Sudanese Journal of Ophthalmology

: 2016  |  Volume : 8  |  Issue : 1  |  Page : 30--35

Correlation between severity of Type 2 diabetes mellitus and corneal morphology using specular microscopy in Indian population: A case–control study

Eesha Jayant Shukla, Anjali Darius Nicholson, Anamika H Agrawal, Darshana B Rathod 
 Department of Ophthalmology, Topiwala National Medical College and BYL Nair Charitable Hospital, Mumbai, Maharashtra, India

Correspondence Address:
Eesha Jayant Shukla
Flat No. 12/B Wing, Shri Sai Krupa, 59, Prarthana Samaj Road, Vile Parle East, Mumbai-400 057, Maharashtra


Context: Globally, there are 366 million patients living with Type 2 diabetes mellitus (DM), of which, 62.4 million are living in India alone. Thus, diabetic keratopathy is an important entity for all ophthalmologists. Aims: The study aims to identify variables associated with Type 2 DM, which correlate well with specular microscopy parameters. Settings and Design: This is a population-based case–control study aimed at determining the correlation between severity of DM Type 2 and corneal morphology. Subjects and Methods: After obtaining informed consent, detailed history of onset, duration, and treatment of Type 2 DM was ascertained. A comprehensive eye examination was done including specular microscopy using noncontact technique. Blood investigations were performed by the endocrine department for documenting whether disease was controlled or not. Statistical Tests: Unpaired t-test and linear regression analysis were used to determine the effects of multiple variables on the various specular microscopy parameters. Results: There were 100 eyes of diabetic patients with the mean age of 52.02 years, and 100 eyes of nondiabetic subjects with the mean age of 54.76. Short-term glycemic control parameter such as fasting blood sugar (FBS) and postlunch blood sugar (PLBS) showed no correlation with three specular microscopy parameters. However, HbA1c and severity of diabetic retinopathy influenced those parameters significantly. Conclusions: Corneal endothelium in diabetic patients was compromised, and it correlated well with the long-term glycemic control (HbA1c) but not with the short-term glycemic control (FBS, PLBS) or the duration of disease.

How to cite this article:
Shukla EJ, Nicholson AD, Agrawal AH, Rathod DB. Correlation between severity of Type 2 diabetes mellitus and corneal morphology using specular microscopy in Indian population: A case–control study.Sudanese J Ophthalmol 2016;8:30-35

How to cite this URL:
Shukla EJ, Nicholson AD, Agrawal AH, Rathod DB. Correlation between severity of Type 2 diabetes mellitus and corneal morphology using specular microscopy in Indian population: A case–control study. Sudanese J Ophthalmol [serial online] 2016 [cited 2022 Aug 9 ];8:30-35
Available from:

Full Text


Diabetes mellitus (DM) can affect almost all structures of the eye. Globally, there are 366 million patients living with Type 2 diabetes mellitus (DM), of which, 62.4 million are living in India alone.[1],[2] Patients can develop not only diabetic retinopathy (DR) but also corneal damage such as endothelial defects, punctate epithelial keratopathy, recurrent corneal erosions, and persistent epithelial defects.[3]

Animal studies have also shown that corneal endothelial cells in diabetic rats have morphological abnormalities. These abnormalities include a decrease in endothelial cell density (CD) and hexagonality, as well as increased polymegathism, polymorphism, and central corneal thickness (CCT).[4] This study aims to further investigate if there are any correlations between the corneal morphology and other factors such as duration of diabetes, glycosylated hemoglobin (HbA1c) value, and the stage of DR.

Research has indicated that diabetic corneas have poor surgical outcomes as compared to nondiabetics. However, the type, duration, and severity of the disease have not been correlated with the corneal specular microscopic parameters in a tertiary health care set up in Urban India. Our study aims to achieve the same.

The aims and objectives of the study are to assess the influence of severity of the diabetic disease process on corneal endothelial morphologic parameters using specular microscopy, as compared to nondiabetic, healthy control corneas.

 Subjects and Methods

This study is a case–control study conducted at a tertiary care center for 15 months. Prior approval from the Institutional Ethics Committee was taken, and informed consent was obtained from each subject. This study was conducted in accordance with the tenets of the Declaration of Helsinki. About 50 Type 2 DM patients (100 eyes) participated as cases and 50 age matched subjects participated as controls. Exclusion criteria included patients with Type 1 DM, age <40 years, patients with any preexisting ocular pathology. All patients underwent complete ophthalmic and general physical examination with necessary laboratory studies.

Clinical history included duration of diabetes, whether they took oral hypoglycemic agents or injectable insulin, any other microvascular or macrovascular systemic complication (e.g., cardiovascular disease, diabetic foot, neuropathy, etc.).

Ophthalmic evaluation included thorough slit lamp examination, indirect ophthalmoscopy; slit lamp biomicroscopy with 90D and specular microscopy examination with Specular Microscope EM-3000, Topcon Medical Systems (TMS), Oakland, NJ, USA.

Fundus findings among cases were classified according to the early treatment for diabetic retinopathy study (ETDRS) classification.

Fasting blood sugar (FBS), postlunch blood sugar (PLBS), glycosylated hemoglobin (HbA1c) were the parameter to gauge severity and duration of glycemic control.

At the end of 15 months, the collected data were tabulated, statistically analyzed, and the severity of Type 2 DM was correlated with corneal morphology using the parameters of the specular microscope.


A case–control study was undertaken on 50 cases (100 eyes) and 50 controls (100 eyes). The CCT, the endothelial CD, and the coefficient of variation (CV) was studied. All the following parameters were assessed after classifying patients (cases) according to the ETDRS classification of DR. Cases were further classified according to age and sex as well as the duration of Type 2 DM. The statistical analysis was done using SPSS software (SPSS Statistics for Windows, Version 19.0. Armonk, NY: IBM Corp. USA). Unpaired t-test and linear regression analysis were used to determine effects of multiple variables on the various specular microscopy parameters.

The mean age among cases was 54.76 years and among controls was 52.02 years with a standard deviation of 9.2 and 7.3, respectively.

[Table 1] shows a linear regression analysis showing the relationship between age and CCT. With a P value of 0.043 and 0.039, a significant relationship is established between the two variables. There was an age-related decline in the thickness of the diabetic and nondiabetic corneas which was almost uniform amongt the cases as well as controls.{Table 1}

[Table 2] compares the average CCT among cases and controls. The mean CCT are highlighted in each group. It is found as per the above values that the control group had thinner corneas as compared to the cases. The P value between Group A and Group C was 0.021, and it was 0.026 between Groups B and C, thus making the above difference statistically significant.{Table 2}

[Table 3] shows the average endothelial CD among cases and controls. The mean values of each group are highlighted. It is seen from the above results that although there is not much of a difference between the CD of controls and cases without proliferative diabetic retinopathy (PDR), there is indeed a significant difference between controls and cases with PDR. The P value between Group A and Group C is < 0.05 making it statistically very significant, whereas the P value between Group B and Group C is 0.214 which is not significant.{Table 3}

[Table 4] gives a correlation between the CV amongst the cases and controls. The CV indicates polymorphism (variation of shape) and polymegathism (variation of size) of the corneal endothelium. The CV is higher in cases as compared to controls, and more so among those cases with PDR as opposed to those without. The P value for Group A and Group C is 0.024, whereas the P value for Group B and Group C is 0.037, which makes this correlation statistically significant.{Table 4}

[Table 5] demonstrates the CCT and CD values in males and females found in our study among the cases and the controls. With a P > 0.05 in all the above groups, no statistically significant correlation is found between the corneal endothelial morphology among males and females in the two groups.{Table 5}

In our study, out of 50 cases (100 eyes), there were 28 cases with mild nonproliferative diabetic retinopathy (NPDR), 15 cases with moderate NPDR, 4 cases with severe NPDR, and 3 cases with PDR.

[Table 6] shows that the mean CCT in microns and the mean CV is higher in patients with proliferative DR as opposed to those without, whereas the mean CD is lesser.{Table 6}

[Table 7] gives the Pearson correlation matrix, to determine if there is a significant effect of the glycemic controls in our cases on CCT, CD, and CV. These parameters of glycemic control are the FBS level, the PLBS which depict short-term glycemic control and the HbA1c (glycosylated hemoglobin), which depicts the long-term glycemic control. P values were calculated for each using regression analysis.{Table 7}

It was found that the short-term glycemic control, i.e., FBS and PLBS did not influence the CCT, CD, and CV significantly with a P > 0.05 in each group.

However, the HbA1c was positively correlated with the CCT. An increase in HbA1c was associated with an increase in CCT. It was also associated with an increase in CV and a decrease in CD, as is demonstrated by the r value in the above tables. This difference was statistically significant as the P value in each group associated with HbA1c is <0.05

[Table 8] depicts the distribution of cases depending upon the duration of the disease. As is clearly seen, our study had 52% patients with the disease duration between 1 and 5 years. We had only 28% of the total patients with disease duration more than 5 years. Hence, long-term sequelae of the disease on the morphology of the corneal endothelium could not be assessed in this particular study.{Table 8}

[Table 9] shows a regression analysis of the duration of diabetes among the cases and the CCT and CD values. With a P value of 0.73 and 0.72, we were not able to establish a strong correlation between the duration of the disease and its effect on the endothelial morphology.{Table 9}


Mean age in our study among cases was 54.76 years and mean age in controls was 52.02 years. All the subjects in the study were more than 40 years old.

A significant relationship is established between CCT and CD in cases versus controls, but there was an age-related decrease in CCT and CD in both the groups. This was in concordance with the study carried out by Galgauskas et al.[5] which found that CD and CCT decrease with age. The loss of endothelial cells is greater than the decrease in corneal thickness over the years, but the tendency of fewer endothelial cells in thin corneas remains.

Older people usually have thinner corneas. It was found that CCT declines about 4 μm (in male corneas) to 5 μm (in female corneas) every 10 years.[6]

CCT was significantly more in cases with PDR as compared to those without PDR, as indicated by a P value of 0.021. The increased corneal thickness in the diabetic patients may be due to either an increased water content or an increased dry weight content of the cornea or both. However, the presence of small folds in Descemet's membrane may be taken as evidence for an increased hydration of the cornea. The presence of folds in Descemet's layer is a well-known clinical feature of corneal edema, and it arises because the cornea can swell only in the posterior direction, thereby decreasing the area of the posterior corneal surface. The occurrence of minute folds in Descemet's membrane in diabetic patients has been reported by Henkind and Wise et al.[7]

The association between increased corneal thickness and severe retinal complications suggests that the corneal thickness may be an indicator of the risk of retinal complications in diabetic individuals. Other studies such as by Yazgan et al.[8] and Ravalico et al.[9] also reported a significant increase in CCT in diabetic patients.

In our study, we found that the mean CD in cases with PDR was 2396.0 and in cases without PDR was 2811.7 as compared to the mean control CD of 2791.8. Thus, the CD was lesser among cases with PDR as compared to controls and this difference was statistically significant with a P < 0.05. This was similar to the reduction found by Inoue et al.[10] in their study of Type 2 diabetics in Japan. Shenoy et al.[11] in Oman and Lee et al. in Korea also reported significant reductions in endothelial density in their studies on diabetics and insulin dependent diabetics, respectively.

It was found in our study that the mean CV among controls was 39.1, and the mean among cases with PDR was 40.7. The mean CV among cases without PDR was 36.1. Thus, the mean CV was highest amongst cases with PDR as compared to controls and cases without PDR. This difference was statistically significant with a P < 0.05.

The number of endothelial cells with polymegathism and polymorphism was significantly greater among eyes of diabetic patients. There were less corneal endothelial cells with polymorphism in nondiabetic patients. Polymegathism and polymorphism of corneal endothelial cells seems to be positively associated with DM Type 2. The presence of polymegathism, polymorphism, and reduction in density of corneal endothelial cells in Type 2 diabetic patients clearly shows that diabetes affects the corneal endothelium.

It is thought that intracellular accumulation of sorbitol, which acts as an osmotic agent leads to swelling of the endothelial cells. The Krebs cycle slows down with a consequent reduction in ATP production which is necessary for endothelial pump function.

This eventually results in morphological and permeability changes in the corneas of diabetic patients. The CV reflects polymegathism and polymorphism in the endothelial cells.

This study also found that the average size and CV of corneal endothelial cells is significantly increased in diabetics. The decrease in endothelial CD and increase in polymegathism correlate well as the existing endothelial cells have to enlarge to fill the gaps between adjacent cells, thereby increasing the CV.

There were a total of 26 males and 24 females among the cases and 31 males and 19 females among the controls. No statistically significant difference was found between the male and female patients as regard CCT and CD in our study.

This was similar to the findings in the study by Kaji et al.[12] which also allowed the assessment of the influence of sex on the corneal endothelium cell parameters. No significant relationship was observed in this study. However, Snellingen et al.[13] in South Asia found that early childhood development is 2.9% higher in women than in men.

We further divided all the cases as those with PDR and those without PDR (which included cases with mild, moderate and severe NPDR). When the aforementioned two groups were compared with age matched controls it was found that a statistically significant difference was present between the CD, CV, and CCT. However, studies conducted by Matsuda et al.[14] showed that there was no significant difference in the degree of these endothelial changes among the three diabetic groups. However another study done by Lee et al. showed that the corneal endothelial CD significantly decreased and the CV in cell size significantly increased for high risk PDR patients undergoing phacoemulsification, in contrast to normal persons at the postoperative 6 months.

In the diabetic group, lower cell counts were associated with higher HbA1c values (P < 0.05). Also higher values of HbA1c were associated with higher CV, indicating a greater degree of polymorphism and polymegathism in patients with poor glycemic control. The HbA1c was also positively correlated with the CCT. Higher the HbA1c value, greater was the CCT among cases.

This is in contrast to the results of the study conducted by Yee et al.[15] However, the aforementioned study was carried out on dogs who had Alloxan-induced DM. Hence, whether we can extrapolate the same conclusions to a human study remains to be confirmed with further research, using a larger database of reliable HbA1c values.

Our study found no significant effect of glycemic control on the morphology of the corneal endothelium. Similar such results have also been obtained in the study conducted by Matsuda et al.[14]

In our study, out of a total of fifty patients, we had ten patients in group 1, 26 patients in Group 2, and 14 patients in Group 3. The mean duration of diabetes in our study was 4.16 years with a standard deviation of 3.57.

There was a very weak correlation between duration of disease and the corneal endothelial morphology and it was not statistically significant. However, studies by Choo et al.[16] found that diabetes of over 10 years' duration showed thicker corneas, a lower corneal endothelium density, lower hexagonality ratios, and higher CV in cell size than those patients having diabetes of under 10 years' duration. With regard to the corneal endothelial morphology, the CV in cell size appears to be the most sensitive variable. Overall, this fact is likely to suggest that polymegathism and polymorphism may precede a decrease in CD.

In addition, the fact that our study failed to show any correlation between the duration of the disease and CCT, CV, etc., could be attributed to the fact that we had very few cases with duration of disease more than 10 years. Long-term prospective studies are essential for establishing such a relationship.

Busted et al.[17] showed that the diabetic corneal thickness was significantly thicker than the normal corneal thickness, but there was no significant relation between CCT of diabetes and the duration of the. This difference is probably due to the fact that the mean age of their subjects (34 years) was lower. A more active compensation of the corneal endothelial pump action seems to be present in younger diabetics than in older diabetics, because the corneal thickness of diabetics depends on the increased hydration of the cornea.


We conclude that it is important to critically assess the corneal endothelium with specular microscopy in all patients with Type 2 DM. It should be a part of the routine preoperative assessment in all patients undergoing intraocular surgery, as predicting postoperative corneal complications would be possible based upon the parameters of specular microscopy. It will be possible to identify preoperatively, a high-risk cornea which is prone to early decompensation based on the findings of specular microscopy.


Department of Ophthalmology, B.Y.L Nair Charitable Hospital, Mumbai Central, India.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1Diabetes Atlas. International Diabetes Federation; 2016. Available from: [Last cited on 2016 Mar 21].
2Anjana RM, Pradeepa R, Deepa M, Datta M, Sudha V, Unnikrishnan R, et al. Prevalence of diabetes and prediabetes (impaired fasting glucose and/or impaired glucose tolerance) in urban and rural India: Phase I results of the Indian Council of Medical Research-INdia DIABetes (ICMR-INDIAB) study. Diabetologia 2011;54:3022-7.
3Lutty GA. Effects of diabetes on the eye. Invest Ophthalmol Vis Sci 2013;54:ORSF81-7.
4Fukushi S, Merola LO, Tanaka M, Datiles M, Kinoshita JH. Reepithelialization of denuded corneas in diabetic rats. Exp Eye Res 1980;31:611-21.
5Galgauskas S, Norvydaite D, Krasauskaite D, Stech S, Ašoklis RS. Age-related changes in corneal thickness and endothelial characteristics. Clin Interv Aging 2013;8:1445-50.
6Urban B, Raczynska D, Bakunowicz-Lazarczyk A, Raczynska K, Kretowska M. Evaluation of corneal endothelium in children and adolescents with type 1 diabetes mellitus. Mediators Inflamm 2013;2013:913754.
7Henkind P, Wise GN. Descemet's wrinkles in diabetes. Am J Ophthalmol 1961;52:371-4.
8Yazgan S, Celik U, Kaldirim H, Ayar O, Elbay A, Aykut V, et al. Evaluation of the relationship between corneal biomechanic and HbA1C levels in type 2 diabetes patients. Auckland, NZ. Clinical Ophthalmology 2014;8:1549-53. doi:10.2147/OPTH.S67984.
9Ravalico G, Tognetto D, Palomba M, Calderini S, Vattovani O. Corneal endothelial function in diabetes: A fluorophotometric study. Ophthalmologica 1994;208:179-84.
10Inoue K, Kato S, Inoue Y, Amano S, Oshika T. The corneal endothelium and thickness in type II diabetes mellitus. Jpn J Ophthalmol 2002;46:65-9.
11Shenoy R, Khandekar R, Bialasiewicz A, Al Muniri A. Corneal endothelium in patients with diabetes mellitus: A historical cohort study. Eur J Ophthalmol 2009;19:369-75.
12Kaji Y, Usui T, Oshika T, Matsubara M, Yamashita H, Araie M, et al. Advanced glycation end products in diabetic corneas. Invest Ophthalmol Vis Sci 2000;41:362-8.
13Snellingen T, Rao GN, Shrestha JK, Huq F, Cheng H. Quantitative and morphological characteristics of the human corneal endothelium in relation to age, gender, and ethnicity in cataract populations of South Asia. Cornea 2001;20:55-8.
14Matsuda M, Ohguro N, Ishimoto I, Fukuda M. Relationship of corneal endothelial morphology to diabetic retinopathy, duration of diabetes and glycemic control. Jpn J Ophthalmol 1990;34:53-6.
15Yee RW, Matsuda M, Kern TS, Engerman RL, Edelhauser HF. Corneal endothelial changes in diabetic dogs. Curr Eye Res 1985;4:759-66.
16Choo M, Prakash K, Samsudin A, Soong T, Ramli N, Kadir A. Corneal changes in type II diabetes mellitus in Malaysia. Int J Ophthalmol 2010;3:234-6.
17Busted N, Olsen T, Schmitz O. Clinical observations on the corneal thickness and the corneal endothelium in diabetes mellitus. Br J Ophthalmol 1981;65:687-90.