ABSTRACT
AIMS: Hospital complications and hyperglycemia are common in elderly patients during hospitalization. Our aim was to analyze the relationship between hyperglycemia and hospital complications in an ageing population. METHODS: We conducted an observational study to evaluate the association between maximum blood glucose (MBG) levels and hospital complications. Patients were stratified according to the quartiles of MBG levels. Diabetes mellitus (DM) was determined by patient history and/or admission glycated hemoglobin (HbA1c) level ≥6.5%. Hyperglycemia in patients without DM was defined as stress-induced hyperglycemia (SH). The composite primary end-point included frequent complications and/or all-cause hospital mortality. RESULTS: Among 461 patients, mean age 80±7.5years, 238 (51.6%) patients had DM, 20 had undiagnosed DM, and 162 (35.1%) developed hospital complications. Patients with complications had higher mean daily BG levels (215±84 vs 195±85mg/dl, P<.01). The incidence of complications was directly associated with severity of hyperglycemia according to the quartiles of MBG levels in patients without DM, namely SH (<140 mg/dl, 22.2%; 140-185mg/dl, 40%; 186-250mg/dl, 47%; >250mg/dl, 60%; P=.002), but not in patients with DM (<140mg/dl, 26.3%; 140-185mg/dl, 40.4%; 186-250mg/dl, 35.6%; >250mg/dl, 37.4%; P=.748). In the multivariate analyses, SH was independently associated with complications: OR 2.60 (CI95%: 1.2-5.6), 2.82 (CI95%: 1.2-6.5), 5.50 (CI95%: 1.4-20.8) for the second, third and fourth quartile respectively (P=.01), as compared to the first quartile. We found no association with readmissions and all-cause mortality. CONCLUSIONS: SH in elderly patients is associated with hospital complications, but not with all-cause mortality, compared to patients with diabetes or normoglycemia.
