A minority of T1D occurs without evidence of autoimmunity as in the cause of the insulin deficiency, which nonetheless can be profound and develop rapidly. This is referred to as idiopathic diabetes, previously called “type 1b diabetes.” Included in the category of idiopathic diabetes is “fulminant type 1 diabetes,”23 most commonly described in Asian patients24 and usually presenting after a viral infection or during pregnancy. Onset is acute, usually with DKA despite HbA1c levels that are near normal because of the rapid onset of the hyperglycemia. Of special relevance to treating health-care providers in the emergency room and hospital, death from DKA may occur within 24 h if insulin therapy is not initiated immediately upon presentation.23
Type 2 Diabetes
T2D accounts for the majority of diabetes in the world, accounting for ~90% of all cases. Uncontrolled hyperglycemia in T2D often goes undiagnosed for many years because of the absence of symptoms or presence of vague symptoms. In the hospitalized setting, DKA can occur, but it is almost always associated with stress of another illness, such as infection or ischemia. Undiagnosed diabetes is particularly common in patients admitted for myocardial infarction.25 An inpatient admission can be an important opportunity for diagnosis and initiation of treatment in such high-risk individuals. Even patients with previously well-controlled T2D usually will require discontinuation of prior diabetes oral agents and noninsulin injectable therapies during hospitalizations and treatment with IV or subcutaneous insulin therapies to optimally and safely control hyperglycemia.
Individuals with T2D are resistant to insulin and have relative, as opposed to absolute, insulin deficiency. Over time, they can become profoundly insulin deficient. Depending on the individual and the situation, insulin and c-peptide levels may be high, normal, or low. Autoimmune destruction of the β-cells does not occur, and patients are antibody negative. The risk of developing T2D increases with age, obesity, sedentary lifestyle, and positive family history. It occurs more frequently in women with previous gestational diabetes and polycystic ovarian syndrome; postmenopause, it is seen in individuals with dyslipidemia and hypertension; and it is seen in many ethnic groups (African American, Native American, Hispanic, Pacific Islander, and Asian American). Approximately 85% of individuals with T2D are obese or overweight, and those who are not obese often have an increased percentage of body fat distributed in the abdominal region. The obesity definition is ethnicity-related. For example, although Caucasians are considered obese with a BMI >30 kg/m2, the World Health Organization defines obesity <30 kg/m2 for most Asians and the recent ADA standards changed the BMI cut point for screening overweight Asian Americans for T2D to 23 kg/m2, from the previous 25 kg/m2.5
Multiple genetic mutations have been associated with T2D but in clinical practice it is not possible to identify a specific genetic abnormality. One specific example of a poorly understood genetic entity of T2D is atypical diabetes, also called Flatbush diabetes. This is a ketosis-prone diabetes, initially described in African Americans who presented with DKA, but the subsequent disease course more closely resembles classic T2D.26 The underlying pathogenesis is unclear, but studies have shown a transient secretory defect of β-cells at the time of presentation with remarkable recovery of insulin-secretory capacity.27 Ketosis-prone diabetes also has been described in other ethnicities.
Monogenic Diabetes Syndromes
These patients represent a small fraction of diabetes (<5%), which is the result of a single genetic defect and generally presents before the age of 25 years. They usually are negative for the antibodies commonly found in T1D. The two main subtypes of monogenic diabetes are neonatal diabetes and maturity onset diabetes of youth (MODY). When diagnosed within the first 6 months of life, diabetes is called neonatal diabetes, which is not a form of T1D. The diabetes in these neonates may be transient or permanent, with the latter most commonly having a mutation on the gene encoding the Kir6.2 subunit of the β-cell KATP channel. Despite the early onset, these individuals can be well managed with sulfonylureas instead of insulin as children and, if eventually diagnosed later, even as adults. For this reason, all patients diagnosed with diabetes before the age of 6 months should have genetic screening for neonatal diabetes, even if the history of age of onset is discovered decades later.
MODY is a heterogeneous group of antibody-negative, autosomal-dominant inherited, youth-onset disorders of the β-cell. MODY is characterized by impaired insulin secretion but no (or minimal) defects in insulin action. To date, six different gene mutations are identified on different chromosomes, each one resulting in a different clinical entity.28 One of these genes encodes the enzyme glucokinase (associated with MODY 2), whereas the other five loci encode transcription factors. The most common is hepatic nuclear factor 1-a (associated with MODY 3). Differentiating MODY from T1D is important given the autosomal-dominant inheritance of the former and the observation that many of these patients can be controlled with sulfonylureas. Furthermore, those with glucokinase MODY generally need no therapy (except during pregnancy). Specialty commercial lab testing is now available to identify the gene mutations for clinical (nonresearch) diagnosis to accurately diagnose youth who have MODY and to facilitate prompt identification of other potentially affected family members.
Pancreatic Diabetes
Many diseases of the pancreas affect endocrine function. Formally termed “pancreatic diabetes,” the etiologies, degree of insulin sensitivity, and the subsequent risk for hypoglycemia vary.
For example, cystic fibrosis (CF) patients have a high frequency of diabetes called cystic fibrosis–related diabetes (CFRD). This occurs in 40–50% of adult patients with CF.29 Insulin sensitivity is generally normal or only slightly decreased except in the setting of acute illness when insulin resistance can be severe.30 The frequent intervention of lung transplantation requiring use of corticosteroids and calcineurin inhibitors further increases the prevalence of diabetes in this population. Conversely, the diabetes secondary to chronic pancreatitis mainly occurs from the destruction of islet cells by pancreatic inflammation. There is also an idiopathic variety of chronic, calcific pancreatitis associated with malnutrition that has been termed “tropical chronic pancreatitis.” Both of these latter forms of chronic pancreatitis also are associated with glucagon deficiency and thus more marked sensitivity to insulin with increased risks of hypoglycemia associated with insulin therapy.
Hereditary hemochromatosis is another etiology of pancreatic diabetes, with up to 23% of hemochromatosis patients in one study diagnosed with diabetes.31
Surgical pancreatectomy is particularly common in hospitals with busy oncology centers. Immediately after surgery, insulin deficiency in these patients can be reasonably easy to control with basal-bolus insulin therapy. These patients, however, are extremely prone to the risk of devastating hypoglycemia because of glucagon deficiency, and reducing insulin dosing by 20–25% at discharge from the hospital is suggested.
Ideally, patients scheduled for pancreatectomy should meet with a diabetes team or endocrinologist before the surgery and focused diabetes education is needed after this procedure. Because the pancreatectomy is due to a malignancy (or in some cases pancreatic dysplasia32), the diabetes often is not the main focus of the patient or the family. Especially for those patients who have a good prognosis, the importance of good glycemic control and avoidance of hypoglycemia, even more so than the newly diagnosed patient with T1D, should be stressed. The glucagon deficiency, besides leading to abnormal glucose counterregulation, also results in overall increased insulin sensitivity, meaning extra precaution is required for exercise. Although exact data are not available, there are anecdotal reports of death resulting from hypoglycemia in this population possibly related to physical activity, lack of blood glucose testing, alcohol, or a combination. Continuous glucose monitoring should be strongly considered for these patients.
Drug-Induced Diabetes
Many drugs are known to cause diabetes. Some drugs, such as streptozotocin and IV pentamidine can permanently destroy pancreatic β-cells. More commonly, the ubiquitously useful glucocorticoid therapies (including IV, oral, intra-articular,