Hypoglycemia in Type 2 Diabetes

Pathophysiology, frequency, and effects of different treatment modalities

  1. Nicola N. Zammitt, MRCP and
  2. Brian M. Frier, MD
  1. From the Department of Diabetes, Royal Infirmary of Edinburgh, Edinburgh, Scotland, U.K
  1. Address correspondence and reprint requests to Prof. B.M. Frier, Department of Diabetes, Royal Infirmary of Edinburgh, 51 Little France Crescent, Edinburgh EH16 4SA, Scotland, U.K. E-mail: brian.frier{at}luht.scot.nhs.uk

The importance of strict glycemic control to limit the risk of diabetic vascular complications is indisputable, but many barriers obstruct its attainment. Hypoglycemia is recognized to be a major limitation in achieving good control in type 1 diabetes (1) but has been considered to be a minor problem of the treatment modalities used for type 2 diabetes (2). This may be a misperception based on inadequate information. The burden of covert hypoglycemia associated with oral antidiabetic agents may be underestimated, and with the increasing use of insulin to treat type 2 diabetes, the actual prevalence of hypoglycemia is likely to escalate.

The frequency and pathophysiology of hypoglycemia in type 2 diabetes and the relationship to different therapies was reviewed by conducting a literature search using the bibliographic database PubMed to identify publications in English from 1984 until 2005 related to hypoglycemia associated with treatment of type 2 diabetes, and the bibliographies of relevant articles were scrutinized for additional citations. Search terms included “type 2 diabetes,” “NIDDM,” “non-insulin-dependent diabetes,” “hypoglycemia,” and “hypoglycaemia.”


Normal physiological responses to hypoglycemia

The human brain primarily uses glucose as its source of energy. Under normal conditions, the brain is unable to synthesize or store glucose and is exquisitely vulnerable to glucose deprivation. To protect the integrity of the brain, several physiological mechanisms have evolved to respond to and limit the effects of hypoglycemia (36).

In humans, the initial response to a decline in blood glucose is suppression of endogenous insulin secretion followed by release of counterregulatory hormones, of which glucagon and epinephrine (adrenaline) are the most potent. When blood glucose falls in a nondiabetic adult, the secretion of counterregulatory hormones and the onset of cognitive, physiological, and symptomatic changes occur at reproducible blood glucose thresholds (4,7) within a defined hierarchy (5) (Fig. 1). Subjective recognition of the symptoms of hypoglycemia is fundamental to effective self-management and to prevent progression in severity (9,10). Symptoms are generated at arterialized blood glucose concentrations around 2.8–3.2 mmol/l (50–58 mg/dl) and in young adults have been classified as neuroglycopenic, autonomic, and malaise (11). Hypoglycemic symptoms are idiosyncratic and age specific (10).

The effects of ageing on the responses to hypoglycemia

Despite the increasing incidence of type 2 diabetes in young people, this condition is primarily associated with advancing age. It is pertinent therefore to consider the specific effects of ageing on the responses to hypoglycemia before examining how type 2 diabetes affects these processes.

With increasing age, the symptoms of hypoglycemia may become less intense (12,13) and the symptom profile is modified (9,12,14). In a small British study (12) that compared responses to hypoglycemia in seven (five male) nondiabetic adults, aged 65–80 years, with six (three male) younger people, aged 24–49 years, the symptom scores were significantly lower in the older group; autonomic and neuroglycopenic symptoms were affected equally. A Canadian study (15) comparing symptom responses to hypoglycemia in 10 young (5 male, aged 25–30 years) and 9 older nondiabetic subjects (5 male, aged 67–84 years) implicated attenuation of autonomic activation as the cause of the diminished symptomatic response. A further study (16) by the same group reported that symptomatic responses were similar in 10 elderly people with (7 male, aged 72 ± 1 years) and 10 without (6 male, aged 74 ± 1 years) type 2 diabetes, suggesting that the decreased symptom intensity observed in their first study was associated with increasing age, independent of any effects of diabetes.

In young adults, symptomatic responses to hypoglycemia are generated at a blood glucose level that is higher than the level at which cognitive function becomes impaired. This allows sufficient time to take corrective action before severe neuroglycopenia supervenes (5). The difference between these glycemic thresholds is ∼1.0 mmol/l (18 mg/dl). In a study comparing seven healthy older men (aged 65 ± 3 years) with seven younger male control subjects (aged 23 ± 2 years), symptoms and cognitive dysfunction occurred almost simultaneously at 3.0 ± 0.2 mmol/l (54 ± 4 mg/dl) in the older subjects (13) (Fig. 2). The juxtaposition of these thresholds may limit the time available to self-treat and increase the risk of developing incapacitating neuroglycopenia.

Does the ageing process modify the counterregulatory hormonal responses to hypoglycemia? Early studies (1721) that compared these responses in young and older nondiabetic adults yielded conflicting results, and many of the participants had comorbidities that confounded interpretation of the data. Hypoglycemia was induced mostly with an intravenous bolus injection of insulin, producing variable degrees of hypoglycemia, and principally examined cortisol and growth hormone responses (1721), so these results are of limited value.

A modest attenuation of blood glucose recovery from hypoglycemia may occur in older nondiabetic adults, in whom the rise in plasma epinephrine is slower than in younger subjects (22). The glycemic thresholds for the secretory responses of glucagon and epinephrine to hypoglycemia occurred at a blood glucose of 3.3 mmol/l (59 mg/dl) in 10 young nondiabetic adults (5 male, aged 25–30 years) compared with 2.8 mmol/l (50 mg/dl) in 9 older adults (5 male, aged 67–84 years) (15).

With advancing age, the magnitude of the counterregulatory response may be determined by the hypoglycemic nadir. In clamp studies comparing elderly and young nondiabetic subjects aged (means ± SE) 65 ± 1 and 24 ± 1 years, respectively, the magnitude of the glucagon and epinephrine responses was lower in the elderly group during mild hypoglycemia (blood glucose 3.3 mmol/l; 59 mg/dl), but in both age-groups equivalence was achieved at a lower blood glucose (2.8 mmol/l; 50 mg/dl), indicating preservation of these responses to more profound hypoglycemia (23). The rate of insulin clearance from the circulation declines with age (2426), which may enhance the risk of hypoglycemia in elderly people.

If symptomatic and counteregulatory responses to hypoglycemia are modified by advancing age, it is not known at what age this occurs and whether the effects differ in male and female subjects. Considerable biological variability between individuals may occur in nondiabetic and diabetic adults with respect to the effects of ageing. This introduces a potentially confounding variable in studies of people with type 2 diabetes, which is not only a heterogeneous disorder but affects a wide age range. Few investigations of hypoglycemia in type 2 diabetes have taken age into account when designing studies and analyzing results, and few studies have included patients aged >70 years (Tables 1 and 2).

The effects of type 2 diabetes on the responses to hypoglycemia

Counterregulatory responses to hypoglycemia have been investigated less systematically in type 2 than in type 1 diabetes (1,27,28). Although various counterregulatory hormonal deficiencies have been described in type 2 diabetes, these were mostly mild, and epinephrine secretion was invariably preserved. Interpretation of the early studies is limited by heterogeneity of study design (29), differences in blood glucose nadir between the diabetic and control groups (30,31), a lack of age-matched control subjects (31), and the disparate methods used to induce hypoglycemia (3035).

Three studies of counterregulatory responses to hypoglycemia in people with type 2 diabetes, treated with either diet or oral medication, have shown that counterregulatory hormonal release occurs at higher blood glucose levels than in nondiabetic control subjects (36,37) and people with type 1 diabetes (38). In one of these studies, the influence of glycemic control on the counterregulatory response to hypoglycemia was assessed in 11 subjects (9 male) with type 2 diabetes (aged 56 ± 7 years), who were either diet treated or were taking sulfonylureas, and compared with 10 subjects (5 male) with type 1 diabetes (aged 27 ± 6 years) and 2 nondiabetic control groups matched for age and body weight (38). Hypoglycemia was induced using a stepped glucose clamp. Counterregulatory hormones were secreted at higher blood glucose levels in the subjects with type 2 than in those with type 1 diabetes (Fig. 3). Two potential confounding factors in this study merit discussion. First, male subjects were overrepresented in the group with type 2 diabetes, which may have influenced the magnitude of hormonal response. In both nondiabetic and type 1 diabetic subjects, female subjects have lower counterregulatory responses to hypoglycemia than male subjects (3942), although no information is available in type 2 diabetes. Secondly, six of the subjects with type 2 diabetes required a high rate of insulin infusion (3–6 mU · kg−1 · min−1) to achieve hypoglycemia. Evidence exists in nondiabetic and type 1 diabetic subjects to implicate hyperinsulinemia per se in suppressing the release of glucagon in response to hypoglycemia (4346) while increasing catecholamine and cortisol release (47), although this has been disputed (48). This putative effect of hyperinsulinemia is undetermined in type 2 diabetes.

Type 2 diabetes may therefore confer greater protection against hypoglycemia, particularly when glycemic control is suboptimal, because the counterregulatory responses commence at higher blood glucose levels than observed in the nondiabetic state or in people with type 1 diabetes (37,38). However, improving glycemic control with insulin therapy shifts the threshold for the counterregulatory response to a lower blood glucose level (37,38) (Fig. 3); a similar phenomenon is observed in type 1 diabetes when glycemic control is intensified (1,8,49).

In type 1 diabetes, deficiency of the secretory response of glucagon to hypoglycemia is an early acquired abnormality of counterregulation. The catecholamine response compensates for several years but declines with time (1). In type 2 diabetes, the glucagon response to hypoglycemia has been diversely reported as being either modestly diminished (16,30,50) or preserved (33,34,37,38). People with type 2 diabetes constitute a disparate group, within which the ability of an individual to secrete glucagon in response to hypoglycemia may be related to the degree of insulin deficiency. Most investigators reporting preservation of the glucagon response have studied people with type 2 diabetes who were unlikely to be insulin deficient (33,34,36,38), and, with one exception (16), all of these studies have examined counterregulatory responses in middle-aged subjects in their 5th or 6th decade. However, most people with type 2 diabetes are aged >60 years, and these studies have therefore neglected to consider or account for the effect of ageing on counterregulation. The counterregulatory responses to hypoglycemia were examined in 15 nondiabetic control subjects (7 male, aged 50 ± 6 years) and in 13 people with type 2 diabetes, 7 of whom were receiving treatment with oral antidiabetic agents (3 male, aged 56 ± 6 years), while 6 had been treated with insulin for at least 5 years and were insulin deficient as demonstrated by C-peptide measurements (3 male, aged 57 ± 6 years) (51). The glucagon response to hypoglycemia was intact in the tablet-treated patients and in the nondiabetic control subjects but was almost absent in the insulin-deficient patients (Fig. 4), demonstrating the presence of acquired counterregulatory abnormalities in association with insulin deficiency.

A condition labeled HAAF (hypoglycemia-associated autonomic failure) has been described in type 1 diabetes (52,53), whereby recurrent hypoglycemia provokes failure of the centrally mediated sympatho-adrenal response so causing counterregulatory deficiency and impaired awareness of hypoglycemia. Are people with insulin-deficient type 2 diabetes at risk of developing HAAF? In the study by Segel, Paramore, and Cryer (51), a hypoglycemic clamp performed on the 1st day of the study was followed by another period of hypoglycemia later in the day. When these subjects with type 2 diabetes were exposed to further hypoglycemia on the following day, the plasma glucose levels required to activate the glucagon, catecholamine, and symptomatic responses were lower than in the first hypoglycemic clamp (51). Thus, antecedent hypoglycemia can modify the glycemic thresholds for responses to hypoglycemia in type 2 diabetes and may promote HAAF (53).

In many people with type 2 diabetes who have insulin resistance, the lipolytic effects of epinephrine outweigh the effects of insulin on adipose tissue (50). Plasma free fatty acids increase in response to hypoglycemia in type 2 diabetes (30,33,50) but do not in type 1 diabetes (54). Epinephrine secretion during hypoglycemia may therefore have a greater protective effect in insulin-resistant patients by promoting metabolic substrate release rather than storage. Epinephrine also stimulates release of glucose from the kidney, and, in people who have a deficient glucagon response to hypoglycemia, this may compensate for their impaired hepatic glucose output (55). Thus, in type 2 diabetes, defensive mechanisms to hypoglycemia may be more effective than in type 1 diabetes.

Morbidity of hypoglycemia in type 2 diabetes and in the elderly

Hypoglycemia may cause serious morbidity, provoking major vascular events such as stroke, myocardial infarction, acute cardiac failure, and ventricular arrhythmias (5658). When the patient receives treatment, the precipitating role of hypoglycemia may not be recognized, particularly if medical attendants are unfamiliar with the age-related differences in the manifestations of hypoglycemia. In a 7-year review of 102 cases of hypoglycemic coma secondary to either insulin or glyburide (glibenclamide), 92 patients had type 2 diabetes, 7 sustained physical injury, 5 died, 2 suffered myocardial ischemia, and 1 patient had a stroke as a consequence of severe hypoglycemia (59). The morbidity associated with hypoglycemia, such as impaired consciousness and convulsions, can be particularly debilitating in the elderly, who are at increased risk of injury and bone fractures because of general frailty and the presence of comorbidities, such as osteoporosis (57).

In elderly people of either sex who have diabetes, unsteadiness and weakness are commonly reported symptoms of hypoglycemia (60), and a group of neurological symptoms affecting vision and coordination have been identified in addition to autonomic and neuroglycopenic symptoms (10,14). Consequently, the manifestations of hypoglycemia in elderly people may be mistaken for other conditions, such as transient ischemic attacks or vaso-vagal episodes. Many elderly people with type 2 diabetes possess little knowledge of the symptoms and treatment of hypoglycemia (60,61). This lack of knowledge extends to their relatives and caregivers (62). Inadequate retention of information may be a consequence of age-related cognitive decline, but irrespective of age, knowledge of diabetes and its treatment decreases with time (63), so regular educational reinforcement is required.


Mild hypoglycemia is usually defined by the ability to self-treat, while episodes requiring external assistance are defined as severe. The frequency of hypoglycemia has been examined most extensively in people with type 1 diabetes, in whom mild hypoglycemia occurs on average around twice weekly (64,65). In studies in northern Europe of unselected populations with type 1 diabetes, the estimated incidence of severe hypoglycemia ranges from 1.0 to 1.7 episodes per patient per year (6567). The annual prevalence is between 30% (66,68,69) and 40% (67). These unselected cohorts included people at high risk of severe hypoglycemia, such as those with impaired awareness of hypoglycemia (70). They differ from the atypical participants of the Diabetes Control and Complications Trial (68), who had been selected because they had a low risk of severe hypoglycemia (71) and in whom the observed incidence of severe hypoglycemia was lower, ranging from 0.19 to 0.62 episodes per patient per year.

It is difficult to derive equivalent figures for people with type 2 diabetes because of the heterogeneity of this disorder. Most people with type 2 diabetes are middle aged or elderly; accurate measures of the frequency of hypoglycemia are probably underestimated in the latter (57). Definitions of hypoglycemia differ between studies, hindering comparison, and most studies have reported retrospective data. In people with type 1 diabetes, recall of severe hypoglycemia is relatively robust over a period of 1 year, but recall of mild hypoglycemia is unreliable after an interval of 1 week (64,72). In people with insulin-treated type 2 diabetes, recall of severe hypoglycemia also appears to be preserved over a period of 1 year (73), but reliability of their recall of mild hypoglycemia is unknown. Data from various epidemiological studies are shown in Tables 1 and 2.

Epidemiological data from interventional trials

The U.K. Prospective Diabetes Study (2) reported the prevalence of hypoglycemia in different treatment groups of people with type 2 diabetes, ranging in age from 25 to 65 years. A higher frequency of hypoglycemia was associated with intensive, compared with conventional, treatment with either sulfonylureas or insulin. With intensive treatment, hypoglycemia occurred most frequently in the insulin-treated patients, and the prevalence of hypoglycemia was lower in the 1st decade of the study than in later years (2) (Fig. 5). The prevalence of hypoglycemia was lower when the groups were analyzed on an “intention-to-treat” basis because an increasing number of patients in the conventional treatment groups required the addition of treatment with sulfonylureas or insulin as their glycemic control deteriorated. Although patients were questioned about the occurrence of hypoglycemia at every 4-monthly review, only the most severe episode was documented, so underestimating the overall frequency. The lack of accurate information on the incidence of hypoglycemia is an unfortunate lacuna among the wealth of material provided by this large study.

In the U.S., the Veterans Affairs Cooperative Study in Type 2 Diabetes, examining glycemic control and complications, compared a simple (“standard”) insulin regimen administered once daily with an intensive (“stepped”) regimen (74). The participants in this trial differed from those of the U.K. Prospective Diabetes Study in that they had diabetes of shorter duration ([mean ± SD] 7.8 ± 4 years), they were all insulin-treated male subjects, and they were followed-up for only 18–35 months. The overall incidence of severe hypoglycemia was 0.02 episodes per patient per year, and no significant difference was observed between the standard and stepped treatment groups. The frequency of mild hypoglycemia was significantly higher in the intensively treated group (stepped vs. standard: 16.5 vs. 1.5 episodes per patient per year), but in the standard treatment group blood glucose was monitored less frequently so mild hypoglycemia may have been underreported (74).

Data obtained in clinical trials in which treatment has been applied in an unconventional manner should not be extrapolated to the wider diabetic population. Thus, the U.K. Prospective Diabetes Study and Veterans Affairs Cooperative Study in Type 2 Diabetes trials, while frequently cited, are not representative of current treatment regimens and do not demonstrate the true frequency and risk of hypoglycemia.

Epidemiological data from observational studies

Several observational studies have recorded (mostly retrospectively) the frequency of hypoglycemia in the setting of a hospital outpatient clinic. A study (75) in England of 219 people with type 2 diabetes treated with sulfonylureas and/or metformin observed that 20% of those taking sulfonylureas had experienced symptoms of hypoglycemia in the preceding 6 months. In Atlanta, a 6-month, retrospective, cross-sectional survey of 1,055 predominantly female, African-American patients with type 2 diabetes, treated with oral antidiabetic drugs or insulin, completed serial questionnaires to estimate the frequency of hypoglycemia (76). A quarter of the group had experienced at least one episode of hypoglycemia during the study period. The prevalence of hypoglycemia rose with escalating therapeutic requirements, with the highest rate being associated with insulin. Severe hypoglycemia occurred in 0.5% of patients, all of whom had been treated with insulin. The study is limited by its reliance on patient recall of hypoglycemia and the ethnicity and sex of the study group. A population-based study in Tennessee examined episodes of hypoglycemia retrospectively over a 4-year period in 19,932 Medicaid patients, aged ≥65 years, who had type 2 diabetes (77). This study reported incidences of 1.23 episodes per 100 person-years of “serious” hypoglycemia with sulfonylureas and 2.76 episodes per 100 person-years with insulin treatment, but the strict definition of “serious” hypoglycemia (an episode having a fatal outcome or requiring hospital treatment) may have underestimated the frequency of severe events.

A retrospective study in Turkey examined 165 patients treated with insulin, 114 of whom had type 2 diabetes (78). Hospital case notes were examined for a record of hypoglycemia requiring assistance or hospital admission. This historical approach is likely to have substantially underestimated the overall frequency of severe hypoglycemia, and the incidence of severe hypoglycemia was only 0.15 episodes per patient per year, both in type 1 and insulin-treated type 2 diabetes (78). The authors interpreted these findings as indicating that severe hypoglycemia occurred with a similar magnitude in insulin-treated type 2 diabetic patients as in type 1 diabetes. Although the low incidence of severe hypoglycemia suggests incomplete data collection, particularly in type 1 diabetes, it is possible that the incidence of severe hypoglycemia necessitating emergency medical intervention is similar in these groups. This was certainly true in a population survey in a region of Scotland, in which all episodes of severe hypoglycemia that were attended by the emergency medical services were identified over a 12-month period (79). A total of 244 episodes of severe hypoglycemia had been treated in 160 patients with diabetes. Severe hypoglycemia had required emergency treatment in 7.1% of patients with type 1 diabetes, in 7.3% of patients with insulin-treated type 2 diabetes, and in 0.8% of patients taking oral antidiabetic agents. In type 1 diabetes, severe hypoglycemia is often treated at home, and less than one-third of episodes are thought to need the assistance of the emergency medical services (80). People with insulin-treated type 2 diabetes who suffer severe hypoglycemia may be more likely to require emergency assistance than people with type 1 diabetes, and this was confirmed by a prospective survey in the same region, where the occurrence of hypoglycemia was monitored in a cohort of 267 people with insulin-treated diabetes (both type 1 and type 2) over a period of 1 month (81). The prevalence of all hypoglycemia (mild and severe) in the group with insulin-treated type 2 diabetes was 45% with an incidence of 16.4 episodes per patient per year (42.9 episodes per patient per year in type 1 diabetes). Their incidence of severe hypoglycemia was 0.35 episodes per patient per year (1.15 episodes per patient per year in type 1 diabetes). In the group with type 1 diabetes, only 1 in 10 of those experiencing severe hypoglycemia required emergency service treatment compared with 1 in 3 of the group with type 2 diabetes (81). Although the annual incidence in this study (81) was extrapolated from prospective data collected over a short period of 1 month, the calculated annual rates for people with type 1 diabetes are consistent with those recorded in other European studies (6467). However, the frequency of severe hypoglycemia recorded in people with type 2 diabetes was higher than anticipated (81). Although plasma C-peptide levels were not measured, it is likely that most of these subjects with insulin-treated type 2 diabetes were insulin deficient and were therefore at greater risk of hypoglycemia than people treated with oral antidiabetic agents.

A retrospective Scottish survey in Edinburgh of 215 people with insulin-treated type 2 diabetes observed that the frequency of hypoglycemia increased with duration of insulin therapy and of diabetes and was inversely proportional to HbA1c (A1C) concentration (82). The annual prevalence of severe hypoglycemia was 15% with an overall incidence of 0.28 episodes per patient per year. A retrospective study performed a decade earlier in Edinburgh had assessed the incidence of severe hypoglycemia in 600 unselected insulin-treated diabetic patients. The incidence in the 56 people with type 2 diabetes was 0.73 episodes per patient per year compared with 1.7 episodes per patient per year in the 544 with type 1 diabetes (66). A further survey in the same center compared the frequency of severe hypoglycemia in 86 people with insulin-treated type 2 diabetes with 86 people with type 1 diabetes, matched for duration of insulin treatment and dose (83). The frequency of severe hypoglycemia was comparable in the two groups, and a direct relationship was found between increasing frequency of severe hypoglycemia and increasing duration of treatment with insulin (r = 0.39, P < 0.001).

Moderators of hypoglycemia in type 2 diabetes

There is no evidence to suggest that in type 2 diabetes the principal causes of hypoglycemia (too much insulin or insulin secretagogue, physical exertion or inadequate carbohydrate consumption) differ from type 1 diabetes. Several factors such as sleep, consumption of alcohol, caffeine and various medications, and the timing of exercise, that are known to affect the risk of hypoglycemia in type 1 diabetes, are an unknown quantity in type 2 diabetes. Treatment with insulin for >10 years is an important predictor of increased risk of severe hypoglycemia in type 2 diabetes (81). When people with type 2 diabetes become insulin deficient, their frequency of severe hypoglycemia approaches that experienced by people with type 1 diabetes (83).

Impaired awareness of hypoglycemia is a major risk factor for severe hypoglycemia in type 1 diabetes (70) but is less common in people with type 2 diabetes (83). One retrospective survey of 215 individuals with insulin-treated type 2 diabetes showed that only 8% had impaired awareness estimated by a validated scoring system (70), but those so affected had a ninefold greater incidence of hypoglycemia than those with intact awareness (82). Continuous glucose monitoring systems have been used to detect asymptomatic hypoglycemia in type 1 diabetes, but to date their use in type 2 diabetes has been limited. In a prospective study (84), asymptomatic hypoglycemia <3 mmol/l (<60 mg/dl) was detected in 47% of 30 individuals (17 male, aged 58 ± 11 years) with type 2 diabetes (9 on oral agents, 21 on intensive insulin therapy) compared with 63% of 40 patients with type 1 diabetes (18 male, aged 36.5 ± 12 years). An Australian study examined the frequency of hypoglycemia over two 72-h periods using continuous monitoring in 25 patients treated with sulfonylureas (21 male, aged 73.9 ± 4.4 years). Readings of <2.2 mmol/l (<40 mg/dl) for at least 15 min were recorded in 56% of subjects and none were perceived (85). Impaired awareness of hypoglycemia may be more prevalent in type 2 diabetes than is appreciated.

The Diabetes Outcomes in Veterans Study in the U.S. was a prospective observational trial (86) designed to validate a statistical model for predicting hypoglycemia. The model was tested on a predominantly male cohort of people with insulin-treated type 2 diabetes. Participants performed blood glucose profiles for 8 weeks, and episodes of hypoglycemia were prospectively reported over 1 year. The probability of all hypoglycemia was greater in those who had a low mean blood glucose with a high SD, suggesting that the variability of blood glucose values is as important as A1C values in predicting the risk of hypoglycemia in insulin-treated type 2 diabetes.

Compared with the nondiabetic state, people with type 2 diabetes have a normal rate of exercise-related skeletal muscle glucose uptake but an impaired hepatic glucose output (87), which can result in hypoglycemia during physical exertion. Exercise in type 2 diabetes results in improved insulin sensitivity (88) and reduced postprandial plasma glucose levels (89,90). Improvements in insulin resistance persist for up to 16 h after the period of activity (91), thus exposing the individual to a continuing risk of hypoglycemia. The combination of moderate exercise and ingestion of alcohol did not result in acute hypoglycemia, either after a light meal or after fasting in 12 (8 male) untrained middle-aged subjects with type 2 diabetes who were C-peptide positive (92). Alcohol impairs counterregulatory responses to hypoglycemia in type 1 diabetes (93) but does not appear to delay recovery from hypoglycemia in type 2 diabetes (94).

Risk factors for severe hypoglycemia in people with type 2 diabetes treated with sulfonylureas include age, a past history of vascular disease, renal failure, reduced ingestion of food, alcohol consumption, and interactions with other drugs (59,73,95102).


Oral antidiabetic agents

Hypoglycemia with oral antidiabetic agents is predominantly associated with the insulin secretagogues. Hypoglycemia is not a common side effect of treatment with metformin, thiazolidinediones, or α-glucosidase inhibitors, although it has been occasionally reported in association with metformin when food intake is limited (2,103). The frequency of hypoglycemia is lower in people treated with sulfonylureas than in those treated with insulin (2,76,79) but is probably underestimated (104).

The risk of hypoglycemia of each sulfonylurea relates to its pharmacokinetic properties (104108) and is highest with long-acting sulfonylureas such as chlorpropamide, glyburide (glibenclamide), and long-acting glipizide (101,109111). Glyburide is associated with significantly more episodes of severe hypoglycemia than gliclazide (112) because its hypoglycemic effects last for 24 h (111) as a consequence of the presence of active metabolites (111,113). Glyburide also impairs the glucagon response to hypoglycemia in nondiabetic volunteers (114) and in people with type 2 diabetes (56,115).

Although glipizide is associated with fewer episodes of hypoglycemia, over a 7-year period the Swedish Adverse Drug Reactions Advisory Committee reported 19 cases of severe hypoglycemia that presented with coma or reduced consciousness, with two fatalities (116). Renal impairment and advanced age were identified as risk factors for severe hypoglycemia. In most cases, the severe hypoglycemia had occurred within 1 month of commencing the drug and was not related to dose, suggesting that the response was idiosyncratic.

Efforts have been made to find a sulfonylurea that provides good glycemic control with a low risk of hypoglycemia. Glimepiride, a long-acting sulfonylurea, may partly fulfil this role as it has a lower affinity for the β-cell receptor than glyburide (117), and its insulin secretory capacity is lower in both the fasting (118) and postprandial (119) states. A population-based study in Germany examined the incidence of hypoglycemia in patients with type 2 diabetes who had attended a hospital emergency department over a 4-year period (120). A total of 145 episodes of severe hypoglycemia were treated, and 45 of these patients were receiving treatment with sulfonylureas. Although glimepiride had been prescribed more frequently than glyburide, it was implicated in 6 episodes of severe hypoglycemia, compared with a total of 38 severe events associated with glyburide. In patients with renal impairment, glimepiride can cause prolonged hypoglycemia (121), but it is thought to be safer than other sulfonylureas (122).

A modified release preparation of gliclazide may have a lower risk of hypoglycemia than glimepiride. A multicenter, double-blind, controlled trial in Europe compared the efficacy and safety of modified release gliclazide with glimepiride over a 6-month period (123). The study included people at greater risk of hypoglycemia, such as those aged >65 years (35%) and people with renal impairment. Both groups achieved a reduction of A1C of around 1.0%, with fewer patients reporting hypoglycemia with modified release gliclazide (3.7%) compared with glimepiride (8.9%). Severe hypoglycemia did not occur.

The oral glucose prandial regulators, repaglinide and nateglinide, are insulin secretagogues that have a rapid onset of action but do not stimulate insulin secretion in the fasting state and provoke less hypoglycemia than the sulfonylureas (124127). Repaglinide has been compared with glipizide, gliclazide, and glyburide in separate double-blind, randomized, 1-year studies (125127). Mean A1C concentrations did not differ between any of the treatment groups, and in all sulfonylurea groups the prevalences of hypoglycemia (3.3%) were comparable. In the repaglinide group the prevalence of hypoglycemia was 1.3% with equivalent glycemic control. In a randomized multicenter trial comparing repaglinide with nateglinide, slightly lower A1C values were achieved after 16 weeks on repaglinide, but 7% of patients had experienced mild hypoglycemia compared with none in the nateglinide group (128).

Alternative insulin regimens

Basal insulins can be used safely in combination with oral antidiabetic agents in people with type 2 diabetes. In a systematic review of randomized controlled trials comparing insulin monotherapy and combination therapy with oral agents, 13 of 14 studies did not show any significant difference in hypoglycemia rates between the different regimens (129). In an observational study in our own center of 41 people with type 2 diabetes treated with bedtime NPH (isophane) insulin and oral antidiabetic drugs, 49% had experienced infrequent mild hypoglycemia since commencing insulin, with an incidence of four episodes per patient per year and no episodes of severe hypoglycemia (130). Insulin analogs appear to limit hypoglycemia. In some studies, the risk of hypoglycemia has been reported to be lower with long-acting insulin glargine (131134) and insulin detemir (135) when compared with NPH insulin. Glargine was also associated with a lower frequency of hypoglycemia than premixed insulins (136,137). Rapid-acting insulin analogs, such as lispro and glulisine, were also associated with a lower frequency of hypoglycemia in people with type 2 diabetes when compared with short-acting (soluble) regular insulins (138140).

While continuous subcutaneous insulin infusion (CSII) is beneficially used in selected participants with type 1 diabetes, at present this method of insulin delivery is not commonly employed in people with type 2 diabetes. In a randomized trial of 121 male participants with type 2 diabetes, CSII was compared with multiple dose insulin. Comparable glycemic control was obtained with both regimens, with a lower incidence of mild hypoglycemia in the CSII group (28.4 vs. 9.5 events per patient-year, P < 0.001) (141), although no effect was observed on the incidence of severe hypoglycemia. A 12-month prospective randomized study in 107 adults with insulin-treated type 2 diabetes showed no significant difference between CSII and multiple dose insulin in the rates of mild or severe hypoglycemia (142).

Studies of alternative formulations of insulin, which can be administered by inhalation, include a 6-month randomized trial of 299 participants with type 2 diabetes in which inhaled insulin was compared with subcutaneous insulin. Glycemic control was comparable and inhaled insulin was associated with a relative risk of all hypoglycemia of 0.89 (95% CI 0.82–0.97) when compared with subcutaneous insulin (143).

New agents for the treatment of type 2 diabetes

A detailed discussion of new treatment modalities for type 2 diabetes is beyond the scope of this review. Analogs of glucagon-like peptide-1 are associated with improvements in glycemic control (144148). Although they may provoke reactive hypoglycemia in nondiabetic volunteers (149), they do not appear to cause hypoglycemia in people with type 2 diabetes (150,151).


Few studies of hypoglycemia in people with type 2 diabetes have addressed the potential effects of ageing per se, but the available evidence suggests that it modifies the counterregulatory and symptomatic responses to hypoglycemia. In older people, effective self-treatment of hypoglycemia may be compromised by the juxtaposition of the glycemic thresholds for onset of symptoms and cognitive dysfunction, which occur almost simultaneously, and these age-related changes will be relevant to many people with type 2 diabetes. Most studies that have examined the responses to hypoglycemia in type 2 diabetes have overlooked the potential effects of ageing on counterregulation by selecting middle-aged subjects. The paucity of data from elderly people is of concern, as this age-group is at greatest risk from the morbidity of hypoglycemia, particularly as their presenting features are often misinterpreted and they may not receive prompt treatment. In type 2 diabetes, counterregulatory responses to hypoglycemia commence at higher blood glucose levels than those observed in nondiabetic adults or in people with type 1 diabetes, and this may have a protective effect. Blood glucose thresholds are influenced by glycemic control, and when A1C is reduced with insulin therapy, they are shifted to lower blood glucose levels. With progressive insulin deficiency, people with type 2 diabetes develop counterregulatory deficiencies and impaired symptomatic awareness, similar to type 1 diabetes.

Hypoglycemia has been considered to be a mild and infrequent side effect of treatment in type 2 diabetes, but insufficient and misleading information may have encouraged this misperception. It occurs most frequently with insulin therapy, but sulfonylurea-induced hypoglycemia is also a significant problem. Hypoglycemia is less frequent with the second generation sulfonylureas. Glimepiride, modified release gliclazide, and the prandial glucose regulators may also limit hypoglycemia risk.

Variations in study design, heterogeneity of study populations, and differing definitions of hypoglycemia have confounded attempts to derive accurate overall figures for the frequency of hypoglycemia in type 2 diabetes. Although less common than in type 1 diabetes, the frequency of hypoglycemia in insulin-treated type 2 diabetes progressively rises with increasing duration of insulin treatment. The use of insulin analogs may limit, but does not eradicate, the risk of hypoglycemia. In insulin-treated type 2 diabetes, the frequency of hypoglycemia must not be underestimated, particularly in the elderly, in whom the morbidity of hypoglycemia poses particular problems, and the mortality may be unrecognized.

Figure 1—

Glycemic thresholds for secretion of counterregulatory hormones and onset of physiological, symptomatic, and cognitive changes in response to hypoglycemia in the nondiabetic human. Reproduced from Frier and Fisher (8) in Hypoglycaemia in Clinical Diabetes. Reproduced with permission of John Wiley and Sons, Chichester, U.K.

Figure 2—

Glycemic thresholds for subjective symptomatic awareness of hypoglycemia and for the onset of cognitive dysfunction in young and elderly nondiabetic males. Based on data derived from Matyka et al. (13). Figure reproduced from McAulay and Frier in Diabetes and Old Age (57), with permission of John Wiley and Sons, Chichester, U.K.

Figure 3—

Relationship between blood glucose threshold (mmol/l) for epinephrine secretion in response to hypoglycemia and total HbA1 (%) in type 2 (▪) and type 1 (▴) diabetes. Type 2 diabetes: r = 0.82, P < 0.01; type 1 diabetes: r = 0.63, P < 0.05; P = 0.05 between groups. Reproduced from Levy et al. 1998 (38) with permission of the American Diabetes Association.

Figure 4—

Mean (±SE) plasma glucagon concentrations during hyperinsulinemic stepped hypoglycemic clamps in nondiabetic subjects (•) (n = 15) and in patients with type 2 diabetes treated with oral antidiabetic agents (○) (n = 7) and with insulin (□) (n = 6). P = 0.0252 for nondiabetic vs. type 2 diabetic subjects treated with insulin therapy. Reproduced from Segel, Paramore, and Cryer (51) with permission of the American Diabetes Association.

Figure 5—

Major and any hypoglycemic episodes per year by intention-to-treat analysis and actual therapy for intensive and conventional treatment. Reproduced from the U.K. Prospective Diabetes Study 33 (2), with permission of the Lancet.

Table 1—

Epidemiological data on hypoglycemia in type 2 diabetes, expressed as incidence

Table 2—

Epidemiological data on hypoglycemia in type 2 diabetes, expressed as prevalence


N.N.Z. was supported by a grant from the Juvenile Diabetes Research Foundation.


  • B.M.F. has been a member of an advisory panel for and has received honoraria/consulting fees from Eli Lilly, Sanofi-Aventis, GlaxoSmithKline, and Takeda.

    A table elsewhere in this issue shows conventional and Système International (SI) units and conversion factors for many substances.

    • Accepted September 12, 2005.
    • Received May 16, 2005.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
  73. 73.
  74. 74.
  75. 75.
  76. 76.
  77. 77.
  78. 78.
  79. 79.
  80. 80.
  81. 81.
  82. 82.
  83. 83.
  84. 84.
  85. 85.
  86. 86.
  87. 87.
  88. 88.
  89. 89.
  90. 90.
  91. 91.
  92. 92.
  93. 93.
  94. 94.
  95. 95.
  96. 96.
  97. 97.
  98. 98.
  99. 99.
  100. 100.
  101. 101.
  102. 102.
  103. 103.
  104. 104.
  105. 105.
  106. 106.
  107. 107.
  108. 108.
  109. 109.
  110. 110.
  111. 111.
  112. 112.
  113. 113.
  114. 114.
  115. 115.
  116. 116.
  117. 117.
  118. 118.
  119. 119.
  120. 120.
  121. 121.
  122. 122.
  123. 123.
  124. 124.
  125. 125.
  126. 126.
  127. 127.
  128. 128.
  129. 129.
  130. 130.
  131. 131.
  132. 132.
  133. 133.
  134. 134.
  135. 135.
  136. 136.
  137. 137.
  138. 138.
  139. 139.
  140. 140.
  141. 141.
  142. 142.
  143. 143.
  144. 144.
  145. 145.
  146. 146.
  147. 147.
  148. 148.
  149. 149.
  150. 151.
  151. 151.
| Table of Contents