Diabetes is divided into two major types. Type I, or insulin-dependent diabetes mellitus, had been called “juvenile-onset diabetes”; it is characterized by an immediate need for insulin at the time of diagnosis. Type II, or noninsulin-dependent diabetes, had been described previously as “adult-onset diabetes.” However, there are many subvarieties within the disorder known as diabetes mellitus, which makes this differentiation less accurate and blurs these distinctions. An examination of the features that distinguish the varieties is beyond the scope of this chapter. Nevertheless, be aware that several diverse types of disorders are included within the name diabetes mellitus, and they may not be the same. These disorders share a historic name derived from an abnormality in glucose metabolism. For this reason, however, the ocular manifestations of diabetes mellitus were described originally as if they were one disease. It is now known that they represent different disease processes. It has been found that some of the ophthalmic features of the different kinds of diabetes mellitus vary, as do their responses to therapy.

Type I diabetes mellitus is less common and has its highest rate of occurrence in the Scandinavian countries. There is a gradient across Europe of a decreasing incidence of insulin-dependent diabetes for children younger than 16 years of age, which starts in Finland (30 new cases per 100,000 per year) and diminishes while approaching the Mediterranean coast of Italy (5 new cases per 100,000 per year). An exception is a “hot spot” on the island of Sardinia.⁴

In contrast, type II diabetes has a different distribution in the world. Type II diabetes is found most often among certain indigenous peoples, such as the natives of the Nauru Islands in the Pacific Ocean and the Pima Indians of North America. In these two groups, the prevalence of type II diabetes approaches 50% of the adult population. A high peak also is found among the other Native American populations but does not reach the numbers encountered among the Pima Indians.⁵

Ten percent or more of individuals with diabetes mellitus develop visual impairment within 15 years of diagnosis.⁶ In the United States in 1980, about 8% of the individuals who were legally blind were blind from diabetes. In the population aged 65 years and older, this percentage doubles.⁷

The duration of diabetes, age of patient, and age of patient at the time of diagnosis all have been found to be related to the development of diabetic retinopathy.^9,10 In patients with type I diabetes, retinopathy was infrequent (less than 8%) for the first 3 years after diagnosis. However, after 20 or more years of diabetes, 99% of patients had retinopathy. In addition, after 30 years of diabetes, over 50% of these patients developed proliferative diabetic retinopathy. In comparison, for patients with type II diabetes, at 3 years from the time of diagnosis, almost 23% had some retinopathy, and 2% of them had proliferative diabetic retinopathy. However, after 20 years or more of diabetes, fewer type II patients were found to have any form of retinopathy (60%) or proliferative retinopathy (5%) than the type I patients.^9,10

The relation of blood pressure to diabetic retinopathy remains uncertain. Although numerous studies imply no direct relation, recent reports indicate that when diastolic blood pressure is routinely elevated, diabetic retinopathy is stimulated to develop and progress.¹¹

As expected, hyperglycemia is the most important risk factor for the development and progression of diabetic retinopathy. The Diabetes Control and Complications Trial determined that intensive medical therapy, with the goal of maintaining glucose levels within the normal range, could prevent or slow the progression of diabetic retinopathy in patients with type I diabetes mellitus.¹² Similar features were discovered for patients with type II diabetes by studies in England.¹³

Protein glycosylation is the nonenzymatic reaction of glucose with the lysine residues of protein. Repeated or continuously high levels of serum glucose are necessary to produce this type of protein glycosylation. The resulting chemical compound rearranges itself and is transformed into an irreversible structure called an Amadori product. When proteins and Amadori products combine, they form “advanced glycosylation end products.” In many tissues, advanced glycosylation end products have been found to cause the release of biologically active molecules. These biologically active molecules produce microvascular changes identical to the features found in diabetic retinopathy. Aminoguanidines are one class of agents that have been found to inhibit formation of advanced glycosylation end products in animal studies. Since aminoguanidines can prevent and reverse early microvascular abnormalities in experimental animals, in theory they can prevent diabetic retinopathy.^14,15 Whether this is true in humans remains a subject of intense study.

In the presence of high concentrations of sugars, the aldose reductase enzyme converts the sugars to alcohols. Certain cells, such as the pericytes and endothelial cells of the retinal vasculature, are known to have high concentrations of this enzyme. Therefore, in the presence of high levels of blood glucose, alcohols are formed within the pericytes and endothelial cells, which can result in the death of those cells. In histopathologic studies, this type of cell death has been described as one of the first stages in the development of diabetic retinopathy¹⁶ (Fig. 1). The aldose reductase enzyme and the cellular after effects of its activity have been implicated in many of the ocular problems associated with diabetes mellitus.¹⁷ Several different kinds of medications have been developed to inhibit the ocular effect of this enzyme. In experimental animal studies, they have worked well. However, multiple human trials have reveal that such agents are not suitable for regular use.^18,19

As described earlier, the Diabetes Control and Complications Trial demonstrated a relation between increased glycosylated hemoglobin levels and the development or progression of diabetic retinopathy. Since hemoglobin is regularly and continuously produced in the bone marrow, elevated blood glucose levels at the time of hemoglobin production result in increased levels of glucose-bound hemoglobin. Oxygen cannot become free from glycosylated hemoglobin as easily as it can from nonglycosylated hemoglobin. This results in a relative tissue hypoxia. When there is more glycosylated hemoglobin, there is less oxygenation of tissues throughout the body. This tissue hypoxia may be the final common pathway for the development of diabetic retinopathy. The Diabetes Control and Complications Trial demonstrated that maintaining a glycosylated hemoglobin level at 7 or less can prevent the development of diabetic retinopathy or slow the progression of this disorder, if it is present.^12,20 Current theories imply that the relative hypoxia associated with elevated glycosylated hemoglobin levels stimulates vascular endothelial growth factor (VEGF) production.²¹ This may be another feature related to the development of diabetic retinopathy.

The morphologic changes that develop in the diabetic patient's corneal epithelial cells are well known. These include polymorphism, polymegethism, irregular cellular distribution, and stunting of surface cell microvilli.²⁵ In diabetes mellitus, changes in the thickness and composition of the basement membranes occur throughout the body and in the cornea. Corneal epithelial basement membrane thickening and other discontinuities have been described in the eyes of diabetic patients.²⁶ The epithelial barrier function may be altered in the cornea of diabetic patients. The corneal epithelial zonulae serve as a diffusion barrier,²⁷ and the corneal epithelial permeability is increased by a factor of about five in diabetes mellitus patients.²⁸

In addition to these corneal epithelial problems, the corneal endothelium has an increased incidence of dysfunction in individuals with diabetes.²⁹ Similarly, there are changes in endothelial cell morphologic features in such patients. The corneal endothelium normally consists of a monolayer of cells arranged in a regular hexagonal pattern. In diabetes, there is inadequate cell volume regulation associated with cytoskeletal abnormalities. Computer-assisted morphometric analysis has found quantitative changes in cell variation (polymegathism) and cell shape variations (pleomorphism).³⁰ As a result, there are morphologic changes in Descemet's membrane that influence the success rate of corneal transplantation surgery.³¹ Because of these changes, there is great concern regarding refractive surgery in individuals with diabetes mellitus. Although not examined as a specific factor in scientific studies of refractive surgery, most reports of such investigations have excluded patients with systemic diseases (such as diabetes mellitus), which can influence corneal healing.³² Many recent studies consider diabetes mellitus to be a contraindication for such procedures.³³

The greatest visual threat to patients having both glaucoma and diabetes is the development of neovascular glaucoma. Over 50 years ago, it was hypothesized that ischemic retinal tissue released a diffuseable angiogenic substance that stimulated neovascularization.³⁶ However, only recently was VEGF found to be an agent that satisfies this description.³⁷ It has been discovered that significantly higher VEGF levels occur in the ocular fluids of patients with proliferative diabetic retinopathy when compared with patients without that disorder. Extremely elevated levels of VEGF were found in the ocular fluids of patients with neovascular glaucoma.

The clinical course of neovascular glaucoma can be divided into several stages.³⁸ In the prerubeosis stage, there is no evidence of clinically detectable new vessel formation in the iris or anterior angle; however, there is proliferative diabetic retinopathy in the posterior portion of the eye. In the preglaucoma stage (rubeosis iridis), new vessels are visible by clinical examination on the iris or in the anterior chamber angle. These vessels are seen first, most often, in the peripupillary region. Although the anterior chamber angle is open in these patients, vessels can be seen taking a direct path across the peripheral iris-ciliary body band and scleral spur. In the next stage, the open-angle glaucoma stage, there are increased vessels on the iris stroma. By gonioscopic examination, the anterior chamber angle still appears to be open, although there more new vessels in the angle. At this stage, a fibrovascular membrane develops, which covers the iris and grows into the anterior chamber angle to inhibit aqueous outflow. During the final stage of this disorder, the angle-closure glaucoma stage, the fibrovascular membrane contracts to produce changes on the iris and anterior chamber angle. The peripheral iris is pulled forward to create a partial or total peripheral anterior synechia. This synechial closure of the angle produces the intraocular pressure elevation.

The management of this disorder is related to the stage of the disease. In the prerubeosis stage, treatment of the underlying proliferative diabetic retinopathy by panretinal photocoagulation causes regression of the abnormality. The rubeosis iridis stage also is treated by panretinal photocoagulation for the underlying diabetic retinopathy. In cases where panretinal photocoagulation is not possible, panretinal cryoablation, or endophotocoagulation, is a reasonable alternative. For patients in the openangle stage, it is first necessary to treat the elevated intraocular pressure. This should be attempted with medications and, after the pressure is brought to a normal level, panretinal photocoagulation needs to be applied. If the anterior segment neovascularization can be eliminated with panretinal photocoagulation, such patients are good candidates for filtration surgery. For patients in the angle-closure stage, the first therapy should be to alleviate pain and reduce intraocular pressure. Panretinal photocoagulation still remains the primary operation of choice, even at this late stage. Such treatment reduces the amount of anterior segment neovascularization. In some cases, filtration can be done after that. However, these patients often need implant drainage surgery.³⁹

In addition to these refractive problems, there are significant changes in accommodative activity in diabetic patients. Some of this results from the changes in lens hydration and in the lens capsule. However, in addition, many diabetes patients have glycogen deposition within the ciliary body. This reduces the ciliary body's ability to function and starts presbyopia at an earlier age than in nondiabetic patients.

As described earlier, in the hyperglycemic state there is an elevation of aqueous glucose levels. The glucose diffuses through the lens capsule and increases its concentration within the crystalline lens. Several pathways for glucose metabolism exist within the lens. However, most intralenticular enzyme systems are easily saturated, and the excess glucose accumulates within the lens. This results in a glycosylation of some lens proteins. Some of the glucose is converted by the intracellular enzyme aldose reductase into sorbitol, an alcohol. The accumulation of sorbitol within the lens causes further osmotic change, alterations in lens permeability, and cataract formation.

A special type of cortical cataract has been seen in many diabetic patients. This abnormality was noted to have a type of “snowflake” appearance with bilateral, widespread subcapsular changes. Often, it had an abrupt onset and an acute course. These cataracts were seen most often in young persons with uncontrolled diabetes mellitus. These lens changes matured rapidly and resulted in total opacification over a period of a few weeks. This “true diabetic cataract” currently is rare because there are better means of diabetes control.

In contrast to the unique type of juvenile-onset cataract, the more typical aging-type senescent lens opacities are more common in diabetic patients. In large, population-based epidemiologic studies, diabetic patients have been found to be at an increased risk for regular senescent cataract formation. The Health and Nutrition Examination Survey and the Framingham Eye Study found an incidence of typical senile cataracts to be increased by a factor of three or more in diabetic individuals compared with an age-matched nondiabetic population.⁴⁰

In general, diabetic retinopathy is not a contraindication for cataract surgery. In some cases, lens extraction may be needed to improve retinal surveillance.⁴¹ However, cataract surgery can stimulate additional problems for a patient with diabetes mellitus. Neovascular glaucoma is a recognized complication of cataract surgery in diabetic patients. Cataract surgery, particularly intracapsular cataract surgery or any type of cataract surgery that results in a capsulotomy, allows the angiogenic factors produced by the retina to flow from the posterior segment forward to reach the iris. In that situation, the angiogenic factors that are causing the proliferative diabetic retinopathy stimulate neovascular glaucoma.

The greatest clinical problem in early diabetic retinopathy is associated with a less common disorder that occurs with areas of retinal thickening. This is macular edema secondary to diabetes, and it can cause significant visual loss. Many patients with these disorders have their associated systemic disease under a poorer level of control than is desirable. By improving the systemic disease therapy, returning the blood glucose to normal, correcting coexisting renal abnormalities, and bringing the blood pressure under control, many of these patients can have the diabetic macula edema reversed or have the area of diabetic macula edema reduced. Therefore, laser treatment often is limited to the residual disease after the systemic therapy has been made as good as the patient can tolerate. Photocoagulation to the origin of the leakage, or “grid treatment” to areas of diffuse leakage, can reverse the edema in many individuals (Fig. 2). One study found that the rate of visual loss can be reduced by 50% by treating a select category of macular edema known as “clinically significant macula edema,”⁴³ which was defined as intraretinal edema containing any one of the following features:

  Thickening of the retina at or within 500 μm of the center of the macula
  Hard exudates at or within 500 μm of the center of the macula when associated with thickening of the adjacent retina
  A zone, or zones, of retinal thickening that are 1500 μm or larger in diameter when any portion of the zone is within 1500 μm of the center of the macula

Although laser surgery benefits many patients with macula edema, the benefit is greatest in this select population containing clinically significant macula edema. In such patients, treatment is directed to the following:

  All focal leaks that are 500 μm or more from the center of the macula
  Any focal leaks that are more than 300 μm but less than 500 μm from the center of the macula, if such treatment will not destroy the perifoveal capillary network
  Areas of diffuse leakage that are treated with a grid of moderate laser intensity when each laser spot has a diameter of 200 μm or less and there is an area of 100 μm or more of untreated retina between each spot
  Zones of capillary nonprofusion (except for the foveal avascular zone) that are treated in a grid pattern

After years of background diabetic retinopathy, patients progress to develop a more visually threatening disorder known as proliferative diabetic retinopathy (Fig. 3). The Diabetic Retinopathy Study, performed between 1972 and 1975, found that blindness from proliferative diabetic retinopathy can be significantly reduced by satisfactory photocoagulation treatment. In the Diabetic Retinopathy Study, 1758 patients were studied with randomization of laser treatment to one eye and the fellow eye assigned as a control. In 28 months, blindness developed in 16% of the untreated control eyes that had neovascularization at the start of the study. Such a result was found in only 6.4% of treated eyes.⁴⁴ This therapeutic result was so positive that this has become the standard of therapy in the United States. In general, this treatment consists of 1200 moderate-intensity photocoagulation spots applied to the midperiphery. This therapy is performed at a laser spot size of 500-μm diameter for each application, and each spot is applied at 500 μm from its nearest neighbor. This allows enough clear, untreated retina between each application to permit areas of residual function that reduce the risk of a compromised peripheral visual field. Often, this treatment is not applied in one sitting because of patient and physician discomfort. Instead, it is initiated and becomes a part of a series of two or three treatment sessions, each about 1 week apart (Fig. 4). In addition to preventing blindness, the Diabetic Retinopathy Study was able to identify the features of diabetic retinopathy that are associated with an increased risk of blindness. These risk factors were described as follows:

  Presence of vitreous or preretinal hemorrhage
  Presence of neovascularization
  Location of neovascularization on or near (within 1 disc diameter) the optic disc
  Severity of the new vessels (severity, in this situation, means that the neovascularization covers more than one third of a disc area when it is on or near the disc; or the neovascularization covers an area of 2 disc diameters or more when the disc is not involved)

In the presence of a single risk factor, the danger of developing blindness in the near future is 6.7%. When two risk factors are present, this becomes 8.5%. When three risk factors are present, there is a great increase to 26.7%. There is a further increase to 36.9% when all four risk factors are present.⁴⁵ In contrast, retinal photocoagulation, as performed in the diabetic retinopathy study, reduces this risk of becoming blind to 6.4%. That is, the risk of developing blindness does not disappear completely but becomes substantially less.

The problem remains, however, that such therapy has inherent dangers. Every form of treatment contains some risk to the patient. In the best of situations, the complication rate of such panretinal photocoagulation surgery approaches 9%. These dangers involve changes in peripheral visual fields from the laser spots themselves or visual distortion caused by a wrinkling of the inner limiting membrane of the retina. Therefore, a patient who has 20/20 vision and no visual complaints but has the ophthalmic manifestations of diabetes detected by clinical examination will have about a 9% risk of reduced vision after treatment. Although this level of reduced vision is much less than would occur without treatment, patients are reluctant to undergo therapy when the known complication rate is this high. Therefore, many practitioners have found it prudent to wait for the development of at least three of the risk factors described earlier before initiating intervention. When three are present, the risk-benefit ratio is so strongly in favor of treatment that there is little reason to hesitate. However, recent discoveries have raised questions as to whether such a delay is in the best interest of every patient. This is addressed later in this section.

In most patients, the risk factors regress within 2 months after treatment; in a few, they do not; and in others who have an initial regression of disease, these risk factors recur about 6 to 9 months after the initial treatment. In such cases, it has become common practice to apply another 800 photocoagulation marks to the midperiphery of the retina to cause the regression of diabetic retinopathy once more. In most clinical situations, such therapy is reapplied when the patient's eye develops three risk factors. Nevertheless, it is uncommon for a patient to need more than three repeat treatments of this variety.

To determine the most appropriate time to perform photocoagulation for diabetic retinopathy, the National Institutes of Health started the Early Treatment of Diabetic Retinopathy Study (ETDRS) in April 1980. This resulted in a better understanding of the treatment of clinically significant macula edema as described earlier.⁴³ However, years after the completion of the ETDRS, when there was a better understanding of the differences between type I and type II diabetes mellitus, the data from that study were reassessed. At this time, the patients with type II diabetic retinopathy were found to have a response that was measurably different from those with type I diabetes. In patients with type I diabetes mellitus, treatment that was applied as described earlier, before the development of high-risk characteristics, offered no benefit compared with waiting until the onset of high-risk characteristics. In addition, waiting until the discovery of the high-risk characteristics deferred the potential onset of the known complications of such therapy. However, in patients with type II diabetes, waiting for the development of three high-risk characteristics was associated with a reduced benefit. In addition, the deferral of the onset of complications in patients with type II diabetes mellitus was not equal to this benefit reduction. In brief, it was found that in some patients with type II diabetes, treatment before the onset of high-risk characteristics increases benefits and reduces complications.⁴⁶ It is assumed that this is because many type II diabetes patients have additional, unmeasureable microvascular disease.

As a nondiabetic person ages, the vitreous gel develops multiple sites of liquefaction. These fluid-filled cavities gradually fuse; with normal eye motion, this fluid migrates to separate the retina from the residual gel-like vitreous. In the nondiabetic person, these separations continue to progress, except in the areas where the vitreous fibrils are firmly adherent to the vitreous base. In contrast, in individuals with diabetes mellitus, these changes occur soon after the onset of proliferative diabetic retinopathy. In addition to the vitreous base, there are other sites of firm adhesion of the vitreous fibrils; this is specifically noted in the areas of neovascularization origin.⁵⁰ These areas of vitreous gel adhesion to the retinal surface become regions of vitreoretinal traction. As the vitreous gel collapses further, tractional forces develop throughout this vitreoretinal complex that distort the normal retinal architecture. This vitreous collapse stresses the neovascular proliferative tissue and results in ruptured blood vessels and vitreous hemorrhage. Although some of this blood can reabsorb and vision can return, in many patients, this is the first sign of progressive visual loss. Techniques of pars plana vitrectomy can remove such blood and return vision to eyes with these problems. However, questions remain as to which patients should undergo operation and how soon should they receive surgery after this type of blindness occurs. To resolve some of these questions, the National Institutes of Health initiated a multiple medical center collaborative prospective study known as the Diabetic Retinopathy Vitrectomy Study. In one of their first reports, an examination was made of the natural history of eyes that had a “blinding vitreous hemorrhage” (best corrected vision of 20/200 or worse). A few eyes were able to return to good vision without intervention in the first months after the blinding episode. However, after 4 months, only about 9.5% of such blind eyes returned to relatively good vision without surgical intervention.⁵¹ In addition, it was found that the chance of returning good vision with surgical intervention started to drop off soon after that. For individuals that had type I diabetes mellitus, surgical intervention at about 4 months after the blinding episode was associated with a final vision of 20/40 or greater in 36% of patients compared with only 12% of patients that had conventional management (i.e., surgical intervention initiated at 6 months or longer after the blinding episode).⁵² As expected, the results for patients with type I and type II diabetes mellitus appeared to be different. Although both types of patients were benefited by such intervention, there was a greater benefit associated with intervening earlier in the patients that have type I diabetes mellitus.⁵³

In 1992, it was found that 3.5% of the U.S. population had diabetes mellitus, and 15% of all health care costs in this country were spent for the treatment of this disorder.⁵⁸ Concerns therefore were raised regarding the cost-effectiveness of preventing the eye problems associated with diabetes mellitus. Many studies demonstrate that implementing screening and treatment programs for patients with type I diabetes mellitus is beneficial to the patients and is cost effective to the country.^59,60 In this manner, it was discovered that if all such patients received appropriate care, more than 79,000 person-years of sight would be saved. Similarly, it was found that screening and treatment for diabetic retinopathy also was beneficial and cost-effective for patients with type II diabetes mellitus.⁶¹ Therefore, the following clinical screening programs have been suggested for patients with diabetes mellitus⁶²:

1. Beginning 5 years after the onset of diabetes, all patients with type I diabetes mellitus should receive a detailed retinal evaluation.
   
2. Patients with type II diabetes mellitus should have an initial detailed retinal evaluation shortly after the diagnosis of diabetes.
   
3. After these initial examinations, these patients need repeated, detailed retinal examinations on a yearly basis and intervention when appropriate.
   

Despite all this, the following dilemma remains. More than 30 years ago, diabetes mellitus became the most common cause of new blindness in adults in the United States.^1,2 Methods of surgical intervention were developed in 1976 that could reduce the risk of diabetic-related blindness⁴⁴ by about 50%. Since then, medical therapeutic techniques were discovered that could prevent diabetic eye disorders or slow the rate of progression of these disorders if they were already in existence.^12,13 Nevertheless, diabetes mellitus remains the most common cause of new blindness in adults in the United States.² What can you do to prevent this epidemic of blindness?

1. Operational Research Department, National Society to Prevent Blindness: Vision Problems in the U.S.: A Statistical Analysis. New York, National Society to Prevent Blindness, 1980

2. National Diabetes Data Group: Diabetes in America, NIH publication no. 95-1468. Washington, DC, National Institutes of Health, 1995

3. World Health Organization: Diabetes Mellitus: Report of a WHO Study Group, technical report #727. Geneva, World Health Organization, 1985

4. King H, Rewers M: Global estimates for prevalence of diabetes mellitus and impaired glucose tolerance in adults. Diabetes Care 16:157, 1993

5. Alberti KG, Mazza ER (eds): Frontiers of Diabetes Research: Current Trends in NIDDM. Amsterdam, Elsevier, 1989

6. Gruber W, Lander T, Leese B (eds): Diabetes Health Economics Study Group: The Economics of Diabetes and Diabetes Care. Geneva, International Diabetes Federation and World Health Organization, 1997

7. National Society to Prevent Blindness: Vision Problems in the U.S. Data Analysis: Definitions, Data Sources, Detailed Data Tables, Analysis, Interpretations. New York, National Society to Prevent Blindness, 1980

8. Cruickshanks KG, Zadheim CM, Moss SE: Genetic marker associations with proliferative retinopathy in persons diagnosed with diabetes before thirty years of age. Diabetes 41:879, 1992

9. Klein R, Klein BEK, Moss S: The Wisconsin Epidemiology Study of Diabetic Retinopathy. II. Prevalence and risk of diabetic retinopathy when age at diagnosis is less than 30 years. Archiv Ophthalmol 102:520, 1984

10. Klein R, Klein BEK, Moss S: The Wisconsin Epidemiology Study of Diabetic Retinopathy. III. Prevalence and risk of diabetic retinopathy when age at diagnosis is 30 or more years. Archiv Ophthalmol 102:527, 1984

11. Klein R, Klein BEK, Moss SE: Is blood pressure a predictor of the incidence of progression of diabetic retinopathy? Archiv Intern Med 149:2427, 1989

12. The Diabetes Control and Complications Trial Research Group: The effect of intensive treatment of diabetes on the development and progression of long term complications in insulin dependent diabetes mellitus. N Engl J Med 329: 977, 1993

13. United Kingdom Prospective Diabetes Study [UKPDS] Group: Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type II diabetes. Lancet 352:837, 1998

14. Brownlee M, Cerami A, Vlassara H: Advanced glycation end products in tissue and the biochemical basis of diabetic complications. N Engl J Med 318:1315, 1988

15. Vlassara H, Brownlee M, Manogue K: Advanced glycation end product (AGE) binding to its macrophage receptor stimulates of a multifunctional growth promoting monokine. Science 240:1546, 1988

16. Cogan DG, Toussaint D, Kuwabara T: Retinal vascular patterns. IV. Diabetic retinopathy. Arch Ophthalmol 66:366, 1961

17. Cogan DG, Kinoshita JH, Kador PF et al: Aldose reductase and complications of diabetes. Ann Intern Med 101:82, 1984

18. Sorbinil Retinopathy Trial Research Group: A randomized trial of sorbinil, an aldose reductase inhibitor, in diabetic retinopathy. Arch Ophthalmol 108:1234, 1990

19. Feman SS, Leonard-Martin TC, Redman: The Vanderbilt Classification System in the evaluation of diabetic retinopathy patients treated with Alredase. Trans Am Ophthalmol Soc 94:433, 1996

20. The Diabetes Control and Complications Trial Research Group: The rleationship of glycemic exposure (HbA1c) to the risk of development and progression of retinopathy in the diabetes control and complications trial. Diabetes 44:968, 1995

21. Aiello LP, Northrup JM, Keyt BA: Hypoxic regulation of vascular endothelial growth factor in retinal cells. Arch Ophthalmol 113:1538, 1995

22. Schwartz DE: Corneal sensitivity in diabetics. Archiv Ophthalmol 91:174, 1974

23. Friend J, Thoft RA: The diabetic cornea. Int Ophthalmol Clin 24:111, 1984

24. Schultz RO, Peters MA, Sobocinski K: Diabetic corneal neuropathy. Trans Am Ophthamol Soc 81:107, 1983

25. Tsubota K, Chiba K, Shimazaki J: Corneal epithelium in diabetic patients. Cornea 10:156, 1991

26. Taylor HR, Kimsey RA: Corneal epithelial basement membrane changes and type. Invest Ophthalmol Vis Sci 20: 548, 1981

27. Parrish CM: The cornea in diabetes mellitus. Feman SS (ed): Ocular Problems and Diabetes Mellitus, p 179. Boston, Blackwell Scientific, 1992

28. Goebbels M, Spitznas M, Oldendoerp J: Impairment of corneal epithelial barrier function in diabetics. Graefes Arch Clin Exp Ophthalmol 227:142, 1989

29. Busted N, Olsen T, Schmitz O: Clinical observations on the corneal thickness and the corneal endothelium in diabetes mellitus. Br J Ophthalmol 65:687, 1981

30. Schultz RO, Matsuda M, Yee RW, Edelhauser HS: Corneal endothelium changes in type I and type II diabetes mellitus. Am J Ophthalmol 98:401, 1984

31. Shetlar DJ, Bourne DM, Campbell J: Morphologic evaluation of Deseemet's membrane and corneal endothelium in diabetes mellitus. Ophthalmology 96:247, 1989

32. Waring GO III: Examination and selection of patients for refractive keratotomy. In Waring GO III (ed): Refractive Keratotomy for Myopia and Astigmatism, p 309. St Louis, Mosby Year Book, 1992

33. Pallikaris IG, Siganos DS: Excimer laser in situ keratomileusis and photorefractive keratectomy for correction of high myopia. J Refract Corneal Surg 10:498, 1994

34. Becker B: Diabetes mellitus and primary open-angled glaucoma. Am J Ophthamol 71:1, 1971

35. Klein BE, Klein R, Jensen SC: Open-angle glaucoma in older onset diabetes: The Beaver Dam Eye Study. Ophthalmology 101:1173, 1994

36. Michaelson IC: The mode of development of the vascular system of the retina with some observations on its significance for certain retinal diseases. Trans Ophthalmol Soc UK 68:137, 1949

37. Aiello LP, Avery RL, Arrig PG: Vascular endothelium growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med 331:1480, 1994

38. Shields MB: Glaucoma in diabetic patients. Feman SS (ed): Ocular Problems and Diabetes Mellitus, p 207. Boston, Blackwell Scientific, 1992

39. Molteno ACB, VanRooyen MNB, Bartholomew RS: Implants for draining neovascular glaucoma. Br J Ophthalmol 61:120, 1977

40. Ederer F, Hiller R, Taylor HR: Senile lens change in diabetes in two population studies. Am J Ophthalmol 91:381, 1981

41. Flanagan DW: Current management of established diabetic eye disease. Eye 7:302, 1993

42. Feman SS: The natural history of the first clinically visible features of diabetic retinopathy. Trans Am Ophthalmol Soc 92:744, 1994

43. Early Treatment Diabetic Retinopathy Study Group: Photocoagulation for diabetic macula edema. Arch Ophthalmol 103:1796, 1985

44. Diabetic Retinopathy Study Research Group: Preliminary report on effect of photocoagulation therapy. Am J Ophthalmol 81:383, 1976

45. Diabetic Retinopathy Study Research Group: Four risk factors for severe visual loss in diabetic retinopathy. Arch Ophthalmol 97:654, 1979

46. Ferris F: Early photocoagulation in patients with either type I or type II diabetes. Trans Am Ophth Soc 94:505, 1996

47. Kalebic J, Garbisa S, Glaser B, Liotta L: Basement membrane degradation by migrating endothelial cells. Science 221:281, 1983

48. Davis M: Vitreous contraction in proliferative diabetic retinopathy. Arch Ophthalmol 54:741, 1965

49. Wallow IHL, Geldener DS: Endothelial fenestrae in proliferative diabetic retinopathy. Invest Ophthalmol Vis Sci 19:1176, 1980

50. Tagawa H, McMeel JW, Furukawa H: Role of the vitreous in diabetic retinopathy. Ophthalmology 93:596, 1986

51. Diabetic Retinopathy Vitrectomy Study Research Group: Early vitrectomy for severe proliferative diabetic retinopathy in eyes with useful vision. Ophthalmology 95:1307, 1988

52. Diabetic Retinopathy Vitrectomy Study Research Group: Early vitrectomy for severe vitreous hemorrhage in diabetic retinopathy: Four year results of the randomized trial. Arch Ophthalmol 108:958, 1990

53. Diabetic Retinopathy Vitrectomy Study Research Group: Early vitrectomy for severe proliferative diabetic retinopathy in eyes with useful vision: Clinical application of results of the randomized trial. Ophthalmology 95:1321, 1988

54. Feman SS, Leonard-Martin TC, Semchyshyn TM: The topographic distribution of the first sites of diabetic retinal neovascularization. Am J Ophthalmol 125:704, 1998

55. Moss SE, Klein R, Kessler SD, Richie KA: Comparison between ophthalmoscopy and fundus photography in determining the severity of diabetic retinopathy. Ophthalmology 92:62, 1985

56. American Diabetes Association: Eye care guidelines for patients with diabetes mellitus. Diabetes Care 11:745, 1988

57. American Academy of Ophthalmology, Quality of Care Committee, Retinal Panel: Diabetic Retinopathy: Preferred Practice Pattern. San Francisco, American Academy of Ophthalmology, 1989

58. Rubin RJ, Altman WM, Mendelson DN: Health care expenditures for people with diabetes mellitus. J Clin Endocrinol Metab 78:809, 1994

59. Javitt JC, Canner JK, Sommer A: Cost effectiveness of current approaches to the control of retinopathy in type I diabetes. Ophthalmology 96:255, 1989

60. Javitt JC, Aiello LP, Bassi LJ: Detecting and treating retinopathy in patients with type I diabetes mellitus: Savings associated with improved implementation of current guidelines. Ophthalmology 98:1565, 1991

61. Javitt JC, Aiello LP, Chang YP: Preventive eye care in people with diabetes is cost saving to the federal government: Implications for health care reform. Diabetes Care 17:909, 1994

62. American Diabetes Association: Policy statement on diabetic retinopathy: Clinical practice recommendations. Diabetes Care 21(suppl 1):157, 1998