Evaluation of Macular Thickness, Retinal Nerve Fiber Layer and Ganglion Cell Layer Thickness in Patients among Type 2 Diabetes Mellitus Using Optical Coherence Tomography.

BACKGROUND: Vascular abnormalities and microvasculopathy are one of the widely accepted factors of diabetic retinopathy (DR). Retinal Neuronal dysfunction and neurodegeneration are also important components in the pathogenesis of DR. However recent investigations show neurodegenerative alterations before the appearance of microvascular changes in patients having DR. AIMS AND
OBJECTIVES: (1) To measure the macular thickness, retinal nerve fiber layer thickness, and ganglion cell complex thickness among patients with type 2 diabetes mellitus using optical coherence tomography. (2) To compare the macular thickness, retinal nerve fiber layer thickness and ganglion cell complex thickness in type 2 diabetic patients with and without DR with normal controls using optical coherence tomography.
MATERIALS AND METHODS: Thirty Patients with type 2 diabetes mellitus without DR, 30 having mild and moderate DR and 30 healthy normals are taken considering the inclusion and exclusion criteria. Macular thickness, retinal nerve fiber layer (RNFL) thickness, ganglion cell layer-inner plexiform layer (GCL-IPL) thickness was measured in each individual and it was compared using one way ANOVA test, post hoc test and Pearson correlation was performed to evaluate the linear correlation between variables and calculated P < 0.05 was regarded as its significance.
RESULTS: The average RNFL thickness was 90.27 ± 5.57 and 107.7 ± 5.32 um in diabetic patients and controls respectively (P < 0.001). Furthermore, for two different groups of diabetic patients, the average RNFL thickness was 89.92 ± 6.62 um in the no DR group and 78.6 ± 3.93 in the DR group (P = 0.339). The average GCL-IPL thickness was 82.65 ± 2.25 um and 92.10 ± 2.41 um in diabetic patients and controls, respectively (P < 0.001). Furthermore, for two different groups of diabetic patients, the average GCL-IPL thickness was 82.22 ± 2.11 um in the no DR group and 71.55 ± 2.34 in the DR group (P = 0.535). The average macular thickness was 238.03 ± 4.42 and 277.9 ± 5.85 um in diabetic patients and controls, respectively (P < 0.001). Furthermore, for two different groups of diabetic patients, the average macular thickness was 236.56 ± 4.10 um in the no DR group and 242.8 ± 4.95 um in the DR group (P = 0.585).
CONCLUSION: There was a statistically significant reduction of mean RNFL, GCL-IPL and macular thickness in type 2 diabetic patients with no DR compared with a homogenous control group indicating neuroretinal changes occur before vascular changes of DR. Copyright:

INTRODUCTION

Diabetes mellitus is no longer considered an epidemic that can be ignored, with 80% of patients being concentrated in low- and middle-income countries. Diabetes mellitus is one among the heterogeneous group of diseases, characterized by the state of chronic hyperglycemia, resulting from a diversity of etiologies both environmental and genetic, which leads to serious damage to the heart, blood vessels, eyes, kidneys, and nerves.

Around 422 million people worldwide have diabetes, the majority living in low- and middle-income countries, and 1.6 million deaths are directly attributed to diabetes every year.[1] It is estimated to rise to 592 million by 2035.[2] World Health Organization estimates show diabetes was the seventh-leading cause of death in 2016.[3] Between 2000 and 2016, there was 5% increase in premature mortality from diabetes.[3]

According to World Health Organization reports, India tops the world in the number of diabetics.[4] The prevalence of diabetes in South East Asian region is 88 million and it is estimated to be 153 million by 2045. The prevalence of diabetes in India is 8.9% in the 40–79 year age group and around 77 million are affected with diabetes mellitus.[5] The increase is attributed to the rapid epidemiological transition along with urbanization and socioeconomic development.

Diabetic retinopathy (DR) is characterized by changes in the retinal microvasculature and entails endothelial cells and pericyte loss and the resulting features of them encompass microaneurysms, hemorrhages, capillary nonperfusion, and neovascularization.[6] Vascular abnormalities and micro-vasculopathy are one of the widely accepted factors of DR. Accordingly, the early clinical signs of DR include microaneurysms and retinal micro-hemorrhages.[7] Recent research, however have found that neuronal dysfunction, neurodegeneration which ought to be considered as very important component in the pathogenesis of DR, because of its correlation with the microvascular dysfunction.[8] This in turn is supported by the evidences from the animal studies and diabetic donors which shows that retinal neuronal cell degeneration occurs early in the course of diabetes.[9]

The evolution of new technologies such as Heidelberg retinal tomography 3, glaucoma diagnostics-variable corneal compensation, and optical coherence tomography (OCT) has made evaluation of the optic nerve head, the peripapillary area, the macula and retinal nerve fiber layer (RNFL) revolutionary. The resolution and reproducibility of such technologies almost give a near histological evaluation of the tissue or area which we study in the retina. The high resolution spectral domain OCT (SD-OCT) allows measurement of the thickness of all individual retinal layers, including retinal nerve fiber layer and ganglion cell layer.[10] Moreover, more than half of the ganglion cells of the retina are cantered on fovea and thus macula scan with ganglion cell analysis could be early indicator of this disease. OCT can be a useful test in this regard.

Even though there are studies exploring the same, the exact knowledge on neurodegeneration and retinal nerve layer thinning in diabetes is still not fully understood. Moreover, there is a paucity of studies on the same in this part of the country. Hence, this study was conducted to measure the macular thickness, retinal nerve fiber layer thickness, and ganglion cell complex thickness using optical coherence tomography and to compare the macular thickness, retinal nerve fiber layer thickness, and ganglion cell complex thickness with that of normal controls among type 2 diabetics (with and without retinopathy), aged 40 years and above attending the retina clinic.

MATERIALS AND METHODS

Study design

A comparative case-control study was conducted among type 2 diabetes mellitus patients attending Vinayaka Mission's Kirupananda Variyar Medical College and Hospital, Salem.

Study duration

The study was conducted for a total period of 1½ years from March 2019 to September 2020.

Study population

Inclusion criteria

Patients with type 2 diabetes mellitus aged 40 years and above attending the Department of Ophthalmology, Vinayaka Mission's Kirupananda Variyar Medical College and Hospital, Salem were included as cases. They were divided into those with retinopathy (Mild and Moderate DR) and those without retinopathy.

Exclusion criteria

Patients with Type 1 diabetes mellitus. High myopia (more than − 3D) and hypermetropia (+3D), glaucoma, any type of previous retinal treatment (laser photocoagulation, vitrectomy, intravitreal steroids, and/or antiangiogenic drugs), other ocular diseases such as uveitis, history of ocular trauma or neoplasm, and diabetic macular edema.

Controls

Healthy controls of more than or equal to 40 years of age and of both sex were randomly recruited from individuals accompanying patients.

Sample size and sampling

Assuming the RNFL thickness in diabetics as 86.18 (8.44) (86.74 [11.58] in patients without retinopathy vs. 85.62 [11.10] in patients with retinopathy) and 91.79 (4.77) among normal controls, 45 at 80% power and 95% confidence level, the sample size was estimated as 24 in each arm, using OpenEpi. Hence a sample of 30 was enrolled in each arm during the study duration, hence giving a total sample size of 90.

All patients with type 2 diabetes attending the Department of Ophthalmology have been recruited as cases, purposively until the sample size was reached. Simultaneously healthy controls were randomly selected from the accompanying individuals.

Study procedure

After obtaining permission from the institute ethics committee and informed consent from the participants, the recruited participants were informed about the nature and purpose of the study. The sociodemographic and clinical history of the cases and controls were then obtained using a predefined proforma.

Following that relevant ocular history was obtained and then, the participants were examined thoroughly for visual acuity using Snellen's chart and intraocular pressure was measured using Goldman's Application Tonometer. Anterior segment was then examined by slit-lamp examination and dilated fundus using slit lamp with 90D lens, direct and indirect ophthalmoscope. Indirect ophthalmoscope was used to grade the level of severity of the retinopathy and stereoscopic slit lamp biomicroscopy of the disc and macula using +90D lens. Fundus photographs were taken with fundus camera. DR was graded using the ETDRS system.

All participants were tested after pupil dilatation using the eye drops containing 0.8% tropicamide and 5% phenylephrine hydrochloride using an RS-330 Retina scan duo SD-OCT (NIDEK OCT system), with a scan speed of 53,000 axial scans per second and wavelength of 880 nm. RS-330 Retina scan Duo SD-OCT (NIDEK) was used to acquire a macular scan using macular cube 512 × 128 scan protocol. Macular thickness, retinal nerve fiber layer thickness, and ganglion cell layer thickness are measured using macular map scan and optic disc map scan.

Investigations such as HbA1C, fasting, and postprandial sugar were also done to evaluate the control status of diabetics.

Statistical analysis

Data were analyzed using Statistical Package for Social Sciences V21 for windows. Categorical variables such as gender and type of treatment are expressed as frequency and percentages. Continuous variables such as age and duration of diabetes are expressed as mean (standard deviation) or median (interquartile range), depending on the type of distribution. One-way ANOVA method was used to compare the macular thickness RNFL and GC-IPL thickness between the three groups. Post hoc test was performed using Bonferroni correction and P < 0.05 was regarded as significant. Chi-square test was used to compare the categorical variables and independent t-test was used to compare continuous variables. Pearson's correlation is used to determine the correlation between the continuous variables.

Approval of research review board

Ethical approval was sought from the Institutional Ethics Committee. Informed consent was obtained priorly from the study participants before data collection. Confidentiality was maintained by limiting the identifying variables to the minimum. Datas were analyzed in aggregate and accessed only to the authors.

RESULTS AND OBSERVATIONS

A total of 90 participants, 60 patients with diabetes were taken as cases and 30 normal individuals were chosen as controls. Among the diabetics, 30 patients with DR and 30 patients without DR. A total of 180 eyes (120 among the diabetics and 60 among the controls) were used for the comparison between the cases and controls.

Table 1 compares the baseline characteristics of the participants in terms of gender and age. It can be noted that the gender is comparable between the groups but the mean age was found to be significantly higher among cases when compared to healthy controls (P < 0.001). Among the cases, patients with retinopathy were found to significantly older when compared to patients without retinopathy (P < 0.001).

Table 1

Sociodemographic characteristics of the participants (n=90)

Characteristics	Diabetic cases (n=60)		Controls (n=30)	P
	Without retinopathy (n=30)	With retinopathy (n=30)
Gender, n (%)
Males	18 (60.0)	17 (56.7)	13 (43.3)	0.392
Females	12 (40.0)	13 (42.3)	17 (56.7)
Age (years), mean (SD)	56.6 (6.3)	63.3 (4.3)	47.8 (4.4)	<0.001*

*One way ANOVA. SD: Standard deviation

Table 2a shows that there was significant difference in mean macular thickness between the three groups in all the fields (P < 0.001).

Table 2a

Macular thickness in the diabetic patients and the controls (n=180)

Macular thickness (µM)	Diabetic eyes (n=120), mean (SD)		Control eyes (n=60), mean (SD)	P*
	With retinopathy (n=60)	Without retinopathy (n=60)
Nasal outer	246.2 (5.6)	265.7 (3.6)	278.0 (4.8)	<0.001
Nasal inner	262.2 (6.9)	287.6 (3.9)	302.1 (4.5)	<0.001
Temporal outer	230.4 (3.0)	243.7 (5.0)	259.2 (4.3)	<0.001
Temporal inner	255.6 (6.2)	274.9 (3.4)	288.2 (4.9)	<0.001
Superior outer	239.8 (3.6)	250.3 (3.3)	268.7 (4.6)	<0.001
Superior inner	260.8 (5.6)	282.9 (4.4)	301.1 (6.3)	<0.001
Inferior outer	230.7 (3.3)	240.5 (3.2)	261.3 (5.7)	<0.001
Inferior inner	254.3 (5.8)	272.4 (4.6)	291.3 (13.4)	<0.001
Central subfield	205.5 (4.6)	218.6 (8.4 )	251.5 (4.2)	<0.001

*One way ANOVA. SD: Standard deviation

Table 2b shows the Bonferroni post hoc analysis of the macular thickness between the groups. It can be noted that thinning is significant in DR patients when compared to patients without retinopathy and the controls. Similarly, macula was significantly thinner in patients without retinopathy when compared with healthy controls.

Table 2b

Bonferroni post hoc analysis for the macular thickness between the groups (n=180)

Macular thickness (µM)	Diabetic retinopathy versus diabetic nonretinopathy	Diabetic retinopathy versus controls	Diabetic nonretinopathy versus controls
Nasal outer	<0.001	<0.001	<0.001
Nasal inner	<0.001	<0.001	<0.001
Temporal outer	<0.001	<0.001	<0.001
Temporal inner	<0.001	<0.001	<0.001
Superior outer	<0.001	<0.001	<0.001
Superior inner	<0.001	<0.001	<0.001
Inferior outer	<0.001	<0.001	<0.001
Inferior inner	<0.001	<0.001	<0.001
Central subfield	<0.001	<0.001	<0.001

Table 3a shows that there was a significant difference in mean retinal nerve fiber thickness between the three groups in all the fields (P < 0.001).

Table 3a

Retinal nerve fiber layer thickness in diabetic patients and control (n=180)

Retinal nerve fiber thickness (µM)	Diabetic eyes (n=120), mean (SD)		Control eyes (n=60), mean (SD)	P*
	With retinopathy (n=60)	Without retinopathy (n=60)
Nasal	57.6 (3.7)	65.7 (4.8)	82.6 (6.1)	<0.001
Temporal	56.7 (3.7)	65.2 (4.0)	71.4 (5.0)	<0.001
Superior	98.9 (6.6)	114.7 (6.6)	136.7 (5.3)	<0.001
Inferior	101.2 (6.9)	115.5 (6.9)	140.1 (4.9)	<0.001

*One way ANOVA. SD: Standard deviation

Table 3b shows the Bonferroni post hoc analysis of the retinal nerve fiber thickness between groups. It can be noted that thinning is significant in DR patients when compared to patients without retinopathy and the controls. Similarly, the macula was significantly thinner in diabetics without retinopathy when compared with the healthy controls.

Table 3b

Bonferroni post hoc analysis for the retinal nerve fiber thickness between the groups (n=180)

Retinal nerve fiber thickness (µM)	Diabetic retinopathy versus diabetic nonretinopathy	Diabetic retinopathy versus controls	Diabetic nonretinopathy versus controls
Nasal	<0.001	<0.001	<0.001
Temporal	<0.001	<0.001	<0.001
Superior	<0.001	<0.001	<0.001
Inferior	<0.001	<0.001	<0.001

Table 4a shows that there was a significant difference in mean ganglion cell layer-inner plexiform layer (GCL-IPL) thickness between three groups in all the fields (P < 0.001).

Table 4a

Ganglion cell layer-inner plexiform layer thickness in diabetic patients and control (n=180)

GCL-IPL thickness (µM)	Diabetic eyes (n=120), mean (SD)		Control eyes (n=60), mean (SD)	P*
	With retinopathy (n=60)	Without retinopathy (n=60)
Superior	71.6 (3.0)	82.9 (2.3)	92.3 (2.2)	<0.001
Superonasal	71.2 (2.8)	82.6 (2.2)	91.9 (2.5)	<0.001
Superotemporal	71.7 (2.7)	82.6 (2.0)	92.5 (2.4)	<0.001
Inferior	71.4 (3.0)	82.1 (2.5)	91.9 (2.7)	<0.001
Inferonasal	71.9 (2.5)	83.2 (1.9)	91.9 (2.4)	<0.001
Inferotemporal	71.5 (2.5)	82.5 (2.6)	92.6 (2.3)	<0.001

*One way ANOVA. GCL-IPL: Ganglion cell layer-inner plexiform layer, SD: Standard deviation

Table 4b shows the Bonferroni post hoc analysis of the GCL-IPL thickness between the groups. It can be noted that thinning is significant in DR patients when compared to the patients without retinopathy and the controls. Similarly, macula was significantly thinner in patients without retinopathy when compared with the healthy controls.

Table 4b

Bonferroni post hoc analysis for the ganglion cell layer-inner plexiform layer thickness between the groups (n=180)

Retinal nerve fiber thickness (µM)	Diabetic retinopathy versus diabetic nonretinopathy	Diabetic retinopathy versus controls	Diabetic nonretinopathy versus controls
Superior	<0.001	<0.001	<0.001
Superonasal	<0.001	<0.001	<0.001
Superotemporal	<0.001	<0.001	<0.001
Inferior	<0.001	<0.001	<0.001
Inferonasal	<0.001	<0.001	<0.001
Inferotemporal	<0.001	<0.001	<0.001

Table 5 shows that there was no significant change in macular thickness with a change in grade of retinopathy.

Table 5

Comparison of macular thickness with the grade of diabetic retinopathy (n=60)

Macular thickness (µM)	Grade of retinopathy, mean (SD)		P
	Mild (n=30)	Moderate (n=30)
Nasal outer	246.1 (5.6)	246.3 (5.6)	0.927
Nasal inner	262.6 (6.5)	261.9 (7.4)	0.726
Temporal outer	230.3 (2.8)	230.4 (3.3)	0.900
Temporal inner	253.8 (6.3)	257.5 (5.7)	0.051
Superior outer	240.1 (3.1)	239.4 (4.1)	0.434
Superior inner	260.8 (5.7)	260.9 (5.6)	0.964
Inferior outer	230.9 (3.1)	230.5 (3.6)	0.674
Inferior inner	253.7 (5.5)	254.9 (6.1)	0.413
Central subfield	204.7 (4.6)	206.3 (4.5)	0.177

SD: Standard deviation

Table 6 shows that as the grade of retinopathy increases, there is a significant decrease in the RNFL thickness in the temporal quadrant (P = 0.002), but there was no significant change in RNFL thickness with changes in grades of retinopathy in any other quadrants.

Table 6

Comparison of retinal nerve fiber layer thickness with the grade of diabetic retinopathy (n=60)

Macular thickness (µM)	Grade of retinopathy, mean (SD)		P
	Mild (n=30)	Moderate (n=30)
	Nasal	59.0 (3.8)		56.2 (2.9)	0.002
Temporal	56.7 (4.0)	56.7 (3.5)	0.973
Superior	98.2 (6.3)	99.7 (6.9)	0.397
Inferior	102.3 (7.6)	100.1 (6.0)	0.227

SD: Standard deviation

Table 7 shows that there was no significant change in GCL-IPL thickness with change in grade of retinopathy.

Table 7

Comparison of ganglion cell layer-inner plexiform layer thickness with the grade of diabetic retinopathy (n=60)

Macular thickness (µM)	Grade of retinopathy, mean (SD)		P
	Mild (n=30)	Moderate (n=30)
Superior	71.6 (3.1)	71.7 (2.9)	0.865
Superonasal	71.8 (2.8)	71.3 (2.8)	0.821
Superotemporal	71.8 (2.7)	71.6 (2.8)	0.780
Inferior	71.1 (2.9)	71.7 (3.0)	0.419
Inferonasal	71.4 (2.8)	72.4 (2.2)	0.158
Inferotemporal	71.9 (2.3)	71.1 (2.7)	0.171

SD: Standard deviation

DISCUSSION

There is a need for DR to be prevented in cases with diabetes to avoid permanent visual loss and thus, the early diagnosis and management of functional changes related to DR can help a long way in preventing retinal degeneration. Moreover, there are conflicting evidence for the retinal degeneration occurring before DR and hence this study could be a preliminary step to understand the phenomenon.

Our study showed that the mean age of the participants was 47.8 years and these results are similar to the studies conducted by Mehboob et al.,[11] Ng et al.[12] and Pincelli Netto et al.,[13] in which the mean age group of the diabetic patients was 44.63 years. 58.2 years and 57.3 years respectively. This is in line with theory that diabetes is found to be more common in the age group of 40–60 years. A study by Koopman et al.[14] had reported that the mean age at presentation of diabetes to the hospital had reduced from 52 to 46 years and which is true with our study too. These changes may represent earlier onset of type 2 diabetes or earlier detection or combination of both the effects.

The mean duration of the disease was 11.0 years in patients without retinopathy and 17.5 years in patients with DR. A study by Fong et al.[15] has reported that the onset of retinopathy was 60% after 10 years of onset and 80% after 15 years of onset of the disease. Our study finding clearly supports the hypothesis. Majority of the patients were on oral hypoglycemic drugs in both groups. However, diabetic control was less in the cases with retinopathy which might explain the reason behind uncontrolled diabetes and retinopathy as a result. However, the difference was not significant to prove our hypothesis. Fasting blood sugar and postprandial blood sugar was not under control both in patients without DR and in patients with DR and generally, it is believed that 30% of the patients will have their blood sugar under control and our findings were contrary to our belief. HbA1C% was under control in around 25% of the patients without DR and only in 7% of the patients with DR. The poor glycemic control could also point towards the development of retinopathy among these patients.

A study by Funatsu et al.[16] had reported that when the HbA1C% decreases by more than 2%, there will be the rapid progression of DR. The explanation for this could be some molecular mechanisms in which insulin could stimulate neovascularization. Vascular endothelial growth factor is one of the main mediators of neovascularization or ischemia. On best-corrected visual acuity, there was a statistical significance between the controls and the patients with diabetes mellitus. Visual impairment was higher among the patients with diabetes. Even among the patients with diabetes, visual impairment was significantly higher among those with DR when compared to those without DR. The differences in values may be due to associated macular edema which often accompanies NPDR and PDR group. These findings are similar to the findings of various studies across the world.[17]

Our study showed that the macular thickness was significantly lower in patients with diabetes when compared with the controls. On post hoc analysis, the macular thickness showed statistical difference in patients with DR when compared to patients without DR. These findings are consistent with the findings of studies conducted by Borooah et al.,[18] and Shawky et al.[19] Similarly, the retinal nerve fiber thickness found to be significantly lower in the diabetic group when compared to the normal controls. On post hoc analysis, significance was present between patients with DR and the patients without DR, with patients with DR having thinner retinal nerve fiber layer. The significant thinning was present for GCL-IPL thickness.

GCL-IPL thickness was significantly lower among the patients with diabetes when compared to the normal controls. Even among the diabetics, the significance was present between the DR group and nonDR group. Exact mechanisms for inner retinal loss are not yet clear but have been related to lower perfusion and higher metabolic demands of the inner retina which makes it more vulnerable to the metabolic stress induced by hyperglycemia.[19] It can also be stated that the thinning of the inner layers of the retina is related to disease duration also. A study by Afef et al.[20] also reported similar results.

Strengths and limitations

One of the limitations of the study is that it involved a relatively small sample size and hence the problem of external validity. It can still be argued that the study could be generalizable to a similar setting. Another limitation is that there was no proper randomization of the participants which might have resulted in the selection bias. The diagnosis of the retinal thickness using optical coherence tomography adds to the strength of the study. On best-corrected visual acuity, there was a statistical significance between the controls and the patients with diabetes mellitus. Visual impairment was higher among the patients with diabetes. Even among the patients with diabetes, visual impairment was significantly higher among those with DR when compared to those without DR.

CONCLUSION

The mean age of the participants was 47.8 years and the mean duration of disease was 11.0 years in patients without retinopathy and 17.5 years in patients with DR. HbA1C% was under control in around 25% of the patients without DR and only in 7% of the patients with DR. The macular thickness, RNFL thickness, and GCL-IPL thickness were significantly lower in the patients with diabetes when compared with the controls. On post hoc analysis, the macular thickness, RNFL thickness and GCL-IPL thickness showed statistical difference in patients with DR when compared to patients without DR. There was a significant negative correlation between age, duration of diabetes, and HbA1C levels and the macular, RNFL, and GCL-IPL thickness in our study. However, there was no association between the grade of retinopathy and the retinal thickness. This study detected morphological changes in DM patients using OCT, confirmed that the loss of neural tissue begins in the early stages of diabetes. This supports the presence of the neurodegenerative process in eyes of patients with diabetes and warrants neuroprotective intervention to prevent neurodegeneration. OCT may represent a valuable tool for identifying early signs of neurodegeneration in diabetic eyes. It will help in monitoring the patient. Further, studies are necessary to understand whether ganglion cell neuroretinal degeneration and microvascular damages are pathogenically linked and whether neuro retinal degeneration represents a target in diabetes treatment to prevent DR.

Financial support and sponsorship

This is a self-financed study.

Conflicts of interest

There are no conflicts of interest.