Diabetes Research and Clinical Practice, 10 (1990)
251-263
257
Elsevier DIABET
00447
Development of diabetic retinopathy and its associated risk factors in type 2 diabetic patients in Osaka district, Japan: a long-term prospective study Akira Sasaki, Naruto
Horiuchi,
Kyoichi Hasewgawa
and Masuko
Uehara
Osaka Medical Center for Adult Diseases. Osaka, Japan
(Received 11 May 1990) (Revision received 3 July 1990) (Accepted 14 July 1990)
Summary Incidence rates of diabetic retinopathy and its associated risk factors were studied in a long-term prospective study involving 976 type 2 (non-insulin dependent) diabetic patients that showed no retinopathy at entry to the study. 322 of the patients (one-third of the subjects studied) developed diabetic retinopathy during the observation period (average length 8.3 years). The incidence rate per 1000 personyears was 39.8. The rate was significantly related to age at onset of diabetes, to fasting plasma glucose (FPG) level at entry, and to type of treatment. The incidence rate was also related to the duration of the disease. Glycemic control clearly played a role in the occurrence of retinopathy. Patients with FPG 3 200 mg/dl had the highest incidence rate, while patients with FPG < 140 mg/dl showed the lowest incidence rate throughout the observation period. Furthermore, a longer period between onset of diabetes and development of retinopathy was observed in patients with FPG < 140 mg/dl compared to patients with FPG > 200 mg/dl. The findings suggest that strict glycemic control may be of particular value to reduce the incidence of retinopathy and to delay its appearance in type 2 diabetic patients. Key words: Diabetic retinopathy
; Risk factor; Type 2 diabetes ; Prospective
Introduction Previous studies concerning the development of diabetic retinopathy pointed out the relation to duration of diabetes [l-8] and poor glycemic control as major risk factors [ 5-7,9-151, whereas ____ Address for correspondence: Professor A. Sasaki, Osaka Prefectural College of Nursing, 1-41 Tezukayama-Higashi 2 Chome, Sumiyoshi, Osaka, Japan. 016%8227/90/$03.50
0 1990 Elsevier Science Publishers
study
the role of hypertension, obesity, elevated lipid levels, impaired insulin response and cigarette smoking in the development of diabetic retinopathy are less certain. We found previously, in a retrospective study, that glycemic control and duration of diabetes played important roles as risk factors for diabetic retinopathy in type 2 diabetic patients [ 161. However, retrospective studies have certain shortcomings, so a prospective study with a long-term observation period
B.V. (Biomedical
Division)
258
was required for determining the natural history of diabetic retinopathy in Japan, while such studies have already been reported in the West [3,5,6,12,13]. The present study was carried out using a long-term prospective design for determining the incidence rate of diabetic retinopathy and its associated risk factors in type 2 diabetic patients in Japan, who showed no retinopathy upon entry to the study.
Subjects and methods Subjects The subjects studied were 976 type 2 (non-insulin dependent) diabetic patients (608 males and 368 females) who were first seen at the Osaka Medical Center for Adult Diseases, Osaka, Japan, during the period 1960-l 979. They were followed until the end of 1984. None of them showed retinopathy at entry, and they all had fasting plasma glucose levels >, 140 mg/dl or a 2 h level of 50 g OGTT > 180 mg at entry. Mean age ( k SD) was 52.1 + 10.9 years, and mean duration of diabetes was 3.0 + 4.6 years at entry. Methods Diabetic retinopathy was diagnosed by ocular fundus test performed by an ophthalmologist at least once a year as a rule. Glucose level was determined by the ferricyanide method using venous plasma. Albuminuria was measured by Albustix method. Glycemic control at entry was defined as the mean value of fasting plasma glucose (FPG) levels measured during the first year after the initiation of treatment. Obesity was expressed as percent of actual body weight divided by the standard body weight for Japanese adults [ 171. Classification of ECG findings were based on the criteria of the Whitehall Study [ 181 in which abnormalities in Q/QS and ST/T, LBBB and positive exercise test were defined as ischemic change. Incidence rate of diabetic retinopathy per 1000 person-years was computed according to the factors presented at entry to the study, since
observation periods differed from patient to patient. Life-table method was used for the computation of cumulative incidence rate. Chi-square test and Student’s t-test were used for statistical analysis.
Results Development of diabetic retinopathy and related risk factors 322 cases (i.e., one-third of the subjects studied) developed diabetic retinopathy during the observation period, which was 8.3 years on average. Table 1 shows the incidence rates of diabetic retinopathy per 1000 person-years, according to the factors present at the start of the follow-up period. The overall incidence rate of retinopathy was 39.8 per 1000 person-years, 38.2 for males and 42.5 for females, with no significant difference between the sexes. The rate was related significantly to the age at onset of diabetes, with higher incidence rate for younger age at onset. The rate was also related significantly to the FPG at entry, with a sharp increase as the FPG level rose above 110 mg/dl. The incidence of retinopathy was also related to the type of treatment; the incidence rate was significantly higher in groups treated with the oral hypoglycemic agent or with insulin compared with the diet-treated group. However, systolic blood pressure, ischemic ECG changes, obesity, serum cholesterol, albuminuria and cigarette smoking showed no significant relation to the incidence of retinopathy. The relation between incidence rate of retinopathy and the length of the observation period was analysed. The rates were 25.08 for patients with observation periods less than 5 years, 39.52 for patients with observation periods between 5-9years, 48.39 for observation periods of lo-14 years, and 58.38 for observation periods >, 15 years. Fig. 1 shows those relations with respect to FPG levels. The rates were affected signif?cantly by the FPG levels. It is noteworthy that patients with FPG 3 200 mg/dl always had the highest incidence rates regardless of the length
259 TABLE 1 Incidence
rate per 1000 person-years
of diabetic retinopathy
by factors at the baseline
Factors at baseline
No. of
subjects
Person-years observed
Cases with retinopathy
Incidence” rate
All subjects
916
8092.9
322
39.8
Sex
males females
608 368
5177.5 2915.4
198 124
38.2 42.5
Age at onset
t45 45 - 54 55 - 55 65 +
333 328 246 69
2883.0 2662.5 2012.0 535.6
130 115 62 15
45.1* 43.2 30.8 28.0
< 140 140 - 159 160+
522 304 134
4352.2 2548.1 1025.2
166 107 42
38.1 42.0 41.0
Fasting glucose level (mg/dl)
200 mg/dl) always showed the highest incidence rates of
262
retinopathy, regardless of the length of the observation period, whereas those patients that showed good glycemic control (FPG levels at entry incidence < 140 mg/dl) had lower rates throughout the observation period. The influence of initial FPG level on the cumulative incidence rate for retinopathy was maintained during the observation period. Thus, our findings support many recent studies on relations between retinopathy and fasting glucose level or glycated hemoglobin level, emphasizing the importance of glycemic control for reducing the risk of retinopathy [5,7,9-151. Relations of retinopathy to blood pressure [2,4,10,13,19] or to cigarette smoking [ 14,201 were, however, not apparent in the present study. It is noteworthy that the cumulative incidence rate of retinopathy in patients with FPG level > 200 mg/dl at entry seemed to reach a plateau at about 70% at the end of the follow-up period (Fig. 2). This may imply that some type 2 diabetic patients are resistent to the development of retinopathy, in spite of both a long duration of diabetes and poor glycemic control. Perhaps predispositional or genetic factors are responsible for such differences [ 161. The period from onset of diabetes to development of retinopathy was estimated as 9.2 years on average in the present study, indicating that a considerable time was required for development of retinopathy after the onset of diabetes. Nathan et al. [ 71 reported that the incidence of retinopathy increased 10 years after the onset of type 2 diabetes, which agrees with the present findings. However, it should be noted that the interval between onset of diabetes and development of retinopathy was dependent on glycemic control, particularly in patients with early onset of diabetes. There was a clear-cut tendency for young patients with good glycemic control to show a longer duration before developing retinopathy. The present prospective study has confirmed the results of our previous retrospective study [ 161, in that both the duration of type 2 diabetes and a poor glycemic control are major risk factors
for diabetic retinopathy. Evidently, long-term exposure to hyperglycemia is a fundamental condition for the development of retinopathy in type 2 diabetes. Although the duration of diabetes is an uncontrollable risk factor, it is usually possible to regulate blood glucose level in type 2 diabetes. Strict control of blood glucose level can be expected both to reduce the incidence of retinopathy and to delay its appearance. Improvement of glycemic control should, therefore, be a primary target in the treatment of type 2 diabetes so as to reduce the likelihood of retinopathy.
References
6
I
8
9
10
Skyler, J. (1979) Complications of diabetes mellitus: relationship to metabolic dysfunction. Diabetes Care 2, 499-509. West, K.M., Erdreich, I.J. and Stober, J.A. (1980) A detailed study of risk factors for retinopathy and nephropathy in diabetes. Diabetes 29, 501-508. Pirrart, J. (1978) Diabetes mellitus and its degenerative complications: a prospective study of 4400 patients observed between 1947 and 1973. Diabetes Care 1, 168-188,252-263. Knowler, W.C., Benett, P.H. and Ballinitine, E.J. (1980) Increased inidence of retinopathy in diabetics with elevated blood pressure. A six-year follow-up study in Pima Indians. New Engl. J. Med. 302, 645-650. Ballard, D.J., Melton, L.J.III, Dwyer, MS., Trautmann, J.C., Chu, C.P., CYFallon, W.M. and Palumbo, P.J. (1986) Risk factors for diabetic retinopathy: a population-based study in Rochester, Minnesota. Diabetes Care 9, 334-342. Krolewski, A.S., Warram, J.H.. Rand. , L.I.., Christlieb, A.R., Busick, E.J. and Kahn, C.R. (1986) Risk of proliferative diabetic retinopathy in juvenile-onset type 1 diabetes: A 40-year follow-up study. Diabetes Care 9, 443-452. Nathan, D.M., Signer, D.E., Godine, J.E., Harrington, C.H. and Perlmuter, L.C. (1986) Relationship in older type II diabetics. Association with glucose control. Diabetes 35, 797-801. Jerneld, B. and Algvere, P. (1985) Prevalence of retinopathy in diabetics treated with oral antihyperglycemic agents. Acta Ophthalmol. 63, 535-540. Tchobroutsky, G. (1978) Relation of diabetic control to microvascular development of complications. Diabetologia 15: 143-152. Dornan, T.I., Mann, J.I. and Turner, R.C. (1982) Factors protective against retinopathy in insulin-dependent
263 diabetics free of retinopathy. Br. Med. J. 285,1073-1077. 11 Maneschi, F., Cassar, L., Lowy, C. and Kohner, E.M. ( 1981) Development of diabetic microangiopathy and diabetic control. Diabete Metab. 7: 181-187. 12 Pettitt, D.J., Knowler. W.C., Lisse, J.R. and Benett, P.H. (1980) Development of retinopathy and proteinuria in relation to plasma glucose concentration in Pima Indians. Lancet ii, 1050-1052. 13 Teuscher, A., Schnell, H. and Wilson, P.W.F. (1988) Incidence of diabetic retinopathy and relationship to baseline glucose and blood pressure. Diabetes Care 11, 246-251. 14 Kingsley, L.A., Dorman, J.S., Doft, B.H., Orchard, T.J., LaPort, R.E., Kuller, L.H. and Drash, A.L. (1988) An epidemiologic approach to the study of retinopathy: the Pittsburgh diabetic morbidity and retinopathy studies. Diabetes Res. Clin. Pratt. 4, 99-109. 15 Klein, B.K.E., Moss, S.E. and Klein, R. (1987) Longitudinal measure of glycemic control and diabetic retinopathy. Diabetes Care 10. 273-277.
16 Sasaki, A., Horiuchi, N., Hasegawa, K. and Uehara, M. (1984) Risk factors for diabetic retinopathy in Japanese diabetic patients with special reference to metabolic control. J. Japan Diab. Sot. 27, 1115-1122. 17 Minowa, S.,Takahashi, S., Mayuzumi, N. and Miyashita, H. (1962) Studies on the standard body weight for Japanese adults. Japan Med. J. No. 1988. 24-28. 18 Fuller, J.H., McCartney, P., Jarrett, R.J., Keen, H., Rose, G., Shipley, M.J. and Hamilton, P.J.S. (1979) Hyperglycemia and coronary heart disease: The Whitehall Study. J. Chron. Dis. 32, 721-727. 19 Ishihara. M., Yukimura, Y., Aizawa. T.. Yamada, T., Ohto, K. and Yoshizawa, K. (1987) High blood pressure as risk factor in diabetic retinopathy development in NIDDM patients. Diabetes Care 10, 20-25. 20 Pactkau. M.E., Boyd, T.A.S., Winship, B. and Grace, M. (1977) Cigarrette smoking and diabetic retinopathy. Diabetes 26, 46-49.
251-263
257
Elsevier DIABET
00447
Development of diabetic retinopathy and its associated risk factors in type 2 diabetic patients in Osaka district, Japan: a long-term prospective study Akira Sasaki, Naruto
Horiuchi,
Kyoichi Hasewgawa
and Masuko
Uehara
Osaka Medical Center for Adult Diseases. Osaka, Japan
(Received 11 May 1990) (Revision received 3 July 1990) (Accepted 14 July 1990)
Summary Incidence rates of diabetic retinopathy and its associated risk factors were studied in a long-term prospective study involving 976 type 2 (non-insulin dependent) diabetic patients that showed no retinopathy at entry to the study. 322 of the patients (one-third of the subjects studied) developed diabetic retinopathy during the observation period (average length 8.3 years). The incidence rate per 1000 personyears was 39.8. The rate was significantly related to age at onset of diabetes, to fasting plasma glucose (FPG) level at entry, and to type of treatment. The incidence rate was also related to the duration of the disease. Glycemic control clearly played a role in the occurrence of retinopathy. Patients with FPG 3 200 mg/dl had the highest incidence rate, while patients with FPG < 140 mg/dl showed the lowest incidence rate throughout the observation period. Furthermore, a longer period between onset of diabetes and development of retinopathy was observed in patients with FPG < 140 mg/dl compared to patients with FPG > 200 mg/dl. The findings suggest that strict glycemic control may be of particular value to reduce the incidence of retinopathy and to delay its appearance in type 2 diabetic patients. Key words: Diabetic retinopathy
; Risk factor; Type 2 diabetes ; Prospective
Introduction Previous studies concerning the development of diabetic retinopathy pointed out the relation to duration of diabetes [l-8] and poor glycemic control as major risk factors [ 5-7,9-151, whereas ____ Address for correspondence: Professor A. Sasaki, Osaka Prefectural College of Nursing, 1-41 Tezukayama-Higashi 2 Chome, Sumiyoshi, Osaka, Japan. 016%8227/90/$03.50
0 1990 Elsevier Science Publishers
study
the role of hypertension, obesity, elevated lipid levels, impaired insulin response and cigarette smoking in the development of diabetic retinopathy are less certain. We found previously, in a retrospective study, that glycemic control and duration of diabetes played important roles as risk factors for diabetic retinopathy in type 2 diabetic patients [ 161. However, retrospective studies have certain shortcomings, so a prospective study with a long-term observation period
B.V. (Biomedical
Division)
258
was required for determining the natural history of diabetic retinopathy in Japan, while such studies have already been reported in the West [3,5,6,12,13]. The present study was carried out using a long-term prospective design for determining the incidence rate of diabetic retinopathy and its associated risk factors in type 2 diabetic patients in Japan, who showed no retinopathy upon entry to the study.
Subjects and methods Subjects The subjects studied were 976 type 2 (non-insulin dependent) diabetic patients (608 males and 368 females) who were first seen at the Osaka Medical Center for Adult Diseases, Osaka, Japan, during the period 1960-l 979. They were followed until the end of 1984. None of them showed retinopathy at entry, and they all had fasting plasma glucose levels >, 140 mg/dl or a 2 h level of 50 g OGTT > 180 mg at entry. Mean age ( k SD) was 52.1 + 10.9 years, and mean duration of diabetes was 3.0 + 4.6 years at entry. Methods Diabetic retinopathy was diagnosed by ocular fundus test performed by an ophthalmologist at least once a year as a rule. Glucose level was determined by the ferricyanide method using venous plasma. Albuminuria was measured by Albustix method. Glycemic control at entry was defined as the mean value of fasting plasma glucose (FPG) levels measured during the first year after the initiation of treatment. Obesity was expressed as percent of actual body weight divided by the standard body weight for Japanese adults [ 171. Classification of ECG findings were based on the criteria of the Whitehall Study [ 181 in which abnormalities in Q/QS and ST/T, LBBB and positive exercise test were defined as ischemic change. Incidence rate of diabetic retinopathy per 1000 person-years was computed according to the factors presented at entry to the study, since
observation periods differed from patient to patient. Life-table method was used for the computation of cumulative incidence rate. Chi-square test and Student’s t-test were used for statistical analysis.
Results Development of diabetic retinopathy and related risk factors 322 cases (i.e., one-third of the subjects studied) developed diabetic retinopathy during the observation period, which was 8.3 years on average. Table 1 shows the incidence rates of diabetic retinopathy per 1000 person-years, according to the factors present at the start of the follow-up period. The overall incidence rate of retinopathy was 39.8 per 1000 person-years, 38.2 for males and 42.5 for females, with no significant difference between the sexes. The rate was related significantly to the age at onset of diabetes, with higher incidence rate for younger age at onset. The rate was also related significantly to the FPG at entry, with a sharp increase as the FPG level rose above 110 mg/dl. The incidence of retinopathy was also related to the type of treatment; the incidence rate was significantly higher in groups treated with the oral hypoglycemic agent or with insulin compared with the diet-treated group. However, systolic blood pressure, ischemic ECG changes, obesity, serum cholesterol, albuminuria and cigarette smoking showed no significant relation to the incidence of retinopathy. The relation between incidence rate of retinopathy and the length of the observation period was analysed. The rates were 25.08 for patients with observation periods less than 5 years, 39.52 for patients with observation periods between 5-9years, 48.39 for observation periods of lo-14 years, and 58.38 for observation periods >, 15 years. Fig. 1 shows those relations with respect to FPG levels. The rates were affected signif?cantly by the FPG levels. It is noteworthy that patients with FPG 3 200 mg/dl always had the highest incidence rates regardless of the length
259 TABLE 1 Incidence
rate per 1000 person-years
of diabetic retinopathy
by factors at the baseline
Factors at baseline
No. of
subjects
Person-years observed
Cases with retinopathy
Incidence” rate
All subjects
916
8092.9
322
39.8
Sex
males females
608 368
5177.5 2915.4
198 124
38.2 42.5
Age at onset
t45 45 - 54 55 - 55 65 +
333 328 246 69
2883.0 2662.5 2012.0 535.6
130 115 62 15
45.1* 43.2 30.8 28.0
< 140 140 - 159 160+
522 304 134
4352.2 2548.1 1025.2
166 107 42
38.1 42.0 41.0
Fasting glucose level (mg/dl)
200 mg/dl) always showed the highest incidence rates of
262
retinopathy, regardless of the length of the observation period, whereas those patients that showed good glycemic control (FPG levels at entry incidence < 140 mg/dl) had lower rates throughout the observation period. The influence of initial FPG level on the cumulative incidence rate for retinopathy was maintained during the observation period. Thus, our findings support many recent studies on relations between retinopathy and fasting glucose level or glycated hemoglobin level, emphasizing the importance of glycemic control for reducing the risk of retinopathy [5,7,9-151. Relations of retinopathy to blood pressure [2,4,10,13,19] or to cigarette smoking [ 14,201 were, however, not apparent in the present study. It is noteworthy that the cumulative incidence rate of retinopathy in patients with FPG level > 200 mg/dl at entry seemed to reach a plateau at about 70% at the end of the follow-up period (Fig. 2). This may imply that some type 2 diabetic patients are resistent to the development of retinopathy, in spite of both a long duration of diabetes and poor glycemic control. Perhaps predispositional or genetic factors are responsible for such differences [ 161. The period from onset of diabetes to development of retinopathy was estimated as 9.2 years on average in the present study, indicating that a considerable time was required for development of retinopathy after the onset of diabetes. Nathan et al. [ 71 reported that the incidence of retinopathy increased 10 years after the onset of type 2 diabetes, which agrees with the present findings. However, it should be noted that the interval between onset of diabetes and development of retinopathy was dependent on glycemic control, particularly in patients with early onset of diabetes. There was a clear-cut tendency for young patients with good glycemic control to show a longer duration before developing retinopathy. The present prospective study has confirmed the results of our previous retrospective study [ 161, in that both the duration of type 2 diabetes and a poor glycemic control are major risk factors
for diabetic retinopathy. Evidently, long-term exposure to hyperglycemia is a fundamental condition for the development of retinopathy in type 2 diabetes. Although the duration of diabetes is an uncontrollable risk factor, it is usually possible to regulate blood glucose level in type 2 diabetes. Strict control of blood glucose level can be expected both to reduce the incidence of retinopathy and to delay its appearance. Improvement of glycemic control should, therefore, be a primary target in the treatment of type 2 diabetes so as to reduce the likelihood of retinopathy.
References
6
I
8
9
10
Skyler, J. (1979) Complications of diabetes mellitus: relationship to metabolic dysfunction. Diabetes Care 2, 499-509. West, K.M., Erdreich, I.J. and Stober, J.A. (1980) A detailed study of risk factors for retinopathy and nephropathy in diabetes. Diabetes 29, 501-508. Pirrart, J. (1978) Diabetes mellitus and its degenerative complications: a prospective study of 4400 patients observed between 1947 and 1973. Diabetes Care 1, 168-188,252-263. Knowler, W.C., Benett, P.H. and Ballinitine, E.J. (1980) Increased inidence of retinopathy in diabetics with elevated blood pressure. A six-year follow-up study in Pima Indians. New Engl. J. Med. 302, 645-650. Ballard, D.J., Melton, L.J.III, Dwyer, MS., Trautmann, J.C., Chu, C.P., CYFallon, W.M. and Palumbo, P.J. (1986) Risk factors for diabetic retinopathy: a population-based study in Rochester, Minnesota. Diabetes Care 9, 334-342. Krolewski, A.S., Warram, J.H.. Rand. , L.I.., Christlieb, A.R., Busick, E.J. and Kahn, C.R. (1986) Risk of proliferative diabetic retinopathy in juvenile-onset type 1 diabetes: A 40-year follow-up study. Diabetes Care 9, 443-452. Nathan, D.M., Signer, D.E., Godine, J.E., Harrington, C.H. and Perlmuter, L.C. (1986) Relationship in older type II diabetics. Association with glucose control. Diabetes 35, 797-801. Jerneld, B. and Algvere, P. (1985) Prevalence of retinopathy in diabetics treated with oral antihyperglycemic agents. Acta Ophthalmol. 63, 535-540. Tchobroutsky, G. (1978) Relation of diabetic control to microvascular development of complications. Diabetologia 15: 143-152. Dornan, T.I., Mann, J.I. and Turner, R.C. (1982) Factors protective against retinopathy in insulin-dependent
263 diabetics free of retinopathy. Br. Med. J. 285,1073-1077. 11 Maneschi, F., Cassar, L., Lowy, C. and Kohner, E.M. ( 1981) Development of diabetic microangiopathy and diabetic control. Diabete Metab. 7: 181-187. 12 Pettitt, D.J., Knowler. W.C., Lisse, J.R. and Benett, P.H. (1980) Development of retinopathy and proteinuria in relation to plasma glucose concentration in Pima Indians. Lancet ii, 1050-1052. 13 Teuscher, A., Schnell, H. and Wilson, P.W.F. (1988) Incidence of diabetic retinopathy and relationship to baseline glucose and blood pressure. Diabetes Care 11, 246-251. 14 Kingsley, L.A., Dorman, J.S., Doft, B.H., Orchard, T.J., LaPort, R.E., Kuller, L.H. and Drash, A.L. (1988) An epidemiologic approach to the study of retinopathy: the Pittsburgh diabetic morbidity and retinopathy studies. Diabetes Res. Clin. Pratt. 4, 99-109. 15 Klein, B.K.E., Moss, S.E. and Klein, R. (1987) Longitudinal measure of glycemic control and diabetic retinopathy. Diabetes Care 10. 273-277.
16 Sasaki, A., Horiuchi, N., Hasegawa, K. and Uehara, M. (1984) Risk factors for diabetic retinopathy in Japanese diabetic patients with special reference to metabolic control. J. Japan Diab. Sot. 27, 1115-1122. 17 Minowa, S.,Takahashi, S., Mayuzumi, N. and Miyashita, H. (1962) Studies on the standard body weight for Japanese adults. Japan Med. J. No. 1988. 24-28. 18 Fuller, J.H., McCartney, P., Jarrett, R.J., Keen, H., Rose, G., Shipley, M.J. and Hamilton, P.J.S. (1979) Hyperglycemia and coronary heart disease: The Whitehall Study. J. Chron. Dis. 32, 721-727. 19 Ishihara. M., Yukimura, Y., Aizawa. T.. Yamada, T., Ohto, K. and Yoshizawa, K. (1987) High blood pressure as risk factor in diabetic retinopathy development in NIDDM patients. Diabetes Care 10, 20-25. 20 Pactkau. M.E., Boyd, T.A.S., Winship, B. and Grace, M. (1977) Cigarrette smoking and diabetic retinopathy. Diabetes 26, 46-49.
