Type 2 Diabetes

1 Type 2 DiabetesEtiology and reversibilityROYTAYLOR,MD,FRCPR eversal of type 2 Diabetes to normalmetabolic control by either bariat-ric surgery or hypocaloric diet allowsfor the time sequence of underlying path-ophysiologic mechanisms to be reverse order, the same mechanisms arelikely to determine the events leading to theonset of hyperglycemia and permit insightinto the etiology of type 2 Diabetes . Within7 days of instituting a substantial negativecalorie balance by either dietary interven-tion or bariatric surgery, fasting plasmaglucose levels can normalize. This rapidchange relates to a substantial fall in liver fatcontent and return of normal hepatic in-sulin sensitivity. Over 8 weeks,first phaseand maximal rates of insulin secretionsteadily return to normal, and this changeis in step with steadily decreasing pancre-atic fat content. The difference in timecourse of these two processes is information on the intracellulareffects of excess lipid intermediaries ex-plains the likely biochemical basis,which simplifies both the basic under-standing of the condition and the con-cepts used to determine appropriatemanagement.

2 Recent large, long-durationpopulation studies on time courseof plasma glucose and insulin secretionbefore the diagnosis of Diabetes are con-sistent with this new 2 Diabetes has long been regardedas inevitably progressive, requiring in-creasing numbers of oral hypoglycemicagents and eventually insulin, but it isnow certain that the disease process canbe halted with restoration of normalcarbohydrate and fat metabolism. Type2 Diabetes can be understood as a poten-tially reversible metabolic state precipi-tated by the single cause of chronicexcess intraorgan 2 Diabetes has long been knownto progress despite glucose-loweringtreatment, with 50% of individuals re-quiring insulin therapy within 10 years(1). This seemingly inexorable deteriora-tion in control has been interpreted tomean that the condition is treatable butnot curable. Clinical guidelines recognizethis deterioration with algorithms of se-quential addition of therapies.

3 Insulin re-sistance andb-cell dysfunction areknown to be the major pathophysiologicfactors driving type 2 Diabetes ; however,these factors come into play with very dif-ferent time courses. Insulin resistance inmuscle is the earliest detectable abnor-mality of type 2 Diabetes (2). In contrast,changes in insulin secretion determineboth the onset of hyperglycemia and theprogression toward insulin therapy (3,4).The etiology of each of these two majorfactors appears to be distinct. Insulin re-sistance may be caused by an insulin sig-naling defect (5), glucose transporterdefect (6), or lipotoxicity (7), andb-celldysfunction is postulated to be caused byamyloid deposition in the islets (8), oxi-dative stress (9), excess fatty acid (10), orlack of incretin effect (11). The demon-stration of reversibility of type 2 diabetesoffers the opportunity to evaluate the timesequence of pathophysiologic events dur-ing return to normal glucose metabolismand, hence, to unraveling the of type 2 diabetesby bariatric surgerydThefirsthint that type 2 Diabetes is a fully re-versible syndrome came from bariatricsurgery.

4 Almost a quarter century ago,Pories et al. (12) demonstrated that bloodglucose levels normalized in obese peoplewith type 2 Diabetes undergoing bariatricsurgery and that 10 years later, almost90% remained free of Diabetes . The phe-nomenon was more recently tested in arandomized prospective study comparinggastric banding with intensive medicaltherapy for type 2 Diabetes (13). This leastinvasive type of surgery was most suitablefor the randomized study, although it wasassociated with lower rates of Diabetes re-versal than other procedures. Mean fast-ing plasma glucose fell to normal levels inthe surgically treated group but declinedonly modestly in the intensive medicaltreatment group despite oral agents andinsulin (Fig. 1) (13). Remission of diabe-tes was related to the degree of weight lossrather than to group allocation and wasachieved in 73% of the surgical group and13% of the intensive medical treatmentgroup because surgery was more effectivein achieving weight loss as previouslyFigure 1dA: Fasting plasma glucose andweight change 2 years after randomizationeither to gastric banding or to intensive medi-cal therapy for weight loss and glucose plotted with permission from Dixon et al.

5 (13).B: Early changes in fasting plasma glu-cose level following pancreatoduodenal bypasssurgery. A decrease into the normal range wasseen within 7 days. Reproduced with permis-sion from Taylor (98).ccccccccccccccccccccccccccccccccccc ccccccccccccccFrom the Magnetic Resonance Centre, Institute of Cellular Medicine, Newcastle University, Newcastle uponTyne, author: Roy Taylor, 6 September 2012 and accepted 1 November : 2013 by the American Diabetes Association. Readers may use this article as long as the work is properlycited, the use is educational and not for profit, and the work is not altered. See for ,VOLUME36, APRIL20131047 REVIEW ARTICLE described (14). Type 2 Diabetes can be re-versed by applying a surgical procedurethat diminishes fat , the observation that nor-malization of glucose in type 2 diabetesoccurred within days after bariatric sur-gery, before substantial weight loss (15),led to the widespread belief that surgeryitself brought about specific changes me-diated through incretin hormone secre-tion (16,17).

6 This reasoning overlookedthe major change that follows bariatricsurgery: an acute, profound decrease incalorie intake. Typically, those undergo-ing bariatric surgery have a mean bodyweight of;150 kg (15) and would there-fore require a daily calorie intake of; MJ/day (3,200 kcal/day) forweight maintenance (18). This intake de-creases precipitously at the time of sur-gery. The sudden reversal of traffic intofat stores brings about a profound changein intracellular concentration of fat me-tabolites. It is known that under hypo-caloric conditions, fat is mobilizedfirstfrom the liver and other ectopic sitesrather than from visceral or subcutaneousfat stores (19). This process has beenstudied in detail during more moderatecalorie restriction in type 2 Diabetes over8 weeks (20). Fasting plasma glucose wasshown to be improved because of an 81%decrease in liver fat content and normali-zation of hepatic insulin sensitivity withno change in the insulin resistance of type 2 diabetesby diet alonedIf the rapid changesin metabolism following bariatric surgeryare a consequence of the sudden changein calorie balance, the defects in bothinsulin secretion and hepatic insulin sen-sitivity of type 2 Diabetes should becorrectable by change in diet alone.

7 Totest this hypothesis, a group of peoplewith type 2 Diabetes were studied beforeand during a 600 kcal/day diet (21).Within 7 days, liver fat decreased by30%, becoming similar to that of the con-trol group, and hepatic insulin sensitivitynormalized (Fig. 2). The close associationbetween liver fat content and hepatic glu-cose production had previously been es-tablished (20,22,23). Plasma glucosenormalized by day 7 of the this 8-week study,b-cellfunction was tested by a gold standardmethod that used a stepped glucose infu-sion with subsequent arginine bolus (21).In type 2 Diabetes , the glucose-inducedinitial rapid peak of insulin secretion(thefirst phase insulin response) typicallyis absent. This was confirmed at baselinein the study, but thefirst phase responseincreased gradually over 8 weeks of avery-low-calorie diet to become indistin-guishable from that of age- and weight-matched nondiabetic control maximum insulin response, as elicitedby arginine bolus during hyperglycemia,also normalized.

8 Pancreas fat contentdecreased gradually during the study pe-riod to become the same as that in thecontrol group, a time course matchingthat of the increase in bothfirst phaseand total insulin secretion (Fig. 3). Fatcontent in the islets was not directly mea-sured, although it is known that islets takeup fat avidly (24) and that islet fat contentclosely reflects total pancreatic fat contentin animal models (25). Although a cause-and-effect relationship between raised in-traorgan fat levels and metabolic effecthas not yet been proven, the time coursedata following the dietary interventionstudy are highly suggestive of a causallink (21).New perspectives on insulinresistanceMuscleWhole-body insulin resistance is the ear-liest predictor of type 2 Diabetes onset,and this mainly reflects muscle insulinresistance (26). However, careful separa-tion of the contributions of muscle andliver have shown that early improvementin control of fasting plasma glucose levelis associated only with improvement inliver insulin sensitivity (20,21).

9 It is clearthat the resumption of normal or near-normal diurnal blood glucose controldoes not require improvement in muscleinsulin sensitivity. Although thisfindingmay atfirst appear surprising, it is sup-ported by a wide range of earlier observa-tions. Mice totally lacking in skeletalmuscle insulin receptors do not developdiabetes (27). Humans who have thePPP1R3 Agenetic variant of muscle glyco-gen synthase cannot store glycogen inmuscle after meals but are not necessarilyhyperglycemic (28). Many normoglyce-mic individuals maintain normal bloodglucose levels with a degree of muscle in-sulin resistance identical to those withtype 2 Diabetes (29).Although a defect in mitochondrialfunction is associated with extremes ofinsulin resistance in skeletal muscle (30),this does not appear to be relevant to theetiology of type 2 Diabetes . No defect ispresent in early type 2 Diabetes but ratheris directly related to ambient plasma glu-cose concentration (31).

10 Observed ratesof mitochondrial ATP production can bemodified by increasing or decreasingplasma fatty acid concentration (32,33).Additionally, the onset of insulin stimula-tion of mitochondrial ATP synthesis isslow, gradually increasing over 2 h, andquite distinct from the acute onset of in-sulin s metabolic effects (34). Although itremains possible that secondary mito-chondrial effects of hyperglycemia andexcess fatty acids exist, there is no evi-dence for a primary mitochondrial defectunderlying type 2 physiologic importance of mus-cle insulin resistance is likely to operateover a period of many years. The pres-ence of long-standing muscle insulinresistance will not of itself cause bloodglucose levels to rise, but raised plasmainsulin levels will expedite accumulationFigure 2dEffect of a very-low-calorie diet intype 2 Diabetes on fasting plasma glucose level(A), basal hepatic glucose production (HGP)(B), and hepatic triacylglycerol content (C).