Hypoglycemic-Related Hospitalizations Associated with Increased Post-ESRD Mortality

Thirty to 40% of patients with diabetes mellitus will develop chronic kidney disease (CKD), and 44% of cases of incident end-stage renal disease (ESRD) in the United States are associated with diabetic kidney disease. The annual mortality rate of diabetic patients with ESRD is greater than among those with other causes of ESRD, creating a need to identify modifiable risk factors in patients with diabetic ESRD, particularly in the early phases of transitioning from non-dialysis dependent CKD to ESRD requiring maintenance dialysis, when a mortality peak has been observed.

There are few definitive data on optimal glycemic control in the management of diabetic patients with CKD. Early studies found a lowering of risk for microvascular complications with intensive glycemic control; however, contemporary studies have seen a lack of benefit as well as adverse cardiovascular outcomes in patients with long-standing type 2 diabetes. There are also few data on the risk factors and sequelae of hypoglycemia in diabetic patients with kidney disease, particularly in patients progressing to ESRD.

Connie M. Rhee, MD, MSc, and colleagues conducted an observational cohort study designed to examine the relationship between pre-ESRD hypoglycemia events during the predialysis transition period and post-ESRD mortality in a population of US veterans with diabetes. The researchers also sought to identify the clinical characteristics associated with the risk for hypoglycemia in the period prior to transition to dialysis. Results of the study were reported in the American Journal of Kidney Diseases [2018;72(5):701-710].

The study population included patients from the Transition of Care in CKD study, a retrospective cohort study that examined US veterans transitioning to ESRD from October 1, 2007, to September 30, 2011. The outcome of post-ESRD all-cause mortality was examined relative to pre-ESRD hospitalizations related to hypoglycemia. The outcome of pre-ESRD hypoglycemia-related hospitalization was examined relative to antidiabetic medication regimens.

Of the 30,156 patients who met eligibility criteria, 5.9% experienced one or more hospitalization(s) related to hypoglycemia prior to initiating dialysis therapy; 94.1% had 0 hospitalizations; 5.1% had one hospitalization; 0.6% had two hospitalizations; and 0.2% had three or more hypoglycemia-related hospitalizations. Of the patients hospitalized for hypoglycemia, 0.1%, 98.3%, and 1.6% had type 1, type 2, and unknown diabetes type, respectively. The overall mean and median estimated glomerular filtration rate (eGFR) slope was –10.5 and –8.2 mL/min/1.73 m2 per year, respectively

Patients with any hypoglycemia-related hospitalizations tended to be younger, less likely to be white, more likely to be receiving in-center hemodialysis, and had a higher Charlson comorbidity index score. They were also more likely to have congestive heart failure or cerebrovascular disease, had higher values of hemoglobin A1c (HbA1c) and prelude-averaged eGFR, and lower serum albumin. They were more likely to be using insulin (including rapid-, short-, intermediate-,and long-acting types) and oral antidiabetic drugs (including sulfonylureas, thiazolidinediones, and metformin) compared with patients with no hypoglycemia-related hospitalizations.

Results of case-mix, logistic regression analyses found a higher likelihood of hypoglycemia-related hospitalization(s) among patients of Hispanic ethnicity, those with congestive heart failure or cerebrovascular disease, higher HbA1c, and prelude averaged eGFR values, and those using rapid-, short-, intermediate-, and long-acting insulin. There was a lower likelihood of hypoglycemia-related hospitalization(s) among patients of white race and those with higher serum albumin levels. Following adjustment for additional covariates, the patterns of association largely persisted.

In case-mix analyses in a subcohort of 28,047 patients who were prescribed zero to two oral antidiabetic drugs with or without insulin, those who were prescribed an increasing number of oral antidiabetic medications without insulin had an incrementally higher risk for hypoglycemia-related hospitalization(s) compared with those in a reference group who were prescribed neither oral antidiabetic drugs nor insulin.

In comparison with the same reference group, patients prescribed insulin had higher risk for hypoglycemia-related hospitalization(s); the risk was even higher among patients concomitantly prescribed one or two oral antidiabetic drugs. The pattern of findings persisted following adjustment for expanded case-mix plus laboratory covariates. Compared with the same reference group, among patients whose medication regimens included sulfonylureas and/or insulin, there was an association with higher risk for hypoglycemia.

In comparisons of patients with hypoglycemia-related hospitalization(s), using patients with no hypoglycemia-related hospitalizations as the reference group, there was an independent association between incrementally higher mortality risk and hypoglycemia-related hospitalization(s) (adjusted hazard ratios, 1.21, 95% confidence interval [CI], 1.12-1.30; 1.47, 905% CI, 1.19-1.82; and 2.07, 95% CI, 1.46-2.95 for one, two, and three or more hypoglycemia-related hospitalizations, respectively).

The researchers cited some limitations to the study, including defining hypoglycemic events using hospitalization records, perhaps excluding mild episodes that did not require hospitalization; prescription data may not have reflected the actual intake of medications; lack of assessment of the relationship between hypoglycemia and short-term death; and no indication of a causal association between hypoglycemia-related hospitalization(s) and mortality.

“In conclusion, we observed that hypoglycemia-related hospitalization(s) during the pre-ESRD prelude period were associated with higher post-ESRD mortality among diabetic incident patients with ESRD, and this risk was incrementally stronger with increasing frequency of hypoglycemia. Further studies are needed of management strategies that prevent hypoglycemia episodes among diabetic patients with NDD-CKD transitioning to dialysis therapy to evaluate whether reducing such episodes reduces the exceedingly high death rate in ESRD,” the researchers said.

Takeaway Points

  1. Researchers conducted an observational cohort study to determine the relationship between hypoglycemia-related hospitalizations in the period leading to the transition to dialysis therapy and post-end stage renal disease (ESRD) mortality.
  2. The cohort included US veterans with diabetes and chronic kidney disease who transitioned to dialysis therapy from October 1, 2007, to September 30, 2011; data were obtained from the national Veterans Affairs database.
  3. There was a dose-dependent relationship between the frequency of pre-ESRD hypoglycemic-related hospitalizations and post-ESRD mortality in the population of incident patients with ESRD with diabetes.