Increased glucose variability is associated with an increased length of hospitalization and short- and long-term mortality in hospitalized adults with and without diabetes, according to findings published in The Journal of Clinical Endocrinology & Metabolism.
Amit Akirov, MD, of the Institute of Endocrinology, Rabin Medical Center-Beilinson Hospital in Israel, and colleagues evaluated historical prospectively collected data on 20,303 adults (mean age, 70 years) hospitalized from 2011 to 2013 to determine the association between glucose variability, length of stay and mortality. Follow-up was a median 1,022 days, and 42% of participants had a diagnosis of diabetes.
Coefficient of variance and standard deviation (SD) of glucose values during hospital were used to assess glucose variability.
Participants without diabetes had a longer length of stay (mean, 7 days) compared with participants with diabetes (mean, 6 days; P < .001). Longer length of stay was associated with higher SD and coefficient of variance in all participants. Coefficient of variance of glucose significantly interacted with length of admission (P < .0001), and mean length of admission was longer in participants with higher coefficient of variance of glucose in all participants. Similarly, longer hospital stay in all participants was associated with increased SD tertile.
In-hospital mortality was higher in participants without diabetes (9%) compared with participants with diabetes (7.1%). Increased hospital mortality was associated with higher SD tertile and coefficient of variance tertile in all participants.
Length of stay and in-hospital mortality were both independently affected by glucose levels during hospitalization and glucose variability.
“Our findings suggest that increased [glucose variability], whether in patients with or without pre-existing [diabetes], is associated with longer length of hospitalization and increased short- and long-term mortality risk,” the researchers wrote. “Mortality risk is higher in patients with increased [glucose variability] and high median glucose values during the hospitalization, with more significant differences in patients without [diabetes], compared to patients with [diabetes]. Randomized controlled interventional trials are needed to investigate the impact of minimizing [glucose variability] on morbidity and mortality.” – by Amber Cox
Disclosure: The researchers report no relevant financial disclosures.