KDOQI Clinical Practice Guidelines for Chronic Kidney Disease: Evaluation, Classification, and Stratification



The risk of cardiovascular disease, retinopathy, and other diabetic complications is higher in patients with diabetic kidney disease than in diabetic patients without kidney disease.


The onset of diabetes is characterized by metabolic and hemodynamic disturbances that increase vascular permeability, raise systemic blood pressure, and alter the regulation of intracapillary pressure. In the kidney, these changes may lead to increased trafficking of plasma proteins across the glomerular membrane and to the appearance of protein in the urine. The presence of urinary protein not only heralds the onset of diabetic kidney disease, but it may contribute to the glomerular and tubulointerstitial damage that ultimately leads to diabetic glomerulosclerosis.574 The strong relationship between proteinuria and a constellation of other diabetic complications supports the view that elevated urinary protein excretion reflects a generalized vascular disturbance that affects many organs, including the eyes, heart, and nervous system.575

This guideline describes the association of cardiovascular (macrovascular), retinal (microvascular), and other (principally neuropathic) complications of diabetes with levels of albumin/protein in the urine. It highlights the strong relationship between progressive diabetic kidney disease and the development of other diabetic complications and emphasizes the importance of monitoring and treating diabetic chronic kidney disease patients for these other complications.


Microalbuminuria refers to levels of urinary albumin excretion below those detected by standard dipstick methods and macroalbuminuria refers to higher levels of urinary albumin excretion. Microalbuminuria is present when the albumin excretion rate is 30 to 300 mg/24 hours (20 to 200 µg/min) or the albumin-to-creatinine ratio is 30 to 300 mg/g.576 Proteinuria generally refers to a positive dipstick test for protein or to a daily output of protein above a certain cut point, typically ≥ 500 mg/d. Thus, macroalbuminuria and proteinuria may be relatively equivalent measures of urinary protein excretion (see Guideline 5). Nevertheless, differences in methods of measurement and the lack of standardized definitions or terminology often make comparisons between studies difficult.

Definitions of Diabetic Complications Other Than Chronic Kidney Disease

Cardiovascular disease. Cardiovascular disease is not a specific complication of diabetes per se, since it occurs frequently in nondiabetic individuals. Diabetes and an array of metabolic disorders associated with it, however, increase the risk of cardiovascular disease in diabetic patients and may accelerate the process of atherosclerosis. For the purposes of this guideline, cardiovascular disease refers to coronary heart disease, cerebrovascular disease, peripheral vascular disease, congestive heart failure, and left ventricular hypertrophy. The American Diabetes Association provides clinical practice recommendations for screening and treatment of cardiovascular disease in diabetes526.

The variety of measures of cardiovascular disease used in different studies may limit the interpretability of the findings reviewed for this guideline, although the nearly uniformly positive association suggests this may not be an important limitation. On the other hand, cardiovascular disease itself may increase the level of urinary albumin/protein. Thus, the extent to which chronic diabetic glomerulosclerosis is an independent risk factor for the development of cardiovascular disease may be difficult to determine with certainty, especially in congestive heart failure, without demonstrating diabetic kidney damage at the tissue level.

Retinopathy. The earliest change of diabetic retinopathy that can be seen with the ophthalmoscope is the retinal microaneurysm. Other changes found in nonproliferative retinopathy include retinal hemorrhages, hard exudates, cotton-wool spots, intraretinal microvascular abnormalities (IRMA), and venous abnormalities. Growth of abnormal blood vessels and fibrous tissue that extends from the retinal surface or optic nerve characterizes the proliferative stage of diabetic retinopathy. With experience, these changes can be identified readily by direct ophthalmoscopy, preferably through dilated pupils. Stereoscopic fundus photographs, however, produce a more reliable and reproducible assessment of diabetic retinopathy. The Airlie House Classification scheme, or a modification of this scheme, is commonly used to classify the level of retinopathy in epidemiological studies; the more severely involved eye is used for classification. The American Diabetes Association provides clinical practice recommendations for screening and treatment of diabetic retinopathy.526

Neither the diagnosis nor classification of retinopathy was uniform in the studies reviewed for this guideline. Some studies performed retinal photographs (from 2 to 7 fields, depending on the study) and others relied on ophthalmoscopic examinations through dilated pupils. Moreover, retinopathy was graded by the Airlie House Classification scheme (or a modification of this scheme) in some studies and by less precisely defined clinical criteria in others. Beyond methodological issues, the absence of retinopathy in some subjects with elevated albuminuria/proteinuria may reflect the presence of nondiabetic kidney disease, particularly in older type 2 diabetic patients. These factors undoubtedly contributed, at least in part, to the reported variability of the association between retinopathy and albuminuria/proteinuria.

Neuropathy. Diabetic neuropathy is perhaps one of the most difficult complications of diabetes to measure. Although 60% to 70% of people with either type of diabetes are affected, many investigators in the past used non-standardized methods for measuring neuropathy. The lack of standardized nomenclature and criteria for diabetic neuropathy undoubtedly diminished the quality of the data available for review. Accordingly, studies examining the relationship between the level of urinary albumin/protein and diabetic neuropathy often yielded confusing and conflicting results.

In 1988, a joint conference of the American Diabetes Association and the American Academy of Neurology adopted standardized nomenclature and criteria for the diagnosis of neuropathy in diabetes.577 The classification was divided into subclinical and clinical neuropathy (Table 126).

Subclinical neuropathy is defined as an abnormal electrodiagnostic test, quantitative sensory threshold, or autonomic function test in the absence of clinical signs and symptoms. Clinical neuropathy is defined as an abnormal test associated with clinical signs and/or symptoms. The American Diabetes Association provides clinical practice recommendations for screening and treatment of diabetic neuropathy.526

Strength of Evidence

Given the numerous studies and general agreement on the relationship between proteinuria and complications of diabetes other than chronic kidney disease, review articles were used as the primary source of information for this guideline. Since reviews often reported the associations qualitatively, individual studies were included to provide quantitative estimates of the association. Reference was also made to individual studies of non-Caucasian patients, since many reviews reported only results from studies in Caucasians. Given the low rate or absence of type 1 diabetes in many non-Caucasians, the impact of ethnicity on the relationship between proteinuria and other diabetic complications was examined only in those with type 2 diabetes. Selection of individual studies for this guideline was not subject to the systematic review process used in other KDOQI guidelines and is intended to be illustrative rather than comprehensive.

Cardiovascular disease is related to the level of proteinuria or albuminuria in diabetic kidney disease (Table 127 and Figs 51 and 52) (R, C).

Figure 51

Cardiovascular mortality with diabetes. Relative cardiovascular mortality in type 1 (left panel) and type 2 (right panel) diabetes according to the level of urinary protein excretion in the WHO Multinational Study of Vascular Disease in Diabetes.597 Death rate ratios were adjusted for age, duration of diabetes, systolic blood pressure, serum cholesterol concentration, and smoking history by Poisson regression.

Figure 52

Microalbuminuria and cardiovascular morbidity with type 2 diabetes. Crude association of microalbuminuria and cardiovascular morbidity or mortality in type 2 diabetes. The results are presented with (total) and without (subtotal) the study that included subjects with clinical proteinuria. Adapted and reprinted with permission.586

Increased cardiovascular mortality was linked with elevated urinary albumin excretion in type 2 diabetes in 1984578,579 and with type 1 diabetes in 1987.580 This association was confirmed subsequently in many studies and described in numerous review articles,581-596 and it is reported in diverse racial/ethnic groups (Table 127).

The association between diabetic kidney disease and cardiovascular disease is generally considered stronger in type 2 than in type 1 diabetes at all levels of albuminuria/proteinuria, due in large part to the older age of the type 2 diabetic patients. In the WHO Multinational Study of Vascular Disease in Diabetes,597 however, similar cardiovascular death rates were reported in the 1,188 patients with type 1 and the 3,234 patients with type 2 diabetes from ten centers worldwide (Fig 51). These results may be influenced by the racial/ethnic mix of the sample cohort, since some populations included in the cohort with high rates of type 2 diabetes, such as the Pima Indians, have lower rates of cardiovascular disease than Caucasians with type 2 diabetes.601

A meta-analysis of 11 cohort studies involving 2,138 patients with type 2 diabetes586 provides compelling evidence that even modest elevations of urinary albumin excretion are associated with increased cardiovascular risk. In this review, patients with microalbuminuria had an overall crude odds ratio for cardiovascular morbidity and mortality of 2.0 (95% confidence interval, 1.4 to 2.7) compared to those with normal urinary albumin excretion (Fig 52).

Retinopathy is related to the level of proteinuria or albuminuria in diabetic kidney disease (Table 128)

(R, C). Review articles evaluated for this guideline included patients from clinic and population-based studies of type 1 and type 2 diabetes.581,585,590,591,595,602 These articles reflect the widely recognized positive association between the level of albuminuria/proteinuria and retinopathy in both types of diabetes. Discordance between these two diabetic complications, however, occurs frequently,603,604 particularly in type 2 diabetes, because of the coexistence of nondiabetic kidney disease. Nevertheless, the incidence of proliferative retinopathy increases dramatically with the development of elevated urinary albumin/protein excretion.605 A greater frequency of retinopathy with higher levels of urinary albumin/protein excretion is also reported in various racial/ethnic groups (Table 128), and the relationship may be stronger in some groups than in others.606

Other diabetic complications (for example, neuropathy) may be related to the level of proteinuria or albuminuria in diabetic kidney disease (R). Less is known about the strength of the association between urinary albumin/protein excretion and neuropathy than about the other complications of type 1 and type 2 diabetes. The review articles evaluated for this guideline comment briefly that some studies found a relationship whereas others did not.581,591 Much of the confusion is undoubtedly attributable to non-standardized definitions of diabetic neuropathy. In 1988, consensus was achieved on a standardized classification scheme (vide supra), but there are still few reviews available that comment on the relationship between albuminuria/proteinuria and diabetic neuropathy by these criteria.

A large number of published guidelines and position statements are available to guide the practitioner in the prevention, detection, evaluation and treatment of diabetic complications (Table 129).

Guidelines regarding angiotensin-converting-enzyme inhibitors or angiotensin-receptor blockers and strict blood pressure control are particularly important since these agents may prevent or delay some of the adverse outcomes of both kidney and cardiovascular disease (R).


Data in the elderly and in racial/ethnic groups other than non-Hispanic whites are sparse, particularly in review articles, making the findings reported in this guideline difficult to extrapolate with confidence to the elderly and to different populations around the world. Moreover, after the development of kidney failure, much of the available data do not differentiate type 1 from type 2 diabetes.

Clinical Applications

Studies in both type 1617 and type 2618 diabetes indicate that nearly all of the excess mortality associated with diabetes is found in those with elevated urinary albumin/protein excretion. Much of the excess mortality, particularly in type 2 diabetes, is attributable to cardiovascular disease rather than kidney failure, indicating the importance of identifying and treating the other complications of diabetes in these patients and the importance of close monitoring of proteinuria and kidney function to identify those at increased risk. The evidence reviewed to date suggests that the appearance of elevated albuminuria/proteinuria is associated with a higher risk of the non-kidney complications of diabetes even as patients progress towards chronic kidney disease. The association between albuminuria/proteinuria and cardiovascular disease, diabetic retinopathy, and diabetic neuropathy described in this guideline supports the recommendation that patients with diabetic nephropathy be carefully examined for the presence of other diabetic complications and that proper care for these complications be initiated. This recommendation is based on opinion derived from a review of the available evidence.

Implementation Issues

Implementation of coordinated patient management to address the diversity of potential complications is one of the greatest challenges of diabetes care. Recommendations regarding management of diet, exercise, glycemia, blood pressure, lipids, neuropathy, retinopathy, and cardiovascular disease must all be considered in addition to those for kidney disease. Although the challenges for health care providers are formidable, they may seem overwhelming to those with diabetes. One of the objectives of the National Diabetes Education Program, a Program managed jointly by the National Institute of Diabetes and Digestive and Kidney Diseases and the Centers for Disease Control and Prevention, is to promote an integrated patient-centered approach to diabetes care with the goal of reducing the morbidity and mortality associated with diabetes and its complications (www.ndep.nih.gov).

Research Recommendations

Much of the understanding about the relationships between diabetic nephropathy and cardiovascular disease, retinopathy, and neuropathy comes from studies in Caucasians. Yet the epidemic of diabetes affects many racial/ethnic groups worldwide. Since race/ethnicity may influence not only the risk of diabetes, but the severity and type of diabetic complications that develop, further characterization of the impact of diabetes in different populations is needed. Further characterization of these relationships in the elderly is also needed. Moreover, the extent to which aggressive treatment of diabetic complications modulates the progression of kidney disease needs to be examined, since recent studies suggest that improvements in the treatment of cardiovascular disease in patients with type 2 diabetes have contributed to an increase in diabetic kidney failure.619