The Association between Glycemic Control in People with Diabetes and Symptoms at the End of Life: A Prospective Observational Consecutive Cohort Study

Background: Diabetes is increasingly prevalent globally, including in palliative care. Guidelines vary as to the ideal glycemic goals for patients near the end of life. The relationship between hyperglycemia and attributable symptoms late in life remains ill defined. Objective: To pilot the association between blood glucose level (BGL) and symptoms (nausea, fatigue, pain, and appetite) and mortality in palliative care patients with diabetes. Design: This prospective observational consecutive cohort study consisted of 17 patients with diabetes admitted to an inpatient palliative care unit. Repeat measures of BGL and symptom distress scores using the patient-reported Symptom Assessment Scale (SAS) were recorded during a five-month period as was patient mortality. The association between BGL and SAS domains was assessed using negative binomial regression and the association between mortality and high versus low BGL was determined using log-rank statistics and Kaplan-Meier curves. Results: All patients had malignancy: 15 had type 2 diabetes and 2 had steroid-induced diabetes. A total of 121 patient observation days were included in the analysis. BGL was inversely associated with patient-reported SAS for nausea (incidence rate ratio [IRR] = 0.83, 95% confidence interval [CI] = 0.70-0.99, p = 0.04), but not other symptoms. Insulin usage was also associated with decreased nausea (IRR = 0.24, 95% CI = 0.09-0.60, p = 0.002). Survival did not differ between low- and high-BGL groups. Conclusion: These findings warrant a larger multisite consecutive cohort study and a re-exploration of current clinical practice. Ultimately, interventional trials comparing strict versus more liberal glycemic control on symptom management and survival are the ideal design to better understand differing levels of glycemic control at the end of life.