Abstract Body: Introduction: Severe hypoglycemia (an episode of low glucose requiring assistance) is a rare but serious complication of type 2 diabetes. Diabetes management guidelines for older adults with cognitive impairment recommend less stringent glycemic targets and avoiding insulin and sulfonylureas to reduce the risk of hypoglycemia. However, the risk of severe hypoglycemia according to cognitive status and older adults with diabetes in the general population is poorly characterized.
Hypothesis: Older adults with diabetes and mild cognitive impairment or dementia (vs having diabetes and no cognitive impairment) will (1) frequently use high-risk glucose-lowering medications and (2) have an elevated risk of severe hypoglycemia in a community-based population.
Methods: We conducted a prospective cohort analysis using data from the Atherosclerosis Risk in Communities (ARIC) Study, with visit 5 (2011–2013) as baseline. We limited our population to the 1,632 participants with diabetes (defined by medication use or self-reported physician diagnosis) and valid cognitive status. Severe hypoglycemia events were ascertained from diagnostic codes. We used the nonparametric cumulative incidence function with a competing risk of death to obtain cumulative incidence estimates for severe hypoglycemia according to baseline cognitive status. We used the Fine-Gray subdistribution hazards model to obtain adjusted (for age, race, and sex) associations of cognitive status with the cumulative incidence of severe hypoglycemia.
Results: The study population had a mean age of 75 year (SD, 5), 44% male; and 65% White adults, 7% had dementia, and 25% had MCI. Approximately 50% of participants with dementia or MCI were currently using high-risk glucose-lowering agents (insulin or sulfonylureas). The 5-year risks of mortality were 55%, 26%, and 13% in those with dementia, MCI, or normal cognition, respectively. After accounting for the competing risk of mortality, severe hypoglycemia was strongly associated with dementia and MCI (Figure). These associations persisted after adjustment for age, race, and sex (Figure).
Conclusions: Cognitive impairment was associated with more than a 2-fold increase in the 5-year risk of severe hypoglycemia. In older adults with diabetes, insulin or sulfonylureas was common in individuals with cognitive impairment despite clinical recommendations for discontinuation in most patients. These results suggest the need to improve guideline adherence.