and complications trial. A randomized, controlled trial. Ann Intern Med 1998;128:517–523
21. Turner R, Stratton I, Horton V, et al. UKPDS 25: autoantibodies to islet-cell cytoplasm and glutamic acid decarboxylase for prediction of insulin requirement in type 2 diabetes. Lancet 1997;350:1288–1293
22. Pozzilli P, Guglielmi C, Caprio S, Buzzetti R. Obesity, autoimmunity, and double diabetes in youth. Diabetes Care 2011;34(Suppl. 2):S166–S170
23. Hanafusa T, Imagawa A. Fulminant type 1 diabetes: a novel entity requiring special attention by all medical practitioners. Nature Clin Pract Encrinol 2007;3:36–45
24. WHO Expert Consultation. Appropriate body mass index for Asian populations and its implications for policy and intervention strategies. Lancet 2004;363:157–163
25. Norhammar A, Tenerz A, Nilsson G, et al. Glucose metabolism in patients with acute myocardial infarction and no previous diagnosis of diabetes mellitus: a prospective study. Lancet 2002;359:2140–2144
26. Winter WE, Maclaren NK, Riley WJ. Maturity-onset diabetes of youth in black Americans. N Engl J Med 1987;316:285–291
27. Mauvas-Jarvis F, Sobngwi E, Porcher R, et al. Ketosis-prone type 2 diabetes in patients of Sub-Saharan African origin clinical pathophysiology and natural history of b-cell dysfunction an insulin resistance. Diabetes 2004;53:645–653
28. Fajans SS, Bell GI, Polonsky KS. Molecular mechanisms and clinical pathophysiology of maturity-onset diabetes of the young. N Engl J Med 2001;345:971–980
29. Moran A, Dunitz J, Nathan B, et al. Cystic fibrosis-related diabetes: current trends in prevalence, incidence, and mortality. Diabetes Care 2009;32:1626–1631
30. Ode LK, Moran A. New insights into cystic fibrosis-related diabetes in children. Lancet Diabetes Endocrinol 2013;1:52–58
31. McClain DA, Abraham D, Rogers J, et al. High prevalence of abnormal glucose homeostasis secondary to decreased insulin secretion in individuals with hereditary hemochromatosis. Diabetologia 2006;49:1661–1669
32. Templeton AW, Brentnall TA. Screening and surgical outcomes of familial pancreatic cancer. Surg Clin North Am 2013;93:629–645
33. Goncalves P, Araujo JR, Martel F. Antipsychotics-induced metabolic alterations: focus on adipose tissue and molecular mechanisms. Eur Neuropsychlpharmacol 2015;25:1–16
34. Jin H, Meyer JM, Jeste DV. Phenomenology of and risk factors for new-onset diabetes mellitus and diabetic ketoacidosis associated with atypical antipsychotics: an analysis of 45 published cases. Ann Clin Psychiatry 2002;14:59–64
35. Garg A. Acquired and inherited lipodystrophies. N Engl J Med 2004;350:1220–1234
36. Lacube A, Hernadez C, Genesca, Simo R. Glucose abnormalities in patients with hepatitis C virus infection. Diabetes Care 2006;29:1140–1149
Chapter 3
Perils of Glycemic Variability and Rapid Correction of Chronic Hyperglycemia
Susan S. Braithwaite, MD,1 and Irl B. Hirsch, MD2
1Presence Saint Joseph Hospital, Chicago, IL; Presence Saint Francis Hospital, Evanston, IL; West Suburban Medical Center, Oak Park, IL; Westlake Hospital, Melrose Park, IL; Clinical Professor of Medicine, University of Illinois at Chicago, Chicago, IL. 2Professor of Medicine University of Washington Medical Center–Roosevelt, Seattle, WA.
DOI: 10.2337/9781580406086.03
Introduction
The objective of this chapter is to discuss evidence among hospitalized patients supporting or discrediting each of two propositions, in contexts other than hyperglycemic emergencies:
• Glycemic variability, independent of hyperglycemia and hypoglycemia, may causally contribute to risk of harm among hospitalized patients.
• Aggressive correction of chronic hyperglycemia may cause short-term harm for hospitalized patients experiencing uncontrolled diabetes.
Under each of several definitions, glycemic variability, independently from hypoglycemia and hyperglycemia, in the hospital has been recognized to be a risk factor associated with adverse outcomes.1–12 Although supporting evidence has been collected mostly in the intensive care setting, limited data collected in the general hospital setting suggest a similar relationship.10 Glycemic variability is not restricted to patients having preexisting diabetes, whereas rapid correction of chronic hyperglycemia occurs only among patients having preexisting uncontrolled diabetes.1 With respect to predictive value or physiology, it is not clear to what extent glycemic variability actually resembles rapid correction of chronic hyperglycemia. Therefore, we do not classify a one-time hospital-based correction of chronic hyperglycemia as an example of glycemic variability. Present-day therapeutic tools may permit the provider to control the rate of correction of chronic hyperglycemia. Importantly, we have only early evidence suggestive of therapeutic approaches that could reduce glycemic variability in the hospital.
Glycemic Variability
Glycemic Variability and Outcomes in the Inpatient Setting
The goal of this section is to discuss evidence that harms occurring in the hospital are associated with glycemic variability, independent of hyperglycemia and hypoglycemia (Table 3.1). The definition of variability and choice of metrics may determine whether or not a pattern of glycemia within a population is identified as showing increased patient-level glycemic variability.2–12 Intuitively, for this discussion, glycemic variability is understood as a propensity of a single patient to develop repeated episodes of excursions of hyperglycemia or troughs of hypoglycemia over a relatively short period of time that exceed the amplitude expected in normal physiology.13
Table 3.1—Outcomes Associated with Glycemic Variability
*At least 3 BG measurements were required.
See also Braithwaite.13
To characterize or compare groups of patients, it is desirable to choose metrics that will quantify the variability experienced by typical group members. The choice of metrics used to identify variability has been controversial. When standard deviation (SD) or coefficient of variability (CV) is used as a variability metric, all of the data points are used, thus optimizing the power of the metric to make comparisons. For inferential or predictive purposes, SD should be applied to data sets exhibiting Gaussian distribution. Recognizing the predictive value of CV with respect to risk for hypoglycemia, and depending on whether the absolute magnitude of excursions is important or the magnitude of excursions relative to the mean, some authorities favor CV over SD.14,15 The SD is highly correlated with the mean BG.16 In a study of 18,563 patients having myocardial infarction, an association of variability with mortality risk, noted by five metrics in unadjusted analyses, was not upheld after reexamination with adjustment for patient factors, including mean BG.7 Recognizing that a single major change of glycemia during an observation interval can yield a high SD, we favor caution on the use of the SD or CV during intervals of rapidly changing average of blood glucose. Alternative metrics should be considered, such that glycemic variability can be differentiated by the chosen metrics from rapid correction of chronic hyperglycemia.5,13 Nevertheless, some of the earliest and strongest evidence linking variability to outcomes in the critical care setting is based on the use of SD.2,3
During an era in which the inpatient use of continuous glucose monitoring (CGM) has not yet become a standard of care, the study of inpatient glycemic variability has suffered from irregularity of timing and infrequency of monitoring of blood glucose. When applied to CGM, the problem