Dry eye disease in diabetics versus non-diabetics: Associating dry eye severity with diabetic retinopathy and corneal nerve sensitivity

Abstract

Purpose:

To compare the prevalence of dry eye disease (DED) and assess corneal nerve sensitivity (CNS) in diabetic and non-diabetic patients. To study the association of severity of DED in patients with diabetic retinopathy (DR) and CNS in DED.

Methods:

A cross-sectional prospective comparative study was conducted on 400 patients attending the ophthalmology OPD. The patients above 18 years of age were divided into two groups—diabetic (T2DM) and non-diabetic. All patients were subjectively assessed for DED based on Standard Patient Evaluation of Eye Dryness (SPEED) questionnaire and objectively, using Schirmer’s II test and Tear Film Break-Up Time (TBUT). Visual acuity assessment, anterior segment, and posterior segment evaluation were done.

Results:

Considering the SPEED score, Schirmer II values, TBUT values, and Dry Eye Work Shop (DEWS) II diagnostic criteria, mild DED was seen in 23% diabetic and 22.25% non-diabetic groups, moderate DED in 45.75% diabetic and 9.75% non-diabetic groups, and severe DED in 2% diabetic and 1.75% non-diabetic groups. Moderate DED was more common within all grades of DR. CNS was reduced more in diabetic group and also in patients with higher degree of DED.

Conclusion:

Prevalence of DED is more in the patients with T2DM. CNS was reduced more in patients with T2DM and in patients with moderate DED. Our study also correlated that severity of DR affects the severity of DED.

Dry eye disease (DED) is a multifactorial ocular surface disorder caused due to a vicious cycle of tear film instability, hyperosmolarity, and chronic inflammation. It leads to ocular discomfort due to irritation and visual disturbances.[1] Globally, around 5%–34% of population has been diagnosed with dry eye. Studies in India have shown the prevalence of DED to be 18.4%–54.3% across various geographical domains.[2]

Type 2 diabetes mellitus (T2DM) is a known factor causing DED. The manifestations associated with diabetes, with respect to DED, are due to exocrine dysfunction of lacrimal gland, decreased sensitivity to stimulatory signals, or abnormality in the production of tear proteins.[3] Insulin resistance or deficiency and long-standing hyperglycemia can lead to a reduction in the meibomian gland epithelial cells and goblet cells resulting in a deficient tear film.[4]

T2DM leads to structural anomalies in the corneal nerve fibers leading to decreased sensitivity.[5] A consequence of chronic T2DM is autonomic neuropathy.[3] This causes a reduction in blink rate which leads to excess evaporation of the tear film increasing ocular discomfort.[6]

Since there are multiple mechanisms which can influence the severity of DED, severity of T2DM also needs to be considered in the pathogenesis of DED.

Although there are few studies that have drawn an association between T2DM and DED, very few studies were found that correlated the severity of both DR and DED.

Our study aims to compare the prevalence of DED and to assess the corneal nerve sensitivity in diabetic and non-diabetic patients, and also attempts to correlate the severity of DR and DED. Furthermore, the study also tries to find a relationship between the corneal nerve sensitivity (CNS) and severity of DED.

Methods

The study was a cross-sectional prospective comparative study conducted at a tertiary healthcare center in Southern India. A written and informed consent was taken from all the participants, after explaining about the study. The study adhered to the Declaration of Helsinki and was approved by the Institutional Ethics Committee. Participants were divided into two groups—diabetic group and non-diabetic groups.

Inclusion criteria

1. Patients above 18 years of age,

2. For diabetic group—patients suffering from T2DM, irrespective of their glycemic control and duration of DM.

Exclusion criteria

History of,

1. Topical/systemic ocular medication,

2. Lid and adnexal disorders, corneal pathologies,

3. Any ocular surgery or trauma,

4. Contact lens users,

5. Cigarette smokers,

6. Vitamin A deficienc,

7. Media opacities restricting posterior segment evaluation.

General assessment included demographic data, detailed history of T2DM which consists of duration and treatment, if applicable, and history of other systemic illness. HbA1c levels were noted during the assessment. T2DM patients were classified as normal (HbA1c <5.7%), prediabetes (HbA1c 5.7% - 6.4%), and diabetes (HbA1c >6.5%).[7]

Ocular examination included assessment of visual acuity using Snellen’s chart, anterior segment evaluation using slit-lamp biomicroscope, and posterior segment evaluation by indirect ophthalmoscopy. Grading of DR was done by Early Treatment of Diabetic Retinopathy Study (ETDRS) classification.

Four quadrants of the cornea were tested for sensitivity using cotton wisp.

Subjective assessment of dry eye symptoms was done using the Standard Patient Evaluation of Eye Dryness (SPEED) questionnaire which is a valid and repeatable instrument.[8]

The SPEED questionnaire assessed the following symptoms as shown in Fig. 1.

Figure 1.

SPEED Questionnaire

The maximum score achieved on SPEED questionnaire is 28 points. Scoring was done as follows:

0-4 points: Mild dry eye symptoms,

5-7 points: Moderate dry eye symptoms,

8+ points: Severe dry eye symptoms.

Schirmer’s test

Sterile strips of Whatman No. 41 paper measuring 5 mm by 35 mm were used for the Schirmer’s II test. The patient was asked to wait in a room with reduced ventilation. Proparacaine hydrochloride 0.5% ophthalmic solution was administered to the patient’s eyes. The strips were folded at 90° at the marked area and inserted at the junction of medial two-thirds and lateral one-third of the lower eyelid. The wetting of the strips was recorded at the end of 5 mins.

Interpretation:

>15 mm of wetting: Normal,

10 mm – 15 mm: Mild dry eye,

5 mm – 10 mm: Moderate dry eye,

0 mm – 5 mm: Severe dry eye.

Tear Film Break-up Time (TBUT)

The patient was positioned in front of a slit lamp. One drop of antibiotic solution was applied to a sterile fluorescein strip. The lower palpebral conjunctiva was stained by placing the wet fluorescein end of the strip in the cul-de-sac. The patient was instructed to blink several times to ensure even spreading of the dye on the tear film. The patient was then instructed to look straight without blinking, and tear film was examined using a cobalt blue filter. The tear film break-up time was calculated as the number of seconds that elapsed between the last blink and the first dry spot in the tear film.

Interpretation:

TBUT >10 seconds = Normal tear film stability,

5 – 10 seconds = Marginal tear film stability,

<5 seconds = Low tear film stability.

Grading of DED was done as follows:

• Normal - Schirmer’s II test value >15 mm and TBUT >10 seconds, SPEED value 0 points,

• Mild DED – Schirmer’s II test value between 10 and 15 mm and TBUT of 10 seconds, SPEED value 1–4 points,

• Moderate DED – Schirmer’s II test value of 5 and 10 mm and TBUT of 5 – 10 seconds, SPEED value 5–7 points,

• Severe DED – Schirmer’s II test value of <5 mm of wetting and TBUT of <5 seconds, SPEED value >8 points.

Statistical analysis

Data was analyzed using statistical software R version 4.2.1. and Microsoft Excel. Categorical variables were represented by frequency and percentage. Continuous variables were given in Mean ± SD/Median (Min, Max) form. Normality of variable was checked by Shapiro–Wilk test and QQ plot. Mann–Whitney U-test was used to compare the distributions of variables over groups. Chi-square test was used to check the dependency between categorical variables. P value less than or equal to 0.05 indicates statistical significance.

Results

A total of 400 patients in this study were divided into two groups of 200 each, Group 1—diabetic individuals and Group 2—non-diabetic individuals, based on their diabetic status. The collected sample size was seen in the age group of 30–87 years, the mean age being 55.02 ± 12.21 years. The number of males was marginally more than females, males being 53.5% and females being 46.5%.

Diabetic patients were segregated according to the duration of T2DM into four groups, duration <5 years, 5–10 years, 10–15 years, and >15 years.

In patients with <5 years of T2DM, no form of DED was seen. In patients with 5–10-year duration, 56.5% showed mild DED. Patients, with 10–15 years and >15 years duration, predominantly showed moderate DED with 46.4% and 39.34%, respectively.

Based on the criteria given by Center for Disease Control (CDC), with respect to HbA1c, 20.16%, patients in the normal category showed mild DED. Moderate DED was the predominantly seen in both prediabetes and diabetes group, as 46.34% and 77.35%, respectively.

SPEED scoring was done, and the two study groups were classified as normal, mild DED, moderate DED, and severe DED, based on the SPEED score [Fig. 2].

Figure 2.

Distribution of dry eye based on SPEED score over groups

According to Schirmer’s II test, 29.5% diabetics and 66% non-diabetics had normal values and no DED. Mean value of Schirmer’s II test was 11.61 ± 5.69 mm in diabetic patients and 17.89 ± 7.07 mm in non-diabetic patients. Mild DED was seen in 13.75% diabetics and 16.75% non-diabetics. Moderate DED was indicated in 43.25% diabetics and 13.25% non-diabetics. 13.5% diabetics and 4% non-diabetics showed severe DED.

On the basis of TBUT, low tear film stability was noted in 8.5% diabetics and 3% non-diabetics. Marginal tear film stability was seen in 42% diabetics and 10.5% non-diabetics. However, 49.5% diabetics and 86.25% non-diabetics showed normal tear film stability. Mean TBUT values were 12.43 ± 5.32 sec in diabetic patients and 16.46 ± 4.55 sec in non-diabetic patients.

The final grading of DED was done taking into consideration, the results of SPEED questionnaire evaluation, Schirmer’s II test and TBUT in accordance with DEWS II Report [Fig. 3].

Figure 3.

Distribution of dry eye disease over groups

Fundus examination was done, and 400 eyes of the 200 patients included in the diabetic group were classified using the ETDRS Classification into the following groups. According to the ETDRS Classification, 272 eyes (68%) showed no evidence of diabetic retinopathy. A total of 31 eyes (7.75%) had changes of mild non-proliferative diabetic retinopathy (NPDR), 66 eyes (16.5%) showed changes of moderate NPDR, and 11 eyes (2.75%) showed changes of severe NPDR. Proliferative diabetic retinopathy was seen in 19 eyes (4.75%), and clinically significant macular edema (CSME) or diabetic maculopathy was seen in 1 eye (0.25%).

In diabetic patients, the grades of DR and the severity of DED were taken into consideration to assess if any association existed [Fig. 4].

Figure 4.

Distribution of grade of diabetic retinopathy according to severity of dry eye in diabetic patients who had dry eye disease

Corneal nerve sensitivity (CNS) assessment was done in both the groups—diabetics and non-diabetics, and the following results were noted [Table 1].

Table 1.

Corneal nerve sensitivity over groups

Variables	Sub- category	Groups		Total (n=800)
		Diabetic (n=400)	Non-diabetic (n=400)
Corneal nerve sensitivity	Normal	212 (53%)	388 (97%)	600 (75%)
	Reduced	188 (47%)	12 (3%)	200 (25%)

Association of CNS was done with the severity of DED, and the following was noted [Fig. 5].

Figure 5.

Distribution of corneal nerve sensitivity in patients with dry eye

Discussion

Ocular surface and its disorders have been in discussion in great detail for the past few years. Diabetes mellitus-associated dry eye syndrome (DMDES) has been studied for the past few decades and closely monitored for obtaining a better understanding of the pathological processes intertwining the two.[1] T2DM is a metabolic disorder, and dysregulation of lacrimal functional unit (LFU) may be through various mechanisms. The following initial events may cause alterations in the tear film and ocular surface of diabetic patients: a) chronic hyperglycemia, b) corneal nerve damage, and c) impairment of insulin action.[9]

This study aimed to compare the prevalence of DED and to assess the corneal nerve sensitivity in patients with T2DM and those without the co-morbidity. Furthermore, we also found a correlation between the severity of both these disorders.

The study included 400 patients, in the age group ranging from 30 to 87 years, with the mean age of 55.02 years. The patients were divided into two groups, diabetic and non-diabetic, and the median age in the groups was 56 years in the non-diabetic group and 55.5 years in the diabetic group, and hence was comparable.

The male-to-female ratio was approximately similar, with a P value of 0.6884, indicating that there was no statistically significant difference between the two groups.

Duration of T2DM was taken into consideration and patients suffering for >10 years were significantly affected with Moderate DED. This indicates that duration of T2DM, is an important risk factor for development of ocular surface disorders too, just as seen in other T2DM related complications. Similar findings were noted in a study by Lyu et al.[9]

Higher grades of DED were more commonly noted in patients with a poor glycemic control. A study by Dutta et al.,[10] concluded there is a positive correlation between prevalence of dry eye and poor glycemic control in a patient of diabetes. This was in contradiction with another study done by Pamu et al.,[11] who concluded that there was no clinically significant correlation between the two.

While most other studies have used Ocular Surface Disease Index score (OSDI) as a way to subjectively evaluate symptoms of DED, we used SPEED questionnaire, since it is a reliable, repeatable and has a non-tedious format which sufficiently analyses the complaints of the patients.

On evaluating DED subjectively using SPEED Questionnaire and objectively using Schirmer’s II test and TBUT, most patients with T2DM were noted to suffer from moderate DED. Although the correlation between T2DM and DED has been recognized in previous studies, the fact that moderate DED is more common in T2DM patients has been shown in this study.

Corneal nerve sensitivity, was found to be normal in most of the cases, but a reduction of sensitivity, when noted, was significantly more in patients with T2DM. Also, patients with moderate DED had higher rates of reduced sensitivity, indicating that increased severity of DED in addition to T2DM can be a factor causing neural changes [Fig 5]. In a study by Manchikanti V et al., the corneal nerve morphology was studied and effect of diabetes was significantly noted.[3] Another study by Kheirkhah et al.[12] concluded that there was an improvement in the sub-basal nerve fiber layer in patients with DED post therapy thus proving the damage that can be caused by increased severity of DED.

Severity of DR was also found to significantly associated with severity of DED. Higher degrees of DED were seen in patients with worse forms of DR. This was seen in a similar study done by Khurana et al.,[13] who concluded that prevalence and severity of DED was more in patients with DR.

Our study found that factors which exaggerate the process of DED include the longer duration of T2DM, poorer glycemic index, and severity of DR in addition to autonomic dysfunction caused by T2DM.

Conclusion

In this study we concluded that the prevalence of DED, specifically moderate DED, is much more common in the patients suffering from T2DM than in non-diabetics. Corneal nerve sensitivity was also found to be reduced more so in patients with T2DM than without. Furthermore, among patients suffering from DED, those with moderate DED had greater affected corneal nerve sensitivity. Our study also correlated that severity of DR affects the severity of DED. Diabetes mellitus associated Dry Eye Disease (DMDED) is an upcoming entity and needs to be explored to prevent further visual loss.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.