Diabetes mellitus is a group of metabolic diseases characterized by hyperglycaemia resulting from defects in insulin secretion, insulin action, or both. It affects millions of people all over the world. WHO has labelled India as “The diabetic capital of the world” as it has the highest number of diabetics in the world (1).Diabetes mellitus is associated with a number of ocular complications which can lead even to blindness. Diabetic retinopathy, neovascular glaucoma, cataract, refractive errors, ptosis, cranial nerve palsies, and lid infections are typical ocular complications in diabetic patients. Problems involving the ocular surface, dryness in particular, have been reported.(2–7) Diabetic patients suffer from a variety of corneal complications, including superficial punctuate keratopathy, corneal ulceration, and persistent epithelial defects. Dry eyes are an important contributor to these ocular surface related problems. The mechanism responsible for dry eye is thought to be autonomic dysfunction and neuropathy related to diabetes.(8) Some studies have found that diminished corneal sensation in diabetes is strongly related to peripheral neuropathy.(9–11)

Various earlier studies have reported qualitative and quantitative tear film abnormalities in diabetics. The prevalence of dry eyes found in diabetics varies over a wide range in various studies from as less as 18.1% (2) to 70%(12). So, there is lack of a general consensus on the exact prevalence of dry eye in diabetes. Moreover, ocular surface examination is ignored and much importance is given to retinopathy.

Although several studies have documented tear secretion and tear film dysfunction in diabetes mellitus, few studies have been conducted in the population of Rajasthan. In addition, there are few studies that report the association between dry eye and corneal sensation and the correlation that these two entities have with diabetic retinopathy and the available literature shows conflicting results.

The present study was undertaken to evaluate the amount of tear production, the stability of the tear film and the condition of the ocular surface in diabetic individuals in order to detect possible tear film anomalies. This study was aimed at finding prevalence of dry eye in patients with and without diabetes in our hospital. Status of corneal sensitivity was also assessed in these two groups and prevalence of diminished corneal sensitivity was determined. Furthermore, an attempt was made to find out relationship between severity of dry eyes and corneal sensitivity with that of diabetic retinopathy.

Objectives:

• To study the prevalence of dry eye in the general population and in patients with diabetes mellitus.
• To assess the status of corneal sensitivity in the general population and in patients with diabetes mellitus.
• To find correlation of dry eye and corneal sensitivity with diabetic retinopathy

A hospital based cross-sectional study was undertaken from November 2022 to January 2024. The individuals enrolled for this study were all patients who are above age of 40 years and who attended the ophthalmology outpatient department.

Inclusion criteria: All patients of either sex who are above the age of 40 years and who did not have any exclusion criteria.

Exclusion criteria:

1. Patients with systemic disease and local ocular disease/ surface abnormalities, as assessed by history and clinical examination, other than diabetes mellitus. Which are known to cause dry eye or ocular surface abnormalities.
2. Contact lens wearers.
3. Patients who have undergone ocular surgeries in the past.
4. Patients on local or systemic medications, which are known to cause dry eyes/ocular surface disorders.

A detailed history of patient regarding name, age, sex, occupation, address, drug history, presenting symptom, duration, progression, associated conditions and past history was recorded.

Details of history regarding diabetes such as symptoms, duration, and type of treatment was enquired. Fasting blood sugar values and post prandial blood sugar values were recorded

Questionnaire:

An eight-item questionnaire of ocular symptoms relating to dry eye was used which included the following questions:

1. Do your eyes ever feel dry?
2. Do you ever feel a gritty or sandy sensation in your eye?
3. Do your eyes ever have a burning sensation?
4. Are your eyes ever red
5. Do your eyes ever feel sticky?
6. Do your eyes ever feel watery or tearing?
7. Do you notice much crusting on your lashes? and
8. Do your eyes ever get stuck shut?

Presence of one more symptom often or all the time was taken as positive.

Examination:

Ocular examination included recording visual acuity with snellen’s chart. Detailed anterior segment examination was done under slit lamp. Condition of lids, meibomian glands, conjunctival surface (dryness, wrinkling, sheen) and corneal surface was noted. Cornea was evaluated in detail for its sheen, surface (superficial punctate keratitis SPK/mucous plaques/filamentary keratitis). Sensation was recorded after Schirmers I test with a Cochet-Bonnet esthesiometer and graded as normal (5mm to 6mm) or reduced (<5mm).

Tear film evaluation was done in the following order-Tear meniscus height was recorded as normal or low (under slit lamp, thin beam) Precorneal tear film was observed for presence of debris (mucous/oil droplets/debris).

Tear break up time measurement:

No anesthesia was used. A dry fluorescein strip was touched to the inferior fornix with the patient looking up. The cornea was scanned under low slit lamp magnification using a blue cobalt filtered light. The patient was instructed to blink once or twice and then stare straight ahead without blinking. The time of appearance of the first dry spot formation (small black spots within the blue-green field) from the last blink was taken as the TBUT value.

Fluorescein staining of cornea was graded from 0-3.

0-no staining of corneal epithelial surface.

1 – Mild staining occupying < 1/3 of corneal epithelial surface. (Figure 1)

2 – Moderate staining occupying < ½ of corneal epithelial surface. (Figure 2)

3 – Severe staining of > ½ of the corneal epithelial surface. (Figure 3)

Schirmers test (basal and reflex tearing):

It was performed by placing a precut strip of filter paper in the inferior cul-de-sac; Patient was asked to blink normally, and the amount of wetting of the paper strip after 5 minutes was measured. Wetting of ≤10 mm was taken as abnormal. (Figure 4)

The basal secretion test was performed following the instillation of topical anesthetic (4% xylocaine drops) and the placement of a thin strip of filter paper in the inferior cul-de-sac. Measurement of less than 5 mm was taken as abnormal, 5-10 mm was considered equivocal.

Rose Bengal staining: (Figure 5)

A moistened strip of rose Bengal was applied to the inferior cul-de-sac, under without anesthesia. Van Bijsterveld scoring system was used to grade the staining of cornea and conjunctiva, based on a scale of 0-3 in 3 areas: nasal conjunctiva, temporal conjunctiva, and cornea. With this system, the maximum possible score is 9, a score of 3.5 or

Greater was considered positive for dry eye.

Dry eye was defined as having one more symptoms( often or all the time present) along with one or more positive clinical findings (based on slit lamp examination) and one or more positive clinical tests(tear break up time of ≤10 seconds, Schirmer test score ≤ 10mm, with anesthesia ≤5mm, fluorescein score of ≥1, rose bengal stain score of ≥3.5. Asymptomatic patients with positive signs or positive tests were also considered in the diagnosis. Dry eye was graded into three types-mild, moderate, and severe.

Patients were classified as mild dry eye when there was Schirmer test of less than 10 mm in 5 minutes, T BUT less than 10 seconds and less than one quadrant of staining of the cornea, moderate dry eye when there was a Schirmer test of 5 to 10 mm in 5 minutes, T BUT of 5 to 10 seconds with punctuate staining of more than one quadrant of the corneal epithelium and severe dry when there was diffuse punctuate or confluent staining (with fluorescein and rose Bengal) of the corneal epithelium, often with filaments and diffuse punctae or confluent staining of the conjunctival epithelium. The Schirmer values in these patients is less than 5 mm in 5 minutes and T BUT less than 5 seconds. Intraocular pressure- was recorded by Goldmann Applanation tonometry.

Detailed fundus examination:

 A dilated fundus examination was done by direct ophthalmoscopy and with a Goldmann 3 mirror fundus contact lens. Retinopathy if present was classified according to ETDRS guidelines

Out of a total of 100 participants 57 were males and 43 were females. The ages of the participants ranged from 40 years to 72 years with mean age of 55.02 ± 9.33 years. Thirty eight patients belonged to age range between 40 years to 50 years. 30 patients were in the age range between 51 years and 60 years. 30 patients were in the range 61 years to 70 years and 2 patients were in the range 71 to 80 years. (Figure 6)

The total number of male diabetics who had dry eyes was 16 (43.2%). The total number of female diabetics who had dry eyes was 13 (48.1%).

Out of total number of non-diabetics 2 (5.5%) had dry eyes, out of which 1 (50%) had mild dry eye and another 1 (50%) patient had moderate dry eyes. There was no non diabetic patient with severe dry eye. None of the non-diabetic male had dry eyes while only 2 female non diabetic (12.5%) had dry eyes.

Out of total diabetic patients 15 (23.4%) had diminished corneal sensitivity. (Odds ratio= 0.19) All patients with diminished corneal sensitivity had dry eyes. Out of diabetics with diminished corneal sensation 11 (73.3%) of patients had moderate dry eye and 4 (26.7%) had severe dry eyes (p<0.0001). (Figure 13)

In the present study a significant correlation was found both with the presence of dry eye and reduced corneal sensitivity in patients with type 2 diabetes mellitus. Neither of these two parameters was found to be correlated with diabetic retinopathy. However, these factors were found to be significantly interrelated. Age and duration of diabetes were also found to correlate with dry eye and reduced corneal sensitivity.

The proportion of patients with dry eye in diabetics was found to be significantly higher (45.3%) compared to that of non-diabetics (5.5%) (p < 0.0001). This finding correlated with various other  studies.(2–4,6–8,10,12,16,28,29) . However the prevalence shows a wide variation in these studies from 18.1% (2) to 70%.(12) This large degree of variation may be due to different diagnostic criteria, different study population, different questionnaires and diagnostic tests. Li and associates(28) compared the status of dry eye both in diabetics and in age and sex matched controls. They found that 19.8% of diabetics had dry eyes compared to 8% of controls. The Beaver dam study cohort(2) found that diabetes was “independently and significantly associated with dry eye” with an odds ratio of 1.38. The prevalence of dry eye in diabetics was found to be 18.1%. This slightly lower prevalence could possibly be attributed to the fact that the diagnosis of dry eye was based on subjective complaints alone and not by objective tests, so some cases of dry eye might have been missed. Kaiserman et al(6) found  that about 20.6% of diabetics use artificial lubricants to get relief from dry eye. In this study also no objective tests to assess dry eyes were done and only a review of electronic records of patients was done to come to this conclusion. Although the percentage of diabetic patients having dry eye was low in these studies, it was significantly higher compared to that of non diabetics.

Compared to the aforementioned studies some have found a higher percentage of diabetics who have dry eye. Seifart and associates(12) compared the dry eye status in type 1 and type 2 diabetics. They found prevalence of dry eye in type 2 diabetics as high as 70% while in patients with type 1 diabetes it was 57%. Manaviat et al(8) found that 54.3% of 199 diabetic subjects suffered from dry eyes. Dry eye status was not compared with controls in this study.

Several other studies also discuss the close relationship of dry eye and diabetes. A study published by Wang et al(5) found that there was greater association of dry eye in diabetes with complications (such as neuropathy and nephropathy). Dogru and associates(7) found that TBUT and Schirmer’s tests values were significantly lower in the diabetic group than non diabetics. In their analysis, the conjunctiva in diabetics showed goblet cell loss and conjunctival squamous metaplasia, both of which are related to dry eyes. These findings were similar to those seen by Yoon et al(4) who assessed keratoepitheliopathy, TBUT, Schirmer value, and conjunctival impression cytology in diabetics and non diabetics and came to the conclusion that diabetic patients with poor metabolic control, neuropathy, and advanced stages of retinopathy should be examined for tear film and ocular surface abnormalities.

Saito and associates(10) found that both total and reflex tear secretion were reduced in diabetics as compared to non diabetics. Inoue and associates(16) found a greater degree of nonuniformity of tear lipid layer, corneal sensitivity , and tear breakup time in diabetics compared to patients without diabetes and attributed the keratoepitheliopathy seen in diabetes to these factors. They also found that Schirmers test and rose bengal staining had more diagnostic value in detecting dry eye than TBUT. This finding was also shared by Han et al(30) who found a positive correlation between Schirmers test and symptom based diagnosis of dry eye. Schirmers was found to have a high specificity but low sensitivity in this study.

In patients without diabetes the dry eye prevalence was very low compared to their diabetic counterparts in the present study. Seifart and associates(12) found a prevalence of dry eye symptoms in 9.3% of non diabetic patients while objective dry eye tests were abnormal in 5.8% of non diabetics. Moss et al(2) found a prevalence of 14.1% but only subjective complaints of dry eye were taken into consideration. Li and associates(28) found a prevalence of dry eye of 8% in the non diabetic population.

In our study we found that most patients with mild and moderate dry eye had no diabetic retinopathy (p<0.0001), while no such correlation could be made with respect to severe dry eye due to a small number of patients in this group. Several other studies have found a correlation between dry eye and diabetic retinopathy.(8,19,31) Li and associates(28) found a significantly worse dry eye score with increasing severity of retinopathy. Patients with retinopathy were also divided based on whether they had undergone photocoagulation or note. All the dry eye parameters were significantly worse in patients who had undergone photocoagulation. 

There are some studies, however that found no correlation between dry eye and diabetic retinopathy.(4,7,10,29) Dogru et al(7) found that TBUT, Schirmers test and impression cytology to be abnormal in diabetics more frequently, but their findings did not correlate with presence and severity of diabetic retinopathy. Yoon and associates(4) investigated keratoepitheliopathy, tear film break-up time, total tear secretion and basal tear secretion in diabetics along with other parameters. They found that there were no significant differences of parameters between the no DR and the NPDR groups, except for total tear secretion. The mean total tear secretion found by Schirmer’s test was 15.57±4.99 mm and 13.04 ± 3.93mm in no DR and NPDR groups respectively. However, there were statistically significant differences of parameters between the no DR and the PDR groups, and between the NPDR and the PDR groups. Because of this increased risk for patients with advanced diabetic retinopathy, they advocate the examination of patients with advanced diabetic retinopathy for tear film and ocular surface changes.

Corneal sensitivity in diabetics is closely related to the development of peripheral neuropathy.(7) We found that corneal sensitivity was significantly reduced in diabetic patients when compared to normal patients (OR: 0.19). Several other studies have also obtained similar findings.(3,4,7,9,10,21,22) A study by Dogru et al(7) in diabetic patients found that the mean corneal sensitivity was significantly lower in diabetic patients, diabetic patients with peripheral neuropathy, and poorly controlled diabetics compared with non diabetic control subjects. Murphy et al(21) compared corneal sensitivity in diabetic and non diabetics in young, middle aged and older group of patients. In their study no significant difference was found between the diabetic and nondiabetic subjects when the young and middle groups were compared, but a significant difference was noted between the older groups. In a study by Zalentein(22) et al corneal sensitivity to different modalities of stimulus was determined in one randomized eye in 52 patients with diabetes and in 27 healthy subjects. They assessed corneal sensitivity for mechanical, chemical and cold stimuli and found that sensitivity thresholds to selective mechanical, chemical and cold stimulation were significantly higher in diabetic patients compared to controls.

In the present study it was found that diminished corneal sensitivity correlated with the presence and severity of diabetic retinopathy (p=0.005). Similar results were obtained in other studies.(10,11,17) Gerald et al(17) found that eyes with normal fundi had normal sensitivity. Those with background retinopathy had a mildly decreased sensitivity, while those with proliferative retinopathy showed a greater loss of corneal sensation. Eyes that had received treatment with panretinal photocoagulation demonstrated the most reduction in corneal sensitivity. They thus suggested that corneal sensitivity could be used as a screening tool for proliferative diabetic retinopathy. Saito et al(10) also found a significant correlation with increased severity of retinopathy. In our study, all the 4 patients with PDR had normal corneal sensation. Tavakoli et al(9) assessed the neuropathy deficit score in 147 diabetics and 18 age matched controls using both a Cochet-Bonnet aesthesiometer and a non- contact corneal aesthesiometer. With the Cochet Bonnet esthesiometer they found diminished sensitivity in both diabetics and non diabetics but the prevalene was much higher in those with diabetic retinopathy. With the non-contact corneal esthesiometer, however, only patients with moderate and severe retinopathy showed evidence of diminished corneal sensitivity. It appears that the Cochet bonnet aesthesiometer may be more sensitive in picking up mild cases of diminished corneal sensation.

However some studies have found no positive correlation between diminished corneal sensation and severity of diabetic retinopathy.(4,7,22,32)

Diminished corneal sensation may lead to dry eye by reducing reflex tear secretion and by producing corneal surface irregularities. In the present study we found that there was a significant correlation between diminished corneal sensitivity and the presence of dry eye (p<0.0001). All the patients with normal corneal sensitivity had mild or no dry eye while patients with diminished corneal sensitivity had moderate to severe dry. This shows that the presence of diminished corneal sensation significantly influenced the severity of dry eye. This was also found by some of the other studies.(23–25,26,27) Bourcier(27) found that corneas of patients with dry eye were less sensitive and this significantly correlated with the intensity of fluorescein and Lissamine green corneal staining but not with the results of the Schirmer test. In a study by Jose et al, twenty-one patients with dry eye and 20 healthy volunteers were studied for corneal innervations using in vivo confocal corneal microscopy, and corneal sensitivity.(26) A statistically significant decrease in the number and density of subbasal nerves and the density of superficial epithelial cells was observed in dry eyes. A significant decrease was found in dry eyes with respect to mechanical, chemical, and thermal sensitivity of cornea compared to normal patients. In another study by the same author(24) a comparative study between healthy volunteers and patients with dry eyes showed that the corneas of the patients with dry eye had more bead like formations. They felt that this could be the cause as well as result of reduced corneal sensivity in dry eye.

Age was seen to be significantly associated with dry eyes in the present study (p<0.0001). The mean age of patients without dry eyes was 50.4±7.49 years and for those with dry eyes were 62.9±5.47 years. We found that the prevalence of dry eye in the older age group (>60 years) was 3 times more than in patients with age less than 60 years. Similar results were obtained by several authors.(14,15,30,33) In the Beaver Dam Eye study, Moss et al found a positive correlation between dry eye and age only up to 70 years of age.(33) However Sahai and associates found that dry eye prevalence was maximum in those above 70 years of age. They also found that after older patients (>70 years), patients in the age group of 31-40 years also showed a significantly high prevalence of dry eye.(14) Similar finding was shared by Li and associates who found that the 40–49 year age group reported most dry eye symptoms, although a significant increase in dry eye symptoms was found with increasing age. However, only subjective dry eye symptoms were considered in their study.(28) We did not come across such finding of a higher incidence in the younger age group in our study.

In our study duration of diabetes was found to be significantly associated with the presence and severity of dry eye (p<0.0001). The mean duration of diabetes in patients without dry eye was 25.37±12.18 months and that for patients with dry eyes was 76.13±20.60 months. Similarly patients with a longer duration of diabetes were seen to have more severe dry eye. Similar results were obtained in some other studies.(8,18) Manaviat and associates found that out of 108 patients with dry eye syndrome the mean duration of diabetes was 11.48 ± 7.4 years whereas this was 9 ± 6.5years in subjects without dry eye syndrome.(8) Some of the other studies found no correlation between duration of diabetes and dry eyes.(4,7,28,29)

In the present study there was a difference between the frequency of dry eye in males (43.24%) and females (48.14%) but it was not statistically significant (p=0.5127). Similar results were found in several other studies.(4,6,8,34) However some studies found that there was a statistically significant difference between genders.(14,15,56,30,34,35) All these studies found that dry eyes were more common in females as compared to males.

Corneal sensitivity was found to be significantly associated with age in the present study. (p<0.0001). The mean age of patients with diminished corneal sensation was 67.06±3.55 years and for patients with normal corneal sensation it was 52.69±7.46 years. A similar correlation with age was seen in many other studies.(9,13,24,26,27) Millodot(13) found no association of age with corneal sensitivity upto 40 years of age; however corneal sensitivity diminished significantly after that. Tavakoli et al(9) found that age correlated with sensation tested with the Cochet-Bonnet aesthesiometer but not with the non-contact corneal aesthesiometer. Jose et al found that the number of sub-basal fibers in the cornea decrease with age.(24) Murphy et al(21) found that there is a gradual reduction in corneal sensitivity with increasing age in both nondiabetic subjects and diabetic subjects

.

However some studies found that age does not correlate with corneal sensitivity.(11,22) Zalentein et al found that sensation thresholds to mechanical, chemical, heat and cold stimulation became higher proportionally with age in non diabetic subjects while in contrast, sensation thresholds for the different stimulus modalities in diabetic patients were independent of age.(22)

Corneal sensitivity was found to have a significant association with duration of diabetes in our study (p<0.0001). The mean duration of diabetes in patients with diminished corneal sensitivity was 89.2±19.22 months and that for patients with normal corneal sensation was 35.87±20.40 months. Similar findings were obtained in some other studies.(9,20)

However several other studies found no correlation between duration of diabetes and reduced corneal sensitivity.(4,7,17,21,22)

To conclude, dry eye and reduced corneal sensitivity was significantly associated with diabetes mellitus which is supported by other published studies. Neither of these were found to correlate with the degree of diabetic retinopathy. The published literature gives conflicting reports regarding this association. Sample size and selection bias may be a factor. Nevertheless there is enough evidence to suggest a thorough ocular evaluation of all diabetics that should include evaluation for evidence of tear dysfunction and corneal sensitivity.

The significant correlation between dry eye and corneal sensitivity in patients with type 2 diabetes mellitus found in this study correlates with the findings of other studies in the literature. Since these conditions are a potential cause for significant visual morbidity, both individually and even more so combined, evaluation of all diabetic patients for dry eye and diminished corneal sensation would help to identify these potential risk factors. Neither of these conditions can be reversed but their harmful effects can be obviated by timely and appropriate treatment. Both of these conditions are simple to detect in an ophthalmology outpatient department. So assessment of dry eye and corneal sensitivity should be an integral part of ophthalmic evaluation in patient with diabetes mellitus.