Diabetic eye disease in southern India
It is still a popular misconception that diabetes mellitus is a disease which affects the affluent and has resulted from an overindulgent lifestyle. Recent studies have shown that diabetes is not a disease restricted to the industrial nations. In fact its prevalence appears to be highest among the populations of developing countries, and the migrant and minority communities in industrialised countries.1
The preliminary stage in implementing training for health workers in screening for diabetic eye disease is to appreciate the magnitude of the problem. This study proposed to investigate the extent of the problem in southern India by focusing on one specialist centre, the Aravind Eye Hospital in Madurai, Tamil Nadu.
The medical records of all new patients presenting to the Aravind Eye Hospital Out-Patients Department (Paying Section) between 1st January 1993 and 30th June 1993 were obtained. The medical records of those patients who reported having diabetes mellitus or who were found to be diabetic at presentation by urine test/blood test were selected. The following information was recorded for each of these patients.
- Age and Sex
- Duration of diabetes (if known)
- Treatment history (if any)
- Presenting complaint
- Any referral letter
- Visual acuity for both eyes
- Schiotz tonometer readings
- Systolic blood pressure
- Diastolic blood pressure
- Ocular findings on examination
Diabetic retinopathy was classified thus:
(a) Background retinopathy: the presence of hard exudates or dot and blot haemorrhages
(b) Pre-proliferative retinopathy: the presence of cotton-wool spots (micro-infarcts), venous irregularities (beading, reduplication, loops), multiple haemorrhages and intraretinal micro-vascular abnormalities.
(c) Proliferative diabetic retinopathy: the presence of new vessels on the disc or elsewhere in the retina.
(d) Maculopathy: the presence of hard exudates within one disc diameter of the macula, with or without concurrent macular oedema.
The data was analysed on Apple Macintosh™ computer by the nonparametric techniques available on the Stats Works software™ package.
Over the specified six month period, 47,499 new patients were seen in the OPD. 1,863 new patients had a diagnosis of diabetes mellitus. Of these, 194 patients (10%) were found to have no abnormality on ocular examination. 987 patients (53%) had cataracts in one or both eyes. 684 patients (37%) were diagnosed as having diabetic retinopathy, of whom 388 patients had maculopathy. 106 patients (6%) were found to have glaucoma (Table 1). In addition, four patients had pterygia, seven had corneal ulcers, two had dacryocystitis, twelve had nerve palsies, two had retinal vein occlusion, and one had retinitis pigmentosa.
Age and sex
The mean age of the patients was 55.4 years (Table 2). 61.6% of the patients were male and 38.4% were female.
Almost all the cataract patients complained of defective vision. Of the remaining 876 patients, the following reasons were obtained for the patient coming to hospital:
- Defective vision: 84%
- Irritation/itching/watering: 5%
- Pain/headache: 4%
- ‘Check-up’: 7%
Only 27% of the patients’ records included letters of referral from other health professionals. Referrals were slightly more common for patients with diabetic retinopathy than for those with no retinopathy, but in both cases tended to be in the late stages of disease, e.g., advanced proliferative retinopathy.
9% of the patients had visual acuity equivalent to counting fingers or worse (i.e., detecting hand movements, perception of light only, or no perception of light) in one eye or both eyes. This figure was higher for the patients with cataract. Only 31% of the patients had visual acuity of 6/18 or better in both eyes (with or without spectacles).
Anterior segment findings
987 patients were found to have cataracts on slit lamp examination. Of these, 719 were classified as immature cataracts and 268 as mature cataracts. The mean age of these patients was 60.5 years (range 35-105). In the majority of these patients, fundoscopy could not be performed adequately. This may represent a considerable amount of diabetic retinopathy which was not detected.
106 patients were found to have glaucoma by tonometry/fundoscopy/visual field analysis. Of these, only 65 patients (61%) had Schiotz tonometer readings above 20mmHg.
Posterior segment findings
684 patients (37%) were diagnosed as having diabetic retinopathy. 416 patients (22%) had background retinopathy, 81 patients (4%) had pre-proliferative retinopathy, and 187 patients (10%) had proliferative retinopathy. In addition, 388 of these patients (21%) had maculopathy (Table 1).
The differences in the mean duration of diabetes between the patients with no diabetic retinopathy and those with retinopathy were significant at the p<0.05 level.
Hypertension (taken as a blood pressure >140/90 according to WHO definition) was found in 283 patients. Hypertensive changes were detected in 52 patients. Those with diabetic retinopathy were twice as likely to have hypertension/ hypertensive retinopathy. A significant association was found between diastolic blood pressure and maculopathy (p<0.05 level).
Table 1: Diabetes Mellitus: 1,863 New Patients: Main Associated Eye Disease
Anterior Segment(one or both eyes)
|Eye Disease||Number (Percentage)|
Posterior Segment (one or both eyes)
|Eye Disease||Number (Percentage)|
|Diabetic retinopathy||684 (37%)|
|Background retinopathy||416 (22%)|
|Pre-proliferative retinopathy||81 (4%)|
|Proliferative retinopathy||187 (10%)|
|Maculopathy (+ other retinopathy)||388 (21%)|
This retrospective study was limited to data available in the medical records. The data may also be biased, as only information from paying patients (i.e., those who were relatively affluent) was available. Data on smoking habits, the level of diabetic control, and cholesterol levels, which are risk factors for diabetic retinopathy were not available. However, its strength lies in the large number of patients reviewed.
Although a study of hospital out-patients is a survey of a selected group, a few comparisons can be made with community-based studies and some conclusions can be reached from the results already outlined.
As expected, the number of male patients exceeded the number of female patients, and the number of cases of background retinopathy far exceeded the number of cases of proliferative retinopathy. In comparison with the types of retinopathy found in population-based studies done in England,2 Denmark,3 and the United States,4 there were two notable differences. Firstly, the proportion of proliferative diabetic retinopathy was high. Secondly, the number of cases of diabetic retinopathy with maculopathy was exceedingly high.
That the severity of retinopathy increases with duration of diabetes is well established,3 and the results of this study are in accordance with this finding (Table 2). Although elevated diastolic blood pressure has been related to diabetic retinopathy in other studies (e.g., Klein et al. 1984),3 this study also found it to be significantly associated with maculopathy. Regardless of such differences, it can be suggested that good hypertensive control is a worthwhile aim in the prevention of diabetic eye disease.
One of the notable differences between the findings of this study and those studies of diabetes in developed countries is the extremely high incidence of cataracts amongst patients presenting to the Aravind Eye Hospital. In this selected population, this finding can be attributed partially to the high incidence of cataract in southern India (Indian Council for Medical Research, 1991), and partially to this hospital’s established national and international reputation for cataract surgery.
Although the large number of patients with advanced cataracts does account for the finding of generally low visual acuity to some extent, it does not explain it entirely. No significant difference was found in the visual acuities of those patients presenting with cataracts and those presenting with other pathologies. The vast majority of patients (84%) present because of defective vision, and most patients appear to be presenting only when their visual acuity is so poor that it prevents them from functioning in daily life. It, therefore, becomes alarmingly apparent that such late presentation, for example in the case of diabetic retinopathy, is often too late. The importance of early detection by trained health workers is thus emphasised.
Table 2. Diabetes mellitus: 1863 new patients: retinopathy, age, duration
|Retinopathy||Mean age (range)||Mean duration of diabetes (range)|
|None||58.3 (19 to 85)||4.8 years|
|Background||57.9 (25 to 90)||9.4 years|
|Pre-proliferative||55.5 (36 to 81)||10.4 years|
|Proliferative||54.0 (30 to 87)||12.8 years|
Diabetic eye disease is becoming an increasing problem in developing countries, due to longer life expectancy and a higher incidence of diabetes. In this study, an average of 45 patients/month presented with advanced diabetic retinopathy. Only a few of these patients were referred for examination, suggesting that there is a need to improve awareness amongst the community and health care professionals of the risk of this serious, and to a large extent treatable, cause of blindness.
My thanks to Dr G Venkataswamy, Director, Aravind Eye Hospital, and his colleagues, for allowing access to medical records, and for their advice, support and encouragement throughout.
1 King H, Rewers M. Diabetes in adults is now a third world problem.Bull WHO 1991; 69(6): 643-8.
2 McLeod B, Thompson J, Rosenthal A. The prevalence of retinopathy in the insulin-requiring diabetic patients of an English country town.Eye 1988; 2: 424-30.
3 Nilsson S, Nilson J, et al. The Kristianstad survey II.Acta Med Scand 1967; Suppl 469: 1-42.
4 Klein R, Klein B. The Wisconsin epidemiological study of diabetic retinopathy.Arch Ophthalmol 1984; 102: 520-6.
5 Burditt A, Caird F. The natural history of diabetic retinopathy.Quart J Med 1968; 37: 303-317.