Diabetologia 9 Springer-Verlag 1992

Transcription

1 Diabetlgia (1992) 35: Diabetlgia 9 Springer-Verlag 1992 Fr debate Pathgenesis f Type 2 (nn-insulin dependent) diabetes mellitus: a balanced verview* R. A. DeFrnz Divisin f Diabetes, The University f Texas Health Science Center and Audie L. Murphy Veterans Administratin Hspital, San Antni, Texas, USA Fllwing an vernight fast the majrity f glucse dispsal ccurs in insulin-independent tissues, the brain (-5 % ) and splanchnic rgans (-25 %), while nly 25 % ccurs in insulin-dependent tissues, primarily muscle [1-4]. Basal glucse utilizatin (-2 mg.kg -~.min 1) is precisely matched by glucse prductin by the liver [1-4]. Fllwing glucse ingestin, the balance between uptake and utput is disrupted and maintenance f glucse hmestasis depends upn three prcesses that must ccur in a c-rdinated fashin: (1) insulin secretin; (2) stimulatin f glucse uptake by splanchnic (liver and gut) and peripheral (primarily muscle) tissues in respnse t hyperinsulinaemia plus hyperglycaemia; (3) suppressin f hepatic glucseprductin. It lgically fllws that abnrmalities at the level f the Beta cell, muscle, and/r liver can lead t the develpment f glucse intlerance. The full blwn syndrme f Type 2 (nn-insulin-dependent) diabetes mellitus requires the simultaneus presence f tw defects, insulin resistance and impaired Beta-cell functin. In Type 2 diabetes the primary r inherited defect mst likely represents impaired tissue (muscle and/r liver) sensitivity t insulin. Eventually, hwever, the Beta cell fails t maintain a sufficiently high rate f insulin secretin t cmpensate fr the insulin resistance, and vert diabetes mellitus ensues. Insulin secretin in Type 2 (nn-insulin-dependent) diabetes Study f Beta-cell functin in Type 2 diabetes has demnstrated a cnsistent pattern which reveals a cmplex inter- The sectin "Fr debate..." is pen t cntributins dealing with issues f a particularly debatable nature in diabetlgy. Cntributins are published either standing by themselves r accmpanied by invited cmments. Other cmments frm the readership may be published as Letters t the Editr. Manuscripts intended fr publicatin in this sectin f the jurnal are accepted at the discretin f the Edit>in-Chief and may be subject t a referee prcedure. * This Fr debate article has been written in respnse t an article previusly published in Diabetlgia n this subject. J.E. Gerich (1991) Diabetlgia 34:67-61 play between insulin secretin and insulin sensitivity. In individuals with impaired glucse tlerance (IGT) and mild diabetes, the fasting plasma insulin cncentratin is invariably increased and basal insulin secretin is enhanced [5-7] (Fig. 1). As the fasting glucse increases frm 4.5 t 7.8 mml/1 (8 t 14 mg/dl) fasting plasma insulin prgressively rises. When the fasting glucse exceeds 7.8 mml/1 (14 mg/dl), insulin secretin drps ff precipitusly (Fig. 1). The inability f the pancreas t maintain its high rate f insulin secretin has imprtant pathphysilgic implicatins, since it is at this pint (fasting glucse = 7.8 mml/1 = 14 mg/dl) that hepatic glucse prductin increases in abslute terms and begins t cntribute t the elevatin in fasting plasma glucse [7]. In Type 2 diabetic subjects, when glucse-stimulated insulin secretin is pltted against the fasting glucse, the same inverted "U" shaped curve is bserved [1] (Fig. 2). The pancreatic functin curves displayed in Figures 1 and 2 are cnsistent with the natural histry f IGT and Type 2 diabetes in man [1, 8-13] and in the rhesus mnkey [14, 15] (Fig. 2). Even befre the develpment f IGT, insulin resistance is well-established (Fig.2). Prgressin frm nrmal t IGT t Type 2 diabetes with mild fasting hyperglycaemia (6.7 t 7.8 mml/1 = 12 t 14 mg/dl) is assciated with a marked increase in bth fasting and glucse-stimulated plasma insulin levels [1, 8, 9, 13] (Fig. 2). Overt fasting hyperglycaemia ( > 7.8 mml/1 = 14 mg/dl) results frm an inability f the Beta cell t maintain its high rate f insulin secretin (Fig. 2). A similar pattern f insulin secretin ccurs during the develpment f diabetes in the Rhesus mnkey [14, 15]. These studies [1, 9, 1, 12, 13-16] cnclusively dcument that hyperinsulinaemia precedes Type 2 diabetes and exclude the pssibility that insulinpenia initiates the prcess f diabetes. Studies in ethnic grups at high risk f develping Type 2 diabetes [8, 9, 16-2] and in first-degree relatives f Type 2 diabetic individuals [18, 21, 22] als dcument that hyperinsulinaemia predicts the develpment f Type 2 diabetes. In summary, the earliest stages f Type 2 diabetes are characterized by augmented insulin secretin, which represents a cmpensatry respnse t insulin resistance.

2 ~ 39 ==.=_ 25 2O u_.- 1 J [ I I I I I i 6 I Fasting plasma glucse (mg/dl) Fig.1. Relatinship between fasting plasma glucse cncentratin and fasting plasma insulin cncentratin in nrmal weight cntrl subjects, in individuals with impaired glucse tlerance, and in Type 2 (nn-insulin-dependent) diabetic subjects with varying degrees f fasting hyperglycaemia. As fasting plasma glucse cncentratin rises frm baseline t 7.8 mml/1 (14 mg/dl) there is a prgressive increase in the fasting insulin cncentratin. Thereafter, further rises in the fasting glucse are assciated with a prgressive decline in fasting insulin. In diabetic subjects with fasting glucse cncentratins in excess f mml/1 (2-22 mg/dl), fasting insulin declines t values bserved in cntrl subjects. Reprduced frm reference 7 with permissin The cause f late-nset Beta-cell failure remains unknwn. Beta-cell mass is nly mdestly reduced (by 2-4 %) [23, 24] and genetic mutatins f the insulin gene are uncmmn [25]. A likely explanatin fr the acquired defect in insulin secretin relates t the cncept f "glucse txicity", in which chrnic sustained hyperglycaemia in a (genetically) predispsed Beta cell leads t impaired insulin secretin [1, 26, 27]. We nw shall shift frm the pancreas t the insulin sensitive tissues, muscle and liver. Hwever, we shall return t the Beta cell t examine the dynamic interactin between insulin actin and insulin secretin [1, 7-9, 12, 15, 21, 28-3], since it is the disruptin f this finely regulated balance which leads t the develpment f vert diabetes mellitus. Insulin resistance in Type 2 diabetes Lngitudinal and crss-sectinal studies have cnclusively dcumented that hyperinsulinaemia antedates the develpment f Type 2 diabetes [1, 8-21, 31, 32], and studies emplying the euglycaemic insulin clamp have demnstrated that prgressin frm nrmal t IGT is assciated with the develpment f severe insulin resistance [1, 8, 9,13-15,21]. The defect in insulin actin is seen at all plasma insulin cncentratins spanning the physilgic and pharmaclgic range [33-35]. These bservatins prvide cnvincing evidence that insulin resistance, nt impaired insulin secretin, initiates the prcess f Type 2 diabetes in man. Site f insulin resistance Fllwing the stimulatin f insulin secretin, whle bdy glucse hmestasis is dependent upn three tightly cupled mechanisms: (1) suppressin f hepatic glucse R. A. DeFrnz: Pathgenesis f Type 2 diabetes prductin; (2) augmentatin f splanchnic (hepatic plus gastrintestinal) glucse uptake; (3) stimulatin f glucse dispsal by peripheral tissues, primarily muscle. The cntributin f each f these prcesses t the insulin resistance in Type 2 diabetes will be reviewed. Hepatic glucse prductin (HGP) De Frnz et al. have shwn that the liver f healthy subjects prduces glucse at mg. kg-1. min-1 in the pstabsrptive state [1, 2, 4, 33, 36-41]. Type 2 diabetic subjects with mderate fasting hyperglycaemia demnstrate a cnsistent increase (-.5 mg. kg 1. min- 1) in basal HGP (Fig.3), which is clsely crrelated (r =.847, p <.1) with the degree f fasting hyperglycaemia [1, 2, 4, 33, 36-41]. When extraplated ver 24 h the liver f a 7 kg patient with mild diabetes prduces an additinal 5 g f glucse each day. These results indicate that in Type 2 diabetic individuals with vert fasting hyperglycaemia ( > 7.8 mml/1 = 14 mg/dl) excessive HGP is an imprtant determinant f fasting hyperglycaemia. Similar results have been published by thers [15, 3, 34, 35, 42-45]. In the pstabsrptive state, plasma insulin cncentratin is increased 2-3 fld in diabetic individuals [1, 2, 4, 7, 33, 36-41] (Fig. 1). Since hyperinsulinaemia is a pwerful inhibitr f HGP [1, 2, 33, 46], it is clear that hepatic resistance t insulin is present and cntributes t the excessive basal utput f glucse by the liver. In respnse t insulin, suppressin f HGP is impaired in Type 2 diabetic patients with mderate t severe fasting hyperglycaemia [33-35] but this defect is less evident in diabetic,~176 t 8F ='~ 4 I O,... O 2 L 4 "I~ 3 2 Q. ==~ 1.-~ I I I I I I CON OB OB- OB- OB i Insulin-mediated 2 glucse uptake (mg/rff. rain) GLU DIAB DIAB INTOL Hi INS L INS Fig.2. Summary f the plasma glucse (bttm panel) and plasma insulin (tp panel, pen circles) respnses during 1 g OGTT and tissue sensitivity t insulin (tp panel, clsed circles) in cntrl, bese nn-diabetic, bese glucse intlerant, bese hyperinsulinaemic diabetic, and bese hypinsulinaemic diabetic subjects. See text fr a detailed discussin. Reprduced frm reference 1with permissin 15 1

3 R. A. DeFrnz: Pathgenesis f Type 2 diabetes 391 c- O O_ T i~iiiii~i~iii~ii!!ii~!iiiiiii'iiiii!,!ii ~i~ii - 9 *~ (3 %~176 ~ insulin resistance under euglycaemic cnditins. In respnse t insulin leg glucse uptake prmptly increased M-fld in cntrl subjects [4]. In cntrast, in Type 2 diabetic subjects the nset f insulin actin was markedly delayed and leg glucse uptake was decreased by 4-5 % [4]. Similar results have been reprted using the frearm catheterizatin technique [47-55]. Since in man adipcytes are quite inert [56], the primary leg and frearm tissue respnsible fr glucse remval must be muscle and it can be calculated that impaired muscle glucse uptake accunts fr the great majrity (-9 %) f the decrease in ttal bdy glucse dispsal during an insulin clamp study in Type 2 diabetic subjects [1, 4]. Glucse dispsal during ral glucse tlerance test (OGTT) 1.5 I I [ [ I I J J I I J I [ [ [ [ J I I [ J 5 I :5 Fasting plasma glucse (mg/dl) Fig. 3. Summary f hepatic glucse prductin in 77 nrmal weight Type 2 (nn-insulin-dependent) diabetic subjects (pen circles) with fasting plasma glucse cncentratin ranging frm 5.8 t 16.7mml/1 (15 t 3mg/dl). Seventy-tw age- and weightmatched cntrl subjects are indicated by the dsed circles. In the 33 diabetic subjects with fasting plasma glucse levels belw 7.8 retl/1 (1.4 mg/dl) (shaded area), the mean rate f hepatic glucse prductin was identical t cntrl subjects. In diabetic subjects with fasting plasma glucse cncentratin abve 7.8 mml/1 (14 mg/dl), there was a prgressive rise in hepatic glucse prductin that crrelated clsely (r =.847, p <.1) with the fasting plasma glucse cncentratin. Reprduced frm reference 1 with permissin patients with mild fasting hyperglycaemia [1, 2, 4, 36-41]. Frm the quantitative standpint, hwever, impaired suppressin f HGP can accunt fr nly a small percentage (5-1 %) f the defect in whle bdy glucse metablism during an englycaemic insulin clamp study [1, 2, 4, 33-41]. Splanchnic (hepatic) glucse uptake Emplying hepatic vein catheterizatin, DeFrnz et al. [4] have dcumented net glucse release frm the splanchnic area in the pstabsrptive state in bth cntrl and Type 2 diabetic subjects, reflecting glucse prductin by liver. In respnse t insulin there was a prmpt suppressin f splanchnic glucse utput (reflecting the inhibitin f HGP) and after 2 h there was a small net splanchnic uptake f glucse, which averaged -.5 mg.kg -1-min -1 in bth cntrl and diabetic subjects. These results demnstrate that impaired splanchnic (liver plus gut) glucse uptake cannt explain the insulin resistance bserved in Type 2 diabetic subjects under euglycaemic hyperinsulinaemic cnditins. Peripheral (muscle) glucse uptake Using the insulin clamp technique with femral vein/artery catheterizatin t quantitate leg glucse exchange, muscle tissue has been shwn t be the primary site f The nrmal rute f glucse entry is via the gastrintestinal tract. Emplying hepatic vein catheterizatin with a dual istpe technique, Ferrannini et al. [57] have examined the cntributin f splanchnic and peripheral tissues t glucse dispsal in healthy subjects (Table 1). During 3.5 h fllwing glucse (68 g)ingestin: (1) 19 g r 28 % f the ral lad was taken up by splanchnic tissues, (2) 48 g r 71% was dispsed f by peripheral tissues; (3) f the 48 g taken up by peripheral tissues, the brain (an insulin-independent tissue) accunts fr -15 g (-1 mg. kg -1.min -1) r 22 % f the ttal glucse lad [58]; (4) basal HGP declined by 53 %. Remarkably similar percentages have been reprted fr splanchnic glucse uptake (24-29 %) and suppressin f HGP (5-6 %) in nrmal subjects by all ther investigatrs [44, 52, 59-61]. Fur studies [51,59-61] have examined the cntributin f skeletal muscle t the dispsal f ral glucse and the results have varied frm a lw f 26 % f the ingested glucse lad [59] t a high f 56 % [61], with a mean f 45 %. These results emphasize several imprtant differences between ral vs i.v. glucse administratin. Fllwing glucse ingestin: (1) HGP is less cmpletely suppressed; (2) peripheral tissue (primarily muscle) glucse uptake is quantitatively less Table 1. Balance sheet fr the dispsitin f an ral glucse lad in nrmal subjects. Summarized frm reference 57 Glucse Mean SEM Range (g) (g) A. Ingested B. Appearing in peripheral plasma (ral) C. Released by the liver D. Taken up by splanchnic tissues E. Taken up by peripheral tissues 48 _ F. Remaining in the glucse space G. Unrecvered 18 _ H. "Saved" by the liver True net splanchnic balance (D - C) 4 _ Splanchnic "cnservatin" (D + H) Splanchnic verall cntributin t glucse hmestasis (D + H - C)

4 392 R.A. DeFrnz: Pathgenesis f Type 2 diabetes Table 2. Summary f glucse metablism fllwing glucse ingestin in Type 2 (nn-insulin-dependent) diabetic patients. The number f arrws indicates the magnitude f change Authr Ref. Test Bdy Insulin Splanchnic Suppressin Tissue Incremental Glucse Frearm Incremental weight respnse a glucse f HGP Rd tissue clearance glucse frearm uptake Rd uptake glucse uptake Ferrannini 62 OGTT nrmal weight sl$ N $ $$ $.1, $$ - - Firth 44 OGTF Obese $ sl$ $$ 1" $$ $$ N sl$ Mitraku 52 OGTT Obese N, delayed sl$ $$ N $ $$ N $ Firth 64 MM Obese N sl$ $$ sl1" $$ $$ N $ McMahn 63 MM Obese $?$ $$? $$ $$ $ $ a In all studies the plasma insulin respnse was deficient when viewed relative t the plasma glucse cncentratin OGTT, rai glucse tlerance test; MM, mixed meal; N, nrmal; sl, slight; HGR hepatic glucse prductin; Rd, glucse dispsal imprtant; (3) splanchnic glucse uptake is quantitatively much mre significant. Five studies have examined the dispsitin f an ral glucse lad in Type 2 diabetes [44, 52, 62-64] and all have revealed very similar results (Table 2). As riginally demnstrated by Ferrannini et al. [62], the disturbance in glucse metablism in Type 2 diabetic patients is accunted fr by tw factrs: (1) decreased tissue glucse uptake (44 vs 6 ver 3.5 h), (2) impaired suppressin f HGP (17 vs 1 ver 3.5 h). Splanchnic glucse uptake was similar in diabetic and cntrl grups. Thus, inapprpriate suppressin f HGP (79 ver 3.5 h) accunts fr apprximately ne-third f the defect in ttal bdy glucse hmestasis, while reduced peripheral, presumably muscle, glucse uptake (14 g ver 3.5 h) accunts fr the remaining tw-thirds. Essentially identical results have been reprted by thers [44, 52, 63, 64], including Gerich and clleagues [52]. It shuld be nted that the duble tracer technique is assciated with significant variability since it invlves the subtractin f tw large numbers (Ra f 3Hglucse minus Ra f ~4C-glucse) t calculate suppressin f HGR Therefre, small differences in suppressin f HGP between labratries are likely t have little physilgic meaning. The inherent variability f the methd, variatin in patients characteristics and differences in insulin secretry respnse can easily explain the 1-2 % differences in suppressin f HGP reprted by varius investigatrs [44, 52, 62-64]. The imprtant message is that everyne has fund that suppressin f HPG is impaired in Type 2 diabetes and that this defect can accunt fr abut ne-third t ne-half f the disturbance in whle bdy glucse hmestasis. Splanchnic glucse uptake has been reprted t be nrmal [62], slightly decreased [44, 52, 64], r increased [63] and des nt appear t cntribute significantly t the impairment in ral glucse tlerance. Special cmment is warranted cncerning whle bdy tissue glucse dispsal fllwing glucse ingestin. In abslute terms mst, but nt all [62], studies have shwn it t be nrmal [52] r slightly increased [44, 63, 64] (Table 2). Hwever, the efficiency f glucse dispsal, i.e. the glucse clearance, is severely reduced. Mst imprtantly, it is nt the abslute glucse dispsal rate, but rather the increment in glucse dispsal abve baseline which determines the rise in plasma glucse abve its fasting value. In every published study [44, 52, 62-64] the incremental respnse in whle bdy glucse uptake was mderately t severely reduced in Type 2 diabetes (Table 2). Similar results have been reprted fr frearm muscle glucse uptake [44, 52, 63, 64] (Table 2). Thus, all published results [44, 52, 62-64], including thse by Gerich and clleagues [52], are ttally cnsistent and pint ut the imprtant cntributin f impaired muscle glucse dispsal in Type 2 diabetes. The cnclusin by Gerich [65] that muscle des nt play an imprtant rle in impaired ral glucse tlerance in Type 2 diabetes stems frm the failure t recgnize that it is the incremental increase in muscle glucse dispsal that determines the rise in plasma glucse cncentratin abve baseline. In summary, results f OGTT indicate that bth impaired suppressin f hepatic glucse prductin and decreased tissue, muscle, glucse uptake cntribute apprximately equally t the glucse intlerance f Type 2 diabetes. Hwever, nne f the currently available studies [44, 52, 62-64, 66] allws ne t define whether the defects in hepatic and peripheral (muscle) glucse metablism are the result f insulin resistance, diminished insulin secretin, r an impairment in the mass actin effect f glucse (i. e. glucse resistance) t prmte its wn uptake. Fasting hyperglycaemia in Type 2 diabetes: pancreas vs muscle vs liver Once vert fasting hyperglycaemia (>7.8mml/l= 14 mg/dl) has develped, HGP is elevated in abslute terms and crrelates clsely with the severity f fasting hyperglycaemia [1,2, 4, 3, 33, 34, 36-42, 45]. These studies d nt establish whether, in the earliest stages f Type 2 diabetes, fasting hyperglycaemia results frm excessive HGR decreased efficiency f tissue glucse uptake, r sme cmbinatin f the tw. Recently, this questin has been addressed by DeFrnz et al. [7] in 77 lean Type 2 diabetic patients. In 33 diabetic patients with fasting glucse cncentratin 7.8 mml/l (Fig. 3, shaded area), HGP ( mg.kg-1.min -1) was virtually identical t cntrl subjects ( mg. kg- 1. min- z). Hwever, this "nrmal" basal rate f HGP was maintained at the expense f a tw-fld greater fasting plasma insulin cncert-

5 R. A. DeFrnz: Pathgenesis f Type 2 diabetes 2,5 ~ I ~-~ 1,9 ~ D~ gl5 I O ) ~ (~ 13 i,l ~i ::::~;;:. ~ I I I f I::~:.~ i::~ I, I I I I i I 5 I 15 2 Fasting plasma glucse (mg/dl),~, q ~, I 25 3 Fig. 4. Summary f the metablic clearance rate f glucse in 77 nrmal weight Type 2 (nn-insulin-dependent) diabetic subjects (pen circles) with fasting plasma glucse cncentratins ranging frm 5.8 t 16.7 mml/1 (15 t 3 mg/dl). Seventy-tw age- and weightmatched cntrl subjects are shwn by the pen circles. In the 33 diabetic subjects with fasting plasma glucse levels less than 7.8 mml/1 (14 mg/dl) (shaded area), the glucse clearance rate fell precipitusly and was inversely crrelated (r = -.697, p <.1) with the increase in plasma glucse levels. At fasting plasma glucse levels abve 7.8 mml/1 (14 mg/dl), the rate f decline in glucse clearance began t slw and reached a plateau at glucse levels abve 1 mml/1 (18 mg/dl). Reprduced frm reference 1 with permissin tratin (2 + 2 vs gu/ml, p <.1) (Fig. 1). These bservatins are underscred by the recent reprts by Erikssn et al. [21] and Gulli et al. [67] and indicate that hepatic insulin resistance is well-established early in the curse f Type 2 diabetes. Only the reprt by Gerich [65] has suggested that HOP may be increased in patients with IGT r mild Type 2 diabetes. Hwever, the number f patients in this study [65] is very small, patient characteristics are prly described, the patient ppulatin is bese, treatment regimen is nt defined, fasting plasma insulin cncentratin is nt prvided, and OGTT results are nt given. Mst imprtantly, in six f ten individuals with fasting glucse levels between 6-7 mml/1 ( mg/dl) (see Fig. 1 f ref. 65), HGP was clearly within the nrmal range and the mean HGP ( ~tml. kg- 1 min- 1) was nt significantly (t= 1.42; p =.18; calculated frm the data presented in Fig. 1 frm ref. 65) elevated cmpared t 19 cntrl subjects ( pml.kg -1. min 1) with fasting glucse between 5-6 mml/1 (9-18 mg/dl). Thus, even the data f Gerich d nt supprt his cntentin [65]. Mrever, the earlier publicatin by Gerich et al. [68] t which he refers did nt examine diabetic patients with fasting glucse levels between 6-7 mml/1. Lastly, the besity index f the diabetic (BMI = kg/m 2) patients in the earlier Gerich's publicatin [68] was much greater than in the mre recent article (BMI = kg/m2), and it is bvius that the patient ppulatins in these tw papers [65, 68] were quite different. Therefre, it is nt apprpriate t equate the metablic disturbances described in the earlier paper [68] with the patient ppulatin described in the mre recent publicatin [65]. In summary neither the previus [68] nr present [65] publicatins justify the statement that in "individuals with fasting plasma glucse cncentratins between 6 and 7 mml/1, rates f glucse prductin are elevated" [65]. Rather, the data f Gerich et al. [65] demnstrate that HGP is nt significantly elevated in the early stages f Type 2 diabetes. Similar results have been reprted in a Type 2 diabetic primate mdel [14,15]. During the pstabsrptive state glucse uptake by all tissues equals HGP and, when viewed in abslute terms, is increased in Type 2 diabetes. Hwever, since bth fasting glucse and insulin cncentratins are elevated, efficiency f tissue glucse uptake, i.e. glucse clearance rate, is markedly reduced early in the curse f Type 2 diabetes [7] (Fig.4). As the fasting glucse rises frm 5.8 t 7.8 mml/1 (15 t 14 mg/dl), glucse clearance declines linearly (r =.7, p <.1) whereas HGP remains cnstant (Fig. 3). Hwever, with fasting glucse levels abve 7.8 mml/1 (14 mg/dl), the restraining effect f hyperinsulinaemia n the liver is lst and HGP increases prgressively (r =.847, p <.1) (Fig. 3), while the decline in whle bdy glucse clearance plateaus at glucse levels between mml/1 (14-18 mg/dl) (Fig. 4). Why is fasting hyperinsulinaemia sufficient t prevent excessive HGP (Fig. 3), yet inadequate t maintain a nrmal basal rate f tissue glucse clearance (Fig. 4) in diabetics with fasting glucse < 7.8 mml/l. This paradx is explained by the distinctive dse-respnse relatinships between plasma insulin cncentratin vs hepatic glucse prductin and tissue glucse dispsal [29, 33-35]. Small increments in plasma insulin (8 t 27 gu/ml) in nrmal subjects, suppress HGP by 68 %, but have n stimulatry effect n whle bdy glucse uptake [33]. Thus, in Type 2 diabetic subjects fasting hyperinsulinaemia is sufficient t ffset the hepatic insulin resistance but des nt stimulate tissue glucse uptake [29, 33, 69, 7], resulting in a decline in basal glucse clearance (Fig. 4). In Type 2 diabetic subjects DeFrnz et al. [4] have shwn that basal leg (muscle) glucse clearance is similar t cntrl subjects. In cntrast, Gerich et al. [68] and Firth et al. [44] have reprted mdestly reduced rates f glucse clearance by frearm (muscle) tissue in Type 2 diabetes in the pstabsrptive state. Using hepatic vein catheterizatin, DeFrnz et al. have demnstrated that part f the decrease in whle bdy glucse clearance resides within the splanchnic tissues (liver plus gastrintestinal) [4]. Hwever, frm a purely quantitative standpint tissues in additin t liver [4] and muscle [44, 68] must als cntribute t the decline in glucse clearance. The brain represents a prime candidate t explain this decrease. In the pstabsrptive state, 5-6 % f glucse dispsal ccurs in cerebral tissues. Brain glucse uptake is insulin-independent, saturates at plasma glucse cncentratins f -2.2 mml/1 (4 mg/dl), and remains nrmal in Type 2 diabetic subjects despite fasting hyperglycaemia [58]. It fllws, therefre, that brain (and cnsequently whle bdy) glucse clearance must decline in Type 2 diabetic subjects with prgressive increases in basal glucse cncentratin. It remains unknwn whether tissues, in additin t muscle, splanchnic and brain, als cntribute t the 393

6 394 Plasma glucse Plasma insulin ~. } '~ f 9O 15 ~= Hepatic glucse prductin Glucse clearance 1.9 ~ 2. a_" E ~ E Fig.5. Effect f vernight insulin infusin t nrmalize basal rate f hepatic glucse prductin in 19 nrmal weight Type 2 (nn-insulindependent) diabetic subjects (shaded bars) and in 72 age- and weight-matched cntrl subjects (crss-hatched bars). Despite similar rates f hepatic glucse prductin and a tw-fld greater increase in plasma insulin cncentratin (p <.1), fasting plasma glucse remained significantly elevated in Type 2 diabetic patients vs cntrl subjects. The decreased glucse clearance in Type 2 diabetic patients indicates a diminished efficiency f tissue glucse uptake. Reprduced frm reference 7 with permissin decline in basal glucse clearance in Type 2 diabetic subjects. The defects in basal HGP and tissue glucse dispsal are best appreciated by studying Type 2 diabetic subjects after vernight insulin infusin t nrmalize basal HGP [7] (Fig. 5). Despite similar rates f HGP in cntrl and diabetic subjects and a tw-fld greater plasma insulin cncentratin in the latter, fasting glucse remained elevated in diabetic individuals ( vs mml/1 = vs mg/dl, p <.1). The persistent hyperglycaemia is explained by a decreased efficiency f tissue glucse remval (glucse clearance=l.71 vs 2. ml. kg- 1. min- 1, p <.1). Further evidence fr impaired tissue glucse uptake cmes frm studies in which Type 2 diabetic subjects received vernight insulin infusin t nrmalize fasting plasma glucse cncentratin [7] (Fig. 6). At identical plasma glucse cncentratins, the abslute rate f tissue glucse uptake was significantly reduced in the Type 2 diabetic subjects, even thugh the plasma insulin cncentratin was mre than tw-fld elevated in the frmer grup. These data unequivcally demnstrate that tissue glucse dispsal is impaired in Type 2 diabetes. In cntrast t the suggestin f Gerich et al. [65, 68], these results cannt be explained simply by failure f the brain t passively enhance its uptake f glucse in respnse t a prgressive rise in fasting plasma glucse cncentratin. Frm currently available data it is nt pssible t establish which defect, i.e. hepatic insulin resistance r decreased efficiency f tissue glucse remval, develps first in the evlutin f fasting hyperglycaemia in Type 2 diabetes mellitus. Three equally plausible sequences can R. A. DeFrnz: Pathgenesis f Type 2 diabetes be pstulated. First, bth defects develp in parallel. As hyperglycaemia ensues (due bth t excessive HGP and decreased tissue glucse uptake), basal insulin secretin is stimulated (Fig. 1). The resultant hyperinsulinaemia restres HGP t baseline, while hyperglycaemia returns tissue glucse uptake t nrmal. This sequence wuld result in nrmal basal rates f HGP and tissue glucse uptake but at the expense f fasting hyperglycaemia and fasting hyperinsulinaemia [1, 7]. Since the small increment in plasma insulin has n stimulatry effect n tissue glucse uptake, whle bdy glucse clearance falls. Secnd, decreased efficiency f tissue glucse uptake culd represent the primary defect. As fasting plasma glucse rises, insulin secretin is enhanced; the resultant hyperglycaemia returns tissue glucse uptake t nrmal, while hyperinsulinaemia has tw ppsing actins: (1) inductin f hepatic insulin resistance by dwn-regulating bth receptr and pst-receptr events [71, 72]; (2) suppressin f HGR Because these tw metablic actins f insulin ffset each ther, basal HGP remains unaltered. Third, hepatic insulin resistance culd initiate fasting hyperglycaemia by causing a small, imperceptible rise in plasma glucse. This wuld stimulate insulin secretin, returning HGP t nrmal. ]?issue glucse uptake wuld remain unaltered because the small rise in plasma insulin is insufficient t augment tissue glucse uptake [29, 33, 69, 7]. Cnsequently, tissue glucse clearance wuld fall. It shuld be nted that the statement by Gerich [65] that "fr plasma glucse t increase, glucse prductin must exceed glucse uptake" is nt, strictly speaking, crrect. Fr plasma glucse t increase, glucse prductin must transiently exceed glucse uptake r tissue glucse uptake must transiently decrease belw the rate f glucse prductin. Studies perfrmed under steady cnditins after a metablic perturbatin has ccurred and the system has [ 9 :2t =- 1.8 ~ 1.7 v Plasma glucse Glucse dispsal Plasma insulin Glucse clearance '~ Fig.6. Effect f vernight insulin infusin t nrmalize the fasting plasma glucse cncentratin in 11 nrmal weight Type 2 (nninsulin-dependent) diabetic patients (shaded bars) and in 72 ageand weight-matched cntrl subjects (crss-hatched bars). Despite identical plasma insulin cncentratins the abslute rate f whle bdy tissue glucse uptake in the pst-absrptive state was significantly reduced in the Type 2 diabetic grup (p <.1). Reprduced frm reference 7 with permissin

7 R. A. DeFrnz: Pathgenesis f Type 2 diabetes re-equilibrated cannt recnstruct the sequence f events that led t the establishment f the new steady state. In summary, in Type 2 diabetic subjects with mild fasting hypergtycaemia ( < 7.8 mml/1 = 14 mg/dl) bth decreased efficiency f tissue glucse uptake and hepatic insulin resistance cntribute t the elevated pstabsrptive plasma glucse cncentratin. Dynamic interactin between insulin actin and insulin secretin in Type 2 diabetes Type 2 diabetic subjects are characterized by bth tissue (muscle and liver) insulin resistance and impaired insulin secretin. T fully appreciate the evlutin f the fullblwn diabetic cnditin, it is necessary t examine the dynamic interactin between insulin actin and insulin secretin in the same individual ver a wide range f insulin sensitivity. Three grups have prvided such infrmatin [1, 8, 9, 12, 13, 28, 4, 41, 73]. In besity (Fig, 2) DeFrnz, Felber and clleagues [1, 13, 41, 73] have shwn that weight gain is assciated with a marked reductin (4-5 % ) in insulin sensitivity. Nnetheless, glucse tlerance remains nrmal because the Beta cell apprpriately augments its insulin secretry capacity t ffset the insulin resistance. Prgressin frm nrmal t IGT is assciated with further reductin in insulin sensitivity. Hwever, glucse tlerance is nly mildly impaired because f a further cmpensatry increase in insulin secretin. Onset f vert diabetes is heralded by a mdest decline in insulin secretin withut any additinal wrsening f the insulin resistance. This mdest decline in insulin secretin, in the presence f severe insulin resistance, results in frank diabetes (Fig.2). Prgressin frm mderate t severe diabetes is assciated with a further reductin in insulin secretin withut any change in insulin sensitivity (Fig. 2). These bservatins underscre the critical interactin between insulin resistance and insulin secretin in the develpment f vert Type 2 diabetes. The sequence f events described abve has been cnfirmed in Pima Indians [8, 9, 12], lean Caucasians [13, 34], and mnkeys [14, 15]. In summary, in the earliest stage f Type 2 diabetes bth hepatic and peripheral tissue resistance t insulin is well-established and is ffset by the presence f cmpensatry hyperinsulinaemia. Overt diabetes develps nly in individuals whse pancreas is unable t meet the increased and sustained demand fr insulin secretin. Acknwledgements. The wrk reviewed in this paper was supprted by NIH grant DK 2492, a VA Merit Award, GCRC grant # M1- RR-1346, the GRECC, and the VA General Research Service. Ms. S. Cnter and Ms. L. Olivarri prvided expert secretarial assistance in the preparatin f the manuscript. References 1. DeFrnz RA (1988) Lilly Lecture. The triumvirate: beta cell, muscle, liver. A cllusin respnsible fr NIDDM. Diabetes 37: DeFrnz RA, Ferrannini E (1987) Regulatin f hepatic glucse metablism in humans. Diab Metab Rev 3: DeFrnz RA, Jact E, Jequier E, Maeder E, Wahren J, Felber JP (1981) The effect f insulin n the dispsal f intravenus glucse: results frm indirect calrimetry. Diabetes 3: DeFrnz RA, Gunnarssn R, Bjrkman O, Olssn M, Wahren J (1985) Effects f insulin n peripheral and splanchnic glucse metablism in nn-insulin dependent diabetes mellitus. J Clin Invest 76: Faber OK, Damsgaard EM (1984) Insulin secretin in type II diabetes. Acta Endcrin1262 [Suppl]: Faber OK, Markussen J, Nalthani VK et al. (1978) Kinetics f human cnnecting peptide in nrmal and diabetic subjects. J Clin Invest 62: DeFrnz RA, Ferrannini E, Simnsn DC (1989) Fasting hyperglycemia in nn-insulin-dependent diabetes mellitus: cntributins f excessive hepatic glucse prductin and impaired tissue glucse uptake. Metablism 38: Saad MF, Knwler WC, Pettitt D J, Nelsn RG, Mtt DM, Bennett PH (1989) Sequential changes in serum insulin cncentratin during develpment f nn-insulin-dependent diabetes. Lancet I: Lillija S, Mrt DM, Hward BV (1988) Impaired glucse tlerance as a disrder f insulin actin. Lngitudinal and crss-sectinal studies in Pima Indians. N Engl J Med 318: Wan-am JH, Martin BC, Gleasn RE, Seldner JS (1987) Slw glucse remval rate but nt insulin secretin predicts develpment f NIDDM in ffspring f tw NIDDM parents. Diabetes 36 [Suppl 1]: 14A (Abstract) 11. Yudkin JS, Alberti KGMM, McClarity DG, Swai ABM (199) Impaired glucse tlerance-is it a risk factr fr diabetes r a diagnstic ragbag? BMJ 31: Saad MF, Knwler WC, Pettitt D J, Nelsn RG, Mtt DM, Bennett PH (1988) The natural histry f impaired glucse tlerance in the Pima Indians. N Engl J Med 319: Felber JP, Jallut D, Glay A, Munger R, Frascarl R Jequier E (1989) Obesity t diabetes. A lngitudinal study f glucse metablism in man. Diabetes 2 [Suppl 1] 221 A (Abstract) 14. Hansen BC, Bdkin NH (1986) Hetergenity f insulin respnses: phases leading t Type2 (nn-insulin-dependent) diabetes mellitus in the rhesus mnkey. Diabetlgia 29: Bdkin NL, Metzger BL, Hansen BC (1989) Hepatic glucse prductin and insulin sensitivity preceding diabetes in mnkeys. Am J Physi1256:E676 E Sicree RA, Zimmet R King HO, Cventry JO (1987) Plasma insulin respnse amng Nauruans. Predictin f deteriratin in glucse tlerance ver 6 years. Diabetes 36: Haffner SM, Stern MR Mitchell BD, Hazula HE Pattersn JK (199) Incidence f type II diabetes in Mexican Americans predicted by fasting insulin and glucse levels, besity, and bdy-fat distributin. Diabetes 39: Haffner SM, Stern MR Hazula HE Mitchell BD, Pattersn JK (1988) Increased insulin cncentratins in nn-diabetic ffspring f diabetic parents. N Engl J Med 319: Balkan B, King H, Zimmet R Raper LR (1985) Factrs assciated with the develpment f diabetes in the Micrnesian ppulatin f Nauru. Am J Epidemi1122: Byk E J, Keane EM, Marshall JA, Hamman RF (1991) Higher insulin and C-peptide cncentratins in Hispanic ppulatin at high risk fr NIDDM. San Luis Valley Diabetes Study. Diabetes 4: Erikssn J, Franssila-Kallunki A, Ekstrand A et at. (1989) Early metablic defects in persns at increased risk fr nn-insulin-dependent diabetes mellitus. N Engl J Med 321: H LT, Chang ZY, Wang JT et al. (199) Insulin insensitivityin ffspring f parents with type 2 diabetes mellitus. Diab Med 7: Westermark R Wilander E (1978) The influence f amylid depsits n the islet vlume in maturity-nset diabetes mellitus. Diabetlgia 15: Stefan Y, Orci L, Malalsse-Lagae E Perrelet A, Patet Y, Unger R (1982) Quantitatin f endcrine cell cntent in the pancreas f nn-diabetic and diabetic humans. Diabetes 31:

8 Steiner DE Tager HS, Chan S J, Nanj T, Sanke T, Rubenstein AH (199) Lessns learned frm mlecular bilgy f insulingene mutatins. Diab Care 13: Rssetti L, Giaccari A, DeFrnz RA (199) Glucse txicity. Diab Care: 13: Leahy JL (199) Natural histry f beta cell dysfunctin in nninsulin-dependent diabetes mellitus. Diab Care 13: Reaven GM, Hllenbeck CB, Chen YDI (1989) Relatinship between glucse tlerance, insulin secretin, and insulin actin in nn-bese individuals with varying degrees f glucse tlerance. Diabetlgia 32: Bnadnna RC, Grp L, Kraemer N, Ferrannini E, Del Prat S, DeFrnz RA (199) Obesity and insulin resistance in man. A dse respnse study. Metablism 39: Bgardus C, Lillija S, Hward BV, Reaven G, Mtt D (1984) Relatinships between insulin secretin, insulin actin, and fasting plasma glucse cncentratin in nn-diabetic and nn-insulin-dependent subjects. J Clin Invest 74: Reaven GM, Miller R (1968) Study f the relatinship between glucse and insulin respnses t an ral glucse lad in man. Diabetes 17: Zimmet R Whitehuse S, Alfrd F, Chishlm D (1978) The relatinship f insulin respnse t a glucse stimulus ver a wide range f glucse tlerance. Diabetlgia 15: Grp LC, Bnadnna RC, Del Prat S (1989) Glucse and free fatty acid metablism in nn-insulin dependent diabetes mellitus. Evidence fr multiple sites f insulin resistance. J Clin Invest 84: Firth R, Bell R Rizza R (1987) Insulin actin in nn-insulin-dependent diabetes mellitus: the relatinship between hepatic and extrahepatic insulin resistance and besity. Metablism 36: Campbell P J, Mandarin LJ, Gerich JE (1988) Quantificatin f the relative impairment in actins f insulin n hepatic glucse prductin and peripheral glucse uptake in nn-insulin-dependent diabetes mellitus. Metablism 37: DeFrnz RA, Ferrannini E, Kivist V (1983) New cncepts in the pathgenesis and treatment f nn-insulin-dependent diabetes mellitus. Am J Med 75 (1 A): D efrnz RA, Deibert D, Hendler R, Felig P (1982) Insulin sensitivity and insulin binding t mncytes in maturity-nset diabetes. J Clin Invest 63:93% DeFrnz RA, Simnsn D, Ferrannini E (1982) Hepatic and peripheral insulin resistance: a cmmn feature in nn-insulindependent and insulin dependent diabetes. Diabetlgia 23: Simnsn D, Ferrannini E, B evilacqua Set al. (1984) Mechanism f imprvement in glucse metablism fllwing chrnic glyburide therapy. Diabetes 33: Glay A, DeFrnz RA, Ferrannini E et al. (1988) Oxidative and nn-xidative glucse metablism in nn-bese Type 2 (nn-insulin-dependent) diabetic patients. Diabetlgia 31: Glay A, Felber JR Jequier E, DeFrnz RA, Ferrannini E (1988) Metablic basis f besity and nninsulin-dependent diabetes mellitus. Diab Metab Rev 4: Henry RR, Wallace R Olefsky JM (1986) Effects f weight lss n mechanisms f hyperglycemia in bese nn-insulin dependent diabetes mellitus. Diabetes 35: Best JD, Judzewitsch RG, Pfeiffer MA, Beard JC, Halter JB, Prte D (1982) The effect f chrnic sulfnyl urea therapy n hepatic glucse prductin in nn-insulin-dependent diabetes mellitus. Diabetes 31: Firth RG, Bell PM, Marsh HM, Hansen I, Rizza RA (1986) Pstprandial hyperglycemia in patients with nn-insulin-dependent diabetes mellitus. J Clin Invest 77: Garvey WT, Olefsky JM, Griffin J, Hamman RF, Klterman OG (1985) The effect f insulin treatment n insulin actin in type II diabetes mellitus. Diabetes 34: DeFrnz RA, Ferrannini E, Hendler R, Felig R Wahren J (1983) Regulatin f splanchnic and peripheral glucse uptake by insulin and hyperglycemia. Diabetes 32:35-45 R. A. D efrnz: Pathgenesis f Type 2 diabetes 47. Butterfield WJH, Whichelw MJ (1965) Peripheral glucse metablism in cntrl subjects and diabetic patients during glucse, glucse-insulin, and insulin sensitivity tests. Diabetlgia 1: Revers R, Fink R, Griffin J, Olefsky J, Klterman O (1984) Influence f hyperglycemia in insulin's in viv effects in type II diabetes. J Clin Invest 73: Kalant N, Leibvici D, Fukushima J, Ozaki S (1982) Insulin respnsiveness f superficial frearm tissues in Type 2 (nn-insulin-dependent) diabetes. Diabetlgia 22: Campbell R Mandarin L, Gerich J (1988) Quantificatin f the relative impairment in actins f insulin n hepatic glucse prductin and peripheral glucse intake in nn-insulin-dependent diabetes mellitus. Metablism 37: Jacksn RA, Perry G, Rgers J, Advni U, Pilkingtn TRE (1973) Relatinship between the basal glucse cncentratin, glucse tlerance, and frearm glucse uptake in maturity nset diabetes. Diabetes 22: Mitraku A, Kelley D, Veneman T et al. (199) Cntributin f abnrmal muscle and liver glucse metablism t pstprandial hyperglycemia in NIDDM. Diabetes 39: Butterfield WJH,Whichelw MJ (1965) Peripheral glucse metablism in cntrl subjects and diabetic patients during glucse, glucse-insulin and insulin sensitivity tests. Diabetlgia 1: Zierler KL, Rabinwitz D (1963) Rles f insulin and grwth hrmne, based n studies f frearm metablism in man. Medicine 42: Capald B, Napli R, Dimarin L, Picardi A, Riccardi G, Sacca L (1988) Quantitatin f frearm glucse and free fatty acid (FFA) dispsal in nrmal subjects and type 2 diabetic patients: evidence against an essential rle fr FFA in the pathgenesis f insulin resistance. J Clin Endcrinl Metab 67: Marin R Rebuffe-Scrive, Smith U, Bjrntrp P (1987) Glucse uptake in human adipse tissue. Metablism 36: Ferrannini E, Reichard GA, Bjrkman O et al. (1985) The dispsal f an ral glucse lad in nrmal subjects. A quantitative study. Diabetes 34: Grill V (199) A cmparisn f brain glucse metablism in diabetes as measured by psitrn emissin tmgraphy r by arterivenus techniques. Ann Med 22: Kelley D, Mitraku A, Marsh H et al.: Skeletal muscle glyclysis, xidatin, and strage f an ral glucse lad. J Clin Invest 81: Jacksn RA, Rshania RD, Hawa MI, Sire BM, DiSilvi L (1986) Impact f glucse ingestin n hepatic and peripheral glucse metablism in man: an analysis based n simultaneus use f the frearm and duble istpe techniques. J Clin Endcrinl Metab 63: Katz LD, Glickman MG, Rapprt S, Ferrannini E, DeFrnz RA (1983) Splanchnic and peripheral dispsal f ral glucse in man. Diabetes 32: Ferrannini E, Simnsn DC, Katz LD et al. (1988) The dispsal f an ral glucse lad in patients with nn-insulin dependent diabetes. Metablism 37: McMahn V, Marsh HM, Rizza RA (1989) Effects f basal insulin supplementatin n dispsitin f mixed meal in bese patients with NIDDM. Diabetes 38: Firth RG, Bell PM, Rizza RA (1986) Effects f tlazamide and exgenus insulin n insulin actin in patients with nn-insulindependent diabetes mellitus. N Engl J Med 314: Gerich JE (1991) Is muscle the majr site f insulin resistance in Type 2 (nn-insulin-dependent) diabetes mellitus? Diabetlgia 34: Chen Y-DI, Jeng CY, Hllenbeck CB, Wu MS, Reaven GM (1988) Relatinship between plasma glucse and insulin cncentratin, glucse prductin, and glucse dispsal in nrmal subjects and patients with nn-insulin-dependent diabetes. J Clin Invest 82: Gulli G, Haffner S, Ferrannini E, DeFrnz RA (199) What is inherited in NIDDM? Diabetes 39 [Suppl 1]: 116 A (Abstract)

9 R. A. DeFrnz: Pathgenesis f Type 2 diabetes 68. Gerich JE, Mitraku A, Kelly D et al. (199) Cntributin f impaired muscle glucse clearance t reduced pstabsrptive systemic glucse clearance in NIDDM. Diabetes 39: Zierler KL, Rabinwitz D (1964) Effect f very small cncentratins f insulin n frearm metablism. Persistence f its actin n ptassium and free fatty acids withut its side effect n glucse. J Clin Invest 43: Natali A, Santr D, Palmb C, Cerri M, Ghine S, Ferrannini E (1991) Impaired insulin actin n skeletal muscle metablism in essential hypertensin. Hypertensin 17: Marshall S, Olefsky J (198) Effect f insulin incubatin n insulin binding, glucse transprt, and insulin degradatin by islated adipcytes. Evidence f hrmne-induced desensitizatin at the receptr and pst-receptr level. J Clin Invest 66: Amatruda JM, Newmeyer HW, Chang CL (1982) Insulin-induced alteratins in insulin binding and insulin actin in primary cultures f rat hepatcytes. Diabetes 31: Felber JR Ferrannini E, Glay A (1987) Rle f lipid xidatin in the pathgenesis f the insulin resistance f besity and type II diabetes. Diabetes 36: Prf. R. A. D efrnz Divisin f Diabetes The University f Texas Health Science Center at San Antni 773 Flyd Curl Drive San Antni, Texas USA