Antipsychotics, delirium and glucose in older patients

Antipsychotic drugs are widely used in older patients to reduce psychotic and behavioral symptoms in delirium. It has been suggested that cardiometabolic adverse effects of antipsychotic drugs contribute to the increased mortality risk seen in older patients using these drugs. Both delirium and antipsychotic drug use have been associated with disturbances in glucose levels, but the interplay between antipsychotic treatment, delirium, and glucose has not been clarified. This thesis aimed to answer whether antipsychotic treatment is associated with alterations in glucose levels, whether glucose variability is associated with the onset of delirium and whether glucose variability is altered during delirium in older patients. We have shown that antipsychotic drug use was associated with an increased risk of hospitalization for hypoglycemia in older patients with diabetes, especially in the first 30 days of treatment and with higher doses. We showed that haloperidol use was not associated with an increased risk of hyperglycemia or hypoglycemia in hospitalized older patients. However, hospital initiated haloperidol was associated with hyperglycemia. Our results suggest that closer monitoring of blood glucose levels may be indicated after starting antipsychotic drugs in diabetic older patients or after starting haloperidol in a hospital setting. Furthermore, we showed in a prospective, observational cohort study, which was prematurely ended due to recruitment problems, that one of the two non-diabetic patients with delirium in our study had the highest measured glucose variability and the most hyperglycemia episodes. Our investigation supports the hypothesis that delirium is associated with high glucose variability. But it also illustrates the difficulties encountered when conducting research involving older patients. We have shown that diabetes was not associated with Intensive Care Unit (ICU) delirium. Hypoglycemia did not significantly increase the risk of transition to ICU delirium. Only in patients without diabetes the occurrence of hyperglycemia and the occurrence of both hyperglycemia and hypoglycemia on the same day was associated with transition to ICU delirium. Finally, we have demonstrated that mean glucose concentrations, its standard deviation, mean absolute glucose change, daily delta and the risk of hyperglycemia were unaltered during delirious days compared to non-delirious days in non-diabetic and diabetic patients at the ICU. Furthermore, delirium was associated with hypoglycemia in patients with diabetes. The investigations conducted in this thesis increased our knowledge of the interplay between antipsychotic drugs, delirium, and glucose homeostasis in older outpatients, hospitalized older patients admitted to non-ICU wards, and ICU patients. The findings for ICU patients described in this thesis support the hypothesis that delirium acts on glucose homeostasis and vice versa. The interplay between delirium and glucose was different for patients with and without diabetes, indicating another underlying mechanism. Glucose dysregulation should be added to the list of adverse events associated with antipsychotic treatment in older patients. While the absolute risk of glucose dysregulation during antipsychotic use seems small, its impact in the frail older population may be huge.