Hyperglycemia is a fundamental cause of vascular complications of diabetes. When diabetes is complicated by CKD, the risk of cardiovascular events increases significantly.
According to the K/DOQI, diabetic patients who reach CKD stage 3 are more likely to die of cardiovascular disease than to progress to kidney failure. Elevated HbA1c is an independent risk factor, after adjustment for other known risk factors, for diabetic cardiovascular disease. Prospective analysis of data from the UKPDS demonstrated a relation between HbA1c and cardiovascular risk. And data from the Atherosclerosis Risk in Communities Study showed that each percentage point increase in HbA1c throughout the range of levels has been associated with a rise in coronary heart disease and cardiovascular events.
Reduction in HbA1c levels could potentially reduce the risk (Am Heart J. 2006;152:27-38). ADA guidelines have indicated that improved glycemic control may lower the risk of MI and cardiovascular death in diabetes patients.
Lowering the HbA1c may not be adequate to reduce the cardiovascular risk associated with hyperglycemia in patients with advanced diabetes. Whether reaching glycemic goals in type 2 diabetes could prevent cardiovascular events has been addressed in two large randomized double-blind trials in patients at high risk for cardiovascular disease.
Neither the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial (in which HbA1c targets of less than 6.5% and 7%-7.9% were compared in more than 10,000 patients) nor the Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial found any reduction in major cardiovascular events with glycemic control.
A more recent cohort study of type 2 diabetes reported diminished cardiovascular benefit from intense glycemic control in patients with significant comorbidities (Ann Intern Med. 2009;151:854-860). In a soon-to-be released article, Alicic and Tuttle suggest that the disappointing results of these trials compared with the DCCT/UKPDS trials may derive from the inclusion of lower-risk patients, less intense glycemic control targets, and longer duration of follow-up in the earlier studies.
Clinical trials forming the basis of treatment guidelines have excluded patients with significant kidney impairment, and therapeutic targets for HbA1c in advanced stages of kidney disease are uncertain. Nonetheless, USRDS data indicate that use of insulin and oral diabetic drugs has increased in the past few years, suggesting a trend toward more aggressive treatment.
Notably, the relationships between HbA1c and clinical outcomes in diabetic dialysis patients have recently been called into question by complementary observational studies. In the first large national US database analysis, we were unable to demonstrate the expected parallel association between increasing HbA1c levels and mortality risk over one year.
Subsequently, Kalantar-Zadeh et al reported analysis of a second large national ESRD database and found a 41% greater death risk for both all-cause and cardiovascular death in patients with HbA1c levels of 10% or more versus 5%-6% (Diabetes Care. 2007;30:1049-1055).
These studies and others indicate that the relationship between glycemic control and survival outcomes in diabetic ESRD patients appear for the most part to be weak (Am J Kidney Dis. 2008;52:766-777). While the HbA1c level seems to be less precise in ESRD, the more likely explanation is that other comorbidity factors are dominating hyperglycemia in affecting macrovascular disease progression.
Risks of aggressive lowering
Increasing attention is being given to the risks of hypoglycemia (blood glucose less than 70 mg/dL) in the diabetic CKD/ESRD population. Diabetes itself is characterized by acute glucose fluctuations from postprandial peaks to nadirs, and glycemic variability is associated with metabolic alterations and risks of complications (Diabetes Care. 2008;31[Suppl. 2]:S150-S154). The ADA recommends a goal HbA1c of less than 7.0% or as close to normal as possible without unacceptable hypoglycemia.
However, with increasing pressure to achieve tight glycemic control targets, hypoglycemia is becoming more prevalent. In the general diabetic population, increased risk of hypoglycemia has been demonstrated in the ICU patient (JAMA. 2008;300:933-944). Results from the ACCORD trial showed that hypoglycemia was the major risk associated with intensive treatment and that the annualized rate of hypoglycemic episodes requiring medical assistance tripled (3% vs. 1%).
In the ADVANCE trial, severe hypoglycemia was nearly twice as common in the intensive-control as in the standard-control group, although the overall risk was low. Half of patients in the low HbA1c group had at least a minor hypoglycemic event during the study. Patients treated with insulin comprise the group at highest risk, but more information is becoming available on hypoglycemia in type 2 diabetes and in the elderly.
Patients with CKD/ESRD who develop hypoglycemia do not necessarily have diabetes. ESRD unrelated to diabetes, for example, is the second most common cause of hypoglycemia in hospitalized patients and has a high mortality rate. Increased risk of hypoglycemia (and poor defense against it) in CKD stages 3 to 5 is related to reduced renal clearance of insulin and decreased gluconeogenesis by impaired kidneys (during starvation, normal kidneys become a major source of glucose production).
Factors contributing to hypoglycemia in CKD patients include diminished food intake; weight loss; protein-energy wasting; alcohol consumption without food; and, in some cases, medications. Insulin sensitivity is improved by dialysis. In most instances, the term “spontaneous hypoglycemia” is a misnomer in ESRD patients.
The risk of hypoglycemia in ESRD patients is more than doubled by the additional presence of diabetes (Curr Diab Rep. 2009;9:466-472). In diabetic ESRD patients, glucose levels are lower on dialysis days, and the risk of hypoglycemia is greater within 24 hours of dialysis (Diabetes Care. 2009;32:1137-1142).
In the previously noted analysis by Kalantar-Zadeh, up to one-third of prevalent HD patients with diabetes had a normal to low HbA1c level. In another survey, half of diabetic chronic dialysis patients suffered hypoglycemia during a three-month period (Ther Apher Dial. 2009;13:95-102). We recently reported that increased risk of all-cause hospitalization occurred with glycemic variability (deviations of HbA1c greater than one percentage point) in a large database of diabetic ESRD patients. At particular risk were patients whose overall glycemic control was below the conventional HbA1c target of 7%.
There is growing recognition that low HbA1c values may actually be associated with higher mortality rates in diabetic patients on chronic HD. Additional factors contributing to hypoglycemia risk in diabetic CKD/ESRD include too much medication, inadequate or variable carbohydrate intake, and taking medications at the wrong time. Oral agents, such as glyburide, or overtreatment of peritoneal dialysis patients who are on icodextrin exchanges may also be causative.
Little is known about the potential impact of recent industry reductions in dialysate dextrose on provoking hypoglycemia in HD patients. In some ESRD patients, such changes may lead to resolution of hyperglycemia peaks and decrease in HbA1c levels into the normal range, a condition that has recently been termed “burnt-out diabetes.”
Optimal glycemic management in CKD/ESRD remains elusive. The expected benefits of tight glycemic control need to be weighed against the risk involved in the presence of advanced kidney disease. Until more studies lead to evidence-based treatment guidelines targeting diabetic CKD/ESRD patients, a strategy of individualizing the HbA1c target, establishing a risk/benefit “profile,” and less stringent glucose control with increasing patient age, comorbidities, and reduced life expectancy is the best approach.
Glycemic control in the CKD/ESRD patient will be a topic of discussion in April at the National Kidney Foundation 2010 Spring Clinical Meetings in Orlando, Fla. For more information, visit www.nkfclinicalmeetings.org.
Dr. Williams is a clinical investigator and senior staff physician at Joslin Diabetes Center and Co-Director of Dialysis at Beth Israel Deaconess Medical Center, both in Boston.