Risk of Hypoglycemia in Older Adults Stratified by eGFR

Patients with lower estimated glomerular filtration rates (eGFRs) are at higher risk for hypoglycemia, increasing the risk for cardiac and neurologic dysfunction and impaired quality of life. Higher concentration of urinary albumin has been identified as an independent risk factor for the development of hypoglycemia; older age and use of antihyperglycemic medications may also be risk factors.

There are few data on the risk for hypoglycemia by eGFR stage, level of albuminuria, and use of antihyperglycemic medications. Meryl Hodge, BHSc, and colleagues recently conducted a population-based retrospective cohort study designed to test the hypothesis that a graded association in which patients with lower eGFR would have higher risk for a hospital encounter with hypoglycemia. The researchers also expected the association to be evident in patients who were users and nonusers of antihyperglycemic medications, with absolute incidence higher when those medications were used. Study results were reported in the American Journal of Kidney Diseases [2017;70(1):59-68].

Study participants were older adults in Canada (≥65 years of age); the primary outcome of interest was the 3-year incidence rate of a hospital encounter with hypoglycemia, either in the emergency department or as an inpatient. The rate of hypoglycemia was classified according to eGFR state and, in a subpopulation, by eGFR and urine albumin-creatinine ratio (ACR).

The primary cohort included 329,748 patients who were accrued from April 1, 2002, to March 21, 2010, when the first achieved one of the following: (1) evidence of maintenance dialysis therapy, defined a receipt of at least one code for dialysis separated by ≥90 days but <180 days or (2) evidence of stable kidney function, defined by two outpatient eGFR values separated by ≥90 days to less than 1 year within 5 mL/min/1.73 m2 or ≤5% of each other. The index date was the date a person first met enrollment criteria. Exclusion criteria included invalid or missing identification number, age <66 years at the index date, receipt of a kidney transplant, and at least one dialysis code prior to the accrual period.

Patients were stratified by eGFR stage: ≥90, 60 to <90, 45 to <60, 30 to <45, 15 to <30, and <15 mL/min/1.73 m2 or on maintenance dialysis therapy. They were also stratified by use of antihyperglycemic medications; the medications considered were only those available on the provincial formulary at the time of study accrual.

Patients who met the requirement of a single serum creatinine measurement with a urine ACR value in the year prior to or on the creatinine test date were included in the secondary cohort. The exclusion criteria were the same as those for inclusion in the primary cohort. Patients in the secondary cohort were stratified into eGFR and ACR risk categories of low, moderate, high, or very high risk; they were also stratified by use of antihyperglycemic medications.

Among the 329,748 patients in the primary cohort, compared with those with eGFRs ≥90 mL/min/1.73 m2, those with lower eGFRs were older, had higher burden of comorbid conditions, and utilized more medications and other healthcare resources. Among patients in the secondary cohort 9n=201,239), higher risk patients were older, had a higher burden of comorbid conditions, and utilized more medications and other healthcare resources compared with patients in the low risk eGFR and ACR category.

In both users and nonusers of antihyperglycemic medications in the primary cohort, the incidence rate of hypoglycemia per 10,000 person-years increased in a graded manner in patients with lower eGFRs (P<.001). In the secondary cohort, among both users and nonusers of antihyperglycemic medications, there was also an increase in a graded manner in higher eGFR and ACR risk categories in the incidence rate of hypoglycemia per 10,000 person-years (P<.001).

In both cohorts, the 3-year cumulative incidence of hypoglycemia was higher in patients with lower kidney function. There was also a graded relationship seen by ACR stage; the relationship persisted following adjustment for eGFR. The incidence rate of hypoglycemia among patients on maintenance dialysis therapy with diabetes was higher than among patients on maintenance therapy without diabetes (439.68 and 30.43 per 10,000 person-years, respectively; P<.001). Among patients who did not use antihyperglycemic medications, the rate of hypoglycemia was higher among those on hemodialysis than among patients on peritoneal dialysis.

During the 3-year follow-up period, only a small portion of patients had evidence of a recurrent hypoglycemia-associated hospital encounter (0.8% of users of antihyperglycemic medications and 0.01% of non-users across both the primary and secondary cohorts).

Study limitations cited by the authors included the inability to document home episodes of hypoglycemia, not accounting for recurrent hypoglycemic events in calculating incidence rates, and the generalizability of the results being limited to those >65 years of age.

“There was a step-wise increase in risk for hypoglycemia in patients with lower eGFRs. Our clear and quantifiable estimates may help support the patient-provider dialogue and inform future research in this vulnerable population,” the researchers said.

Takeaway Points

  1. Researchers conducted a study to determine the 3-year incidence rates of hypoglycemia in older adults stratified by estimated glomerular filtration rates (eGFRs) and by urine albumin-creatinine ratio (ACR).
  2. In both users and nonusers of antihyperglycemic medications, there was a graded increase in risk for hypoglycemia by eGFR stage. A similar risk relationship was seen in ACR category.
  3. The researchers said their results may help in the patient-provider dialogue and may also inform future studies to prevent hypoglycemia in older adults with reduced kidney function.