R E V I E W

A R T I C L E

Pathogenesis of NIDDM
A Balanced Overview
RALPH A. DEFRONZO, MD RLCCARDO C . BONADONNA, MD ELEUTERIO FERRANNINI, MD

Non-insulin-dependent diabetes mellitus (NIDDM) results from an imbalance between insulin sensitivity and insulin secretion. Both longitudinal and cross-sectional studies have demonstrated that the earliest detectable abnormality in NIDDM is an impairment in the body's ability to respond to insulin. Because the pancreas is able to appropriately augment its secretion of insulin to offset the insulin resistance, glucose tolerance remains normal. With time, however, the P-cell fails to maintain its high rate of insulin secretion and the relative insulinopenia (i.e., relative to the degree of insulin resistance) leads to the development of impaired glucose tolerance and eventually overt diabetes mellitus. The cause of pancreatic "exhaustion" remains unknown but may be related to the effect of glucose toxiciry in a genetically predisposed P-cell. Information concerning the loss of first-phase insulin secretion, altered pulsatility of insulin release, and enhanced proinsulin-insulin secretory ratio is discussed as it pertains to altered p-cell function in NIDDM. Insulin resistance in NIDDM involves both hepatic and peripheral, muscle, tissues. In the postabsorptive state hepatic glucose output is normal or increased, despite the presence of fasting hyperinsulinemia, whereas the efficiency of tissue glucose uptake is reduced. In response to both endogenously secreted or exogenously administered insulin, hepatic glucose production fails to suppress normally and muscle glucose uptake is diminished. The accelerated rate of hepatic glucose output is due entirely to augmented gluconeogenesis. In muscle many cellular defects in insulin action have been described including impaired insulin-receptor tyrosine kinase activity, diminished glucose transport, and reduced glycogen synthase and pyruvate dehydrogenase. The abnormalities account for disturbances in the two major intracellular pathways of glucose disposal, glycogen synthesis, and glucose oxidation. In the earliest stages of NIDDM, the major defect involves the inability of insulin to promote glucose uptake and storage as glycogen. Other potential mechanisms that have been put forward to explain the insulin resistance, include increased lipid oxidation, altered skeletal muscle capillary density/fiber type/blood flow, impaired insulin transport across the vascular endothelium, increased amylin, calcitonin gene-related peptide levels, and glucose toxicity.

FROM THE DIVISION OF DIABETES, THE UNIVERSITY OF TEXAS HEALTH SCIENCE CENTER AT SAN ANTONIO; THE AUDIE L. MURPHY VA HOSPITAL, SAN ANTONIO, TEXAS; AND THE INSTITUTE OF PHYSIOLOGY, C.N.R., PISA, ITALY. ADDRESS CORRESPONDENCE TO RALPH A. DEFRONZO, MD, PROFESSOR OF MEDICINE, CHIEF, DIABETES DIVISION, THE UNIVERSITY OF TEXAS HEALTH SCIENCE CENTER AT SAN ANTONIO, 7703 FLOYD CURL DRIVE, SAN ANTONIO, I X 7 8 2 8 4 - 7 8 8 6 .

fter an overnight fast, insulinindependent tissues, the brain (50%), and splanchnic organs (25%) account for most of the totalbody glucose disposal. Insulin-independent tissues, primarily muscle (14), are responsible for the remaining 25% of glucose utilization. Basal glucose uptake is 2 mg kg" 1 min" 1 , and this is precisely matched by the release of glucose from the liver (1-4; Fig. 1). After glucose ingestion or infusion, this delicate balance between tissue glucose uptake and hepatic glucose output is disrupted and the maintenance of normal glucose homeostasis is dependent on three processes that occur in a coordinated and tightly integrated fashion (Table 1). In response to glucose, pancreatic insulin secretion is stimulated and the combination of hyperinsulinemia plus hyperglycemia promotes glucose uptake by splanchnic Giver and gut) and peripheral (primarily muscle) tissues and suppresses hepatic glucose production. It follows, therefore, that defects at the level of the P-cell, muscle, and/or liver can lead to the development of glucose intolerance or overt diabetes mellitus. In this review we emphasize the concept that the full-blown syndrome of noninsulin-dependent diabetes mellitus (NIDDM) requires the simultaneous presence of two major defects, insulin resistance and impaired P-cell function. In most NIDDM individuals impaired tissue (muscle and/or liver) sensitivity to insulin represents the primary or inherited defect. If the P-cell fails to maintain a sufficiently high rate of insulin secretion to offset the insulin resistance, fasting hyperglycemia and overt diabetes mellitus ensue. This sequence of events is characteristic of both obese and lean diabetic individuals. In some NIDDM patients the primary defect may start at the level of the P-cell and manifest itself as an impairment in insulin secretion; insulin resistance develops concomitantly with or subsequent to the disturbance in insulin secretion. Such individuals are unusual and are represented by

318

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH 1992

DeFronzo

Table 1Factors responsible for the maintenance of normal glucose tolerance in healthy subjects
INSULIN SECRETION TISSUE GLUCOSE UPTAKE PERIPHERAL (PRIMARILY MUSCLE) SPLANCHNIC (LIVER PLUS GUT) SUPPRESSION OF HEPATIC GLUCOSE PRODUCTION

Although adipose tissue is insulin sensitive, it is responsible for the disposal of only 1-2% of an infused or ingested glucose load.

the lean diabetic. It should be emphasized that, whichever defect, i.e., diminished insulin secretion or insulin resistance, initiates the development of NIDDM it subsequently leads to the emergence of the second abnormality. Importantly, both defects must be present simultaneously before significant glucose intolerance will ensue. Although excessive hepatic glucose metabolism, and specifically an increase in gluconeogenesis, plays an important role in maintaining the diabetic state once it has become firmly established, it is uncertain what role the liver plays in the development of early fasting hyperglycemia in NIDDM patients. INSULIN SECRETION IN NIDDM |3-Cell function in NIDDM has been the subject of much controversy (1,5-7).

However, recent studies have demonstrated a consistent pattern that reveals a complex interplay between insulin secretion, insulin sensitivity, and the severity of hyperglycemia. In Table 2 we summarized published studies in which insulin secretion has been quantitated in normal-weight NIDDM subjects with fasting hyperglycemia (5). Because obesity causes insulin resistance, which in turn elicits a compensatory increase in insulin secretion (1,6,8), we focus initially on the lean NIDDM patient to examine the pure impact of diabetes on insulin secretion. In subsequent sections we explore the interaction between obesity and diabetes on insulin sensitivity and insulin secretion.

Fasting Plasma Insulin

20

IR

l5

60

I O O

140

180 220 260 300

Fasting Plasma Glucose (mg/dl)

Fasting insulin concentrations

The fasting plasma insulin concentration has invariably been found to be normal or increased in NIDDM (Table 2). Even in those studies where normal fasting plasma insulin levels have been reported, they uniformly have been in the high normal range. Moreover, basal insulin secretion, estimated in NIDDM patients with C-peptide, is increased (9,10). Recently, we measured the plasma insulin concentration in the basal state and during an oral glucose tolerance test (OGTT) in 77 normal-weight NIDDM subjects (11; Fig. 2). The relationship between fasting glucose and insulin levels is complex, resembling an inverted "U." As the fasting glucose increases from 80 to 140 mg/dl, there is a progressive rise in fasting plasma insulin, which represents a compensatory response by the pancreas to offset the deterioration in glucose metabolism. When the fasting glucose exceeds 140 mg/dl, insulin secretion drops off precipitously. The inability of the pancreas to maintain its high rate of insulin secretion has important pathophysiological implications, because it is at this point (fasting glucose 140 mg/dl) that hepatic glucose production increases in absolute terms and begins to make a major contribution to the

Figure 2Relationship between the fasting plasma glucose concentration and the fasting plasma insulin concentration in normal-weight control subjects, in individuals with impaired glucose tolerance, and in non-insulin-dependent diabetic subjects with varying degrees of fasting hyperglycemia. As the fasting plasma glucose concentration rises from baseline to 140 mg/dl, there is a progressive increase in the fasting insulin concentration. Thereafter, further rises in the fasting glucose level are associated with a progressive decline in fasting insulin. In diabetic subjects with fasting glucose concentrations >200-220 mg/dl, the fasting insulin level declines to values observed in control subjects. From DeFronzo et al. (11). by Metabolism.

Table 2 Summary of plasma response during glucose tolerance nonobese non-insulin-dependent subjects with fasting hyperglycemia

insulin tests in diabetic

PLASMA INSULIN RESPONSE TO GLUCOSE FASTING INSULIN DECREASED NORMAL INCREASED EARLY* LATE TOTAL

Other Glycogenolysis 1.5 mg/kgmin Glycerol (2%) ruvote(l%) Loctate(l6%) AminoAcids(6%) Glucose Oxidation Glycolysis

Splanchnic Glucose Uptake

0 27 5

21 6 5

13 12 7

16 11 5

TOTAL BODY GLUCOSE UPTAKE

Figure 1Schematic representation of glucose production and glucose utilization in nondiabetic humans in the postabsorptive state. Drawn from the data presented in refs. 2-4 and 149.

The 32 publications from which the above summary was prepared can be found in ref. 5. *The early phase of insulin secretion refers to the 0 - to 10-min period during the intravenous glucose tolerance test and the 0- to 60-min period during the oral glucose tolerance test.

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH 1992

319

Pathogenesis ofMDDM

elevation in fasting plasma glucose concentration (11). The simultaneous presence of fasting hyperglycemia and fasting hyperinsulinemia indicates the presence of severe insulin resistance. In the postabsorptive state both hepatic and peripheral tissues share in this insulin resistance (see subsequent discussion).

Glucose-stimulated insulin secretion

In lean NIDDM subjects glucose-stimulated insulin secretion has variably been reported to be normal, increased, or decreased (1,5; Table 2). It must be emphasized, however, that even a normal or increased plasma insulin response is deficient when viewed relative to the level of hyperglycemia and degree of insulin resistance. Numerous studies, including our own (1,12-22), have shown that the relationship between insulin secretion during an OGTT and the severity of diabetes is complex and resembles that seen in the fasting state. During a 2-h oral glucose load (100 g), a nondiabetic individual (fasting glucose 80 mg/dl) has a mean plasma insulin concentration of 45 fiU/ml. With the onset of fasting hyperglycemia the (3-cell recognizes that the glucose homeostatic mechanism has become disrupted and augments its insulin secretory capacity in an attempt to offset the disturbance in glucose metabolism (Fig. 3). Thus, an individual with a fasting glucose of 120 mg/dl (i.e., impaired glucose tolerance [IGT] or early diabetes mellitus) secretes approximately twice as much insulin as a person with normal glucose tolerance. When the fasting glucose exceeds 120 mg/dl, there is a progressive decline in insulin secretion, and a diabetic person with a fasting glucose of 150-160 mg/dl secretes an amount of insulin that is similar to that in a healthy nondiabetic individual. Note, however, that a normal insulin response in the presence of hyperglycemia is markedly abnormal. With further increments in fasting glucose (> 150-160 mg/dl) the plasma insulin response, when viewed in absolute terms, becomes

mild fasting hyperglycemia (<140 mg/ dl), there is a further increase in both Mean Plasma fasting and glucose-stimulated insulin Insulin Response secretion (1,22,28,30,31) and a worsenDuring OGTT ing of the insulin resistance (1; Fig. 4). (yaU/ml) The onset of overt fasting hyperglycemia (> 140 mg/dl) results from an inability of 120 160 Fasting Plasma Glucose the P-cell to maintain its high rate of Cone, (mg/dl) insulin secretion, both in the postabsorpFigure 3Starling's curve of the pancreas for tive and stimulated states (Fig. 4). A similar picture of insulin secretion has been insulin secretion. In normal-weight patients with impaired glucose tolerance and mild diabetes reported with the development of diabemellitus, the plasma insulin response to ingestedtes in the rhesus monkey (33,34). These studies (1,22,28,30,31,32) conclusively glucose (OGTT) increases progressively until the fasting glucose reaches 120 mg/dl. Thereafter, document that hyperinsulinemia prefurther increases in the fasting glucose concen- cedes the development of NIDDM in hutration are associated with a progressive declinemans and exclude the possibility that in insulin secretion. Drawn from the data pre- insulinopenia initiates the process of disented in refs. 1, 4, 5, 134-136, and 167. abetes mellitus. Consistent with this sequence of events, studies in ethnic groups who are at high risk to develop NIDDM (20,22,27,35-40) and in firstdeficient. This inverted U-shaped curve degree relatives of NIDDM individuals is similar to that observed in the postab- (37,41,42) have shown that hyperinsusorptive state (Fig. 2). Finally, when the linemia predicts the eventual developfasting glucose level exceeds 200220 mg/dl, the plasma insulin response becomes markedly blunted, although the basal rate of insulin secretion remains elevated and fasting hyperinsulinemia 250 Insulinpersists (9,10,23). Even with severe fastO..-..-0 ing hyperglycemia, 24-h insulin profiles in NIDDM subjects reveal normal integrated insulin and C-pep tide areas (24 26). These normal values result from the combination of high fasting and decreased postprandial insulin/C-peptide secretory rates. The pancreatic function curves described above (Figs. 2 and 3) are consistent with the natural history of impaired glucose tolerance and NIDDM in humans (1,22,27-32) and the rhesus Figure 4Summary of plasma insulin (A, monkey (33,34). The progression from open circles) and plasma glucose (B) responses normal to IGT is associated with a during a 100-g oral glucose tolerance test marked increase in both fasting and glu- (OGTT) and tissue sensitivity to insulin (A, cose-stimulated plasma insulin levels closed circles) in control, obese nondiabetic, obese (1,22,27,31,32). In these studies glucose-intolerant, obese hyperinsulinemic dia(1,22,27,31,32) conversion to IGT is as- betic, and obese hypoinsulinemic diabetic subsociated with the development of severe jects. See text for a detailed discussion. From insulin resistance (Fig. 4). As the person DeFronzo (1). by the American Diabetes Aswith IGT progresses to NIDDM with sociation.
Mean Plasma Insulin During OOTT (jiUVml) 120 100 80 Msdlatsd Gtucose Plasma Glucose During OGTT CON OB OBOBGLU DIAB INTOL Hi INS

320

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

ment of IGT and NIDDM. Most recently, hyperinsulinemia has been shown to predict the development of NIDDM in whites (43,44).

response has been shown to be normal in the relatives of patients with NIDDM (57) and in the normal glucose-tolerant insulin-resistant offspring of two diabetic parents (58).

First-phase insulin secretion

In humans (45) and animals (46) insulin secretion is biphasic, with an early burst of insulin release within the first 10 min, followed by a progressively increasing phase of insulin secretion that persists as long as the hyperglycemic stimulus is present. This biphasic response is not easily identifiable after oral glucose but can be readily demonstrated after intravenous glucose. It has been suggested that loss of the first phase of insulin secretion is the earliest detectable abnormality in patients who are destined to develop NIDDM (47,48). From the data reviewed in Table 2 it is obvious that the early phase of insulin secretion during both the OGTT and the intravenous glucose tolerance test is reduced in most NIDDM subjects with fasting plasma glucose levels > 110120 mg/dl. The loss of the first phase of insulin secretion has important pathogenetic consequences, because this early burst of insulin release plays an important role in priming those insulin target tissues, especially the liver, that are responsible for the maintenance of normal glucose homeostasis (49-52). However, loss of first-phase insulin secretion does not appear to be the primary defect responsible for NIDDM. Recent studies (27,28) have shown that progression from normal to IGT or to NIDDM is associated with a reduction in insulin sensitivity and an increase in insulin secretion with an intact first-phase response. Although the first-phase insulinsecretory response is characteristically lost in patients with well-established NIDDM (Table 2), this defect does not occur until the fasting plasma glucose concentration rises to 115-120 mg/dl (53). Moreover, tight metabolic control partially restores the defect in first-phase insulin response (54-56), indicating that the defect is acquired and not inherited. In the same vein, first-phase insulin

Pulsatile insulin secretion

In humans (59) and animals (60) insulin is secreted in a pulsatile fashion and some evidence suggests that pulsatile insulin delivery is more effective in promoting glucose disposal than continuous administration (61). However, this has been refuted by others (62). A recent study reported that loss of oscillatory insulin secretion is a characteristic feature of relatives of patients with NIDDM (58) and suggested that this may be the earliest lesion in the natural history of NIDDM. Somewhat differing results have been reported by Polonsky et al. (63). In patients with established NIDDM, the number of pulses secreted during a 24-h period did not differ from control subjects, although the pulse amplitude was significantly reduced and their relationship to meals was distorted (63). The meaning of these somewhat discrepant observations remains uncertain because there is no conclusive evidence that loss of pulsatile insulin release has any adverse consequences on glucose homeostasis.

use of a specific proinsulin assay, several laboratories have demonstrated that in NIDDM individuals with moderate to severe hyperglycemia, proinsulin makes up a larger fraction of the immunoassayable insulin (64,67-69). One study suggested that proinsulin accounts for >50% of circulating radioimmunoassayable insulin (64). However, in this study the diabetic subjects had severe fasting hyperglycemia (>225 mg/dl), only a single early time point (30 min) during the OGTT was analyzed, and proinsulin/ insulin was assayed on samples that had been stored for a prolonged period. Under such conditions, one would expect a preferential loss of insulin compared with proinsulin and its split products. This would artifactually increase the insulin-proinsulin ratio.

Nonetheless, the observations described above have raised several important questions. When does the increase in proinsulin occur and are patients with IGT and/or NIDDM with mild fasting hyperglycemia characterized by true insulin deficiency? To address this issue, Yoshioka et al. (69) measured the insulin and proinsulin responses in nondiabetic healthy subjects, in subjects with IGT, and in NIDDM patients with a wide range of fasting plasma glucose levels. Subjects with IGT and mild fasting hyperglycemia (<140 mg%) manifested significant increases in both fasting and Proinsulin It has been suggested that NIDDM indi- stimulated insulin and proinsulin levels. viduals may be more insulinopenic than Although the plasma proinsulin concenpreviously appreciated because of the tration increased disproportionately to presence of high circulating levels of that of insulin, the absolute amount of proinsulin and 32-33 split proinsulin, "true" biologically active insulin secreted which are biologically much less active was nonetheless markedly increased. than insulin but cross-react substantially These observations indicate that during in the routine insulin radioimmunoassay the earliest stages of development of (64). In nondiabetic subjects the abso- NIDDM, diabetic patients are characterlute amount of proinsulin secreted is ized by hyperinsulinemia. Recently, small compared with insulin (65). How- Leahy et al. (70) examined the dynamics ever, because the metabolic clearance of proinsulin secretion in the partially rate of proinsulin is significantly less pancreatectomized rat model. The than that of insulin (66), it accounts for amount of proinsulin in pancreatic ex 10% of the total amount of circulating tracts was significantly elevated in parinsulin that is measured in the basal and tially pancreatectomized rats compared glucose-stimulated states (65). With the with control rats, and an increased pro-

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

321

Pathogenesis of NIDDM

insulin-insulin ratio was observed in the portal vein. However, as in diabetic humans (69), the percentage of increase in portal vein proinsulin in the rat was small (3-10%). The authors suggested that chronic hyperglycemia was responsible for the augmented rates of proinsulin synthesis and secretion in this partially pancreatectomized diabetic rat model (70). In summary, it appears that in diabetic humans circulating levels of both "true" biologically active insulin and proinsulin are increased. Only in moderately severe to severe NIDDM individuals (fasting glucose > 160-180 mg/dl) is there a decrease in the biologically active insulin component.

Etiology of insulin deficiency in NIDDM

The progression from normal glucose tolerance to IGT to NIDDM with mild fasting hyperglycemia (< 120-140 mg/ dl) is characterized by progressive hyperinsulinemia (Figs. 2-4). However, once the fasting glucose concentration exceeds 140 mg/dl there is a decline in both fasting and glucose-stimulated plasma insulin levels. These observations suggest that the decline in |3-cell function is acquired. In this section we review those pathogenetic factors that have been implicated in this progressive impairment in insulin secretion. P-Cell number is a critical determinant of the amount of insulin that is secreted by the pancreas. Most (71-76) but not all (77) studies have demonstrated a modest reduction (20-40%) in f$-cell mass in patients with well-established NIDDM. Morphologically, the islets of Langerhans appear normal and there is no evidence of insulitis. The etiology of the decreased (3-cell mass remains undefined. Advancing age cannot explain the loss of functioning (3-cell number (71-76). Obesity, another insulin-resistant state, is characterized by an increase in (3-cell mass (75). Thus, the finding of even a modest reduction in (3-cell mass is most impressive. Nonetheless, it is important to emphasize that a

>80-90% decrease in (3-cell mass is required before sufficient insulinopenia develops to cause overt diabetes mellitus (78). Moreover, it is unknown whether, in the earliest stages of the natural history of NIDDM, |3-cell mass is diminished whatsoever. It seems likely, therefore, that factors in addition to (3-cell loss must be responsible for the impairment in insulin secretion. A basic genetic defect in the insulin gene has been proposed to explain the disturbance in insulin secretion in NIDDM, but there is little evidence to support this hypothesis. As reviewed by Permutt and Elbein (79), mapping of the insulin gene for restriction-fragmentlength polymorphisms failed to reveal any significant abnormalities. Most recently, the insulin gene has been cloned and its DNA sequence determined (80). Six families have been identified in whom a point mutation in the insulin gene has lead to the development of mild hyperinsulinemic diabetes that is similar, in some respects, to NIDDM in humans (81). However, such mutations appear to be rare and are unlikely to account for most patients with NIDDM. In Pima Indians and Nauruans, two populations with a high prevalence of NIDDM, the coding sequence of the insulin gene has been shown to be normal (82). A topic of considerable interest involves the genes that activate insulin-gene transcription. Because 5'-flanking region cis-regulatory elements have been described, it is possible that deficiency of gene products that activate (or conversely, excess of gene products which inhibit) the promoter region of the insulin gene may be responsible for the development of hypoinsulinemia in NIDDM. Recently, it has been suggested that amylin may contribute to the defect in insulin secretion in NIDDM. This hormone is a 37 amino acid peptide that is structurally similar to calcitonin generelated peptide. It is produced by the P-cell, packaged with insulin in secretory granules, and cosecreted into the sinusoidal space (83-86). This peptide

has been shown to be the precursor for the amyloid deposits that frequently are observed in patients with NIDDM ( 8 5 89). At very high dosages amylin has been shown to inhibit insulin secretion by the perfused rat pancreas in vitro (90) and to precede the appearance of glucose intolerance in spontaneously diabetic monkeys (91). In this latter model the severity of the diabetes is correlated closely with the amount of amylin that is deposited within the pancreas (91). After its secretion, amylin accumulates extracellularly in close contact with the (3-cell, and it has been postulated that these amylin deposits might cause f$-cell dysfunction and eventually death by impairing the transport of nutrients from the plasma to the (3-cell or by interfering with the glucose-sensing and/or insulinsecreting apparatus of the 3-cell ( 8 5 87). Although an attractive hypothesis, Bloom et al. (92,93) failed to find any inhibitory effect of amylin on insulin secretion when the peptide was infused in pharmacological dosages in rats, rabbits, and humans. In summary, the evidence that amylin is responsible for the defect in f}-cell dysfunction in NIDDM in humans is not very strong. Similarly, calcitonin gene-related peptide, which shares a 46% homology with amylin, has been shown to have no effect on insulin secretion when infused intravenously in rats (94). The newest hormone that has been implicated in the impairment of insulin secretion in NIDDM is galanin (95). This 29 amino acid peptide is released by pancreatic sympathetic nerve terminals in response to neural stimulation, inhibits both basal and stimulated insulin secretion, and is a glucagon secretagogue (95). What role, if any, this novel peptide plays in the pathogenesis of NIDDM remains to be determined. The most likely explanation for the acquired defect in insulin secretion relates to the concept of glucose toxicity (1,7,96,97). Much evidence in support of this hypothesis has been produced both in humans and animals. In NIDDM patients, tight metabolic control, how-

322

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

ever, achieved (i.e., diet, insulin therapy, sulfonylureas), leads to an improvement in insulin secretion (7,96,98-104 and in the article byj. Leahy, this issue, p. 442). Although it is well established that sulfonylurea agents enhance insulin secretion, weight loss has no direct effect on P-cell function (105), and insulin administration actually inhibits insulin secretion (106). The observation that improved glycemic control, independent of the method via which it is achieved, leads to improved (3-cell function (96), provides strong clinical support that hyperglycemia exerts a deleterious effect on insulin secretion. A more rigorous test of the glucose toxicity hypothesis requires that the plasma glucose concentration be lowered without altering circulating substrate levels (other than glucose) with an agent that does not have any direct effect on cellular metabolism. This goal can be achieved by the chronic administration of phlorizin, a potent inhibitor of renal tubular glucose transport (96). In diabetic animals phlorizin restores normoglycemia without altering plasma insulin, amino acid, free fatty acid, or other substrate/hormone concentrations (107). In rats made diabetic by removal of 90% of the pancreas, mild fasting hyperglycemia and moderately severe glucose intolerance develop (107). In control rats a square wave of intravenous glucose (hyperglycemic clamp) elicits the typical biphasic pattern of insulin secretion, with an early burst of insulin release within the first 10 min followed by a progressively increasing pattern of insulin release over the subsequent 50 min. In partially pancreatectomized rats the first phase of insulin secretion was totally lost and the second phase was markedly impaired (Fig. 5). This rat model closely resembles insulinopenic NIDDM in humans. Restoration of normoglycemia by phlorizin treatment returned both the first and second phases of insulin secretion to normal (107; Fig. 5). These results indicate that impaired (3-cell function in partially pancreatectomized rats is

6 5 4 3 2 1

First Phase (0 -10 min)

PLASMA INSULIN RESPONSE B (ng/ml 6 per g of pancreas) 5

4 3 2 1 0

Second Phase (10 - 60 min)

CONTROL

PANX

Figure 5First-phase (A) and second-phase (B) insulin secretion in control rats, in 90% partially pancreatectomized rats, and in pancretectomized rats treated with phlorizin. Phlorizin completely corrected the defects in first- and second-phase insulin secretion. From Rossetti et al. (96). by the American Diabetes Association.

functional in nature and cannot be attributed to "P-cell death." Otherwise, one would not observe a recovery of P-cell function with improved glycemic control. Additional support for this hypothesis is derived from the P-cell response to nonglucose stimuli. Thus, arginine infusion causes a threefold greater rise in insulin secretion in partially pancreatectomized diabetic versus control rats, and this potentiated insulin response is returned to normal after correction of the hyperglycemia with phlorizin (96,107). This potentiated insulin response to amino acids and other nonglucose stimuli most likely explains the normal daylong plasma insulin levels in NIDDM patients, despite the presence of moderate to severe fasting hyperglycemia (24-26). Weir and Leahy et al. (108-110) provided additional support for the glucose toxicity hypothesis. In both the par-

tially pancreatectomized and neonatal streptozocin-induced diabetic rat models the insulin response to hyperglycemia in perfused pancreatic tissue was markedly impaired in diabetic versus control animals. In contrast, the insulin secretory response to arginine, isobutyl methylxanthine, and isoproterenol was either normal or increased. In an elegant series of studies these authors demonstrated that the defect in insulin secretion was not the consequence of diminished P-cell mass alone, but resulted from superimposed mild hyperglycemia (111). A sustained rise in the plasma glucose concentration of as little as 15 mg/dl caused a 75% inhibition of insulin secretion by the in vitro perfused pancreas. These results provide support for the concept that a minimal elevation in the mean plasma glucose concentration, in the presence of a reduced p-cell mass, can lead to a major impairment in insulin secretion by the remaining pancreatic tissue (96). A deleterious effect of hyperglycemia on P-cell function also has been demonstrated in normal rats after chronic (72-h) glucose infusion (112, 113). The studies of Giroix et al. (114), Kergoat et al. (115), Grill et al. (116), and Imanura et al. (117) are in agreement with those of Weir and Leahy et al. The potential mechanism(s) responsible for the downregulation of insulin secretion have been reviewed by Rossetti and DeFronzo (96). From the in vivo and in vitro studies reviewed above, it seems reasonable to postulate that chronic hyperglycemia is responsible, at least in part, for the P-cells' inability to respond to an acute hyperglycemic challenge. This concept, which we have referred to as glucose toxicity, has far reaching implications (1,96). First, it means that hyperglycemia no longer simply can be viewed as a manifestation, i.e., a laboratory marker, of diabetes mellitus. Rather, the hyperglycemia must be considered as a pathogenetic factor that impairs insulin secretion and, therefore, perpetuates the diabetic state. Second, the glucose toxic-

DIABETES CARE, VOLUME 15,

NUMBER 3 ,

MARCH

1992

323

Pathogenesis of NIDDM

Diet (17)

Sulfonylureas (26) After Therapy

Insulin (7)

50 -

Plasma Insulin (nU/ml)

25

nous insulin, which inhibits endogenous insulin secretion; and 3) sulfonylureas, which augment insulin secretion. With all three treatment schedules, the ability of the fi-cell to secrete insulin uniformly improved (Fig. 6). The only common feature that links these three therapeutic interventions is a reduction in the plasma glucose concentration. These results strongly suggest that it is the normalization of the plasma glucose concentration and removal of glucose toxicity that leads to the improvement in insulin secretion.

Summary: insulin secretion NIDDM involves defects in both insulin secretion and insulin action (1). During B the very earliest stages in the natural his400 r tory of NIDDM, insulin secretion is augmented compared with age- and weightmatched control subjects. Although one could postulate that enhanced insulin secretion by the pancreatic (3-cells represents the primary abnormality in NIDDM Plasma and that the insulin resistance develops Glucose 2 5 secondary to chronic hyperinsulinemia (mg/dl) (32), there is little proof to support this 200 pathogenetic sequence. Most of the available evidence suggests that insulin resistance is the primary metabolic disturbance in NIDDM and that the augmented f$-cell response represents a compensatory adaptation by the fi-cells to offset the defect in insulin action. Until 2 3 0 I this juncture, we have focused our attenTime (hours) tion on the (3-cell. We now review what is known about defect(s) in the insulinsensitive tissues, muscle, and liver. SubFigure 6Plasma insulin (A) and glucose (B) concentrations in non-insulin-dependent diabetic sequently, we return to the fJ-cell to expatients before and after treatment with diet, sulfonylureas, and insulin. Improved glycemic control, regardless of the means by which it was achieved, uniformly led to an improvement in insulin amine the dynamic interaction between secretion. From Kosaka et al. (99). by Diabetologia. insulin action and insulin secretion (1,8,11,12,22,27,30-34,41,118,119), because it is the disruption of this finely ity hypothesis has important therapeutic the plasma glucose concentration and are regulated balance that leads to the emerimplications and may help to explain the associated with a concomitant increase in gence of overt diabetes mellitus with fastuniform improvement in insulin secre- insulin secretion. The results of Kosaka ing hyperglycemia. tion observed after many diverse maneuvers, all of which have in common the ability to lower the plasma glucose concentration. Thus, acute and chronic caloric restriction, intensified insulin therapy, and sulfonylurea drugs all reduce et al. (99) are particularly relevant to this argument (Fig. 6). NIDDM subjects with moderately severe fasting hyperglycemia (>200 mg/dl) were treated with each of three regimens: I) weight loss, which has no effect on insulin secretion; 2) exogeINSULIN RESISTANCE IN NIDDM As discussed in the preceding sections, longitudinal and crosssectional studies have provided conclusive evidence that hyperinsulinemia

324

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH

1992

DeFronzo

pendent on three tightly coupled mechantedates the development of NIDDM anisms: 1) suppression of hepatic glucose (1,13-22,27,28,30,32). Moreover, nuproduction, 2) augmentation of glucose merous studies that used the euglycemic uptake by splanchnic (hepatic plus gasinsulin-clamp technique (45) have demtrointestinal) tissues, and 3) stimulation onstrated that the progression from norGlucose Uptake of glucose uptake by peripheral tissues. mal to 1GT is associated with the devel- (mg/kgmin) Glucose utilization, in turn, is dependent opment of severe insulin resistance on enhanced flux through two major (1,22,27,31,41), whereas both fasting metabolic pathways: glycolysis (glucose and glucose-stimulated insulin responses oxidation and anaerobic glycolysis) and are markedly increased (Figs. 2 - 4 ) . glycogen synthesis. In the following disThese observations provide convincing cussion, the contribution of each of these evidence that insulin resistance, not imNIDDM Controls processes to insulin resistance in NIDDM paired insulin secretion, initiates the is reviewed. process of NIDDM in humans. Figure 7Insulin-mediated whole-body gluHimsworth et al. (120), with the cose uptake in 38 normal-weight non-insulin- Hepatic glucose production (HGP). combined OGTT/intravenous insulin tol- dependent diabetic (NIDDM) patients (right bar)With tritiated glucose, DeFronzo et al. erance test, were the first to demonstrate and in 33 age- and weight-matched control sub- showed that the liver of healthy subjects that insulin sensitivity was diminished in jects (left bar). Tissue sensitivity to insulin was p r o d u c e s glucose at 1 . 8 - 2 . 2 diabetic patients. Subsequent investiga- reduced by 40% in the NIDDM group. Each mg kg" 1 min" 1 in the postabsorptive tors who used various techniques (insu- individual diabetic subject is represented by a state (1,2,5,132-136,138). This flux lin-tolerance test [121,122], quadruple solid circle. Drawn from the data presented in provides glucose to meet the obligatory needs of the brain and other neural tisinfusion technique [123-126], insulin- refs. 1,4, 5, 134-136, and 167. sues that utilize glucose at a constant rate suppression test [127,128], forearm perof - 1 . 0 - 1 . 2 mg kg" 1 min" 1 (3,142). fusion technique [129,130], modified in8); and 3) in diabetic patients with overt Brain-glucose uptake accounts for sulin-tolerance test [131], the minimal fasting hyperglycemia, even maximally model [119], radioisotope turnover 50% of glucose disposal during the stimulating plasma insulin concentramethodology [131a]) provided addipostabsorptive state, is insulin indepentions do not elicit a normal glucose mettional support for the presence of insulin dent and therefore occurs at the same abolic response under euglycemic condiresistance in NIDDM. With the use of the rate during absorptive and interabsorptions (137-141). more physiological insulin-clamp techtive periods, and is not altered in nique, DeFronzo et al. (45) provided the NIDDM (142). After glucose ingestion, Site of insulin resistance most conclusive documentation of insuinsulin is released into the portal vein After the stimulation of insulin secretion, lin resistance in NIDDM. In the largest and carried to the liver where it binds to whole-body glucose homeostasis is depublished series involving >50 normalspecific receptors on the hepatocyte and weight diabetic patients with mild fasting suppresses hepatic glucose output. Failhyperglycemia (150 8 mg/dl), a 3 5 500-, ure of the liver to perceive this signal, 40% decrease in insulin-mediated glui.e., secondary to insulin resistance, re400sults in impaired suppression of HGP cose disposal was demonstrated in Total Body 3006-. Glucose and resultant hyperglycemia. In NIDDM NIDDM subjects (1,4,7,132-136; Fig. Uptake 2007), and there was a nearly complete sep- (mg/m -min) subjects with moderate fasting hyperglyaration between the diabetic and control cemia a consistent increase in basal HGP 100 groups. Three additional points are worhas been of 0.5 mg kg min o SO 100 150 200 250 thy of comment: 1) in lean NIDDM subdemonstrated. When extrapolated over Insulin Cone jects with more severe fasting hypergly24 h, the liver of a 70-kg diabetic subject cemia (198 10 mg/dl) the severity of Figure 8Dose-response curve relating the with mild to moderate fasting hyperglyinsulin resistance was only slightly (10plasma insulin concentration to the rate of insu- cemia produces an additional 50 g of 20%) greater than in diabetic subjects lin-mediated whole-body glucose uptake in con-glucose each day. In NIDDM the increase with mild fasting hyperglycemia (1,136trol (closed circles) and non-insulin-dependentin basal HGP is closely correlated 139); 2) the impairment in insulin action diabetes mellitus (open circles) subjects. (r = 0.847, P < 0.001) with the degree is observed at all plasma insulin concen- *P < 0.01 vs. control subjects. From Groop et of fasting hyperglycemia (1,2,4,7,132trations, spanning the physiological and al. (138). by the Journal of Clinical Investi- 136,138; Fig. 9). These results indicate pharmacological range (1,137-141; Fig. gation. that in NIDDM subjects with overt fast2

DIABETES CARE, VOLUME 15,

NUMBER 3 ,

MARCH

1992

325

Paihogenesis of NIDDM

20o

3.00 r = 0.847 p < 0.001 3.00 Hepatic Glucose (mgftg-min) 2.90

0

Glycogenolysis

0 o
o
oo
o

Hepatic Glucose Output (mg/m*min)

Hepatic 15_ Glucose Output (umol/kg-min) 1 5Gluconeogonesls

o
o

0 2.00

'&

9
,(0

<*
0

100

150

200

250

Portal Insulin Cone (uU/ml)

1.90 100 ISO 200 230 900

Fasting Plasma Glucose (mj'dl)

Figure 9 Summary of hepatic glucose production in 77 normal-weight non-insulindependent diabetic subjects (open circles) with fasting plasma glucose concentrations ranging from 105 to >300 mg/dl. Seventy-three age- and weight-matched control subjects are shown by the closed circles. In the 33 diabetic subjects with fasting plasma glucose levels < J 4 0 mg/dl (shaded area), the mean rate of hepatic glucose production was identical to control subjects. In diabetic subjects with fasting plasma glucose concentrations >140 mg/dl, there was a progressive rise in hepatic glucose production that correlated closely (r = 0.847, P < 0.001) with the fasting plasma glucose concentration. From DeFronzo et al. (11). by Metabolism.

Figure 10Dose-response curve relating the plasma insulin (portal vein) concentration to the suppression of hepatic glucose production in control (closed circles) and non-insulin-dependent diabetic (open circles) subjects with moderately severe fasting hyperglycemia. *P < 0.05, **P < 0.01 vs. control subjects. From Groop et al. (138). by the Journal of Clinical Investigation.

Figure 11 Contribution of gluconeogenesis and glycogenolysis in control deft bar) and noninsulin-dependent diabetic (right bar) subjects. All of the increase in total hepatic glucose production is accounted for by an excessive rate of gluconeogenesis. From Consoli et al. (157). by the American Diabetes Association.

ing hyperglycemia (>140 mg/dl) excessive HGP is an important factor in the development of fasting hyperglycemia. The close relationship between fasting plasma glucose concentration and HGP in N1DDM subjects is consistent with results published by other investigators (12,34,105,140-148). In control subjects there was little variation in the basal rate of hepatic glucose output and, not surprisingly, there was no correlation between basal HGP and fasting glucose concentration. In the postabsorptive state the plasma insulin concentration was increased two- to threefold in diabetic individuals (1,2,4,7,132-136,138; Fig. 2). Because hyperinsulinemia is a powerful inhibitor of HGP (1,2,138,149,150), it is clear that hepatic resistance to insulin is present in the postabsorptive state and

contributes to the excessive output of glucose by the liver. Because hyperglycemia also suppresses HGP (149-151), there may also be glucose resistance with respect to the inhibitory effect of hyperglycemia on hepatic glucose output. This later possibility has received little attention in NIDDM. In NIDDM subjects with fasting glucose concentrations >140 mg/dl all investigators consistently have found an elevated rate of basal HGP. However, apparently conflicting results have appeared concerning the suppression of HGP by insulin. DeFronzo et al. (1,2,4, 5,132-136) have reported a >90% inhibition of HGP by insulin, whereas others have reported an impaired suppression of HGP in NIDDM (140,141,152). These apparently discrepant results are explained by at least three factors. First, the dose-response curve relating plasma insulin to the inhibition of HGP is steep, with an ED50 of - 3 0 - 4 0 |xU/ml (138, 141; Fig. 10). Thus, at high plasma insulin concentrations, 100 |xU/ml, one is close to the Vmax for insulin's suppressive effects on HGP and a defect in insulin action might be masked. In contrast, at lower plasma insulin levels impaired suppression of HGP by insulin is readily

demonstrable. Second, the severity of hepatic insulin resistance appears to be related to the severity of diabetes. Thus, in NIDDM subjects with more severe fasting hyperglycemia, the ability of insulin to suppress HGP is impaired (138; Fig. 10). Third, the use of tritiated glucose to quantitate HGP under non-steady-state conditions underestimates total-body glucose disposal and overestimates the suppression of HGP (153,154). Although this problem can be overcome by maintaining the plasma glucose specific activity constant during the insulin clamp (155), this approach has not been applied to study HGP in NIDDM. The glucose, which is released by the liver, can be derived either from glycogenolysis or gluconeogenesis (2). Studies by Waldhausl et al. (156) showed that uptake of gluconeogenic precursors, especially lactate, is increased in NIDDM subjects. Consistent with this observation, Consoli et al. (157), with radioisotope turnover techniques, showed that 90% of the increase in HGP above baseline can be accounted for by accelerated gluconeogenesis (Fig. 11). Moreover, the increase in gluconeogenesis was closely correlated with the elevated fasting plasma glucose concentration. The mechanism(s) responsible for the increase in hepatic gluconeogenesis have

326

DIABETES CARE, VOLUME 15,

NI'MBFR 3,

MARCH

1992

DeFronzo

1.0 0.6 Net Splanchnic

-0.5 Balance (mg/kg-mln) .1.0

Qlucoso

-1.6
20 40 60 80 100 120 140 160 180 Time (Minutes)

Figure 12 Time course of change in the net splanchnic glucose balance in non-insulindependent diabetic (open circles) and control (closed circles) subjects. The difference between diabetic and control subjects is small, statistically insignificant, and cannot account for the marked impairment in total body glucose metabolism observed during the euglycemic insulin-clamp study. Also, note that the total amount of glucose disposed of by the splanchnic area represents <10% of the infused glucose load in both groups. From DeFronzo et al. (4). by the Journal of Clinical Investigation.

yet to be denned but include hyperglucagonemia (158-160), increased circulating levels of gluconeogenic precursors (lactate, alanine, glycerol) (157,161), increased free fatty acid (FFA) oxidation (138), enhanced sensitivity to glucagon, decreased sensitivity to insulin (138,140, 141,152), and/or increased activity of phosphoenol pyruvate carboxykinase, pyruvate carboxylase, or glucose-6phosphatase (the key regulatory enzymes in the gluconeogenic pathway). Splanchnic (gastrointestinal plus hepatic) glucose uptake. To examine the contribution of the splanchnic Giver plus gastrointestinal) tissues to overall glucose homeostasis DeFronzo et al. (4) used the hepatic vein catheterization technique in combination with the euglycemic insulin clamp. In the postabsorptive state there was a net release of glucose from the splanchnic area in both the control and NIDDM subjects (Fig. 12), reflecting glucose production by the liver. In response to insulin there was a prompt suppression of splanchnic glucose output (reflecting the inhibition of hepatic glucose production), and by 20 min the net glucose balance across the splanchnic region

was zero (i.e., there was no net uptake or release). After 2 h of sustained hyperinsulinemia there was a small net uptake of glucose by the splanchnic area (i.e., positive balance) that averaged 0.5 mg kg" 1 min" 1 . This uptake was virtually identical to the rate of splanchnic glucose uptake observed in the basal state, indicating that the splanchnic tissues, like the brain, are insensitive to insulin at least with respect to the stimulation of glucose uptake (2,3,149). Importantly, there was no difference between diabetic and control subjects in the amount of glucose that was taken up by the splanchnic tissues at any time during the insulin-clamp study. These studies illustrate another important point; i.e., that under conditions of euglycemic hyperinsulinemia very little glucose is taken up by the splanchnic (and, therefore, hepatic) tissues (Fig. 12). During the insulin clamp whole-body glucose uptake averaged 7 mg kg" 1 min" 1 and of this only 0.5 mg kg" 1 min" 1 , or 7%, was disposed of by the splanchnic region. Because the difference in insulin-mediated total-body glucose uptake between the NIDDM and control groups during the euglycemic clamp study was 2.5 mg kg" 1 min" 1 , from a purely quantitative standpoint it is clear that a defect in splanchnic (hepatic) glucose removal can not account for the magnitude of impairment in total-body glucose uptake. However, after glucose ingestion, both the oral route of administration and the resultant hyperglycemia conspire to enhance splanchnic (hepatic) glucose uptake (2,149,162), and under these conditions diminished hepatic glucose uptake has been shown to contribute to the impairment in glucose tolerance in NIDDM (2,143,163). In summary, it can be concluded that in NIDDM individuals with modest fasting hyperglycemia (< 150-160 mg/ dl), neither a defect in suppression of HGP nor a defect in hepatic glucose uptake can account for the major impairment (>40-50%) in tissue sensitivity to

Leg Glucose Uptake

(mg/kg leg wl per min)

20

40

60 80 100 120 140 160 180 Time (Minutes)

Figure 13 Time course of change in leg glucose uptake in non-insulin-dependent diabetic (open circles) and control (closed circles) subjects. In the postabsorptive state glucose uptake in the diabetic group was significantly greater than in control subjects. However, the ability of insulin (euglycemic insulin clamp) to stimulate leg glucose uptake was markedly delayed and reduced by 50% in the diabetic subjects. From DeFronzo et al. (4). by the Journal of Clinical Investigation.

insulin, which is observed during the euglycemic insulin clamp study. Peripheral (muscle) glucose uptake. During a euglycemic insulin clamp study, suppression of HGP is either normal or slightly impaired in NIDDM subjects with modest fasting hyperglycemia (150 mg/dl), whereas splanchnic glucose uptake is normal (Figs. 10 and 12). Therefore, by exclusion, peripheral (extrahepatic) tissues must be the primary site of insulin resistance. This has been confirmed by simultaneous femoral artery/vein catheterization to quantitate leg glucose exchange (4; Fig. 13). In the basal state leg glucose uptake was slightly but significantly (P < 0.01) increased in the diabetic group, a finding consistent with observations by other investigators (146,164-166). This increase in basal tissue glucose uptake also is consistent with results obtained from radioisotope turnover studies (1,2,4,5,12, 105,132-136,143-148,157,164-166). This accelerated rate of tissue glucose uptake in NIDDM subjects is due to the mass action effect of hyperglycemia that passively drives glucose into cells (2,151). However, the metabolic fate of the glucose that is taken up is not normal. In the postabsorptive state glucose

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

327

Paihogenesis ofNIDDM

oxidation is impaired in NIDDM subjects (1,136,167), and there is no net flux of glucose into glycogen. Moreover, muscle glycogen synthase activity and glycogen synthesis have been shown to be impaired in NIDDM (168-170). Therefore, a disproportionate amount of the excessive glucose that is taken up by the muscle and other tissues in the postabsorptive state must be converted to lactate (4, 171173), which is subsequently released and can serve as a substrate to drive gluconeogenesis by the liver (2, 156,174,175). Note, however, that from the quantitative standpoint tissues other than muscle must supply the precursors to sustain the elevated rate of basal hepatic gluconeogenesis. Although Capaldo, Sacca et al. (171) found a twofold increase in muscle lactate release during postabsorptive conditions, DeFronzo et al. (4) and Consoli et al. (172,174) found more modest increases (30%), whereas Kelley and Mandarino et al. (173) found no significant increase. Also, a similar cycle (glucose^lactate>glucose) is likely to occur locally within the splanchnic region (i.e., gut glucose uptake and conversion to lactate, release of lactate into the portal vein, uptake of lactate by the liver and conversion to glucose; 2), and this accelerated Cori cycle activity can be viewed as an important mechanism that provides substrate to sustain the accelerated rate of HGP in diabetic individuals with well-established fasting hyperglycemia. Most recently, attention has focused on the biochemical and anatomical heterogeneity within the liver itself (176). Thus, perivenous hepatocytes take up glucose and convert it to lactate, whereas periportal liver cells take up lactate and other gluconeogenic precursors, converting them to glucose (176). Thus, one could envision a self-contained Cori cycle within the liver itself where glucose is taken up at an accelerated rate (due to the mass action effect of hyperglycemia), as documented by DeFronzo et al. (4), and locally converted to lactate which

serves as a substrate to drive gluconeogenesis (176). During insulin-clamp studies performed in healthy control subjects, hyperinsulinemia elicits a prompt stimulation of leg glucose uptake, which reaches a plateau value of 10 mg kg leg wt~ 1 min~ * during the last hour of the study (4; Fig. 13). In contrast, in NIDDM subjects the onset of insulin action is delayed for 40 min and the absolute rate of insulin-mediated leg glucose uptake is reduced by 40-50%, although the insulin-infusion period is extended for an additional 60 min in the NIDDM group to allow insulin to more fully express its biological effects (4). By extrapolating from leg muscle to totalbody muscle glucose uptake, one can account for >90% of the impairment in insulin-mediated whole-body glucose disposal (1,4). These results provide conclusive evidence that the primary site of insulin resistance during the euglycemic insulin-clamp studies performed in NIDDM subjects resides in peripheral tissues, primarily muscle. Similar results have been reported with the forearm perfusion technique (129,165,171,173, 177-181). In adult humans adipocytes are resistant to the stimulatory effect of insulin on glucose metabolism, and < l - 2 % of an infused or ingested glucose load is taken up by fat tissue (182, 183). In a recent publication Marin et al. (184) suggested that this value may be somewhat higher, 4%. Even if one chooses to use this estimate, it is clear that the primary leg and forearm tissue responsible for most (>90%) glucose uptake must be muscle (184). A summary of insulin-mediated whole-body glucose utilization under conditions of euglycemic hyperinsulinemia insulin is shown in Fig. 14. Net splanchnic glucose uptake (quantitated by hepatic venous catheterization) is similar in control and NIDDM subjects and averages 0.5 mg kg" 1 min" 1 . Adipose tissue glucose uptake (not directly measured) is assumed to represent 2% of total glucose disposal (182-184).

7 -Splanchnic 6 5 mg/kg-min 4 3 2 1 0 Controls NIDD

Figure 14Summary of glucose metabolism during euglycemic insulin (+100 fxU/ml) clamp studies performed in normal-weight non-insulindependent diabetic (NIDD) and control subjects. See text for a more detailed discussion. From DeFronzo (I). by the American Diabetes Association.

Brain glucose uptake (1.0-1.2 mg kg" 1 min" 1 in the basal state; 142,185, 186) is unaffected by hyperinsulinemia (142). Muscle glucose uptake (extrapolated from leg catheterization data) in control subjects represents ~ 7 5 % of total glucose metabolism (1,3,4) and in NIDDM subjects accounts for essentially all of the impairment in insulin-mediated glucose uptake. Even if adipose tissue took up absolutely no glucose, it could, at best, explain only a small fraction of the defect in whole-body glucose metabolism. Glucose disposal during OGTT Although the insulin clamp (45) provides precise quantitation of glucose metabolism under fixed conditions of hyperinsulinemia and euglycemia, the normal route of glucose entry into body is via the gastrointestinal tract. Therefore, it is of interest to examine whether the mechanisms responsible for impaired glucose disposal during the euglycemic insulin clamp are similar to those after glucose ingestion. To address this question, Ferrannini et al. (187) used hepatic vein catheterization in combination with

328

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH

1992

DeFronzo

the ingested glucose load (188) to a high of 56% (189), with a mean value of 45%. Some of this variability results from estiMEAN SE RANGE mates of the fraction of body mass that is GLUCOSE (G) (G) muscle. This fraction is age and sex de1. INGESTED 68 3 55-93 pendent and can vary from as high as 50 4 32-56 2. APPEARING IN PERIPHERAL PLASMA (ORAL) 38%, determined by dissecting cadavers 15 2 5-20 3 . RELEASED BY THE LIVER (191), to 26%, determined by neutron 19 4 0.7-34 4. TAKEN UP BY SPLANCHNIC TISSUES activation (192). The data of Katz et al. 48 6 5 . TAKEN UP BY PERIPHERAL TISSUES 28-83 (190) have been revised according to this - 2 to 15 latter estimate. The results summarized 2 2 6. REMAINING IN THE GLUCOSE SPACE 8-31 7. UNRECOVERED 18 3 above indicate several important differ12-26 18 2 8. "SAVED" BY THE LIVER ences between the tissues responsible for -17 to 17 the maintenance of normal glucose ho43 TRUE NET SPLANCHNIC BALANCE ( 4 3) 37 4 20-62 SPLANCHNIC "CONSERVATION" ( 4 + 8 ) meostasis during intravenous versus oral SPLANCHNIC OVERALL CONTRIBUTION TO GLUCOSE glucose in nondiabetic subjects. In re22 4 1-42 HOMEOSTASIS ( 4 + 8 - 3 ) sponse to glucose ingestion 1) HGP is less completely suppressed, 2) peripheral Based on ref. 187. tissue (primarily muscle) glucose uptake is quantitatively less important, and 3) splanchnic glucose uptake is quantitatively much more significant. The less a dual isotope technique to examine the cose uptake (24-29%) and suppression complete suppression of HGP after glucontribution of splanchnic and periph- of HGP (50-60%) in nondiabetic subcose ingestion may be related to activaeral tissues to overall glucose disposal in jects by other investigators. Of the peof local sympathetic nerves that intion healthy subjects (Table 3). During the ripheral tissues that participate in the nervate the liver (193). 3.5 h after glucose (68 g) ingestion, 19 g disposal of ingested glucose, the contrior 28% of the oral load was taken up by bution of the brain (an insulin indepenFive studies used the doublethe splanchnic tissues, whereas 48 g or dent tissue) is fixed at 1.0-1.2 mg tracer technique to examine the disposi71% was disposed of by peripheral tis- kg" 1 min" 1 and accounts for 15-18 tion of an oral glucose load in NIDDM sues; during the same period basal HGP g or 20-30% of total glucose utiliza- (144,146,163,172,194) (Table 4). In declined by 53%. With the use of the tion after glucose ingestion (142,172, four of the studies forearm catheterizasame double tracer (146,172,188,189) 188-190). Four studies (172,188-190) tion was also used to look at muscle and hepatic vein catheter (190) tech- examined the role of skeletal muscle in glucose uptake (144,146,172,194). niques, remarkably similar percentages the disposal of oral glucose and the re- Note, however, that significant differhave been reported for splanchnic glu- sults have varied from a low of 26% of ences in body weight, severity of diabeTable 3Balance sheet for the disposition of an oral glucose load in nondiabetic subjects

Table 4 Summary of glucose metabolism following glucose ingestion in non-insulin-dependent diabetic patients

SUPPRESSION SPLANCHNIC INSULIN REF. FERRANNINI FIRTH MlTRAKOU FlRTH MCMAHON GLUCOSE UPTAKE O F HEPATIC GLUCOSE PRODUCTION ABSOLUTE TISSUE R d

INCREMENTAL TISSUE GLUCOSE CLEARANCE FOREARM GLUCOSE UPTAKE

INCREMENTAL FOREARM GLUCOSE UPTAKE

No.
163 144 172 146 194

TEST

BODY WEIGHT NORMAL WEIGHT OBESE OBESE. OBESE OBESE

RESPONSE*

OGTT OGTT OGTT

si 4

4
N , DELAYED

MM MM

si 4 si 4 si 4

tt

4 4 4 4 4 44 4 4

4 4
N

t t

sit

44 44 4 44 44

44 44 4 4 44 4 4

N N N

sl4

4 4 4

The number of arrows indicates the magnitude of change; OGTT, oral glucose tolerance test; MM, mixed meal; N, normal; si, slight. *In all studies the plasma insulin response was deficient when viewed relative to the plasma glucose concentration; Rd, rate of glucose disposal.

DIABETES CARE, VOLUME 15, NUMBER 3 , MARCH 1992

329

Pathogenesis of NIDDM

tes, plasma insulin response, and composition of the meal existed among the five studies (Table 4). Nonetheless, a very uniform picture emerges concerning the tissues and mechanisms responsible for the impairment in oral glucose tolerance (Table 4). The data of Ferrannini et al. (163) are illustrative of the metabolic abnormalities in NIDDM. During the 3.5 h after glucose ingestion, glucose tolerance was markedly impaired in NIDDM versus control subjects (mean plasma glucose 298 23 vs. 147 7 mg/dl). The disturbance in glucose metabolism in NIDDM patients was accounted for by two factors: 1) after glucose ingestion, tissue glucose uptake was significantly reduced (44 vs. 60 g over 3.5 h; Fig. 15) and 2) HGP continued at a higher rate (17 vs. 10 g over 3.5 h; Fig. 15). Splanchnic glucose uptake was similar in diabetic and control groups. These results indicate that the combined effect of hyperglycemia plus hyperinsulinemia to suppress HGP after glucose ingestion is impaired in NIDDM, and this difference in HGP (7 g over 3.5 h) accounts for

GLUCOSE fim/3 5 hr)'

Some special comment is warranted concerning whole-body tissue glucose disposal after glucose ingestion. When viewed in absolute terms, most studies have shown it to be normal (172) or increased (144,146,194; Table 4). However, when viewed in terms of the Figure 15Schematic representation of he- prevailing plasma glucose concentration, patic glucose production and tissue glucose up-the efficiency of glucose disposal, i.e., take in non-insulin-dependent diabetic (cross- glucose clearance, is severely reduced. hatched bars) and control (open bars) subjects Most importantly, it is not the absolute after glucose ingestion. Data have been extrap- rate of glucose disposal, but rather the olated to a 3.5-h period. See text for a more increment in glucose disposal above baseline that determines the rise in detailed discussion. From Ferrannini et al. (163). plasma glucose concentration above its by Metabolism.
TOTAL GLUCOSE ORAL ENDOGENOUS TISSUE GLUCOSE GLUCOSE GLUCOSE APPEARANCE' DISPOSAL

I1

J L

approximately one-third of the defect in total-body glucose homeostasis, whereas the defect in peripheral, presumably muscle, glucose uptake (14 g over 3.5 h) accounts for the remaining portion of the disturbance in overall glucose metabolism. From the quantitative standpoint, nearly identical results have been reported by others (144,146,172,194). Note that the double-tracer technique is associated with significant variability because it involves the subtraction of two large numbers (Ra of [3H] glucose minus Ra of [14C] glucose) to calculate suppression of HGP. Therefore, small differences in suppression of HGP between laboratories are likely to have little physiological meaning. The inherent variability of the method, variation in patient characteristics, and differences in insulin secretory response easily can explain the 10 20% differences in suppression of HGP reported by various investigators (144,146,163,172,194). The important message is that everyone has found that suppression of HGP is impaired in NIDDM and that this defect can account for about one-third to onehalf of the disturbance in whole-body glucose homeostasis following glucose ingestion. Splanchnic glucose uptake has been reported to be normal (163), slightly decreased (144,146, 172), or increased (194) and does not appear to contribute significantly to the impairment in oral glucose tolerance.

fasting value. In NIDDM subjects the basal rate of whole-body tissue glucose uptake is significantly enhanced (1,11). Consequently, in every published study the incremental response in tissue glucose uptake was moderately to severely reduced in NIDDM (144,146,163,172, 194; Table 4). Similar results have been reported for forearm muscle glucose uptake (144,146,172,194). Although the absolute rate of forearm glucose uptake after glucose ingestion is normal, the increment above baseline is impaired (144, 146,172,194; Table 4). Similar conclusions concerning tissue glucose uptake can be drawn from the data of Chen et al. (195), although these later investigators did not label the oral glucose load with [14C]glucose. Thus, all published results (144,146,163,172,194), including those by Gerich et al. (172), are totally consistent and indicate the important contribution of impaired muscle glucose disposal in NIDDM. The conclusion by Gerich (196) that muscle does not play an important role in impaired oral glucose tolerance in NIDDM stems from the failure to recognize that it is the incremental increase in muscle glucose disposal that determines the rise in plasma glucose concentration above baseline. In summary, the results of the OGTT indicate that both impaired suppression of HGP and decreased tissue (muscle) glucose uptake contribute approximately equally to the glucose intolerance of NIDDM. Diminished splanchnic glucose uptake is not an important contributory factor to the impaired disposal of an oral glucose load in NIDDM. Based on the preceding discussion, it is obvious that there are important differences between OGTT and euglycemic insulin clamp with respect to the quantitative importance of the liver versus peripheral tissues (muscle) in the glucose intolerance of NIDDM. During the insulin clamp the defect in suppression of HGP is quantitative small,

330

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

mia, and meal-stimulated hyperinsulinemia antedate the onset of 1GT and NIDDM in humans. Once overt fasting hyperglycemia (>140 mg/dl) has developed, HGP is elevated and correlates closely with the severity of fasting hyperglycemia (1,2,4,5,12,132-136,138,140, 143-148). It is important to emphasize, however, that by the time the fasting glucose concentration has risen to 140 mg/dl, the diabetic state has been present for a very long period, > 5 - 1 0 yr. Thus, these studies cannot establish whether, in the earliest stages of NIDDM, fasting hyperglycemia results from an excessive rate of HGP, decreased efficiency of tissue glucose uptake, or some combination of the two. This question was addressed by DeFronzo et al. (11). Basal HGP and tissue glucose uptake were quantitated in 77 lean NIDDM patients and compared with 72 age-, sex-, and weight-matched healthy individuals. Thirty-three NIDDM subjects had fasting plasma glucose levels below 140 mg/dl and 44 above 140 mg/dl (Fig. 9). The cutoff point of 140 mg/dl was chosen for two reasons: 1) it defines the level at which NIDDM is diagnosed according to the National Diabetes Data Group (198) and 2) it represents the plasma glucose concentration at which diabetic subjects tip over the top of Starling's curve of the pancreas and fasting plasma insulin levels begin to decline (Fig. 2). In the 33 diabetic individuals with fasting plasma glucose concentrations <140 mg/dl (Fig. In summary, the results obtained 9, shaded area), the 1 rate of 1 HGP with the OGTT are in agreement with (1.85 0.03 mg kg" min" ) was those from the euglycemic insulin clamp virtually identical to 1control 1subjects and, in addition, emphasize the impor- (1.84 0.02 mg kg" min" ). Howtant contribution of the liver to the im- ever, this normal basal rate of HGP was pairment in glucose homeostasis in maintained at the expense of a twofold greater fasting plasma insulin concenNIDDM. tration (20 2 vs. 11 1 jxU/ml, P < 0.001), indicating that hepatic insuFASTING HYPERGLYCEMIA IN lin resistance already is well established NIDDM: PANCREAS VERSUS early in the course of NIDDM. These LongiMUSCLE VERSUS LIVER observations are underscored by the retudinal and cross-sectional (1,22,27, ports of Eriksson et al. (41) and Gulli et 28,30,31-36) studies have demonal. (58) in individuals who are at high strated conclusively that the triad of risk to develop NIDDM later in life. In insulin resistance, fasting hyperinsuline-

whereas the impairment in tissue (muscle) glucose uptake is largely responsible for the defect in total-body glucose homeostasis. In contrast, after glucose ingestion both impaired suppression of HGP and diminished tissue (muscle) glucose uptake contribute approximately equally to the disturbance in wholebody glucose homeostasis in NIDDM. None of the available studies allows one to define whether the defects in hepatic and peripheral (muscle) glucose metabolism during OGTT are the result of insulin resistance, diminished insulin secretion, or an impairment in the mass action effect of glucose (i.e., glucose resistance) to promote its own uptake because plasma glucose and insulin concentrations varied widely between subjects and insulin resistance was not measured (144,146,163,172,194,195). Lastly, it should be noted that the impaired suppression of HGP is likely to assume added importance in everyday life where mixed meals are normally consumed. Under such conditions plasma amino acid, FFA, and glucagon concentrations rise, rather than fall as occurs during an OGTT. Hyperglucagonemia, increased delivery of gluconeogenic precursors, and enhanced FFA oxidation would be expected to further augment gluconeogenesis and lead to an even greater impairment in the suppression of HGP. In fact, such a sequence of events has been demonstrated by Shank et al. (197).

the former study (41), the normal glucose-tolerant first-degree relatives of NIDDM patients were markedly insulin resistant (insulin-clamp technique), yet basal HGP was perfectly normal. Similar findings were reported by Gulli et al. (58) in the normal glucose-tolerant offspring of two diabetic parents. In both studies (41,58) fasting hyperinsulinemia was present. These observations provide convincing evidence that in the earliest stages of NIDDM, HGP is not increased when viewed in absolute terms. Nonetheless, the presence of a normal rate of HGP in the face of fasting hyperinsulinemia establishes the presence of hepatic insulin resistance. Only the report by Gerich et al. (196) has suggested that HGP may be increased in patients with IGT and mild NIDDM. Note, however, that the number of patients in the Gerich study is small, the patient characteristics are poorly described, the patient population is obese, the treatment regimen of the diabetic subjects is not defined, the fasting plasma insulin concentration is not provided, and no information concerning their OGTT is given. Most importantly, in 6 of 10 individuals with fasting glucose levels between 6 and 7 mM (108-126 mg/dl) (Fig. 1 of ref. 196), the rate of HGP clearly was within the normal range and the mean rate of HGP (11.7 0.3 ixmol kg" 1 min" 1 ) was not significantly (t = 1.42, P = 0.18, calculated from the data presented in Fig. 1 of ref. 196) elevated compared with the 19 control subjects (11.2 0.2 jjunol kg" 1 min" 1 ) with fasting glucose levels between 5 and 6 mM (90-108 mg/dl). Thus, even the data of Gerich (196) do not support his contention. Note that in a previous publication by the same group, Gerich et al. (164) did not specifically examine individuals with fasting glucose levels between 6 and 7 mM. In control subjects the fasting glucose was 5.3 0.1 mM (mean 2SD 4.5-6.1 mM), whereas in diabetic subjects it was 10.6 0.6 mM. Thus, the range of plasma glucose concentrations (6-7 mM), which are of most

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

331

Pathogenesis of NIDDM

interest, were not examined. Moreover, the obesity index of the diabetic (body mass index [BMI] 27.8 0.6 kg/m2) patients in the earlier publication of Gerich et al. (164) was much greater than reported in the more recent review article (BMI 25.8 0.8 kg/m2; 196), and it is obvious that the patient populations in these two papers (164,196) must be quite different. Therefore, it is not appropriate to equate the metabolic disturbances described in the earlier paper (164) with the patient population reported in the more recent publication (196). In summary, neither the previous (164) nor present (196) publication of Gerich justifies the statement that in "individuals with fasting plasma glucose concentrations between 6 and 7 mM, rates of glucose production are elevated" (196). Rather, the data of Gerich et al. (196) are consistent with the previously published results of DeFronzo et al. (11), Eriksson et al. (41), and Gulli et al. (58) and indicate that HGP is not elevated in the early stages of NIDDM in humans. Similar results have been reported in a type II diabetic primate model (33,34). During the postabsorptive state the rate of HGP equals the rate of glucose uptake by all of the tissues in the body. It follows, therefore, that tissue glucose utilization, when viewed in absolute terms, is increased in NIDDM. However, when viewed in the context of the elevated plasma glucose and insulin concentrations, it is clear that the efficiency of tissue glucose uptake, i.e., the metabolic clearance rate of glucose, must be reduced early in the course of NIDDM (11; Fig. 16). As the fasting plasma glucose concentration rises from 105 to 140 mg/ dl, glucose clearance declines linearly (r = 0.700, P < 0.001; Fig. 16), whereas HGP remains constant (Fig. 9). As the diabetic condition worsens, the contributions of excessive HGP and diminished tissue glucose clearance to fasting hyperglycemia reverse themselves. With fasting plasma glucose levels > 140 mg/ dl, the restraining effect of hyperinsulinemia on the liver is lost and HGP increases

2.3 "* 2.3 2.1

s

*&:
%^*9>

Glucoso Clearance 1.9 ml/kg-mln)

1.7 1.3

V*p
o

O &

o
0

0 0 0

13 1.1 09

o c 0 J$)
0

00

0)
0

Fasting Plasma Glucose (mg/dl)

Figure 16 Summary of the metabolic clearance rate of glucose in 77 normal-weight noninsulin-dependent diabetic subjects (open circles) with fasting plasma glucose concentrations ranging from 105 to >300 mg/dl. Seventy-two ageand weight-matched control subjects are shown by the open circles. In the 33 diabetic subjects with fasting plasma glucose concentrations <140 mg/dl (shaded area), the glucose clearance rate fell precipitously and was inversely correlated (r = -0.697, P < 0.001) with the increase in plasma glucose concentration. At fasting plasma glucose levels >140 mg/dl, the rate of decline in glucose clearance began to slow and reached a plateau at glucose levels >180 mg/dl. From DeFronzo et al. (11). by Metabolism.

prevent excessive HGP (Fig. 9), yet inadequate to maintain a normal basal rate of tissue glucose clearance (Fig. 16). This paradox is readily appreciated by examining the dose-response relationship between the plasma insulin concentration versus HGP and tissue glucose disposal (8,138,140,141,152). When the plasma insulin concentration is increased from 8 to 15 to 27 |xU/ml in nondiabetic subjects, basal HGP is suppressed by 33 and 68%, respectively, whereas whole-body glucose uptake fails to increase above baseline (138). Thus, in NIDDM subjects, as the fasting plasma insulin concentration increases (Fig. 2), the resultant hyperinsulinemia is sufficient to offset any hepatic insulin resistance and to maintain a normal absolute rate of hepatic glucose output. However, the small increment in fasting plasma insulin is not capable of stimulating tissue glucose uptake (138,200-202), and the glucose clearance drops progressively (Fig. 16). The decline in whole-body glucose clearance in the postabsorptive state can only partially be explained by a decrease in glucose uptake by peripheral tissues (11). In NIDDM subjects DeFronzo et al. (4) showed that basal leg glucose uptake is significantly increased. Moreover, because the rise in leg glucose uptake was proportional to the increase in fasting glucose concentration, leg (including muscle) glucose clearance was virtually identical in control and NIDDM subjects (4). In contrast, Gerich et al. (164) and Firth et al. (146) reported slightly, although significantly, reduced rates of glucose clearance by forearm tissues in NIDDM in the postabsorptive state. With the hepatic vein catheter technique, DeFronzo et al. (4) showed that at least part of the decline in whole body glucose clearance resides within the splanchnic tissues Giver plus gastrointestinal). However, from a purely quantitative standpoint tissues, in addition to those within the splanchnic region (4) and muscle (146,164), also must contribute to the decline in whole-body glu-

progressively (r = 0.847, P < 0.001; Fig. 9), whereas the decline in wholebody glucose clearance plateaus at glucose levels between 140 and 180 mg/dl (Fig. 16). At glucose levels >180 mg/dl the whole-body glucose disposal rate rises in direct proportion to the elevated fasting glucose concentration and the glucose clearance remains constant, although reduced by 40-50% compared with the postabsorptive state. Because the renal tubular threshold for glucose reabsorption is 180 mg/dl (199), the constancy of glucose clearance at plasma glucose concentration above this level most likely is explained by the development of glucouria. Two important questions remain to be answered concerning glucose metabolism in the postabsorptive state. First, why is the development of fasting hyperinsulinemia (Fig. 2) sufficient to

332

DIABETES CARE, VOLUME

15,

NUMBER 3 ,

MARCH

1992

DeFronzo

Plasma Glucose

Plasma Insulin
25 20

(mg/dl)

15 10 5

Hepatic Glucose Production 1.9

1.8 (mg/kg*min) 1.7

Glucose Clearance
2.0 1.9 1.8 (ml/kg*min) 1.7 1.6

1.6 1.5

Figure 17The effect of overnight insulin infusion to normalize the basal rate of hepatic glucose production in 19 normal-weight non-insulin-dependent diabetic (NIDDM) subjects (shaded bars) and in 72 age- and weight-matched control subjects (cross-hatched bars). Despite similar rates of hepatic glucose production and a >2-fold increase in the plasma insulin concentration (P < 0.01), the fasting plasma glucose remained significantly elevated in NIDDM vs. control subjects. The decreased glucose clearance in NIDDM subjects indicates a diminished efficiency of tissue glucose uptake. From DeFronzo et al. (11). by Metabolism.

cose clearance. The brain and other neural tissues represent a prime candidate to explain the decrease in glucose clearance. In the postabsorptive state, 5 0 60% of total glucose disposal occurs in the cerebral tissues. Brain-glucose uptake is insulin independent and saturates at a plasma glucose concentration of - 4 0 mg/dl (142,185,186). It is not surprising, therefore, that brain-glucose uptake remains normal in NIDDM individuals, despite fasting hyperglycemia (142), and that the brain- (and consequently whole body) glucose clearance declines in NIDDM subjects with progressive increases in the basal glucose concentration. It remains to be defined whether tissues, in addition to muscle, splanchnic, and brain, also contribute to

the decline in basal glucose clearance in NIDDM. It should be emphasized, however, that the lower absolute rate of glucose disposal in NIDDM patients studied at comparable levels of fasting glucose and a twofold greater plasma insulin concentration as control subjects (11) indicates that the decline in whole-body glucose clearance must be accounted for by tissues other than the brain. The defects in basal HGP and tissue glucose disposal are better appreciated by studying NIDDM subjects after an overnight infusion of insulin designed to normalize basal HGP (11). To accomplish this goal a peripheral plasma insulin concentration of 24 |xU/ml (2-fold greater than control subjects) was required (Fig. 17). In nondiabetic subjects

a similar infusion of insulin inhibited basal HGP by >60% (Fig. 10). Despite similar rates of HGP in control and diabetic subjects, the fasting glucose concentration remained elevated in the latter (105 3 vs. 92 1 mg/dl, P < 0.01). The persistent elevation in the basal plasma glucose can only be explained by a decreased efficiency of tissue glucose removal, as reflected by the decreased glucose clearance (1.71 vs. 2.00 mlkg" 1 min" 1 , P < 0.01). Further evidence in support of an impairment in tissue glucose uptake comes from studies in which NIDDM individuals received an overnight insulin infusion to normalize the fasting plasma glucose concentration (11; Fig. 18). At identical plasma glucose concentrations, the absolute rate of tissue glucose uptake was significantly reduced in NIDDM versus control subjects (1.68 vs. 1.84, P < 0.01), although the plasma insulin concentration was more than twofold elevated in the former group. These data convincingly demonstrate that under comparable conditions of euglycemia the tissues of NIDDM subjects are unable to metabolize glucose at rates comparable to control subjects. In contrast to the suggestion of Gerich (164, 196), these results cannot simply be explained by a failure of the brain to passively enhance its uptake of glucose in response to a progressive rise in the fasting plasma glucose concentration. From available data it is not possible to establish which defect, i.e., hepatic insulin resistance or decreased efficiency of tissue glucose removal, develops first in the evolution of fasting hyperglycemia in NIDDM. Three equally plausible sequences can be postulated. First, it is possible that both defects develop in parallel. As hyperglycemia ensues (due both to excessive HGP and decreased tissue glucose uptake), basal insulin secretion is stimulated (Fig. 2). The resultant hyperinsulinemia restores HGP to baseline, whereas the resultant hyperglycemia returns tissue glucose uptake to normal. This sequence of events would result in normal basal rates of

DIABETRS CARE, VOLUME 15,

NUMBER 3, MARCH

1992

333

Pathogenesis of NIDDM

(204-209), and 2) it has a direct suppressive effect on HGP. Because these two metabolic actions of insulin offset each other, the absolute basal rate of HGP remains unaltered. In addition, hepatic insulin resistance could initiate the (ixU/ml) development of fasting hyperglycemia in NIDDM by causing a small imperceptible rise in plasma glucose concentration. This, in turn, would stimulate insulin secretion, returning the elevated rate of basal HGP to normal. Tissue glucose upGlucose Disposal Glucose Clearance take, however, would remain unaltered 1.9 2.1 because the small rise in plasma insulin concentration is insufficient to stimulate 2.0 tissue glucose uptake (8,138,200-202). 1.9 (ml/kg*min) This would lead to an "apparent" decline in the efficiency of tissue glucose re1.8 moval, i.e., tissue glucose clearance would fall. Gerich (196) stated that "for 1.7 plasma glucose to increase, glucose production must exceed glucose uptake." Figure 18The effect of overnight insulin infusion to normalize the fasting plasma glucose Strictly speaking, this is not correct. concentration in 11 normal-weight non-insulin-dependent diabetic (NIDDM) subjects (shaded bars) What is true is that for the plasma gluand in 72 age- and weight-matched control subjects (cross-hatched bars). Despite identical plasma cose concentration to increase, glucose insulin concentrations, the absolute rate of whole-body tissue glucose uptake in the postabsorptive production must transiently exceed glustate was significantly reduced in the NIDDM group (P < 0.01). From DeFronzo et al. (11). by cose uptake or, conversely, tissue glucose Metabolism. uptake must transiently decrease below the rate of glucose production (see preceding discussion). Studies performed under steady-state conditions after a HGP and tissue glucose uptake but at the and insulin levels in the development of metabolic perturbation has occurred and expense of fasting hyperglycemia and IGT and NIDDM) with normal serum the system has reequilibrated cannot refasting hyperinsulinemia (1,11). This is calcium and phosphate concentrations construct the sequence of events that led analogous to the development of second- (i.e., analogous to the normal rate of to the establishment of the new steady ary hyperparathyroidism in patients with HGP and tissue glucose uptake in state. Thus, it is not possible to define impaired renal function. As the glomer- NIDDM). Importantly, because the small which defect, i.e., hepatic or peripheral ular filtration rate declines, there is a increment in plasma insulin concentrainsulin resistance, comes first in the natsmall imperceptible rise in the serum tion has no or little stimulatory effect on ural history of NIDDM, or if both defects phosphate concentration, which in turn tissue glucose uptake (8,138,200-202), develop simultaneously and in a parallel leads to a small imperceptible decline in whole-body glucose clearance falls. fashion. the serum calcium concentration (203). Alternatively, decreased effiThe resultant hypocalcemia causes stim- ciency of tissue glucose uptake could It is interesting to speculate about ulation of the parathyroid hormone, represent the primary defect responsible which factors might be responsible for which enhances renal phosphate excre- for fasting hyperglycemia in NIDDM. As the progressive rise in HGP at fasting tion and returns the serum phosphate the plasma glucose concentration rises, plasma glucose concentrations >140 concentration to normal. The restoration insulin secretion is enhanced and the re- 160 mg/dl. One obvious explanation is of normophosphatemia leads to a rise in sultant hyperglycemia returns tissue glu- the progressive decline in basal plasma serum calcium concentration to baseline cose uptake to normal, whereas the re- insulin level that occurs in NIDDM pavalues. As this scenario is repeated, the sultant hyperinsulinemia has two tients with fasting plasma glucose conresult is a significant increase in serum opposing actions: 1) it induces hepatic centrations > 140-160 mg/dl. Consisparathyroid hormone concentration (i.e., insulin resistance by downregulating tent with this thesis, basal HGP and analogous to the elevated plasma glucose both receptor and postreceptor events fasting insulin levels in this group were

Plasma Glucose

Plasma Insulin

334

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

strongly and inversely correlated (11). Other factors that could account for the rise in HGP include a worsening of hepatic insulin resistance with progressive severity of diabetes or an accelerated rate of hepatic gluconeogenesis. In summary, the results reviewed above demonstrate that in NIDDM subjects with mild fasting hyperglycemia (<140 mg/dl) both a decrease in the efficiency of tissue glucose uptake and hepatic insulin resistance are present and contribute to the elevated postabsorptive plasma glucose concentration. In absolute terms HGP is not increased, whereas tissue glucose clearance is reduced. In NIDDM subjects with more severe fasting hyperglycemia (> 140 mg/dl) glucose clearance falls further, whereas HGP is increased and correlates closely with the elevation in fasting glucose concentration.

jects. All subjects received a 100-g oral glucose load to provide a measure of glucose tolerance and insulin secretion. Insulin sensitivity was quantitated with the euglycemic insulin-clamp (4-100 |xU/ml) technique (45), and indirect calorimetry (210) was used to measure rates of glucose oxidation and nonoxidative glucose disposal. The latter measure has been shown to primarily reflect glycogen synthesis (168,211,212). An integrated summary of the results are presented in Fig. 4. In obese nondiabetic subjects, tissue sensitivity to insulin was markedly reduced by 40 50%, yet oral glucose tolerance remained normal because the (3-cell was capable of appropriately augmenting its insulin secretory capacity to offset the defect in insulin action. The progression to IGT was associated with a further reduction in insulin-mediated glucose disposal (Fig. 4). However, glucose tolerance was only mildly impaired because the (3-cell DYNAMIC INTERACTION was able to further augment its secretion BETWEEN INSULIN ACTION AND INSULIN SECRETION IN NIDDM of insulin to counteract the deterioration NIDDM subjects are characterized by ab- in insulin sensitivity. Further progression normalities in both tissue (muscle and from IGT to overt diabetes mellitus was liver) sensitivity to insulin and pancreatic heralded by a modest decline in insulin insulin secretion. To fully appreciate the secretion without any worsening of the evolution of the full-blown diabetic con- insulin resistance. However, this modest dition, it is necessary to examine the dy- decline in insulin secretion, in the presnamic interaction between insulin action ence of severe insulin resistance, resulted and insulin secretion in the same indi- in marked glucose intolerance and frank vidual over a wide range of insulin sen- diabetes mellitus (Fig. 4). The obese disitivity. Recently, three groups have abetic person has tipped over the top of provided such information (1,22,27- Starling's curve of the pancreas (1; Figs. 32,118,135,136,167). DeFronzo (1) and 2 and 3). Although the plasma insulin Felber et al. (31,136,167) studied four response remained increased compared groups of subjects who were subdivided with control subjects, it was not approas follows: group 1, 47 obese subjects priately elevated for the degree of insulin with a mean ideal body weight of 148% resistance. Lastly, the obese diabetic (this group was further subdivided into group with a low insulin response man24 subjects with normal glucose toler- ifested the greatest glucose intolerance, ance and 23 with 1GT); group 2, 35 obese and this was due entirely to a progressive (ideal body weight 147%) subjects with decline in insulin secretion without any diabetes mellitus (fasting glucose >140 change in insulin sensitivity (Fig. 4). The mg/dl) (group 2 was further subdivided preceding sequence of events leading into those with a hyperinsulinemic re- from obesity with normal glucose tolersponse and those with a hypoinsulinemic ance to obesity with IGT to obesity with response during an OGTT); group 3, 13 diabetes has been confirmed in a longnormal-weight age-matched control sub- term prospective study (31) involving

the same subjects who previously participated in the longitudinal study (1,136,167). These observations underscore the critical interaction between insulin resistance and insulin secretion. Both abnormalities must be looked at in concert. Insulin resistance alone is not sufficient to cause frank diabetes mellitus. The onset of overt diabetes requires a concomitant defect in insulin secretion. This construct is entirely consistent with the classic overfeeding studies of Sims et al. (213). Healthy lean subjects who received an increased caloric intake gained 20-30 lb over 3 - 5 mo, and this was associated with the development of moderate to severe insulin resistance in muscle tissue (forearm catheterization technique). Nonetheless, glucose tolerance remained normal because the pancreas was capable of augmenting its insulin secretory capacity to precisely counterbalance the impairment in insulin action. The sequence of events described above has been confirmed in both humans and monkeys (22,27,30-34,118). Reaven et al. (118) studied five groups of lean subjects with a wide range of glucose tolerance. As reported by DeFronzo et al. (1,136,167) in obese individuals (Fig. 4), in lean individuals' (118) progression from normal glucose tolerance to IGT was associated with the development of severe insulin resistance, which was offset by an increase in insulin secretion (Fig. 19). The emergence of NIDDM was associated with a stepwise decline in insulin secretion, which paralleled the decrease in glucose tolerance (Fig. 19); tissue sensitivity to insulin did not change or decreased only slightly as subjects progressed from IGT to overt NIDDM (Fig. 19). A similar sequence of events has been demonstrated in a prospective study in Pima Indians (22,27,30,32). This study (22,27,30,32) is of particular importance because in vivo insulin action is inherited as a familial trait in this population (212). The same sequence of events leading to the development of NIDDM in whites (1,41,118,136,167) and Pima Indians

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

335

Pathogenesis of NIDDM

Mean Plasma Insulin Cone. (nU/ml)

100
50
400 -I

T
T
T

Glucose 3 0 0 ' Clearance 200 (ml/m?min)

CON

<8

8-15
Fasting Plasma Glucose

Figure 19Insulin-mediated glucose clearance (measured with the insulin suppression test) and the plasma insulin response (measured with oral glucose tolerance test) in control (CON) subjects, subjects with impaired glucose tolerance (1GT), and non-insulin-dependent diabetic individuals (shaded bars) with varying severity of glucose intolerance. See text for a more detailed discussion. The fasting plasma glucose concentrations is given in mM. From Reaven et al. (118). by Diabetologia.

(22,27,30,32) has been described in the rhesus monkey (33,34). With advancing age, the monkey becomes obese and develops a diabetic condition that closely mimics NIDDM in humans. In this primate model insulin resistance represents the earliest detectable abnormality. However, because of the compensatory increase in insulin secretion, fasting plasma glucose concentration, basal HGP, and glucose tolerance remain normal. Three recent studies have attempted to define the primary defect responsible for NIDDM by examining insulin action and insulin secretion in the normal glucose-tolerant first-degree relatives of NIDDM patients (41) and in the offspring of one (28) and two diabetic parents (58), respectively. The rationale underlying this approach is based on the assumption that NIDDM is an inherited disease and that the study of normal glucose-tolerant individuals (who are at high risk to develop diabetes later in life;

214-220) of affected diabetic family members will reveal the basic metabolic defect in NIDDM. In the studies of Eriksson et al. (41), Gulli et al. (58), and Warram et al. (28) insulin secretion (both the 1st and 2nd phases) in response to oral and intravenous glucose was increased. In contrast, insulin sensitivity was reduced by 3040%. These results demonstrate that insulinopenia does not precede the development of NIDDM in humans. Rather, insulin resistance represents the earliest detectable defect in NIDDM. As discussed above, it has become fashionable to view the insulin resistance in NIDDM as the primary defect and hyperinsulinemia as a compensatory response. Although somewhat speculative, it is entirely plausible that the increase in insulin secretion represents the primary disturbance and that the insulin resistance results from a downregulation of both receptor and postreceptor events

by chronic sustained hyperinsulinemia (204-210). Consistent with this hypothesis, Lillioja et al. (32) recently documented that in Pima Indians with IGT of a similar degree to that in whites, the plasma insulin response during an OGTT was 50% greater. This greater hyperinsulinemic response could not be explained by differences in plasma glucose levels during the OGTT, obesity, age, sex, or severity of insulin resistance between the two groups. Thus, it is possible that, at least in Pima Indians, augmented (3-cell function represents the initial or primary disturbance in NIDDM and that the insulin resistance develops secondarily to chronic hyperinsulinemia. In summary, the results reviewed in this section indicate that in the earliest detectable stage of NIDDM, i.e., normal glucose-tolerant offspring of two diabetic parents and normal glucose-tolerant relatives of NIDDM individuals, insulin resistance is well established and is offset by the presence of compensatory hyperinsulinemia. Overt diabetes mellitus develops only in individuals whose (3- cells are unable to meet the increased and sustained demand for insulin secretion. CELLULAR MECHANISMS OF INSULIN RESISTANCE Stimula tion of glucose metabolism by insulin requires that the hormone must first bind to specific receptors that are present on the cell surface of all insulin-target tissues (221-224). During the process of insulin binding and internalization, a "second messenger" is generated (225227). Unfortunately, the precise nature of this second messenger remains controversial at present. Many candidates have been suggested, but, as of yet, none has gained universal acceptance. Considerable evidence suggests that at least one of the second messengers for insulin action, tyrosine kinase, is an enzyme complex, which is an integral part of the receptor itself (221,223,224,228,229). Once the second messenger for insulin action has been generated, it initiates a series of events that promote intracellular

336

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

glucose metabolism. The initial step involves activation of the glucose-transport system with the resultant influx of glucose into insulin-target tissues, primarily muscle (230-235). The free glucose that has entered the cell subsequently is phosphorylated by hexokinase and metabolized by a series of enzymatic steps that are under the control of insulin. Of these, the most important are glycogen synthase (which controls glycogen synthesis) and pyruvate dehydrogenase (which regulates glucose oxidation).

Insulin receptor
In NIDDM individuals both receptor and postreceptor defects have been shown to contribute to the insulin resistance. Numerous studies have demonstrated that insulin binding to monocytes and adipocytes from NIDDM patients is reduced on mean by - 2 0 - 3 0 % (236-246). The reduction in insulin binding is due to a decrease in the number of insulin receptors without alteration in insulin-receptor affinity. In addition to a decrease in the number of cell surface receptors, various defects in insulin-receptor internalization and processing have been described (235,245,247). However, some caution should be used in interpreting these insulin binding studies because the two major organs for insulin action in vivo are the liver and muscle, and few studies have examined insulin binding to these tissues in humans. In obese diabetic individuals insulin binding to solubilized receptors obtained from skeletal muscle biopsies has been shown to be normal when expressed per milligram of protein (248,249). Several other lines of evidence indicate that diminished insulin binding cannot satisfactorily explain the defect in insulin action in NIDDM: 1) decreased insulin-receptor number cannot be demonstrated in approximately one-third of all NIDDM subjects, especially those with high fasting plasma glucose levels (239,250-262); 2) Lonnroth et al. (257), Olefsky et al. (258), and others have failed to find a correlation between the reduction in insulin binding

and the severity of insulin resistance in NIDDM patients; and 3) studies that have examined the dose-response relationship between insulin-mediated glucose uptake and the plasma insulin concentration have demonstrated the presence of a severe postreceptor defect in insulin action (138,140,141,152,259). In patients with IGT and/or with very mild diabetes, the dose-response curve was shifted to the right but very high plasma insulin concentrations were able to overcome the insulin resistance and elicit a normal glucose metabolic response. This picture is most consistent with a receptor defect in insulin action (260), and this has been documented in isolated adipocytes (152). In contrast, in diabetic patients with moderate to severe fasting hyperglycemia, the dose-response curve was both shifted to the right and exhibited a decrease in maximum rate of insulin-stimulated glucose disposal (Fig. 8). These results are most consistent with a postreceptor defect in insulin action (260), and indeed no decrease in insulin binding could be demonstrated in this group of diabetic patients (152). In summary, the above results indicate that in diabetic subjects with moderate to severe fasting hyperglycemia, postbinding defects in insulin action are responsible for the observed insulin resistance. In individuals with IGT or mild fasting hyperglycemia, a modest defect in insulin binding may exist. These individuals are usually hyperinsulinemic, and it is likely that the mild decrease in insulin binding results from a downregulation of insulin-receptor number by sustained hyperinsulinemia (204,205,207). In the most general sense, postbinding defects in insulin action can be explained by one of three metabolic disturbances: impaired generation of insulin's second messenger, diminished glucose transport into cells, and a postglucose transport abnormality in some critical enzymatic step involved in glucose utilization.

Insulin second messenger

During the last decade there has been an explosion of knowledge about the structure-function of the insulin receptor and the mechanisms that couple the receptor to the intracellular sites of insulin action (221-224,229,235). Nonetheless, the precise identification of the second messenger for insulin action has remained elusive. Phosphorylation/dephosphorylation of key intracellular proteins appears to be an important signaling mechanism that couples insulin binding to the intracellular action of insulin ( 2 2 1 224,235,238,261). The insulin receptor is a complex glycoprotein that is comprised of two a-subunits and two P-subunits that are linked by disulfide bonds. The a-subunit of the insulin receptor faces outward and contains the insulin binding domain, whereas the (3-subunit faces inward and expresses insulinstimulated kinase activity directed toward its own tyrosine residues (221224,235). The a-subunit is totally extracellular, whereas the P-subunit is a transmembrane protein. Considerable evidence indicates that insulin-receptor phosphorylation of the (3-subunit, with subsequent activation of insulin-receptor tyrosine kinase, represents an important second messenger for many of the hormone's varied actions (221-224,235). In vitro mutagenesis experiments have shown that insulin receptors, which completely lack tyrosine kinase activity, are ineffective in mediating insulin stimulation of cellular metabolism (262264). Similarly, mutations in any of the three major phosphorylation sites (at residues 1146, 1150, and 1151) diminish insulin-receptor kinase activity and lead to a decrease in insulin's acute metabolic effects (265). Because of the potentially important role of tyrosine kinase in the insulin-signaling mechanism, numerous investigators have examined tyrosine kinase activity in various cell types (skeletal muscle, adipocytes, hepatocytes, and erythrocytes) from normal-weight and obese NIDDM subjects (240,244,246, 248,252,266-270). In nondiabetic sub-

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH

1992

337

Pathogenesis ofMDDM

jects insulin-receptor tyrosine kinase activity increased linearly with the glucose disposal rate throughout the physiological range of plasma insulin concentrations (269). In diabetic and obese subjects most investigators found a severe (>50%) reduction in tyrosine kinase activity that could not be explained by an alteration in insulin-receptor number. The study of Freidenberg et al. (268) is particularly informative with regard to this. In obese NIDDM patients weight loss, which decreased the fasting glucose from 205 to 118 mg/dl, led to a near normalization of insulin-receptor tyrosine kinase activity. This observation suggests that the defect in tyrosine kinase is acquired secondarily to some aspect of poor metabolic control, hyperinsulinemia, or insulin resistance, all of which improved after weight loss. The human insulin-receptor gene has been cloned (271) and shown to be quite large, with 4.1 kb of coding sequence that is distributed over 22 axons. At least eight point mutations have been identified (228,235,271,272). Each mutation is associated with a specific defect in insulin-receptor function, and several are characterized by impaired insulinreceptor tyrosine kinase activity (228,235,272). Common to all of these insulin-receptor gene defects is the presence of severe insulin resistance, and several excellent reviews of these insulinresistance syndromes have been published recently (219,235,273). Although frank NIDDM may occur, it is not a universal finding. Nonetheless, this raises the possibility that an abnormality in the insulin-receptor gene may be responsible for the typical NIDDM in humans. Although restriction-fragmentlength polymorphism in the insulinreceptor gene has been reported in a small percentage of NIDDM subjects with insulin resistance (274-277), this has not been a reproducible finding (278-282), and the insulin-receptor cDNA sequence in NIDDM patients homozygous for at least one of these restriction-fragment-length polymorphisms

has been shown to be normal (283). Until recently, information concerning the cDNA sequence of the insulin-receptor gene has been limited to one white and three Pima Indians with NIDDM, in whom it was shown to be normal (283285). Because insulin resistance is an inherited characteristic in Pima Indians (213), the insulin-receptor gene is an unlikely candidate to explain the development of NIDDM in this insulin-resistant population. With single-stranded conformation polymorphism analysis, O'Rahilly et al. (286) examined the five exons that encode the tyrosine kinase domain of the insulin receptor in 30 unrelated white NIDDM individuals. Exons 19, 20, and 21 were entirely normal, whereas one missense mutation was discovered in exon 18. Five abberant patterns were detected in exon 18 and these were explained by three different nucleotide substitutions, two of which were common silent polymorphisms. The authors concluded that two of these previously unreported amino acid substitutions in the highly conserved tyrosine kinase domain could represent potentially important candidates for the insulin-resistance gene in NIDDM. These results, if applicable to the population of NIDDM patients in general, indicate that a structural abnormality in the insulin-receptor gene can account for only a small percentage ( < 2 - 3 % at most) of typical NIDDM individuals. In a recent study McGuire et al. (287) examined the activity of protein tyrosine phosphatase activity in skeletal muscle from seven insulin-resistant subjects with IGT. After insulin binds, insulin-receptor protein tyrosine kinase is activated and phosphorylates tyrosyl residues on many proteins, bringing about insulin's many and varied actions. The extent of tyrosine phosphorylation represents a balance between the antagonistic action of protein tyrosine phosphatases and the stimulatory effect of protein tyrosine kinases. In the insulinresistant subjects studied by McGuire et al. (287) basal protein tyrosine phos-

phatase activity was increased and failed to suppress normally in response to insulin. These observations, if confirmed in non-Pima Indian NIDDM subjects, suggest a potential role for excessive protein tyrosine phosphatase activity in the development of insulin resistance in NIDDM individuals. In addition to impaired stimulation of tyrosine kinase activity by insulin and impaired suppression of protein tyrosine kinase by insulin, production of an inhibitor of insulin-receptor tyrosine kinase has been described in at least one NIDDM patient with severe insulin resistance (288). How commonly such defects are found in the insulin-receptor tyrosine kinase signaling mechanism in garden variety NIDDM remains to be determined. Before ending any discussion about the role of insulin-receptor tyrosine kinase in the insulin resistance of NIDDM, it is important to note that biologically relevant substrates for the kinase have yet to be defined. Therefore, the true significance of any tyrosine kinase defect remains to be elucidated. In addition to tyrosine kinase, a number of other potential messengers for insulin action have been proposed. These include interaction of the insulin receptor with GTP binding proteins (221), stimulation of a specific membrane-bound phospholipase that results in the generation of phosphoinositides and diacylglycerol (225), activation of a membrane-bound serine kinase (289), and conformational changes in the insulin receptor (229). However, it remains unclear whether defects in any of these signaling mechanisms contribute to the insulin resistance in NIDDM. Glucose transport After the generation of the second messenger for insulin action, glucose transport is activated. As initially described simultaneously by Cushman et al. (230) and Kono et al. (290), this effect of insulin is brought about by the translocation of a large intracellular pool of glucose transporters (associated with low-den-

338

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

Table 5Classification of glucose transport and hexokinase activity according to their tissue to observe an increase in intracellular distribution and functional regulation free glucose after insulin administration, impaired hexokinase activity has generally been dismissed as a cause of NIDDM HEXOKINASE in humans. However, recent observaCOUPLER ORGAN GLUCOSE TRANSPORTER CLASSIFICATION tions, that suggest an expansion of the GLUT1 HK I Glucose dependent intracellular free glucose pool in reBRAIN HK 1 ERYTHOCYTE GLUT1 Glucose dependent sponse to insulin (154), indicate that the HK 11 GLUT4 Glucose dependent role of hexokinase in the pathogenesis of ADIPOCYTE HK II MUSCLE GLUT4 Glucose dependent NIDDM deserves further examination. HKIVL LIVER GLUT2 Glucose sensor Glucose-transport activity in HKIVB GLUT2 Glucose sensor B-CELI. NIDDM has been studied extensively GUT GLUT3Symporter Sodium dependent and uniformly has been found to be deKIDNEY GLUT3Symporter Sodium dependent creased in both adipocytes (152,251, 256,259,301-304) and muscle (305). Garvey et al. (303) showed that adipocytes from NIDDM subjects demonstrate sity microsomes) to the plasma mem- is present in muscle and adipocytes. It is a marked depletion of intracellular glubrane. During the last 5 yr the glucose located primarily in the plasma mem- cose transporters and a diminished abiltransporter has been the focus of inten- brane where its concentration changes ity of insulin to elicit a normal translosive investigation and its DNA sequence little after the addition of insulin. GLUT1 cation response. Moreover, the intrinsic has been determined (235,291-293). has a low Km, ~ 1 mM, and is well suited activity of the GLUT4 transporter after Five different facilitative glucose trans- for its function, which is to mediate basal insertion into the plasma cell membrane porters with distinctive tissue distribu- glucose uptake; it is found in association was impaired (303). Similar findings tions have been described (235,291; Ta- with hexokinase 1 ( 2 9 5 , 2 9 6 , 2 9 8 ) . have been reported by Sinha et al. (306) ble 5). GLUT4 is the insulin-regulatable GLUT2 predominates in the liver and in adipocytes obtained from morbidly transporter. It is found in the insulin- pancreatic (3-cells, where it is found in obese NIDDM subjects. In a follow-up sensitive tissues (muscle and adipocytes), association with a specific hexokinase IV study Garvey et al. (307) showed that has a Km of ~ 5 mM, which is close to (295,296,299,300). It has a high Km, the depletion of total glucose transporter that of the plasma glucose concentration 1 5 - 2 0 mM, and, as a consequence, number in adipocytes reflects loss from (230-235,291,294), and is associated the free glucose concentration in cells all subcellular fractions, including with a specific hexokinase II (295,296). expressing this transporter rises in direct plasma membranes and both low- and After exposure to insulin, the concentra- proportion to the increase in plasma glu- high-density microsomes. Note that the tion of GLUT4 in the plasma membrane cose concentration. This characteristic profound loss of GLUT4 transporters of adipocytes and muscle increases allows these cells to respond as glucose was closely correlated with a parallel remarkedly and there is a reciprocal de- sensors. Thus, the (3-cell senses the am- duction of GLUT mRNA, leading the aucline in the intracellular low-density mi- bient glucose concentration and adjusts thors to conclude that pretranslational crosomal GLUT4 pool. Recent studies its output of insulin, whereas the liver suppression of GLUT4 -transporter gene that use electron microscopy with immu- reads the plasma glucose level and regu- expression was an important cause of the nogold labeling and a specific monoclo- lates its output of glucose. In summary, insulin resistance in NIDDM. Of particnal antibody have localized the GLUT4 each tissue has a specific glucose trans- ular interest, subjects with IGT manitransporter to the triad (transverse tu- porter and associated hexokinase that al- fested a similar reduction in GLUT4bules) and sacroplasmic tubules in mus- lows it to uniquely carry out its special- transporter number and mRNA levels, cle (297). No GLUT4 transporters were ized function to maintain whole-body indicating that a pretranslational defect found in the plasma membrane or vas- glucose economy. From Table 5, it is in GLUT4 expression may be an early cular endothelium. This anatomical loca- clear that the Km of the glucose trans- defect in the evolution of NIDDM. On tion suggests that glucose transport into porter and hexokinase are similar and as the basis of the results obtained in adimuscle cells may occur across the trans- such function as a tightly coupled trans- pocytes (see above discussion), it was verse tubule membrane. GLUT1 consti- port system. This means that a defect in assumed that a decreased expression of tutes the predominant glucose trans- hexokinase would be indistinguishable the GLUT4 transporter was responsible porter in the insulin-independent from a glucose-transporter defect at the for the insulin resistance of NIDDM. This tissues, brain and erythrocytes, but also whole-body level. Because of the failure assumption is consistent with observa-

DIABETES CARE, VOLUME 15,

NUMBER 3 ,

MARCH

1992

339

Pathogenesis of MDDM

tions in streptozocin and 90% partially pancreatectomized alloxan-induced diabetic rats (308-310). In these insulinopenic models, a pretranslational defect in the GLUT4 transporter was documented in adipocytes. However, both Pedersen et al. (311) and Groop et al. (312) have shown that muscle tissue obtained from lean NIDDM subjects contains normal levels of GLUT4 mRNA (measured by Northern blotting) and GLUT4 protein (determined by immunoblotting). Most recently, the glucose transporter has been directly sequenced in six white and Mexican-American NIDDM subjects and shown to be normal (313). The results described above indicate that the genetic material encoding the major insulin-responsive glucose transporter and the transcription and translation of the gene are not impaired in NIDDM. However, they do not exclude a defect in the intrinsic activity of the glucose transporter. With regard to this, Kahn et al. (314) have shown that after 7 days of streptozocin-induced diabetes in rats, the GLUT4 mRNA is dramatically reduced in both muscle and adipocytes, whereas the GLUT4-transporter protein is reduced in adipocytes, but remains normal in muscle. These observations underscore two important points: 1) the GLUT4 transporter undergoes tissuespecific regulation and 2) the defect in muscle glucose transport cannot be explained by a decrease in GLUT4-transporter number and must be due to a decrease in the intrinsic activity of the glucose transporter itself, due perhaps to glucose toxicity (1,96). Consistent with this hypothesis, correction of hyperglycemia with phlorizin has been shown to restore glucose-transport activity to normal (310). With a novel double-tracer technique, which uses the simultaneous injection of [3H]D-glucose and [14C]0methyl glucose into the brachial artery (315), Bonadonna et al. (316) have shown that the ability of insulin to stimulate forearm muscle glucose uptake is severely impaired in NIDDM. Based on

previous observations in NIDDM individuals (311,312), this defect in glucose transport cannot be explained by a reduction in the total number of glucosetransporter units. However, the findings of Bonadonna et al. (316) do not define 1) whether the intrinsic activity of the GLUT4 transporter is impaired or whether there is a defect in translocation and 2) whether the defect in glucose transport is inherited or secondary to downregulation of the glucose-transport system by chronic hyperglycemia (1,96,304,317; see subsequent discussion on glucose toxicity). With regard to this question, the studies of Gulli et al. (58) and Eriksson et al. (41) are particularly germane. In the offspring of two diabetic parents (58) and in the firstdegree relatives of patients with NIDDM (41) marked insulin resistance is present and primarily accounted for by a defect in glucose storage, i.e., glycogen formation. These observations indicate that, in the very earliest stages of development of NIDDM, glucose transport is not impaired. If glucose transport was diminished, one would expect proportional decreases in both glucose oxidation and glucose storage. Rather, the results suggest a specific defect in the glycogen synthetic pathway or in the signal mechanism coupling the receptor and glycogen synthase. Alternatively, it is possible that there is a differential sensitivity of the glycogen synthetic and glucose oxidative pathways to a decrease in glucose transport, i.e., the former is more sensitive to a reduction in intracellular glucose-6phosphate than the latter.

GLUCOSE UPTAKE 2 (mg/m' mln)

Glucose Slorage

CON

NORMAL OBESE HYPER- HYPOWEIGHT NONINSULINEMIC - DIABETIC - OBESE DIABETIC

Figure 20Insulin-mediated rates (100 /xU/ml euglycemic insulin clamp) of whole-body glucose uptake (total height of bar), glucose oxidation (lower part of each bar), and nonoxidative glucose disposal (upper part of each bar) in control (CON), normal-weight diabetic, obese nondiabetic, and obese diabetic subjects. Obese diabetic patients were further subdivided into hyperinsulinemic and hypoinsulinemic groups based on their plasma insulin response during a 100-g oral glucose tolerance test. *P < 0.001 vs. control subjects. From DeFronzo (1). by the American Diabetes Association.

Glycogen synthesis
After glucose has been transported into the cell it is disposed of by one of two metabolic pathways: glycogen formation and glycolysis (Fig. 20). Most (>90%) glycolysis is accounted for by glucose oxidation (Fig. 20), whereas a small amount (<10%) represents anaerobic glycolysis, i.e., lactate production. In the low range of physiological hyperinsulinemia, these pathways are of equal

quantitative importance (202). With increasing plasma insulin concentrations, glycogen synthesis becomes the predominant pathway of glucose disposal (202). With indirect calorimetry (210), DeFronzo, Felber, and Ferrannini et al. (1,3,135,136,138,167,202,318) directly measured total-body glucose oxidation during euglycemic insulin-clamp studies. Subtraction of the rate of glucose oxidation from the rate of whole-body insulin-mediated glucose disposal (determined from the insulin clamp) yields the rate of glucose storage or nonoxidative glucose disposal. The latter primarily reflects glycogen synthesis (3,168, 211,212) because net glucose conversion to lipid is negligible (182,183,202,210) and little of the glucose taken up by muscle is released as lactate (3,171173,319). In all insulin-resistant states, including obesity and normal-weight diabetes, reduced glucose storage represents the major intracellular abnormality responsible for the defect in insulin action. Glucose oxidation, although re-

340

DIABETES CASE,

VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

duced, is quantitatively less impaired (Fig. 20). Similar findings have been published by Bogardus et al. (12) and Sims et al. (213) in obese and N1DDM subjects. Note that the severity of the defects in whole-body insulin resistance and nonoxidative glucose disposal are of similar magnitude in obese and normalweight diabetic individuals (Fig. 20). Moreover, the combination of obesity and diabetes confers a degree of insulin resistance that is only slightly greater than that observed with either obesity alone or diabetes alone (320). This is true whether the obese diabetic subject maintain a hyperinsulinemic response or have tipped over the top of Starling's curve and are insulinopenic (1). Similar findings have been reported by Reaven et al. (139). In obese diabetic subjects, as was observed in nondiabetic obese and lean diabetic subjects, impaired nonoxidative glucose disposal (glucose storage) represents the major intracellular pathway responsible for the insulin resistance (Fig. 20). The natural history of the evolution of the intracellular defects, i.e., impaired nonoxidative glucose disposal and diminished glucose oxidation, in 173 obese individuals encompassing a wide spectrum of glucose tolerance has been reported by Golay et al. (136). The development of obesity with normal glucose tolerance or 1GT is characterized by the emergence of a severe defect in glucose storage when the subjects are studied under euglycemic-hyperinsulinemic conditions (insulin clamp) (Fig. 21, A). However, the defect could be overcome by hyperglycemia (i.e., OGTT) (Fig. 21, B). The further progression of obesity with IGT to overt diabetes mellitus is associated with the inability of hyperglycemia to compensate for the defect in insulin-mediated glucose storage (Fig. 21). These findings indicate that the ability of insulin to promote nonoxidative glucose disposal (glycogen synthesis) represents a characteristic and early defect in the development of insulin resistance in both obesity and NIDDM. The emergence of frank diabetes mellitus

Insulin Clamp

1779
mg/mlnm2 o >

80 Decrement in Non-Oxidative -100-1 Glucose Disposal (mg/min-m 2 )

21013 mg/minm2

40

60 Time (min)

80

Figure 21 Decrement (compared to normalweight control subjects) in nonoxidative glucose disposal (which primarily reflects glycogen synthesis) during the oral glucose tolerance test (OGTT) (B) and a 100 jiU/ml euglycemic insulin clamp (A) in obese normal glucose-tolerant (OB), obese glucose-intolerant (IGT), and obese diabetic (DIAB) subjects. The absolute value for nonoxidative glucose disposal during the OGTT and euglycemic insulin clamp in normal-weight control subjects is shown to the far right of the zero line. The development of impaired glucose tolerance was heralded by a defect in insulinmediated nonoxidative glucose disposal (insulin clamp). In subjects who progress from impaired glucose tolerance to frank diabetes, the ability of hyperglycemia (OGTT) to compensate for the defect in insulin-mediated glycogen synthesis is lost. *P < 0.01 vs. normal-weight control subjects; **P < 0.05 vs. normal-weight control subjects and P < 0.01 vs. obese diabetic subjects. From Golay et al. (136). by Diabetes Metabolism and Review.

Figure 22Effect of combined hyperinsulinemia (100 mV/ml) and hyperglycemia (200 mg/ dl) on muscle glycogen synthesis in control (open circles) and non-insulin-dependent diabetic (closed circles) subjects. The ability of insulin to stimulate glycogen formation was delayed and markedly decreased in diabetic patients. The time course and magnitude of decrease in leg muscle glycogen formation closely paralleled that of total leg glucose uptake (see Fig. 13). From Shulman et al. (211). by the New England Journal of Medicine.

subjects the rate of nonoxidative glucose disposal (glucose storage) correlated closely with the rate of glycogen synthesis as determined by nuclear magnetic resonant imaging spectroscopy (Fig. 23). In the NIDDM subjects, however, the onset of insulin-stimulated glycogen synthesis was markedly delayed (Fig. 22) and closely mimicked the delay that was observed in leg glucose uptake (Fig. 13). The rate of glycogen synthesis in NIDDM subjects was reduced by 50% and parwith fasting hyperglycemia is associated alleled the decrease in total leg glucose with a major reduction in insulin- uptake both in time and magnitude (Fig. mediated glucose storage that cannot be 13), and accounted for essentially all of offset by hyperglycemia (136,318; Fig. the decrease in whole-body glucose dis21). In a recent study Shulman et al. posal because glucose oxidation was nor(211) used nuclear magnetic resonant mal under the experimental conditions imaging spectroscopy to directly quanti- of combined hyperglycemia/hyperinsutate glycogen synthesis in the gastrone- linemia. The results summarized above mius muscle of the leg. They demon- and depicted in Figs. 20-22 provide strated a decreased incorporation of [XH, convincing evidence that impaired gly13 C] glucose into muscle glycogen under cogen synthesis represents a major defect steady-state conditions of hyperglycemia in normal glucose-tolerant obese suband hyperinsulinemia (Fig. 22). In all jects, in individuals with IGT, and in

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

341

Pathogenesis of NIDDM

(331). In insulin-resistant Pima Indians, dians. Impaired inhibition of cAMPKida et al. (332) demonstrated a close dependent protein kinase represents an? correlation between the decrease in gly- other potentially important defect in cogen synthase activity and reduced fast- insulin action that could explain the obing and insulin-stimulated activity of gly- served dimunition of glycogen synthase cogen synthase phosphatase (332). activity in skeletal muscle (Fig. 24). Most Glycogen synthase exists in an active or recently, Kida et al. (334) examined this 9> dephosphorylated form and an inactive question in insulin-resistant Pima Indi3 50 or phosphorylated form (Fig. 24). The ans and showed that the ability of insulin interconversion of these two forms is reg- to stimulate cAMP-dependent protein 0 10 20 30 40 50 60 ulated by insulin, which activates glyco- kinase in vivo in skeletal muscle is imNonoxidative Glucose Metabolism (jimol/kg body wt/min) gen synthase phosphatase and inhibits paired and correlates well with the defect glycogen synthase kinase (Fig. 24). A de- in glycogen synthase activation. Figure 23Correlation between insulinfect in glycogen synthase phosphatase, as In summary, results from nuclear mediated muscle glycogen synthesis (measured described by Kida et al. (332), would be magnetic resonant spectroscopy, muscle by nuclear magnetic resonance spectroscopy) and expected to cause a greater proportion of biopsies, and indirect calorimetry demnonoxidative glucose disposal (measured by inglycogen synthase to remain in the phos- onstrate that the ability of insulin to augdirect calorimetry) in control (open circles) and non-insulin-dependent diabetic (closed circles) phorylated or inactive form and could ment glycogen synthesis is impaired in patients (r = 0.89, P < 0.001). From Shulman explain the impairment in glycogen syn- the following groups: normal glucoseet al. (211). by the New England Journal of thesis in Pima Indians with NIDDM. tolerant insulin-resistant offspring of two Caesin kinase II, an insulin-regulated en- diabetic parents, normal glucose-tolerant Medicine. zyme that is important in the activation insulin-resistant first-degree relatives of of glycogen synthase phosphatase, has diabetic individuals, obese individuals been suggested to be responsible for the with normal glucose tolerance, lean and patients with overt diabetes mellitus. defect in muscle glycogen synthesis. obese subjects with IGT, and lean and Moreover, the family studies of Gulli et However, a recent study by Maeda et al. obese NIDDM patients. In all of these al. (58) and Eriksson et al. (41) suggest (333) demonstrated that both basal and groups impaired glycogen synthesis has that the earliest detectable metabolic de- insulin-stimulated caesin kinase II activ- been related to a decrease in basal and fect responsible for the insulin resistance ity is normal in muscle biopsies obtained insulin-stimulated glycogen synthase acin individuals who are destined to de- from insulin-resistant diabetic Pima In- tivity. Defects in both glycogen synthase velop NIDDM is an impairment in glycogen synthesis. Glycogen synthase is the key inINSULIN sulin-regulated enzyme that controls muscle glycogen formation (321-323). Cyclic AMP In NIDDM subjects the ability of insulin Synthase Protein to stimulate glycogen synthase in muscle Phosphorylase Kinases Kinase and adipocytes is impaired (12,168Kinase 170,323-330). Both Bogardus et al. , Glycogen (170,330) and Damsbo et al. (327) have Glycogen Glycogen Glycogen Glycogen shown that the decrease in glycogen synPhosphorylase Phosphorylase Synthase Synthase thase activity correlates closely with both (inactive) (active) (active) (inactive) the defects in nonoxidative glucose disUDPG .G-1-P posal (measured by indirect calorimetry) Phosphatase and muscle glycogen formation (measured with muscle biopsy). Similar corG-G-P relations between impaired insulinmediated glucose disposal, diminished Glucose activation of glycogen synthase, and decreased muscle glycogen formation during the insulin clamp have been reported Figure 24Schematic representation of the control of glycogen synthesis. Sites of insulin regulation in poorly controlled NIDDM patients of glycogen synthase are indicated. See text for a more detailed discussion.
30

342

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

phosphatase and cAMP-dependent protein kinase have been implicated as causes of the impairment in glycogen synthase. Regulation of glycogen phosphorylase appears to be normal in N1DDM (335). Now that the human muscle glycogen synthase gene has been sequenced (336), it will be interesting to determine whether its DNA code is normal in NIDDM patients.

Glucose oxidation
Glucose oxidation represents the other major fate of glucose disposal (206), and flux through this pathway also has been shown to be impaired in NIDDM individuals (1,135,136,138,167,318; Fig. 23). Importantly, glucose oxidation can not be normalized even if total glucose flux into the cells is normalized with pharmacological levels of insulin and glucose in NIDDM patients with wellestablished fasting hyperglycemia (324; S. Del Prato, R. C. B., E. Bonora, G. Guilli, A. Solini, M. Shank, R. A. DeF., unpublished observations). A key regulator of glucose oxidation is pyruvate dehydrogenase, an enzyme whose activity is regulated by insulin (323,338). Two previous studies have examined pyruvate dehydrogenase activity in NIDDM patients, and the ability of insulin to stimulate this enzyme complex in both adipocytes (325) and muscle (173) was found to be impaired. However, it should be remembered that both obesity and moderate to severe diabetes mellitus are associated with accelerated rates of FFA turnover/oxidation which, according to the Randle cycle (339,340), would be expected to inhibit pyruvate dehydrogenase and consequently glucose oxidation (see subsequent discussion). In lean NIDDM patients with mild fasting hyperglycemia (135) and insulin-resistance subjects with normal glucose tolerance or IGT (41), insulin-mediated glucose oxidation is normal. This observation suggests that the defect in glucose oxidation in more severely diabetic individuals is an acquired abnormality, secondary to a decrease in circulating insulin levels

(Figs. 2 and 3) or to an increased rate of lipolysis and elevated plasma FFA levels/ intracellular lipid oxidation that occur with the onset of insulin deficiency. Lastly, note that a defect in glycolysis before the pyruvate dehydrogenase step also could account for the impairment in glucose oxidation. With regard to this, phosphofructokinase is an insulin-regulated enzyme (324). Unfortunately, little is known about the activity of phosphofructokinase in NIDDM. Del Prato et al. (unpublished observations) showed that if glucose transport into the cell is normalized with a combination of hyperglycemia and hyperinsulinemia, total glycolytic flux is similar in diabetic and control subjects. However, glucose oxidation remains low, whereas anaerobic glycolysis, i.e., lactate production, is excessively enhanced. Although these findings do not exclude a defect in phosphofructokinase, they at least indicate that any abnormality that is present can be overcome by sufficiently elevating the plasma glucose and insulin levels. In contrast, no level of hyperinsulinemia or hyperglycemia appears to be capable of overcoming the defect in pyruvate dehydrogenase once it has become fully established. In summary, it appears that both binding and postbinding defects in insulin action contribute to the insulin resistance in NIDDM. However, it is now clear that diminished insulin binding occurs primarily in individuals with IGT or with very mild diabetes and results secondarily to a downregulation of the insulin receptor by chronic sustained hyperinsulinemia. Moreover, the magnitude of the decrease in insulin-receptor number is small and cannot explain the marked degree of insulin resistance in NIDDM. In NIDDM patients with fasting glucose levels >140 mg/dl, postbinding defects are primarily responsible for the insulin resistance. Many postbinding defects have been documented, including diminished insulin-receptor tyrosine kinase activity, decreased glucose transport, impaired glycogen synthase activ-

ity, and reduced pyruvate dehydrogenase activity. From the quantitative standpoint, reduced glucose transport and impaired glycogen synthesis represent the major defects in NIDDM. Family studies suggest that a defect in the glycogen synthetic pathway, or possibly in glucose transport, represents the earliest detectable abnormality in NIDDM.
OTHER M E C H A N I S M S OF INSULIN RESISTANCE I n

the

preceding discussion we focused on the primary intracellular disturbances in muscle and liver that are responsible for the insulin resistance in NIDDM. In addition to these abnormalities, there are a number of other interesting hypotheses that have been proposed to explain the defect in insulin action in NIDDM.

lipid oxidation and insulin resistance in NIDDM

Approximately 30 yr ago, Randle et al. (339,340) initially proposed the concept of substrate inhibition, whereby increased FFA oxidation restrains glucose oxidation in muscle by altering the redox potential of the cell and inhibiting several key enzymatic steps within the glycolytic cascade (Fig. 25). Excessive oxidation of FFA leads to the intracellular accumulation of acetyl-CoA, which is a powerful inhibitor of pyruvate dehydrogenase (339-342). In addition, the accelerated rate of FFA oxidation increases the NADH-NAD ratio, leading to an inhibition of the Krebs cycle and the accumulation of citrate. Citrate is a powerful inhibitor of phosphofructokinase (339,340,343), and inhibition of this enzyme causes product inhibition of the early steps involved in glucose metabolism. Eventually, glucose-6-phosphate builds up and inhibits hexokinase, resulting in a decrease in glucose transport into the cell. The sequence of events described above impairs both glucose oxidation (direct inhibition of pyruvate dehydrogenase and Krebs cycle) and glycogen formation (secondary to decreased glucose transport). Enhanced

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

343

Pathogenesis of NIDDM

FATTY ACID-* FATTY ACYL C o A C M . O A C Y L CoA + AcCoA

NAD GLUCOSE r->HK I -GLUCOSE-6-P
GLYCOGEN

GLUCOSE OXIDATION (mg/kg mln)

NON-OXIDATIVE GLUCOSE DISPOSAL (mg/kg mln)

Figure 26Plasma free fatty acid (FFA) concentration and total-body lipid oxidation (meaKC CITRATE sured by indirect calorimetry) in the basal state and during a 100 /ill/ml euglycemic insulin NAD NAD clamp performed with and without intralipid (IL) infusion. IL was infused at 2 rates to maintain (low IL infusion) or increase (high IL infusion) the basal plasma FFA concentration. The highFigure 25Randle cycle. Enhanced free fatty acid oxidation depletes NAD stores, leading to an dose IL infusion maintained the rate of totalinhibition of the Krebs cycle (KC) and a resultant increase in intracellular citrate and acetyl-CoA body lipid oxidation constant at the basal value. concentrations. Accumulation ofacetyl-CoA and citrate leads to the inhibition ofpyruvate dehydroFrom Thiebaud et al. (345). by Metabolism. genase (PDH) and phosphofructokinase (PFK), respectively. Build up of glucose-6-P inhibits hexokinase, leading to an inhibition of glucose transport into the cell. The decrease in glucose transport, in combination with an inhibitory effect of fatty acetyl-CoA on glycogen synthase, results in diminished glycogen formation. Dashed lines represent sites of inhibition of glucose metabolism. See text for a and lipid oxidation, as well as imtions more detailed description. paired suppression of FFA/Hpid oxida-

FRUCTOSE-6-P * PFK 4 FRUCTOSE-1,6-P J, PDH --1 GLYCERALDEHYDE-3-P--- PYRUVATE ACETYL CoA NAD NAD

BASAL

INSULIN

INSULIN + IL (Low)

INSULIN + IL (High)

FFA oxidation also has been shown di- tion (Fig. 26). Infusion of intralipid rectly to inhibit glycogen synthase activ- during the insulin clamp, to maintain or ity in liver by causing a dissociation of its increase the plasma FFA level (Fig. 26), subunits (344). Whether a similar defect inhibits insulin-mediated stimulation of also occurs in muscle is unknown. both glucose oxidation and glucose storTherefore, from a theoretical standpoint, age (Fig. 27). These data demonstrate an elevated rate of FFA oxidation can that the Randle cycle operates in vivo in reproduce all of the major intracellular humans in response to physiological abnormalities (decreased glucose trans- changes in the plasma FFA concentraport, decreased glycogen synthesis, de- tion. creased glucose oxidation) that have In obese nondiabetic and obese been described in NIDDM. In healthy NIDDM subjects, Felber, Golay, Dehumans a physiological elevation in the Fronzo et al. demonstrated that the fastplasma FFA concentration stimulates ing plasma FFA concentration is elevated FFA oxidation, which in turn inhibits and correlates with the increased rate of both glucose oxidation and glucose stor- whole-body lipid oxidation (measured age (glycogen synthesis), providing ex- by indirect calorimetry) (1,135,136,138, perimental validation of the Randle cycle 167,318,353-356). After glucose inges(345-352). In nondiabetic subjects a tion, suppression of plasma FFA levels physiological increment in the plasma and lipid oxidation are impaired, despite insulin level (+100 fiU/ml) causes a 5 0 - the presence of hyperinsulinemia (1,135, 60% decline in plasma FFA concentra- 136,138,167,318,353-356). Increased tion and a parallel decline in lipid oxida- postabsorptive plasma FFA concentra-

tion in response to insulin, have been reported by others in nondiabetic obese and obese diabetic individuals (12,357 359). Felber, DeFronzo et al. used the euglycemic insulin clamp to more pre-

PLASMA FFA

(utnol/l)

LIPID OXIDATION

(mg/kg mln)

of

BASAL

INSULIN

INSULIN + IL (Low)

INSULIN IL (High)

Figure 27Inhibitory effect of intralipid infusion and enhanced lipid oxidation on insulinmediated rates of glucose oxidation and nonoxidative glucose disposal (glycogen synthesis). See legend to Fig. 26 for description of the experimental design. From Thiebaud et al. (345). by Metabolism.

344

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH

1992

DeFronzo

normal fat mass and the presence of basal and meal-stimulated hyperinsu40 linemia/hyperglycemia appear to be sufPLASMA FFA ficient to prevent an increase in lipid (pmol/l) oxidation. However, even in normalweight NIDDM subjects, as insulinope20 p<0.01 nia develops the restraining effect of insulin on lipolysis is lost, plasma FFA levels rise, and even nonobese NIDDM subjects eventually manifest an increase in lipid oxidation (138). p<0.02 The most comprehensive study 800 p<0.01 LIPID examining FFA/lipid metabolism in obesity and NIDDM was published by OXIDATION 2 Groop et al. (138,360). These investiga(mg/m min) tors infused insulin at five different rates 400 (4, 10, 20, 40, and 100 mU m~ 2 min" 1 ) in lean and obese nondiabetic subjects and in lean and obese diabetic subjects. Studies were performed with BASAL INSULIN [14C]palmitate to trace plasma FFA metabolism and with indirect calorimetry to Figure 28 Plasma free fatty acid (FFA) concentration and total-body lipid oxidation in the basal quantitate total-body lipid oxidation. In and insulin-stimulated (100 fiU/ml euglycemic insulin clamp) states in control (solid bars), normalboth obesity and diabetes the ability of weight non-insulin-dependent diabetic (NIDDM) (stippled bars), and obese NIDD (cross-hatched physiological and pharmacological conbars) subjects. Elevated basal plasma FFA concentration and lipid oxidation and impaired suppression centrations of insulin to suppress plasma of plasma FFA and lipid oxidation by insulin were observed only in obese NIDD subjects. Drawn from FFA levels and to inhibit total FFA turnthe data presented in refs. 135, 136, and 167. over, FFA/lipid oxidation, and nonoxidative FFA disposal (reesterification) was impaired. Thus, obesity and diabetes cisely examine the effect of insulin on lationship between lipid oxidation and represent insulin-resistant states, both plasma FFA and lipid oxidation in obese glucose storage during the insulin clamp with regard to FFA/lipid metabolism and and NIDDM subjects (1,135,136,138, (r = - 0 . 3 5 , P < 0.05) was also ob- glucose metabolism. In general, NIDDM 167,318). In obese nondiabetic individ- served but this correlation was much subjects were more resistant to the supuals, in obese subjects with IGT, and in weaker than between lipid and glucose pressive effects of insulin on FFA turnover and oxidation than were obese nonobese diabetic individuals the fasting oxidation. diabetic subjects, and within the NIDDM plasma FFA concentration and basal rate Taken in to to, the results described of lipid oxidation are elevated compared above suggest that elevated rates of lipid group obese diabetic subjects were more with control subjects and fail to suppress oxidation contribute to the defect in glu- resistant than lean diabetic subjects. In normally in response to insulin (Fig. 28). cose oxidation and, to a lesser extent, in both lean and obese NIDDM subjects the In all three obese groups (normal glucose glucose storage in insulin-resistant obese ability of insulin to augment tissue glutolerant, impaired glucose tolerant, and diabetic individuals. However, in nor- cose metabolism (both glucose oxidation diabetic), insulin-mediated rates of mal-weight NIDDM subjects with mild and nonoxidative glucose disposal) and whole-body glucose uptake were re- fasting hyperglycemia, basal rates of lipid to inhibit HGP production was markedly duced due to defects in both glucose oxidation have been reported to be nor- impaired (138,360). The elevated rates oxidation and nonoxidative glucose dis- mal and to suppress normally in re- of FFA/lipid oxidation showed a strong posal (Fig. 20). As predicted by the Ran- sponse to insulin (135). Thus, it is diffi- inverse correlation with the decreased dle cycle, whole-body lipid and glucose cult to invoke increased Randle cycle rates of basal and insulin-stimulated gluoxidation were strongly and inversely activity as a cause of the impairment in cose oxidation and a positive correlation correlated with each other during both glucose oxidation and nonoxidative glu- with the increased rate of basal HGP and the basal state and the insulin clamp cose disposal in normal-weight diabetic impaired suppression of HGP by insulin. (r = -0.80, P < 0.001). An inverse re- patients (Fig. 20). In such individuals a Increased FFA oxidation also
P<

L 0 1

Controls NIDD - Normal Wt NIDD - Obese

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH 1992

345

Paihogenesis ofNIDDM

augments HGP. Addition of fatty acids to cultured hepatocytes and to the isolated perfused liver stimulates gluconeogenesis (361-363), and in healthy subjects the rate of HGP in the basal state and throughout the range of physiological plasma insulin concentrations is closely related to the rate of FFA oxidation (364). Furthermore, FFA infusion in nondiabetic humans under conditions that simulate the diabetic state (346,350,352) and in obese insulinresistant subjects (365) stimulates HGP, most likely secondary to a stimulation of gluconeogenesis. As discussed earlier, in NIDDM subjects the basal plasma FFA level and basal rates of FFA/lipid oxidation are increased and both are strongly correlated with the increase in fasting plasma glucose concentration and the increased basal rate of HGP (1,12,136, 138,360,366). Conversely, it has been shown that reduction by acipimox (an inhibitor of lipolysis) of the elevated plasma FFA concentration and FFA oxidation in NIDDM subjects leads to an enhanced suppression by insulin of the excessive rate of HGP (367). The relationship between plasma FFA concentration, lipid oxidation, and HGP in obesity and NIDDM can be explained as follows: J) increased plasma FFA concentration (by mass action) enhances cellular FFA uptake (368,369); the increase in intracellular FFA leads to a stimulation of lipid oxidation and the accumulation of acetyl-CoA, which in turn inhibits pyruvate dehydrogenase and stimulates pyruvate carboxylase, an important enzyme in the gluconeogenic pathway (2,361,363); the net result is an enhanced flux of three carbon precursors into gluconeogenesis; 2) the augmented rate of lipid oxidation also provides a continued source of energy (ATP) and substrate to drive gluconeogenesis (363,365); 3) the hepatic uptake of circulating gluconeogenic precursors is elevated in obesity (370) and in NIDDM (156); 4) NIDDM subjects have increased plasma glucagon levels (371-374), and hepatic sensitivity to the stimulatory effect

of glucagon on glucose production may be increased; and 5) the inhibitory influence of insulin on gluconeogenesis is much more resistant than its retraining action on glycogenolysis (157). Lastly, it should be emphasized that an increase in plasma FFA concentration and FFA oxidation need not be associated with an increase in HGP, although gluconeogenesis is enhanced (375). This is especially true in nondiabetic subjects in whom, except in very unique experimental conditions (346), elevation of the plasma FFA level rarely leads to an accelerated rate of hepatic glucose output (345,364,376). As recently shown by Clore et al. (375), the FFA-stimulated rise in hepatic gluconeogenesis is precisely offset by an inhibition of glycogenolysis, resulting in no net change in HGP. This hepatic autoregulation in response to FFA infusion is similar to that observed during the infusion of gluconeogenic precursors (377-380). Thus, in dogs and nondiabetic humans intravenous administration of alanine, lactate, and glycerol augment gluconeogenesis but fail to increase total HGP because of a reciprocal decline in gluconeogenesis. The situation in insulinresistant states, such as obesity, appears to be different (365). Here, FFA infusion impairs the suppression of HGP by insulin. The effect of FFA infusion on net HGP and gluconeogenesis has not been examined in NIDDM subjects. In summary, an increase in body fat mass, as occurs in obese individuals, and the presence of insulin resistance and/or insulinopenia, as occurs in NIDDM, are associated with an accelerated rate of lipolysis and an increase in the plasma FFA concentration (Fig. 29). By mass action effect, the elevated plasma FFA level enhances cellular FFA uptake and stimulates lipid oxidation. In muscle the accelerated rate of fat oxidation reduces insulinmediated glucose disposal, primarily by inhibiting glucose oxidation, and to a lesser and more variable extent by impairing glycogen synthesis. At the level of the liver, it

f LIPOLYSIS

I
t FFA MOBILIZATION

1GLUCOSE UTILIZATION

tGLUCONEOGENESIS

HYPERGLYCEMIA

Figure 29 Schematic representation of the effect of increased plasma free fatty acid oxidation on muscle glucose metabolism and hepatic glucose production. See text for a more detailed discussion. From DeFronzo (1). by the American Diabetes Association.

stimulates gluconeogenesis and increases hepatic glucose output.

Skeletal muscle capillary density fiber type, blood flow, and endothelial transport: determinants of insulin action
Muscle and liver represent the primary tissues responsible for the insulin resistance in NIDDM (1). During euglycemichyperinsulinemic conditions, muscle accounts for >90% of the defect in insulin-mediated glucose disposal, whereas diminished muscle glucose uptake and impaired suppression of HGP contribute approximately equally to the disturbance in glucose tolerance after glucose ingestion (see earlier discussion). Much investigation has focused on the basic biochemical mechanisms responsible for the disturbance in insulin-mediated muscle glucose metabolism, and abnormalities in insulin-receptor function, glucose transport, glucose oxidation, and glycogen synthesis have been described. Recent studies by Lillioja et al. (381) have suggested that skeletal muscle capillary density and fiber type may also contribute to the insulin resistance in NIDDM individuals. In both whites and Pima In-

346

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

dians, insulin-mediated glucose disposal was found to correlate with capillary density, and the authors suggested that the diffusion distance from capillary to muscle cells, or some associated biochemical change related to the diffusion distance, contributed to the insulin resistance (381). A positive correlation between the severity of insulin resistance (insulin-clamp technique) and the decrease in the number of type 1 oxidative slow-twitch fibers was also reported (381). Type 1 fibers are very sensitive to insulin, appear to be genetically determined, and are relatively fixed in number (381). Conversely, insulin-mediated glucose uptake was inversely related to the number of type 2B fast-twitch glycolytic fibers that are insulin resistant. These findings suggest an anatomical basis for the insulin resistance in NIDDM. Whether the alteration in muscle fiber composition represents a primary disturbance that contributes to the insulin resistance or represent a histologic change that occurs in response to the insulin resistance remains to be determined, although the authors favor the former interpretation (381). For insulin to reach its target tissues, the hormone must first diffuse across the vascular endothelium, and a defect in insulin diffusion has been suggested as a potential rate-limiting step for insulin action (119). King and Johnson (382) demonstrated that insulin transport across capillary endothelial cells is receptor mediated and rapid. In dogs Yeon et al. (383) have shown a positive correlation between the insulin concentration in thoracic duct lymph and glucose utilization during conditions of euglycemic hyperinsulinemia (insulinclamp technique). On the basis of this observation, they suggested that transcapillary insulin transport is responsible for the normal delay in the onset of insulin action in healthy individuals and may contribute to the excessive delay and diminished response to insulin in NIDDM subjects (383). Unfortunately, most of the thoracic duct lymph is de-

rived from the splanchnic area, and this is confirmed by the authors' findings of higher glucose concentrations (the liver produces glucose) and much lower insulin concentrations (the liver extracts insulin) in thoracic duct lymph compared with plasma (383). This observation raises serious questions about the physiological significance of the positive correlation between thoracic duct lymph and whole-body (which primarily reflects muscle) insulin-mediated glucose disposal. Moreover, after exposure to pharmacological plasma insulin concentrations for 120 min, maximally stimulated glucose uptake is reduced by 50% in NIDDM versus control subjects (138,152). Under these steady-state conditions of supraphysiological hyperinsulinemia, a defect in insulin transport across the vascular endothelial cells cannot be rate limiting for insulin action. Although an interesting hypothesis, it remains to be proved whether insulin diffusion across the capillary endothelium plays any role in the onset of insulin action in nondiabetic humans or contributes to the defect in insulin action of diabetic humans. More studies, including quantitation of muscle lymph insulin concentration and direct measurement of insulin diffusion across the muscle capillary endothelial barrier, are needed to establish whether diffusion of insulin out of the vascular system is rate limiting for the hormone's action. Most recently, a stimulation of muscle blood flow has been proposed as an important determinant of the increase in tissue glucose uptake after insulin administration (384). Tissue (muscle) glucose uptake is the product of the blood flow and the arteriovenous blood glucose concentration difference. In nondiabetic subjects Laasko et al. (384) provided data that about half of the increase in insulin-mediated leg (muscle) glucose utilization in nondiabetic subjects is due to an increase in blood flow, whereas the other half results from an intrinsic stimulation of muscle glucose uptake. In obese nondiabetic subjects leg glucose

uptake was reduced throughout the range of both physiological and pharmacological plasma insulin concentrations. Impaired stimulation of leg blood flow and reduced arteriovenous glucose concentration difference contributed approximately equally to the decrease in total leg glucose uptake in obesity. In a preliminary publication the same group has reported a similar defect in leg blood in NIDDM individuals (385). Although a vasodilatory effect of insulin on forearm blood flow has been demonstrated by several other investigators (386,387), most published reports have failed to demonstrate any effect of insulin on forearm or leg blood flow in nondiabetic or NIDDM subjects ( 3 , 4 , 1 6 4 , 1 7 1 173,179-181,200,201,316,319,388). The explanation for the discordant results concerning the effect of insulin on muscle blood flow has not been established. The studies of Laasko et al. (384) used a thermodilution technique to measure leg blood flow, whereas most studies in which no effect of insulin on blood flow was demonstrated used plethysmography or the dye dilution technique. However, it is unlikely that differences in methodology can explain the varied results, because the increase in forearm blood flow by Anderson et al. (386) was observed using plethysmography. A more likely explanation for the apparently discrepant results concerning the effect of insulin on limb blood flow may relate to the dual effects of the hormone on the vasculature. In the study by Anderson et al. (386) systemic insulin infusion in nondiabetic humans caused a stimulation of muscle sympathetic nerve activity (measured by microneurography) and a 42% increase in circulating norepinephrine levels. The expected consequence of this sympathetic nervous system stimulation should have been vasoconstriction and a decrease in forearm blood flow. To the contrary, forearm blood flow increased due to vasodilation. This may have been due to a direct local vasodilatory action of insulin on muscle blood flow, as suggested by Creager et al.

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

347

Pathogenesis of N1DDM

(387), or to the release of some vasodilatory factor(s). Depending on the balance between the direct/indirect vasodilatory effects of insulin and the hormone's stimulatory effect on the sympathetic nervous system, one could observe either no change in limb blood flow (the usual response), an increase in limb blood Qess commonly observed), or even a decrease in limb blood flow. Amylin and calcitonin gene-related peptide (CGRP) Amylin is a 37 amino acid peptide that is cosecreted with insulin by the pancreatic P-cells (87,89,338,389-391). Although this hormone was initially postulated to be responsible for the impairment in insulin secretion in N1DDM (see earlier discussion), subsequent studies generally have not provided support that amylin is a physiologically relevant modulator of insulin secretion (390,391). More recently, it has been suggested that amylin may, in part, contribute to the impairment in insulin sensitivity in NIDDM. Both in vivo and in vitro studies have demonstrated that amylin causes insulin resistance by interfering with glycogen synthesis (392-396). A similar defect in insulin action has been reported with CGRP (393,395), whose amino acid sequence shares 50% homology with amylin. Unfortunately, all the studies in which amylin or CGRP was infused used pharmacological levels of the hormones, making their physiological interpretation difficult. Moreover, two recent studies failed to demonstrate any effect of amylin on p e r i p h e r a l glucose disposal (397,398). Lastly, Bulter et al. (399) failed to demonstrate any difference in either basal or meal-stimulated plasma amylin concentrations between NIDDM and control subjects. Because amylin is not known to be concentrated in muscle, its role in the insulin resistance of NIDDM seems remote. In contrast, local CGRP concentrations in skeletal muscle may be considerably greater than in plasma because it is secreted by sensory afferent nerve fibers (400), which are

PLASMA " GLUCOSE (mg/IOOml)

GLUCOSE < INFUSION RATE (mg/minfcg)

glucose, and the effect of phlorizin on plasma glucose concentration is entirely secondary to its effect on the kidney to g 5-SS55-5S--3-..S..S..3 O,oDet.cs fi*"~* ! I I ll controls inhibit renal tubular glucose absorption. The pancreatectomized rats had mild fasting hyperglycemia, a markedly abI S^ ; -__---J -5 -A -Jo,ab.t.cs normal meal tolerance test, and a deficient plasma insulin response to the mixed meal compared with the control rats. Phlorizin treatment returned the fasting plasma glucose concentration and i/JLU postmeal glucose profile to normal without any change in plasma insulin concentration or other hormone/substrate levels. After phlorizin was discontinued, glucose intolerance promptly returned and the plasma insulin response reFigure 30Effect of streptozocin-induced dimained unchanged and markedly defiabetes mellitus on insulin-mediated glucose upcient. If hyperglycemia is an important take (insulin-clamp technique) in dogs. Insulideterminant of insulin action in vivo, the nopenia and hyperglycemia were associated with diabetic animals would manifest insulin the development of severe insulin resistance. resistance and phlorizin would be exFrom Bevilacqua et al. (403). by Metabolism. pected to improve the defect in insulinmediated glucose disposal, although insulin secretion remained deficient. This abundant in skeletal muscle (401). This is precisely what was observed (Fig. 31). raises the possibility that part of the in- Phlorizin treatment completely normalsulin resistance in NIDDM may be neu- ized insulin sensitivity in diabetic rats, rogenic in origin and related to an excessive secretion of CGRP.
20 40 60 80 TIME (mln)

Hyperglycemia and glucose toxicity In experimental animal models it is well established that an initial defect in insulin secretion can lead to the development of insulin resistance (93,96,137,317, 402-405). If dogs are made diabetic with streptozocin, fasting hyperglycemia ensues and insulin-mediated glucose uptake becomes markedly impaired (403; Fig. 30). To delineate whether insulin deficiency or some other metabolic derangement that is secondary to the insulinopenia is responsible for the defect in insulin action, Rossetti et al. (317) studied 90% partially pancreatectomized rats before and after treatment with phlorizin (an inhibitor of renal tubular glucose transport) to normalize the plasma glucose profile. Importantly, phlorizin has no effect on any circulating hormone or substrate level other than

CONTROL

PANX

PANX

PANX PHLORIZIN

PHLORIZIN

Figure 31Insulin-mediated glucose uptake during a 100 fiU/ml euglycemic insulin clamp performed in control (solid bar) and 90% pancreatectomized (stippled bar) rats. Phlorizin treatment (cross-hatched bars) restored tissue sensitivity to insulin to normal. Withdrawal of phlorizin (open bar) was associated with the return of severe insulin resistance. From Rossetti et al. (317). by the Journal of Clinical Investigation.

348

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

vided support for an important role of chronic hyperglycemia in the development of insulin resistance. Yki-Jarvinen et al. (421) studied well-controlled p- 3 IDDM patients who were receiving chronic subcutaneous insulin infusion therapy. Subjects participated in two euglycemic insulin-clamp protocols. In the initial study the plasma glucose concenO tration was maintained at the basal level (99 6 mg/dl) for 24 h before perform"5 1ing the insulin clamp; during the 24 h E before the second study the plasma glucose concentration was elevated to 281 16 mg/dl. These authors showed that as little as 24 h of hyperglycemia was CON DIAB DIAB + CON DIAB DIAB + sufficient to induce a 20% decline in the PHLOR PHLOR rate of insulin-mediated glucose disposal (Fig. 33). With a more physiological inFigure 32Glucose transport in the basal (open bars) and insulin-stimulated states (closed bars) crement in plasma glucose concentration in control (CON) and 90% pancreatectomized diabetic (DIAB) rats before and after phlorizin (+40 mg/dl) for 3 days in healthy young (PHLOR) therapy. Restoration of normoglycemia with phlorizin returned glucose-transport activity to subjects, Del Prato et al. (422) demonnormal without any change in plasma insulin concentration. From Kahn et al. (310). by the Journal strated a similar impairment in insulin of Clinical Investigation. action that was entirely accounted for by a defect in nonoxidative glucose disposal. These results suggest that both and discontinuation of phlorizin therapy insulin sensitivity. These observations glucose transport and postglucose transwas associated with a return of the insu- are consistent with a large number of in port abnormalities contribute to the imlin resistance (317). In control rats phlo- vitro and in vivo studies. When muscle pairment in insulin action after exposure rizin had no effect on insulin sensitivity. cells and adipocytes are incubated with a to chronic hyperglycemia. Per Unit Cell Surface Area As discussed in a preceding section, numerous studies have shown that impaired glucose transport plays an important role in the insulin resistance of NIDDM. Both hyperglycemia (downregulation) and insulin (upregulation) are important regulators of the glucosetransport system (96,230-234,310). With 3-O-methylglucose to quantitate glucose transport in isolated adipose cells, Rossetti et al. (310,317) documented that insulin-stimulated glucose transport was diminished in partially pancreatectomized diabetic rats and that the defect in glucose transport could be completely reversed with phlorizin treatment (Fig. 32). The increase in glucose transport was unrelated to changes in GLUT4 mRNA content or GLUT4 transporter number (310). Importantly, the improvement in glucose-transport activity correlated closely with the increase in high medium glucose concentration (10-20 mM), there is a progressive decline in glucose-transport activity, which is associated with a loss of cytochalasin B binding activity in both plasma and lowdensity microsomal membrane fractions. The defect in glucose transport correlates closely with the development of insulin resistance (406-412). Conversely, glucose deprivation in cultured muscle cells, adipocytes, and fibroblasts leads to an upregulation of the glucose-transport system (406,409,413-416). Chronic hyperglycemia also inhibits glucose transport in insulin-independent tissues, including brain (417,418), through a mechanism that involves downregulation of the glucose-transport system (419). Conversely, chronic hypoglycemia, i.e., upregulation, has the opposite effect on brain glucose transport (420). Studies in humans also have proIn summary, many studies that used both in vivo and in vitro techniques have provided impressive evidence that chronic physiological hyperglycemia leads to the development of insulin resistance, both through a downregulation of the glucose-transport system, as well as from an acquired defect in the postglucose transport steps involved in insulin action. Thus, hyperglycemia must be viewed not only as a manifestation of diabetes mellitus but also as a selfperpetuating cause of the diabetic condition. It follows logically from the above discussion that one would expect normalization of the plasma glucose profile, regardless of the means, to lead to an improvement in tissue sensitivity to insulin in diabetic patients. With the insulin-clamp technique, many investigators have examined the effect of diet (weight

Per Cell

DIABETES CARE, VOLUME

15,

NUMBER 3 ,

MARCH

1992

349

Paihogenesis of NIDDM

10

80

60

Glucose Uptake (mg/kg*min) 40

SteadyState Plasma Insulin OiU/ml)

20
Control Post Glucose Infusion Control Post Glucose Infusion

F i g u r e 33Effect of 24 h of sustained hyperglycemia on insulin-mediated glucose metabolism (two left bars) in well-controlled insulin-dependent diabetic subjects. Despite higher steady-state plasma insulin concentrations (two right bars) during the insulin clamp, whole-body glucose uptake was significantly reduced after 24 h of sustained hyperglycemia. Drawn from the data presented in ref. 421 and reproduced from ref. 96.

loss) (100,105,423-425), sulfonylureas (426-431), and insulin therapy (96,98, 99,101,104,423,426,432-441) on insulin action in insulin-dependent diabetic and NIDDM individuals. The results of these studies provide many interesting insights. First, all three therapeutic interventions effectively reduce the fasting plasma glucose concentration, improve glucose tolerance, and enhance insulin secretion (Table 6). However, the improvement in insulin sensitivity was small compared with the complete normalization of insulin-mediated glucose disposal after restoration of normoglycemia with phlorizin in experimental diabetes in the rat (Fig. 31). Second, tight glycemic control with insulin in insulindependent diabetic patients (436-441) uniformly improved insulin sensitivity, whereas a similar degree of glycemic control in NIDDM subjects had a minimal effect to improve insulin action (96, 98,99,101,104,423-426,432-435; Table 7). In NIDDM patients insulin sensitivity was decreased on mean by 70% compared with control subjects, and improved glycemic control with insulin

caused only a small statistically insignificant rise in tissue sensitivity to insulin (Table 7). In contrast, intensive insulin therapy consistently augmented insulin action in IDDM individuals (Table 7). Thus, the failure to observe a significant

amelioration of the insulin resistance in NIDDM cannot be attributed to the insulin therapy itself (204-209). At first glance one might interpret these results in NIDDM as evidence against the glucose toxicity hypothesis. However, it should be remembered that the primary inherited abnormality in NIDDM is insulin resistance. Moreover, the defect in insulin action becomes maximally manifest early in the natural history of NIDDM (see previous discussion and Fig. 4). Thus, when another severely insulin-resistant state, i.e., obesity, is superimposed on NIDDM, the insulin resistance is only slightly greater than observed with diabetes alone or obesity alone (1,136,139; Fig. 20). Therefore, it follows that superimposition of the insulin resistance from glucose toxicity on the insulin resistance that is inherited in NIDDM would not cause much of an additional deterioration in insulin action. Conversely, strict glycemic control and removal of glucose toxicity should not produce a major improvement in insulin sensitivity. In contrast, restoration of normoglycemia in insulin-dependent diabetes in humans and experimental diabetic animal modelswhere the insulin

Table 6Insulin secretion in non-insulin-dependent diabetic patients after tight glycemic control with intensified insulin regimen

PLASMA I[NSUL1N FASTING C3LUCOSE DURATION OF TREATN (MM) BEFORE AFTER CONCENTRATION (PM) BEFORE AFTER PLASMA INSULIN CONCENTRATION (% INCREASE) REF.

MENT (WK)

13 10 7 13 6 7 10 10

2 4 1.5 4 2 2 13 4

12.4 15.1 12.6 12.0 16.1 12.6 10.0 16.2

5.8 7.1 6.2 6.5 7.0 8.1 7.2 6.0

84 96 276 282 168* 222 102 162*

186 300 726 570 810* 240 126 246*

129 56 162 102 382 8 21 52

56 432 101 104 304 423 444 445

Based on ref. 96. * Plasma insulin concentration represents the mean value during an oral glucose tolerance test or mixed meal.

350

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH

1992

DeFronzo

Table 7Insulin sensitivity in non-insulin-dependent

diabetic (NIDDM) and insulin-dependent

diabetic (IDDM) patients after tight glycemic control with intensified insulin regimen

FASTING CSLUCOSE DURATION

INSULIN SENSITIVITY

(MM)
BEFORE AFTER

OF TREATMENT ( W K )

BEFORE

AFTER

REF.

N1DDM

14 10 8 7 8 13 7

5
19
MEAN

3 4 13 1.5 4 4 2 6 4

16.0 15.1 10.1 12.7 14.2 12.0 12.6 12.8 8.8 13.4 8.6 17.1 9.2 4.5 12.3

8.7 7.1 13.3 6.2 5.5 6.5 8.1 6.9 5.6 5.9 6.0 6.8 7.3 4.2 6.4

IDDM

11 8 10 14 7 5 9

26

16-32
6
12 1 10 12

MEAN

25 30 35 15 14 26 36 50 42 30 37 56 65 64 54 59 51 55

43 47 46 15 23 40 40 69 60 42 65 76 82 87 78 67 97 79

98 432 144 101 433 104 423 434 435 436 437 148 439 440 331 441

Primary cellular defect: insulin resistance (Fig. 34) Insulin resistance is a universal finding in resistance is acquired, not inherited sician, he/she will have had diabetes for patients with established NIDDM. In the would be expected to enhance insulin- many years, and defects in both insulin normoglucose-tolerant first-degree relamediated glucose disposal, and this is action and insulin secretion will be tives of NIDDM subjects, in the normosubstantiated by a large body of pub- readily demonstrable (1). At this stage it glucose-tolerant offspring of two diabetic is not possible to reverse the clock to parents, in normal-weight and obese inlished results (Table 7, Fig. 31). define which defect developed first in the dividuals with IGT, and in NIDDM subPATHOGENESIS OF NIDDM Nor- natural history of the disease. Nonethe- jects with mild fasting hyperglycemia mal glucose homeostasis is dependent on less, it is now clear that in any given (110-140 mg/dl), both the basal and a finely balanced dynamic interaction be- diabetic patient whatever defect, i.e., in- stimulated plasma insulin levels are intween tissue sensitivity to insulin and sulin resistance or impaired insulin se- creased and hyperinsulinemia predicts insulin secretion. Even in the presence of cretion, initiates the disturbance in glu- those individuals who are at risk to desevere insulin resistance, a perfectly nor- cose metabolism, it will eventually be velop NIDDM later in life. In each of the mal (3-cell is capable of secreting suffi- followed by the emergence of its coun- above groups tissue sensitivity to insulin, cient amounts of insulin to offset the terpart. Figure 34 depicts and integrates measured with the insulin-clamp techdefect in insulin action. Thus, the evolu- the principal pathogenetic factors re- nique, is diminished. Prospective studies tion of NIDDM requires the presence of sponsible for the emergence of NIDDM have demonstrated that hyperinsulinedefects in both insulin secretion and in- as we currently understand them. The mia and insulin resistance precede the sulin action. Consequently, when an sequence of events outlined in Fig. 34, development of IGT and that IGT repreNIDDM patient first presents to the phy- like the development of NIDDM, is still sents the forerunner of NIDDM (30-

Based on ref. 96. Insulin sensitivity (measured by euglycemic insulin clamp) is expressed as a percentage of the value in healthy control subjects. All values have been converted to milligrams of glucose utilized per meter squared of surface area for uniformity of data presentation. In some instances, it was necessary to extrapolate values from figures presented by the authors.

in a state of dynamic evolution and is meant to be altered as new data become available. Much knowledge has been gained during the period since this topic was last reviewed (1), and the explosion in molecular biology is certain to expand these horizons. As an example, within the last year Bell et al. (442) have linked the gene for diabetes in maturity-onset diabetes in the young families to the adenosine deaminase gene on the long arm of chromosome 20. What relevance, if any, this has to the common day variety of NIDDM remains to be determined. In an accompanying article in this issue (p. 369) Granner et al. have reviewed recent advances in molecular biology as they apply to the study of NIDDM and have highlighted areas of future research. In the following discussion we shall emphasize two points: 1) individuals at risk to develop NIDDM inherit a gene or set of genes that confer insulin resistance and 2) the full-blown diabetic syndrome with overt fasting hyperglycemia develops only if a concomitant insulin secretory defect, be it acquired (as through glucose toxicity) or inherited, is also present.

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH 1992

351

Pathogenesis ofMDDM

Diabetes Gene(s) Glucose Toxictty Impaired Beta Cell Function Glucose-induced Insulin Secretion Hepatic Glucose Production HyperGlycemia Cellular Glucose Uptake
Diabetes Gene(s) ^ ^

Basal Hyper-_ Insulinemia *

PostReceptor Defect

I Glucose TTransport
lnsutin

Obesity Gene(s) T

Resistance 7 )
," ^ " C V

Insulin Binding

Environmental *"""^ ^ ^ . ^

Figure 34Pathogenetic sequence of events leading to the development of glucose intolerance, insulin resistance, and impaired insulin secretion in non-insulin-dependent diabetes. Note that whether the primary defect that initiates the glucose intolerance resides in the (i-cell or in peripheral/ hepatic tissues, development of insulin resistance eventually will ensue or become aggravated, respectively. By the time that overt fasting hyperglycemia (>140 mg/dl) develops, both impaired insulin secretion and severe insulin resistance are present. Broken arrows represent positive feedback loops, that result in self-perpetuation of primary defects. From DeFronzo (1). by the American Diabetes Association.

34,443). Such results provide conclusive evidence that insulin resistance is the inherited defect that initiates the diabetic condition. Recent studies from normal glucose-tolerant first-degree relatives of diabetic individuals and from the offspring of two diabetic parents (groups at high risk to develop NIDDM) indicate that the inherited defect in insulin action results from an abnormality in the glycogen synthetic pathway, or perhaps in glucose transport (28,41,58). Because of the compensatory increase in insulin secretion by the pancreas, initially the insulin resistance by itself is not sufficient to diminish glucose uptake by muscle or to increase HGP. As the insulin resistance progresses, muscle glucose uptake becomes impaired and an excessive rise in postmeal plasma glucose concentration becomes evident. However, the hyperinsulinemia is sufficient to maintain the fasting plasma glucose concentration

and HGP within the normal range, although a longer time is required to restore normoglycemia after each meal. Eventually, however, the insulin resistance becomes sufficiently severe that the compensatory hyperinsulinemia is no longer adequate to maintain the fasting plasma glucose concentration at the basal level. The development of fasting and postprandial hyperglycemia further stimulates (3-cell secretion and the resultant hyperinsulinemia causes a downregulation of both insulin-receptor number and postreceptor events involved in insulin action, thereby exacerbating the insulin resistance. In some individuals the persistent stimulus to the fJ-cell to oversecrete insulin will lead to a deterioration in p-cell function (Figs. 2-4). Recent studies have indicated that chronic hyperglycemia i.e., glucose toxicity, may be responsible for the acquired defect in insulin secretion. It is likely that those in-

dividuals in whom hyperglycemia causes a progressive impairment in insulin secretion have an underlying genetic defect in fJ-cell function that is unmasked by the persistent hyperglycemia. The onset of insulinopenia is associated with the emergence of and/or exacerbation of postreceptor defects in insulin action. Many intracellular events (i.e., glucose transport, glycogen synthase, pyruvate dehydrogenase) involved in glucose metabolism are dependent on the surge of insulin that occurs 3 - 4 times/day in response to nutrient ingestion. If the insulin response becomes deficient, the activity of the glucose-transport system becomes severely impaired and a number of key intracellular enzymatic steps involved in glucose metabolism become depressed. In addition, the restraining effect of insulin on lipolysis is lost. At this juncture plasma FFA levels rise and FFA oxidation increases, further contributing to the defects in intracellular glucose disposal. In addition, there is compelling evidence that hyperglycemia can downregulate the glucose-transport system, as well as many other cellular events involved in insulin action (glucose toxicity). The pathogenetic sequence described above and depicted in Fig. 34 can explain all of the clinical and laboratory features observed in NIDDM. Insofar as the cellular defect is generalized, both hepatic and peripheral tissues (and possibly the (3-cells themselves) will manifest the insulin resistance and the numerous metabolic alterations that are characteristic of the diabetic state could be related to one and the same primary defect. In the preceding discussion we have emphasized the genetic basis of the insulin resistance in NIDDM. However, acquired factors, such as obesity and decreased physical activity, also can contribute to the development of impaired insulin action in NIDDM. Primary insulin secretory defect (Fig. 34) Most evidence documents that hyperinsulinemia precedes the development of

352

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

IGT and N1DDM (1,20,22,27,28,3034,443). Moreover, in populations that are genetically predisposed to develop NIDDM, hyperinsulinemia is a characteristic finding. These observations strongly argue against insulinopenia as the initiating cause of NIDDM. However, there clearly are some diabetic subjects in whom either the first and/or second phases of insulin secretion are diminished. In such individuals the impairment in insulin secretion will lead to an excessive and prolonged rise in plasma glucose concentration. Because the postprandial hyperglycemia presents a persistent stimulus to the fS-cell, the total amount of insulin secreted in response to the meal may actually be increased. Initially, the postprandial hyperinsulinemia may be sufficient to return the fasting plasma glucose concentration to normal, although a longer period of time will be required to restore euglycemia. As the diabetic state worsens (as a consequence of the progressive nature of the f$-cell defect), the plasma insulin response (although still increased in absolute terms) will become inadequate to return the plasma glucose concentration to its basal level and fasting hyperglycemia will ensue. The presence of fasting hyperglycemia will present a continuous stimulus to the pancreas to oversecrete insulin throughout the day and fasting hyperinsulinemia will develop. The elevated fasting insulin concentration will cause a decrease in the number of insulin receptors on insulin-target tissues and downregulate the intracellular events involved in insulin action. Such a pathogenetic sequence of events can explain all of the laboratory features in the patient with IGT (i.e., fasting euglycemia, an impaired but not overtly diabetic glucose tolerance test, fasting hyperinsulinemia, an increased insulin response to glucose) and in the patient with early diabetes mellitus (mild fasting hyperglycemia, a diabetic glucose tolerance test, fasting hyperinsulinemia, normal or increased plasma insulin response to glucose). In such individuals basal HGP remains within the

normal range because of the inhibitory action of fasting hyperglycemia and fasting hyperinsulinemia on glucose release by the liver, whereas the efficiency of tissue glucose uptake would be markedly reduced because the modest rise in fasting insulin concentration is not sufficient to stimulate tissue glucose uptake. As the plasma insulin response becomes progressively more impaired, either as a result of the initial f$-cell defect or pancreatic exhaustion or glucose toxicity, both the early and late phases of insulin secretion become absolutely deficient. Nonetheless, the basal plasma insulin level remains elevated because of the persistent stimulus presented to the P-cell by the fasting hyperglycemia. With the onset of insulineopenia, be it relative or absolute, marked fasting hyperglycemia ensues due to an accelerated rate of HGP and a further reduction in tissue glucose clearance. Another important pathogenetic disturbance that emerges with the development of insulinopenia is a postreceptor defect in insulin action. The eventual clinical picture is representative of a typical NIDDM patient. In summary, the sequence of events, beginning with a defect in insulin secretion and leading to the emergence of insulin resistance, appears to be quite plausible based on available information. What remains unknown is how commonly a primary (3-cell defect with accompanying insulinopenia initiates the sequence of events leading to the development of NIDDM in humans. CONCLUSION In summary, the large and seemingly disparate body of information concerning the pathogenesis of NIDDM can be distilled down to the interaction of the two classic defects, insulin deficiency and insulin resistance. Either defect can be primary and is sufficient to induce hyperglycemia. On clinical and epidemiological grounds, most evidence indicates that insulin resistance most commonly initiates the sequence of events leading to NIDDM. As long as the

pancreas sustains a sufficiently high insulin secretory response to counterbalance the insulin resistance, glucose tolerance remains normal or is only mildly impaired. However, once the (3-cell begins to fail, glucose tolerance deteriorates rapidly and overt diabetes mellitus ensues. Thus, the development of NIDDM requires the presence of two fundamental defects, i.e., insulin resistance and impaired insulin secretion, which disrupt the delicate balance by which insulintarget tissues communicate with the fJ-cells and vice versa.
References 1. DeFronzo RA: Lilly Lecture. The triumvirate: P-cell, muscle, liver. A collusion responsible for NIDDM. Diabetes 37: 667-87, 1988 2. DeFronzo RA, Ferrannini E: Regulation of hepatic glucose metabolism in humans. Diabetes Metab Rev 3:415-59, 1987 3. DeFronzo RA, Jacot E, Jequier E, Maeder E, Wahren J, Felber JP: The effect of insulin on the disposal of intravenous glucose: results from indirect calorimetry. Diabetes 30:1000-1007, 1981 4. DeFronzo RA, Gunnarsson R, Bjorkman O, Olsson M, Wahren J: Effects of insulin on peripheral and splanchnic glucose metabolism in non-insulin dependent diabetes mellitus. J Clin Invest 76:149-55, 1985 5. DeFronzo RA, Ferrannini E, Koivisto V: New concepts in the pathogenesis and treatment of non-insulin-dependent diabetes mellitus. Am J Med 75:52-81, 1983 6. Porte D: Beta cells in type 11 diabetes mellitus. Diabetes 40:166-80, 1991 7. Leahy JL: Natural history of beta cell dysfunction in NIDDM. Diabetes Care 13:992-1010, 1990 8. Bonadonna RC, Groop L, Kraemer N, Ferrannini E, Del Prato S, DeFronzo RA: Obesity and insulin resistance in man: a dose response study. Metabolism 39:452-59, 1990 9. Faber OK, Damsgaard EM: Insulin secretion in type II diabetes. Ada Endo-

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

353

Pathogenesis ofNIDDM

crinol Suppl 262:47-50, 1984 10. Faber OK, Markussen J, Naithani VK, Blix PM, Kuzuya H, Horwitz DL, Rubenstein AH, Rossing N: Kinetics of human connecting peptide in normal and diabetic subjects. J Clin Invest 62: 197-203, 1978 11. DeFronzo RA, Ferrannini E, Simonson DC: Fasting hyperglycemia in noninsulin-dependent diabetes mellitus: contributions of excessive hepatic glucose production and impaired tissue glucose uptake. Metabolism 38:387-95, 1989 12. Bogardus C, Lillioja S, Howard BV, Reaven G, Mott D: Relationships between insulin secretion, insulin action, and fasting plasma glucose concentration in non-diabetic and non-insulindependent subjects. J Clin Invest 74: 1238-46, 1984 13. Reaven GM, Miller R: Study of the relationship between glucose and insulin responses to an oral glucose load in man. Diabetes 17:560-69, 1968 14. Welborn TA, Stenhouse NS, Johnson CJ: Factors determining serum insulin response in a population sample. Diabetologia 5:263-66, 1969 15. Savage PJ, Dippe SE, Bennett PH, Gorden P, Roth J, Rushforth NB, Miller M: Hyperinsulinemia and hypoinsulinemia. Diabetes 24:362-68, 1975 16. Zimmet P, Whitehouse S, Alford F, Chisholm D: The relationship of insulin response to a glucose stimulus over a wide range of glucose tolerance. Diabetologia 15:23-27, 1978 17. Reaven GM, Miller RG: An attempt to define the nature of chemical diabetes using a multidimensional analysis. Diabetologia 16:17-24, 1979 18. Martin FIR, Wyatt GB, Griew AR, Hauranelia M, Higginbotham LI: Diabetes mellitus in urban and rural communities in Papua, New Guinea: studies of prevalence and plasma insulin. Diabetologia 18:369-74, 1980 19. Zimmet P, Whitehouse S: Biomodality of fasting and two-hour glucose tolerance distributions in a Micronesian population. Diabetes 27:793-800, 1978 20. Sicree RA, Zimmet P, King HO, Coven-

try JO: Plasma insulin response among Nauruans: prediction of deterioration in glucose tolerance over 6 yr. Diabetes 36:179-86, 1987 21. Zimmet P, Whitehouse S, KissJ: Ethnic variability in the plasma insulin response to oral glucose in Polynesian and Micronesian subjects. Diabetes 28: 624-28, 1979 22. Saad MF, Knowler WC, Pettitt DJ, Nelson Mott DM, Bennett PH: Sequential changes in serum insulin concentration during development of non-insulindependent diabetes. Lancet 1:1356-59, 1989 23. Kingston ME, Skoog WC: Maintenance of basal insulin secretion in severe noninsulin-dependent diabetes. Diabetes Care 9:232-35, 1986 24. Reaven GM, Hollenbeck C, Jeng C-Y, Wu MS, Chen Y-DI: Measurement of plasma glucose, free fatty acid, lactate, and insulin for 24 h in patients with NIDDM. Diabetes 37:1020-24, 1988 25. Garvey WT, Olefsky JM, Rubenstein AH, Kolterman OG: Day-long integrated urinary C-peptide excretion. Diabetes 37:590-599, 1988 26. Liu G, Coulston A, Chen Y-DI, Reaven GM: Does day-long absolute hypoinsulinemia characterize the patient with non-insulin-dependent diabetes mellitus? Metabolism 32:754-56, 1983 27. Lillioja S, Mott DM, Howard BV, Bennett PH, Yki-Jarvinen H, Freymond D, Nyomba BL, Zurlo F, Swinburn B, Bogardus C: Impaired glucose tolerance as a disorder of insulin action: longitudinal and cross-sectional studies in Pima Indians. N Engl J Med 318:1217-25, 1988 28. Warram JH, Martin BC, Gleason RE, Soeldner JS: Slow glucose removal rate but not insulin secretion predicts development of NIDDM in offspring of two NIDDM parents (Abstract). Diabetes 36 (Suppl. 1):14A, 1987 29. Yudkin JS, Alberti KGMM, McClarity DG, Swai ABM: Impaired glucose tolerance. Is it a risk factor for diabetes or a diagnostic ragbag? Br Med J 301:397401, 1990 30. Saad MF, Knowler WC, Pettitt DJ, Nelson RG, Mott DM, Bennett PH: The

natural history of impaired glucose tolerance in the Pima Indians. N Engl J Med 319:1500-505, 1988 31. Felber JP, Jallut D, Golay A, Munger R, Frascarolo P, Jequier E: Obesity to diabetes: a longitudinal study of glucose metabolism in man (Abstract). Diabetes 20 (Suppl. 1):221A, 1989 32. Lillioja S, Nyomba BL, Saad MF, Ferraro R, Castillo C, Bennett PH, Bogardus C: Exaggerated early insulin release and insulin resistance in a diabetesprone population: a metabolic comparison of Pima Indians and Caucasians. J Clin Endocrinol Metab 73:866-76, 1991 33. Hansen BC, Bodkin NH: Heterogenity of insulin responses: phases leading to type 2 (non-insulin-dependent) diabetes mellitus in the rhesus monkey. Diabetologia 29:713-19, 1986 34. Bodkin NL, Metzger BL, Hansen BC: Hepatic glucose production and insulin sensitivity preceding diabetes in monkeys. Am] Physiol 256:E676-81, 1989 35. Haffner SM, Stern MP, Hazula HP, Pugh JA, Patterson JK: Hyperinsulinemia in a population at high risk for non-insulindependent diabetes mellitus. N Engl J Med 315:220-24, 1986 36. Haffner SM, Stern MP, Mitchell BD, Hazula HP, Patterson JK: Incidence of type II diabetes in Mexican Americans predicted by fasting insulin and glucose levels, obesity, and body-fat distribution. Diabetes 39:283-88, 1990 37. Haffner SM, Stern MP, Hazula HP, Mitchell BD, Patterson JK: Increased insulin concentrations in non-diabetic offspring of diabetic parents. N Engl J Med 319:1297-301, 1988 38. Balkau B, King H, Zimmet P, Raper LR: Factors associated with the development of diabetes in the Micronesian population of Nauru. Am J Epidemiol 122:594-605, 1985 39. Boyko EJ, Keane EM, Marshall JA, Hamman RF: Higher insulin and C-peptide concentrations in Hispanic population at high risk for NIDDM: San Luis Valley Diabetes Study. Diabetes 40: 509-15, 1991 40. Kadowaki T, Miyake Y, Hagura R, Akanuma Y, Kajinuma H, Kuzuya N, Takaku F, Kosaka K: Risk factors for

354

DIABETES CARE, VOLUME 15, NUMBER 3, MARCH 1992

DeFronzo

worsening to diabetes in subjects with impaired glucose tolerance. Diabetologia 26:44-49, 1984 41. Eriksson J, Franssila-Kallunki A, Ekstrand A, Saloranta C, Widen E, Schalin C, Groop L: Early metabolic defects in persons at increased risk for noninsulin-dependent diabetes mellitus. N EnglJ Med 321:337-343, 1989 42. Ho LT, Chang ZY, Wang JT, Li SH, Liu YF, Chen Y-Dl, Reaven GM: Insulin insensitivity in offspring of parents with type 2 diabetes mellitus. Diabetic Med 7:31-34, 1990 43. Charles MA, Fontbonne A, Thibult N, WametJ-M, Rosselin GE, Eschwege E: Risk factors for N1DDM in white population: Paris prospective study. Diabetes 40:796-99, 1991 44. Elbein SC, Maxwell TM, Schumacher MC: Insulin and glucose levels and prevalance of glucose intolerance in pedigrees with multiple diabetic siblings. Diabetes 40:1024-32, 1991 45. DeFronzo RA, Tobin JD, Andres R: Glucose clamp technique: a method for quantifying insulin secretion and resistance. AmJ Physiol 6:E214-23, 1979 46. Curry DL, Bennett LL, Grodsky GM: Dynamics of insulin secretion by the perfused rat pancreas. Endocrinology 83: 572-84, 1968 47. Efendic S, Grill V, Luft R, Wajngot A: Low insulin response: a marker of prediabetes. Adv Exp Med Biol 246:167-74, 1988 48. Cerasi E, Luft R, Efendic S: Decreased sensitivity of the pancreatic beta cells to glucose in pre-diabetic and diabetic subjects. Diabetes 21:224-34, 1972 49. Calles-EscandonJ, Robbins DC: Loss of early phase of insulin release in humans impairs glucose tolerance and blunts thermic effect of glucose. Diabetes 36: 1167-72, 1987 50. Bruce DG, Chisholm DJ, Storlien LH, Kraegen EW: Physiological importance of deficiency in early prandial insulin secretion in N1DDM. Diabetes 37:73644, 1988 51. Steiner KE, Bowles CR, Mouton SM, Williams PE, Cherrington AD: The relative importance of first and second phase insulin secretion in countering

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

the action of glucagon on glucose turnover in the conscious dog. Diabetes 31: 964-72, 1982 Luzi L: Effect of the loss of first phase insulin secretion on glucose production and disposal in man. AmJ Physiol 257: E241-46, 1989 Brunzell JD, Robertson RP, Lerner RL, Hazzard WR, Ensinck JW, Bierman EL, Porte D: Relationships between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endorcinol 46:222-29, 1976 Vague P, Moulin J-P: The defective glucose sensitivity of the B cell in insulin dependent diabetes: improvement after twenty hours of normoglycemia. Metabolism 31:139-42, 1982 Savage PJ, Bennion LJ, Flock EV: Dietinduced improvement of abnormalities in insulin and glucagon secretion and in insulin receptor binding in diabetes mellitus. J Clin Endocrinol Metab 48: 999-1007, 1979 Kosaka K, Kuzuya T, Akanuma Y, Hagura R: Increase in insulin response after treatment of overt maturity onset diabetes mellitus is independent of the mode of treatment. Diabetologia 18:2328, 1980 O'Rahilly S, Turner RC, Matthews DR: Impaired pulsatile secretion of insulin in relatives of patients with non-insulin dependent diabetes. N EnglJ Med 318: 1225-30, 1988 Gulli G, Haffner S, Ferrannini E, DeFronzo RA: What is inherited in N1DDM (Abstract)? Diabetes 39 (Suppl. 1):116A, 1990 Lang DA, Matthews DR, Peto J, Turner RC: Cyclic oscillations of basal plasma glucose and insulin concentrations in human beings. N EnglJ Med 301:102327, 1979 Goodner CJ, Walike BC, Koerker DJ, Brown AC, Chideckel EW, Palmer J, Kalnasy L: Insulin, glucagon, and glucose exhibit synchronous, sustained oscillations in fasting monkeys. Science 195:177-79, 1977 Paolisso G, Sgambato S, Passariello N, Sheen A, D'Onofrio F, Lefebvre PJ: Greater efficacy of pulsatile insulin type

I diabetes critically depends on plasma glucose levels. Diabetes 36:566-70, 1987 62. Weigle DS, Rumbaoa AV, Goodner CJ: Lack of evidence for improvement in long-term glycemic control by pulsatile insulin infusion in streptozocin-induced diabetic baboon. Diabetes 40: 349-57, 1991 63. Polonsky KS, Given BD, Hirsch LJ, Tillil H, Shapiro ET: Abnormal patterns of insulin secretion in non-insulin dependent diabetes mellitus. N Engl J Med 318:1231-39, 1988 64. Temple RC, Carrington CA, Luzio SD, Owens DR, Schneider AE, Sobey WJ, Hales CN: Insulin deficiency in noninsulin-dependent diabetes. Lancet 1:293-95, 1989 65. Horowitz DL, Starr Jl, Mako ME, Blaekard WG, Rubenstein AH: Proinsulin, insulin, and C-peptide concentrations in human protal and peripheral blood. J Clin Invest 55:1278-83, 1975 66. Glauber HS, Revers RR, Henry R, Schmeiser L, Wallace P, Kolterman OG, Cohen RM, Rubenstein AH, Galloway JA, Frank BH, Olefsky AH: In vivo glucose deactivation of pro-insulin action on glucose disposal and hepatic glucose production in normal man. Diabetes 35: 311-17, 1986 67. Ward WK, LaCava EC, Paquette TL, Beard JC, Wallum BJ, Porte D: Disproportionat elevation of immunoreative pro-insulin in type 2 (non-insulindependent) diabetes mellitus and in experimental insulin resistance. Diabetelogia 30:698-702, 1987 68. Deacon CF, Schleser-Mohr, Ballmann M, Williams B, Conlon JM, Creutzfeldt W: Preferential release of pro-insulin relative to insulin in non-insulin dependent diabetes mellitus. Ada Endocrinol 119:549-54, 1988 69. Yoshioka N, Kuzuya T, Matsuda A, Taniguchi M, lwamoto Y: Serum proinsulin levels at fasting and after oral glucose load in patients with type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia 31:355-60, 1988 70. LeahyJL, Halban PA, Weir GC: Relative hypersecretion of proinsulin in rat model of N1DDM. Diabetes 40:985-89, 1991

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

355

Pathogenesis of NIDDM

71. Westermark P, Wilander E: The influence of amyloid deposits on the islet volume in maturity-onset diabetes mellitus. Diabetologia 15:417-21, 1978 72. Gepts W, Lecompte PM: The pancreatic islets in diabetes. Am] Med 70:105-14, 1981 73. Maclean N, Ogilvie RF: Quantitative estimation of the pancreatic islet tissue in diabetic subjects. Diabetes 4:367-76, 1955 74. Saito K, Yaginuma N, Takahashi T: Differential volumetry of alpha, beta, and delta cells in the pancreatic islets of diabetic and nondiabetic subjects. TohokuJ Exp Med 129:273-83, 1979 75. Kloppel G, Lohr M, Habich K, Oberholzer M, Heitz PU: Islet pathology and the pathogenesis of type 1 and type 2 diabetes mellitus revisisted. Survey Synthesis Pathol Res 4:110-25, 1985 76. Clark A, Wells CA, Buley ID, CruickshankJK, Vanhegan RI, Matthews DR, Cooper GJS, Holman RR, Turner RC: Islet amyloid, increased alpha-cells, reduced beta-cells and exocrine fibrosis: quantitative changes in the pancreas in type 2 diabetes. Diabetes Res 9:151-59, 1988 77. Stefan Y, Orci L, Malaisse-Lagae F, Perrelet A, Patel Y, Unger R: Quantitation of endocrine cell content in the pancreas of nondiabetic and diabetic humans. Diabetes 31:694-700, 1982 78. Eisenbarth GS, ConnellyJ, SoeldnerJS: The "natural" history of type I diabetes. Diabetes Metab Rev 3:873-91, 1987 79. Permutt MA, Elbein SC: Insulin gene in diabetes: analysis through RFLP. Diabetes Care 13:364-72, 1990 80. Bell GI, Pictet RL, Rutter WJ, Cordell B, Tischer E, Goodman HM: Sequence of the human insulin gene. Nature (Lond) 284:26-32, 1980 81. Steiner DF, Tager HS, Chan SJ, Nanjo T, Sanke T, Rubenstein AH: Lessons learned from molecular biology of insulin-gene mutations. Diabetes Care 13: 600-607, 1990 82. Raben N, Barbetti F, Cama A, Lesniak MA, Lillioja S, Zimmet P, Serjeantson SW, Taylor SI, Roth J: Normal coding sequence of insulin gene in Pima Indians and Nauruans, two groups with

highest prevalence of type 11. Diabetes 40:118-22, 1991 83. Lukinius A, Willander E, Westermark GT, Engstrom U, Westermark P: Colocalization of islet amyloid polypeptide and insulin in the beta cell secretory granules of the human pancreatic islets. Diabetologia 32:240-44, 1989 84. Ogawa A, Harris V, McCorkle SK, Under RH, Luskey KL: Amylin secretion from the rat pancreas and its selective loss after streptozotocin treatment. J Clin Invest 85:973-76, 1990 85. Johnson KH, O'Brien TD, Betysholtz C, Westermark P: Islet amyloid, isletamyloid polypeptide, and diabetes mellitus. N Engl J Med 321:513-18, 1989 86. Clark A: Islet amyloid and type 2 diabetes. Diabetic Med 6:561-67, 1989 87. Nishi M, Sanke T, Nagamatsu S, Bell GI, Steiner DF: Islet amyloid polypeptide: a new beta cell secretory product related to islet amyloid deposits. J Biol Chem 265:4173-76, 1990 88. Westermark P, Wernstedt C, Wilander E, Hayden DW, O'Brien TD, Johnson KH: Amyloid fibrils in human insulinoma and islets of Langerhans of the diabetic cat are derived from a neuropeptide-like protein also present in normal islet cells. Proc Natl Acad Sci USA 84:3881-85, 1985 89. Cooper GJS, Willis AC, Clark A, Turner RC, Sim RB, Rein KBM: Purification and characterization of a peptide from amyloid-rich pancreases of type 2 diabetic

93. Bretherton-Watt D, Gilbey SG, Ghatei MA, Beacham J, Bloom SR: Failure to establish islet amyloid polypeptide (amylin) as a circulating beta cell inhibiting hormone in man. Diabetologia 33: 115-17, 1990 94. Yamaguchi A, Chiba T, Morishita T, Nakamura A, Inui T, Yamatani T, Kadowaki S, Chihara K, Fukase M, Fujita T: Calcitonin gene-related peptide and induction of hyperglycemia in conscious rats in vivo. Diabetes 39:168-74, 1990

95. Dunning BE, Taborsky GJ: Galaninsympathetic neurotransmitter in endocrine pancreas. Diabetes 37:1157-62, 1988 96. Rossetti L, Giaccari A, DeFronzo RA: Glucose toxicity. Diabetes Care 13:610 30, 1990 97. Unger RH, Grundy S: Hyperglycaemia as an inducer as well as a consequence of impaired islet cell function and insulin resistance: implications for the management of diabetes. Diabetologia 28: 119-21, 1985 98. GarveyWT, OlefskyJM, Griffin J, Hamman RF, Kolterman OG: The effect of insulin treatment on insulin action in type II diabetes mellitus. Diabetes 34: 222-34, 1985 99. Kosaka K, Kuzuya T, Akanuma Y, Hagura R: Increase in insulin response after treatment of overt maturity onset diabetes mellitus is independent of the mode of treatment. Diabetologia 18:2328, 1980 100. Savage PJ, Bennion LJ, Flock EV: Dietpatients. Proc Natl Acad Sci USA 84: induced improvement of abnormalities 8628-32, 1987 in insulin and glucagon secretion and in 90. Ohsawa H, Kanatsuka A, Yamaguchi T, insulin receptor binding in diabetes Makino H, Yoshida S: Islet amyloid mellitus. J Clin Endocrin Metab 48:999polypeptide inhibits glucose-stimulated 1007, 1979 insulin secretion from isolated rat pan101. Hidaka H, Nagulesparan M, Klimes I, creatic islets. Biochem Biophys Res ComClark R, Sasaki H, Aronoff SL, Vasquez mun 160:961-67, 1989 B, Rubenstein AH, Unger RH: Improve91. Howard CF: Longitudinal studies on ment of insulin secretion but not insuthe development of diabetes in individlin resistance after short term control of ual macaca nigra. Diabetologia 29:301plasma glucose in obese type II diabet306, 1986 ics. J Clin Endocrinol Metab 54:217-22, 92. Ghatei MA, Datta HK, Zaidi M, Breth1982 erton-Watt D, Wimalawansa SJ, Maclntyre I, Bloom SR: Amylin and amylin- 102. Hadden DAR, Montgomery DA, Skelly RJ, Trimble ER, Weaver JA, Wilson EA, amide lack an acute effect on blood Buchanan KD: Maturity onset diabetes glucose and insulin. ] Endocrinol 124: mellitus. Br Med J 3:276-78, 1975 R9-11, 1990

356

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

103. Stanik S, Marcus R: Insulin secretion 113. LeahyJL, Weir GC: Evolution of abnormal insulin secretory responses during improves following dietary control of 48-h in vivo hyeprglycemia. Diabetes plasma glucose in severely hyperglyce37:217-22, 1988 mic obese patients. Metabolism 29:34650, 1980 114. Giroix MH, Portha B, Kergoat M, Bailbe D, Picon L: Glucose insensitivity and 104. Andrews WJ, Vasquez B, Nagulesparan amino-acid hypersensitivity of insulin M, Klimes I, FoleyJ, Unger R, Reaven release in rats with non-insulin-depenGM: Insulin therapy in obese, nondent diabetes. Diabetes 32:445-51, insulin-dependent diabetes induces im1983 provements in insulin action and secretion that are maintained for two weeks 115. Kergoat M, Bailbe D, Portha B: Insulin after insulin withdrawal. Diabetes 33: treatment improves glucose-induced 634-42, 1984 insulin release in rats with NIDDM induced by streptozocin. Diabetes 36: 105. Henry RR, Wallace P, OlefskyJM: Ef971-77, 1977 fects of weight loss on mechanisms of hyperglycemia in obese non-insulin- 116. Grill V, Westberg M, Ostenson CG: dependent diabetes mellims. Diabetes Beta cell insensitivity in a rat model of 35:990-98, 1986 non-insulin-dependent diabetes. J Clin 106. Faber OK, Damsbo P, Schalg T, SvendInvest 80:664-69, 1987 sen RA, Regeur L, DeFronzo RA: Beta 117. Imamura T, Koffler M, Helderman JH, cell function during insulin treatment. Prince D, Thirlby R, Inman L, Unger In Recent Advances in Obesity and DiaRH: Severe diabetes induced in subtobetes Research. Melchionda N, Kuzuya tally depancreatectomized dogs by susH, Shade D, Eds. New York, Raven, tained hyperglycemia. Diabetes 37: 1983, p. 77-82 600-609, 1988 107. Rossetti L, Shulman GI, Zawalich W, 118. Reaven GM, Hollenbeck CB, Chen DeFronzo RA: Effect of chronic hyperY-Dl: Relationship between glucose tolglycemia on in vivo insulin secretion in erance, insulin secretion, and insulin partially pancreatectomized rats. J Clin action in non-obese individuals with Invest 80:1037-44, 1987 varying degrees of glucose tolerance. 108. LeahyJL, Bonner-Weir S, Weir GC: AbDiabetologia 32:52-55, 1989 normal glucose regulation of insulin se- 119. Bergman RN: Lilly Lecture. Toward cretion in models of reduced P-cell physiological understanding of glucose mass. Diabetes 33:667-73, 1984 toleranceminimal-model approach. 109. Bonner-Weir S, Trent DF, Weir GC: Diabetes 38:1512-26, 1989 Partial pancreatectomy in the rat and 120. Himsworth HP, Kerr RB: Insulinsubsequent defect in glucose-induced sensitive and insulin-insensitive types insulin release. J Clin Invest 71:1544of diabetes mellitus. Clin Sci 4:120-52, 53, 1983 1937 110. Leahy JL, Bonner-Weir S, Weir GC: 121. Alford FP, Martin FI, Pearson MF: The Minimal chronic hyperglycemia is a significance of interpretation of mildly critical determinant of impaired insulin abnormal oral glucose tolerance. Diabesecretion after an incomplete pancretologia 7:173-80, 1971 atectomy. J Clin Invest 81:1407-14, 122. Beck-Nielsen H, Pedersen O, Sorensen 1988 NS: Effects of dietary changes on cellu111. LeahyJL, Cooper HE, Weir GC: Imlar insulin binding and in vivo insulin paired insulin secretion associated with sensitivity. Metabolism 29:482-87, near normoglycemia. Diabetes 36:4591980 64, 1987 123. Shen SW, Reaven GM, Farquhar JW: 112. LeahyJL, Cooper HE, Deal DA, Weir Comparison of impedance to insulinGC: Chronic hyperglycemia is associmediated glucose uptake in normal ated with impaired glucose influence on subjects with latent diabetes. J Clin Ininsulin secretion. J Clin Invest 77:908vest 49:2151-60, 1970 15, 1986 124. Ginsberg H, Kimmerling G, OlefskyJM,

Reaven GM: Demonstration of insulin resistance in untreated adult-onset diabetic subjects with fasting hyperglycemia. J Clin Invest 55:454-61, 1975 125. Kimmerling G, Javorski C, OlefskyJM, Reaven GM: Locating the site(s) of insulin resistance in patients with nonketotic diabetes mellitus. Diabetes 25: 673-78, 1976 126. Reaven GM: Banting Lecture. Role of insulin resistance in human disease. Diabetes 37:1595-607, 1988 127. Harano Y, Ongaku S, Hidaka H, Haneda K, Kikkawa R, Shigeta Y, Abe H: Glucose, insulin, and somatostatin infusion for the determination of insulin sensitivity. J Clin Endocrinol Metab

45:1124-27, 1977 128. Ratzman KP, Besch W, Witt S, Schulz B: Evaluation of insulin resistance during inhibition of endogenous insulin and glucagon secretion by somatostatin in non-obese subjects with impaired glucose tolerance. Diabetologia 21:19297, 1981 129. Butterfield WJH, Whichelow MJ: Peripheral glucose metabolism in control subjects and diabetic patients during glucose, glucose-insulin, and insulin sensitivity tests. Diabetologia 1:43-53, 1965 130. Zierler KL, Rabinowitz D: Roles of insulin and growth hormone, based on studies of forearm metabolism in man. Medicine 42:385-402, 1963 131. Wangot A, Roovete A, Vranic M, Luft R, Efendic S: Insulin resistance and decreased insulin response to glucose in lean type II diabetes. Proc Natl head Sci USA 79:4422-27, 1982 131a.Bowen HF, Moorhouse JA: Glucose turnover and disposal in maturity-onset diabetes. J Clin Invest 52:3033-45, 1973 132. DeFronzo RA, Deibert D, Hendler R, Felig P: Insulin sensitivity and insulin binding to monocytes in maturityonset diabetes. J Clin Invest 63:939-46, 1982 133. DeFronzo RA, Simonson D, Ferrannini E: Hepatic and peripheral insulin resistance: a common feature in non-insulin-dependent and insulin dependent diabetes. Diabetologia 23:313-19, 1982

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

357

Pathogenesis ofMDDM

134. Simonson D, Ferrannini E, Bevilacqua 143. Kurtz F: Mechanism of metformin acS, Smith D, Barrett E, Carlson R, Detion in non-insulin-dependent diabetes Fronzo RA: Mechanism of improvemellitus. Diabetes 36:632-40, 1987 ment in glucose metabolism following 144. Firth RG, Bell PM, Rizza RA: Effects of chronic glyburide therapy. Diabetes 33: tolazamide and exogenous insulin on 838-45, 1984 insulin action in patients with noninsulin-dependent diabetes mellitus. N 135. Golay A, DeFronzo RA, Ferrannini E, EnglJ Med 314:1280-86, 1986 Simonson DC, Thorin D, Acheson K, Thiebaud D, Curchod B, Jequier E, Fel- 145. Best JD, Judzewitsch RG, Pfeiffer MA, ber JP: Oxidative and non-oxidative Beard JC, Halter JB, Porte D: The effect glucose metabolism in non- obese type of chronic sulfonyl urea therapy on he2 (non-insulin-dependent) diabetic papatic glucose production in non-insutients. Diabetologia 31:585-91, 1988 lin-dependent diabetes mellitus. Diabetes 31:333-38, 1982 136. Golay A, FelberJP, Jequier E, DeFronzo RA, Ferrannini E: Metabolic basis of obesity and noninsulin-dependent diabetes mellitus. Diabetes Metab Rev 4:727-47, 1988 DallAglio E, Chang H, Hollenbeck CB, Mondon CE, Sims C, Reaven GM: In vivo and in vitro resistance to maximal insulin-stimulated glucose disposal in insulin deficiency. Am J Physiol 249: E312-16, 1985 Groop LC, Bonadonna RC, Del Prato S, Ratheiser K, Zych K, Ferrannini E, DeFronzo RA: Glucose and free fatty acid metabolism in non-insulin dependent diabetes mellitus: evidence for multiple sites of insulin resistance. J Clin Invest 84:205-13, 1989 Hollenbeck CB, Chen Y-DI, Reaven GM: A comparison of the relative effects of obesity and non-insulin dependent diabetes mellitus on in vivo insulin stimulated glucose utilization. Diabetes 33:622-24, 1984 Firth R, Bell P, Rizza R: Insulin action in non-insulin-dependent diabetes mellitus: the relationship between hepatic and extrahepatic insulin resistance and obesity. Metabolism 36:1091-95, 1987 Campbell PJ, Mandarino LJ, GerichJE: Quantification of the relative impairment in actions of insulin on hepatic glucose production and peripheral glucose uptake in non-insulin-dependent diabetes mellitus. Metabolism 37:1521, 1988 Grill V: A comparison of brain glucose metabolism in diabetes as measured by position emission tomography or by arteriovenous techniques. Ann Med 22: 171-75, 1990 146.

glucose, [3-3H] glucose, and [3-2H] glucose: differences in the apparent pattern of insulin resistance depending upon the isotope used. J Clin Invest 78: 1479-86, 1986 154. Ferrannini E, Smith JD, Cobelli C, Toffalo G, Pilo A, DeFronzo RA: The effect of insulin on the distribution and disposition of glucose in man: insights from physiologic modeling. J Clin Invest 76:357-64, 1985

137.

147.

138.

148.

149.

139.

150.

140.

151.

141.

152.

142.

153.

155. Koivisto V, Yki-Jarvinen H, Kolaczynski J, DeFronzo RA: No evidence for isotope discrimination of tritiated glucose tracers in measurements of glucose Firth RG, Bell PM, Marsh HM, Hansen turnover rates in man. Diabetologia 33: I, Rizza RA: Postprandial hyperglyce168-73, 1990 mia in patients with non-insulin-dependent diabetes mellitus. J Clin Invest 156. Waldhausl W, Bratusch-Marrain P, 77:1525-32, 1986 Gasic S, Kom A, Nowatony P: Insulin Jackson RA, Hawa MI, Jaspan JB, Sim production rate, hepatic insulin retenBM, DiSilvio L, Featherbe D, Kurtz AB: tion, and splanchnic carbohydrate meMechanisms of metformin action in tabolism after oral glucose ingestion in non-insulin-dependent diabetes mellihyperinsulinemic type II (non-insulintus. Diabetes 36:632-40, 1987 dependent) diabetes mellitus. Diabetologia 23:6-15, 1982 GarveyWT, OlefskyJM, Griffin J, Ham157. Consoli A, Nurjahn N, Capani F, German RF, Kolterman OG: The effect of ichj: Predominant role of gluconeogeninsulin treatment on insulin action in esis in increased hepatic glucose protype II diabetes mellitus. Diabetes 34: duction in N1DDM. Diabetes 38:550222-34, 1985 56, 1989 DeFronzo RA, Ferrannini E, Hendler R, 158. Unger RH, Aguilar-Parada E, Mueller Felig P, Wahren J: Regulation of WA, Eisentraut AM: Studies of pancresplanchnic and peripheral glucose upatic alpha-cell function in normal and take by insulin and hyperglycemia. Didiabetic subjects. J Clin Invest 49:837abetes 32:35-45, 1983 45, 1970 DeFronzo RA, Ferrannini E: Influence of plasma glucose and insulin concen- 159. Baron AD, Schaeffer L, Shragg P, Kolterman OG: Role of hyperglucagonemia tration on plasma glucose clearance in in maintenance of increased rates of heman. Diabetes 31:583-688, 1982 patic glucose output in type II diabetes. Cherrington AD, Stevenson RW, Diabetes 36:274-83, 1987 Steiner KE, Davis MA, Myers SR, Adkins BA, Abumrad NN, Williams PE: 160. Orskov C, JeppesenJ, Masbad S, Hoist J: Proglucagon products in plasma of Insulin, glucagon, and glucose as regunon-insulin dependent diabetics and lators of hepatic glucose uptake and nondiabetic controls in the fasting state production in vivo. Diabetes Metab Rev and after oral glucose and intravenous 3:307-32, 1987 arginine. J Clin Invest 87:415-23, 1991 Kolterman OG, Gray RS, Griffin J, Burstein P, InselJ, Scarlett JA, Olefsky 161. Felig P, Wahren J, Hendler R: Influence of maturity-onset diabetes on splanchJM: Receptor and postreceptor defects nic balance after oral glucose ingestion. contribute to the insulin resistance in Diabetes 27:121-26, 1978 non-insulin-dependent diabetes melli162. Adkins BA, Myers SR, Hendrick GK, tus. J Clin Invest 68:957-69, 1981 Stevenson RW, Williams PE, CherBell PM, Firth RG, Rizza RA: Assessrington AD: Importance of the route of ment of insulin action in insulin depenglucose delivery to hepatic glucose updent diabetes mellitus using [6-14C]

358

DIABETES CARE, VOIUME 15,

NTMBFR 3, MARCH

1992

DeFronzo

163.

164.

165.

166.

167.

168.

169.

170.

171.

take in the conscious dog. J Clin Invest 79:557-65, 1987 Ferrannini E, Simonson DC, Katz LD, Reichard G, Bevilacqua S, Barrett EJ, Olsson M, DeFronzo RA: The disposal of an oral glucose load in patients with non-insulin dependent diabetes. Metabolism 37:79-85, 1988 Gerich JE, Mitrakou A, Kelly D, Mandarino L, Nurjhan N, Reilly J, Jenssen T, Veneman T, Consoli A: Contribution of impaired muscle glucose clearance to reduced postabsorptive systemic glucose clearance in N1DDM. Diabetes 39: 211-15, 1990 Revers R, Fink R, Griffin J, OlefskyJ, Kolterman O: Influence of hyperglycemia in insulin's in vivo effects in type 11 diabetes. J Clin Invest 73:664-72, 1984 Capaldo B, Santoro D, Riccardi G, Perrotti N, Sacca L: Direct evidence for a stimulatory effect of hyperglycemia per se on peripheral glucose disposal in type 11 diabetes. J Clin Invest 77:128590, 1986 FelberJP, Ferrannini E, Golay A, Meyer HU, Thiebaud D, Curchod B, Maeder E, Jequier E, DeFronzo RA: Role of lipid oxidation in the pathogenesis of the insulin resistance of obesity and type 11 diabetes. Diabetes 36:1341-50, 1987 Young A, Bogardus C, Wolfe-Lopez D, Mott D: Muscle glycogen synthesis and disposition of infused glucose in humans with reduced rates of insulinmediated carbohydrate storage. Diabetes 37:303-307, 1988 Kida Y, Esposito-DelPuente A, Bogardus C, Mott D: Insulin resistance is associated with reduced fasting insulinstimulated glycogen synthase phosphatase activity in human skeletal muscle. J Clin Invest 85:476-81, 1990 Bogardus C, Lillioja A, Stone K, Mott D: Correlation of muscle glycogen synthase activity and in vivo insulin action in man.J Clin Invest 73:1185-90, 1984 Capaldo B, Napoli R, Di Bonito P, Albano G, Sacca L: Glucose and gluconeogenic substrate exchange by the forearm skeletal muscle in hyperglycemic and insulin-treated type 11 diabetic patients.J Clin Endocrinol Metab 71:122023, 1990

studies of forearm metabolism in man. 172. Mitrakou A, Kelley D, Veneman T, Medicine 42:385-402, 1963 Jensen T, Pangburn T, Reilly J, Gerich J: Contribution of abnormal muscle and 182. Bjorntorp P, Sjostrom L: Carbohydrate liver glucose metabolism to postpranstorage in man: speculations and some dial hyperglycemia in N1DDM. Diabetes quantitative considerations. Metabolism 39:1381-90, 1990 27 (Suppl. 2):1853-65, 1978 173. Kelley DE, Mandarino LJ: Hyperglyce- 183. Bjorntorp P, Berchtold P, Holm J: The glucose uptake of human adipose tissue mia normalizes insulin-stimulated skelin obesity. Eur J Clin Invest 1:480-85, etal muscle glucose oxidation and storage in noninsulin-dependent diabetes 1971 mellitus. J Clin Invest 86:1999-2007, 184. Marin P, Rebuffe-Serive S, Smith U, 1990 Bjorntorp P: Glucose uptake in human adipose tissue. Metabolism 36:1154174. Consoli A, Nurjhan N, Reilly JJ, Bier 60, 1987 DM, Gerich JE: Mechanism of increased gluconeogenesis in noninsulin-depen- 185. Reinmuth OM, Scheinberg P, Bourne B: dent diabetes mellitus: role of alterTotal cerebral blood flow and metaboations in systemic, hepatic, and muscle lism. Arch Neiirol 12:49-66, 1965 lactate and alanine metabolism. J Clin 186. Huang SC, Phelps ME, Hoffman EJ, SidInvest 86:2038-45, 1990 eris K, Selin CJ, Kuhl DE: Non-invasive determination of local cerebral meta175. Davis MA, Williams PE, Cherrington bolic rate of glucose in man. Am] PhysAD: The effect of glucagon on lactate metabolism in conscious dog. Am J iol 238:E69-82, 1980 Physio! 248:E463-70, 1985 187. Ferrannini E, Reichard GA, Bjorkman 176. Jungerman K: Metabolic zonation of O, Wahren J, Pilo A, Olsson M, Deliver parenchyma: significance for regFronzo RA: The disposal of an oral gluulation of glycogen metabolism, glucocose load in normal subjects: a quantineogenesis, and glycolysis. Diabetes tative study. Diabetes 34:580-88, 1985 Metab Rev 3:269-93, 1987 188. Kelley D, Mitrakou A, Marsh H, Sch177. Kalant N, Leibovici D, Fukushima J, wenk F, Benn J, Sonnenberg G, ArOzaki S: Insulin responsiveness of sucangeli M, Aok T, Sorensen J, Berger M, perficial forearm tissues in type II (nonSonksen P, Gerich J: Skeletal muscle insulin-dependent) diabetes. Diabetologlycolysis, oxidation, and storage of an gia 22:239-44, 1982 oral glucose load. J Clin Invest 81:1563 178. Campbell P, Mandarino L, Gerich J: 71, 1988 Quantification of the relative impair- 189. Jackson RA, Roshania RD, Hawa MI, ment in actions of insulin on hepatic Sim BM, DiSilvio L: Impact of glucose glucose production and peripheral gluingestion on hepatic and peripheral cose intake in non-insulin-dependent glucose metabolism in man: an analysis diabetes mellitus. Metabolism 37:15based on simultaneous use of the fore22, 1988 arm and double isotope techniques. J 179. Jackson RA, Perry G, Rogers J, Advoni Clin Endocrinol Metab 63:541-49, U, Pilkington TRE: Relationship be1982 tween the basal glucose concentration, 190. Katz LD, Glickman MG, Rapoport S, glucose tolerance, and forearm glucose Ferrannini E, DeFronzo RA: Splanchnic uptake in maturity onset diabetes. Diaand peripheral disposal of oral glucose betes 22:751-61, 1973 in man. Diabetes 32:675-79, 1983 180. Butterfield WJH, Whichelow MJ: Pe- 191. Clarys JP, Martin AF, Frinkwater DT: ripheral glucose metabolism in control Gross tissue weights in the human body subjects and diabetic patients during by cadaver dissection. Hum Biol 56: glucose, glucose-insulin and insulin 459-73, 1984 sensitivity tests. Diabetologia 1:43-53, 192. Cohn SH, Vartsky D, Yasumura S, Sa1965 witsky A, Zani I, Vaswani A, Ellis J: 181. Zierler KL, Rabinowitz D: Roles of inCompartmental body composition sulin and growth hormone, based on based on total-body nitrogen, potas-

DLABETES CARE, VOLUME

15,

NUMBER 3 ,

MARCH

1992

359

Pathogenesis o/NIDDM

193.

194.

195.

196.

197.

198.

199.

200.

201.

202.

203.

sium, and calcium. Am J Physiol 239: E524-28, 1980 Shimazu T: Neuronal regulation of hepatic glucose metabolism in mammals. Diabetes Metab Rev 3:185-206, 1987 McMahon V, Marsh HM, Rizza RA: Effects of basal insulin supplementation on disposition of mixed meal in obese patients with N1DDM. Diabetes 38: 291-303, 1989 Chen Y-Dl, Jeng CY, Hollenbeck CB, Wu MS, Reaven GM: Relationship between plasma glucose and insulin concentration, glucose production, and glucose disposal in normal subjects and patients with non-insulin-dependent diabetes. J C/in Invest 82:21-25, 1988 Gerich JE: Is muscle the major site of insulin resistance in type II diabetes? Diabetologia 34:607-10, 1991 Shank M, Solini A, Barzilai N: Peripheral and hepatic glucose metabolism following mixed meal ingestion in NIDDM (Abstract). Diabetes 39 (Suppl. 1):118A, 1990 National Diabetes Data Group: Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 28:1039-57, 1979 DeFronzo RA, Their SO: Inherited disorders of renal tabular function. In The Kidney. Brenner BM, Rector FC, Eds. Philadelphia, PA, Saunders, 1986, p. 1297-340 Zierler KL, Rabinowitz D: Effect of very small concentrations of insulin on forearm metabolism: persistence of its action on potassium and free fatty acids without its side effect on glucose. J Clin Invest 43:950-62, 1964 Andres R, Baltzan MA, Cader G, Zierler KL: Effect of insulin carbohydrate metabolism and on potassium in the forearm of man. J Clin Invest 41:108-14, 1962 Thiebaud D, Jacot E, DeFronzo RA, Maeder E, Jequier E, Felber JP: The effect of graded doses of insulin on total glucose uptake, glucose oxidation, and glucose storage in man. Diabetes 31: 957-63, 1982 Slatopolsky E, Caglar S, Gradowska L: On the prevention of secondary hyperparathyroidism in experimental chronic renal disease using "proportional re-

duction" of dietary phosphorus intake. Kidney Int 2:147-51, 1972 204. Marshall S, Olefsky J: Effect of insulin incubation on insulin binding, glucose transport, and insulin degradation by isolated adipocytes: evidence of hormone-induced desensitization at the receptor and post-receptor level. J Clin Invest 66:763-72, 1980 205. Amatruda JM, Newmeyer HW, Chang CL: Insulin-induced alterations in insulin binding and insulin action in primary cultures of rat hepatocytes. Diabetes 31:145-48, 1982 206. Mandarino L, Barker B, Rizza R, Genest J, Gerich JE: Infusion of insulin impairs human adipocyte glucose metabolism in vitro without decreasing adipocyte insulin receptor binding. Diabetologia 27:358-63, 1984 207. Rizza RA, Mandarino LA, Genest J, Baker BA, Gerich JE: Production of insulin resistance by hyperinsulinemia in man. Diabetologia 28:70-75, 1985 208. Marangou AG, Weber KM, Boston RC, Aitken PM, Heggie JCP, Kirsner LG, Best JD, Alford FP: Metabolic consequences of prolonged hyperinsulinemia in humans: evidence for induction of insulin insensitivity. Diabetes 35:1383-89, 1986 209. Sheehan P, Leonetri F, Rosenthal N: Effect of prolonged hyperinsulinemia on glucose metabolism (Abstract). Diabetes 35 (Suppl. 1):16A, 1986 210. Simonson DC, DeFronzo RA: Measurement of substrate oxidation and energy expenditure in man by indirect calorimetry: practical and theoretical consideration. Am] Physiol 258:E399-412, 1990 211. Shulman GI, Rothman DL, Jue T, Stein P, DeFronzo RA, Shulman RG: Quantitation of muscle glycogen synthesis in normal subjects and subjects with noninsulin-dependent diabetes by 13C nuclear magnetic resonance spectroscopy. N EnglJ Med 322:223-28, 1990 212. Lillioja A, Mott DM, Zawadzki JK, Young A A, Abbott WG, Bogardus C: Glucose storage is a major determinant of in vivo "insulin resistance" in subjects with normal glucose tolerance. J Clin Endocrinol Metab 62:922-27, 1986 213. Sims EAH, Danford E, Horton ES, Bray

GA, GlennonJA, Salans LB: Endocrine and metabolic effects of experimental

obesity in man. Recent Prog Horm Res

29:457-96, 1973 214. Bamett AH, Eff C, Leslie RD, Pyke DA: Diabetes in identical twins: a study of 200 pairs. Diabetologia 20:87-93, 1981 215. Newman B, Selby JV, King MC, Slemenda C, Fabsitz R, Friedman GD: Concordance for type 2 (non-insulindependent) diabetes mellitus in male twins. Diabetologia 30:763-68, 1987 216. Kobberling J, Tillil H: Empirical risk figures for first-degree relatives of noninsulin-dependent diabetics. In The Genetics of Diabetes Mellitus. Kobberling J,

217.

218.

219.

220.

221.

222.

223.

224.

225.

226.

Tattersall R, Eds. London, Academic, 1982, p. 201-10 Harris H: The familial distribution of diabetes mellitus: a study of the relatives of 1241 diabetic propositi. Ann Eugen 15:95-110, 1950 Viswanathan M, Mohan M, Snehalatha C, Ramachandran A: High prevalence of type 2 (non-insulin-dependent) diabetes among the offspring of conjugal type 2 diabetic parents. Diabetologia 28: 907-10, 1985 O'Rahilly SO, Wainscoat JS, Turner RC: Type 2 (non-insulin-dependent) diabetes mellitus: new genetics for old nightmares. Diabetologia 31:407-14, 1988 Simpson NED: Heritabilities of liability to diabetes when sex and age at onset are considered. Ann Hum Genet 32: 283-90, 1969 Kahn CR, White MF: The insulin receptor and the molecular mechanism of insulin action. J Clin Invest 82:115154, 1988 Olefsky JM: Perspectives in diabetes. The insulin receptor: a multifunctional protein. Diabetes 39:1009-15, 1990 Rosen OM: Banting Lecture. Structure and function of insulin receptors. Diabetes 38:1508-14, 1989 Carpentier JL: The cell biology of the insulin receptor. Diabetologia 32:62763, 1989 Saltiel AR: Second messengers of insulin action. Diabetes Care 13:244-53, 1990 Larner J: Banting Lecture. Insulinsignaling mechanisms: lessons learned

360

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

from the old testament of glycogen metabolism and the new testament of molecular biology. Diabetes 37:262-75, 1987 227. Denton RM: Early events in insulin ac-

228.

229.

230.

231.

232.

233.

234.

235.

236.

237.

238.

239.

binding and insulin action in humans. Danish Med Bull 31:1-32, 1984 240. Caro JF, Ittoop O, Pories WJ, Meelheim D, Flickinger FT, Jenquin M, Silverman JF, Khazanie PG, Sinka MK: Studies on the mechanism of insulin resistance in tions. Adv Cyclic Nucleotide Protein Phosthe liver from humans with non-insuphorylation Res 26:293-341, 1986 lin-dependent diabetes. Insulin action Taylor SI, Kadowaki T, Kadowaki H, and binding in isolated hepatocytes, inAccili D, Cama A, McKeon C: Mutasulin receptor structure, and kinase actions in insulin-receptor gene in insutivity. J Clin Invest 78:249-58, 1986 lin-resistant patients. Diabetes Care 13: 257-75, 1990 241. Comi RJ, Grunberger G, Gorden P: ReLane MD, Flores-Riveros JR, Hresko lationship of insulin binding and insuRC, Kaestner KH, Liao K, Janicot M, lin-stimulated tyrosine kinase activity is Hoffman RD, McLenithan JC, Kastelic altered in type II diabetes. J Clin Invest T, Christy RJ: Insulin-receptor tyrosine 79:453-62, 1987 kinase and glucose transport. Diabetes 242. Freidenberg GR, Henry RR, Klein HH, Care 13:565-74, 1990 Reichart DR, Olefsky JM: Decreased kiCushman SW, Wardzala LJ: Potential nase activity of insulin receptors from mechanism of insulin action on glucose adipocytes of non-insulin-dependent transport in the isolated rat adipose cell: diabetic subjects. J Clin Invest 79:240apparent translocation of intracellular 50,1987 transport systems to the plasma mem- 243. Takayama S, Kahn CR, Kubo K, Foley brane. J Biol Chem 255:4758-62, 1980 JE: Alterations in insulin receptor autoKlip A, Paquet MR: Glucose transport phosphorylation in insulin resistance: and glucose transporters in muscle and correlation with altered sensitivity to their metabolic regulation. Diabetes glucose transport and anti-lypolisis to Care 13:228-40, 1990 insulin. J Clin Endocrinol Metab 66:992Kasanicki MA, Pilch PF: Regulation of 90, 1988 glucose-transporter function. Diabetes 244. Theis RS, Molina JM, Ciaraldi TP, FreiCare 13:219-25, 1990 denberg GR, Olefsky JM: Insulin-recepThorens B, Charron MJ, Lodish HF: tor autophosphorylation and endogeMolecular physiology of glucose transnous substrate phosphorylation in porters. Diabetes Care 13:209-16, 1990 human adipocytes from control, obese, Kahn BB, Flier JS: Regulation of glucose and NIDDM subjects. Diabetes 39:250transporter gene expression in vitro and 58, 1990 in vivo. Diabetes Care 13:548-60, 1990 245. Trichitta V, Brunetti A, Chiavetta A, Bell Gl: Lilly Lecture. Molecular defects Benzi L, Papa V, Vigneri R: Defects in in diabetes mellitus. Diabetes 40:413insulin-receptor internalization and 22, 1990 processing in monocytes of obese subOlefsky JM: Insulin resistance and injects and obese NIDDM patients. Diasulin action: an in vitro and in vivo betes 38:1579-84, 1989 perspective. Diabetes 30:148-62, 1981 246. Caro JF, Sinha MK, Raju SM, Ittoop O, Rizza RA, Mandarino LJ, Gerich JE: Pories WJ, Flickinger EG, Meelheim D, Mechanism and significance of insulin Dohm GL: Insulin receptor kinase in resistance in non-insulin- dependent dihuman skeletal muscle from obese subabetes mellitus. Diabetes 30:990-95, jects with and without non-insulin1981 dependent diabetes. J Clin Invest 79: DePirro R, Fusco A, Lauro R, Testa 1, 1330-37, 1987 Ferreti F, DeMartinis C: Erythrocyte in- 247. Molina JM, Ciaraldi TP, Olefsky JM: sulin receptors in non-insulin-depenDecreased activation rate of insulindent diabetes mellitus. Diabetes 29:96mediated glucose transport in adipo99, 1980 cytes from obese and NIDDM subjects Pedersen O: Studies of insulin receptor (Abstract). Clin Res 36:157A, 1988

248. Arner P, Pollare T, Lithell H, Livingston JN: Defective insulin receptor tyrosine kinase in human skeletal muscle in obesity and type 11 (non-insulin-dependent) diabetes mellitus. Diabetologia 30: 437-40, 1987 249. Dohm GL: Insulin receptor kinase in human skeletal muscle from obese subjects with and without non-insulindependent diabetes. J Clin Invest 79: 1330-37, 1987 250. Arner P, Einarsson K, Ewerth S, Livingston J: Studies on the human liver insulin receptor in non-insulin-dependent diabetes mellitus. J Clin Invest 77: 1716-18, 1986 251. Kashiwagi A, Verso MA, Andrews J, Vasquez B, Reaven G, Foley JE: In vitro insulin resistance of human adipocytes isolated from subjects with non-insulindependent diabetes mellitus. J Clin Invest 72:1246-54, 1983 252. Bolinder J, Ostman J, Amer P: Postreceptor defects causing insulin resistance in normoinsulinemic non-insulin-dependent diabetes mellitus. Diabetes 31:911-16, 1982 253. Hidaka H, Nagulesparan M, Klines I, Clark R, Sasaki H, Aronoff SL, Vasquez B, Rubenstein AH, Unger RH: Improvement of insulin secretion but not insulin resistance after short term control of plasma glucose in obese type II diabetes
mellitus. J Clin Endocrinol Metab 54:

217-22, 1982 254. Nankervis A, Proietto J, Aitken P, Harwood M, Alford F: Differential effects of insulin therapy on hepatic and peripheral insulin sensitivity in type 11 (noninsulin-dependent) diabetes. Diabetologia 23:320-25, 1982 255. Okamoto M, Kuzuya H, Seino Y, Ikeda M, Imura H: Insulin binding to erythrocytes in diabetic patients. Endocrinol J 28:169-73, 1981 256. Seltzer H: Are insulin receptors clinically significant? J Lab Clin Med 100: 821-51, 1982 257. Lonnroth P, Digirolamo M, Krotkiewski M, Smith U: Insulin binding and responsiveness in fat cells from patients with reduced glucose tolerance and type II diabetes. Diabetes 32:748-54, 1983

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

361

Paihogenesis ofNIDDM

258. OlefskyJM, Reaven GM: Insulin binding in diabetes: relationships with plasma insulin levels and insulin sensitivity. Diabetes 26:680-88, 1977 259.

260.

261.

262.

263.

264.

265.

266.

267.

action in humans. Am J Physiol 258: E964-74, 1990 268. Freidenberg GR, Reichart D, Olefsky JM, Henry RR: Reversibility of defective adipocyte insulin receptor kinase activMandarino LJ, Campbell PJ, Gottesman ity in non-insulin dependent diabetes IS, Gerich JE: Abnormal coupling of mellitus: effect of weight loss. J Clin insulin receptor binding in non-insulinInvest 82:1398-406, 1988 dependent diabetes. Am J Physiol 247: E688-92, 1984 269. Freidenberg GR, Suter SL, Henry RR, Reichart D, OlefskyJM: In vivo stimuKahn CR: Role of insulin receptors in lation of the insulin receptor kinase in insulin resistant states. Metabolism 29: human skeletal muscle: correlation 455-66, 1980 with insulin-stimulated glucose disAvruch J, Memenoff RA, Pierce M, posal during euglycemic clamp studies. Kwok YC, BlackshearJP: Protein phosJ Clin Invest 87:2222-29, 1991 phorylation as a mode of insulin action. In Molecular Basis of Insulin Action. 270. Maegawa H, Shigeta Y, Egawa K, KobaCzech M, Ed. New York, Plenum, 1985, yashi M: Impaired autophosphorylation p. 263-296 of insulin receptors from abdominal skeletal muscles in nonobese subjects McClain DA: Insulin action in cells exwith NIDDM. Diabetes 40:815-19, 1991 pressing truncated or kinase-defective insulin receptors: dissection of multiple 271. Seino S, Seino M, Bell GI: Human insulin-receptor gene. Diabetes 39:129hormone signaling pathways. Diabetes 32, 1990 Care 13:302-14, 1990 Chou CK, Dull TJ, Russell DS, Gherzi 272. Kadowaki T, Kadowaki H, Rechler MM, Serrano-Rios M, Roth J, Gorden P, TayR, Lebwohl D, Ullrich A, Rosen OM: lor SI: Five mutant alleles of the insulin Human insulin receptors mutated at the receptor gene in patients with genetic ATP-binding site lack protein tyrosine forms of insulin resistance. J Clin Invest kinase activity and fail to mediate pos86:254-62, 1990 treceptor effects of insulin. J Biol Chem 273. Moller DE, Flier JS: Insulin-resistance 262:1842-47, 1987 mechanisms, syndrome and implicaEbina Y, Araki E, Tiara F, Shimada F, tions. N EnglJ Med 325:938, 1991 Mori M, Criak CS, Siddle K, Pierce SB, Roth RA, Rutter WJ: Replacement of 274. McClain DA, Henry RR, Ullrich A, Olefsky JM: Restriction-fragment-length lysine residue 1030 in the putative polymorphism in insulin-receptor gene ATP-binding region of the insulin reand insulin resistance in NIDDM. Diaceptor abolishes insulin- and antibodybetes 37:1071-75, 1988 stimulated glucose uptake and receptor kinase activity. Proc Natl Acad Sci USA 275. Raboudi SH, Mitchell BD, Stern MP, Eifter CW, Haffner SM, Hazuda HP, 84:704-708, 1987 Frazier ML: Type II diabetes mellitus Ellis LE, Clauser E, Morgan ME, Roth and polymorphism of insulin-receptor RA, Rutter WJ: Replacement of insulin gene in Mexican Americans. Diabetes receptor tyrosine residues 1162 and 38:975-79, 1989 1163 compromises insulin-stimulated kinase activity and uptake of 2-deoxy- 276. Xiang KS, Cox NJ, Sanz N, Huang P, Karam JH, Bell GI: Insulin receptor and glucose. Cell 45:721-32, 1986 apolipoprotein genes contribute to deBrillon DJ, Freidenberg GR, Henry RR, velopment of NIDDM in Chinese AmerOlefskyJM: Mechanism of defective inicans. Diabetes 38:17-23, 1989 sulin-receptor kinase activity in NIDDM: evidence for two receptor 277. Oelbaum RS, Bouloux PMG, Li SR, Baroni MG, Stocks J, Galton DJ: Insulin populations. Diabetes 38:397-403, receptor gene polymorphisms in type 2 1989 (non-insulin-dependent) diabetes melNyomba BL, Ossowski VM, Bogardus litus. Diabetologia 34:260-64, 1991 C, Mott DM: Insulin-sensitive tyrosine kinase relationship with in vivo insulin 278. Elbein SC, Borecki I, Corsetti L, Fajans

SS, Hansen AT, Nerup J, Province M, Permutt MA: Linkage analysis of the human insulin receptor gene and maturity onset diabetes of the young. Diabetologia 30:641-47, 1987 279. O'Rahilly S, Trembath RC, Patel P, Galton DJ, Turner RC, Wainscoat JS: Linkage analysis of the human insulin receptor gene in type II (non-insulin-dependent) diabetic families and a family with maturity onset diabetes of the young. Diabetologia 31:792-97, 1988 280. Elbein SC, Ward WK, Beard JC, Permutt MA: Familial NIDDM: moleculargenetic analysis and assessment of insulin action and pancreatic (3-cell function. Diabetes 37:337-82, 1988 281. Cox NJ, Epstein PA, Spielman RS: Linkage studies on NIDDM and the insulin and insulin-receptor genes. Diabetes 38: 653-58, 1989 282. Li SR, Oelbaum RS, Stocks J, Galton DJ: DNA polymorphisms of the insulinreceptor gene in Japanese subjects with non-insulin-dependent diabetes mellitus. Hum Hered 38:273-76, 1988 283. KusariJ, OlefskyJM, Strahl C, McClain DA: Insulin-receptor cDNA sequence in NIDDM patients homozygous for insulin-receptor gene RFLP. Diabetes 40: 249-54, 1991 284. Moller DE, Yokota A, Flier JS: Normal insulin receptor cDNA sequence in Pima Indians with non-insulin-dependent diabetes mellitus. Diabetes 38: 1496-500, 1989 285. Cama A, Patterson AP, Kadowaki T, Kadowaki H, Siegel G, D'Ambrosio D, Lillioja S, Roth J, Taylor SI: The amino acid sequence of the insulin receptor is normal in an insulin-resistant Pima Indian. J Clin Endocrinol Metab 70:115561, 1990 286. O'Rahilly S, Choi WH, Patel P, Turner RC, Flier JS, Moller DE: Detection of mutations in insulin-receptor gene in NIDDM patients by analysis of singlestranded conformation polymorphisms. Diabetes 40:777-82, 1991 287. McGuire MC, Fields RM, Nyomba BL, Raz I, Bogardus C, Tonks NK, Sommercornj: Abnormal regulation of protein tyrosine phospatase activities in skeletal muscle of insulin-resistant humans. Di-

362

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

amino acid sequence of rat brain hexabetes 40:939-42, 1991 okinase, deduced from the cloned 288. Sbraccia P, Goodman PA, Maddux BA, cDNA, and proposed structure of a Wong KY, Chen Y-Dl, Reaven GM, mammalian hexokinase. Proc Natl Acad Goldfine ID: Production of inhibitor of Sci USA 86:2563-67, 1989 insulin-receptor tyrosine kinase in fibroblasts from patient with insulin re- 299. Magnuson MA, Andreone 1L, Printz RL, sistance and N1DDM. Diabetes 40:295Koch S, Granner DK: The glucokinase 99,1991 gene: structure and regulation by insulin. Proc Natl Acad Sci USA 8 6 : 4 8 3 8 289. Yu KT, Khalaf N, Czech MP: Insulin 42, 1989 stimulates a membrane-bound serine kinase that may be phosphorylated on 300. Magnuson MA, Shelton KD: An altertyrosine. Proc Natl Acad Sci USA 84: nate promoter in the glucokinase gene 3972-76, 1987 is active in the pancreatic beta cell. J Biol Chem 264:15936-42, 1989 290. Kono T, Robinson FW, Blevins TL, Ezaki O: Evidence that translocation of 301. Ciaraldi TP, Kolterman OG, Scarlett JA, the glucose transport activity is the maKao M, Olefsky JM: Role of the glucose jor mechanism of insulin action on glutransport system in the postreceptor decose transport fat cells. J Biol Chem 257: fect of non-insulin-dependent diabetes 10942-47, 1982 mellitus. Diabetes 31:1016-22, 1982 291. Bell Gl, Kayano T, Buse JB, Burant J, 302. Foley JE, Kashawagi A, Verso MA, Reaven G, Andrews J: Improvement in TakedaJ, Lin D, Fukumoto H, Seino S: in vitro action after one month of insuMolecular biology of mammalian glulin therapy in obese non-insulin-decose transporters. Diabetes Care 13: pendent diabetes.J Clin Invest 72:1901198-206, 1990 909, 1983 292. Mueckler MM, Caruso C, Baldwin SA, Panico M, Blench 1, Morris HR, Allard 303. Garvey WT, Huecksteadt TP, Mattaei S, Olefsky JM: Role of glucose transporters JW, Lienhard GE, Lodish HF: Sequence in the cellular insulin resistance of type and structure of a human glucose trans11 non-insulin-dependent diabetes milporter. Science 229:941-45, 1985 letus. J Clin Invest 81:1528-36, 1988 293. Birnbaum MJ, Haspel HC, Rosen OM: Cloning and characterization of cDNA 304. Scarlett JA, Kolterman OG, Ciaraldi TP, Kao M, Olefsky JM: Insulin treatment encoding the rat brain glucose-transreverses the postreceptor defect in adiporter protein. Proc Natl Acad Sci USA pocyte 3-O-methyl-glucose transport in 83:5784-88, 1986 type 11 diabetes melliltus. J Clin Endo294. Kahn BB, Cushman SW: Subcellular crinol Metab 56:1195-201, 1983 translocation of glucose transporters: role in insulin action and its perturba- 305. Dohm GL, Tapscott EB, Pories WJ, Dabbs DJ, Flickinger EG, Meelheim D, tion in altered metabolic states. Diabetes Fushiki T, Atkinson SM, Elton CW, Metab Rev 1:203-27, 1985 Caro JF: An in vitro human muscle 295. Colowick SP: The hexokinases. In The prepartion suitable for metabolic studEnzymes. Vol. 9. Boyer PD, Ed. New ies: decreased insulin stimulation of York, Academic, 1973, p. 1-48 glucose transport in muscle from mor296. Nishi S, Susumu S, Bell Gl: Human bidly obese and diabetic subjects. J Clin hexokinase: sequences of amino- and Invest 82:486-94, 1988 carboxyl-terminal halves are homolo306. Sinha MK, Raineri-Maldonado C, Bugous. Biochem Biophys Res Comm 157: chanan C, Pories WJ, Carter-Su C, Pilch 937-43, 1988 PF, Caro JF: Adipose tissue glucose 297. Friedman JE, Dudek RW, Whitehead transporters in N1DDM: decreased levDS, Downes DL, Frisell WR, Caro JF, els of muscle/fat isoform. Diabetes 40: Dohm GL: lmmunolocalization of glu472-77, 1991 cose transporter GLUT4 within human skeletal muscle. Diabetes 40:150-54, 307. Garvey WT, Maianu L, Huecksteadt TP, Bimbaum MJ, Molina JM, Ciaraldi TP: 1991 Pretranslational suppression of a glu298. Schwab DA, Wilson JE: Complete

308.

309.

310.

311.

312.

313.

314.

315.

cose transporter protein causes insulin resistance in adipocytes from patients with non-insulin-dependent diabetes mellitus and obesity. J Clin Invest 87: 1072-81, 1991 Garvey WT, Huecksteadt TP, Bimbaum MJ: Pretranslational suppression of an insulin-responsive glucose transporter in rats with diabetes mellitus. Science 245:60-63, 1989 Berger J, Biswas C, Vicario P, Strout HV, Saperstein R, Pilch PF: Decreased expression of the insulin-responsive glucose transporter in diabetes and fasting. Nature (Lond) 340:70-72, 1989 Kahn B, Shulman GJ1, DeFronzo RA, Cushman SW, Rossetti L: Normalization of blood glucose in diabetic rats with phlorizin treatment reverses insulin resistant glucose transport in adipose cells without restoring glucose transporter gene expression. J Clin Invest 87:561-70, 1991 Pedersen O, Bak JF, Andersen PH, Lund S, Moler DE, Flier JS, Kahn BB: Evidence against altered expression of GLUT1 or GLUT4 in skeletal muscle of patients with obesity or N1DDM. Diabetes 39:865-70, 1990 Eriksson J, Koranyi L, Bourey R, Schalm-Jantti C, Widen E, Mueckler M, Permutt AM, Groop LC: Insulin resistance in type 2 (non-insulin-dependent) diabetic patients and their relatives is not associated with a defect in the expression of the insulin-responsive glucose transporter (GLUT-4) gene in human skeletal muscle. Diabetologia. In press KusariJ, Berma US, Buse JB, Henry RR, Olefsky JM: Analysis of the gene sequences of the insulin receptor and the insulin-sensitive glucose transporter (GLUT-4) in patients with commontype non-insulin-dependent diabetes mellitus. J Clin Invest 88:1323-30, 1991 Kahn BB, Rossetti L, Lodish HF, Charron MJ: Decreased in vivo glucose uptake but normal expression of GLUT1 and GLUT4 in skeletal muscle of diabetic rats. J Clin Invest 87:2197--206, 1991 Cobelli C, Saccomani MP, Ferrannini E,

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

363

Pathogenesis of NIDDM

DeFronzo RA, Gelfand R, Bonadonna R: A compartmental model to quantitate in vivo glucose transport in the human forearm. Am J Physiol 257: E943-58, 1989 316. Bonadonna RC, Del Prato S, Cobelli C, Saccomani MP, Stocco A, Bier D, Ferrannini E, DeFronzo RA: Glucose transport in skeletal muscle is insulin resistant in type 2 diabetes (Abstract). Diabetologia 33 (Suppl.):A23, 1990 317. Rossetti L, Smith D, Shulman Gl, Papachristou D, DeFronzo RA: Correction of hyperglycemia with phlorizin normalizes tissue sensitivity to insulin in diabetic rats. J Clin Invest 79:1510-15, 1987 318. FelberJP, Golay A, Jequier E, Curchod B, Temler E, DeFronzo RA, Ferrannini E: The metabolic consequences of longterm human obesity, lntj Obes 12:37789, 1988 319. Natali A, Buzzigoli G, Taddei S, Santoro D, Cerri M, Pedrinelli R, Ferrannini E: Effects of insulin on hemodynamics and metabolism in human forearm. Diabetes 39:490-98, 1990 320. Reaven GM, Chen Y-Dl, Donner CC, Fraze E, Hollenbeck CB: How insulin resistant are patients with non-insulindependent diabetes mellitus? J Clin Endocrinol Metab 61:32-36, 1985 321. Yki-Jarvinen H, Mott D, Young A A, Stone K, Bogardus C: Regulation of glycogen synthase and phosphorylase activity by glucose and insulin in human skeletal muscle. J Clin Invest 80:95100, 1987 322. Kruszynska YT, Home PD, Alberti KGMM: In vivo regulation of liver and skeletal muscle glycogen synthase activity by glucose and insulin. Diabetes 35: 662-67, 1986 323. Mandarino LJ, Wright KS, Verity LS, Nichols J, Bell JM, Kolterman OG, Beck-Nielsen H: Effects of insulin infusion on human skeletal muscle pyruvate dehydrogenase, phosphofructokinase, and glycogen synthase: evidence for their role in oxidative glucose metabolism. J Clin Invest 80:655-63, 1987 324. Thorburn AW, Gumbiner B, Bulacan F, Wallace P, Henry RR: Intracellular glucose oxidation and glycogen synthase

325.

326.

327.

328.

329.

330.

331.

332.

333.

activity are reduced in non-insulindependent (type II) diabetes independent of impaired glucose uptake. J Clin Invest 85:522-29, 1990 Mandarino LJ, MadarZ, Kolterman OG, Bell JM, Olefsky JM: Adipocyte glycogen synthase and pyruvate dehydrogenase in obese and type II diabetic subjects. Am] Physiol 25LE489-96, 1986 Wright KS, Beck-Nielsen H, Kolterman OG, Mandarino LJ: Decreased activation of skeletal muscle glycogen synthase by mixed-meal ingestion in NIDDM. Diabetes 37:436-40, 1988 Damsbo P, Vaag A, Hother-Nielsen O, Beck-Nielsen H: Reduced glycogen synthase activity in skeletal muscle from obese patients with and without type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia 34:239-45, 1991 Thorburn AW, Bumbiner B, Bulacan F, Brechtel G, Henry RR: Multiple defects in muscle glycogen synthase activity contribute to reduced glycogen synthesis in non-insulin dependent diabetes mellitus. J Clin Invest 87:489-95, 1991 Freymond D, Bogardus C, Okubo M, Stone K, Mott D: Impaired insulinstimulated muscle glycogen synthase activation in vivo in man is related to low fasting glycogen synthase phosphatase activity. J Clin Invest 82:1503509, 1988 Bogardus C, Lillioja S, Stone K, Mott D: Correlation between muscle glycogen synthase activity and in vivo insulin action in man. J Clin Invest 73:1185-90, 1984 Kruszynski YT, Petranyi G, Home PD, Taylor R, Alberti KGMM: Muscle enzyme activity and insulin sensitivity in type I (insulin-dependent) diabetes mellitus. Diabetologia 29:699-705, 1986 Kida Y, Esposito-DelPuente A, Bogardus C, Mott D: Insulin resistance is associated with reduced fasting and insulin-stimulated glycogen synthase phosphatase activity in human skeletal muscle. J Clin Invest 85:476-81, 1990 Maeda R, Raz I, Zurlo F, Sommercorn J: Activation of skeletal muscle casein kinase II by insulin is not diminished in subjects with insulin resistance. J Clin

Invest 87:1017-22, 1991 334. Kida Y, Nyomba BL, Bogardus C, Mott DM: Defective insulin response of cyclic adenosine monophosphate-dependent protein kinase in insulin-resistant humans. J Clin Invest 87:673-79, 1991 335. Bogardus C: Skeletal muscle and insulin action in vitro in man. In Current Concept Series. Kalamazoo, MI, Upjohn, 1987, p. 1-28 336. Browner MF, Nakano K, Bang AG, Fleffenzk RJ: Human muscle glycogen synthese with DNA sequence: a negatively charged protein with an asymetric charge distribution. Proc Natl Acad Sci USA 86:1443-47, 1989 337. Del Prato S, Bonadonna RC, Bonora E, Gulli G, Solini A, Shank M, DeFronzo RA: Characterization of cellular defects of insulin action in type 2 (non-insulindependent) diabetes mellitus. Submitted for publication. 338. Buffington CK, Stentz FB, Kitabchi AE: Activation of pyruvate dehydrogenase complex by porcine and biosynthetic human insulin in cultured human fibroblasts. Diabetes 33:681-85, 1984 339. Randle PJ, Garland PB, Hales CN, Newsholme EA: The glucose fatty acid cycle: its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1:785-89, 1963 340. Randle PJ, Newsholme EA, Garland PB: Regulation of glucose uptake by muscle: effects of fatty acids, ketone bodies and pyruvate, and of alloxan-diabetes and starvation on the uptake and metabolic fate of glucose in rat heart and diaphragm muscles. BiochemJ 93:65265, 1964 341. Taylor SI, Mukherjee C, Jungas RL: Regulation of pyruvate dehydrogenase in isolated rat liver mitochondria. J Biol Chem 250:2028-35, 1975 342. Taylor SI, Mukherjee C, Jungas RL: Studies on the mechanism of activation of adipose tissue pyruvate dehydrogenase by insulin .JBiol Chem 2 48:73 - 81, 1973 343. Jeanrenaud B, Halimi S, van de Werve G: Neuro-endocrine disorders seen as triggers of the triad: obesity-insulin resistance-abnormal glucose tolerance. Diabetes Metab Rev 1:261-92, 1985

364

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

344. Wititsuwannakul D, Kim K: Mechanism of palmityl coenzyme A inhibition or liver glycogen synthase. J Biol Chem 252:7812-17, 1977 345. Thiebaud D, DeFronzo RA, Jacot E, Golay A, Acheson K, Maeder E, Jequier E, Felber JP: Effect of long-chain triglyceride infusion on glucose metabolism in man. Metabolism 31:1128-36, 1982 346. Ferrannini E, Barrett EJ, Bevilacqua S, DeFronzo RA: Effect of fatty acids on glucose production and utilization in man. J Clin Invest 72:1737-47, 1983 347. Felber JP, Vannotti A: Effect of fat infusion on glucose tolerance and insulin plasma levels. Med Exp 10:153-56, 1964 348. RousselleJ, Buckert A, Pahud P, Jequier E, Felber JP: Relationship between glucose oxidation and glucose tolerance in man. Metabolism 31:866-70, 1982 349. Vouillamoz D, Temler E, Jequier E, Felber JP: Importance of substrate competition in the mechanism of insulin resistance in man. Metabolism 36:71520,1987 350. Boden G, Jadali F: Effects of lipid on basal carbohydrate metabolism in normal men. Diabetes 40:686-92, 1991 351. Yki-Jarvinen H, Puhakainen 1, Saloranta C, Groop L, Taskinen M-R: Demonstration of a novel feedback mechanism between FFA oxidation from intracellular and intravascular sources. Am J Physiol 260:E680-89, 1991 352. Lee KU, Lee HK, Koh CS, Min HK: Artificial induction of intravascular lipolysis by lipid-heparin infusion leads to insulin resistance in man. Diabetologia 31:285-90, 1988 353. Meyer HU, Curchod B, Maeder E, Pahud P, Jequier E, Felber JP: Modifications of glucose storage and oxidation in nonobese diabetics, measured by continuous indirect calorimetry. Diabetes 29:752-56, 1980 354. Golay A, Felber JP, Meyer HU, Curchod B, Maeder E, Jequier E: Study on lipid metabolism in obesity diabetes. Metabolism 33:111-16, 1984 355. Felber JP, Meyer HU, Curchod B, lselin H, Rousselle J, Maeder E, Pahud P, Jequier E: Glucose storage and oxida-

356.

357.

358.

359.

360.

361.

362.

363.

tion in different degrees of human obesity measured by continuous indirect calorimetry. Diabetologia 20:39-44, 1981 Felber JP, Magnenat G, Casthelaz M, Geser CA, Muller-Hess R, Dekalbermatten N, Ebiner JR, Curchod R, Pittet P, Jequier E: Carbohydrate and lipid oxidation in normal and diabetic subjects. Diabetes 26:693-99, 1977 Taskinen MR, Bogardus C, Kennedy A, Howard BV: Multiple disturbances of free fatty acid metabolism in noninsulin dependent diabetes. J Clin Invest 76: 637-44, 1985 Lillioja S, Bogardus C, Mott DM, Kennedy A, Knowler WC, Howard BV: Relationship between insulin mediated glucose disposal and lipid metabolism in man. J Clin Invest 75:1106-15, 1985 Chen Y-Dl, Golay A, Swislocki ALM, Reaven GM: Resistance to insulin suppression of plasma free fatty acid concentrations and insulin stimulation of glucose uptake in non-insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 64:7-21, 1987 Groop LC, Saloranta C, Shank M, Bonadonna RC, Ferrannini E, DeFronzo RA: The role of free fatty acid metabolism in the pathogenesis of insulin resistance in obesity and noninsulindependent diabetes mellitus. J Clin Endocrinol Metab 72:96-107, 1991 Ruderman NB, Toeus CJ, Shafir E: Role of free fatty acids in glucose homeostasis. Arch Intern Med 123:299-313, 1969 Blumenthal SA: Stimulation of gluconeogenesis by palmitic acid in rat hepatocytes: evidence that this effect can be dissociated from provision of reducing equivalents. Metabolism 32:971-76, 1983 Williamson JR, Kreisberg RA, Felts PW: Mechanism for the stimulation of gluconeogenesis of fatty acids in perfused
rat liver. Proc Natl Acad Sci USA 56:

247-54, 1966 364. Bonadonna RC, Groop LC, Zych K, Shank M, DeFronzo RA: Dose-dependent effect of insulin on plasma free fatty acid turnover and oxidation in humans. Am] Physiol 259:E736-50,1990

365. Bevilacqua S, Bonadonna R, Buzzigoli G, Boni C, Ciociaro D, Maccari F, Giorico A, Ferrannini E: Acute elevation of free fatty acid levels leads to hepatic insuilin resistance in obese subjects. Metabolism 36:502-506, 1987 366. Golay A, Swislocki ALM, Chen Y-Dl, Reaven GM: Relationships between plasma free fatty acid concentration, endogenous glucose production, and fasting hyperglycemia in normal and non-insulin-dependent diabetic individuals. Metabolism 36:692-96, 1987 367. Saloranta C, Franssila-Kallunki A, Ekstrand A, Taskinen M-R, Groop L: Modulation of hepatic glucose production by non-esterified fatty acids in type 2 (non-insulin-dependent) diabetes mellitus. Diabetolgia 34:409-15, 1991 368. Fritz IB: Factors influencing the rates of long chain fatty acid oxidation and synthesis in mammalian systems. Physiol Rev 41:52-116, 1961 369. Groop LC, Bonadonna RC, Shank M, Petrides AS, DeFronzo RA: Role of free fatty acids and insulin in determining free fatty acid and lipid oxidation in man. J Clin Invest 87:83-89, 1991 370. Felig P, Wahren J, Hendler R: Splanchnic glucose and amino acid metabolism in obesity. J Clin Invest 53:582-90, 1974 371. Baron AD, Schaeffer L, Shragg P, Kolterman OG: Role of hyperglucagonemia in maintenance of increased rates of hepatic glucose output in type II diabetics. Diabetes 36:274-83, 1987 372. Reaven GM, Chen Y-Dl, Golay A, Swislocki AL, Jaspan JB: Documentation of hyperglucagonemia throughout the day in nonobese and obese patients with non-insulin dependent diabetes mellitus. J Clin Endocrinol Metab 64: 106-10, 1987 373. Unger RH, Aguilar-Parada E, Mueller WA, Eisentrant AM: Studies of pancreatic alpha cell function in normal and diabetic subjects. J Clin Invest 49:83739, 1970 374. Boden G, Soriano M, Hoeldtke RD, Owen GE: Counterregulatory hormone release and glucose recovery after hypoglycemia in non-insulin-dependent diabetic patients. Diabetes 32:1055-59,

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

365

Pathogenesis of NIDDM

1983 375. Clore JN, Glickman PS, Nestler JE, Blackard WG: In vivo evidence for hepatic autoregulation during FFA-stimulated gluconeogenesis in normal humans. AmJ Physiol 26LE425-29, 1991 376. Bonadonna RC, Zych K, Boni C, Ferrannini E, DeFronzo RA: Time dependence of the interaction between lipid and glucose in humans. Am J Physiol 257:E49-56, 1989 377. Diamond MP, Rollings RC, Steiner KE, Williams PE, Lacy WW, Cherrington AD: Effect of alanine concentration independent of changes in insulin and glucagon on alanine and glucose homeostasis in the conscious dog. Metabolism 37:28-33, 1988 378. Jahoor F, Peters EJ, Wolfe RR: The relationship between gluconeogenic substrate supply and glucose production in humans. Am J Physiol 258:E288-96, 1990 379. Jenssen T, Nurjhan N, Consoli A, GerichJE: Failure of substrate-induced gluconeogenesis to increase overall glucose appearance in normal humans. J Clin Invest 86:489-97, 1990 380. Steele R, Winkler B, Altszuler N: Inhibition by infused glycerol of gluconeogenesis from other precursors. Am J Physiol 221:883-88, 1971 381. Lillioja S, Young AA, Culter CL, Ivy JL, Abbott GH, Zawadzki JK, Yki-Jarvinen H, Christin L, Secomb TW, Bogardus C: Skeletal muscle capillary density and fiber type are possible determinants of in vivo insulin resistance in man. J Clin Invest 80:415-24, 1987 382. King GL, Johnson SM: Receptor-mediated transport of insulin across endothelial cells. Science 227:1583-86, 1985 383. Yeon JY, Hope ID, Ader M, Bergman RN: Insulin transport across capillaries is rate limiting for insulin action in dogs. J Clin Invest 84:1620-28, 1989 384. Laakso M, Edelman SV, Brechtel G, Baron AD: Decreased effect of insulin to stimulate skeletal muscle blood flow in obese man: a novel mechanism for insulin resistance. J Clin Invest 85:184452, 1990 385. Laakso M, Edelman S, Brechtel G,

Baron A: Decreased insulin mediated glucose uptake (IMGU) in NIDDM subjects: role of blood flow (Abstract). Diabetes 40 (Suppl. l):100A, 1991 386. Anderson EA, Hoffman RP, Balon TW, Sinkey CA, Mark AL: Hyperinsulinemia produces both sympathetic neural activation and vasodilation in normal humans. J Clin Invest 87:2246-52, 1991 387.

gene-related peptide. Diabetes 39:26064, 1990 396. Frontoni S, Cho SB, Banduch D, Rossetti L: In vivo insulin resistance induced by amylin primarily through inhibition of insulin-stimulated glycogen synthesis in skeletal muscle. Diabetes 40:568-73, 1991

388.

389.

390.

391.

392.

393.

394.

395.

397. Koopmans SJ, van Mansfeld ADM, Jansz HS, Krans HMJ, Radder JK, Frolich M, deBoer SD, Kreutter DK, AnCreager MA, Liang CS, Coffman JD: drews GC, Maassen JA: Amylin-inBeta adrenergic-mediated vasodilator duced in vivo insulin resistance in response to insulin in the human foreconscious rats: the liver is more sensiarm. J Pharmacol Exp Ther 235:709-14, tive to amylin than peripheral tissues. 1985 Diabetologia 34:218-24, 1991 Chisholm DJ, Klassen GA, Dupre J, Pozefsky T: Interaction of secretin and 398. Kassir AA, Upadhyay AK, Lim TJ, Moossa AR, Olefsky JM: Lack of effect insulin on human forearm metabolism. of islet amyloid polypeptide in causing EurJ Clin Invest 5:487-94, 1975 insulin resistance in conscious dogs Nishi MM, Sanke T, Nagamatsu S, Bell during euglycemic clamp studies. DiaGI, Steiner DF: Islet amyloid polypepbetes 40:998-1004, 1991 tide. J Biol Chem 265:4173-76, 1990 Johnson KH, O'Brien TD, Westermark 399. Butler PC, Chou J, Carter WB, Wang YN, Bu BH, Chang D, Chang JK, Rizza P: Newly identified pancreatic protein RA: Effects of meal ingestion on plasma islet amyloid polypeptide: what is its amylin concentration in NIDDM and relationship to diabetes? Diabetes 40: nondiabetic humans. Diabetes 39:752310-14, 1991 55, 1990 Steiner DF, Ohagi S, Nagamatsu S, Bell GI, Nishi M: Is islet amyloid polypep- 400. Mitchell JH, Schmidt RF: Cardiovascular reflex control by afferent fibers from tide a significant facator in pathogenesis skeletal muscle receptors. In Handbook or pathophysiology of diabetes? Diabeof Physiology. The Cardiovascular System: tes 40:305-309, 1991 Peripheral and Organ Blood Flow. Sec. 2, Cooper GJS, Leighton B, Dimitriadis Vol. 3. Shepherd JT, Abboud FM, Eds. GD, Parry-Billings M, Kowalchuk JM, Bethesda, MD, 1983, p. 623-58 Howland K, Rothbard JB, Willis AC, Reid KBM: Amylin found in amyloid 401. Xiao-Ying H: Tachykinins and calcitonin gene-related peptide in relation to deposits in human type II diabetes melperipheral functions of capsaisin-sensilitus may be a hormone that regulates tive sensory neurons. Ada Physiol Scand glycogen metabolism in skeletal musSuppl 551:1-45, 1986 cle. Proc Natl Acad Sci USA 8 5 : 7 7 6 3 402. LevyJ, Gavin JR, Fausto A, Gingerich 66, 1988 RL, Avioli L: Impaired insulin action in Cooper GJS, Leighton B: Pancreatic rats with non-insulin-dependent diabeamylin and calcitonin gene-related peptes. Diabetes 33:901-908, 1984 tide cause resistance to insulin in skeletal muscle in vitro. Nature (Lond) 335: 403. Bevilacqua S, Barrett EJ, Smith D, Simonson DC, Olsson M, Bratusch632-35, 1988 Marrian P, Ferrannini E, DeFronzo RA: Sowa R, Sanke T, HirayamaJ, Tabata H, Hepatic and peripheral insulin resisFuruta H, Nishimura S, Nanjo K: Islet tance following streptozotocin-induced amyloid polypeptide amide causes peinsulin deficiency in the dog. Metaboripheral insulin resistance in vivo in lism 34:817-25, 1985 dogs. Diabetologia 33:118-20, 1990 Molina JM, Cooper GJS, Leighton B, 404. Reaven GM, Sageman WS, Swenson RS: Development of insulin resistance in OlefskyJM: Induction of insulin resisnormal dogs following alloxan-induced tance in vivo by amylin and calcitonin

366

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

DeFronzo

405.

406.

407.

408.

409.

410.

411.

412.

413.

414.

415.

416.

insulin deficiency. Diabetologia 13: 459-62, 1977 Kamielli E, Armoni M, Cohen P, Kanter Y, Rafaeloff R: Reversal of insulin resistance in diabetic rat adipocytes by insulin therapy. Diabetes 36:925-31, 1987 Sasson S, Cerasi E: Substrate regulation of the glucose transport system in rat skeletal muscle. J Biol Chem 261: 16827-33,1986 Sasson S, Edelson D, Cerasi E: In vitro autoregulation of glucose utilization in rat soleus muscle. Diabetes 36:104146, 1987 Garvey WT, Olefsky JM, Matthaei S, Marshall S: Glucose and insulin coregulate the glucose transport system in primary cultured adipocytes. J Biol Chem 262:189-97, 1987 vanPuttenJPM, Krans HMJ: Glucose as a regulator of insulin-sensitive hexose uptake in 3T3 adipocytes. J Biol Chem 260:7996-8001, 1985 Marshall S: Kinetics of insulin action on protein synthesis in isolated adipocytes. J Biol Chem 264:2029-36, 1989 Traxinger RR, Marshall S: Recovery of maximal insulin responsiveness and insulin sensitivity after induction of insulin resistance in primary cultured adipocytes. J Biol Chem 264:8156-63, 1989 Richter EA, Hansen BF, Hansen SA: Glucose-induced insulin resistance of skeletal muscle glucose transport and uptake. BiochemJ 252:733-37, 1988 Kktzein RF, Perdue JF: Induction of sugar transport in chick embryo fibroblasts by hexose starvation: evidence for transcriptional regulation of transport. J Biol Chem 250:593-600, 1985 Ullrey D, Gammon BMT, Kalckar HM: Uptake patterns and transport enhancements in cultures of hamster cells deprived of carbohydrates. Arch Biochem Biophys 167:410-16, 1975 Ullrey DB, Kalckar HM: The nature of regulation of hexose transport in cultured mammalian fibroblasts: aerobic "repressive" control by D-glucosamine. Arch Biochem Biophys 209:168-74, 1975 Haspel C, Wilk EW, Birnbaum MJ,

428. Kolterman OG, Gray RS, Shapiro G, Scarlett JA, Griffin J, Olefsky JM: The abetes mellitus. Proc Natl Acad Sci USA acute and chronic effects of sulfony79:5406-10, 1982 lurea therapy in type II diabetes. Diabetes 33:346-54, 1984 419. Matthaei S, Horuk R, Olefsky JM: Blood-brain glucose transfer in diabetes 429. Simonson DC, Ferrannini E, Bevilacqua mellitus. Diabetes 35:1181-84, 1986 S, Smith D, Barrett E, Carlson R, De420. McCall AL, Fixman LB, Fleming N, Fronzo RA: Mechanism of improveTornheim K, Check W, Ruderman NB: ment in glucose metabolism after Chronic hypoglycemia increases brain chronic glyburide therapy. Diabetes 33: glucose transport. Am J Physiol 251: 838-45, 1984 E442-47, 1986 430. Ma A, Kamp M, Bird D, Howlett V, 421. Yki-Jarvinen H, Helve E, Koivisto VA: Cameron DP: The effects of long term Hyperglycemia decreases glucose upglicazide administration on insulin setake in type 1 diabetes. Diabetes 36: cretion and insulin sensitivity. Aust NZ J Med 19:44-49, 1989 892-96, 1987 422. DelPrato S, Sheehan P, Leonetti F, Si- 431. Groop L, Schalin C, Franssila-Kallunki A, Widen E, Ekstrand A, Eriksson J: monson DC: Effect of chronic physioCharacteristics of non-insulin-depenlogic hyperglycemia on insulin secredent diabetic patients with secondary tion and glucose metabolism (Abstract). failure to oral antidiabetic therapy. AmJ Diabetes 35 (Suppl. 1):196A, 1986 Med 87:183-90, 1989 423. Laasko M, Usitupa M, TakalaJ, Majander H, Reijonen T, Penttila I: Effects of 432. Gormley MJJ, Hadden DR, Woods R, Sheridan B, Andrews WJ: One month's hypocaloric diet and insulin therapy on insulin treatment of type 11 diabetes: the metabolic control and mechanisms of early and medium-term effects followhyperglycemia in obese non-insulining insulin withdrawal. Metabolism 35: dependent diabetic subjects. Metabolism 1029-36, 1986 37:1092-100, 1988 424. ZawadskiJK, Bogardus C, FoleyJE: In- 433. Greenfield M, Lardinois C, Doberne L, Vreman H: Metabolic effects of intensulin action in obese non-insulinsive insulin therapy in patients with dependent diabetes and in their isolated n o n - i n s u l i n - d e p e n d e n t diabetes adipocytes before and after weight loss. (NIDDM) (Abstract). Diabetes 31 Diabetes 36:227-36, 1987 (Suppl. 2):58A, 1982 425. Walshe K, Andrews WJ, Sheridan B, Woods R, Hadden DR: Three months 434. Castillo M, Scheen AJ, Paolisso G, Lefebvre PJ: The addition of glipizide to energy restricted diet does not induce insulin therapy in type II diabetic paperipheral insulin resistance in new ditients with secondary failure to sulfonyagnosed non-insulin-dependent diabetlureas is useful only in the presence of a ics. Horm Metabo Res 19:197-200, significant residual insulin secretion. 1987 Acta Endocrinol 116:364-72, 1987 426. Firth RG, Bell PM, Rizza RA: Effects of tolazamide and exogenous insulin on 435. Yki-Jarvinen H, Nikkila E, Eero H, Taskinen MR: Clinical benefits and insulin action patients with non-insumechanisms of a sustained response to lin-dependent diabetes mellitus. N EngI intermittent insulin therapy in type 11 JMed 314:1280-86, 1986

Cushman SW, Rosen OM: Glucose deprivation and hexose transporter polypeptides of murine fibroblasts. J Biol Chem 261:6778-89, 1986 417. Gjedde A, Crone C: Blood-brain glucose transfer: repression in chronic hyperglycemia. Science 214:456-57, 1981 418. McCall AL, Millington WR, Wurtman RJ: Metabolic fuel and amino acid transport into the brain in experimental di-

427. Mandarino LJ, Gerich JE: Prolonged sulfonylurea administration decreases insulin resistance and increases insulin secretion in non-insulin-dependent diabetes mellitus: evidence for improved insulin action at a postreceptor site in hepatic as well as extrahepatic tissues. Diabetes Care 7 (Suppl. 1):89~99, 1984

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992

367

Pathogenesis ofNIDDM

diabetic patients with secondary drug sensitivity in newly diagnosed type I failure. Am] Med 84:185-92, 1988 diabetics after ketoacidosis and after 436. Beck-Nielsen H, Richelsen B, Hasling C, three months of insulin therapy. J Clin Neilsen OH, Heding L, Sorensen NS: Endocrinol Metab 59:371-78, 1984 Improved in vivo insulin effect during 440. Lager I, Lonnroth P, VonSchenck H, continuous subcutaneous insulin infuSmith U: Reversal of insulin resistance sion in patients with 1DDM. Diabetes in type I diabetes after treatment with 33:832-37, 1984 continuous subcutaneous insulin infusion. Br Med] 287:1001-1004, 1983 437. Simonson DC, Tamborlane WC, Sherwin RS, DeFronzo RA: Improved insu- 441. Gray RS, Cowan P, Duncan LJP, Clarke lin sensitivity in patients with type I BF: Reversal of insulin resistance in type diabetes mellitus after CSII. Diabetes 34 I diabetes following initiation of in(Suppl. 3):80-86, 1985 sulin treatment. Diabetic Med 3:18-23, 1986 438. Yki-Jarvinen H, Koivisto VA: Continuous subcutaneous insulin infusion ther- 442. Bell GI, Xiang K-S, Newman MV, Wu S-H, Wright LG, Fajans SS, Spielman apy decreases insulin resistance in type RS, Cox NJ: Gene for non-insulinI diabetes. ] Clin Endocrinol Metab 58: dependent diabetes mellitus (maturity659-66, 1984 onset diabetes of the young subtype) is 439. Yki-Jarvinen H, Koivisto VA: Insulin

linked to DNA polymorphism on human chromosome 20 q. Proc Natl head Sci USA 88:1484-88, 1991 443. Lillioja S, Bogardus C: Obesity and insulin resistance: lessons learned from the Pima Indians. Diabetes Metab Rev 4:515-40, 1988 444. Samanta A, Burden AC, Jones GR, Clarkson L: The effect of short term intensive insulin therapy in non-insulin-dependent diabetics who had failed on sulphonylurea therapy. Diabetes Res 3:269-71, 1986 445. Hollenbeck CB, Reaven GM: Treatment of patients with non-insulin-dependent diabetes mellitus: diabetic control and insulin secretion and action after different treatment modalities. Diabetic Med 4:311-16, 1987

368

DIABETES CARE, VOLUME 15,

NUMBER 3, MARCH

1992