Prevalence and Risk Factors of Diabetic Retinopathy Among Noninsulin-dependent Diabetic Subjects M u h - S h y C h e n , M.D., C h i e - S h u n g Kao, M.D., Chih-Jen Chang, M.D., Ta-Jen W u , M.D., C h e n - C h u n g Fu, M.D., Chien-Jen C h e n , Sc.D., and Tong-Yuan Tai, M . D . In a population-based study in Taiwan, 11,478 subjects aged 40 years or older were screened for diabetes in one urban and five rural areas. Among the 715 subjects proven to have diabetes, 527 subjects underwent ophthalmoscopy. Diabetic retinopathy was present in 184 of the 527 subjects (35.0%), including background diabetic retinopathy in 157 subjects (30.0%), preproliferative diabetic retinopathy in 15 subjects (2.8%), and proliferative diabetic retinopathy in 12 subjects (2.2%). Diabetic retinopathy was correlated with the duration of diabetes and age at onset of diabetes, type of diabetes treatment, higher serum creatinine levels, and lower serum cholesterol levels. Several other factors, including gender, age, residential area, family income, educational level, control and family history of diabetes, body mass index, physical activity, exercise, cigarette smoking, stroke, ischemie heart disease, leg vessel disease, hypertension, and proteinuria, had no significant association with retinopathy. By multiple logistic regression analysis, duration of diabetes was the most important risk factor related to retinopathy. Diabetic subjects treated with insulin had a higher risk of developing retinopathy than those treated with dietary control (relative risk, 1.57; .05 < P < .10). The univariate analysis disclosed that proliferative diabetic retinopathy was related to older age at
Accepted for publication Sept. 17, 1992. From the Departments of Ophthalmology (Drs. M.-S. Chen and Kao), Public Health (Dr. C.-J. Chen), and Internal Medicine (Drs. Chang, Wu, Fu, and Tai), Col lege of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China. This study was supported by grant DOH75-0299-25 from the Department of Health, Executive Yuan, Taiwan, Republic of China (Dr. Tai). Reprint requests to Tong-Yuan Tai, M.D., No. 7, Chung-Shan South Road, Department of Internal Medi cine, College of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China.
examination, older age at onset of diabetes, type of diabetes treatment, and presence of leg vessel disease. Insulin-treated diabetic subjects also had a higher risk of proliferative diabetic retinopathy than patients in whom diabetes was controlled by diet, with a relative risk of 2.51 (.05 < P < .10) in the multiple logistic regression analysis.
IN RECENT DECADES there has been an increase in the number of individuals with diabe tes. 1 Noninsulin-dependent diabetic patients represent a large portion of such patients. With the economic development in Taiwan during the past few decades, the prevalence of diabetes mellitus among those aged 40 years or older has increased markedly. In the past three surveys performed in 1970, 1979, and 1986 in Taipei City, the prevalence of diabetes was 5.05%, 7.10%, and 8.17%, respectively. 26 In the rural areas of Taiwan, a prevalence of 5.06% was noted during a survey performed simultaneous ly with that of Taipei City in 1986. Diabetic retinopathy has now become a major cause of loss in visual acuity in adults, which is often insidious in noninsulin-dependent diabetes. 6 Epidemiologie studies of diabetic retinopathy have been conducted among different popula tions in many areas of the world. The preva lence of diabetic retinopathy varies widely in different countries. 716 Despite different study procedures and popu lations, several factors, including duration of diabetes, 7 · 8 · 111432 age at the onset of diabe15 18 19 21 22 26 type t e s 7.11,16,17,21,25 a g e a t examination, ' · · · · of d i a b e t e s treatment, 16 · 18,20,21 c o n t r o l of d i a b e t e 8,15,16,18,20-24,26,28-31,33-36 h y p e r t e n s i o n 16.20,23,30,31,37,38
proteinuria,11·16'20,21'23'30,36 serum creatinine lev el,33 serum cholesterol level,32,38 and body mass index,13,16,21,23,81,38 have been found to correlate with the prevalence of diabetic retinopathy. In our population-based study, we investigated
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:723-730, DECEMBER, 1992
723
724
AMERICAN JOURNAL OF OPHTHALMOLOGY
the prevalence of diabetic retinopathy among noninsulin-dependent diabetic patients in Tai wan and determined the factors that correlated with the presence of retinopathy.
Subjects and Methods This survey was conducted from November 1985 to June 1986. Among the 78 subdistricts of Ta-An Precinct in Taipei City, eight subdistricts were randomly selected. Five villages evenly distributed over the island (Suan-Shi, Bar-Ter, Zu-Tan, Su-Fu, and Chih-Ku villages) were also randomly selected. The total number of subjects aged 40 years or older who underwent the diabetic survey in Taipei City was 4,272 (2,194 men and 2,078 women) and that of the five villages was 7,206 (3,582 men and 3,624 women), with a response rate of 65.3% and 72.0%, respectively. With the help of specially trained public health nurses, all subjects were interviewed according to a structured questionnaire. The questionnaire included details on age, gender, height, weight, residential area, monthly fami ly income, educational level, personal and fam ily history of diabetes, degree of physical ac tivity, frequency of exercise, and cigarette smoking. The capillary blood glucose of each partici pant was examined either after an overnight fast or two hours postprandially. Diabetes mellitus was diagnosed according to the World Health Organization criteria.39,40 Diagnosis was confirmed by a 75 -g oral glucose tolerance test in those patients with equivocal diagnostic val ues.40 By using these criteria, diabetes was diag nosed in 715 subjects, with a prevalence rate of 6.23%. Among the 715 subjects, 608 subjects, including 225 subjects in whom diabetes was newly diagnosed and 383 subjects in whom diabetes was previously diagnosed, participat ed in further systemic and laboratory examina tions. Among the 383 patients in whom diabe tes was previously diagnosed, 44 subjects were treated with insulin and 298 subjects were treated with sulfonylureas. Another 41 subjects treated by diet control only were checked for the blood glucose level criteria to confirm their diagnosis. Diabetic control was classified as good, fair, or poor when the fasting plasma glucose levels of diabetic patients were 120 m g / d l or less, 121 to 140 mg/dl, and greater than 140 m g / d l , respectively; or when the two-hour postprandi
December, 1992
al levels were 160 m g / d l or less, 161 to 180 m g / d l , and greater than 180 m g / d l , respective ly.41 Venous blood was collected for serum creatinine, cholesterol, and high-density lipoprotein cholesterol measurements in all subjects. Proteinuria was determined by commercial urine strips. According to the protocol of the World Health Organization Multinational Study on Vascular Disease in Diabetes, 42 large-vessel diseases, in cluding stroke, ischemie heart disease, and leg vessel disease, were investigated. Stroke was defined when neurologic sequelae were present or when unilateral weakness lasted for more than 24 hours. Ischemie heart disease was diag nosed by history and electrocardiographic ab normalities as designated by the Minnesota Code.42 Leg vessel disease was diagnosed by evidence of amputation or a history of intermit tent claudication. Amputation was recorded when the subject had a history of gangrene and necrotic or surgical loss of tissue in the toe, foot, or leg because of arterial obstruction judged by the investigators and that occurred after the onset of diabetes. Hypertension was deemed present when the systolic pressure was 160 mm Hg or greater, or when the diastolic pressure was 95 mm Hg or greater. 43 Body mass index was measured by dividing the weight in kilograms by the square of height in meters. Those subjects with a body mass index of 25 to 30 kg of body weight/m 2 were deemed overweight and those with an index over 30 kg of body weight/m 2 were deemed obese. Physical activity was divided into light (sedentary), moderate, and heavy (laborious) groups. The exercise status (number of times per week) was classified as less than two, two to four, and five or more times per week. The duration of exercise was at least 30 minutes per day. The monthly family income and personal educational level were also investigated. The details of the geographic distribution of study areas and study methods have been described previously. 4,5 Subjects who had not smoked more than 100 cigarettes in their lifetime were classified as nonsmokers; subjects who had smoked more than 100 cigarettes in their lifetime but had stopped smoking before the examination were classified as ex-smokers; and subjects who smoked at the time of the interview were classi fied as current smokers. The ophthalmic examinations were per formed with direct and indirect ophthalmoscopy after dilating the pupils with tropicamide
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
1% and phenylepinephrine 10%. The retinopa thy status was divided into the following five categories: (1) no diabetic retinopathy; (2) background diabetic retinopathy, which was defined as the presence of one or more of the following clinical signs: microaneurysms, punc tate or striate intraretinal hemorrhages, and hard exudates; (3) preproliferative diabetic retin opathy, which was defined as soft exudates, venous beading, and intraretinal microvascular abnormalities; (4) proliferative diabetic reti nopathy, characterized by neovascularization on or within one disk diameter of the disk in extent, neovascularization elsewhere in the ret ina, and fibrous proliferative tissue; and (5) ungradable retinopathy, with media opacities, including leukoma from old trachoma and dense cataracts that precluded gradation of retinopathy. Diabetic subjects were classified according to the most severe changes in the worse eye. Ophthalmoscopy was performed in 527 of the diabetic subjects. All diabetic subjects had noninsulin-dependent diabetes, characterized by the absent symptoms of uncon trolled diabetes, including polyuria, polydipsia, and an evident weight loss at the time of diagnosis, and had not been treated with insu lin during the first year after diagnosis. 4 ' 6,40 The characteristics of 527 participants and 81 nonparticipants were similar. For participants and nonparticipants, the proportions of male gen der were 45.0% and 46.9%, respectively; the mean ages ± standard deviation were 61.1 ± 9.3 years and 62.0 ± 8.9 years, respectively; the mean durations of diabetes ± standard devia tion were 3.7 ± 5.3 and 3.9 ± 5.6 years, respec tively; the percentages of subjects treated with insulin were 7.0% and 9.9%, respectively; the prevalences of hypertension were 31.7% and 35.8%, respectively; and the prevalences of proteinuria were 9.6% and 12.5%, respectively. For participants and nonparticipants, mean fasting plasma glucose levels ± standard devia tion were 160.0 ± 66.9 m g / d l (375 subjects) and 171.5 ± 79.4 m g / d l (61 subjects), respec tively, and mean two-hour postprandial levels ± standard deviation were 232.4 ± 79.0 m g / d l (186 subjects) and 212.4 ± 68.5 m g / d l (20 subjects), respectively. The differences in the previously mentioned characteristics were not significant. There were also no significant dif ferences for residential area, income, educa tional level, age at onset of diabetes, body mass index, physical activity, exercise, cigarette smoking, large-vessel disease, and serum levels of creatinine, cholesterol, and high-density lipoprotein cholesterol.
725
The statistical significance was assessed by using the Student's f-test and the chi-square test. Correlation coefficients were estimated by Spearman's method and multiple logistic re gression analysis was also performed to evalu ate multivariate-adjusted associations between related factors and diabetic retinopathy.
Results The retinopathy status of the 527 participat ing subjects (237 men and 290 women) was determined by age and gender (Table 1). Dia betic retinopathy was present in 184 of the 527 subjects (35.0%). Background diabetic retinop athy was present in 157 subjects (30.0%), pre proliferative diabetic retinopathy was present in 15 subjects (2.8%), and proliferative diabetic retinopathy was present in 12 subjects (2.2%). No diabetic retinopathy was present in 334 subjects (63.3%) and retinopathy was ungrad able in nine subjects (1.7%). The relationship between the prevalence of retinopathy and sociodemographic characteris tics was investigated. The difference in preva lence was not statistically significant between men and women. The prevalence of retinopathy was not statistically different when the subjects were divided according to age (40 to 49 years, 50 to 59 years, 60 to 69 years, and 70 years or older). However, the oldest diabetic subjects had the highest prevalence of proliferative dia betic retinopathy (P < .05 on the basis of a chi-square test). Residential area, monthly fam ily income, and educational level were not significantly associated with retinopathy. Family history of diabetes, body mass index, degree of physical activity, frequency of exer cise, and cigarette smoking were not associated with the prevalence of either total retinopathy or proliferative diabetic retinopathy. Subjects with a longer duration of diabetes had a higher percentage of total retinopathy (Table 2) (P < .01 on the basis of a chi-square test). The prevalence of retinopathy was different among subjects with different types of diabetic treatment (P < .01). The subjects treated with insulin had a higher prevalence of total retinop athy (58.3%) and proliferative diabetic retinop athy (8.3%) than subjects not treated with insu lin (33.8% and 1.9%, respectively) (Table 2). The control of diabetes and the prevalence of total and proliferative diabetic retinopathy were not associated (Table 2). However, for all categories of retinopathy, those subjects with
726
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
TABLE 1 PREVALENCE OF RETINOPATHY STATUS BY AGE AT EXAMINATION AND GENDER IN DIABETIC SUBJECTS RETINOPATHY STATUS OF SUBJECTS
NO DIABETIC RETINOPATHY
AGE AT EXAMINATION
(vns)
40-49 50-59 60-69 70 years or older Total
GENDER
NO.
BACKGROUND
PREPROLIFERATIVE
PROLIFERATIVE
DIABETIC RETINOPATHY
DIABETIC RETINOPATHY
DIABETIC RETINOPATHY
(%)
NO.
{%)
NO.
(%)
NO.
M F M F M F M F
9 30 44 70 66 54 28 33
(50.0) (73.2) (62.9) (67.3) (61.7) (61.4) (66.7) (57.9)
9 11 21 29 36 27 10 14
(50.0) (26.8) (30.0) (27.9) (33.6) (30.7) (23.8) (24.6)
0 0 4 4 2 3 1 1
— — (5.7) (3.8) (1-9) (3.4) (2.4) (1.8)
0 0 1 1 1 3 1 5
M F
147 187
(62.0) (64.5)
76 81
(32.1) (27.9)
7 8
(2.9) (2.8)
3 9
poor control of diabetes seemed to have a high er prevalence as compared with those subjects with good control of diabetes. The association of prevalence of total retinopathy with large-vessel diseases, including stroke, ischemie heart disease, and leg vessel
UNGRADABLE RETINOPATHY
(%) — —
(%) — — — —
NO.
(1.4) (10) (09) (3.4) (2.4) (8.8)
0 0 0 0 2 1 2 4
(1.9) (1.1) (4.8) (7.0)
(1-3) (3.1)
4 5
(1.7) (1.7)
disease, was not statistically significant. How ever, the subjects with leg vessel disease had a higher prevalence of proliferative diabetic retinopathy than the subjects without leg vessel disease (22.2% and 2.0%, respectively; P < .05). The prevalence of total retinopathy
TABLE 2 PREVALENCE OF VARIOUS RETINOPATHY BY DURATION, TREATMENT, AND CONTROL STATUS OF DIABETES RETINOPATHY STATUS OF SUBJECTS BACKGROUND DIABETIC RETINOPATHY VARIABLES
Duration of diabetes (yrs) Newly diagnosed (N=187) < 4 (N=187) 5-9 (N=75) > 10(N=69) Type of diabetes treatment Diet control (N=164) Oral hypoglycιmie agent (N=318) Insulin (N=36) Control of diabetes' Good(N=124) Fair (N=72) Poor (N=322)
NO.
(%)
PREPROLIFERATIVE DIABETIC RETINOPATHY NO.
PROLIFERATIVE DIABETIC RETINOPATHY
(%)
NO.
(%)
TOTAL NO.
(%)
41 57 28 31
(21.9) (30.5) (37.3) (44.9)
1 3 6 5
(0.5) (1.6) (8.0) (7.3)
1 6 3 2
(0.5) (3.2) (4.0) (2.9)
43 66 37 38
(23.0)* (35.3) (49.3) (55.1)
42
(25.6)
2
(1.2)
3
(1.8)
47
(28.7)*
101 14
(31.8) (38.9)
9 4
(2.8) (11.1)
6 3
(1.9) (8.3)
116 21
(36.5) (58.3)
36 20 101
(29.0) (27.8) (31.4)
1 1 13
(0.8) (1.4) (4.0)
3 0 9
(2.4)
40 21 123
(32.3) (29.2) (38.2)
— (2.8)
*P < .01, chi-square test. 'Good, fair, or poor control defined as follows: fasting plasma glucose levels were £ 120 mg/dl, 121 to 140 mg/dl, and > 140 mg/dl, respectively; or two-hour postprandial levels were s 160 mg/dl, 161 to 180 mg/dl, and > 180 mg/dl, respectively.
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
seemed to be higher in the subjects with hyper tension (37.0%) and proteinuria (45.5%) than in the subjects without hypertension and pro teinuria (33.0% and 34.6%, respectively), though the difference was not statistically sig nificant. The subjects with retinopathy had a higher level of serum creatinine than those without retinopathy (mean ± standard deviation, 1.08 ± 0.40 m g / d l and 1.06 ± 0.34 m g / d l , respec tively; P < .05). However, the subjects with retinopathy had a lower level of serum choles terol than those without retinopathy (mean ± standard deviation, 199.6 ± 41.2 m g / d l and 191.0 ± 44.6 mg/dl, respectively; P < .05). The serum high-density lipoprotein cholesterol lev el was not associated with the prevalence of retinopathy. The duration of diabetes was the only factor significantly associated with retinopathy as an alyzed by a multiple logistic regression model. For every one-year increment in the duration of diabetes, the risk of retinopathy increased 1.07 times (P < .005). In other words, for a ten-year increase in diabetes duration, the relative risk of developing retinopathy was as high as 1.95. The insulin-treated diabetic subjects had a higher risk of developing retinopathy than those subjects treated with dietary control (rel ative risk, 1.57; .05 < P < .10) (Table 3). Regarding proliferative diabetic retinopathy, insulin-treated diabetic subjects also had a higher risk than subjects with diabetes con trolled by diet (relative risk, 2.51; .05 < P < .10) (Table 4).
727
Discussion Studies from Australia, 7,8 Denmark, 910 11 12 13 Iceland, Sweden, · and the United States 1416 showed a large difference in prevalence, which ranged from 24% to 70% for total dia betic retinopathy and from 4.5% to 22% for proliferative diabetic retinopathy. Our study showed a prevalence of 35.0% for diabetic reti nopathy and 2.2% for proliferative diabetic retinopathy in noninsulin-dependent diabetic subjects. The prevalence of proliferative diabet ic retinopathy in our study was compared to the data from the Western studies. One study 8 found prevalences of total diabetic retinopathy and proliferative diabetic retinopathy to be 24% and 4.5%, respectively, among adults; one study 10 found prevalences of 40.9% and 3.3%, respectively, among subjects treated with oral hypoglycémie agents; and one study 16 found prevalences of 62.0% and 10.0%, respectively, among subjects older than 30 years. In compar ison, the prevalence of proliferative diabetic retinopathy among Chinese adult patients with noninsulin-dependent diabetes was low, al though the total prevalence of 35% was not remarkably different. The duration of diabetes and the prevalence and severity of retinopathy were closely associ ated (Table 2). The longer the duration of diabe tes, the higher the prevalence of total retinopa thy. The prevalences of total and proliferative diabetic retinopathy were 29.1% and 1.9%,
TABLE 3 MULTIPLE LOGISTIC REGRESSION ANALYSIS OF FACTORS ASSOCIATED WITH PREVALENCE OF RETINOPATHY
FACTORS
COMPARISON
Duration of diabetes Age at onset
Every 1-yr increment Every 1-yr increment Diet vs oral hypoglycιmie agent Diet vs insulin Fair vs good Poor vs good Every 1-mg/dl increment Every 1-mg/dl decrement
Type of diabetes treatment Control of diabetes Serum creatinine Serum cholesterol
*N.S. indicates not significant.
PREVALENCE
95% CONFIDENCE
ODDS RATIO
INTERVAL
P VALUE*
1.07
1.02-1.12
< .005
1.00
0.98-1.03
N.S.
0.81 1.57 0.84 1.15
0.60-1.10 0.95-2.59 0.57-1.23 0.87-1.53
N.S. .05 < P < .1 N.S. N.S.
1.36
0.82-2.27
N.S.
1.00
0.99-1.00
N.S.
728
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
TABLE 4 MULTIPLE LOGISTIC REGRESSION ANALYSIS OF FACTORS ASSOCIATED WITH PREVALENCE OF PROLIFERATIVE DIABETIC RETINOPATHY
FACTORS
COMPARISON
Age at examination Age at onset
Every 1-yr increment Every 1-yr increment Diet vs oral hypoglycιmie agent Diet vs insulin Fair vs good Poor vs good Yes vs no
Type of diabetes treatment Control of diabetes Leg vessel disease
PREVALENCE
95% CONFIDENCE
ODDS RATIO
INTERVAL
P VALUE*
1.10
0.98-1.25
N.S.
0.99
0.89-1.10
N.S.
0.72 2.51 0.56 1.22 1.48
0.33-1.55 0.90-6.99 0.13-2.40 0.49-3.02 0.83-2.65
N.S. .05 < P < .1 N.S. N.S. N.S.
*N.S. indicates not significant.
respectively, for those subjects who had diabe tes for less than five years, and 23.0% and 0.5%, respectively, for subjects in whom diabe tes was newly diagnosed. These data empha sized that the onset of noninsulin-dependent diabetes was difficult to determine precisely, and apparently diagnosis was delayed in many patients. Similar results have also been empha sized in other studies.16·44 Some studies found that the duration of diabetes is the most crucial factor for both total and proliferative diabetic retinopathy. 23 · 25 However, our results showed that duration of diabetes was related to the prevalence of total retinopathy, but not to pro liferative diabetic retinopathy. The difficulty in determining the exact duration of diabetes and the small number of subjects with proliferative diabetic retinopathy explain at least in part the nonsignificant association between duration of diabetes and proliferative diabetic retinopathy. Similar to16·18·20·21 or in contrast to44 the results of other studies, our univariate analysis showed a higher prevalence of total and proliferative diabetic retinopathy in insulin-treated diabetic subjects. In Taiwan, it was difficult to persuade noninsulin-dependent diabetic patients who resisted treatment with oral hypoglycémie agents to consent to treatment with insulin. The subjects who were treated with insulin injection were reluctant to accept treatment with two or more doses a day. Thus, those subjects treated with insulin usually had a longer duration and poorer control of diabe tes.45 Whether good diabetic control reduces the development of diabetic retinopathy is still
controversial. Our cross-sectional study showed no statistically significant relationship between diabetic control and prevalence of reti nopathy, although our results did demonstrate that subjects with good control of diabetes seemed to have a lower prevalence of total retinopathy (32.2%) than those with poor con trol of diabetes (38.2%). The renal dysfunction, as indicated by in creased creatinine levels and proteinuria, has been shown to be associated with retinopa thy 16,20,23,26 ft0fa f a c t 0 rs are assumed to signify generalized microvascular disease in individu als with diabetes. Our data demonstrated that the prevalence of total retinopathy was corre lated with increased creatinine levels and was higher in diabetic subjects with proteinuria. But the associations were not statistically sig nificant after adjusting for duration of diabetes. Recently, more sensitive tests for measuring microalbuminuria showed a correlation to the development of retinopathy. 46 No association 30 or weak association 31 be tween serum cholesterol level and retinopathy was described, although few studies found pos itive association. 32,38 The association between serum cholesterol level and retinopathy was not statistically significant with a multivariateadjusted odds ratio of 1.0 in our investigation. There was no correlation between retinopathy and large-vessel diseases. However, the higher prevalence of proliferative diabetic retinopathy in subjects with leg vessel disease deserves further evaluation. In our series, proliferative diabetic retinopa thy was present in only 12 subjects. The preva-
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
lence o d d s r a t i o s for age at e x a m i n a t i o n , con trol of d i a b e t e s , a n d leg vessel d i s e a s e w e r e 1.10, 1.22, a n d 1.48, r e s p e c t i v e l y . T h e associa tions w e r e n o t statistically significant; h o w e v e r , they m i g h t b e c o m e significant if the s a m p l e size increased. O p h t h a l m o s c o p y is n o t as g o o d as f u n d u s p h o t o g r a p h y for t h e d e t e c t i o n of d i a b e t i c r e t i n o p a t h y . F u n d u s p h o t o g r a p h y can offer p e r m a n e n t r e c o r d s of r e t i n o p a t h y a n d a c c u r a t e l y d o c u m e n t s u b t l e c h a n g e s in r e t i n o p a t h y over time. 47 · 48 Because f u n d u s p h o t o g r a p h y could n o t be p e r f o r m e d in all subjects at the t i m e of o u r large field s t u d y , t h e u s e of o p h t h a l m o s c o p y m i g h t have affected the e s t i m a t e d p r e v a l e n c e of diabetic r e t i n o p a t h y to s o m e e x t e n t . Because ophthalmoscopy was performed by well-exper ienced ophthalmologists w h o did not k n o w the risk factors of d i a b e t i c r e t i n o p a t h y , t h e p r e v a lence o d d s ratios o b s e r v e d b e t w e e n risk factors and retinopathy were considered conservative estimates.
References 1. Entmacher, P. S., Root, H. F., and Marks, H. H.: Longevity of diabetic patients in recent years. Diabe tes 13:373, 1964. 2. Tsai, S. H.: Epidemiology of diabetes mellitus in Taiwan. In Tsuji, S., and Wada, M. (eds.): Diabetes Mellitus in Asia. Amsterdam, Excerpta Medica, 1971, pp. 64-67. 3. Tai, T. Y., Chen, C. Z., and Tsai, S. H.: Epidemi ology of diabetes mellitus among adults in Taipei, Taiwan. In Khalid, B. A. K., and Ng, M. L. (eds.): Proceedings of the Third Congress of Endocrinology. Kuala Lumpur, Asean Federation of Endocrine Socie ties, 1985, F17. 4. Tai, T. Y., Yang, C. L., Chang, C. J., Chang, S. M„ Chen, Y. H„ Lin, B. J., Ko, L. S., Chen, M. S., and Chen, C. J.: Epidemiology of diabetes mellitus among adults in Taiwan, R.O.C. J. Med. Assoc. Thai. 70(suppl. 2):42, 1987. 5. : Epidemiology of diabetes mellitus in Taiwan, R.O.C. Comparison between urban and rural areas. J. Med. Assoc. Thai. 70(suppl. 2):49, 1987. 6. Kahn, H. A., and Hiller, R.: Blindness caused by diabetic retinopathy. Am. J. Ophthalmol. 78:58, 1974. 7. Mitchell, P.: The prevalence of diabetic retinop athy. A study of 1300 diabetics from Newcastle and the Hunter Valley. Aust. J. Ophthalmol. 8:241, 1980. 8. Heriot, W. J., Borger, J. P., Zimmet, P., King, H., Taylor, R., and Raper, L. R.: Diabetic retinopathy in a natural population. Aust. J. Ophthalmol. 11:175, 1983.
729
9. Nielsen, N. V.: Diabetic retinopathy. I. The course of retinopathy in insulin-treated diabetics. A one year epidemiological cohort study of diabetes mellitus. The Island of Falster, Denmark. Acta Oph thalmol. 62:256, 1984. 10. : Diabetic retinopathy. II. The course of retinopathy in diabetics treated with oral hypoglycaemic agents and diet regime alone. A one year epidemiological cohort study of diabetes mellitus. The Island of Falster, Denmark. Acta Ophthalmol. 62:266, 1984. 11. Danielsen, R., Jónasson, F., and Helgason, T.: Prevalence of retinopathy and proteinuria in type 1 diabetics in Iceland. Acta Med. Scand. 212:277, 1982. 12. Jerneld, B., and Algvere, P.: The prevalence of retinopathy in insulin-dependent juvenile-onset dia betes mellitus. A fluorescein-angiographic study. Acta Ophthalmol. 62:617, 1984. 13. : Prevalence of retinopathy in diabetes treated with oral antihyperglycaemic agents. Acta Ophthalmol. 63:535, 1985. 14. Klein, R., Davis, M. D., Moss, S. E., Klein, B. E. K., and DeMets, D. L.: The Wisconsin epidemi ologie study of diabetic retinopathy. A comparison of retinopathy in younger and older onset diabetic per sons. In Vranic, M., Hollenberg, C. H., and Steiner, G. (eds.): Comparison of Type I and Type II Diabetes. New York, Plenum Press, 1985, pp. 321-335. 15. Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., and DeMets, D. L.: The Wisconsin epidemio logie study of diabetic retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagno sis is less than 30 years. Arch. Ophthalmol. 102:520, 1984. 16. : The Wisconsin epidemiologie study of diabetic retinopathy. III. Prevalence and risk of dia betic retinopathy when age at diagnosis is 30 or more years. Arch. Ophthalmol. 102:527, 1984. 17. Burditt, A. F., Caird, F. I„ and Draper, G. J.: The natural history of diabetic retinopathy. Q. J. Med. 37:303, 1968. 18. Miki, E., Fukuda, M., Kuzuya, T., Kosaka, K„ and Nakao, K.: Relation of the course of retinopathy to control of diabetes, age and therapeutic agents in diabetic Japanese patients. Diabetes 18:773, 1969. 19. Kahn, H. A., and Bradley, R. F.: Prevalence of diabetic retinopathy. Br. J. Ophthalmol. 59:345, 1975. 20. Knowler, W. C , Bennett, P. H., and Ballintine, E. J.: Increased incidence of retinopathy in diabetics with elevated blood pressure. N. Engl. J. Med. 302:645, 1980. 21. West, K. M., Erdreich, L. J., and Stober, J. A.: A detailed study of risk factors for retinopathy and nephropathy in diabetes. Diabetes 29:501, 1980. 22. Yanko, L., Goldbourt, U., Michaelson, I. C , Shapiro, A., and Yaari, S.: Prevalence and 15-year incidence of retinopathy and associated characteris tics in middle-aged and elderly diabetic men. Br. J. Ophthalmol. 67:759, 1983. 23. Constable, I. J., Knuiman, M. W., Welborn,
730
AMERICAN JOURNAL OF OPHTHALMOLOGY
T. A., Cooper, R. L., Stanton, K. M., McCann, V. J., and Grose, G. C : Assessing the risk of diabetic retinopathy. Am. J. Ophthalmol. 97:53, 1984. 24. Ishihara, M., Yukimura, Y„ Yamada, T„ Ohto, K., and Yoshizawa, K.: Diabetic complications and their relationships to risk factors in a Japanese popu lation. Diabetes Care 7:533, 1984. 25. Jerneld, B., and Algvere, P.: Relationship of duration and onset of diabetes to prevaiance of dia betic retinopathy. Am. J. Ophthalmol. 102:431,1986. 26. Kalter-Leibovici, O., van Dyk, D. J., Leibovici, L., Loya, N., Erman, A., Kremer, I., Boner, G., Rosenfeld, J. B., Karp, M., and Laron, Z.: Risk factors for development of diabetic nephropathy and retinopa thy in Jewish IDDM patients. Diabetes 40:204, 1991. 27. Palmberg, P., Smith, M , Waltman, S., Krupin, T., Singer, P., Burgess, D., Wendtlant, T., Achten berg, J., Cryer, P., Santiago, J., White, N., Kilo, C , and Daughaday, W.: The natural history of retinopa thy in insulin-dependent juvenile-onset diabetes. Ophthalmology 88:613, 1981. 28. Colwell, J. A.: Effect of diabetic control on retinopathy. Diabetes 15:497, 1966. 29. Dorf, A., Ballintine, E. ]., Bennett, P. H., and Miller, M.: Retinopathy in Pima Indians. Relation ships to glucose level, duration of diabetes, age at diagnpsis of diabetes, and age at examination in a population with a high prevalence of diabetes mellitus. Diabetes 25:554, 1976. 30. Szabo, A. J., Stewart, A. G., and Joron, G. E.: Factors associated with increased prevalence of dia betic retinopathy. A clinical survey. Can. Med. Assoc. J. 97:286, 1967. 31. Nilsson, S. E., Nilsson, J. E., Frostberg, N., and Emilsson, T.: The Kristianstad survey II. Studies in a representative adult diabetic population with special reference to comparison with an adequate control group. Acta Med. Scand. [Suppl.] 469:1, 1967. 32. Osuntokun, B. O.: Diabetic retinopathy in Ni gerians. A study of 758 patients. Br. J. Ophthalmol. 53:652, 1969. 33. Engerman, R., Bloodworth, J. M. B., and Nel son, S.: Relationship of microvascular disease in diabetes to metabolic control. Diabetes 26:760, 1977. 34. Raskin, P.: Diabetic regulation and its relation ship to microangiopathy. Metabolism 27:235, 1978. 35. Siperstein, M. D.: Diabetic microangiopathy and the control of blood glucose. N. Engl. J. Med. 309:1577, 1983.
December, 1992
36. Jerneld, B., and Algvere, P.: Proteinuria and blood glucose levels in a population with diabetic retinopathy. Am. J. Ophthalmol. 104:283, 1987. 37. Kornerup, T.: Blood pressure and diabetic reti nopathy. Acta Ophthalmol. 35:163, 1957. 38. Job, D., Tchobroutsky, G., Eschwege, E., Guyot-Argenton, C , Aubry, J. P., and Déret, M.: Relationships between diabetic retinopathy, blood pressure, body weight and serum lipids levels. Diabetologia 9:74, 1973. 39. World Health Organization: Diabetes mellitus. Report of a WHO study group. WHO Tech. Rep. Ser. 727:8, 1985. 40. Tai, T. Y., Chuang, L. M., Chen, C. J., and Lin, B. J.: Link between hypertension and diabetes melli tus. Epidemiological study of Chinese adults in Tai wan. Diabetes Care 14:1013, 1991. 41. Alberti, K. G. M. M., and Gries, F. A.: Manage ment of non-insulin-dependent diabetes mellitus in Europe. A consensus view. Diabetic Med. 5:275, 1988. 42. The World Health Organization Multinational Study of Vascular Disease in Diabetes: Prevalence of small vessel and large vessel disease in diabetic patients from 14 centers. Diabetologia 28:615, 1985. 43. Gross, F., Pisa, Z., Strasser, T., and Zanchetti, A.: Management of Arterial Hypertension. A Practi cal Guide for Physician and Allied Health Workers. Geneva, World Health Organization, 1984, p. 1. 44. Aiello, L. M., Rand, L. I., Briones, J. C , Wafai, M. Z., and Sebestyen, J. G.: Diabetic retinopathy in Joslin clinic patients with adult-onset diabetes. Ophthalmology 88:619, 1981. 45. Tai, T. Y., Tseng, C. H., Sung, S. M., Huang, R. F., Chen, C. J., and Tsai, S. H.: Retinopathy, neu ropathy and nephropathy in non-insulin-dependent diabetic patients. J. Form. Med. Assoc. 90:936, 1991. 46. Barnett, A. H., Dallinger, K., Jennings, P., Fletcher, J., and Odugbesan, O.: Microalbuminuria and diabetic retinopathy. Lancet 1:53, 1985. 47. Moss, S. E., Klein, R., Kessler, S. D., and Richie, K. A.: Comparison between ophthalmoscopy and fundus photography in determining severity of diabetic retinopathy. Ophthalmology 92:62, 1985. 48. Kinyoun, J. L., Martin, D. C , Fujimoto, W. Y., and Leonetti, D. L.: Ophthalmoscopy versus fundus photographs for detecting and grading diabetic reti nopathy. Invest. Ophthalmol. Vis. Sci. 33:1888, 1992.
Accepted for publication Sept. 17, 1992. From the Departments of Ophthalmology (Drs. M.-S. Chen and Kao), Public Health (Dr. C.-J. Chen), and Internal Medicine (Drs. Chang, Wu, Fu, and Tai), Col lege of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China. This study was supported by grant DOH75-0299-25 from the Department of Health, Executive Yuan, Taiwan, Republic of China (Dr. Tai). Reprint requests to Tong-Yuan Tai, M.D., No. 7, Chung-Shan South Road, Department of Internal Medi cine, College of Medicine, National Taiwan University, Taipei, Taiwan, Republic of China.
examination, older age at onset of diabetes, type of diabetes treatment, and presence of leg vessel disease. Insulin-treated diabetic subjects also had a higher risk of proliferative diabetic retinopathy than patients in whom diabetes was controlled by diet, with a relative risk of 2.51 (.05 < P < .10) in the multiple logistic regression analysis.
IN RECENT DECADES there has been an increase in the number of individuals with diabe tes. 1 Noninsulin-dependent diabetic patients represent a large portion of such patients. With the economic development in Taiwan during the past few decades, the prevalence of diabetes mellitus among those aged 40 years or older has increased markedly. In the past three surveys performed in 1970, 1979, and 1986 in Taipei City, the prevalence of diabetes was 5.05%, 7.10%, and 8.17%, respectively. 26 In the rural areas of Taiwan, a prevalence of 5.06% was noted during a survey performed simultaneous ly with that of Taipei City in 1986. Diabetic retinopathy has now become a major cause of loss in visual acuity in adults, which is often insidious in noninsulin-dependent diabetes. 6 Epidemiologie studies of diabetic retinopathy have been conducted among different popula tions in many areas of the world. The preva lence of diabetic retinopathy varies widely in different countries. 716 Despite different study procedures and popu lations, several factors, including duration of diabetes, 7 · 8 · 111432 age at the onset of diabe15 18 19 21 22 26 type t e s 7.11,16,17,21,25 a g e a t examination, ' · · · · of d i a b e t e s treatment, 16 · 18,20,21 c o n t r o l of d i a b e t e 8,15,16,18,20-24,26,28-31,33-36 h y p e r t e n s i o n 16.20,23,30,31,37,38
proteinuria,11·16'20,21'23'30,36 serum creatinine lev el,33 serum cholesterol level,32,38 and body mass index,13,16,21,23,81,38 have been found to correlate with the prevalence of diabetic retinopathy. In our population-based study, we investigated
©AMERICAN JOURNAL OF OPHTHALMOLOGY 114:723-730, DECEMBER, 1992
723
724
AMERICAN JOURNAL OF OPHTHALMOLOGY
the prevalence of diabetic retinopathy among noninsulin-dependent diabetic patients in Tai wan and determined the factors that correlated with the presence of retinopathy.
Subjects and Methods This survey was conducted from November 1985 to June 1986. Among the 78 subdistricts of Ta-An Precinct in Taipei City, eight subdistricts were randomly selected. Five villages evenly distributed over the island (Suan-Shi, Bar-Ter, Zu-Tan, Su-Fu, and Chih-Ku villages) were also randomly selected. The total number of subjects aged 40 years or older who underwent the diabetic survey in Taipei City was 4,272 (2,194 men and 2,078 women) and that of the five villages was 7,206 (3,582 men and 3,624 women), with a response rate of 65.3% and 72.0%, respectively. With the help of specially trained public health nurses, all subjects were interviewed according to a structured questionnaire. The questionnaire included details on age, gender, height, weight, residential area, monthly fami ly income, educational level, personal and fam ily history of diabetes, degree of physical ac tivity, frequency of exercise, and cigarette smoking. The capillary blood glucose of each partici pant was examined either after an overnight fast or two hours postprandially. Diabetes mellitus was diagnosed according to the World Health Organization criteria.39,40 Diagnosis was confirmed by a 75 -g oral glucose tolerance test in those patients with equivocal diagnostic val ues.40 By using these criteria, diabetes was diag nosed in 715 subjects, with a prevalence rate of 6.23%. Among the 715 subjects, 608 subjects, including 225 subjects in whom diabetes was newly diagnosed and 383 subjects in whom diabetes was previously diagnosed, participat ed in further systemic and laboratory examina tions. Among the 383 patients in whom diabe tes was previously diagnosed, 44 subjects were treated with insulin and 298 subjects were treated with sulfonylureas. Another 41 subjects treated by diet control only were checked for the blood glucose level criteria to confirm their diagnosis. Diabetic control was classified as good, fair, or poor when the fasting plasma glucose levels of diabetic patients were 120 m g / d l or less, 121 to 140 mg/dl, and greater than 140 m g / d l , respectively; or when the two-hour postprandi
December, 1992
al levels were 160 m g / d l or less, 161 to 180 m g / d l , and greater than 180 m g / d l , respective ly.41 Venous blood was collected for serum creatinine, cholesterol, and high-density lipoprotein cholesterol measurements in all subjects. Proteinuria was determined by commercial urine strips. According to the protocol of the World Health Organization Multinational Study on Vascular Disease in Diabetes, 42 large-vessel diseases, in cluding stroke, ischemie heart disease, and leg vessel disease, were investigated. Stroke was defined when neurologic sequelae were present or when unilateral weakness lasted for more than 24 hours. Ischemie heart disease was diag nosed by history and electrocardiographic ab normalities as designated by the Minnesota Code.42 Leg vessel disease was diagnosed by evidence of amputation or a history of intermit tent claudication. Amputation was recorded when the subject had a history of gangrene and necrotic or surgical loss of tissue in the toe, foot, or leg because of arterial obstruction judged by the investigators and that occurred after the onset of diabetes. Hypertension was deemed present when the systolic pressure was 160 mm Hg or greater, or when the diastolic pressure was 95 mm Hg or greater. 43 Body mass index was measured by dividing the weight in kilograms by the square of height in meters. Those subjects with a body mass index of 25 to 30 kg of body weight/m 2 were deemed overweight and those with an index over 30 kg of body weight/m 2 were deemed obese. Physical activity was divided into light (sedentary), moderate, and heavy (laborious) groups. The exercise status (number of times per week) was classified as less than two, two to four, and five or more times per week. The duration of exercise was at least 30 minutes per day. The monthly family income and personal educational level were also investigated. The details of the geographic distribution of study areas and study methods have been described previously. 4,5 Subjects who had not smoked more than 100 cigarettes in their lifetime were classified as nonsmokers; subjects who had smoked more than 100 cigarettes in their lifetime but had stopped smoking before the examination were classified as ex-smokers; and subjects who smoked at the time of the interview were classi fied as current smokers. The ophthalmic examinations were per formed with direct and indirect ophthalmoscopy after dilating the pupils with tropicamide
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
1% and phenylepinephrine 10%. The retinopa thy status was divided into the following five categories: (1) no diabetic retinopathy; (2) background diabetic retinopathy, which was defined as the presence of one or more of the following clinical signs: microaneurysms, punc tate or striate intraretinal hemorrhages, and hard exudates; (3) preproliferative diabetic retin opathy, which was defined as soft exudates, venous beading, and intraretinal microvascular abnormalities; (4) proliferative diabetic reti nopathy, characterized by neovascularization on or within one disk diameter of the disk in extent, neovascularization elsewhere in the ret ina, and fibrous proliferative tissue; and (5) ungradable retinopathy, with media opacities, including leukoma from old trachoma and dense cataracts that precluded gradation of retinopathy. Diabetic subjects were classified according to the most severe changes in the worse eye. Ophthalmoscopy was performed in 527 of the diabetic subjects. All diabetic subjects had noninsulin-dependent diabetes, characterized by the absent symptoms of uncon trolled diabetes, including polyuria, polydipsia, and an evident weight loss at the time of diagnosis, and had not been treated with insu lin during the first year after diagnosis. 4 ' 6,40 The characteristics of 527 participants and 81 nonparticipants were similar. For participants and nonparticipants, the proportions of male gen der were 45.0% and 46.9%, respectively; the mean ages ± standard deviation were 61.1 ± 9.3 years and 62.0 ± 8.9 years, respectively; the mean durations of diabetes ± standard devia tion were 3.7 ± 5.3 and 3.9 ± 5.6 years, respec tively; the percentages of subjects treated with insulin were 7.0% and 9.9%, respectively; the prevalences of hypertension were 31.7% and 35.8%, respectively; and the prevalences of proteinuria were 9.6% and 12.5%, respectively. For participants and nonparticipants, mean fasting plasma glucose levels ± standard devia tion were 160.0 ± 66.9 m g / d l (375 subjects) and 171.5 ± 79.4 m g / d l (61 subjects), respec tively, and mean two-hour postprandial levels ± standard deviation were 232.4 ± 79.0 m g / d l (186 subjects) and 212.4 ± 68.5 m g / d l (20 subjects), respectively. The differences in the previously mentioned characteristics were not significant. There were also no significant dif ferences for residential area, income, educa tional level, age at onset of diabetes, body mass index, physical activity, exercise, cigarette smoking, large-vessel disease, and serum levels of creatinine, cholesterol, and high-density lipoprotein cholesterol.
725
The statistical significance was assessed by using the Student's f-test and the chi-square test. Correlation coefficients were estimated by Spearman's method and multiple logistic re gression analysis was also performed to evalu ate multivariate-adjusted associations between related factors and diabetic retinopathy.
Results The retinopathy status of the 527 participat ing subjects (237 men and 290 women) was determined by age and gender (Table 1). Dia betic retinopathy was present in 184 of the 527 subjects (35.0%). Background diabetic retinop athy was present in 157 subjects (30.0%), pre proliferative diabetic retinopathy was present in 15 subjects (2.8%), and proliferative diabetic retinopathy was present in 12 subjects (2.2%). No diabetic retinopathy was present in 334 subjects (63.3%) and retinopathy was ungrad able in nine subjects (1.7%). The relationship between the prevalence of retinopathy and sociodemographic characteris tics was investigated. The difference in preva lence was not statistically significant between men and women. The prevalence of retinopathy was not statistically different when the subjects were divided according to age (40 to 49 years, 50 to 59 years, 60 to 69 years, and 70 years or older). However, the oldest diabetic subjects had the highest prevalence of proliferative dia betic retinopathy (P < .05 on the basis of a chi-square test). Residential area, monthly fam ily income, and educational level were not significantly associated with retinopathy. Family history of diabetes, body mass index, degree of physical activity, frequency of exer cise, and cigarette smoking were not associated with the prevalence of either total retinopathy or proliferative diabetic retinopathy. Subjects with a longer duration of diabetes had a higher percentage of total retinopathy (Table 2) (P < .01 on the basis of a chi-square test). The prevalence of retinopathy was different among subjects with different types of diabetic treatment (P < .01). The subjects treated with insulin had a higher prevalence of total retinop athy (58.3%) and proliferative diabetic retinop athy (8.3%) than subjects not treated with insu lin (33.8% and 1.9%, respectively) (Table 2). The control of diabetes and the prevalence of total and proliferative diabetic retinopathy were not associated (Table 2). However, for all categories of retinopathy, those subjects with
726
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
TABLE 1 PREVALENCE OF RETINOPATHY STATUS BY AGE AT EXAMINATION AND GENDER IN DIABETIC SUBJECTS RETINOPATHY STATUS OF SUBJECTS
NO DIABETIC RETINOPATHY
AGE AT EXAMINATION
(vns)
40-49 50-59 60-69 70 years or older Total
GENDER
NO.
BACKGROUND
PREPROLIFERATIVE
PROLIFERATIVE
DIABETIC RETINOPATHY
DIABETIC RETINOPATHY
DIABETIC RETINOPATHY
(%)
NO.
{%)
NO.
(%)
NO.
M F M F M F M F
9 30 44 70 66 54 28 33
(50.0) (73.2) (62.9) (67.3) (61.7) (61.4) (66.7) (57.9)
9 11 21 29 36 27 10 14
(50.0) (26.8) (30.0) (27.9) (33.6) (30.7) (23.8) (24.6)
0 0 4 4 2 3 1 1
— — (5.7) (3.8) (1-9) (3.4) (2.4) (1.8)
0 0 1 1 1 3 1 5
M F
147 187
(62.0) (64.5)
76 81
(32.1) (27.9)
7 8
(2.9) (2.8)
3 9
poor control of diabetes seemed to have a high er prevalence as compared with those subjects with good control of diabetes. The association of prevalence of total retinopathy with large-vessel diseases, including stroke, ischemie heart disease, and leg vessel
UNGRADABLE RETINOPATHY
(%) — —
(%) — — — —
NO.
(1.4) (10) (09) (3.4) (2.4) (8.8)
0 0 0 0 2 1 2 4
(1.9) (1.1) (4.8) (7.0)
(1-3) (3.1)
4 5
(1.7) (1.7)
disease, was not statistically significant. How ever, the subjects with leg vessel disease had a higher prevalence of proliferative diabetic retinopathy than the subjects without leg vessel disease (22.2% and 2.0%, respectively; P < .05). The prevalence of total retinopathy
TABLE 2 PREVALENCE OF VARIOUS RETINOPATHY BY DURATION, TREATMENT, AND CONTROL STATUS OF DIABETES RETINOPATHY STATUS OF SUBJECTS BACKGROUND DIABETIC RETINOPATHY VARIABLES
Duration of diabetes (yrs) Newly diagnosed (N=187) < 4 (N=187) 5-9 (N=75) > 10(N=69) Type of diabetes treatment Diet control (N=164) Oral hypoglycιmie agent (N=318) Insulin (N=36) Control of diabetes' Good(N=124) Fair (N=72) Poor (N=322)
NO.
(%)
PREPROLIFERATIVE DIABETIC RETINOPATHY NO.
PROLIFERATIVE DIABETIC RETINOPATHY
(%)
NO.
(%)
TOTAL NO.
(%)
41 57 28 31
(21.9) (30.5) (37.3) (44.9)
1 3 6 5
(0.5) (1.6) (8.0) (7.3)
1 6 3 2
(0.5) (3.2) (4.0) (2.9)
43 66 37 38
(23.0)* (35.3) (49.3) (55.1)
42
(25.6)
2
(1.2)
3
(1.8)
47
(28.7)*
101 14
(31.8) (38.9)
9 4
(2.8) (11.1)
6 3
(1.9) (8.3)
116 21
(36.5) (58.3)
36 20 101
(29.0) (27.8) (31.4)
1 1 13
(0.8) (1.4) (4.0)
3 0 9
(2.4)
40 21 123
(32.3) (29.2) (38.2)
— (2.8)
*P < .01, chi-square test. 'Good, fair, or poor control defined as follows: fasting plasma glucose levels were £ 120 mg/dl, 121 to 140 mg/dl, and > 140 mg/dl, respectively; or two-hour postprandial levels were s 160 mg/dl, 161 to 180 mg/dl, and > 180 mg/dl, respectively.
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
seemed to be higher in the subjects with hyper tension (37.0%) and proteinuria (45.5%) than in the subjects without hypertension and pro teinuria (33.0% and 34.6%, respectively), though the difference was not statistically sig nificant. The subjects with retinopathy had a higher level of serum creatinine than those without retinopathy (mean ± standard deviation, 1.08 ± 0.40 m g / d l and 1.06 ± 0.34 m g / d l , respec tively; P < .05). However, the subjects with retinopathy had a lower level of serum choles terol than those without retinopathy (mean ± standard deviation, 199.6 ± 41.2 m g / d l and 191.0 ± 44.6 mg/dl, respectively; P < .05). The serum high-density lipoprotein cholesterol lev el was not associated with the prevalence of retinopathy. The duration of diabetes was the only factor significantly associated with retinopathy as an alyzed by a multiple logistic regression model. For every one-year increment in the duration of diabetes, the risk of retinopathy increased 1.07 times (P < .005). In other words, for a ten-year increase in diabetes duration, the relative risk of developing retinopathy was as high as 1.95. The insulin-treated diabetic subjects had a higher risk of developing retinopathy than those subjects treated with dietary control (rel ative risk, 1.57; .05 < P < .10) (Table 3). Regarding proliferative diabetic retinopathy, insulin-treated diabetic subjects also had a higher risk than subjects with diabetes con trolled by diet (relative risk, 2.51; .05 < P < .10) (Table 4).
727
Discussion Studies from Australia, 7,8 Denmark, 910 11 12 13 Iceland, Sweden, · and the United States 1416 showed a large difference in prevalence, which ranged from 24% to 70% for total dia betic retinopathy and from 4.5% to 22% for proliferative diabetic retinopathy. Our study showed a prevalence of 35.0% for diabetic reti nopathy and 2.2% for proliferative diabetic retinopathy in noninsulin-dependent diabetic subjects. The prevalence of proliferative diabet ic retinopathy in our study was compared to the data from the Western studies. One study 8 found prevalences of total diabetic retinopathy and proliferative diabetic retinopathy to be 24% and 4.5%, respectively, among adults; one study 10 found prevalences of 40.9% and 3.3%, respectively, among subjects treated with oral hypoglycémie agents; and one study 16 found prevalences of 62.0% and 10.0%, respectively, among subjects older than 30 years. In compar ison, the prevalence of proliferative diabetic retinopathy among Chinese adult patients with noninsulin-dependent diabetes was low, al though the total prevalence of 35% was not remarkably different. The duration of diabetes and the prevalence and severity of retinopathy were closely associ ated (Table 2). The longer the duration of diabe tes, the higher the prevalence of total retinopa thy. The prevalences of total and proliferative diabetic retinopathy were 29.1% and 1.9%,
TABLE 3 MULTIPLE LOGISTIC REGRESSION ANALYSIS OF FACTORS ASSOCIATED WITH PREVALENCE OF RETINOPATHY
FACTORS
COMPARISON
Duration of diabetes Age at onset
Every 1-yr increment Every 1-yr increment Diet vs oral hypoglycιmie agent Diet vs insulin Fair vs good Poor vs good Every 1-mg/dl increment Every 1-mg/dl decrement
Type of diabetes treatment Control of diabetes Serum creatinine Serum cholesterol
*N.S. indicates not significant.
PREVALENCE
95% CONFIDENCE
ODDS RATIO
INTERVAL
P VALUE*
1.07
1.02-1.12
< .005
1.00
0.98-1.03
N.S.
0.81 1.57 0.84 1.15
0.60-1.10 0.95-2.59 0.57-1.23 0.87-1.53
N.S. .05 < P < .1 N.S. N.S.
1.36
0.82-2.27
N.S.
1.00
0.99-1.00
N.S.
728
December, 1992
AMERICAN JOURNAL OF OPHTHALMOLOGY
TABLE 4 MULTIPLE LOGISTIC REGRESSION ANALYSIS OF FACTORS ASSOCIATED WITH PREVALENCE OF PROLIFERATIVE DIABETIC RETINOPATHY
FACTORS
COMPARISON
Age at examination Age at onset
Every 1-yr increment Every 1-yr increment Diet vs oral hypoglycιmie agent Diet vs insulin Fair vs good Poor vs good Yes vs no
Type of diabetes treatment Control of diabetes Leg vessel disease
PREVALENCE
95% CONFIDENCE
ODDS RATIO
INTERVAL
P VALUE*
1.10
0.98-1.25
N.S.
0.99
0.89-1.10
N.S.
0.72 2.51 0.56 1.22 1.48
0.33-1.55 0.90-6.99 0.13-2.40 0.49-3.02 0.83-2.65
N.S. .05 < P < .1 N.S. N.S. N.S.
*N.S. indicates not significant.
respectively, for those subjects who had diabe tes for less than five years, and 23.0% and 0.5%, respectively, for subjects in whom diabe tes was newly diagnosed. These data empha sized that the onset of noninsulin-dependent diabetes was difficult to determine precisely, and apparently diagnosis was delayed in many patients. Similar results have also been empha sized in other studies.16·44 Some studies found that the duration of diabetes is the most crucial factor for both total and proliferative diabetic retinopathy. 23 · 25 However, our results showed that duration of diabetes was related to the prevalence of total retinopathy, but not to pro liferative diabetic retinopathy. The difficulty in determining the exact duration of diabetes and the small number of subjects with proliferative diabetic retinopathy explain at least in part the nonsignificant association between duration of diabetes and proliferative diabetic retinopathy. Similar to16·18·20·21 or in contrast to44 the results of other studies, our univariate analysis showed a higher prevalence of total and proliferative diabetic retinopathy in insulin-treated diabetic subjects. In Taiwan, it was difficult to persuade noninsulin-dependent diabetic patients who resisted treatment with oral hypoglycémie agents to consent to treatment with insulin. The subjects who were treated with insulin injection were reluctant to accept treatment with two or more doses a day. Thus, those subjects treated with insulin usually had a longer duration and poorer control of diabe tes.45 Whether good diabetic control reduces the development of diabetic retinopathy is still
controversial. Our cross-sectional study showed no statistically significant relationship between diabetic control and prevalence of reti nopathy, although our results did demonstrate that subjects with good control of diabetes seemed to have a lower prevalence of total retinopathy (32.2%) than those with poor con trol of diabetes (38.2%). The renal dysfunction, as indicated by in creased creatinine levels and proteinuria, has been shown to be associated with retinopa thy 16,20,23,26 ft0fa f a c t 0 rs are assumed to signify generalized microvascular disease in individu als with diabetes. Our data demonstrated that the prevalence of total retinopathy was corre lated with increased creatinine levels and was higher in diabetic subjects with proteinuria. But the associations were not statistically sig nificant after adjusting for duration of diabetes. Recently, more sensitive tests for measuring microalbuminuria showed a correlation to the development of retinopathy. 46 No association 30 or weak association 31 be tween serum cholesterol level and retinopathy was described, although few studies found pos itive association. 32,38 The association between serum cholesterol level and retinopathy was not statistically significant with a multivariateadjusted odds ratio of 1.0 in our investigation. There was no correlation between retinopathy and large-vessel diseases. However, the higher prevalence of proliferative diabetic retinopathy in subjects with leg vessel disease deserves further evaluation. In our series, proliferative diabetic retinopa thy was present in only 12 subjects. The preva-
Vol. 114, No. 6
Diabetic Retinopathy and Noninsulin-dependent Diabetes
lence o d d s r a t i o s for age at e x a m i n a t i o n , con trol of d i a b e t e s , a n d leg vessel d i s e a s e w e r e 1.10, 1.22, a n d 1.48, r e s p e c t i v e l y . T h e associa tions w e r e n o t statistically significant; h o w e v e r , they m i g h t b e c o m e significant if the s a m p l e size increased. O p h t h a l m o s c o p y is n o t as g o o d as f u n d u s p h o t o g r a p h y for t h e d e t e c t i o n of d i a b e t i c r e t i n o p a t h y . F u n d u s p h o t o g r a p h y can offer p e r m a n e n t r e c o r d s of r e t i n o p a t h y a n d a c c u r a t e l y d o c u m e n t s u b t l e c h a n g e s in r e t i n o p a t h y over time. 47 · 48 Because f u n d u s p h o t o g r a p h y could n o t be p e r f o r m e d in all subjects at the t i m e of o u r large field s t u d y , t h e u s e of o p h t h a l m o s c o p y m i g h t have affected the e s t i m a t e d p r e v a l e n c e of diabetic r e t i n o p a t h y to s o m e e x t e n t . Because ophthalmoscopy was performed by well-exper ienced ophthalmologists w h o did not k n o w the risk factors of d i a b e t i c r e t i n o p a t h y , t h e p r e v a lence o d d s ratios o b s e r v e d b e t w e e n risk factors and retinopathy were considered conservative estimates.
References 1. Entmacher, P. S., Root, H. F., and Marks, H. H.: Longevity of diabetic patients in recent years. Diabe tes 13:373, 1964. 2. Tsai, S. H.: Epidemiology of diabetes mellitus in Taiwan. In Tsuji, S., and Wada, M. (eds.): Diabetes Mellitus in Asia. Amsterdam, Excerpta Medica, 1971, pp. 64-67. 3. Tai, T. Y., Chen, C. Z., and Tsai, S. H.: Epidemi ology of diabetes mellitus among adults in Taipei, Taiwan. In Khalid, B. A. K., and Ng, M. L. (eds.): Proceedings of the Third Congress of Endocrinology. Kuala Lumpur, Asean Federation of Endocrine Socie ties, 1985, F17. 4. Tai, T. Y., Yang, C. L., Chang, C. J., Chang, S. M„ Chen, Y. H„ Lin, B. J., Ko, L. S., Chen, M. S., and Chen, C. J.: Epidemiology of diabetes mellitus among adults in Taiwan, R.O.C. J. Med. Assoc. Thai. 70(suppl. 2):42, 1987. 5. : Epidemiology of diabetes mellitus in Taiwan, R.O.C. Comparison between urban and rural areas. J. Med. Assoc. Thai. 70(suppl. 2):49, 1987. 6. Kahn, H. A., and Hiller, R.: Blindness caused by diabetic retinopathy. Am. J. Ophthalmol. 78:58, 1974. 7. Mitchell, P.: The prevalence of diabetic retinop athy. A study of 1300 diabetics from Newcastle and the Hunter Valley. Aust. J. Ophthalmol. 8:241, 1980. 8. Heriot, W. J., Borger, J. P., Zimmet, P., King, H., Taylor, R., and Raper, L. R.: Diabetic retinopathy in a natural population. Aust. J. Ophthalmol. 11:175, 1983.
729
9. Nielsen, N. V.: Diabetic retinopathy. I. The course of retinopathy in insulin-treated diabetics. A one year epidemiological cohort study of diabetes mellitus. The Island of Falster, Denmark. Acta Oph thalmol. 62:256, 1984. 10. : Diabetic retinopathy. II. The course of retinopathy in diabetics treated with oral hypoglycaemic agents and diet regime alone. A one year epidemiological cohort study of diabetes mellitus. The Island of Falster, Denmark. Acta Ophthalmol. 62:266, 1984. 11. Danielsen, R., Jónasson, F., and Helgason, T.: Prevalence of retinopathy and proteinuria in type 1 diabetics in Iceland. Acta Med. Scand. 212:277, 1982. 12. Jerneld, B., and Algvere, P.: The prevalence of retinopathy in insulin-dependent juvenile-onset dia betes mellitus. A fluorescein-angiographic study. Acta Ophthalmol. 62:617, 1984. 13. : Prevalence of retinopathy in diabetes treated with oral antihyperglycaemic agents. Acta Ophthalmol. 63:535, 1985. 14. Klein, R., Davis, M. D., Moss, S. E., Klein, B. E. K., and DeMets, D. L.: The Wisconsin epidemi ologie study of diabetic retinopathy. A comparison of retinopathy in younger and older onset diabetic per sons. In Vranic, M., Hollenberg, C. H., and Steiner, G. (eds.): Comparison of Type I and Type II Diabetes. New York, Plenum Press, 1985, pp. 321-335. 15. Klein, R., Klein, B. E. K., Moss, S. E., Davis, M. D., and DeMets, D. L.: The Wisconsin epidemio logie study of diabetic retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagno sis is less than 30 years. Arch. Ophthalmol. 102:520, 1984. 16. : The Wisconsin epidemiologie study of diabetic retinopathy. III. Prevalence and risk of dia betic retinopathy when age at diagnosis is 30 or more years. Arch. Ophthalmol. 102:527, 1984. 17. Burditt, A. F., Caird, F. I„ and Draper, G. J.: The natural history of diabetic retinopathy. Q. J. Med. 37:303, 1968. 18. Miki, E., Fukuda, M., Kuzuya, T., Kosaka, K„ and Nakao, K.: Relation of the course of retinopathy to control of diabetes, age and therapeutic agents in diabetic Japanese patients. Diabetes 18:773, 1969. 19. Kahn, H. A., and Bradley, R. F.: Prevalence of diabetic retinopathy. Br. J. Ophthalmol. 59:345, 1975. 20. Knowler, W. C , Bennett, P. H., and Ballintine, E. J.: Increased incidence of retinopathy in diabetics with elevated blood pressure. N. Engl. J. Med. 302:645, 1980. 21. West, K. M., Erdreich, L. J., and Stober, J. A.: A detailed study of risk factors for retinopathy and nephropathy in diabetes. Diabetes 29:501, 1980. 22. Yanko, L., Goldbourt, U., Michaelson, I. C , Shapiro, A., and Yaari, S.: Prevalence and 15-year incidence of retinopathy and associated characteris tics in middle-aged and elderly diabetic men. Br. J. Ophthalmol. 67:759, 1983. 23. Constable, I. J., Knuiman, M. W., Welborn,
730
AMERICAN JOURNAL OF OPHTHALMOLOGY
T. A., Cooper, R. L., Stanton, K. M., McCann, V. J., and Grose, G. C : Assessing the risk of diabetic retinopathy. Am. J. Ophthalmol. 97:53, 1984. 24. Ishihara, M., Yukimura, Y„ Yamada, T„ Ohto, K., and Yoshizawa, K.: Diabetic complications and their relationships to risk factors in a Japanese popu lation. Diabetes Care 7:533, 1984. 25. Jerneld, B., and Algvere, P.: Relationship of duration and onset of diabetes to prevaiance of dia betic retinopathy. Am. J. Ophthalmol. 102:431,1986. 26. Kalter-Leibovici, O., van Dyk, D. J., Leibovici, L., Loya, N., Erman, A., Kremer, I., Boner, G., Rosenfeld, J. B., Karp, M., and Laron, Z.: Risk factors for development of diabetic nephropathy and retinopa thy in Jewish IDDM patients. Diabetes 40:204, 1991. 27. Palmberg, P., Smith, M , Waltman, S., Krupin, T., Singer, P., Burgess, D., Wendtlant, T., Achten berg, J., Cryer, P., Santiago, J., White, N., Kilo, C , and Daughaday, W.: The natural history of retinopa thy in insulin-dependent juvenile-onset diabetes. Ophthalmology 88:613, 1981. 28. Colwell, J. A.: Effect of diabetic control on retinopathy. Diabetes 15:497, 1966. 29. Dorf, A., Ballintine, E. ]., Bennett, P. H., and Miller, M.: Retinopathy in Pima Indians. Relation ships to glucose level, duration of diabetes, age at diagnpsis of diabetes, and age at examination in a population with a high prevalence of diabetes mellitus. Diabetes 25:554, 1976. 30. Szabo, A. J., Stewart, A. G., and Joron, G. E.: Factors associated with increased prevalence of dia betic retinopathy. A clinical survey. Can. Med. Assoc. J. 97:286, 1967. 31. Nilsson, S. E., Nilsson, J. E., Frostberg, N., and Emilsson, T.: The Kristianstad survey II. Studies in a representative adult diabetic population with special reference to comparison with an adequate control group. Acta Med. Scand. [Suppl.] 469:1, 1967. 32. Osuntokun, B. O.: Diabetic retinopathy in Ni gerians. A study of 758 patients. Br. J. Ophthalmol. 53:652, 1969. 33. Engerman, R., Bloodworth, J. M. B., and Nel son, S.: Relationship of microvascular disease in diabetes to metabolic control. Diabetes 26:760, 1977. 34. Raskin, P.: Diabetic regulation and its relation ship to microangiopathy. Metabolism 27:235, 1978. 35. Siperstein, M. D.: Diabetic microangiopathy and the control of blood glucose. N. Engl. J. Med. 309:1577, 1983.
December, 1992
36. Jerneld, B., and Algvere, P.: Proteinuria and blood glucose levels in a population with diabetic retinopathy. Am. J. Ophthalmol. 104:283, 1987. 37. Kornerup, T.: Blood pressure and diabetic reti nopathy. Acta Ophthalmol. 35:163, 1957. 38. Job, D., Tchobroutsky, G., Eschwege, E., Guyot-Argenton, C , Aubry, J. P., and Déret, M.: Relationships between diabetic retinopathy, blood pressure, body weight and serum lipids levels. Diabetologia 9:74, 1973. 39. World Health Organization: Diabetes mellitus. Report of a WHO study group. WHO Tech. Rep. Ser. 727:8, 1985. 40. Tai, T. Y., Chuang, L. M., Chen, C. J., and Lin, B. J.: Link between hypertension and diabetes melli tus. Epidemiological study of Chinese adults in Tai wan. Diabetes Care 14:1013, 1991. 41. Alberti, K. G. M. M., and Gries, F. A.: Manage ment of non-insulin-dependent diabetes mellitus in Europe. A consensus view. Diabetic Med. 5:275, 1988. 42. The World Health Organization Multinational Study of Vascular Disease in Diabetes: Prevalence of small vessel and large vessel disease in diabetic patients from 14 centers. Diabetologia 28:615, 1985. 43. Gross, F., Pisa, Z., Strasser, T., and Zanchetti, A.: Management of Arterial Hypertension. A Practi cal Guide for Physician and Allied Health Workers. Geneva, World Health Organization, 1984, p. 1. 44. Aiello, L. M., Rand, L. I., Briones, J. C , Wafai, M. Z., and Sebestyen, J. G.: Diabetic retinopathy in Joslin clinic patients with adult-onset diabetes. Ophthalmology 88:619, 1981. 45. Tai, T. Y., Tseng, C. H., Sung, S. M., Huang, R. F., Chen, C. J., and Tsai, S. H.: Retinopathy, neu ropathy and nephropathy in non-insulin-dependent diabetic patients. J. Form. Med. Assoc. 90:936, 1991. 46. Barnett, A. H., Dallinger, K., Jennings, P., Fletcher, J., and Odugbesan, O.: Microalbuminuria and diabetic retinopathy. Lancet 1:53, 1985. 47. Moss, S. E., Klein, R., Kessler, S. D., and Richie, K. A.: Comparison between ophthalmoscopy and fundus photography in determining severity of diabetic retinopathy. Ophthalmology 92:62, 1985. 48. Kinyoun, J. L., Martin, D. C , Fujimoto, W. Y., and Leonetti, D. L.: Ophthalmoscopy versus fundus photographs for detecting and grading diabetic reti nopathy. Invest. Ophthalmol. Vis. Sci. 33:1888, 1992.
