Incidence, Progression, and Risk Factors for Cataract in Type 2 Diabetes

Abstract

Purpose: We report the 4-year incidence, progression, and risk factors of cataract subtypes in type 2 diabetes.

Methods: A total of 779 subjects completed baseline and 4-year follow-up.

Results: The incidences of nuclear opalescence (NO), nuclear color (NC), cortical cataract (CC), and posterior subcapsular cataract (PSC) were 70%, 55.2%, 25.7%, and 7.3%, respectively. One-step progressions of NO, NC, CC, and PSC were 14.3%, 16.1%, 8.8%, and 8.1%, respectively, and two-step or more progressions were 5.0%, 6.0%, 0.8%, and 6.0%, respectively. Incident NO was seen in patients 50 to 59 (odds ratio [OR] = 3.3), NC in those 50 to 59 (OR = 2.7) and 60 to 69 (OR = 3.9), and CC in those 60 to 69 (OR = 3.3) years old. A lower hemoglobin A_1c (HbA_1c; OR = 0.7), longer diabetes duration (OR = 1.1), and hyperopia (OR = 4.0) were associated with incident PSC. Women (OR = 1.7) and patients with higher total cholesterol (OR = 1.3) at baseline showed one-step NO progression. Patients 60 to 69 (OR = 2.8) and ≥70 (OR = 3.8) years old showed one-step NC progression, while those 60 to 69 years old showed one-step CC progression (OR = 6.3). A lower HbA_1c (OR = 0.3) was associated with one-step PSC progression. A higher low-density lipoprotein (OR = 1.6) was associated with two-step or more NO progression. Patients 60 to 69 years old (OR = 6.7) had a greater risk, while those with hyperopia at baseline (OR = 0.2) had lower risk of two-step or more NC progression. Patients 40 to 49 years old constituted the reference group for all.

Conclusions: The 4-year cumulative incidence of cataract is higher than that of progression. Greater age is a risk factor for incidence and progression of most types of cataract.

Cataract is a major cause of vision impairment in the general population worldwide¹ as well as in people with diabetes.² A large body of data from cross-sectional and longitudinal studies report an association between diabetes and cataract. The Beaver Dam Eye Study (BDES), Blue Mountains Eye Study (BMES), and Visual Impairment Project have documented an association between diabetes and incidence of posterior subcapsular cataract (PSC) and, to a lesser extent, incidence of cortical cataract (CC).^3–13 In our previously reported population-based cross-sectional study,¹⁴ higher age, macroalbuminuria, and higher glycated hemoglobin were reported as risk factors for the prevalence of any cataract. We observed that higher serum triglycerides are associated with nuclear cataract, whereas female sex and poor glycemic control are associated with PSC and nuclear cataract, respectively. However, to our knowledge there is no longitudinal study to date in the Indian population that reports incidence, progression, and risk factors for cataract in subjects with diabetes. Therefore, in the current population-based study, we examined the role of age and several clinical factors and their association with the 4-year incidence and progression of individual cataract types (subtypes) in subjects with type 2 diabetes. 

Methods

The Sankara Nethralaya Diabetic Retinopathy Epidemiology and Molecular Genetic Study (SN-DREAMS II) was a cohort study in which participants of SN-DREAMS I (baseline) were reexamined after 4 years. The study design and research methodology of SN-DREAMS I have been described in detail elsewhere.¹⁵ The Figure shows a schematic of the baseline and follow-up status of the subjects. Of 5999 subjects aged ≥40 years who underwent field enumeration, 1563 had or were suspected to have diabetes. After excluding 138 nondiabetic subjects and 11 with nongradable retinal images, 1414 subjects were examined at baseline. 

Figure

 

Flowchart shows the details of subject recruitment and follow-up.

At follow-up, after excluding 101 subjects who had died, 408 nonparticipants, and 126 who underwent surgery for bilateral cataract or who had missing data, a total of 779 subjects were analyzed for incidence and progression of cataract. Of the 125 excluded subjects, 58 already had undergone cataract surgery in both eyes during initial recruitment in the study at baseline. At the follow-up visit, another 67 patients had undergone cataract surgery in both eyes. The study was approved by the institutional review board and was conducted according to the Declaration of Helsinki. Informed consent was obtained from the subjects after an explanation of the nature and possible consequences of the study. 

A detailed interview was performed at the base hospital by trained bilingual interviewers. All instruments were developed initially in English and later translated into Tamil (regional spoken language), ensuring that the contents and meanings were preserved. The data collected in the medical history included ocular- and systemic-related conditions, family history of diabetes, current treatment of diabetes, smoking, and alcohol intake. Subjects underwent general physical examination that included measurement of height and weight, with which the body mass index (BMI) was calculated using the formula, weight (kg)/height (m²). The blood pressure (BP) was recorded on the right arm with the patient in the sitting position; two readings were taken 5 minutes apart and the mean of the two was considered the BP. Socioeconomic status (SES) was assessed with a multiple-index questionnaire, and the scoring was characterized as low (score, 0–14), middle (15–28), and high (29–42). Biochemical analysis was done on the Merck (Rockville, MD, USA) Micro Lab 120 semi-automated analyzer. Total serum cholesterol (CHOD-POD method), high-density lipoproteins (HDL; CHOD-POD method after protein precipitation), serum triglycerides (CHOD-POD method) and the glycosylated hemoglobin fraction (Bio-Rad DiaSTAT hemoglobin A_1c [HbA_1c] Reagent Kit; Bio-Rad Laboratories, Hercules, CA, USA) were examined. 

Visual acuity was estimated using the modified Early Treatment Diabetic Retinopathy Study chart (Lighthouse Low Vision Products, New York, NY, USA). Objective refraction (Beta 200; Streak Retinoscope, Heine, Germany) always was followed by a subjective refraction. The following definitions were used: emmetropia was defined as spherical equivalent (SE): −0.50 to +0.50 diopter sphere (DS), myopia as SE < −0.50 DS, hyperopia as SE > +0.50 DS, and astigmatism as cylindrical error < −0.50 diopter cylinder at any axis. In our previous cross-sectional study¹⁶ that reported the prevalence of refractive errors and risk factors at baseline in people with diabetes, we found that hyperopia was more prevalent among women with diabetes. Therefore, for the current longitudinal study, we included refractive error (myopia, hyperopia) in comparison with emmetropia. 

All subjects underwent detailed ophthalmic examination, which included cataract grading. The lens opacity was graded according to the Lens Opacity Classification System version III (LOCS III; LOCS chart III; Leo T Chylack, Harvard Medical School, Boston, MA, USA).¹⁷ 

Grading of Cataract

Pupillary dilatation was performed with tropicamide (1%) and phenylephrine hydrochloride (2.5%) drops (instilled twice, if necessary); the subject's eyes were examined with a slit-lamp (SL-120; Carl Zeiss Meditec, Jena, Germany). Comparing each eye with the LOCS III standard photographs (mounted close to the slit-lamp), the examiner identified the specific lens opacity and assigned a severity grade. 

Lenticular opacities were graded by comparison with the standard set of photographs, which were retroilluminated by mounting on a lightbox. With the slit beam at an angle of 45°, the slit height and width were adjusted to approximate the overall brightness of the corneal image and anterior subcapsular zone to nuclear color/opalescence standard N1; 0° brightness was used to visualize all the lens opacities without causing discomfort to the patient. Keeping the slit width at 0.2 mm, nuclear opalescence (NO) and nuclear color (NC) were graded by comparing the slit-lamp examination image with the nuclear standards NO1 to NO6 and NC1 to NC6. CC and PSC were graded by examining the opacity in retroillumination images (0° angle) focused either anteriorly (at the iris plane) or posteriorly (at the plane of the posterior capsule) and comparing these examination pictures to the cortical standards C1 to C5 and posterior subcapsular standards P1 to P5. 

The severity of the lens opacities, according to the photographic standards, then was stratified into four major groups and analyzed: NO, NC, CC, and PSC. Intergrader agreement was determined by having both graders assess the eyes of 50 patients, recruited from the pilot study, who had various grades of cataract. The grading agreements were as follows: NO (k = 0.87), NC (k = 0.83), CC (k = 0.89), and PSC (k = 0.81). The overall average of grading agreement was high (k = 0.85). In patients who had undergone unilateral cataract surgery or had a nongradable lens, the LOCS III score of the fellow eye was considered for the analysis. Those who had undergone bilateral cataract surgery either at baseline or at follow-up were excluded from the analysis. At follow-up, one eligible eye of each subject was considered for the analysis. If both eyes had cataract, the eye with the worse stage of cataract was included for the analysis. The eye chosen at follow-up then was examined at baseline for cataract status. 

We defined the incidence of cataract in those with cataract grade <2, which showed an increase in the cataract grade to grade 2 or above. We defined progression of cataract as an increase in cataract grade from 2 or above by one step and as two steps or more. For instance, the incidence of NO at follow-up was defined as those who changed from NO < 2 at baseline to NO ≥ 2 at follow-up. The control group for incident NO was identified as those who remained NO < 2 at follow-up. The control group for NC, CC, and PSC was identified in a similar way. 

Definitions

Known diabetes was defined as subjects on oral hypoglycemic drugs and/or insulin. Newly diagnosed diabetes was defined as subjects with fasting plasma glucose ≥110 mg% on two occasions. The duration of diabetes was the time interval between the date of diagnosis of diabetes (as noted by a diabetologist or when the antidiabetic treatment was started) and the date of examination in the current study. This information was collected while obtaining medical history. 

Statistical Analysis

Statistical analyses were performed using statistical software (SPSS 14.0; SPSS Science, Chicago, IL, USA). The results are expressed as mean ± SD if the variables were continuous and as number and percentage if the variables were categorical. For measures examined at baseline and follow-up, a paired t-test was used to compare normally distributed continuous variables and a McNemar test was used to compare proportions. For unpaired data, a t-test was used to compare normally distributed continuous variables and a χ²/exact test to compare proportions. Newly diagnosed subjects with diabetes were given a value of 0 for duration of diabetes. Univariate and multiple logistic regression analyses were performed to study various baseline factors associated with incidence and progression for each type of cataract. A P value of <0.05 was considered to be statistically significant. 

Results

Table 1 shows the subject characteristics at baseline and at follow-up; the BMI (P = 0.001), BP (P < 0.001), total cholesterol (P < 0.001), low-density lipoprotein (LDL; P < 0.001), HbA_1c levels (P < 0.001), current smokers (P = 0.002), and alcohol use (P < 0.001) were slightly lower at follow-up when compared to baseline. The difference in triglyceride levels between baseline and follow-up was statistically significant (P < 0.001). There were no significant differences in HDL levels or refractive error between baseline and follow-up visits. One subject required laser photocoagulation at follow-up and two required anti-vascular endothelial growth factor (VEGF) injection. Another patient required anti-VEGF and laser photocoagulation. 

Table 1

 

Summary of the Clinical and Ophthalmic Characteristics at Baseline and at Follow-Up Using Paired Analysis

 

Table 2 shows the cumulative incidence and progression of individual cataract type from baseline to year 4. Among the subtypes examined, the cumulative incidence of NO, NC, CC, and PSC was 70%, 55.2%, 25.7%, and 7.3%, respectively; the one-step progression was 14.3%, 16.1%, 8.8%, and 8.1%, respectively; and the two-step or more progression was 5.0%, 6.0%, 0.8%, and 6.0%, respectively. 

Table 2

 

The 4-year Incidence and Progression of Cataract

 

Incidence of Each Cataract Type

Table 3 presents the baseline characteristics of the study subjects with respect to the cumulative incidence of each type of cataract by univariate analysis. The two groups with and without incident NO differed in terms of age distribution (P = 0.006). More patients with no NO were 40 to 49 years old compared to those with incident NO (86.1% vs. 56.3%, respectively), while significantly more patients with incident NO were 50 to 59 years old compared to those with no NO (40.2% vs. 13.9%, respectively). Those with incident NO had 5.0 mm Hg lower mean diastolic BP compared to those with no NO (P = 0.041). There were significant differences in SES between those with no and incident NC, with approximately 56.4% and 71.1% of subjects, respectively, belonging to the middle SES (P = 0.043). The age distribution differed significantly between those with and without incident NC (P < 0.0001); the proportion of patients 40 to 49 years old was significantly higher among those with no NC (71.3% with no NC versus 41.4% with incident NC), but that of patients 50 to 59 years old was significantly higher in the incident NC group (49.3% incident NC versus 25.5% no NC). 

Table 3

 

Univariate Analysis of Risk Factors for Incidence of Cataract Subtype at 4-Year Follow-Up

 

Only 9.5% and 18.9% of subjects with and without incident CC, respectively, were using alcohol (P = 0.027). There were differences in age distribution these two groups (P = 0.009); a higher proportion of people were 50 to 59 years old (46% with versus 44% with no incident CC), and slightly higher proportion were 60 to 69 years old (19% versus only 7.5%, respectively). The incident PSC group had a greater proportion of women (60.4% with versus 38.4% with no PSC; P = 0.017), and had a 0.2% lower mean HbA_1c levels (P = 0.032), and a majority (60.5%) of subjects with hyperopia than those with no PSC (P = 0.035). 

Progression of Each Cataract Type

Table 4 presents the factors associated with one- and two-step or more progression of each cataract type. 

Table 4

 

Univariate Analysis of Risk Factors for Progression of Cataract Subtype at 4-Year Follow-Up

 

One-Step Progression

There were fewer men among those with one-step progression of NO than among those who had stable NO (41.5% vs. 56.9%, respectively; P = 0.006). Those who showed one-step progression of NO had 0.4 mmol/L higher mean total cholesterol (P = 0.003) and 0.3 mmol/L higher LDL (P = 0.034) levels than those with stable NO. They also differed with respect to age distribution (P = 0.004); fewer patients were 40 to 49 years old in the one-step NC progression group (16.7% vs. 41.7% stable), but significantly more were ≥70 years old in the progression group (9% vs. 3.6%, respectively). Those with one-step progression of NC had 1.0 kg/m² lower BMI (P = 0.043), 3.3 mm Hg lower diastolic BP (P = 0.025), and 1.7 years longer duration of diabetes (P = 0.002) compared to those with stable NC. Hyperopia occurred in 57.4% subjects with NC progression compared to 43.7% with stable NC (P = 0.029). The proportion of subjects with emmetropia was lower in the progression group and the differences approached statistical significance (P = 0.050). A higher proportion of subjects had hyperopic refractive error in the NC progression group (57.4% vs. 43.7% with stable NC). The age distribution was different in those with (P < 0.001) than in those without one-step progression of CC; a lower proportion of subjects were 50 to 59 years old (16% in progression versus 38%, respectively), but a higher proportion were 60 to 69 years old (64% vs. 19.2%, respectively). The mean HbA_1c was 2.1% lower in those with one-step progression of PSC than those with stable PSC (P = 0.004). 

Two-Step or More Progression

Table 4 also shows baseline variables in those with two-step or more progression of NO and NC. The mean LDL was 0.5 mmol/L higher in those with two-step or more progression of NO than those with stable NO (P = 0.010). With respect to NC, there were differences in age distribution between the two groups with and without progression (P = 0.011), whereby a lesser proportion of subjects were 55 to 59 years old (33.3% with progression versus 38.8% with stable NC). However, a high proportion of subjects in the progression group were 60 to 69 years old (44.4% versus 18.3% with stable NC). Only 26.1% of patients with two-step or more progression of NC had hyperopia compared to 41.4% of those with no progression (P = 0.041). 

Baseline Factors Associated With Incidence and Progression of Each Cataract Type

Table 5 presents the baseline factors associated with cumulative incidence and progression of each cataract type using multivariate analysis. Patients 50 to 59 years old had an odds ratio (OR) of 3.3 for incident NO compared to those 40 to 49 years old (P = 0.027). Incident NC was associated with OR = 2.7 (P = 0.001) and OR = 3.9 (P = 0.047) in the 50 to 59 and 60 to 69 years age groups, respectively. The 60 to 69 years age group had greater odds of incident CC than the younger age groups (OR = 3.3, P = 0.004). Incident PSC was associated with lower HbA_1c levels at baseline (OR = 0.7, P = 0.006), longer duration of diabetes (OR = 1.1, P = 0.01), and hyperopia (OR = 4.0, P = 0.017). 

Table 5

 

Risk Factors at Baseline for Incidence and Progression of Cataract Subtype at Year 4 Follow-Up Using Multiple Logistic Regression Analysis

 

With regards to progression, one-step progression of NO was associated with female sex (OR = 1.7, P = 0.02) and total serum cholesterol (OR = 1.3, P = 0.009) at baseline. One-step progression of NC was associated with the 60 to 69 years age group (OR = 2.8, P = 0.004) and an OR of 3.8 for the ≥70 years age group (P = 0.028) compared to the younger age groups. The 60 to 69 years age group had 6.3 times greater odds of one-step progression of CC (P = 0.001). One-step progression of PSC was associated with significantly lower levels of HbA_1c (OR = 0.3) at baseline (P = 0.01). 

A two-step or more progression of NO was associated with higher serum LDL cholesterol levels (OR = 1.6, P = 0.011), whereas two-step or more progression of NC was associated with the 60 to 69 years age group (OR = 6.7, P = 0.002) and lower odds of hyperopia (OR = 0.2, P = 0.021). Because of the lower number of subjects with two-step or more progression of CC and PSC, a multivariate analysis was not performed in the CC and PSC cataract subtypes. 

Discussion

We examined the 4-year incidence and progression of cataract subtypes and the factors associated with them in a south Indian population with type 2 diabetes. The 4-year cumulative incidence of cataract ranged from 7.3% for PSC to a maximum of 70% for NO. The 4-year one-step progression ranged from 8.1% for PSC to 16.1% for NC and a two-step or more progression ranged from 0.8% for CC to 6.0% for NC and PSC. In general, the cumulative incidence of cataract was higher than that of progression of cataract in this population with type 2 diabetes mellitus. The cumulative incidences of NO and NC were higher than those of CC and PSC. Rates of one-step progression were higher than those of two-step or more progression. 

Cataract is the most common but reversible cause of vision loss in the south Indian population with type 2 diabetes mellitus.¹⁴ Diabetes is a well-known risk factor for incidence of cataract, being 15 to 25 times higher in diabetic than in nondiabetic subjects.¹⁸ Previous studies reporting incidence and progression of cataract vary. For instance, the Wisconsin Epidemiologic Study of Diabetic Retinopathy¹⁹ reported a 10-year cumulative incidence of cataract surgery of 24.9% in those with type 2 diabetes. In a study conducted in Iran,²⁰ the incidence of cataract was 33.1% per 1000 person-years for a mean follow-up of 3.6 years. We observed that the cumulative incidences of NO and NC were higher than those of CC and PSC. Similarly, in the BDES,²¹ the incidence of nuclear cataract in the worse eye ranged from 3.5% to 75% in women and 14.3% to 33.3% in men, showing an increasing trend with age. The incidence of CC ranged from 7.4% to 38.2% in women and 0.0% to 66.7% in men, and the incidence of PSC ranged from 3.6% to 20% and 4.6% to 16%, respectively. Therefore, the incidence was less frequent for PSC. The BMES²² reported a 10-year person-specific incidence of 36% for nuclear cataract, 28% for CC, and 9.1% for PSC. Results from the aforementioned studies probably suggest that the incidence of nuclear cataract is higher than that of PSC, which is similar to our study findings. In contrast, other studies reported that PSC^7,18 is more common than other cataract types in diabetic patients. However, the prior study¹⁸ differed in terms of study purpose, in which the study investigated the prevalence of diabetes and glaucoma in a large cataract-extracted population. Therefore, the study may have overlooked patients who did not undergo cataract surgery. Our current study, on the other hand, examined the incidence and progression of cataract in a population with confirmed diagnosis of diabetes. In our study, the rates of incidence and progression of cataract were examined in a population with diabetes. It has been reported that one of several risk factors for nuclear cataract is living in a warmer climate,²³ where there is exposure to sun's rays.²⁴ However, wearing sunglasses or hats regularly when going outdoors is not a cultural practice in south India. In contrast, the study by Bernth-Petersen and Elsa Bach¹⁸ was conducted among a cataract-extracted population in Denmark, where the climate is much colder compared to that in south India. The higher incidence of nuclear cataract in our study compared to that of other studies likely may be related to warmer climatic conditions and exposure to the sun. We observed that the incidence of cataract was higher than that of progression in general. Srinivasan et al.²⁵ stated that greater exposure to ultraviolet rays can lead to higher photo-oxidation in the lens and, therefore, lead to greater incidence of cataract; in addition, nutritional deficiencies also have been hypothesized to have a role in the greater incidence of cataract.²⁵ At the molecular level, glycation of lens proteins, oxidative stress and the polyol pathway have been reported to be linked with the development of cataract.^26,27 

On the other hand, 1-year progression of cataract has been reported to be 54% in a study conducted between 1994 and 1995.²⁵ The BDES²¹ reported progression rates of NC to range from 50% to 75.9% in women and 66.7% to 81% in men, while the progression of CC ranged from 30% to 58.3% and 17.4% to 51.7%, respectively, and the progression of PSC ranged from 0% to 20% and 0.0% to 12%, respectively. Similarly, the rates of progression of nuclear cataract were more than those of CC or PSC. Also, the rates of progression in our study were much lower compared to studies conducted approximately two decades ago. This probably reflected the changes in or better standards of diabetes care currently available. In the literature, treatment strategies for preventing late complications of diabetes have been reported to have changed over the years and, consequently, a decrease has been observed in microvascular complications of diabetes.^28,29 With regards to our study population, the subjects with diabetes were under regular care of their physicians and, therefore, this may explain the slightly but significantly lower glycemic levels. This probably is evident from Table 1, in which subjects at follow-up may have had better glycemic control as shown by slightly but significantly lower mean levels of HbA_1c, BP, and the majority of lipid parameters when compared to baseline. 

With regards to risk factors, we observed that greater age at baseline (especially, the 50 to 59 and 60 to 69 years age groups), is a risk factor for cumulative incidence of nuclear cataract (NO and NC) and CC and for progression of NC and CC but not for PSC. Other studies^30–34 have confirmed the findings that greater age at baseline is associated with incidence of cataract in general. The BDES²¹ also reported the association between age and nuclear and cortical cataract only. This could suggest a natural ageing process of the nucleus.³⁵ However, diabetes possibly may have a greater influence than the normal ageing process in the development of PSC. For instance, a study showed that levels of the enzyme aldose-reductase in red blood cells of patients younger than 60 years with short duration of diabetes were correlated positively with the prevalence of PSC,³⁶ suggesting a likely role of the enzyme in the pathogenesis of PSC than that of age or diabetes duration. It also could be because of relatively shorter duration of diabetes at baseline (Table 3 shows mean duration is approximately 3–6 years) and that all subjects were reexamined after 4 years. 

We observed that hyperopic refractive error at baseline was associated with cumulative incidence of PSC, and a lower risk of two-step or more progression of NC. Myopic changes have been observed predominantly in relation to poor glycemic control,³⁷ while hyperopic refractive error is associated with stricter glycemic control.^38–40 We observed that the baseline HbA_1c levels were 0.2% lower in those with incident PSC, which was statistically significant. A strict control of blood glucose in subjects with diabetes is associated with: (1) hydration of the lens cortex,³⁸ and (2) an increased risk of PSC.⁴¹ Thus, glycemic control may be an important factor to be considered in the evaluation of refractive error and incidence and progression of cataract. 

We also observed that females are associated with greater odds of one-step progression of NO. Cross-sectional and population-based studies^42,43 reported that women are at a greater risk of cataract developing in general compared to men. Sex differences in cataracts may be attributed to differential exposure to lifestyle-related factors, such as exposure to ultraviolet (UV) light, smoking or nutritional supplements. Women have been reported to have a higher risk of PSC¹⁰ or CC.³⁰ Other cross-sectional studies report that women are at a greater risk for nuclear opacities.^44–48 The reason for this finding is not fully understood. However, it has been hypothesized that albumin/total protein and serum triglyceride levels in women and postmenopausal estrogen deficiency⁴⁹ may be a factor associated with greater predisposition to nuclear cataract.⁴⁸ 

Previous study found an association between the incidence of PSC and higher BMI.²⁵ Although we did not find any significant association with either BMI/obesity or other phenotypic profiles, our observation that progression of NO is associated with higher cholesterol levels may be associated indirectly with the BMI. 

We observed that at baseline, there only four subjects had undergone either laser photocoagulation and/or anti-VEGF injection. Because of the smaller numbers, we believe that laser photocoagulation or anti-VEGF is less likely to have influenced our results. 

Our study has certain limitations. The use of insulin or oral hypoglycemic agents (OHA) is reported to be associated with increased risk of incidence of cataract,⁵⁰ more specifically CC⁵¹ in subjects with diabetes. However, we did not examine for the influence of insulin or diet or oral hypoglycemic agents on the incidence and progression of cataract. If included, we possibly may have observed an association between use of insulin or OHA with some of the subtypes of cataract and, consequently, the risk factors may have been different with respect to the subtypes of cataract. In addition, a small number of subjects were in the ≥70 years age group compared to other age groups. Therefore, the significant association between those aged ≥70 years and one-step progression of NC also may represent a random association. Also, we did not analyze the possibility of combinations or overlap of NO, NC, CC, or PSC in our subjects. We performed intergrader agreement for cataract grading by having two graders assess the eyes of 50 patients. Although the agreement was good, we did not perform a continuous evaluation of quality of grading. Therefore, this is likely to involve a bias. In addition, the proportion of subjects who had undergone bilateral cataract surgery (125/779 + 125 = 13%) were excluded from the analysis. If we had included cataract surgery in the current study, it likely may have resulted in higher incidence rates. 

In conclusion, the cumulative incidence of cataract is much higher than that of progression in this south Indian population with diabetes. Greater age at baseline is a risk factor for cumulative incidence and progression of most types of cataract. 

Acknowledgments

Supported by Jamshetji Tata, Mumbai, India. 

Disclosure: S. Srinivasan, None; R. Raman, None; G. Swaminathan, None; S. Ganesan, None; V. Kulothungan, None; T. Sharma, None