Individualized Glycemic Goals and an Expanded Classification of Severe Hypoglycemia in Diabetes

The relationship between a therapeutic decrease in A1C and the prevention or delay of chronic complications in type 1 diabetes is curvilinear, not linear (2,6,7). An A1C decrease from 11% (97 mmol/mol) to 10% (86 mmol/mol) is associated with a larger decrease in risk, whereas a decrease from 8% (64 mmol/mol) to 7% (55 mmol/mol) is associated with a smaller decrease in risk of chronic complications. Indeed, in a 27-year follow-up of Diabetes Control and Complications Trial (DCCT) patients, a rise in mortality began only with an A1C >9% (75 mmol/mol) (7).

Based on data from 4,306 Danish individuals with clinically diagnosed type 1 diabetes, with a median follow-up of 6.8 years, the Steno type 1 risk engine was used to calculate rate ratios for risk factors for a composite cardiovascular disease outcome (the first fatal or nonfatal cardiovascular disease event: ischemic heart disease, ischemic stroke, heart failure, peripheral artery disease) (8). Compared with risk factors such as macroalbuminuria (rate ratio 2.19) and microalbuminuria (rate ratio 1.55), A1C was a weak predictor (rate ratio 1.14) of cardiovascular disease outcome.

In a prospective study of 389 Danish individuals with type 1 diabetes for at least 20 years who were followed for 13 years, cardiovascular mortality was increased significantly only in those with A1C levels in the fourth (highest) quartile, 9.7% (82 mmol/mol) to 14.0% (129 mmol/mol) (9). Mortality was not increased significantly in those with A1C levels in the second and third quartiles.

In a population-based study of data from 33,915 Swedish individuals with type 1 diabetes compared with 169,249 control subjects without diabetes, with 8 years of follow-up in the patients, there was no difference in cardiovascular (or all-cause) mortality between those patients with a mean updated A1C level of 6.9% (52 mmol/mol) or less and those with a mean updated A1C level of 7.0% (55 mmol/mol) to 7.8% (62 mmol/mol) (10).

The major barrier to glycemic control, indeed often the limiting factor in the glycemic management of diabetes, is iatrogenic hypoglycemia caused by treatment with insulin, a sulfonylurea, or a glinide in the setting of hypoglycemia-associated autonomic failure (3–5,11,12). Iatrogenic hypoglycemia causes recurrent morbidity in most people with type 1 diabetes and many with advanced type 2 diabetes and is sometimes fatal. Notably, in addition to the hypoglycemic mortality rates of 4% to 10% in type 1 diabetes summarized earlier (3,12), it is now known that after 27 years of follow-up 8% of the deaths of patients with type 1 diabetes who had participated in the DCCT were the result of hypoglycemia and that severe hypoglycemia with coma, seizure, or both was significantly associated with death in the DCCT patients (13). Iatrogenic hypoglycemia also generally precludes maintenance of euglycemia over a lifetime of diabetes and, therefore, full realization of the benefits of glycemic control, and it impairs defenses against subsequent hypoglycemia and thus causes a vicious cycle of recurrent hypoglycemia (3,11).

Given the premise that glycemic goals should be individualized in diabetes (3–5), it is important to distinguish between those patients not at risk for hypoglycemia and those patients at risk for hypoglycemia.

A nondiabetic A1C level is reasonable in individuals with diabetes who are not at risk for iatrogenic hypoglycemia, i.e., those treated with lifestyle changes and those treated only with drugs that do not cause hypoglycemia, provided there are no important side effects of such drugs. Glucose-lowering drugs that do not cause hypoglycemia when used as monotherapy or together without insulin, a sulfonylurea, or a glinide include metformin, dipeptidyl peptidase 4 inhibitors, glucagon-like peptide 1 receptor agonists, sodium–glucose cotransporter 2 inhibitors, thiazolidinediones, and α-glucosidase inhibitors (14).

In those individuals with diabetes who are at risk for iatrogenic hypoglycemia because of treatment with insulin, a sulfonylurea, or a glinide, a reasonable glycemic goal is the lowest A1C that does not cause severe hypoglycemia and preserves awareness of hypoglycemia, preferably with little or no symptomatic or even asymptomatic hypoglycemia, at a given stage in the evolution of the individual’s diabetes (3). It has been suggested that an A1C range of 7.5% (58 mmol/mol) to 9.0% (75 mmol/mol) would maximize benefit and minimize harm in older individuals (5). Others have suggested an A1C range of 7.5% (58 mmol/mol) to 8.5% (69 mmol/mol) (15). A pragmatic approach to minimizing hypoglycemia in diabetes has been published (16). A less stringent (higher A1C) goal is appropriate for those who have previously experienced hypoglycemia or have potential high risk for hypoglycemia. The most clear example of the former is an individual with a history of severe hypoglycemia (that requiring the assistance of another person) (17). The most clear example of the latter is an individual with impaired awareness of hypoglycemia (hypoglycemia unawareness), which may be reversed by as little as 2 to 3 weeks of scrupulous avoidance of hypoglycemia (16). A less stringent glycemic goal is also appropriate for individuals with a long duration of diabetes because that is associated with an increased risk of hypoglycemia, presumably the result of further reduced insulin secretion (11), and for those with a short life expectancy due to age, severe vascular complications, or severe comorbidities because a long duration of glycemic control is required for demonstrable mortality benefit (4,5). The American Diabetes Association recommends an A1C level <8.5% (69 mmol/mol) in such patients (18). A conservative level would be <8.0% (64 mmol/mol) (10).

Given the importance of a history of severe hypoglycemia in the selection of glycemic goals in individuals with diabetes at risk for hypoglycemia, just discussed, it is important to address the limitations of the current classification of severe hypoglycemia. Since the DCCT, severe hypoglycemia has been considered a hypoglycemic event requiring the assistance of another person (17). That is a useful criterion that should be retained, although it is often not applicable to young children because they generally need assistance for otherwise nonsevere hypoglycemia. But it does not go far enough.

An A1C level provides information about mean glycemia over the previous 2 to 3 months. However, it provides no information about changes in glycemia in the short term or about threatening episodes of documented hypoglycemia. Thus, there is a need for an additional, measurable outcome for people with diabetes who are at risk for iatrogenic hypoglycemia (11,19). To that end, it is proposed to expand the classification of severe hypoglycemia in diabetes by adding a measured glucose concentration <50 mg/dL (2.8 mmol/L) as a second criterion to the current criterion (17), a hypoglycemic event requiring the assistance of another person. Thus, either a hypoglycemic episode requiring assistance or a measured dangerously low glucose concentration would constitute severe iatrogenic hypoglycemia. (It is not proposed to change the classification of nonsevere hypoglycemia [17].)

A plasma glucose concentration of 50 mg/dL (2.8 mmol/L) is a dangerously low level that does not occur under physiological conditions in individuals without diabetes (20). Thus, it is an unequivocally hypoglycemic value. It attenuates glucose counterregulatory defenses—including the sympathoadrenal and symptomatic defenses—against subsequent hypoglycemia in individuals with and without diabetes (11). Thus, it causes defective glucose counterregulation and impaired awareness of hypoglycemia, the components of hypoglycemia-associated autonomic failure in diabetes. In insulin- or sulfonylurea-treated type 2 diabetes with cardiovascular disease, hypoglycemic events with a continuous glucose monitoring glucose concentration <56 mg/dL (3.1 mmol/L) were associated with a 30-fold increased frequency of Holter monitor–detected runs of ventricular tachycardia (21). Importantly, in randomized controlled clinical trials glucose levels in the range of 50 mg/dL (2.8 mmol/L) were associated with increased mortality (22–24). Because recent antecedent hypoglycemia shifts glycemic thresholds for symptoms of hypoglycemia to lower plasma glucose concentrations, it is not unusual for people with diabetes to have no or few symptoms despite a distinctly low glucose level (11,14,25,26). Detection of a dangerously low glucose concentration with no symptoms is a critically important finding because it is indicative of impaired awareness of hypoglycemia, a treatable risk factor for conventionally defined severe iatrogenic hypoglycemia (14,16) caused by an even lower glucose concentration. Clearly, a glucose concentration <50 mg/dL (2.8 mmol/L) is indicative of severe hypoglycemia even if assistance by another person is not required. Thus, it is proposed to expand the classification of severe hypoglycemia in diabetes to include, in addition to a hypoglycemic event requiring the assistance of another person (17), a history of a measured glucose concentration <50 mg/dL (2.8 mmol/L)—a level associated with sudden death and therefore a level that should be scrupulously avoided in the future. The latter association is why I prefer a glucose concentration <50 mg/dL (2.8 mmol/L) as the second criterion for severe hypoglycemia. One could, however, use a slightly higher glucose concentration. For example, the International Hypoglycemia Study Group (27) concluded that a glucose concentration <54 mg/dL (3.0 mmol/L) is sufficiently low to indicate serious, clinically important hypoglycemia in diabetes. That glucose level was also associated with mortality in some studies (23,24), and the above comments about a glucose concentration of 50 mg/dL (2.8 mmol/L) or less generally also apply to one of 54 mg/dL (3.0 mmol/L) or less (27).

This proposed expansion of the classification of severe hypoglycemia in diabetes to add an objective, measurable, and more common dangerously low plasma glucose concentration criterion to that of a need for assistance from another person would permit a diagnosis of severe iatrogenic hypoglycemia in a broader group of persons at risk for dangerous hypoglycemia, including young children who generally need assistance for otherwise nonsevere hypoglycemia.

Duality of Interest. This manuscript was prepared without external support or assistance. The author has served as a consultant to Novo Nordisk in the past year. No other potential conflicts of interest relevant to this article were reported.