Evaluation of corneal backward light scattering in type 2 diabetes mellitus

Amira Elagamy, Najd Abaalhassan, Mohamed Berika

1Department of Optometry and Vision Sciences, College of Applied Medical Sciences, King Saud University, Riyadh 11433, Saudi Arabia

2Ministry of Health Quality and Patient Safety, Riyadh 11433,Saudi Arabia

3Rehabilitation Science Department, College of Applied Medical Sciences, King Saud University, Riyadh 11433, Saudi Arabia

Abstract

● KEYWORDS: corneal backward light scattering;densitometry; type 2 diabetes mellitus; Pentacam

INTRODUCTION

It is anticipated that the number of diabetic patients will reach almost 552 million by the year 2030[1].Saudi Arabia is considered the second highest in the Middle East and is the seventh in the world for the frequency of diabetes mellitus (DM)[2].Type 2 DM constitutes about 90% of cases and is characterized by overweight and insulin resistance[3].Many studies[4-9]evaluated the impact of DM on the corneal endothelium, epithelium, corneal sensitivity and central corneal thickness.However, a few number of studies[10-13]investigated the effect of DM on the corneal backward light scattering and their results were controversial.Yang and Li[10]showed significantly higher densitometry values in 2 to 6 mm central zones of total cornea in the diabetic eyes compared to the control eyes.However, Rammet al[11]confirmed that the corneal densitometry of all corneal zones of the diabetic eyes [insulin dependent diabetes mellitus (IDDM) and noninsulin dependent diabetes mellitus (NIDDM) patients] was significantly reduced than in the control eyes.Yang and Li[10]found a weak relationship between the existence of DM and the corneal densitometry at 0 to 2 mm and 2 to 6 mm corneal zones in the anterior, the posterior and the total corneal depth.Conversely, Calvo-Marotoet al[12]documented no significant associations between the corneal densitometry of the central zones (3 and 5 mm) of the diabetic eyes (IDDM and NIDDM patients) and DM duration.Rammet al[11]demonstrated a significant negative association between hemoglobin A1c(HbA1c) value and the densitometry values of the total central layer (r=-0.424,P=0.044).However, ?zyolet al[13]found no significant correlation between the total corneal densitometry and HbA1c levels.Rammet al[11]confirmed a significant negative correlation between the densitometry values and the diabetic retinopathy (DR) staging (r=-0.271,P=0.019).Their study documented that the diabetic eyes with maculopathy showed lower total densitometry values than eyes without maculopathy.On the contrary, Calvo-Marotoet al[12]showed no significant association between the corneal densitometry of central zones (3 and 5 mm) of the diabetic eyes (IDDM and NIDDM patients) and DR staging.

It was established that the Scheimpflug system was superior in the objective evaluation of back scatter of the entire cornea compared to confocal microscopy which can assess only a sample area.In addition, it is a quick, and reliable method for measuring corneal densitometry[13-14].

The aim of this study was to compare the corneal backward light scattering values in type 2 diabetic patients with those of healthy age and sex-matched non-diabetic controls using Pentacam HR.In addition, the correlation between corneal backscattering values and DM duration, HbA1c levels, status of DR was evaluated.

SUBJECTS AND METHODS

Ethical ApprovalThe study was a hospital-based, nonrandomized, case-control, observational and quantitative study.It got the Institutional Review Board (IRB) approval of Health Sciences Colleges Research on Human Subjects, College of Medicine, King Saud University, Riyadh, Saudi Arabia[30.10.2022 (05.04.1444) Ref.No.22/0811/IRB].It adhered to the tenets of the Declaration of Helsinki.All the participants signed a comprehensive consent after explanation of the possible consequences of the study prior to investigations.

Study DesignAll patients were recruited from Magrabi Hospital, Riyadh, Saudi Arabia by the non-randomized convenience sampling method.The diagnosis of type 2 DM was based on the criteria of the World Health Organization(WHO)[15].Control patients did not have DM that was confirmed by random blood sugar test.The inclusion criterion was age 40-80y.

The study excluded participants who have history of past intraocular surgery, previous refractive surgery, contact lens use within 2wk, corneal abnormalities such as keratoconus,corneal scarring pathology such as infections, dystrophies,trauma, ectatic corneal diseases, or age-related corneal degenerations.Also, patients with intraocular pressure (IOP)more than 21 mm Hg were excluded in this study.

Age, gender, duration of diabetes, most recent HbA1c, along with the status of DR, and existing medical treatment were recorded.All subjects underwent a complete ophthalmologic examination including visual acuity evaluation utilizing Snellen chart, refraction using auto-refractometer (Topcon RK-8800 Autorefractor/keratometer, Japan) and IOP measurement using air puff tonometer, slit-lamp biomicroscopy, and fundus examination.If both eyes met the eligibility criteria, a randomized eye was selected for statistical analysis.

Table 1 Demographics of control and diabetic patients mean±SD (range)

Backward light scattering (densitometry) was measured to assess changes in corneal transparency using Scheimpflug tomography (Pentacam HR; Oculus, Wetzlar, Germany).The imaging is intended to mechanically trace corneal apex and get nearly 25 images above different cornea’s meridians along with the uniform source of blue light.The grayscale unit is used to label the outcome of corneal densitometry; the value ranges from zero (i.e., least light scattering) to a hundred(i.e., greatest light scattering).For the assessment of the local densitometry, the 12 mm diameter area was divided into 4 concentric radial zones: a central zone 2 mm, a second zone from 2 to nearly 6 mm while the next zone from 6 to nearly 10 mm and lastly, a fourth zone from 10 to 12 mm.The corneal depth is divided into: The anterior layer constitutes the anterior 120 μm, while the posterior layer constitutes the most posterior 60 μm of the cornea.The central corneal layer is calculated by deducting the anterior and posterior layers from the total corneal thickness[10,12-13].Three replicate measurements of each eye were documented, and the mean of these measurements were used for statistical analysis.

Statistical AnalysisData was analyzed using statistical software (SPSS version 22.0, SPSS Inc., USA).Descriptive statistics [mean±standard deviation (SD), 95% confidence intervals (CI), minimum, maximum and frequencies] were used for assessing the demographics and the clinical parameters.Normality of the data distributions for the two groups was determined using the Kolmogorov-Smirnov test.Unpairedttest was used for comparison between the diabetic and nondiabetic patients.Pearson correlation test was performed to find the association between the backscattering values and HbA1c levels and DM duration.Spearman correlation test was performed to find the association between the backscattering values and the presence or the absence of DR.A value ofP＜0.05 was considered statistically significant and a value ofP＜0.01 was considered statistically highly significant.

RESULTS

The study included 23 diabetic eyes without DR, 7 eyes with non-proliferative DR (NPDR) and 30 control (non-diabetic)eyes.There was no significant difference in mean age or sex.The demographics and the clinical characteristics of the participants are summarized in Table 1.Mean of the anterior corneal (6-10 mm) densitometry values was significantly higher in the diabetic group (30.9, 95%CI,29.1-32.6) than in the control group (25.7, 95%CI, 23.8-27.6;P=0.047).Also, mean of the anterior corneal total densitometry values was significantly higher in the diabetic group (28.3,95%CI, 27.1-29.5) than in the control group (24.8, 95%CI,23.8-25.9;P=0.036).In addition, mean of the total corneal densitometry was significantly higher in the diabetic group(20.0, 95%CI, 19.3-20.7) than in the control group (17.9,95%CI, 17.2-18.6;P=0.043).Descriptive statistics of the corneal densitometry of the study groups are shown in Table 2.The corneal densitometry of the diabetic eyes demonstrated no significant correlation with HbA1c and DM duration using Pearson’s correlation coefficients (Table 3).Moreover,the corneal densitometry of diabetic eyes demonstrated no significant correlation with the presence or the absence of DR using Spearman correlation coefficients (Table 4).

Table 2 Descriptive statistics for corneal densitometry for study groups

Table 3 Pearson’s correlation coefficients between clinical measurements within diabetic group

Table 4 Spearman’s correlation coefficients between clinical measurements within diabetic group

DISCUSSION

This study was performed to compare the corneal backward light scattering values in type 2 diabetic patients with those of age and sex-matched non-diabetic healthy controls in Riyadh, Saudi Arabia.The diabetic group in this study demonstrated significantly higher mean densitometry values of the anterior 6-10 mm zone (P=0.047), the total anterior layer(P=0.036) and the total cornea (P=0.043) than in the control group.This finding was in agreement with ?zyolet al[13]who found that the total anterior layer and the total corneal densitometry in the 0 to 2 mm zone showed significantly higher values in the diabetic eyes than in the controls(P=0.024,P=0.020 respectively).However, they documented no significant difference in any concentric zones in the center and the posterior layers.As well, their study did not find any significant difference in the total corneal densitometry of the anterior to the posterior layers over the entire 12 mm diameter area, though the values of the diabetic eyes were greater than the controls in their study.They clarified this finding by the normal variations in total corneal diameter that leads to an inclusion of sclera in the measurement in eyes with a smaller corneal diameter.Moreover, the subclinical age-associated deteriorating situations may influence the measurements in their study.Yang and Li[10]showed significantly higher densitometry values in 2 to 6 mm central zones of total cornea in the diabetic eyes compared to the control eyes in 70 to 79y age group (P=0.003).Additionally, they documented a higher total light backscatter at the total corneal thickness at all concentric radial zones.They predicted that the different results reported by the different studies may be due to the difference of the population samples.In addition, they reported that the corneal densitometry measured using Scheimpflug system may be a valuable technique for early screening and management of the diabetic peripheral neuropathy and the measurable assessment of the corneal transparency.

Calvo-Marotoet al[12]calculated the mean central corneal backscatter for each eye from the seven different meridians analyzed (range 70° to 110°).They demonstrated significant differences in the corneal backscatter between the diabetic and the control groups for the 3-mm and 5-mm central zones (P=0.016,P=0.014 respectively).Furthermore, they confirmed significantly higher corneal backscatter for the 3-mm central zone than for the central 5-mm in the control subjects(P＜0.001).Conversely, no significant difference in the corneal backscatter was detected between the 2 zones for the diabetic patients in their study.However, Rammet al[11]confirmed that the corneal densitometry of all corneal zones of the diabetic eyes (IDDM and NIDDM patients) was significantly reduced than in control eyes.However, Tekinet al[16]detected no significant difference in the densitometry values between the diabetic patients with type 1 DM and the controls.They clarified this outcome by the inclusion of patients in their study with controlled HbA1c values and short DM durations.They excluded patients with DR or maculopathy.

Corneal densitometry in diabetic eyes in the current study demonstrated no significant association with DM duration using Pearson’s correlation coefficients.This finding agreed with Calvo-Marotoet al[12]who documented no significant association between the corneal densitometry of the central zones (3 and 5 mm) of the diabetic eyes (IDDM and NIDDM patients) and DM duration.Likewise, Rammet al[11]found no significant association between DM duration and the densitometry values in their study.Conversely, Yang and Li[10]found a weak relationship between the existence of DM and the corneal densitometry at 0 to 2 mm and 2 to 6 mm corneal zones in the anterior, the posterior, and the total corneal depth.As well, ?zyolet al[13]found highly significant positive correlation between the anterior total corneal densitometry values and DM duration (r=0.802,P=0.001).This finding can be attributed to the longer DM duration in their study.The mean of DM duration in their study was 13.0±5.9y (1-16y)compared to that of our study 9.7±5.2y (1-17y).

The corneal densitometry in the diabetic eyes in this study documented no significant relationship with HbA1c using Pearson’s correlation coefficients.This result matched with ?zyolet al[13]who did not find any significant relationship between the total corneal densitometry and HbA1c levels.They elucidated this result by the fact that HbA1c values specifies mean glycemic values in the previous 2 to 3mo only which may vary from time to time and they evaluated the association between the corneal densitometry and HbA1c values obtained at that time only as we did in our study.In opposition, Rammet al[11]demonstrated a significant negative association between the HbA1c value and the densitometry values of the total central layer (r=-0.424,P=0.044).

The corneal densitometry in the diabetic eyes in the present study, demonstrated no significant correlation with presence or absence DR using Spearman correlation coefficients.This outcome may be due to the small sample size and the short DM duration of the patients enrolled in this study.This was in agreement with Calvo-Marotoet al[12]who showed no significant association between the corneal densitometry of the central zones (3 and 5 mm) of the diabetic eyes (IDDM and NIDDM patients) and DR staging.They clarified this finding in their pilot study because of the small sample size and the unconsidered demonstration of the different stages of DR.On the other hand, Rammet al[11]confirmed a significant negative correlation between the densitometry values and DR staging(r=-0.271,P=0.019).Their study documented that the diabetic eyes with maculopathy showed lower total densitometry values than eyes without maculopathy.

Regarding the normative values of the corneal densitometry in the control group, this study found higher anterior corneal total densitometry values than for both the central and the posterior layers.Mean of the anterior, the central, the posterior and the total corneal densitometry values was 24.8 (23.8-25.9), 15.9(15.2-16.5), 13.1 (12.6-13.5) and 17.9 (17.2-18.6) respectively.This finding matched with Ní Dhubhghaillet al[17]who showed significantly higher densitometry values of the anterior layer 25.81±5.14 than for both the central (P＜0.001) and the posterior layers (P＜0.001).Additionally, they demonstrated the lowest densitometry values in the central zone (0-2 mm,16.76±1.87) and the highest values in the peripheral zone (10-12 mm, 27.36±7.47).Their finding was in agreement with the current study which showed densitometry values in the central zone 16.7 (16.3-17.1) compared to 21.4 (20.4-22.5)in the peripheral zone.This outcome can be explained by the normal variability of the white-to-white corneal diameter[18].Ní Dhubhghaillet al[17]further elucidated that finding by the inclusion of portions of the limbus and sclera in the evaluation of the densitometry values of the peripheral zones in subjects with a small corneal diameter may result in higher values.

The augmented light backscattering in the diabetic patients can be attributed to corneal edema, increased corneal thickness and polymegathism or pleomorphism of corneal endothelium[19].Yang and Li[10]speculated that the anterior layer which corresponds to the corneal epithelial and anterior stromal strata as demarcated with corneal densitometry software was the most vulnerable layer in the diabetic cornea for the increased backward light scattering.Moreover, ?zyolet al[13]explained the significant difference of the backscattering in their diabetic group compared to the control group by the existence of ultrastructural alterations in the anterior layer of cornea in the diabetic eyes such as a condensed, multilayered epithelial basement membrane, glycogen granules buildup in the basal epithelial cells, motivation of the keratocytes and prompted stromal collagen crosslinking[20].Furthermore, Quadradoet al[21]documented significant declines in basal epithelial cells, anterior stromal keratocytes and endothelial cell concentrations in cornea of type 2 diabetic patientsviain vivoconfocal microscopy.Additionally, Gruset al[22]reported that the abnormalities of tear film composition and thickness may contribute to the elevated backscattering in the diabetic cornea.Light scattering can be divided into an intraocular scattering(forward light scattering) and light scattered backward(backward light scattering)[12].No accurate relationship between them was detected until nowadays[10,23].Hwanget al[24]demonstrated an increased level of the intraocular scattering in the diabetic corneas compared to the controls and confirmed the association between the severity of DR and the increased intraocular scattering.Schiano Lomorielloet al[25]evaluated early changes of the corneal sub-basal plexus in type 1 diabetic cornea usingin vivoconfocal microscopy and reported a significant affection of the optical quality of the diabetic cornea compared to controls.Patelet al[26]assessed the influence of the intraocular scattering on vision after penetrating keratoplasty and confirmed that the increased stray light may cause visual complaints.Ní Dhubhghaillet al[17]reported that the significant correlation between the increased backward light scattering and the reduced visual quality is not essentially established.

This study is the first one in Saudi Arabia to evaluate the corneal backward light scattering values in type 2 diabetic patients using Pentacam HR.This study documented a statistically significant difference in the densitometry of the anterior corneal (6-10 mm zone), the total anterior cornea and the total cornea of the diabetic group compared to the control group.This study confirmed the potentiality of using optical quality assessment for monitoring of the corneal abnormalities caused by DM.Further prospective longitudinal studies using a large sample size are recommended to confirm these results and to evaluate the correlation between corneal backscattering values and HbA1c levels, DM duration, status of DR.In addition, the effect of the corneal backscatter on the visual quality of the diabetic cornea needs more investigations.Additionally, the relationship between the intraocular scattering and the backward light scattering necessitates verification.

ACKNOWLEDGEMENTS

The authors extend their appreciation to The Deputyship for Research & Innovation, Ministry of Education in Saudi Arabia for funding this research work through the project (No.IFKSUOR3-499-1).

Conflicts of Interest: Elagamy A,None;Abaalhassan N,None;Berika M,None.