A cross-sectional study was carried out in the Department of Ophthalmology on patients suffering from diabetes (n=100), selected by simple random sampling method, admitted to the Acharya Vinobha Bhave Rural Hospital, Sawangi (Meghe), Wardha, Maharashtra in a duration of 2 months from April to June 2014. Indian Ethical Committee approval was taken.

The prevalence of DR in India ranges from 18% to 44.4% in various studies carried out in the past [19, 20]. Considering the prevalence in central rural India (as the study methodology was similar to ours), based on the following assumptions: Prevalence of DR in general population is 44.4% with a precision of 10% and p<0.05 significant, the sample size was calculated to be 98.56 rounded up to 99, further considering it to be 100 samples, using the formula 4PQ/d².

where: p (expected prevalence) = 0.44 and Q (1-p)=0.56 and d (precision) was considered 10%=0.10

All patients of pediatric age group (below 18 years).

The demographic information of each study subject including his/her name, age, sex, occupation, address was taken after obtaining his/her verbal informed consent.

Fasting blood glucose was estimated by performing glucose oxidase and peroxidase method. Then, history containing study variables like duration of diabetes, current insulin intake, alcohol intake, smoking status, and family history of diabetes was recorded. Blood pressure of each subject was measured in right arm, supine position. Two readings were taken half an hour apart and the average of two was taken as a final reading. In our study the patients were considered hypertensive as per JNC VII criteria in which the reported classification suggests that all patients having average blood pressure <120/80 mmHg are normotensive; systolic blood pressure (SBP) 120-139 mmHg or diastolic blood pressure (DBP) 80-89 mmHg are pre-hypertensive; SBP 140-159 mmHg or DBP 90-99 mmHg are stage 1 hypertensive; and SBP>160 mmHg or DBP>100 mmHg are stage 2 hypertensive. The hypertensive patients are further to be given appropriate medications.

Body weight was measured (to the nearest kilogram). Height was measured (to the nearest centimeter). BMI was calculated as weight in kilogram divided by height in meter square [weight (Kg) / Height (m)²]. Based on the BMI individuals were classified as lean (BMI < 18.5), normal (BMI = 18.5 – 24.9), overweight (BMI = 25.0 – 29.9) and Obese (BMI > 30.0).

Visual acuity was measured by Snellen’s chart. A slit lamp was used for anterior segment evaluation including the depth of anterior chamber. IOP measurement was performed by non-contact tonometry. Indirect ophthalmoscopy was done after complete pupillary dilatation by 1% tropicamide eye drops. Classification of retinopathy was based on the findings of the worst eye of each subject.

Biochemical studies including estimation of total serum cholesterol by CHOD/PAP method, high density lipoprotein by direct enzymatic method, serum TG by GPO/PAP method. Glycemic control is measured by HbA1C, but as HbA1C is not standardized in India and due to the unaffordable cost of the test, fasting blood glucose levels was used as an index for current glycemic control instead of HbA1C.

All the data collected was compiled, edited classified as was entered into the proforma. The difference between DR and non DR was statistically analysed on the basis of all the independent variables with the retinopathy and was then compared with the previous data from the articles of various medical journals. Statistical analysis was done by using descriptive and inferential statistics using chi-square test, student’s unpaired t test and one way ANOVA. The software used in the analysis were SPSS 17.0 and GraphPad Prism 5.0 and p<0.05 was considered as level of significance (p<0.05).

A total of 100 diabetic patients were evaluated. Mean age was 56.4±11.2 years. [Table/Fig-1] shows positive correlation between DR and increasing age but it was not statistically significant. [Table/Fig-1,2] show that retinopathy was prevalent in type 2 diabetes compared with those of type 1 (84% vs. 2% for NPDR; and 14% vs. 0% for PDR), majority fortunately being mild to moderate NPDR cases and a few diabetics unfortunately diagnosed to have PDR.

[Table/Fig-3,4] show that duration of DM, total serum cholesterol, serum triglycerides were significantly associated with both presence and severity of DR (p<0.05). There was significant association of presence of DR with HDL (p=0.012), hypertension (p=0.045), current insulin intake (p=0.02) and with FBG (p=0.003), while BMI showed significant association with severity of DR (p=0.040). Sex, alcohol, smoking, type of DM were not associated with DR at statistically significant levels.

Recent studies have reported that the prevalence of DM in India has rapidly increased to become more than 61 million due to rapid transition economically, demographically and nutritionally along with changes in lifestyle in both rural and urban population. This in turn suggests that in a few more years most of these diabetics will show ocular complications (DR) leading to blindness which is a matter of concern. The figures for NPDR were much higher than other studies which showed prevalence rates of 71.79% and 71.88% [11,19]. This could be due to the reason that participants with lesser duration of diagnosed diabetes were higher in comparison to the other studies. 14% prevalence of PDR in this study, much higher than 5.12% in Loni [19] may be due to poor diabetic control or due to late diagnosis of pre-existing diabetes, which are both very common in rural population of India, where the economic status is poor and there is lesser awareness about diabetes and its systemic complications.

Inclusion of both IDDM and NIDDM may explain the wide range of age of DR patients (32 to 85 years) similar to 2 other studies [11,20], unlike most other studies. The studies have shown that onset of DR may depend on the age of onset of diabetes indirectly, which is a reason for screening of all diabetics of this age group. 70% male prevalence of DR in this study, similar to studies carried out in Ahmednagar (64.10%) and Dhaka (58.9%) [12,19], could be because of higher social status of males over lesser prioritized females in rural India.

16.67% of newly diagnosed diabetics showed DR, which is higher than previous studies [5,6] which may be because of an underscore of early diagnosis of DM which in turn decreases chances of prevention of onset of DR, leaving prevention of progression of severity of DR as the only choice. With increase in duration of diabetes, the magnitude and prolonged exposure of hyperglycemia increase, if other risk factors are not prevented or controlled.

33.33% vs. 45.2% type 1 and 43.30% vs. 54.8% of type 2 DM patients suffered from NPDR and PDR respectively in our study vs. one carried out in another developing nation [22]. Also 2% vs. 43.3% and 98% vs. 56.67% of DR patients suffered from type 1 and 2 respectively in our study vs. one carried out in another country [22].

Family history of diabetes mellitus could not be considered for correlation with DR due to unawareness among most patients about the strong link between family history of DM and its systemic complications. There was no association between type 1 and DR which may be explained by the insulin therapy and hypoglycemic drugs which probably must have prescribed at the time of diagnosis of DM with the aim of maximum protection so as to reduce the risk of such complications. There was no significant association found between current alcohol intake and DR. Other studies reported significant association between heavy alcohol consumption for long durations. This shows a limitation in our study being that only history of current alcohol intake was taken without proper data regarding duration and quantity of alcohol consumption. No significant association between smoking and DR was found in this study may be due to less prevalence of tobacco smoking in comparison to other forms in rural India.

This study showed an inverse association between the insulin intake and the presence of DR, similar to 2 other studies in India [2, 9], which may lead to a theory that control of blood glucose levels by insulin intake, delays ocular complications. Our study has shown statistically significant association between FBG and presence of DR (p=0.003). Similar to other study in South Asia [22] 8% DR cases, where FBG was <100mg/dl, may have resulted due to poor metabolic control earlier or because of longer duration of DM. 80% patients having FBG>150mg/dl, did not suffer from DR. This could be due to shorter duration of DM among most of these patients or absence of other associated risk factors. Hypertension was significantly associated with presence of DR (p=0.45) similar to the study done by UKPDS [23]. There was no association found with progression of DR. 64% non- DR and 50% DR patients had hypertension. Anti-hypertensive therapy taken by earlier diagnosed hypertensive diabetics may have caused shift from recently uncontrolled to normal blood pressure levels, thus preventing or delaying effects on retinal capillary endothelial cells, which may explain why more hypertensive diabetic patients did not suffer from DR in comparison to those who did suffer. Another related possible mechanism by which hypertension may cause DR is through VEGF (vascular endothelial growth factor), but if the patient is on anti-VEGF for prevention of systemic complications, the chances of the diabetic hypertensive patient suffering from DR may reduce upto a certain extent. The mean serum total cholesterol, mean serum triglycerides, and mean HDL in any DR patients among the participants of our study versus the reports from the study done at Dhaka were196.36 mg/dl versus 272.33 mg/dl, 166.16 mg/dl versus 282.43 mg/dl, and 39.87 mg/dl versus 38.61 mg/dl [12]. In our study, HDL was only significantly associated with presence and not the severity of DR unlike triglycerides and total cholesterol which were related to both presence and severity of DR. This could be because of the low mean lipid levels of the participants in our study. Assessing these risk factors is important for early management to reduce the onset and progression of the ocular complications of diabetes [24].

There are some limitations in this study. A majority of the participants selected were self-reported diabetics hospitalized for various systemic complications other than DR. Few of the patients in this study were detected to be diabetic only after being hospitalized for other diseases. This may have caused selection bias. Lack of fundus photography may be responsible for missing some early diabetic retinopathy cases and thus an underestimate of its prevalence. Due to the lack of standardization of measuring HbA1c in our country and relatively higher cost of this test, glycemic control over a longer duration, of the study participants could not be measured. Only current control could be measured which would in turn bring out a possibility of underscore of magnitude of diabetic patients with uncontrolled hyperglycemia and its association with presence and severity of DR.

The under estimated diabetic population includes mostly undetected population of rural settings, who are unaware of both the disease and the risk factors which may lead to systemic complications. With improved strategies of health care services, more diabetics will live for a longer duration, which in turn will lead to increased patients of diabetic retinopathy even if most of the risk factors are prevented or controlled. Thus screening of all high risk individuals for diabetes is required and then further should be made aware of methods to eliminate various risk factors and then receive complete ocular examination at regular intervals.