Journal of Diabetes and Its Complications 28 (2014) 291–297

Contents lists available at ScienceDirect

Journal of Diabetes and Its Complications

j ourna l homepage: WWW.JDCJOURNAL.COM

Prevalence and risk factors for diabetic retinopathy in Asian Indians with young onsetType 1 and Type 2 Diabetes

Ramachandran Rajalakshmi a, Anandakumar Amutha a, Harish Ranjani a, Mohammed K. Ali b,Ranjit Unnikrishnan a, Ranjit Mohan Anjana a, K.M. Venkat Narayan b, Viswanathan Mohan a,⁎a Madras Diabetes Research Foundation & Dr. Mohan’s Diabetes Specialities Centre, WHO Collaborating Centre for Non-communicable Diseases Prevention and Control,IDF Centre for Education, Gopalapuram, Chennai, Indiab Hubert Department of Global Health, Rollins School of Public Health, Emory University, Atlanta, GA

Conflict of interest: None declared.⁎ Corresponding author at: Madras Diabetes Resear

Diabetes Specialities Centre, WHO Collaborating CeDiseases Prevention And Control &IDF Centre For EduGopalapuram, Chennai - 600 086. India. Tel.: +91 44 40935.

E-mail address: drmohans@diabetes.ind.in (V. MohaURL: http://www.drmohansdiabetes.com, http://ww

1056-8727/$ – see front matter © 2014 Elsevier Inc. Alhttp://dx.doi.org/10.1016/j.jdiacomp.2013.12.008

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 26 September 2013Received in revised form 29 November 2013Accepted 23 December 2013Available online 6 January 2014

Keywords:Diabetic retinopathyT1DM-YT2DM-YPrevalenceRisk factorsAsian Indians

Aim: To assess the prevalence and risk factors for diabetic retinopathy (DR) in people with young onset type 1(T1DM-Y) and type 2 diabetes (T2DM-Y).Methods: T1DM-Y(n = 150) and T2DM-Y(n = 150) participants, age between 10 and 25 years at diagnosis,had a complete clinical evaluation, biochemical assessment, and four field digital retinal colour photography.The Early Treatment Diabetic Retinopathy Study grading system was used to grade DR. Proliferative diabeticretinopathy (PDR) and diabetic macular edema (DME) were considered as sight threatening DR.Results: The prevalence of any DR was 53.3% [95% CI 45.3–61.3] in T1DM-Y (duration of diabetes: 12.4 ±7.4years) and 52.7% [44.7–60.7] in T2DM-Y (11.8 ± 8.3 years). The age and gender adjusted prevalence of DR,DME and PDRwas 62.5%, 10% and 7.3% in T1DM-Y, whereas it was 65.8%,12.7% and 9.3% in T2DM-Y respectively.In multivariable logistic regression, diabetes duration [Odds ratio (OR) 1.99 per 5 years; CI 1.42–2.79], waistcircumference [1.28 per 5 cm;1.05–1.56] andmicroalbuminuria [2.39 per 50 μg;1.07–5.31]were associatedwithDR in T1DM-Y, and diabetes duration [2.21 per 5 years; 1.61–3.02], diastolic blood pressure [1.54 per

5 mmHg;1.18–2.02], Glycated hemoglobin [1.37 per %;1.07–1.75] and lower stimulated C-peptide [1.54 per0.5 pmol/ml;1.15–2.05;] were associated with DR in T2DM-Y.Conclusion: Over half of the people with young-onset diabetes, regardless of type, have retinopathy within10–12 years of diabetes duration, emphasizing the need for regular eye screening and aggressive control ofglucose and blood pressure to prevent DR.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

The growing burden of diabetes globally has been accompanied bya rise in numbers of young onset type 1diabetes (T1DM-Y) andparticularly young onset type 2 diabetes (T2DM-Y) (AmericanDiabetes Association, 2000; Amutha et al., 2011; InternationalDiabetes Federation, 2011; McMahon et al., 2004; Mohan, Jaydip, &Deepa, 2007; Pinhas-Hamiel & Zeitler, 2005). Younger age of onset ofdiabetes results in a longer life time exposure to hyperglycemia andconsequently a greater risk of developing complications duringproductive years of adulthood (Amutha, Datta, Unnikrishnan, Anjana,

ch Foundation & Dr. Mohan’sntre For Non-Communicablecation, 4, Conran Smith Road,396 8888; fax: +91 44 2835

n).w.mdrf.in (V. Mohan).

l rights reserved.

& Mohan, 2012; Amutha et al., 2011; Hillier & Pedula, 2003; Pinhas-Hamiel & Zeitler, 2007).

Diabetic retinopathy (DR) is a potentially sight-threatening micro-vascular complication of diabetes, and an important cause ofpreventable blindness. While disease duration and levels of glucoseand blood pressure control strongly influence the risk of DR(Chaturvedi et al., 1998; Klein, Klein, Moss, Davis, & DeMets, 1984a;Raman et al., 2009; Rema et al., 2005; The DCCT Research Group,1993), some data indicate that younger age of onset, especially forT2DM, may confer added susceptibility to DR (Eppens et al., 2006;Wong, Molyneaux, Constantino, Twigg, & Yue, 2008). In the multi-ethnic SEARCH study in the US, the prevalence of DR was 17% in T1DMand 42% in T2DM (Mayer-Davis et al., 2012). Although diabetes in theyoung (particularly T2DM) is becoming more frequent in India(Amutha et al., 2011; Mohan et al., 2007), epidemiological data on DRin India are largely limited to adult onset T2DM (Raman et al., 2009;Rema et al., 2005), and to a few studies in T1DM-Y (Rema, Mohan, &Ponnaiya, 1995; Rema, Mohan, Ramachandran, & Viswanathan, 1989).Data in people with T2DM-Y are scarce.

292 R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

We assessed the prevalence and risk factors for retinopathy inpeople with T1DM-Y and T2DM-Y, with onset of diabetes betweenages 10 and 25 years, in a clinic population in India.

2. Research design and methods

Data for this study were collected between January 2010 andJune 2011 from individuals with T1DM-Y (n = 150) and T2DM-Y(n = 150), diagnosed between the ages of 10 and 25 years, withduration of diabetes over 2 years, registered at Dr. Mohan’s DiabetesSpecialties Centre (DMDSC),a tertiary care network in Chennai(formerly Madras) in southern India. The Institutional EthicsCommittee (IEC) approval was obtained prior to the start of thestudy. Written informed consent was obtained according to the localIEC guidelines and assent was obtained from the study subjects lessthan 18 years of age in addition to parental consent.

Diabetes was defined as fasting plasma glucose (FPG) level≥126 mg/dl (7.0 mmol/l) and/or 2-h post-load glucose level≥200 mg/dl (11.1 mmol/l) (Alberti & Zimmet, 1998) or self-reporteddiabetes treated by a physician or on hypoglycemic medications orinsulin. Diabetes was then classified as follows: T1DM-Y, if accompa-nied by abrupt onset of symptoms like polyuria, polydipsia, orunexplained weight loss, diabetic ketoacidosis (DKA), absent insulinreserve as shown by fasting and stimulated C-peptide (b0.3pmol/ml),and requirement of insulin from the time of diagnosis for controlof hyperglycemia; and T2DM-Y, as absence of ketosis, good betacell functional reserve as evidenced by stimulated C-peptide(≥0.6 pmol/ml), absence of pancreatic calculi (on X-ray abdomen),and good response to oral hypoglycemic agents for more than 2 years(Amutha et al., 2011).

Anthropometric measurements included height, weight, and waistcircumference (Deepa et al., 2003). Height was measured incentimetres using a stadiometer. Weight was measured with atraditional spring balance and recorded to the nearest 0.5 kg.Body mass index (BMI) was calculated using the formula: weight(kg)/height squared (in m2). Waist circumference was measuredusing a non-stretchable measuring tape. The participants were askedto stand erect in a relaxed position with both feet together on a flatsurface; one layer of clothing was accepted. Waist girth was measuredas the smallest horizontal girth between the costal margins and theiliac crests at minimal respiration. Blood pressure was recorded in arested sitting position in the right arm with a mercury sphygmoma-nometer and rounded off to the nearest 2 mmHg. Two readings weretaken 5 min apart and the mean of the 2 readings was used.

Fasting plasma glucose (hexokinase method) was measured onHitachi 912 Autoanalyzer (Hitachi, Mannheim, Germany) using kitssupplied by Roche Diagnostics (Mannheim, Germany); glycatedhaemoglobin (HbA1C) by high-pressure liquid chromatographyusing the Variant machine (Bio-Rad, Hercules, Calif., USA); serum totalcholesterol (cholesterol oxidase-peroxidase-amidopyrine method),serum triglycerides (glycerol phosphate oxidase-peroxidase-amidopyr-ine method), and HDL cholesterol (direct method-polyethylene glycol-pretreated enzymes) using Hitachi-912 Autoanalyser (Hitachi,Mannheim, Germany). Low-density lipoprotein (LDL) cholesterol wascalculatedusing theFriedewald formula (Friedewald, Levy,&Fredrickson,1972). Apo lipo-protein A and B were measured by immuno-turbido-metric method (Olympus AU 2700 USA).

Fasting and stimulated (post-breakfast) C-peptide was estimatedby the electro-luminescence method on Elecsys2010 (Hitachi,Mannheim, Germany); glutamic acid decarboxylase (GAD) antibodieswere measured on a Bio-Rad plate reader 680 (USA) using Elisa EuroImmun kit (Lubeck, Germany); and plasma concentration of hs-CRPby turbidometry method (Beckman Coulter AU 480 USA). The intra-and inter-assay co-efficient of variation for the biochemical assaysranged between 3.1% and 7.6%.

Urine samples were collected after an overnight fast. Microalbu-min concentration was measured using an immunoturbidometricassay (Hitachi 902 autoanalyzer; Roche Diagnostics, Mannheim,Germany) (Varghese, Deepa, Rema, & Mohan, 2001). Microalbumi-nuria was defined as a urine albumin excretion of 30–299 mg/μg ofcreatinine (Pradeepa et al., 2010). Nephropathy was defined asmacro-albuminuria i.e., a urine albumin excretion of ≥300 μg/mg ofcreatinine (Pradeepa et al., 2010). The DMDSC laboratory is certifiedby the College of American Pathologists and the Indian NationalAccreditation Board for Testing and Calibration of Laboratories.

2.1. Retinopathy

Four-field digital retinal colour photography was taken by atrained photographer using a Carl-Zeiss Digital Fundus Camera. The 4fields photographed were the macula, optic disc and nasal to the opticdisc, and superior–temporal and inferior–temporal quadrants of eacheye. The grading of retinopathy was done based on the modified EarlyTreatment Diabetic Retinopathy Study (ETDRS) grading system (EarlyTreatment Diabetic Retinopathy Study Research Group, 1991). Eacheye was graded separately by an ophthalmologist trained in ETDRSgrading, at DMDSC. The minimum criterion for diagnosis of DR wasthe presence of at least one definite microaneurysm in any field of theretina. Photographs were assessed and assigned a retinopathy leveland the final diagnosis for each patient was determined from the levelof DR of the worse eye using ETDRS final retinopathy scale (EarlyTreatment Diabetic Retinopathy Study Research Group, 1991; Rema etal., 2005). Briefly, level 10 represents no retinopathy, levels 20 to 50,non-proliferative diabetic retinopathy (NPDR) [Level 20: mild NPDR;levels 30 to 40: moderate NPDR; level 50: severe NPDR] andlevels ≥ 60, proliferative diabetic retinopathy (PDR) (Early TreatmentDiabetic Retinopathy Study Research Group, 1991).

Diabetic macular edema (DME) was defined as retinal thickeningat or within one disc diameter of the centre of the macula or thepresence of definite hard exudates (Early Treatment DiabeticRetinopathy Study Research Group, 1985; Wilkinson et al., 2003).DME could be present in both NPDR and PDR stages, but, once PDRwas diagnosed, the final severity scale was graded as PDR (Rema et al.,2005). Sight-threatening diabetic retinopathy (STDR) was defined asPDR or DME (clinically significant macular edema) in either or botheyes (Younis, Broadbent, Vora, & Harding, 2003). Where clinicallyindicated, optical coherence tomography (OCT) was done andappropriate treatment was offered to the subjects (Otani, Kishi, &Maruyama, 1999).

3. Statistical analysis

All statistical analyses were done using SPSS statistical packageversion 15.0 (SPSS Inc., Chicago, IL, USA). Continuous data areexpressed as mean ± standard deviation while categorical data arepresented as proportions. Student’s t test was used to compare meansof continuous variables between subjects with and without retinop-athy in T1DM-Y and T2DM-Y groups. Chi square test was used tocompare proportions. To assess independent risk factors for diabeticretinopathy and age adjusted prevalence, we used logistic regressionmodels with DR as the dependent variable and reported odds ratios(OR) with 95% confidence intervals (CIs) and percentages. First,variables were independently tested with the outcomes usingbivariable analysis and significant variables (p value b0.05) wereentered into a multivariable analysis. For all statistical tests, p valueb0.05 was considered significant.

4. Results

Diabetic retinopathy was present in 80 (53.3% [95% CI 45.3–61.3])people with T1DM-Y and 79 (52.7% [95% CI 44.7–60.7]) people with

Fig. 1. Prevalence of diabetic retinopathy (DR) in young onset diabetes.

293R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

T2DM-Y (Fig. 1). The mean age of T1DM-Y and T2DM-Y withretinopathy was 31.8 ± 7.7 years and 38.1 ± 9.2 years (p b 0.001).The age and gender adjusted prevalence of retinopathy in T1DM-Yand T2DM-Y was 62.5% and 65.8% respectively.

DMEwas present in 15 (10%) T1DM-Y and 19 (12.7%) T2DM-Y andPDR in 11(7.3%) T1DM-Y and 14 (9.3%) T2DM-Y participants. Theprevalence of PDR and STDR was no different among T2DM-Y andT1DM-Y participants (Fig. 1). Fig. 2 shows the retinal colourphotography of a T1DM-Y patient with PDR and Fig. 3 shows aT2DM-Y patient with DME. Fig. 4 shows the OCT of a T2DM-Y patientwith severe sight threatening DME.

Table 1 show the clinical and biochemical characteristics of thestudy participants with and without diabetic retinopathy (DR). Inpeople with T1DM-Y, those with DR were significantly older, hadlonger duration of diabetes, had significantly higher BMI, waistcircumference, higher systolic and diastolic blood pressures and a

Fig. 2. Retinal colour photography of diabetic retinopathy in young onset type 1diabetes.

greater proportion had microalbuminuria compared with thosewithout DR. In T2DM-Y, in addition to the above, those with DR alsohad significantly higher fasting plasma glucose, glycated hemoglobin(HbA1C), and apolipoprotein A levels as well as significantly lowerfasting and stimulated C-peptide levels when compared with thosewithout DR.

Prevalence of STDR was higher with longer duration of diabetes inboth T1DM-Y and T2DM-Y, with the prevalence peaking in thosewith N 15 years of diabetes. Interestingly, none of the T1DM-Y withduration of diabetes b10 years had STDR but among T2DM-Y people,STDR was prevalent even in those with a known duration of diabetesof b 5 years (Fig. 5). The overall STDR prevalence was higher withlonger duration of diabetes with 44.1% of the T1DM-Y and 52.5% of theT2DM-Y people with N15 years of diabetes having STDR.

The bivariable analysis for T1DM-Y participants showed age,duration of diabetes, waist circumference, systolic and diastolic blood

Fig. 3. Retinal colour photography of diabetic retinopathy in young onset type 2diabetes.

Fig. 4. Optical coherence tomography (OCT) of a T2DM-Y patients with diabetic macular edema.

294 R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

pressure and microalbuminuria to be associated with diabeticretinopathy. In addition to these factors, age of onset of diabetes,fasting plasma glucose, HbA1C and lower stimulated C-peptide wereassociated with DR in participants with T2DM-Y. For every 5 mm Hghigher systolic (SBP) and diastolic blood pressure (DBP), there werehigher odds of DR in both T1DM-Y and T2DM-Y patients (T1DM-YSBP: OR 1.3, 95% CI 1.09–1.48; DBP: OR 1.4, 95% CI 1.09–1.70; T2DM-YSBP:OR 1.3, 95% CI 1.1–1.49, DBP:1.6, 95% CI 1.24–1.97 per 5 mm Hg).In T1DM-Y participants, the age of onset of diabetes was notassociated with DR while in T2DM-Y, every 5-year increase in age ofonset was associated with a 2.2 times odds of having DR.

In multivariable regression analyses, significant independent pre-dictors for DR in participants with T1DM-Y were duration of diabetes,waist circumference and microalbuminuria, whereas in T2DM-Y

Table 1Clinical and biochemical characteristics of T1DM-Y and T2DM-Y with and without diabetic

Variables T1DM-Y

No DR (n = 70) With DR (n = 80)

Age (years) 24.6 ± 8.1 31.8 ± 7.6Duration (years) 8.4 ± 6.5 15.2 ± 6.7Age at onset (years) 16.6 ± 4.1 17.0 ± 4.4Gender, male n (%)a 36(51.4) 48(60.0)Body mass index (kg/m2) 21.1 ± 3.1 23.0 ± 3.9Waist circumference(cm) 73.5 ± 10.6 80.9 ± 10.4Systolic Blood Pressure (mmHg) 116 ± 10 124 ± 18Diastolic Blood Pressure (mmHg) 72 ± 7.0 76 ± 9Fasting plasma glucose (mg/dl)b 163 (10) 163 (11)Glycated hemoglobin (%) 9.0 ± 1.9 8.5 ± 1.6Total Cholesterol (mg/dl) 160 ± 32 168 ± 41Serum Triglycerides (mg/dl)b 69 (5) 81(6)HDL Cholesterol (mg/dl) 48 ± 12 46 ± 11LDL Cholesterol (mg/dl) 97 ± 27 102 ± 32Apo Lipoprotein A (mg/dl) 126 ± 32 129 ± 31Apo Lipoprotein B (mg/dl) 76 ± 18 82 ± 24Microalbuminuria (ug/mg) 11.9 ± 2.5 68.4 ± 13.2HS-CRP (mg/l)b 1.62 (0.37) 1.92(0.37)C-Peptide Fasting (pmol/ml)b 0.28 (0.01) 0.30 (0.03)C-Peptide Stimulated (pmol/ml)b 0.32 (0.01) 0.32 (0.04)Smoking (n %) 1(1.4) 5(6.2)

Data given as mean ± SD, p b 0.05 considered significant.Bold denotes statistical significance.

a Chi square p value given.b Data given as Geometric mean (SE).

participants, duration of diabetes, diastolic blood pressure, HbA1Cand lower stimulated C-peptide values were associated (Table 2).

5. Discussion

The prevalence of DR and STDR was high in both T1DM-Y andT2DM-Y from this clinic population studied in south India, with nosignificant difference in the age adjusted prevalence of diabeticretinopathy between T1DM-Y and T2DM-Y. This is in contrast toearlier epidemiological studies reporting that diabetic retinopathy inT1DM-Y has a different frequency, onset, and course than DR in T2DM(Klein, Klein, Moss, Davis, & DeMets, 1984b, Klein et al., 1984a,b).The high prevalence of DR among T2DM-Y in our study is similar tothat reported from Japan in T2DM-Y subjects (diagnosed b 30years)

retinopathy (DR).

T2DM-Y

p value No DR (n = 71) With DR (n = 79) p value

b0.001 28.0 ± 7.8 38.1 ± 9.1 b0.001b0.001 7.6 ± 5.9 15.6 ± 8.2 b0.0010.571 20.8 ± 3.9 22.3 ± 2.9 0.0110.291 47(66.2) 45(57.0) 0.2460.001 26.5 ± 4.7 27.1 ± 4.2 0.423

b0.001 89.2 ± 11.6 92.6 ± 10.3 0.0640.001 123 ± 12 130 ± 14 b0.0010.006 76 ± 8 81 ± 9 b0.0010.742 141 (6) 168 (8) b0.0010.067 8.1 ± 1.8 9.1 ± 2.0 0.0200.150 158 ± 35 155 ± 38 0.5370.058 120(10) 130(12) 0.3620.192 38 ± 9 40 ± 9 0.3380.243 93 ± 28 89 ± 29 0.4030.586 114 ± 34 126 ± 37 0.0480.113 84 ± 20 82 ± 19 0.417

b0.001 22.7 ± 6.36 55.0 ± 10.4 0.0090.398 2.46 (0.42) 2.20 (0.39) 0.4670.227 0.87 (0.04) 0.71(0.04) 0.0170.182 1.89 (0.10) 1.30 (0.08) b0.0010.133 5(7.0) 5(6.3) 0.861

Fig. 5. Frequency of diabetic retinopathy (DR) and sight threatening diabetic retinopathy (STDR) based upon the duration of diabetes.

295R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

(Okudaira, Yokoyama, Otani, Uchigata, & Iwamoto, 2000) and also inanother study by Wong et al., (2008), which showed that youngerT2DM had more severe grades of DR than the older T2DM patients.

Krakoff et al. (2003) reported that the incidence of retinopathywas lower in the youth-onset group (b20 years at onset of diabetes)than in either adult group (onset of diabetes at 20 to 39 and 40 to 59years of age) in all categories of diabetes duration. Interestingly, wenoted that the age of onset of diabetes did not play a role in T1DM-Y,while age of onset of diabetes between 21 and 25 years increased therisk for DR in T2DM-Y patients.

A younger onset of diabetes with an increasing life span implieslonger duration of diabetes (Klein et al., 1984a; Namperumalsamyet al., 2009; Okudaira et al., 2000), hence a large number of people areat risk of developing DR. In the Wisconsin Epidemiological Study ofDiabetic Retinopathy (WESDR), the prevalence of DR in young peoplewith diabetes for N15 years was 98% (Klein et al., 1984a). In our study,in those with diabetes for N15 years, the prevalence of DR was 82.9%and 90.7% in T1DM-Y and T2DM-Y respectively. The higher preva-lence of DR and STDR in T2DM-Y in those with more than 15 years ofdiabetes, points to the aggressive nature of the disease. However, it is

Table 2Multivariable logistic regression—Factors associated with diabetic retinopathy inT1DM-Y and T2DM-Y.

Variablesa Increment/value OR(95% CI) p value

T1DM-Y

Duration of diabetes (years) 5 years 1.99 (1.42–2.79) b0.001Waist circumference (cm) 5 cm 1.28 (1.05–1.56) 0.014Microalbuminuria (μg/mg) 50 μg/mg 2.39 (1.07–5.31) 0.033

Variablesa Increment/value OR(95% CI) p value

T2DM-Y

Duration of diabetes (years) 5 years 2.21 (1.61–3.02) b0.001Diastolic bloodpressure(mmHg)

5 mmHg 1.54 (1.18–2.02) 0.002

Glycated hemoglobin (%) 1% 1.37 (1.07–1.75) 0.011C-peptide stimulated(pmol/ml)

0.5 pmol/mldecrease

1.54(1.15–2.05) 0.003

For T1DM-Y subjects, along with the above predictors age (per 5 yrs), systolic anddiastolic blood pressures (per 5 mmHg) were included in the analysis and for T2DM-Ysubjects, age (per 5 years), age onset of diabetes (per 5 years), waist circumference(per 5 cm), systolic blood pressure (per 5 mmHg), fasting plasma glucose (per 50 mg)and microalbuminuria (per 50ug/mg) were included.

a Only significant independent predictors are given, p b 0.05 considered significant.

uncertain whether this may be reflecting a longer period ofasymptomatic undiagnosed diabetes in T2DM-Y.

To our knowledge, there have been only a few studies, comparingthe prevalence of retinopathy in T1DM-Y versus T2DM-Y. In the pilotstudy on DR in the SEARCH study for diabetes in youth, the prevalenceof DR was higher in T2DM-Y (42%) than T1DM-Y (17%). However, thisdifference was largely attributed to the ethnic variation, with DR inT2DM-Y being more common in the non-Caucasian populations(e.g., native American, Hispanic, non-Hispanic Black) (Mayer-Daviset al., 2012). In our study, we recruited equal numbers of patients withT1DM-Y and T2DM-Y that were of the same race/ethnicity and withsimilar mean duration of diabetes. Eppens et al. reported that DR wasmore prevalent in adolescents with T1DM-Y than T2DM-Y (20% vs.4%) (Eppens et al., 2006). However in their study, the duration ofdiabetes was very different between the two groups (T1DM-Y =6.8 years, T2DM-Y = 1.3 years) making comparisons difficult.

We found that increased systolic and diastolic blood pressure,regardless of the type of young onset diabetes, increased the risk forretinopathy. Diastolic blood pressure was also an independentpredictor for DR in T2DM-Y patients. The United Kingdom ProspectiveDiabetes Study (UKPDS) study showed a continuous relationshipbetween the risk of DR and systolic blood pressure (UK ProspectiveDiabetes Study Group, 1998b) and lowering blood pressure to a meanof 144/82 mm Hg reduced microvascular complications and visualloss (UK Prospective Diabetes Study Group, 1998a). In the WESDR,diastolic blood pressure was seen as a significant predictor ofprogression of diabetic retinopathy (Klein et al., 1984a). We alsofound that in T2DM-Y, increased HbA1C was associated withprevalent DR as seen in other studies (Early Treatment DiabeticRetinopathy Study Research Group, 1991; Namperumalsamy et al.,2009). However, there was no association seen between presentglycemic status and development of DR in T1DM-Y subjects.This could probably be attributed to the hyperglycemic memoryinfluencing the development of DR in T1DM-Y patients (Zhang, Chen,& Tang, 2012).

We found that lower stimulated C-peptide levels were associatedwith increased risk for retinopathy in T2DM-Y. Klein, Moss, Klein,Davis, and Demets (1990) have shown that diabetes patients usinginsulin, with undetectable or very low plasma C-peptide (b0.3 pmol/ml), are more likely to have DR and also more severe DR than olderonset individuals who have higher levels of C-peptide (N0.3 pmol/ml). The Diabetes Control and Complications Trial (DCCT) showedthat higher and sustained levels of stimulated C-peptide wereassociated with reduced incidence of retinopathy (Steffes, Sibley,

296 R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

Jackson, & Thomas, 2003). This highlights the possible benefit ofsustaining pancreatic β cell function in order to reduce thedevelopment and progression of microvascular complications likeretinopathy (Yoon et al., 2012).

Another risk factor associated with DR in both T1DM-Y and T2DM-Ywas the presence of microalbuminuria. Associations betweenretinopathy and microalbuminuria have also been demonstrated inmany studies (Collins et al., 1995; Farah, Wals, Friedman, Pisacano, &DiMartino-Nardi, 2006; Gall et al., 1991; Matyka, 2008). T1DM-Ypeople with increased waist circumference also had greater risk forDR. Dirani et al. (2011) have shown that in persons with diabetes,higher BMI and larger waist circumference confer a higher risk of DRand are also associated with severity of DR.

Our study has certain limitations. It is cross-sectional, and thuscausal inferences cannot be drawn. Our population was recruited at atertiary care clinic, and therefore these findings may not apply to thepopulation of T2DM-Y and T1DM-Y at large. Moreover they cannot beextrapolated to the country as a whole as referral bias could haveaffected the results. However our study has certain strengths as well.This is the first study from a developing country to compare theprevalence and risk factors of diabetic retinopathy in T1DM-Y versusT2DM-Y with objective measurement of diabetic retinopathy usingstate-of-the-art digital retinal colour photography.

Overall, half the people with youth-onset diabetes of N10 yearsduration have retinopathy. Our findings emphasise the need forregular and periodic screening for diabetic retinopathy and especiallyin patients with pre-existing DR, a shorter interval of ≤1 year may beneeded (Echouffo-Tcheugui, Ali, Roglic, Hayward, & Venkat Narayan,2013). Tight control of glucose and blood pressure from the time ofdiagnosis is important in order to reduce morbidity and disability dueto diabetic eye disease in young onset diabetes.

Acknowledgments

We acknowledge the help of all the optometrists at DMDSC forperforming the digital retinal colour photography.We acknowledge thesupport of the study team members. Most importantly, we wish tothank the subjects who participated in the study. We acknowledge thefinancial support from the Emory Global Health Institute through theGlobal Diabetes Research Center (GDRC) — collaboration betweenMadras Diabetes Research Foundation, Chennai and Emory University,Atlanta, USA. This is the 11th paper from this collaboration (GDRC-11).

References

Alberti, K. G., & Zimmet, P. Z. (1998). Definition diagnosis and classification of diabetesmellitus and its complications. Part 1: Diagnosis and classification of diabetesmellitus, provisional report of a WHO Consultation. Diabetic Medicine, 15, 539–553.

American Diabetes Association (2000). Type 2 diabetes in children and adolescents(consensus statement). Diabetes Care, 23, 381–389.

Amutha, A., Datta, M., Unnikrishnan, R., Anjana, R. M., & Mohan, V. (2012). Clinicalprofile and complications of childhood and adolescent onset type 2 diabetesseen at a diabetes centre in South India. Diabetes Technology & Therapeutics, 14,497–504.

Amutha, A., Datta, M., Unnikrishnan, I. R., Anjana, R. M., Rema, M., Venkat Narayan,K. M., et al. (2011). Clinical profile of diabetes in the young seen between 1992and 2009 at a specialist diabetescentre in south India. Primary Care Diabetes, 5,223–229.

Chaturvedi, N., Sjolie, A. K., Stephenson, J. M., Abrahamian, H., Keipes, M., Castellarin, A.,et al. (1998). Effect of lisinopril on progression of retinopathy in normotensivepeople with type 1 diabetes. The EUCLID Study Group. EURODIAB Controlled Trialof Lisinopril in Insulin-Dependent Diabetes Mellitus. Lancet, 35, 28–31.

Collins, V. R., Dowse, G. K., Plehwe,W. E., Imo, T. T., Toelupe, P. M., & Taylor, H. R. (1995).High prevalence of diabetic retinopathy and nephropathy in Polynesians ofWestern Samoa. Diabetes Care, 18, 1140–1149.

Deepa, M., Pradeepa, R., Rema, M., Mohan, A., Deepa, R., Shanthirani, S., et al. (2003).The Chennai Urban Rural Epidemiology Study (CURES): study design andmethodology (urban component) (CURES-1). Journal of the Association of Physiciansof India, 51, 863–870.

Dirani, M., Xie, J., Fenwick, E., Benarous, R., Rees, G., Wong, T. Y., et al. (2011).Are obesity and anthropometry risk factors for diabetic retinopathy? The

Diabetes Management Project. Investigative Ophthalmology & Visual Science, 52,4416–4421.

Early Treatment Diabetic Retinopathy Study Research Group (1985). Photocoagulationfor diabetic macular edema: Early Treatment Diabetic Retinopathy Study ReportNumber 1. Archives of Ophthalmology, 103, 1796–1806.

Early Treatment Diabetic Retinopathy Study Research Group (1991). Grading diabeticretinopathy from stereoscopic colour fundus photographs: An extension of themodified Airlie House classification. Early Treatment Diabetic Retinopathy StudyReport Number 10. Ophthalmology, 98, 786–806.

Echouffo-Tcheugui, J. B., Ali, M. K., Roglic, G., Hayward, R. A., & Venkat Narayan, K. M.(2013). Screening intervals for diabetic retinopathy and incidence of visual loss: Asystematic review. Jul 2. Diabetic Medicine, http://dx.doi.org/10.1111/dme.12274[Epub ahead of print].

Eppens, M. C., Craig, M. E., Cusumano, J., Hing, S., Chan, A. K. F., Howard, N. J., et al.(2006). Prevalence of diabetes complications in adolescents with type 2 comparedwith type 1 diabetes. Diabetes Care, 29, 1300–1306.

Farah, S. E., Wals, K. T., Friedman, I. B., Pisacano, M. A., & DiMartino-Nardi, J. (2006).Prevalence of retinopathy and microalbuminuria in pediatric type 2 diabetesmellitus. Journal of Pediatric Endocrinology & Metabolism, 19, 937–942.

Friedewald, W. T., Levy, R. I., & Fredrickson, D. S. (1972). Estimation of the concentrationof low-density lipoprotein cholesterol in plasma, without use of the preparativeultracentrifuge. Clinical Chemistry, 18, 499–502.

Gall, M. A., Rossing, P., Skott, P., Damsbo, P., Vaag, A., & Bech, K. (1991). Prevalence ofmicro- and macroalbuminuria, arterial hypertension, retinopathy, and large vesseldisease in European type 2 (non-insulin dependent) diabetic patients. Diabetologia, 34,655–661.

Hillier, T. A., & Pedula, K. (2003). Complications in young adults with early-onset type 2diabetes: Losing the relative protection of youth. Diabetes Care, 26, 2999–3005.

International Diabetes Federation. (2011). Diabetes atlas (Fifth ed.)Brussels, Belgium:International Diabetes Federation.

Klein, R., Klein, B. E., Moss, S. E., Davis, M. D., & DeMets, D. L. (1984a). The WisconsinEpidemiologic Study of Diabetic Retinopathy. II. Prevalence and risk of diabeticretinopathy when age at diagnosis is less than 30 years. Archives of Ophthalmology,102, 520–526.

Klein, R., Klein, B. E., Moss, S. E., Davis, M. D., & DeMets, M. L. (1984b). The WisconsinEpidemiologic Study of Diabetic Retinopathy. III Prevalence and risk of diabeticretinopathy when age at diagnosis is 30 or more years. Archives of Ophthalmology,102, 527–532.

Klein, R., Moss, S. E., Klein, B. E. K., Davis, M. D., & Demets, D. L. (1990). WisconsinEpidemiologic Study of Diabetic Retinopathy: XII. Relationship of C-peptide anddiabetic retinopathy. Diabetes, 39, 1445–1450.

Krakoff, J., Lindsay, R. S., Looker, H. C., Nelson, R. G., Hanson, R. L., & Knowler, C. (2003).Incidence of retinopathy and nephropathy in youth-onset compared with adult-onset type 2 diabetes. Diabetes Care, 26, 76–81.

Matyka, K. A. (2008). Type 2 diabetes in childhood: Epidemiological and clinicalaspects. British Medical Bulletin, 86, 59–75.

Mayer-Davis, E. J., Davis, C., Saadine, J., D'Agostino, R. B., Jr., Dabelea, D., Dolan, L., et al.(2012). SEARCH for Diabetes in Youth Study Group. Diabetic retinopathy in theSEARCH for Diabetes in Youth cohort: A pilot study. Diabetic Medicine, 29,1148–1152.

McMahon, S. K., Haynes, A., Ratnam, N., Grant, M. T., Carne, C. L., Jones, T. W., et al.(2004). Increase in type 2 diabetes in children and adolescents in WesternAustralia. Medical Journal of Australia, 180, 459–461.

Mohan, V., Jaydip, R., & Deepa, R. (2007). Type 2 diabetes in Asian Indian youth. Pediatricdiabetes, 9, 28–34.

Namperumalsamy, P., Kim, R., Vignesh, T. P., Nithya, N., Royes, J., Gijo, T., et al. (2009).Prevalence and risk factors for diabetic retinopathy: A population-based assess-ment from Theni District, south India. British Journal of Ophthalmology, 93, 429–434.

Okudaira, M., Yokoyama, H., Otani, T., Uchigata, Y., & Iwamoto, Y. (2000). Slightlyelevated blood pressure as well as poor metabolic control are risk factors for theprogression of retinopathy in early-onset Japanese type 2 diabetes. Journal ofDiabetes and its Complications, 14, 281–287.

Otani, T., Kishi, S., & Maruyama, Y. (1999). Patterns of diabetic macular edema in opticalcoherence tomography. American Journal of Ophthalmology, 127, 688–693.

Pinhas-Hamiel, O., & Zeitler, P. (2005). The global spread of type 2 diabetes mellitus inchildren and adolescents. Journal of Pediatrics, 146, 693–700.

Pinhas-Hamiel, O., & Zeitler, P. (2007). Acute and chronic complications of type 2diabetes mellitus in children and adolescents. Lancet, 369, 1823–1831.

Pradeepa, R., Anjana, R. M., Unnikrishnan, R., Ganesan, A., & Mohan, V.,& Rema, M.(2010). Risk factors for microvascular complications of diabetes among SouthIndian subjects with type 2 diabetes: The Chennai Urban Rural Epidemiology Study(CURES) Eye Study-5. Diabetes Technology & Therapeutics, 12, 755–761.

Raman, R., Padmaja, K. R., Reddi, S., Gnanamoorthy, R. P., Uthra, S., Kumaramanickavel,G., et al. (2009). Sankara Nethralaya Diabetic Retinopathy Epidemiology andMolecular Genetics Study report. Prevalence of diabetic retinopathy in India.Ophthalmology, 116, 311–318.

Rema, M., Mohan, V., & Ponnaiya, M. (1995). Natural history of retinopathy in FCPD.MODY and IDDM: A follow-up study. International Journal of Diabetes in DevelopingCountries, 15, 45–48.

Rema, M., Mohan, V., Ramachandran, A., & Viswanathan, M. (1989). Retinopathy ininsulin dependent diabetes mellitus (IDDM) in south India. Journal of theAssociation of Physicians of India, 37, 370–373.

Rema, M., Premkumar, S., Anitha, B., Deepa, R., Pradeepa, R., & Mohan, V. (2005).Prevalence of diabetic retinopathy in urban India: The Chennai Urban RuralEpidemiology Study (CURES) Eye Study-1. Investigative Ophthalmology & VisualScience, 46, 2328–2333.

297R. Rajalakshmi et al. / Journal of Diabetes and Its Complications 28 (2014) 291–297

Steffes, M. W., Sibley, S., Jackson, M., & Thomas, W. (2003). Beta-cell function and thedevelopment of diabetes-related complications in the Diabetes Control and ComplicationsTrial, 26. (pp. 832–836), 832–836.

The DCCT Research Group (1993). The effect of intensive treatment of diabetes on thedevelopment and progression of long-term complications in insulin-dependentdiabetes mellitus. New England Journal of Medicine, 329, 977–986.

UK Prospective Diabetes Study Group (1998a). Efficacy of atenolol and captopril inreducing risk of both macrovascular and microvascular complications in type 2diabetes (UKPDS 39). BMJ, 317, 713–720.

UK Prospective Diabetes Study Group (1998b). Tight blood pressure control and risk ofmacrovascular and microvascular complications in type 2 diabetes (UKPDS 38).BMJ, 317, 703–713.

Varghese, A., Deepa, R., Rema, M., & Mohan, V. (2001). Prevalence of microalbuminuriain type 2 diabetes mellitus at a diabetes centre in Southern India. PostgraduateMedical Journal, 77, 399–402.

Wilkinson, C. P., Ferris, F. L., 3rd, Klein, R. E., Lee, P. P., Agardh, C. D., Davis, M., et al.(2003). Global Diabetic Retinopathy Project Group. Proposed international clinicaldiabetic retinopathy and diabetic macular edema disease severity scales.Ophthalmology, 110, 1677–1682.

Wong, J., Molyneaux, L., Constantino, M., Twigg, S., & Yue, D. (2008). Age of onsetinfluences long-term retinopathy risk in type 2 diabetes, independent of traditionalrisk factors. Diabetes Care, 31, 1985–1990.

Yoon, H. J., Cho, Y. Z., Kim, J. Y., Kim, B. J., Park, K. Y., Koh, G. P., et al. (2012). Correlationsbetween glucagon stimulated C-peptide levels and microvascular complications intype 2 diabetes patients. Diabetes & Metabolism Journal, 36, 379–387.

Younis, N., Broadbent, D. M., Vora, J. P., & Harding, S. P. (2003). Incidence of sight-threatening retinopathy in patients with type 2 diabetes in the Liverpool DiabeticEye Study: A cohort study. Lancet, 361, 195–200.

Zhang, L., Chen, B., & Tang, L. (2012). Metabolic memory: Mechanisms and implicationsfor diabetic retinopathy. Diabetes research and clinical practice, 96, 286–293.