Prevalence and Risk Factors for Cataract in Diabetes: Sankara Nethralaya Diabetic Retinopathy Epidemiology and Molecular Genetics Study, Report No. 17

Abstract

Purpose.: To report the prevalence of cataract and its subtypes in patients with type 2 diabetes mellitus and the risk factors associated with these cataracts.

Methods.: One thousand two hundred eighty-three eligible subjects with type 2 diabetes mellitus, enrolled from a cross-sectional study, underwent examination at the base hospital. Lens opacity was graded by a trained ophthalmologist according to the Lens Opacity Classification System (LOCS) III system.

Results.: The age- and sex-adjusted prevalence of cataract in the study was 65.7% (95% confidence interval [CI], 65.6–65.8). Mixed cataracts were more common than monotype ones (41.6% vs. 19.4%). The prevalence of cataract was higher in women, subjects with known diabetes and those with longer duration of diabetes (51.4%, 50.3%, and 64.5%, respectively). The risk factors for any type of cataract were increasing age (odds ratio [OR], 1.14; 95% CI, 1.11–1.16), macroalbuminuria (OR, 4.61; 95% CI, 1.56–13.59) and increasing glycosylated hemoglobin (OR, 1.92; 95% CI, 1.22–3.00); higher hemoglobin (OR, 0.38; 95% CI, 0.22–0.64) was the protective factor. The risk factors for nuclear cataract included increasing age (OR, 1.15) and high serum triglycerides (OR, 6.83). For cortical cataract, increasing age (OR, 1.14) and poor glycemic control (OR, 2.43) were the risk factors; increasing hemoglobin (OR, 0.41) was the protective factor. For posterior subcapsular cataract, the risk factors included increasing age (OR, 1.11), being of the female sex (OR, 9.12), employment (OR, 9.80), and duration of diabetes (OR, 21.37).

Conclusions.: Nearly two thirds of the diabetic population showed evidence of cataract; mixed cataracts were more common than the monotypes ones.

Cataract, which accounts for nearly half of the blind population, is the commonest cause of preventable blindness worldwide. ¹ In India, according to a recent survey in the Rapid Assessment of Avoidable Blindness (RAAB) study, cataract was responsible for 77.5% of avoidable blindness. ² Many studies (clinical, epidemiologic, and basic science) have documented an association between diabetes and cataract. ^{3 –24} The data from three population-based studies—the Beaver Dam Eye Study, the Blue Mountains Eye Study, and the Visual Impairment Project—have also shown an association between diabetes and posterior subcapsular cataract, and less consistently, with cortical cataract. ^{5 –9,11 –16,22} The Blue Mountains Eye Study showed that an impaired fasting glucose, in the absence of clinical diabetes, was also a risk factor for the development of cortical cataract. ⁴ There is evidence that the risk of cataract increases with increasing duration of diabetes and severity of hyperglycemia. ²³  

As the prevalence of type 2 diabetes mellitus (DM) is on the rise, more so in India, ³ it is important to study the prevalence of cataract in this select population. Also, in a clinical setting, a mixed cataract—a combination of nuclear, cortical, and subcapsular cataract—is more common than just a single entity. ¹⁶ Therefore, the association between diabetes and mixed cataract would be of great interest in an epidemiologic study. The present study was undertaken to help in understanding how the interplay of clinical and biochemical variables influence the development of cataract and its subtypes, including mixed cataract. So far, this aspect has not been studied extensively. 

We report the prevalence of cataract and its subtypes in patients with type 2 DM mellitus and the risk factors associated with these cataracts in a population-based study. 

Materials and Methods

Study Protocol

The study design and research methodology has been described, in detail, in our previous report (SN-DREAMS 1). ²⁵ In summary, 5999 individuals from the general population, aged ≥40 years, were enumerated based on multistage random sampling which was stratified on the basis of an economic criterion. Subjects identified as having diabetes (according to the WHO criteria ²⁶ ) underwent a detailed examination at the base hospital. The fundi of all patients were imaged with 45°, four-field, stereoscopic digital photography. The diagnosis of diabetic retinopathy (DR) was based on the modified Klein classification of the Early Treatment Diabetic Retinopathy Study scale. ²⁷ Agreement in grading, which was performed by two independent observers in a masked fashion, was found to be high (κ = 0.83). ²⁵ The study was approved by the Institutional Review Board, and a written consent was obtained from the subjects in accordance with the Declaration of Helsinki. 

A detailed history, including data on demographics, diabetes, and ocular history, was obtained from all patients at the base hospital. Socioeconomic status was assessed with a multiple-index questionnaire, and the scoring was characterized as low (score, 0–14), middle (15–28), and high (29–42). ²⁶ Body mass index (BMI) was calculated by using the formula weight (in kilograms)/height (in meters)². The blood pressure (BP) was recorded, in sitting position, in the right arm; two readings were taken 5 minutes apart, and the mean of the two was taken as the blood pressure. All patients underwent biochemical tests: glycosylated hemoglobin, hemoglobin, serum lipids, and albuminuria. The details of these tests are described elsewhere. ²⁷  

The grading of lens opacity, according to the Lens Opacity Classification System (LOCS) III system, was performed by experienced ophthalmologists (PR, RR). ²⁸ After the pupils were dilated 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. The severity of the lens opacities, according to the photographic standards, was separated into four major groups: nuclear opalescence (NO), nuclear color (NC), cortical (CC), and posterior subcapsular (PSC). In patients who had undergone unilateral cataract surgery or had a nongradable lens, the LOCS III score of the fellow eye was used. Those who had undergone bilateral cataract surgery were excluded from the analysis. 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: NO (κ = 0.87), NC (κ = 0.83), CC (κ = 0.89), and PSC (κ = 0.81). The overall overage of grading agreement was high (κ = 0.85). 

Definitions

Employment Status.

Persons engaged in any activity resulting in production of goods and services that added value to the country's gross domestic product were classified as employed. ²⁹ Those who were not working, but were seeking work, were classified as unemployed. 

Education Level.

Education level was classified as primary, secondary, tertiary, or illiterate. Those who had completed elementary education were categorized under primary education. Similarly, secondary education included those who had completed high school, and tertiary education referred to those who had continued their studies beyond high school. Those who could not read and write were regarded as illiterate. 

Smokers.

Those who had any history of smoking were classified as smokers. 

Hypertension.

Patients with a systolic BP ≥140 mm Hg or a diastolic BP ≥ 90 mm Hg or undergoing antihypertensive therapy were regarded as having hypertension. ²⁵  

Significant Cataracts.

A significant NC was identified by the presence of an LOCS III score of >4 for NO or >4 for NC. Similarly, a significant cortical cataract (CC) was identified by an LOCS III score of >2 for CC, and a significant posterior subcapsular cataract (PSC) was identified by an LOCS III score of >2. ^7,30  

Sight-Threatening DR.

Patients with severe nonproliferative DR, proliferative DR, and clinically significant macular edema were regarded as having sight-threatening, referable DR. ³¹  

Albuminuria.

The patient was considered to have normoalbuminuria if the albumin creatinine ratio (ACR) was less than 30 mg/g; microalbuminuria, if the ACR was between 30 and 299 mg/g; and macroalbuminuria, if the ACR was ≥300 mg/g. ³²  

Diabetic Neuropathy.

Diabetic neuropathy was present if the vibration pressure threshold (measured with a sensitometer) was >20 V. ³³  

Of the 5999 subjects enumerated, 1414 with diabetes (both known and newly diagnosed) were analyzed for the study. The response rate for the first fasting blood sugar estimation and base hospital examination was 96.20% and 85.60%, respectively; 8.7% were nondiabetic after the second blood sugar; and 0.78% retinal images were ungradable. ³ Of the 1414 individuals with diabetes, 125 were excluded, as they had undergone bilateral cataract surgeries. After eyes were excluded for which grading was not possible (two with corneal opacity, three with phthisis bulbi, and one enucleation), 1283 subjects were included in the final analysis (Fig. 1). 

Figure 1.

 

Development of the sample for the cataract substudy.

The association between the variables and cataract subtypes was evaluated initially by contingency tables and was estimated by the odds ratio (OR) and 95% confidence interval (CI). Multiple logistic regression models were used to estimate the adjusted OR and to access the influence of other variables on each cataract type. Variables considered for inclusion in the regression model included those at P < 0.05. Trends in proportions were tested for significance by using the Mantel-Haenszel procedure (SPSS ver. 14.0; SPSS, Chicago, IL). 

Results

The age- and sex-adjusted prevalence of cataract in the study was 65.7%. Table 1 shows the age- and sex-adjusted prevalence of the cataract subtypes. Mixed cataracts were more common than monotype (41.6% vs. 19.4%). Among the monotype cataracts, CC was the most common subtype in patients with type 2 DM (15.1%). In the mixed cataracts, the combination of NC, CC, and PSC was the most common (19.5%). 

Table 1.

 

Prevalence of Cataract and Its Subtypes

Cataract	Prevalence			Age- and Sex-Adjusted Prevalence*
	n	%	95% CI	%*	95% CI
None	669	52.1	49.4–54.9	34.3	34.1–34.4
Any	614	47.9	45.1–50.6	65.7	65.6–65.8
Monotype	226	17.6	15.5–19.7	19.4	19.4–19.5
NC	48	3.7	2.7–4.8	5.0	5.0–5.1
CC	161	12.5	10.7–14.4	15.1	15.0–15.2
PSC	17	1.3	0.7–1.9	1.1	1.0–1.1
Mixed	388	30.2	27.7–32.7	41.6	41.5–41.6
NC+CC	46	3.6	2.6–4.6	7.2	7.1–7.2
NC+PSC	16	1.2	0.6–1.9	2.3	2.2–2.3
CC+PSC	191	14.9	12.9–16.8	15.6	15.5–15.7
NC+CC+PSC	135	10.5	8.8–12.2	19.5	19.3–19.6

*  Age and sex adjusted to population of Chennai, based on census of India 2001.

Table 2 shows the prevalence of cataract in various subgroups. The prevalence of cataract was higher in the women than in the men (51.4% vs. 44.8%, relative risk [RR], 1.31) and in subjects with known diabetes than in those in whom it was newly diagnosed (50.3% vs. 37%, RR 1.73). Cataract prevalence was also higher in the subjects with longer duration of diabetes (>10 years) than in those with shorter duration (64.5% vs. 45%, RR 2.22). 

Table 2.

 

Prevalence of Cataract in the Various Subgroups

	Prevalence of Cataract					RR	95% CI
	N	n	%	95% CI	P
Sex
Men	688	308	44.8	41.0–48.5	0.017	1
Women	595	306	51.4	47.4–55.4		1.31	1.05–1.63
Diabetes status
Newly diagnosed	238	88	37.0	30.8–43.1	<0.0001	1
Known	1045	526	50.3	47.3–53.4		1.73	1.29–2.31
Duration of diabetes
Shorter duration, ≤10 y	1097	494	45.0	42.1–47.9	<0.0001	1
Longer duration, >10 y	186	120	64.5	57.6–71.4		2.22	1.6–3.07

Table 3 shows the baseline characteristics in the cataract (n = 614) and noncataract (n = 669) groups. Compared with the noncataract group, the cataract group consisted of older subjects (60.1 vs. 50.9 years), more women (49.8% vs. 43.2%), subjects with longer duration of diabetes (6.3 vs. 4.2 years), more subjects on insulin (6.5% vs. 3.0%) and oral hypoglycemic agents (69.4% vs. 62.2%), more unemployed subjects (59.9% vs. 45.3%), fewer educated subjects (secondary education, 45.6% vs. 50.7%; tertiary education, 20.4% vs. 32.3%), and fewer subjects with higher socioeconomic status (16.4% vs. 20.5%). On examination, the cataract group had a lower BMI (25 vs. 25.9), lower waist circumference (90.7 vs. 92.2 cm), and lower hemoglobin (13.5 vs. 14.1 g%). The cataract group also contained more subjects with diabetic nephropathy (microalbuminuria 17.9% vs. 14.1%; macroalbuminuria 4.7% vs. 0.9%) or DR (any DR 18.1% vs. 12.7%; sight-threatening DR 2.9% vs. 1.0%). 

Table 3.

 

Baseline Characteristics of Study Population

	No Cataract (n = 669)		Any Cataract (n = 614)
	n (%) or Mean ± SD	95% CI	n (%) or Mean ± SD	95% CI
Age, y*	50.9 ± 7.4	50.3–51.5	60.1 ± 9.5	59.3–60.8
Sex, female*	289 (43.2)	39.4–46.9	306 (49.8)	45.9–53.8
Duration of diabetes, y*	4.2 ± 5.4	3.8–4.6	6.3 ± 6.3	5.8–6.8
Drug history
User of insulin*	20 (3.0)	1.7–4.3	40 (6.5)	4.6–8.5
User of OHA*	416 (62.2)	58.5–65.8	426 (69.4)	65.7–73.0
Smoker	152 (22.7)	19.5–25.9	102 (16.6)	13.5–19.5
Employment
Unemployed*	303 (45.3)	41.5–49.1	368 (59.9)	56.0–63.8
Employed	366 (54.7)	50.9–58.5	246 (40.1)	36.2–43.9
Education
None	25 (3.7)	2.3–5.2	37 (6.0)	4.1–7.9
Primary	89 (13.3)	10.7–15.9	172 (28.0)	24.5–31.6
Secondary*	339 (50.7)	46.9–54.5	280 (45.6)	41.7–49.5
Tertiary*	216 (32.3)	28.7–35.8	125 (20.4)	17.2–23.5
Socioeconomic status
Low (score, 0–14)	64 (9.6)	7.3–11.8	79 (12.9)	10.2–15.5
Middle (score, 15–28)	468 (70.0)	66.5–73.4	434 (70.7)	67.1–74.3
High (score, ≥29)*	137 (20.5)	17.4–23.5	101 (16.4)	13.5–19.4
Anthropometry
Waist, cm*	92.2 ± 9.8	91.5–92.9	90.7 ± 9.7	89.9–91.5
BMI*	25.9 ± 4.1	25.6–26.2	25.0 ± 3.9	24.7–25.3
Hypertension	413 (61.7)	58.0–65.4	394 (64.2)	60.4–67.9
Biochemical tests
Total serum cholesterol, mg%	186.7 ± 41.2	183.6–189.8	185.7 ± 40.5	182.5–188.9
Serum triglycerides, mg%	159.5 ± 110.0	151.2–167.8	150.8 ± 94.1	143.4–158.2
Serum LDL cholesterol, mg%	111.3 ± 36.4	108.5–114.1	111.9 ± 34.4	109.2–114.6
Serum HDL cholesterol, mg%	38.6 ± 10.1	37.8–39.4	39.6 ± 10.1	38.8–40.4
Hb%*	14.1 ± 1.5	13.9–14.2	13.5 ± 1.5	13.4–13.6
HbA1c	8.1 ± 2.2	7.9–8.3	8.3 ± 2.2	8.1–8.5
Microalbuminuria*	94 (14.1)	11.4–16.7	110 (17.9)	13.5–19.1
Macroalbuminuria*	6 (0.9)	0.2–1.6	29 (4.7)	1.1–3.4
Diabetes complications
Neuropathy	78 (11.8)	9.2–14.1	138 (22.7)	19.2–25.8
DR*	85 (12.7)	10.2–15.2	111 (18.1)	15.0–21.1
STDR*	7 (1.0)	0.3–1.8	18 (2.9)	1.6–4.3

 OHA, oral hypoglycemic agents; HbA1c, glycosylated hemoglobin; HB, hemoglobin; STDR, sight-threatening diabetic retinopathy.

*  Statistically significant difference (P < 0.05) between groups.

Table 4 shows the results of univariate and multivariate analyses identifying the risk factors for cataract in the subjects with type 2 DM. With any cataract, increasing age (OR, 1.13), duration of diabetes (OR, 3.34), being of the female sex (OR, 1.31), use of insulin (OR, 3.15) or oral hypoglycemic agents (OR, 1.61), middle socioeconomic status (OR, 1.67), poor glycemic control (OR, 1.51), lower serum HDL cholesterol (OR, 1.40), and presence of diabetes-related complications such as nephropathy (microalbuminuria OR, 1.40; macroalbuminuria OR, 5.79), neuropathy (OR, 2.19), and retinopathy (any retinopathy OR, 1.52; STDR OR, 2.86) were risk factors. Being employed (OR, 0.55), higher education (OR, 0.39), smoking history (OR, 0.68), increased waist circumference (OR, 0.65), increased BMI (OR, 0.56), and higher hemoglobin (OR, 0.33) were protective factors. The multivariate analysis identified increasing age (OR, 1.14), macroalbuminuria (OR, 4.61), and increasing glycosylated hemoglobin (OR, 1.92) as risk factors; higher hemoglobin (OR, 0.38) was the protective factor for any cataract. 

Table 4.

 

Risk Factors for Any Cataract in Subjects with Diabetes

	Univariate Analysis			P	Multivariate Analysis
	No Cataract (n = 669)	Any Cataract (n = 614)			Any Cataract (n = 614)
	n (%) or Mean ± SD	n (%) or Mean ± SD	OR (95% CI)		P	OR (95% CI)
Age	50.9 ± 7.4	60.1 ± 9.5	1.13 (1.11–1.15)	<0.0001	1.14 (1.11–1.16)	<0.0001
Sex, female	289 (43.2)	306 (49.8)	1.31 (1.05–1.63)	0.017	1.15 (0.71–1.86)	0.567
Education (ref: no education)
Primary	89 (13.3)	172 (28.0)	1.31 (0.74–2.30)	0.357	1.91 (0.93–3.92)	0.079
Secondary	339 (50.7)	280 (45.6)	0.56 (0.33–0.95)	0.032	1.27 (0.64–2.51)	0.489
Tertiary	216 (32.3)	125 (20.4)	0.39 (0.22–0.68)	0.001	1.05 (0.51–2.19)	0.889
Employment (ref: unemployed)
Employed	366 (54.7)	246 (40.1)	0.55 (0.44–0.69)	<0.0001	1.15 (0.78–1.69)	0.488
Duration of diabetes (ref: 1st quintile)
2nd quintile (0.42–2.00)	150 (22.4)	117 (19.1)	1.43 (1.02–2.02)	0.040	1.38 (0.84–2.72)	0.208
3rd Quintile (2.25–5.00)	168 (25.1)	154 (25.1)	1.68 (1.21–2.34)	0.002	1.20 (0.72–1.99)	0.483
4th Quintile (5.50–10.00)	103 (15.4)	124 (20.2)	2.21 (1.55–3.17)	<0.0001	1.35 (0.77–2.37)	0.297
5th Quintile (11.00–45.00)	66 (9.9)	120 (19.5)	3.34 (2.27–4.92)	<0.0001	1.25 (0.67–2.32)	0.481
Drug history (ref: no drug history)
User of insulin	20 (3.0)	40 (6.5)	3.15 (1.77–5.59)	<0.0001	1.41 (0.65–3.03)	0.384
User of OHA	416 (62.2)	426 (69.4)	1.61 (1.26–2.06)	<0.0001	0.93 (0.62–1.41)	0.748
Smoker	152 (22.7)	102 (16.6)	0.68 (0.51–0.89)	0.006	0.78 (0.53–1.16)	0.226
SES (ref: higher)
Middle	468 (70.0)	434 (70.7)	1.26 (0.94–1.68)	0.119	1.35 (0.93–1.96)	0.109
Low	64 (9.6)	79 (12.9)	1.67 (1.10–2.54)	0.016	1.46 (0.83–2.57)	0.192
Hypertension	413 (61.7)	394 (64.2)	1.11 (0.88–1.39)	0.367	0.74 (0.55–1.00)	0.048
Complications of diabetes
Microalbuminuria	94 (14.1)	110 (17.9)	1.40 (1.04–1.89)	0.028	0.83 (0.56–1.22)	0.339
Macroalbuminuria	6 (0.9)	29 (4.7)	5.79 (2.38–14.06)	<0.0001	4.61 (1.56–13.59)	0.006
Diabetic neuropathy	78 (11.8)	138 (22.7)	2.19 (1.62–2.98)	<0.0001	0.82 (0.55–1.23)	0.343
Diabetic retinopathy	85 (12.7)	111 (18.1)	1.52 (1.12–2.06)	0.008	1.14 (0.75–1.74)	0.533
STDR	7 (1.0)	18 (2.9)	2.86 (1.18–6.89)	0.019	1.06 (0.33–3.35)	0.921
Waist circumference, cm (ref: 1st quintile)
2nd quintile (84–89)	144 (21.5)	148 (24.1)	0.92 (0.66–1.29)	0.632	1.07 (0.68–1.66)	0.778
3rd Quintile (90–93)	115 (17.2)	101 (16.4)	0.79 (0.55–1.13)	0.196	1.03 (0.61–1.73)	0.925
4th Quintile (94–99)	144 (21.5)	124 (20.2)	0.77 (0.55–1.09)	0.139	0.97 (0.56–1.67)	0.921
5th Quintile (100–144)	145 (21.7)	106 (17.3)	0.65 (0.46–0.93)	0.018	0.87 (0.46–1.65)	0.675
BMI (ref: 1st quintile)
2nd quintile (21.97–24.17)	127 (19.0)	128 (20.8)	0.79 (0.55–1.12)	0.183	1.01 (0.63–1.63)	0.961
3rd Quintile (24.20–26.06)	131 (19.6)	126 (20.5)	0.75 (0.53–1.07)	0.111	1.22 (0.73–2.05)	0.451
4th Quintile (26.08–28.31)	152 (22.7)	113 (18.4)	0.58 (0.41–0.82)	0.002	0.93 (0.53–1.62)	0.794
5th Quintile (28.33–51.95)	152 (22.7)	110 (17.9)	0.56 (0.39–0.80)	0.002	0.94 (0.49–1.79)	0.847
HbA1c (ref: 1st quintile)
2nd quintile (6.4–7.2)	142 (21.2)	115 (18.7)	1.12 (0.79–1.59)	0.516	1.04 (0.67–1.61)	0.866
3rd Quintile (7.3–8.4)	136 (20.3)	130 (21.2)	1.33 (0.94–1.87)	0.111	1.42 (0.91–2.19)	0.120
4th Quintile (8.5–9.9)	125 (18.7)	133 (21.7)	1.48 (1.04–2.09)	0.029	1.92 (1.22–3.00)	0.005
5th Quintile (10.0–18.6)	119 (17.8)	130 (21.2)	1.51 (1.06–2.15)	0.021	1.58 (0.99–2.52)	0.053
Hb% mg% (ref: 1st quintile)
2nd quintile (12.7–13.3)	111 (16.6)	143 (23.3)	0.76 (0.54–1.09)	0.139	0.68 (0.44–1.04)	0.077
3rd Quintile (13.4–14.2)	154 (23.0)	117 (19.1)	0.45 (0.32–0.64)	<0.0001	0.49 (0.32–0.77)	0.002
4th Quintile (14.3–15.0)	149 (22.3)	106 (17.3)	0.42 (0.29–0.60)	<0.0001	0.41 (0.25–0.67)	<0.0001
5th Quintile (15.1–22.3)	160 (23.9)	88 (14.3)	0.33 (0.23–0.47)	<0.0001	0.38 (0.22–0.64)	<0.0001
Total serum cholesterol, mg% (ref: 1st quintile)
2nd quintile (32–36)	130 (19.5)	127 (20.7)	1.10 (0.78–1.55)	0.579	1.08 (0.61–1.91)	0.783
3rd Quintile (37–40)	134 (20.1)	119 (19.4)	1.00 (0.71–1.41)	0.992	1.35 (0.66–2.75)	0.404
4th Quintile (41–47)	136 (20.4)	119 (19.4)	0.99 (0.70–1.39)	0.941	1.69 (0.73–3.89)	0.218
5th Quintile (48–102)	127 (19.0)	124 (20.2)	1.10 (0.78–1.56)	0.584	2.23 (0.78–6.42)	0.136
Serum HDL cholesterol, mg% (ref: 1st quintile)
2nd quintile (153–175)	141 (21.1)	145 (23.6)	1.40 (1.00–1.97)	0.048	1.29 (0.85–1.96)	0.235
3rd Quintile (176–195)	132 (19.8)	116 (18.9)	1.20 (0.85–1.70)	0.304	1.03 (0.65–1.60)	0.912
4th Quintile (196–218)	123 (18.4)	127 (20.7)	1.41 (0.99–1.99)	0.053	1.09 (0.69–1.73)	0.699
5th Quintile (219–378)	119 (17.8)	114 (18.6)	1.31 (0.92–1.86)	0.136	0.98 (0.61–1.58)	0.930
Serum triglycerides, mg% (ref: 1st quintile)
2nd quintile (82–115)	154 (23.1)	134 (21.8)	0.85 (0.60–1.21)	0.378	0.78 (0.50–1.22)	0.285
3rd Quintile (116–148)	129 (19.3)	125 (20.4)	0.95 (0.66–1.36)	0.788	0.88 (0.55–1.41)	0.594
4th Quintile (149–199)	126 (18.9)	126 (20.5)	0.98 (0.69–1.41)	0.923	0.86 (0.53–1.42)	0.565
5th Quintile (200–990)	147 (22.0)	115 (18.7)	0.77 (0.54–1.09)	0.149	0.62 (0.34–1.11)	0.110
Serum LDL cholesterol, mg% (ref: 1st quintile)
2nd quintile (82.88–102.68)	129 (19.3)	126 (20.5)	1.15 (0.81–1.62)	0.435	0.86 (0.51–1.46)	0.575
3rd Quintile (102.86–120.50)	128 (19.2)	131 (21.3)	1.20 (0.85–1.69)	0.294	0.65 (0.33–1.28)	0.214
4th Quintile (120.68–140.66)	141 (21.1)	119 (19.4)	0.99 (0.70–1.40)	0.962	0.38 (0.17–0.86)	0.021
5th Quintile (140.84–288.62)	129 (19.3)	118 (19.2)	1.07 (0.76–1.52)	0.685	0.37 (0.13–1.04)	0.059

 Significant results are shown in bold. STDR, sight-threatening diabetic retinopathy; OHA, oral hypoglycemic agents; SES, socioeconomic status; HbA1c, glycosylated hemoglobin; Hb, hemoglobin.

We also analyzed the data to find out the influence of the subject's sex, duration of diabetes, and the presence of known or newly diagnosed diabetes on the risk factors causing cataract. Table 5 enumerates the risk factors that showed these differences. Increasing age and low hemoglobin were risk factors for both sexes; however, macroalbuminuria was noted to be significant only in the men. For both shorter and longer duration of diabetes (≤10 years vs. >10 years), increasing age was associated with cataract development. Macroalbuminuria, low hemoglobin, and low serum cholesterol were significant with shorter duration of diabetes, and glycosylated hemoglobin was significant with longer duration. Increasing age was significantly related to cataract in both newly diagnosed and known diabetes; however, macroalbuminuria and low hemoglobin were associated with newly diagnosed diabetes. 

Table 5.

 

Differences in the Risk Factors for Any Cataract

	Sex		Duration of Diabetes		Subjects with Diabetes
	Men	Women	Shorter Duration (≤10 years)	Longer Duration (>10 years)	Newly Diagnosed	Known Diabetes
	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)	OR (95% CI)
Age	1.15 (1.12–1.18)*	1.14 (1.10–1.17)*	1.14 (1.11–1.16)*	1.18 (1.10–1.26)*	1.14 (1.12–1.17)*	1.13 (1.08–1.19)*
Macroalbuminuria	11.38 (2.77–46.66)*	1.34 (0.29–6.21)	8.74 (2.35–35.54)*	0.88 (0.09–8.37)	5.39 (1.68–17.33)*	13.55 (0.60–303.36)
HbA1c	1.07 (0.97–1.17)	1.09 (0.98–1.19)	1.04 (0.97–1.12)	1.24 (1.02–1.51)*	1.07 (0.99–1.16)	1.05 (0.89–1.22)
Hb%	0.82 (0.71–0.94)*	0.79 (0.66–0.94)*	0.81 (0.72–0.92)*	0.84 (0.62–1.13)	0.76 (0.68–0.86)*	1.17 (0.88–1.56)
Total serum cholesterol, mg%	0.99 (0.99–1.00)	0.99 (0.99–1.00)	0.99 (0.99–0.99)*	1.00 (0.99–1.02)	0.99 (0.99–1.00)	0.99 (0.98–0.99)

*  Significant difference.

Table 6 shows the results of multivariate analyses for risk factors (only the statistically significant ones) in all types of monotypes and mixed cataracts. For NC, increasing age (OR, 1.15) and high serum triglycerides (OR, 6.83) were the risk factors. For CC, increasing age (OR, 1.14) and poor glycemic control (OR, 2.43) were the risk factors; increased hemoglobin (OR, 0.41) was the protective factor. For PSC, the risk factors included increasing age (OR, 1.11), being female (OR, 9.12), being employed (OR, 9.80), and duration of diabetes (OR, 21.37). 

Table 6.

 

Multivariate Analysis for Risk Factors for Cataract Subtypes

Risk Factors	95% CI	P
Monotype Cataract
NC
Age, per year increase	1.15 (1.09–1.20)	<0.0001
Serum triglycerides
4th quintile (149–199)	5.46 (1.32–22.55)	0.019
5th quintile (200–990)	6.83 (1.22–38.18)	0.029
CC
Age, per year increase	1.14 (1.11–1.18)	<0.0001
HbA1c
3rd quintile (7.3–8.4)	2.09 (1.06–4.12)	0.034
4th quintile (8.5–9.9)	2.59 (1.31–5.13)	0.006
5th quintile (10.0–18.6)	2.43 (1.18–5.02)	0.016
Hb%
4th quintile (14.3–15.0)	0.31 (0.15–0.65)	0.002
5th quintile (15.1–22.3)	0.41 (0.19–0.92)	0.030
PSC
Age, per year increase	1.11 (1.01–1.22)	0.038
Sex, female	9.12 (1.09–76.14)	0.041
Employed	9.80 (1.14–84.07)	0.037
Duration of DM
5th quintile (11.00–45.00)	21.37 (1.19–381.21)	0.037
Mixed Cataract
NC+CC
Age, per year increase	1.55 (1.31–1.84)	<0.0001
Education
Secondary	0.04 (0.00–0.46)	0.010
Hb%
3rd quintile (13.4–14.2)	0.04 (0.00–0.49)	0.013
4th quintile (14.3–15.0)	0.06 (0.00–0.85)	0.037
CC+PSC
Age, per year increase	1.10 (1.07–1.13)	<0.0001
SES
Low	2.39 (1.04–5.47)	0.040
Macroalbuminuria	4.16 (1.19–14.47)	0.025
Neuropathy	0.35 (0.17–0.69)	0.003
Hb%
3rd quintile (13.4–14.2)	0.04 (0.00–0.49)	0.013
4th quintile (14.3–15.0)	0.06 (0.00–0.85)	0.037
Total serum cholesterol
4th quintile (41–47)	3.37 (1.11–10.22)	0.032
5th quintile (48–102)	5.15 (1.32–20.33)	0.018
Serum triglycerides
5th quintile (200–990)	0.43 (0.19–0.98)	0.045
NC+CC+PSC
Age, per year increase	1.33 (1.25–1.41)	<0.0001
Employed	2.56 (1.07–6.10)	0.034
Macroalbuminuria	17.14 (2.21–132.67)	0.007
HbA1c
5th quintile (10.0–18.6)	3.83 (1.33–11.04)	0.013
Hb
3rd quintile (13.4–14.2)	0.24 (0.08–0.67)	0.007
4th quintile (14.3–15.0)	0.09 (0.03–0.32)	<0.0001
5th quintile (15.1–22.3)	0.22 (0.07–0.71)	0.012

 The 1st quintile is the reference in each group. All results were significant. Other biochemical factors-serum total cholesterol, serum LDL cholesterol, serum HDL cholesterol, and serum triglycerides-had no significant association in any group.

For combined NC and CC, the risk factors were increasing age (OR, 1.55), low education (OR, O.04), and low hemoglobin (OR, 0.06). For combined CC and PSC, the risk factors were increasing age (OR, 1.10), low socioeconomic status (OR, 2.39), macroalbuminuria (OR, 4.16), and high serum cholesterol (OR, 5.15); high hemoglobin (OR, 0.06) and high serum triglycerides (OR, 0.43) were protective. For combined NC, CC, and PSC, increasing age (OR, 1.33), being employed (OR, 2.56), macroalbuminuria (OR, 17.14), and poor glycemic control (OR, 3.83) were the risk factors; higher hemoglobin (OR, 0.22) was the protective factor. 

Table 7 summarizes the results of correlation and regression analyses for risk factors with regard to the LOCS III grades of NO, NC, CC, and PSC. Increasing age correlated positively for all types of cataract, the maximum effect being on NO (r = 0.629, β = 4.302) and the least effect on PSC (r = 0.349, β = 2.348). High serum triglycerides correlated negatively with NO (r = 0.026, β = −1.883) and NC (r = 0.02, β = −1.707). For CC, a positive correlation was evident with glycosylated hemoglobin (r = 0.077, β = 0.121), and a negative correlation, with hemoglobin (r = 0.151, β = −0.169). For PSC, a positive correlation was evident with the duration of diabetes (r = 0.138, β = 0.547) and being female (r = 0.099, β = 0.003) and a negative correlation with being employed (r = 0.055, β = −0.018). 

Table 7.

 

Correlation and Regression Analysis for Risk Factors for LOCS III Grades

Risk Factors*	Correlation (r)	Standardized Regression (β)	P
LOCS III NO grades
Age, y	0.629	4.302	<0.0001
Serum triglycerides, mg%	0.026	−1.883	0.357
LOCS III NC grades
Age, y	0.578	3.357	<0.0001
Serum triglycerides, mg%	0.02	−1.707	0.326
LOCS III CC grades
Age, y	0.461	3.175	<0.0001
HbA1c	0.077	0.121	0.006
Hb%	0.151	−0.169	<0.0001
LOCS III PSC grades
Age, y	0.349	2.348	<0.0001
Duration of diabetes, y	0.138	0.547	<0.0001
Employed	0.055	−0.018	0.048
Female	0.099	0.003	0.743

*  Risk factors that were significant in multiple regression analysis for monotype subtypes. Significant results are shown in bold.

Discussion

The prevalence of type 2 DM and its associated complications is increasing at an alarming rate in India. ³ Although one of the most important causes of visual impairment in subjects with type 2 DM is diabetic maculopathy, cataract—an avoidable cause of blindness—also accounts for visual morbidity. This study showed the high prevalence of cataract to be in such populations: nearly two of three persons >40 years of age with type 2 DM. Identifying this cause and managing the same with surgical treatment would be an important step in reducing the economic burden, particularly among the working-age group, in the population with diabetes. 

In our study, we found that the prevalence of mixed cataract was higher (42%), more than twice that of monotype cataract (19%). Of the mixed types, the most common type was a combination of NC, CC, and PSC (∼20%) followed by the combination of CC and PSC (∼16%). Of the monotype ones, the most common cataract CC (∼15%) followed by NC (5%) and PSC (1%). Table 8 shows a comparison of the prevalence rates of these cataracts with those in published reports. We selected reports in which the subject profile was somewhat similar to that in our study, and the methodology adopted was that of LOCS III grading. To our surprise, there are few reports of cataract profiles obtained exclusively in populations with diabetes. The prevalence of any cataract ranges from 35% to 48% in the general population ^14,16,34,35 ; however, in subjects with diabetes, it is higher (∼66%). ³⁶ The prevalence of NC is ∼32% in the general population, ^14,16,34 but it is lower in subjects with diabetes (5% in the present study). PSC is found in ∼14% of the general population, ^16,34 but in just 1% of the present study's population. Similarly, the prevalence of CC is ∼21% in the general population and ∼18% in subjects with diabetes. The Framingham Study also noted that CC was more common in individuals with diabetes. ³⁷ Similarly, the Lens Opacities Case–Control Study observed a lower frequency of NC in patients with diabetes. ¹²  

Table 8.

 

Comparison with Published Data

Study	n	Age (y)	Assessment	Prevalence
				Any Cataract	NC	CC	PSC	Mixed
General population
Tanjong Pagar Survey, Singapore 14	1206	>40	LOCS III	34.7	22.6	23.9	7.0	—
Sumatra Eye study, Indonesia 34	919	>21	LOCS III	23.0	35.7	30.1	15.1	—
Andhra Pradesh Eye study 35	7416	>16	LOCS III and Wilmer	14.4	9.2	5.5	6.0	—
Aravind Comprehensive Eye study 16	5150	>40	LOCS III	47.5	43.5	13.9	19.9	47.7
Subjects with diabetes
Chen et al. 36	578	>30	LOCS III	—	22.5	20.5	19.9	—
Present study	1414	>40	LOCS III	65.7	5.0	15.1	1.1	44.6

In the present report, the multivariate analysis identified four variables—increasing age, poor glycemic control, anemia, and macroalbuminuria—as risk factors associated with any cataract (Table 4). In a subgroup analysis, macroalbuminuria and anemia were risk factors for cataract in subjects with shorter duration of diabetes and in those with newly diagnosed diabetes, whereas poor glycemic control was related to longer duration of diabetes (Table 5). 

The WESDR (Wisconsin Epidemiologic Study of Diabetic Retinopathy) also showed a direct relationship between glycosylated hemoglobin and cataract. ²² Other studies found an association of cataract with anemia and diabetic nephropathy among subjects with diabetes. ^{38 –40} It is possible that, in diabetes patients who have anemia and diabetic nephropathy, an osmotic imbalance and electrolyte disturbances is responsible for the increased prevalence of cataract. A similar association between anemia and cataract was observed by other investigators as well. ^{38 –40} It is likely that disease courses vary at different ages: osmotic imbalance in younger patients (shorter duration and newly diagnosed) and metabolic abnormalities—advanced glycation end (AGE) products—in older ones (longer duration of diabetes). 

On subgroup analysis, more women than men had cataract. Donnelly et al. ⁴¹ noted differences in the albumin/total protein ratio and serum triglyceride level, especially in women, predisposing them to cataract. Postmenopausal estrogen deficiency may also be a contributing factor. Recent epidemiologic data suggest that estrogen and hormone replacement therapy (HRT) play a protective role in reducing the incidence of age-related cataract and cataract surgery. ⁴²  

The prevalence of cataract was higher in those with a longer duration of diabetes and known diabetes, suggesting a more prolonged influence of biochemical cataractogenic stimuli (hyperglycemia). However, the pathways by which hyperglycemia leads to cataract are still unknown but they probably involve a modification of the lens proteins leading to AGE formation or modification of the ATPase pumps, leading to osmotic stress, or both. ⁴³  

For monotype cataract, multivariate analysis showed that predominantly biochemical variables such as anemia, glycemic control, and triglyceride levels were the risk factors. However, for mixed cataract, lifestyle variables such as education, lower socioeconomic status and employment, also played a role as risk factors. We also noted that abnormalities in the lipid metabolism were related to cataract; for example, high serum triglycerides with NC and high serum cholesterol with shorter duration of diabetes. 

Unlike our study, the Framingham Study showed an association between elevated serum triglycerides and PSC. ³⁴ These differences in risk factors, among the different populations, can be related to the different genetic patterns associated with different types of cataract. Similar to our results, an association between early onset of cataract and serum cholesterol was seen in animal studies. In animals, with streptozotocin-induced diabetes, cataract appeared sooner in those fed a cholesterol-rich diet than in those fed an ordinary diet. However, the plasma glucose levels did not differ between the groups. The onset of cataract correlated positively with high plasma total cholesterol, high triglycerides, and high non-HDL cholesterol. ⁴⁴  

The major strengths of this study include its population-based design and the presence of a well-defined population of diabetes in which lens changes were documented by using standardized photographic grading with LOCS III charts. Such documentation minimized the chances of bias in classifying the various types of cataract, and the results can be extrapolated to a specific population with type 2 DM. This kind of information would be extremely useful for health care providers in developing long-term strategies to combat avoidable blindness, worldwide. The present study is unique, as it highlighted the risk factors—both biochemical and lifestyle variables—for both monotypes and mixed cataracts. Modulating some of these variables may delay the occurrence of cataract in populations with type 2 DM; however, such speculation would have to be supported by future studies. 

A limitation of the study was our inability to validate the causal relationship between identified risk factors and cataracts. Also, risk factors, such as sunlight exposure and nutritional history, which may play an important role in cataractogenesis, were not taken into account. 

Footnotes

 Supported by the R. D. Tata Trust, Mumbai, India.

Footnotes

 Disclosure: R. Raman, None; S.S. Pal, None; J.S.K. Adams, None; P.K. Rani, None; K. Vaitheeswaran, None; T. Sharma, None