene have mon cause of moderate and severe vision loss and blind- been associated with DR and DME in both type 1 ness, behind causes such as uncorrected refractive error, and type 2 diabetes [172]. cataracts, macular degeneration and glaucoma. In Africa, Aldose reductase is the rate-limiting enzyme in the DR is the sixth most common cause of visual impair- polyol pathway that contributes to oxidative stress in ment and blindness, behind the above-listed conditions patients with diabetes. The C(−106)T polymorphism was and trachoma [184]. In the USA, the WESDR investi- found on meta-analysis to be associated with risk of DR in gated visual impairment in patients with type 1 diabetes, type 1, but not type 2 diabetes [173]. Genes coding for and found that 25-year cumulative incidence of visual enzymes in antioxidant pathways such as catalase, impairment (defined as poorer than 6/12 best-corrected superoxide dismutase and glutathione peroxidase are visual acuity in the better eye) and severe visual impair- ment (defined as poorer than 6/60 best-corrected visual downregulated in patients with DR compared to pa- tients with diabetes but without DR, but it is unknown acuity in the better eye) to be 13 and 3 %, respectively if certain polymorphisms predispose to this observation [185]. [174]. Vitamin D receptor gene polymorphisms may Recent data in Leeds, UK, found that in 2008 to 2010, also predispose to DR. T to C substitution at the Taq I DR accounted for 6.1–8.3 % of visual impairment certi- site of the Vitamin D receptor gene [175], and T to C fication. Extrapolated to the total population of the substitution at the start codon FokI [176], was associ- metropolitan area in Leeds, this estimates that 30.0 to ated with severe DR in patients with type 1 diabetes. 43.2 people per million per year will become severely A few genome-wide studies have identified novel gene visually impaired due to DR and its sequelae [186]. In loci associated with DR [177–180]. Association of novel Fife, Scotland, between 2000 and 2009, the mean inci- genes related to vascular endothelium proliferation and dence of blindness (defined as above) was 13.8 per capillary permeability, such as PLXDC2 and ARHGAP22, million per year for the total population of the county imply that our understanding of angiogenic and inflam- [187]. In the Sankara Nethralaya Diabetic Retinopathy matory pathways is still incomplete [178]. Interestingly, Epidemiology and Molecular Genetics Study (SN- polymorphism of RP1-90 L14.1, a long intergenic non- DREAMS) in type 2 diabetes, the prevalence of visual coding RNA gene adjacent to CEP162 was found to be impairment and blindness was 4 and 0.1 %, respectively associated with susceptibility to DR [180]. Since CEP162 [188]. is a key protein in cell ciliogenesis [181], it raises the ques- tion if dysregulation of ciliary assembly plays a role in DR Other eye complications of diabetes pathogenesis. While DR and DME are the most important and well- studied diabetes-related eye complication, many patients Epidemiology of diabetes-related vision loss with diabetes are at risk of vision loss from other While treatment options such as pan-retinal laser photo- diabetes-related eye conditions that range from mild vi- coagulation can largely control neovascularization and sion impairment to blindness. Diabetes is associated with prevent blindness, these treatments cannot restore vision, early and rapid development of cataracts, and is hence a and in fact have vision-impairing effects of their own. In- major cause of visual impairment among patients with travitreal agents such as anti-vascular endothelial growth diabetes. The Singapore Malay Eye Study (SiMES) found factor (VEGF) agents do not fully restore vision in all patients with diabetes to be more likely to have cortical patients, and require frequent and costly doses for and posterior subcapsular cataracts [189]. In the WESDR Lee et al. Eye and Vision (2015) 2:17 Page 19 of 25 study and SN-DREAMS study, presence of cataracts were strong as in nephropathy [8]. In the Chennai Urban Rural significant factors contributing to visual impairment and Epidemiology Study, prevalence of coronary heart disease blindness in patients with diabetes [185, 188]. Many pa- was higher among patients with DR as compared to those tients with diabetes require cataract surgery at a relatively without DR [198]. An eight-year cohort study in Japan younger age. In the WESDR, 10-year cumulative incidence found that patients who developed signs of mild DR were of cataract surgery was 8 % in patients with type 1 diabetes already at higher risk of coronary heart disease or stroke and 25 % in patients with type 2 diabetes [190]. While [9]. Factoring presence of DR in the assessment of patients usually a surgical procedure with good outcomes, cataract with diabetes also improved risk assessment of silent myo- surgery is complicated in patients with diabetes as they cardial infarcts [199]. Presence of DR was also associated may develop DME after surgery [191]. with mortality from cardiovascular disease, especially if Although findings have been inconsistent, diabetes has there is concomitant nephropathy [200]. Literature relat- been found to be a risk factor for developing primary ing DR with peripheral vascular disease is sparse, but a glaucoma in some population-based studies [192]. For recent cross-sectional study in China found an association instance, SiMES found an association between ocular between presence of PDR with lower ankle-brachial index hypertension and diabetes, but not glaucoma [189]. and lower toe-brachial index [201]. Neovascular glaucoma, which is both a blinding and painful condition, can also arise from PDR. A recent Conclusions report found that 7.1 % of patients with PDR requiring As this review shows, the epidemiology of DR has been vitrectomy developed neovascular glaucoma 1 year after extensively studied. The use of a common grading system, surgery [193]. Epiretinal membranes, which can cause the ETDRS severity scale and its modifications, has facili- significant visual impairment, were also found to be tated standardized diagnosis and severity classification of more common among patients with diabetes that have DR in multiple epidemiologic studies, allowing compari- undergone cataract surgery [189]. sons of prevalence, incidence, progression and regression of DR. Review of literature published within the past five Relationship of DR and DME with diabetes related years consistently found higher DR prevalence in Western systemic complications countries compared to Middle-East and Asian countries. Microvascular complications Notable exceptions include Saudi Arabia and Singapore, Diabetic nephropathy is closely associated to DR and two of the most affluent countries in Asia, where DR preva- DME, as many of the pathologic processes affecting mi- lence is comparable to that observed in the US and UK. crovasculature in DR are likely to be causative of diabetic Given the increasing affluence of developing economies nephropathy as well. In a cross-sectional study in Korea, such as China and India, the healthcare burden of DR can compared to patients without DR, patients with DR had be expected to be on the uptrend in the decades ahead. 2.11 the odds (95 % CI 1.04–4.26) of having overt diabetic More recently, cross-sectional studies from developing nephropathy, defined as protein excretion of more than countries are being published. Understandably, the sample 300 mg per 24 h or albumin/creatinine ratio greater than sizes of these studies tend to be small, and few are 300 μg/mg [194]. Ischemic diabetic retinopathy, as evi- population-based. However, it is clear that while people in denced by capillary non-perfusion found on fundal fluor- developing countries are at lower risk of developing dia- escein angiogram, was found to be associated with betes, they have an equivalent if not higher risk of devel- progression of diabetic nephropathy. Patients with more oping DR upon onset of diabetes. While traditional causes than or equal to 10 optic disc areas of capillary non- of visual impairment and blindness in developing coun- perfusion had 6.64 times the risk of progression of tries such as cataracts and trachoma are declining, the nephropathy [195]. Increasing severity of DR was associ- prevalence of DR is growing. Gaps in the literature on ated with increasing severity of chronic kidney disease and theepidemiologyofDRincludethelackof population- based cohort studies investigating the incidence, pro- decreased estimated glomerular filtration rate [196]. In a 15-year follow-up study, development of overt nephropa- gression, and regression in Asian and developing-world thy (defined as above) was found to be associated with the populations. In contrast to DR, the epidemiology of DME is much development of DME [197]. Few studies related the devel- opment of neuropathy with DR. However, the SN- less well studied. Existing studies are split between the use DREAMS found an association between neuropathy and of two diagnostic criteria, one for DME and the other for CSME. Since the CSME criteria are substantially stricter visual-impairment in patients with diabetes [188]. than the DME criteria, direct comparisons between these Macrovascular complications studies cannot be made. The lack of a severity scale also precludes the study of progression and regression of The strength of association between DR and macrovascu- DME. The diagnosis of DME itself is more challenging lar complications, such as cardiovascular disease is just as Lee et al. Eye and Vision (2015) 2:17 Page 20 of 25 than DR. While DR can be diagnosed and classified ad- species; SiMES: Singapore Malay Eye Study; SN-DREAMS: Sankara Nethralaya Diabetic Retinopathy Epidemiology and Molecular Genetics Study; TGF- equately with the assessment of non-stereoscopic fundus β: Transforming growth factor beta; TNF-α: Tumor necrosis factor-α; photos, the diagnosis of DME using this same modality is UK: United Kingdom; UKPDS: United Kingdom prospective diabetes study; challenging as macular thickening is difficult to assess in USA: United States of America; VEGF: Vascular endothelial growth factor; VTDR: Vision-threatening diabetic retinopathy; WESDR: Wisconsin non-stereoscopic photographs. There is no consensus on epidemiologic study of diabetic retinopathy. OCT-based severity classification for DME. More research will have to be carried out to overcome these hurdles in Competing interests diagnosis and classification of DME. The authors declare that they have no competing interests. The investigation of risk factors has also revealed in- teresting considerations both in clinical practice and re- Authors’ contributions RL performed the literature review and drafted the manuscript. CS was search. Hyperglycemia remains the most important involved in drafting the manuscript and critically revising it. WTY critically modifiable risk factor for DR, and intensive glycemic revised the manuscript and gave final approval of the version to be control has been proven to have potent and long-lasting published. All authors read and approved the final manuscript. protective effects against development and progression Author details of DR and DME. As the evidence behind hypertension 1 Singapore Eye Research Institute, Singapore National Eye Centre, Singapore, and dyslipidemia as risk factors is weaker than in hyper- Singapore. Yong Loo Lin School of Medicine, National University of Singapore, Singapore, Singapore. Ophthalmology and Visual Sciences glycemia, intensive control of hypertension and dyslipid- Academic Clinical Program, Duke-NUS Graduate Medical School, Singapore, emia should not be sought solely on the basis to prevent Singapore. onset or progression of DR and DME, but taken in con- Received: 11 August 2015 Accepted: 1 September 2015 sideration of other complications (e.g. reduction in ne- phropathy and cardiovascular diseases). Among novel risk factors, increased serum adiponectin References and LPO were found to be associated with greater preva- 1. Cheung N, Mitchell P, Wong TY. Diabetic retinopathy. Lancet. lence of DR. Vitamin D deficiency has also been found to 2010;376(9735):124–36. 2. Bourne RR, Stevens GA, White RA, Smith JL, Flaxman SR, Price H, et al. be associated with DR, but more evidence is needed to Causes of vision loss worldwide, 1990–2010: a systematic analysis. Lancet ascertain efficacy of Vitamin D supplementation in the Glob Health. 2013;1(6):e339–49. prevention of DR. These novel risk factors are promising, 3. Yau JW, Rogers SL, Kawasaki R, Lamoureux EL, Kowalski JW, Bek T, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes but the findings that have been made in cross-sectional Care. 2012;35(3):556–64. studies have to be supported by consistent findings in pro- 4. Lightman S, Towler HM. Diabetic retinopathy. Clin Cornerstone. spective cohort studies. The relationship between these 2003;5(2):12–21. 5. Tong L, Vernon SA, Kiel W, Sung V, Orr GM. Association of macular factors and DME is unknown and is worth exploring. involvement with proliferative retinopathy in type 2 diabetes. Diabet Med. The association between DR and other vascular diseases 2001;18(5):388–94. are important areas of study. DR is strongly associated 6. He F, Xia X, Wu XF, Yu XQ, Huang FX. Diabetic retinopathy in predicting diabetic nephropathy in patients with type 2 diabetes and renal disease: a with nephropathy, which has significant burden on health- meta-analysis. Diabetologia. 2013;56(3):457–66. care systems due to the need for renal replacement ther- 7. Hagg S, Thorn LM, Putaala J, Liebkind R, Harjutsalo V, Forsblom CM, et al. apy. The presence of DR is also associated with vascular Incidence of stroke according to presence of diabetic nephropathy and severe diabetic retinopathy in patients with type 1 diabetes. Diabetes Care. diseases that are disabling, such as stroke and peripheral 2013;36(12):4140–6. vascular disease, or life threatening, such as myocardial 8. Mottl AK, Pajewski N, Fonseca V, Ismail-Beigi F, Chew E, Ambrosius WT, et al. infarction. Physicians and ophthalmologists should there- The degree of retinopathy is equally predictive for renal and macrovascular outcomes in the ACCORD Trial. J Diabetes Complications. 2014;28(6):874–9. fore be aware that patients with DR and DME are receiv- 9. Kawasaki R, Tanaka S, Tanaka S, Abe S, Sone H, Yokote K, et al. Risk of ing appropriate assessment and treatment for these cardiovascular diseases is increased even with mild diabetic retinopathy: the comorbidities. Japan Diabetes Complications Study. Ophthalmology. 2013;120(3):574–82. 10. Thomas RL, Dunstan FD, Luzio SD, Chowdhury SR, North RV, Hale SL, et al. Abbreviations Prevalence of diabetic retinopathy within a national diabetic retinopathy ACCORD: Action to control cardiovascular risk in diabetes; ADVANCE: Action screening service. Br J Ophthalmol. 2015;99(1):64–8. in diabetes and vascular disease; AGE: Advanced glycation end products; 11. Pedro RA, Ramon SA, Marc BB, Juan FB, Isabel MM. Prevalence and apM-1: Adipose most abundant gene transcript-1; CRP: C-reactive protein; relationship between diabetic retinopathy and nephropathy, and its risk CSME: Clinically significant macular edema; DCCT: Diabetes control and factors in the North-East of Spain, a population-based study. Ophthalmic complications trial; DCPD1987: Danish cohort of pediatric diabetes 1987; Epidemiol. 2010;17(4):251–65. DIRECT: Diabetic retinopathy candesartan trials; DME: Diabetic macular 12. Hautala N, Hannula V, Palosaari T, Ebeling T, Falck A. Prevalence of diabetic edema; DR: Diabetic retinopathy; DRCR.net: Diabetic retinopathy clinical retinopathy in young adults with type 1 diabetes since childhood: the Oulu research network; ETDRS: Early Treatment for diabetic retinopathy study; cohort study of diabetic retinopathy. Acta Ophthalmol. 2014;92(8):749–52. FIELD: Fenofibrate intervention and event-lowering in diabetes; 13. Bertelsen G, Peto T, Lindekleiv H, Schirmer H, Solbu MD, Toft I, et al. Tromso IL-6: Interleukin-6; OCT: Optical coherence tomography; OSA: Obstructive eye study: prevalence and risk factors of diabetic retinopathy. Acta sleep apnea; PDR: Proliferative diabetic retinopathy; PPARα: Peroxisome Ophthalmol. 2013;91(8):716–21. proliferator-activated receptor alpha; PKC: Protein kinase C; RAAS: Renin- 14. Knudsen LL, Lervang HH, Lundbye-Christensen S, Gorst-Rasmussen A. The angiotensin-aldosterone system; RAGE: Receptor of advanced glycation end north Jutland county diabetic retinopathy study: population characteristics. products; RASS: Renin-Angiotensin System Study; ROS: Reactive oxygen Br J Ophthalmol. 2006;90(11):1404–9. Lee et al. Eye and Vision (2015) 2:17 Page 21 of 25 15. Dedov I, Maslova O, Suntsov Y, Bolotskaia L, Milenkaia T, Besmertnaia L. 35. Sharew G, Ilako DR, Kimani K, Gelaw Y. Prevalence of diabetic retinopathy in Prevalence of diabetic retinopathy and cataract in adult patients with type Jimma University Hospital, Southwest Ethiopia. Ethiop Med J. 1 and type 2 diabetes in Russia. Rev Diabet Stud. 2009;6(2):124–9. 2013;51(2):105–13. 16. Roy MS, Klein R, O’Colmain BJ, Klein BE, Moss SE, Kempen JH. The 36. Raman R, Gupta A, Krishna S, Kulothungan V, Sharma T. Prevalence and risk prevalence of diabetic retinopathy among adult type 1 diabetic persons in factors for diabetic microvascular complications in newly diagnosed type II the United States. Arch Ophthalmol. 2004;122(4):546–51. diabetes mellitus. Sankara Nethralaya Diabetic Retinopathy Epidemiology 17. Kempen JH, O’Colmain BJ, Leske MC, Haffner SM, Klein R, Moss SE, et al. The and Molecular Genetic Study (SN-DREAMS, report 27). J Diabetes prevalence of diabetic retinopathy among adults in the United States. Arch Complications. 2012;26(2):123–8. Ophthalmol. 2004;122(4):552–63. 37. Kostev K, Rathmann W. Diabetic retinopathy at diagnosis of type 2 diabetes 18. Zhang X, Saaddine JB, Chou CF, Cotch MF, Cheng YJ, Geiss LS, et al. in the UK: a database analysis. Diabetologia. 2013;56(1):109–11. Prevalence of diabetic retinopathy in the United States, 2005–2008. JAMA. 38. Tapp RJ, Shaw JE, Harper CA, de Courten MP, Balkau B, McCarty DJ, et al. 2010;304(6):649–56. The prevalence of and factors associated with diabetic retinopathy in the Australian population. Diabetes Care. 2003;26(6):1731–7. 19. Kung K, Chow KM, Hui EM, Leung M, Leung SY, Szeto CC, et al. Prevalence of complications among Chinese diabetic patients in urban primary care 39. Wong TY, Cheung N, Tay WT, Wang JJ, Aung T, Saw SM, et al. Prevalence clinics: a cross-sectional study. BMC Fam Pract. 2014;15:8. and risk factors for diabetic retinopathy: the Singapore Malay Eye Study. 20. Jee D, Lee WK, Kang S. Prevalence and risk factors for diabetic retinopathy: Ophthalmology. 2008;115(11):1869–75. the Korea National Health and Nutrition Examination Survey 2008–2011. 40. Looker HC, Nyangoma SO, Cromie D, Olson JA, Leese GP, Black M, et al. Invest Ophthalmol Vis Sci. 2013;54(10):6827–33. Diabetic retinopathy at diagnosis of type 2 diabetes in Scotland. 21. Raman R, Rani PK, Reddi Rachepalle S, Gnanamoorthy P, Uthra S, Diabetologia. 2012;55(9):2335–42. Kumaramanickavel G, et al. Prevalence of diabetic retinopathy in India: 41. Klein R, Klein BK, Moss SE, Davis MD, DeMets DL. The wisconsin Sankara Nethralaya diabetic retinopathy epidemiology and molecular epidemiologic study of diabetic retinopathy: Ix. four-year incidence and genetics study report 2. Ophthalmology. 2009;116(2):311–8. progression of diabetic retinopathy when age at diagnosis is less than 30 years. Arch Ophthalmol. 1989;107(2):237–43. 22. Liu L, Wu X, Liu L, Geng J, Yuan Z, Shan Z, et al. Prevalence of diabetic retinopathy in mainland China: a meta-analysis. PLoS One. 42. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic 2012;7(9):e45264. Study of diabetic retinopathy. XIV. Ten-year incidence and progression of 23. Huang OS, Tay WT, Ong PG, Sabanayagam C, Cheng CY, Tan GS, et al. diabetic retinopathy. Arch Ophthalmol. 1994;112(9):1217–28. Prevalence and determinants of undiagnosed diabetic retinopathy and 43. Klein R, Klein BE, Moss SE, Cruickshanks KJ. The Wisconsin Epidemiologic vision-threatening retinopathy in a multiethnic Asian cohort: the Singapore Study of Diabetic Retinopathy: XVII. The 14-year incidence and progression Epidemiology of Eye Diseases (SEED) study. Br J Ophthalmol. 2015. of diabetic retinopathy and associated risk factors in type 1 diabetes. 24. Zheng Y, Lamoureux EL, Lavanya R, Wu R, Ikram MK, Wang JJ, et al. Ophthalmology. 1998;105(10):1801–15. Prevalence and risk factors of diabetic retinopathy in migrant Indians in an 44. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin urbanized society in Asia: the Singapore Indian eye study. Ophthalmology. Epidemiologic Study of Diabetic Retinopathy: XXII the twenty-five-year 2012;119(10):2119–24. progression of retinopathy in persons with type 1 diabetes. Ophthalmology. 2008;115(11):1859–68. 25. Rema M, Premkumar S, Anitha B, Deepa R, Pradeepa R, Mohan V. Prevalence of diabetic retinopathy in urban India: the Chennai Urban Rural 45. Broe R, Rasmussen ML, Frydkjaer-Olsen U, Olsen BS, Mortensen HB, Peto T, Epidemiology Study (CURES) eye study, I. Invest Ophthalmol Vis Sci. et al. The 16-year incidence, progression and regression of diabetic 2005;46(7):2328–33. retinopathy in a young population-based Danish cohort with type 1 26. Al Ghamdi AH, Rabiu M, Hajar S, Yorston D, Kuper H, Polack S. Rapid diabetes mellitus: The Danish cohort of pediatric diabetes 1987 (DCPD1987). assessment of avoidable blindness and diabetic retinopathy in Taif, Saudi Acta Diabetol. 2014;51(3):413–20. Arabia. Br J Ophthalmol. 2012;96(9):1168–72. 46. Younis N, Broadbent DM, Harding SP, Vora JP. Incidence of sight- 27. Papakonstantinou E, Tsinopoulos I, Dimitrakos S, Dehghan MH, Katibeh M, threatening retinopathy in type 1 diabetes in a systematic screening Ahmadieh H, et al. Prevalence and risk factors for diabetic retinopathy in programme. Diabet Med. 2003;20(9):758–65. the 40 to 80 year-old population in Yazd, Iran: the Yazd Eye Study. J 47. Martin-Merino E, Fortuny J, Rivero-Ferrer E, Garcia-Rodriguez LA. Incidence Diabetes. 2015;7(1):139–41. of retinal complications in a cohort of newly diagnosed diabetic patients. 28. Wang FH, Liang YB, Zhang F, Wang JJ, Wei WB, Tao QS, et al. Prevalence of PLoS One. 2014;9(6):e100283. diabetic retinopathy in rural China: the Handan Eye Study. Ophthalmology. 48. Kyto JP, Harjutsalo V, Forsblom C, Hietala K, Summanen PA, Groop PH. 2009;116(3):461–7. Decline in the cumulative incidence of severe diabetic retinopathy in patients with type 1 diabetes. Diabetes Care. 2011;34(9):2005–7. 29. Raman R,Ganesan S,PalSS,KulothunganV,SharmaT.Prevalenceand risk factors for diabetic retinopathy in rural India. Sankara Nethralaya 49. Thomas RL, Dunstan F, Luzio SD, Roy Chowdury S, Hale SL, North RV, et al. Diabetic Retinopathy Epidemiology and Molecular Genetic Study III Incidence of diabetic retinopathy in people with type 2 diabetes mellitus (SN-DREAMS III), report no 2. BMJ Open Diabetes Res Care. attending the diabetic retinopathy screening service for wales: retrospective 2014;2(1):e000005. analysis. BMJ. 2012;344:e874. 30. Katulanda P, Waniganayake YC, Ranasinghe P, Wijetunga WU, Jayaweera M, 50. Younis N, Broadbent DM, Vora JP, Harding SP. Incidence of sight- Wijesinghe NP, et al. Retinopathy among young adults with diabetes threatening retinopathy in patients with type 2 diabetes in the mellitus from a tertiary care setting in Sri Lanka. BMC Endocr Disord. Liverpool Diabetic Eye Study: a cohort study. Lancet. 2014;14:20. 2003;361(9353):195–200. 31. Akhter A, Fatema K, Ahmed SF, Afroz A, Ali L, Hussain A. Prevalence and 51. Stratton IM, Kohner EM, Aldington SJ, Turner RC, Holman RR, Manley SE, et associated risk indicators of retinopathy in a rural Bangladeshi al. UKPDS 50: risk factors for incidence and progression of retinopathy in population with and without diabetes. Ophthalmic Epidemiol. Type II diabetes over 6 years from diagnosis. Diabetologia. 2013;20(4):220–7. 2001;44(2):156–63. 32. Thapa R, Joshi DM, Rizyal A, Maharjan N, Joshi RD. Prevalence, risk factors 52. Jones CD, Greenwood RH, Misra A, Bachmann MO. Incidence and and awareness of diabetic retinopathy among admitted diabetic patients at progression of diabetic retinopathy during 17 years of a population-based a tertiary level hospital in Kathmandu. Nepal J Ophthalmol. 2014;6(11) screening program in England. Diabetes Care. 2012;35(3):592–6. doi:10.3126/nepjoph.v6i1.10760. 53. Tudor SM, Hamman RF, Baron A, Johnson DW, Shetterly SM. Incidence and 33. Kahloun R, Jelliti B, Zaouali S, Attia S, Ben Yahia S, Resnikoff S, et al. progression of diabetic retinopathy in Hispanics and non-Hispanic whites Prevalence and causes of visual impairment in diabetic patients in Tunisia, with type 2 diabetes. San Luis Valley Diabetes Study, Colorado. Diabetes North Africa. Eye (Lond). 2014;28(8):986–91. Care. 1998;21(1):53–61. 34. Mathenge W, Bastawrous A, Peto T, Leung I, Yorston D, Foster A, et al. 54. Varma R, Choudhury F, Klein R, Chung J, Torres M, Azen SP. Four-year Prevalence and correlates of diabetic retinopathy in a population-based incidence and progression of diabetic retinopathy and macular survey of older people in Nakuru, Kenya. Ophthalmic Epidemiol. edema: the Los Angeles Latino Eye Study. Am J Ophthalmol. 2014;21(3):169–77. 2010;149(5):752–61. e1-3. Lee et al. Eye and Vision (2015) 2:17 Page 22 of 25 55. Lee ET, Lee VS, Kingsley RM, Lu M, Russell D, Asal NR, et al. Diabetic 76. Kajiwara A, Miyagawa H, Saruwatari J, Kita A, Sakata M, Kawata Y, et al. retinopathy in Oklahoma Indians with NIDDM. Incidence and risk factors. Gender differences in the incidence and progression of diabetic Diabetes Care. 1992;15(11):1620–7. retinopathy among Japanese patients with type 2 diabetes mellitus: a 56. Cikamatana L, Mitchell P, Rochtchina E, Foran S, Wang JJ. Five-year clinic-based retrospective longitudinal study. Diabetes Res Clin Pract. incidence and progression of diabetic retinopathy in a defined older 2014;103(3):e7–10. population: the Blue Mountains Eye Study. Eye (Lond). 2007;21(4):465–71. 77. Romero-Aroca P, Baget-Bernaldiz M, Fernandez-Ballart J, Plana-Gil N, Soler- 57. Leske MC, Wu SY, Hennis A, Nemesure B, Schachat AP, Hyman L, et al. Nine- Lluis N, Mendez-Marin I, et al. Ten-year incidence of diabetic retinopathy year incidence of diabetic retinopathy in the Barbados Eye Studies. Arch and macular edema. Risk factors in a sample of people with type 1 Ophthalmol. 2006;124(2):250–5. diabetes. Diabetes Res Clin Pract. 2011;94(1):126–32. 58. Tapp RJ, Zimmet PZ, Harper CA, McCarty DJ, Chitson P, Tonkin AM, et al. Six 78. Olsen BS, Sjolie AK, Hougaard P, Johannesen J, Marinelli K, Jacobsen BB, et al. year incidence and progression of diabetic retinopathy: results from the The significance of the prepubertal diabetes duration for the development of Mauritius diabetes complication study. Diabetes Res Clin Pract. retinopathy and nephropathy in patients with type 1 diabetes. J Diabetes 2006;73(3):298–303. Complications. 2004;18(3):160–4. 79. Donaghue KC, Fairchild JM, Craig ME, Chan AK, Hing S, Cutler LR, et al. Do 59. Salinero-Fort MA, San Andres-Rebollo FJ, de Burgos-Lunar C, Arrieta-Blanco all prepubertal years of diabetes duration contribute equally to diabetes FJ, Gomez-Campelo P. Four-year incidence of diabetic retinopathy in a complications? Diabetes Care. 2003;26(4):1224–9. Spanish cohort: the MADIABETES study. PLoS One. 2013;8(10):e76417. 60. Tam VH, Lam EP, Chu BC, Tse KK, Fung LM. Incidence and progression of 80. Harjutsalo V, Maric C, Forsblom C, Thorn L, Waden J, Groop PH. Sex-related diabetic retinopathy in Hong Kong Chinese with type 2 diabetes mellitus. differences in the long-term risk of microvascular complications by age at J Diabetes Complications. 2009;23(3):185–93. onset of type 1 diabetes. Diabetologia. 2011;54(8):1992–9. 61. Song H, Liu L, Sum R, Fung M, Yap MK. Incidence of diabetic retinopathy in 81. Chen S, Jim B, Ziyadeh FN. Diabetic nephropathy and transforming growth a Hong Kong Chinese population. Clin Exp Optom. 2011;94(6):563–7. factor-beta: transforming our view of glomerulosclerosis and fibrosis build- up. Semin Nephrol. 2003;23(6):532–43. 62. Kawasaki R, Tanaka S, Tanaka S, Yamamoto T, Sone H, Ohashi Y, et al. 82. Egan AM, McVicker L, Heerey A, Carmody L, Harney F, Dunne FP. Diabetic Incidence and progression of diabetic retinopathy in Japanese adults with retinopathy in pregnancy: a population-based study of women with type 2 diabetes: 8 year follow-up study of the Japan Diabetes pregestational diabetes. J Diabetes Res. 2015;2015:310239. Complications Study (JDCS). Diabetologia. 2011;54(9):2288–94. doi:10.1155/2015/310239. 63. Fundus photographic risk factors for progression of diabetic retinopathy. ETDRS report number 12. Early Treatment Diabetic Retinopathy Study 83. Diabetes Control and Complications Trial Research Group. Effect of Research Group. Ophthalmology. 1991;98(5 Suppl):823–33. pregnancy on microvascular complications in the diabetes control and 64. Al-Rubeaan K, Abu El-Asrar AM, Youssef AM, Subhani SN, Ahmad NA, Al- complications trial. The Diabetes Control and Complications Trial Research Sharqawi AH, et al. Diabetic retinopathy and its risk factors in a society with Group. Diabetes Care. 2000;23(8):1084–91. a type 2 diabetes epidemic: a Saudi National Diabetes Registry-based study. 84. Ringholm L, Vestgaard M, Laugesen CS, Juul A, Damm P, Mathiesen ER. Acta Ophthalmol. 2015;93(2):e140–7. Pregnancy-induced increase in circulating IGF-I is associated with progression of diabetic retinopathy in women with type 1 diabetes. Growth 65. Pugliese G, Solini A, Zoppini G, Fondelli C, Zerbini G, Vedovato M, et al. Horm IGF Res. 2011;21(1):25–30. High prevalence of advanced retinopathy in patients with type 2 85. Kastelan S, Tomic M, Pavan J, Oreskovic S. Maternal immune system diabetes from the Renal Insufficiency and Cardiovascular adaptation to pregnancy–a potential influence on the course of diabetic Events (RIACE) Italian Multicenter Study. Diabetes Res Clin Pract. retinopathy. Reprod Biol Endocrinol. 2010;8:124. 2012;98(2):329–37. 66. Dutra Medeiros M, Mesquita E, Papoila AL, Genro V, Raposo JF. First diabetic 86. Wong TY, Liew G, Tapp RJ, Schmidt MI, Wang JJ, Mitchell P, et al. Relation retinopathy prevalence study in Portugal: RETINODIAB Study-Evaluation of between fasting glucose and retinopathy for diagnosis of diabetes: the screening programme for Lisbon and Tagus Valley region. Br J three population-based cross-sectional studies. Lancet. Ophthalmol. 2015 doi:10.1136/bjophthalmol-2015–306727 2008;371(9614):736–43. [Epub ahead of print]. 87. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy: a 67. Nathoo N, Ng M, Rudnisky CJ, Tennant MT. The prevalence of diabetic systematic review. JAMA. 2007;298(8):902–16. retinopathy as identified by teleophthalmology in rural Alberta. Can J 88. The effect of intensive treatment of diabetes on the development and Ophthalmol. 2010;45(1):28–32. progression of long-term complications in insulin-dependent diabetes 68. Esteves JF, Kramer CK, Azevedo MJ, Stolz AP, Roggia MF, Larangeira A, et al. mellitus. The Diabetes Control and Complications Trial Research Group. Prevalence of diabetic retinopathy in patients with type 1 diabetes mellitus. N Engl J Med. 1993;329(14):977–86. Rev Assoc Med Bras. 2009;55(3):268–73. 89. Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of intensive therapy. The Diabetes Control and Complications 69. Villena JE, Yoshiyama CA, Sanchez JE, Hilario NL, Merin LM. Prevalence of Trial/Epidemiology of Diabetes Interventions and Complications Research diabetic retinopathy in Peruvian patients with type 2 diabetes: results of a Group. N Engl J Med. 2000;342(6):381–9. hospital-based retinal telescreening program. Rev Panam Salud Publica. 2011;30(5):408–14. 90. Early worsening of diabetic retinopathy in the Diabetes Control and 70. Thomas RL, Distiller L, Luzio SD, Chowdhury SR, Melville VJ, Kramer B, et al. Complications Trial. Arch Ophthalmol. 1998;116(7):874–86. Ethnic differences in the prevalence of diabetic retinopathy in persons with 91. White NH, Sun W, Cleary PA, Danis RP, Davis MD, Hainsworth DP, et al. diabetes when first presenting at a diabetes clinic in South Africa. Diabetes Prolonged effect of intensive therapy on the risk of retinopathy Care. 2013;36(2):336–41. complications in patients with type 1 diabetes mellitus: 10 years after the 71. Kaidonis G, Mills RA, Landers J, Lake SR, Burdon KP, Craig JE. Review of the Diabetes Control and Complications Trial. Arch Ophthalmol. prevalence of diabetic retinopathy in Indigenous Australians. Clin 2008;126(12):1707–15. Experiment Ophthalmol. 2014;42(9):875–82. 92. Lachin JM, White NH, Hainsworth DP, Sun W, Cleary PA, Nathan DM. Effect of intensive diabetes therapy on the progression of diabetic retinopathy in 72. Virgili G, Menchini F, Murro V, Peluso E, Rosa F, Casazza G. Optical patients with type 1 diabetes: 18 years of follow-up in the DCCT/EDIC. coherence tomography (OCT) for detection of macular oedema in patients Diabetes. 2015;64(2):631–42. with diabetic retinopathy. Cochrane Database Syst Rev. 2011;7:CD008081. 73. Klein R, Knudtson MD, Lee KE, Gangnon R, Klein BE. The Wisconsin 93. Hirose A, Furushima D, Yamaguchi N, Kitano S, Uchigata Y. Prediction of Epidemiologic Study of Diabetic Retinopathy XXIII: the twenty-five-year retinopathy at 20 years after onset in younger-onset type 1 diabetes using incidence of macular edema in persons with type 1 diabetes. mean metabolic memory-free HbA1c values: the importance of using Ophthalmology. 2009;116(3):497–503. HbA1c data of total, not partial, diabetes duration. Diabetes Care. 74. Ding J, Wong TY. Current epidemiology of diabetic retinopathy and 2013;36(11):3812–4. diabetic macular edema. Curr Diab Rep. 2012;12(4):346–54. 94. Smith-Palmer J, Brandle M, Trevisan R, Orsini Federici M, Liabat S, Valentine W. Assessment of the association between glycemic variability and 75. Xu J, Xu L, Wang YX, You QS, Jonas JB, Wei WB. Ten-year cumulative diabetes-related complications in type 1 and type 2 diabetes. Diabetes Res incidence of diabetic retinopathy. The Beijing Eye Study 2001/2011. PLoS Clin Pract. 2014;105(3):273–84. One. 2014;9(10):e111320. Lee et al. Eye and Vision (2015) 2:17 Page 23 of 25 95. Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, et al. 117. Cheung D, Switzer NJ, Ehmann D, Rudnisky C, Shi X, Karmali S. The impact Intensive blood glucose control and vascular outcomes in patients with of bariatric surgery on diabetic retinopathy: a systematic review and meta- type 2 diabetes. N Engl J Med. 2008;358(24):2560–72. analysis. Obes Surg. 2015;25(9):1604–9. 96. Gerstein HC, Miller ME, Byington RP, Goff Jr DC, Bigger JT, Buse JB, et al. 118. Tang J, Kern TS. Inflammation in diabetic retinopathy. Prog Retin Eye Res. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2011;30(5):343–58. 2008;358(24):2545–59. 119. Schram MT, Chaturvedi N, Schalkwijk CG, Fuller JH, Stehouwer CD. Markers 97. Fullerton B, Jeitler K, Seitz M, Horvath K, Berghold A, Siebenhofer A. of inflammation are cross-sectionally associated with microvascular Intensive glucose control versus conventional glucose control for type 1 complications and cardiovascular disease in type 1 diabetes–the EURODIAB diabetes mellitus. Cochrane Database Syst Rev. 2014;2:CD009122. Prospective Complications Study. Diabetologia. 2005;48(2):370–8. 98. Jin P, Peng J, Zou H, Wang W, Fu J, Shen B, et al. The 5-year onset and 120. Nguyen TT, Alibrahim E, Islam FM, Klein R, Klein BE, Cotch MF, et al. regression of diabetic retinopathy in Chinese type 2 diabetes patients. PLoS Inflammatory, hemostatic, and other novel biomarkers for diabetic One. 2014;9(11):e113359. retinopathy: the multi-ethnic study of atherosclerosis. Diabetes Care. 2009;32(9):1704–9. 99. Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes: UKPDS 38. UK Prospective Diabetes Study 121. Lim LS, Tai ES, Mitchell P, Wang JJ, Tay WT, Lamoureux E, et al. C-reactive Group. BMJ. 1998;317(7160):703–13. protein, body mass index, and diabetic retinopathy. Invest Ophthalmol Vis 100. Holman RR, Paul SK, Bethel MA, Neil HA, Matthews DR. Long-term follow-up Sci. 2010;51(9):4458–63. after tight control of blood pressure in type 2 diabetes. N Engl J Med. 122. A randomized trial comparing intravitreal triamcinolone acetonide and 2008;359(15):1565–76. focal/grid photocoagulation for diabetic macular edema. Ophthalmology. 101. Chaturvedi N, Porta M, Klein R, Orchard T, Fuller J, Parving HH, et al. Effect of 2008;115(9):1447–9, 1149.e1-10. doi:10.1016/j.ophtha.2008.06.015. candesartan on prevention (DIRECT-Prevent 1) and progression (DIRECT- 123. Beck RW, Edwards AR, Aiello LP, Bressler NM, Ferris F, Glassman AR, et al. Protect 1) of retinopathy in type 1 diabetes: randomised, placebo-controlled Three-year follow-up of a randomized trial comparing focal/grid trials. Lancet. 2008;372(9647):1394–402. photocoagulation and intravitreal triamcinolone for diabetic macular edema. Arch Ophthalmol. 2009;127(3):245–51. 102. Sjolie AK, Klein R, Porta M, Orchard T, Fuller J, Parving HH, et al. Effect of candesartan on progression and regression of retinopathy in type 2 124. Angelo LS, Kurzrock R. Vascular endothelial growth factor and its diabetes (DIRECT-Protect 2): a randomised placebo-controlled trial. Lancet. relationship to inflammatory mediators. Clin Cancer Res. 2008;372(9647):1385–93. 2007;13(10):2825–30. 103. Do DV, Wang X, Vedula SS, Marrone M, Sleilati G, Hawkins BS, et al. Blood 125. Jardeleza MS, Miller JW. Review of anti-VEGF therapy in proliferative diabetic pressure control for diabetic retinopathy. Cochrane Database Syst Rev. retinopathy. Semin Ophthalmol. 2009;24(2):87–92. 2015;1:CD006127. 126. Colucciello M. Current intravitreal pharmacologic therapies for diabetic 104. Preiss D, Seshasai SR, Welsh P, Murphy SA, Ho JE, Waters DD, et al. Risk of macular edema. Postgrad Med. 2015;127(6):640–53. incident diabetes with intensive-dose compared with moderate-dose statin 127. Mitchell P, Bandello F, Schmidt-Erfurth U, Lang GE, Massin P, Schlingemann therapy: a meta-analysis. JAMA. 2011;305(24):2556–64. RO, et al. The RESTORE study: ranibizumab monotherapy or combined with 105. Sheely D, Jialal I. Strategies to lower low-density lipoprotein cholesterol in laser versus laser monotherapy for diabetic macular edema. Ophthalmology. metabolic syndrome: averting the diabetes risk. Metab Syndr Relat Disord. 2011;118(4):615–25. 2013;11(3):149–51. 128. Bressler SB, Almukhtar T, Bhorade A, Bressler NM, Glassman AR, Huang SS, et al. 106. Keech AC, Mitchell P, Summanen PA, O’Day J, Davis TM, Moffitt MS, et al. Repeated intravitreous ranibizumab injections for diabetic macular edema and Effect of fenofibrate on the need for laser treatment for diabetic the risk of sustained elevation of intraocular pressure or the need for ocular retinopathy (FIELD study): a randomised controlled trial. Lancet. hypotensive treatment. JAMA Ophthalmol. 2007;370(9600):1687–97. 2015;133(5):589–97. 107. Chew EY, Ambrosius WT, Davis MD, Danis RP, Gangaputra S, Greven CM, et al. 129. Wells JA, Glassman AR, Ayala AR, Jampol LM, Aiello LP, Antoszyk AN, et al. Effects of medical therapies on retinopathy progression in type 2 diabetes. N Aflibercept, bevacizumab, or ranibizumab for diabetic macular edema. Engl J Med. 2010;363(3):233–44. N Engl J Med. 2015;372(13):1193–203. 108. Massin P, Peto T, Ansquer JC, Aubonnet P, Macu FENSIFT. Effects of 130. Elman MJ, Aiello LP, Beck RW, Bressler NM, Bressler SB, Edwards AR, et al. fenofibric acid on diabetic macular edema: the MacuFen study. Ophthalmic Randomized trial evaluating ranibizumab plus prompt or deferred laser or Epidemiol. 2014;21(5):307–17. triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2010;117(6):1064–77. e35. 109. Elam M, Lovato LC, Ginsberg H. Role of fibrates in cardiovascular disease prevention, the ACCORD-Lipid perspective. Curr Opin Lipidol. 131. Elman MJ, Bressler NM, Qin H, Beck RW, Ferris 3rd FL, Friedman SM, et al. 2011;22(1):55–61. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or 110. Cheung N, Wong TY. Obesity and eye diseases. Surv Ophthalmol. triamcinolone plus prompt laser for diabetic macular edema. 2007;52(2):180–95. Ophthalmology. 2011;118(4):609–14. 111. Henricsson M, Nystrom L, Blohme G, Ostman J, Kullberg C, Svensson M, et al. 132. Wozniak SE, Gee LL, Wachtel MS, Frezza EE. Adipose tissue: the new The incidence of retinopathy 10 years after diagnosis in young adult people endocrine organ? A review article. Dig Dis Sci. 2009;54(9):1847–56. with diabetes: results from the nationwide population-based Diabetes 133. Yamagishi S, Inagaki Y, Amano S, Okamoto T, Takeuchi M. Up-regulation of Incidence Study in Sweden (DISS). Diabetes Care. vascular endothelial growth factor and down-regulation of pigment 2003;26(2):349–54. epithelium-derived factor messenger ribonucleic acid levels in 112. Chaturvedi N, Sjoelie AK, Porta M, Aldington SJ, Fuller JH, Songini M, et al. leptin-exposed cultured retinal pericytes. Int J Tissue React. Markers of insulin resistance are strong risk factors for retinopathy incidence 2002;24(4):137–42. in type 1 diabetes. Diabetes Care. 2001;24(2):284–9. 134. Suganami E, Takagi H, Ohashi H, Suzuma K, Suzuma I, Oh H, et al. Leptin 113. Lu J, Hou X, Zhang L, Jiang F, Hu C, Bao Y, et al. Association between body stimulates ischemia-induced retinal neovascularization: possible role of mass index and diabetic retinopathy in Chinese patients with type 2 vascular endothelial growth factor expressed in retinal endothelial cells. diabetes. Acta Diabetol. 2015;52(4):701–8. Diabetes. 2004;53(9):2443–8. 114. Rooney D, Lye WK, Tan G, Lamoureux EL, Ikram MK, Cheng CY, et al. Body 135. Gariano RF, Nath AK, D’Amico DJ, Lee T, Sierra-Honigmann MR. Elevation of mass index and retinopathy in Asian populations with diabetes mellitus. vitreous leptin in diabetic retinopathy and retinal detachment. Invest Acta Diabetol. 2015;52(1):73–80. Ophthalmol Vis Sci. 2000;41(11):3576–81. 115. Rudrappa S, Warren G, Idris I. Obstructive sleep apnoea is associated with 136. Dossarps D, Petit JM, Guiu B, Cercueil JP, Duvillard L, Bron AM, et al. Body the development and progression of diabetic retinopathy, independent of fat distribution and adipokine secretion are not associated with diabetic conventional risk factors and novel biomarkers for diabetic retinopathy. Br J retinopathy in patients with type 2 diabetes mellitus. Ophthalmic Res. Ophthalmol. 2012;96(12):1535. 2014;51(1):42–5. 116. Mason RH, West SD, Kiire CA, Groves DC, Lipinski HJ, Jaycock A, et al. High 137. Sari R, Balci MK, Apaydin C. The relationship between plasma leptin levels prevalence of sleep disordered breathing in patients with diabetic macular and chronic complication in patients with type 2 diabetes mellitus. Metab edema. Retina. 2012;32(9):1791–8. Syndr Relat Disord. 2010;8(6):499–503. Lee et al. Eye and Vision (2015) 2:17 Page 24 of 25 138. Omae T, Nagaoka T, Tanano I, Yoshida A. Adiponectin-induced dilation of 160. Shimo N, Yasuda T, Kaneto H, Katakami N, Kuroda A, Sakamoto F, et al. isolated porcine retinal arterioles via production of nitric oxide from Vitamin D deficiency is significantly associated with retinopathy in young endothelial cells. Invest Ophthalmol Vis Sci. 2013;54(7):4586–94. Japanese type 1 diabetic patients. Diabetes Res Clin Pract. 139. Omae T, Nagaoka T, Yoshida A. Relationship between retinal blood flow 2014;106(2):e41–3. and serum adiponectin concentrations in patients with type 2 diabetes 161. He R, Shen J, Liu F, Zeng H, Li L, Yu H, et al. Vitamin D deficiency increases mellitus. Invest Ophthalmol Vis Sci. 2015;56(6):4143–9. the risk of retinopathy in Chinese patients with type 2 diabetes. Diabet 140. Srinivasan V, Sulochana KN. Effect of adiponectin on expression of vascular Med. 2014;31(12):1657–64. endothelial growth factor and pigment epithelium-derived factor: an in vitro 162. Jee D, Han K, Kim EC. Inverse association between high blood 25- study. Indian J Pharmacol. 2015;47(1):117–20. hydroxyvitamin D levels and diabetic retinopathy in a representative Korean 141. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V. population. PLoS One. 2014;9(12):e115199. Association of serum adiponectin with diabetic microvascular complications 163. Alcubierre N, Valls J, Rubinat E, Cao G, Esquerda A, Traveset A, et al. Vitamin among south Indian type 2 diabetic subjects - (CURES-133). Clin Biochem. D deficiency is associated with the presence and severity of diabetic 2015;48(1–2):33–8. retinopathy in type 2 diabetes mellitus. J Diabetes Res. 2015;2015:374178. 142. Zietz B, Buechler C, Kobuch K, Neumeier M, Scholmerich J, Schaffler A. 164. Patel S, Chen H, Tinkham NH, Zhang K. Genetic susceptibility of diabetic Serum levels of adiponectin are associated with diabetic retinopathy and retinopathy. Curr Diab Rep. 2008;8(4):257–62. with adiponectin gene mutations in Caucasian patients with diabetes 165. Hallman DM, Huber Jr JC, Gonzalez VH, Klein BE, Klein R, Hanis CL. Familial mellitus type 2. Exp Clin Endocrinol Diabetes. 2008;116(9):532–6. aggregation of severity of diabetic retinopathy in Mexican Americans from 143. Yilmaz MI, Sonmez A, Acikel C, Celik T, Bingol N, Pinar M, et al. Adiponectin Starr County, Texas. Diabetes Care. 2005;28(5):1163–8. may play a part in the pathogenesis of diabetic retinopathy. Eur J 166. Hietala K, Forsblom C, Summanen P, Groop PH. Heritability of proliferative Endocrinol. 2004;151(1):135–40. diabetic retinopathy. Diabetes. 2008;57(8):2176–80. 144. Madsen-Bouterse SA, Kowluru RA. Oxidative stress and diabetic retinopathy: 167. Arar NH, Freedman BI, Adler SG, Iyengar SK, Chew EY, Davis MD, et al. pathophysiological mechanisms and treatment perspectives. Rev Endocr Heritability of the severity of diabetic retinopathy: the FIND-Eye study. Invest Metab Disord. 2008;9(4):315–27. Ophthalmol Vis Sci. 2008;49(9):3839–45. 145. Robison Jr WG, Jacot JL, Katz ML, Glover JP. Retinal vascular changes 168. Clustering of long-term complications in families with diabetes in the induced by the oxidative stress of alpha-tocopherol deficiency contrasted diabetes control and complications trial. The Diabetes Control and with diabetic microangiopathy. J Ocul Pharmacol Ther. Complications Trial Research Group. Diabetes. 1997;46(11):1829–39. 2000;16(2):109–20. 169. Wang H, Cheng JW, Zhu LS, Wei RL, Cai JP, Li Y, et al. Meta-analysis of 146. Kowluru RA. Diabetic retinopathy: mitochondrial dysfunction and retinal association between the -2578C/A polymorphism of the vascular capillary cell death. Antioxid Redox Signal. 2005;7(11–12):1581–87. endothelial growth factor and retinopathy in type 2 diabetes in Asians and 147. Naruse R, Suetsugu M, Terasawa T, Ito K, Hara K, Takebayashi K, et al. Caucasians. Ophthalmic Res. 2014;52(1):1–8. Oxidative stress and antioxidative potency are closely associated with 170. Han L, Zhang L, Xing W, Zhuo R, Lin X, Hao Y, et al. The associations diabetic retinopathy and nephropathy in patients with type 2 diabetes. between VEGF gene polymorphisms and diabetic retinopathy susceptibility: Saudi Med J. 2013;34(2):135–41. a meta-analysis of 11 case–control studies. J Diabetes Res. 148. Wu Y, Tang L, Chen B. Oxidative stress: implications for the development of 2014;2014:805801. diabetic retinopathy and antioxidant therapeutic perspectives. Oxid Med 171. El-Shazly SF, El-Bradey MH, Tameesh MK. Vascular endothelial growth factor Cell Longev. 2014;2014:752387. gene polymorphism prevalence in patients with diabetic macular oedema 149. Yamagishi S, Matsui T. Advanced glycation end products (AGEs), oxidative and its correlation with anti-vascular endothelial growth factor treatment stress and diabetic retinopathy. Curr Pharm Biotechnol. 2011;12(3):362–8. outcomes. Clin Experiment Ophthalmol. 2014;42(4):369–78. 150. Oates PJ, Mylari BL. Aldose reductase inhibitors: therapeutic implications for 172. Kaidonis G, Burdon KP, Gillies MC, Abhary S, Essex RW, Chang JH, et al. diabetic complications. Expert Opin Investig Drugs. 1999;8(12):2095–119. Common Sequence Variation in the VEGFC Gene Is Associated with 151. PKC-DRS Study Group. The effect of ruboxistaurin on visual loss in patients Diabetic Retinopathy and Diabetic Macular Edema. Ophthalmology. 2015. with moderately severe to very severe nonproliferative diabetic retinopathy: doi:10.1016/j.ophtha.2015.05.004. initial results of the Protein Kinase C beta Inhibitor Diabetic Retinopathy 173. Zhou M, Zhang P, Xu X, Sun X. The relationship between aldose reductase Study (PKC-DRS) multicenter randomized clinical trial. Diabetes. C106T polymorphism and diabetic retinopathy: an updated meta-analysis. 2005;54(7):2188–97. Invest Ophthalmol Vis Sci. 2015;56(4):2279–89. 152. Aiello LP, Davis MD, Girach A, Kles KA, Milton RC, Sheetz MJ, et al. Effect of 174. El-Bab MF, Zaki NS, Mojaddidi MA, Al-Barry M, El-Beshbishy HA. Diabetic ruboxistaurin on visual loss in patients with diabetic retinopathy. retinopathy is associated with oxidative stress and mitigation of gene Ophthalmology. 2006;113(12):2221–30. expression of antioxidant enzymes. Int J Gen Med. 2013;6:799–806. 153. Bolton WK, Cattran DC, Williams ME, Adler SG, Appel GB, Cartwright K, et al. 175. Taverna MJ, Sola A, Guyot-Argenton C, Pacher N, Bruzzo F, Slama G, et al. Randomized trial of an inhibitor of formation of advanced glycation end Taq I polymorphism of the vitamin D receptor and risk of severe diabetic products in diabetic nephropathy. Am J Nephrol. 2004;24(1):32–40. retinopathy. Diabetologia. 2002;45(3):436–42. 154. Brzovic-Saric V, Landeka I, Saric B, Barberic M, Andrijasevic L, Cerovski B, et al. 176. Taverna MJ, Selam JL, Slama G. Association between a protein Levels of selected oxidative stress markers in the vitreous and serum of polymorphism in the start codon of the vitamin D receptor gene and diabetic retinopathy patients. Mol Vis. 2015;21:649–64. severe diabetic retinopathy in C-peptide-negative type 1 diabetes. J Clin 155. Pan HZ, Zhang H, Chang D, Li H, Sui H. The change of oxidative stress Endocrinol Metab. 2005;90(8):4803–8. products in diabetes mellitus and diabetic retinopathy. Br J Ophthalmol. 177. Grassi MA, Tikhomirov A, Ramalingam S, Below JE, Cox NJ, Nicolae DL. 2008;92(4):548–51. Genome-wide meta-analysis for severe diabetic retinopathy. Hum Mol Genet. 2011;20(12):2472–81. 156. Pludowski P, Holick MF, Pilz S, Wagner CL, Hollis BW, Grant WB, et al. Vitamin D effects on musculoskeletal health, immunity, autoimmunity, 178. Huang YC, Lin JM, Lin HJ, Chen CC, Chen SY, Tsai CH, et al. Genome-wide cardiovascular disease, cancer, fertility, pregnancy, dementia and mortality-a association study of diabetic retinopathy in a Taiwanese population. review of recent evidence. Autoimmun Rev. 2013;12(10):976–89. Ophthalmology. 2011;118(4):642–8. 157. Albert DM, Scheef EA, Wang S, Mehraein F, Darjatmoko SR, Sorenson CM, et al. 179. Sheu WH, Kuo JZ, Lee IT, Hung YJ, Lee WJ, Tsai HY, et al. Genome-wide Calcitriol is a potent inhibitor of retinal neovascularization. Invest Ophthalmol association study in a Chinese population with diabetic retinopathy. Hum Vis Sci. 2007;48(5):2327–34. Mol Genet. 2013;22(15):3165–73. 158. Ren Z, Li W, Zhao Q, Ma L, Zhu J. The impact of 1,25-dihydroxy vitamin D3 180. Awata T, Yamashita H, Kurihara S, Morita-Ohkubo T, Miyashita Y, Katayama S, on the expressions of vascular endothelial growth factor and transforming et al. A genome-wide association study for diabetic retinopathy in a growth factor-beta(1) in the retinas of rats with diabetes. Diabetes Res Clin Japanese population: potential association with a long intergenic non- Pract. 2012;98(3):474–80. coding RNA. PLoS One. 2014;9(11):e111715. 159. Kaur H, Donaghue KC, Chan AK, Benitez-Aguirre P, Hing S, Lloyd M, et al. 181. Wang WJ, Tay HG, Soni R, Perumal GS, Goll MG, Macaluso FP, et al. CEP162 Vitamin D deficiency is associated with retinopathy in children and is an axoneme-recognition protein promoting ciliary transition zone adolescents with type 1 diabetes. Diabetes Care. 2011;34(6):1400–2. assembly at the cilia base. Nat Cell Biol. 2013;15(6):591–601. Lee et al. Eye and Vision (2015) 2:17 Page 25 of 25 182. Bourne RR, Jonas JB, Flaxman SR, Keeffe J, Leasher J, Naidoo K, et al. 203. Chiang PP, Lamoureux EL, Cheung CY, Sabanayagam C, Wong W, Tai ES, et al. Prevalence and causes of vision loss in high-income countries and in Racial differences in the prevalence of diabetes but not diabetic retinopathy in a Eastern and Central Europe: 1990–2010. Br J Ophthalmol. 2014;98(5):629–38. multi-ethnic Asian population. Invest Ophthalmol Vis Sci. 2011;52(10):7586–92. 183. Keeffe J, Taylor HR, Fotis K, Pesudovs K, Flaxman SR, Jonas JB, et al. 204. Zheng Y, Lamoureux EL, Chiang PC, Anuar AR, Ding J, Wang JJ, et al. Prevalence and causes of vision loss in Southeast Asia and Oceania: Language barrier and its relationship to diabetes and diabetic retinopathy. 1990–2010. Br J Ophthalmol. 2014;98(5):586–91. BMC Public Health. 2012;12:781. 184. Naidoo K, Gichuhi S, Basanez MG, Flaxman SR, Jonas JB, Keeffe J, et al. 205. Katulanda P, Ranasinghe P, Jayawardena R. Prevalence of retinopathy Prevalence and causes of vision loss in sub-Saharan Africa: 1990–2010. Br J among adults with self-reported diabetes mellitus: the Sri Lanka diabetes Ophthalmol. 2014;98(5):612–8. and Cardiovascular Study. BMC Ophthalmol. 2014;14:100. 185. Klein R, Lee KE, Gangnon RE, Klein BE. The 25-year incidence of visual 206. Schellini SA, Carvalho GM, Rendeiro FS, Padovani CR, Hirai FE. Prevalence of impairment in type 1 diabetes mellitus the wisconsin epidemiologic study diabetes and diabetic retinopathy in a Brazilian population. Ophthalmic of diabetic retinopathy. Ophthalmology. 2010;117(1):63–70. Epidemiol. 2014;21(1):33–8. 207. Papali’i-Curtin AT, Dalziel DM. Prevalence of diabetic retinopathy and 186. Arora S, Kolb S, Goyder E, McKibbin M. Trends in the incidence of visual maculopathy in Northland, New Zealand: 2011–2012. N Z Med J. impairment certification secondary to diabetic retinopathy in the Leeds 2013;126(1383):20–8. metropolitan area, 2005–2010. Diabet Med. 2012;29(7):e112–6. 208. Win Tin ST, Kenilorea G, Gadabu E, Tasserei J, Colagiuri R. The prevalence of 187. Hall HN, Chinn DJ, Sinclair A, Styles CJ. Epidemiology of blindness diabetes complications and associated risk factors in Pacific Islands attributable to diabetes in Scotland: change over 20 years in a defined countries. Diabetes Res Clin Pract. 2014;103(1):114–8. population. Diabet Med. 2013;30(11):1349–54. 209. Tsugawa Y, Takahashi O, Meigs JB, Davis RB, Imamura F, Fukui T, et al. New 188. Rani PK, Raman R, Gella L, Kulothungan V, Sharma T. Prevalence of visual diabetes diagnostic threshold of hemoglobin A(1c) and the 3-year impairment and associated risk factors in subjects with type II diabetes incidence of retinopathy. Diabetes. 2012;61(12):3280–4. mellitus: sankara nethralaya diabetic retinopathy epidemiology and 210. Ahmed KR, Karim MN, Bhowmik B, Habib SH, Bukht MS, Ali L, et al. molecular genetics study (SN-DREAMS, report 16). Middle East Afr J Incidence of diabetic retinopathy in Bangladesh: a 15-year follow-up study. Ophthalmol. 2012;19(1):129–34. J Diabetes. 2012;4(4):386–91. 189. Chiang PP, Lamoureux EL, Zheng Y, Tay WT, Mitchell P, Wang JJ, et al. 211. Manaviat MR, Rashidi M, Afkhami-Ardekani M. Four years incidence of Frequency and risk factors of non-retinopathy ocular conditions in people diabetic retinopathy and effective factors on its progression in type II with diabetes: the Singapore Malay Eye Study. Diabet Med. diabetes. Eur J Ophthalmol. 2008;18(4):572–7. 2013;30(2):e32–40. 212. Janghorbani M, Amini M, Ghanbari H, Safaiee H. Incidence of and risk 190. Klein BE, Klein R, Moss SE. Incidence of cataract surgery in the Wisconsin factors for diabetic retinopathy in Isfahan, Iran. Ophthalmic Epidemiol. Epidemiologic Study of Diabetic Retinopathy. Am J Ophthalmol. 2003;10(2):81–95. 1995;119(3):295–300. 213. Perol J, Balkau B, Guillausseau PJ, Massin P. A study of the 3-year incidence 191. Baker CW, Almukhtar T, Bressler NM, Glassman AR, Grover S, Kim SJ, et al. of diabetic retinopathy in a French diabetic population seen at Lariboisiere Macular edema after cataract surgery in eyes without preoperative Hospital, Paris. Diabetes Metab. 2012;38(3):225–9. central-involved diabetic macular edema. JAMA Ophthalmol. 214. Xie XW, Xu L, Wang YX, Jonas JB. Prevalence and associated factors of 2013;131(7):870–9. diabetic retinopathy. The Beijing Eye Study 2006. Graefes Arch Clin Exp 192. Jeganathan VS, Wang JJ, Wong TY. Ocular associations of diabetes other Ophthalmol. 2008;246(11):1519–26. than diabetic retinopathy. Diabetes Care. 2008;31(9):1905–12. 215. Varma R, Bressler NM, Doan QV, Gleeson M, Danese M, Bower JK, et al. 193. Goto A, Inatani M, Inoue T, Awai-Kasaoka N, Takihara Y, Ito Y, et al. Prevalence of and risk factors for diabetic macular edema in the United Frequency and risk factors for neovascular glaucoma after vitrectomy in States. JAMA Ophthalmol. 2014;132(11):1334–40. eyes with proliferative diabetic retinopathy. J Glaucoma. 2013;22(7):572–6. 216. Petrella RJ, Blouin J, Davies B, Barbeau M. Prevalence, demographics, and 194. Lee WJ, Sobrin L, Lee MJ, Kang MH, Seong M, Cho H. The relationship treatment characteristics of visual impairment due to diabetic macular between diabetic retinopathy and diabetic nephropathy in a population- edema in a representative Canadian cohort. J Ophthalmol. based study in Korea (KNHANES V-2, 3). Invest Ophthalmol Vis Sci. 2012;2012:159167. 2014;55(10):6547–53. 195. Lee WJ, Sobrin L, Kang MH, Seong M, Kim YJ, Yi JH, et al. Ischemic diabetic retinopathy as a possible prognostic factor for chronic kidney disease progression. Eye (Lond). 2014;28(9):1119–25. 196. Grunwald JE, Alexander J, Ying GS, Maguire M, Daniel E, Whittock-Martin R, et al. Retinopathy and chronic kidney disease in the Chronic Renal Insufficiency Cohort (CRIC) study. Arch Ophthalmol. 2012;130(9):1136–44. 197. Romero P, Baget M, Mendez I, Fernandez J, Salvat M, Martinez I. Diabetic macular edema and its relationship to renal microangiopathy: a sample of Type I diabetes mellitus patients in a 15-year follow-up study. J Diabetes Complications. 2007;21(3):172–80. 198. Pradeepa R, Surendar J, Indulekha K, Chella S, Anjana RM, Mohan V. Relationship of diabetic retinopathy with coronary artery disease in Asian Indians with type 2 diabetes: the Chennai Urban Rural Epidemiology Study (CURES) Eye Study–3. Diabetes Technol Ther. 2015;17(2):112–8. 199. Cosson E, Nguyen MT, Chanu B, Balta S, Takbou K, Valensi P. The report of Submit your next manuscript to BioMed Central male gender and retinopathy status improves the current consensus and take full advantage of: guidelines for the screening of myocardial ischemia in asymptomatic type 2 diabetic patients. Nutr Metab Cardiovasc Dis. 2013;23(6):557–65. • Convenient online submission 200. Ricardo AC, Grunwald JE, Parvathaneni S, Goodin S, Ching A, Lash JP. • Thorough peer review Retinopathy and CKD as predictors of all-cause and cardiovascular mortality: National Health and Nutrition Examination Survey (NHANES) 1988–1994. Am • No space constraints or color figure charges J Kidney Dis. 2014;64(2):198–203. • Immediate publication on acceptance 201. Chen YW, Wang YY, Zhao D, Yu CG, Xin Z, Cao X, et al. High prevalence of lower extremity peripheral artery disease in type 2 diabetes patients with • Inclusion in PubMed, CAS, Scopus and Google Scholar proliferative diabetic retinopathy. PLoS One. 2015;10(3):e0122022. • Research which is freely available for redistribution 202. Abougalambou SS, Abougalambou AS. Risk factors associated with diabetic retinopathy among type 2 diabetes patients at teaching hospital in Submit your manuscript at Malaysia. Diabetes Metab Syndr. 2015;9(2):98–103. www.biomedcentral.com/submit

Journal

Eye and VisionSpringer Journals

Published: Sep 30, 2015

There are no references for this article.