Diabesity

This page was started in the framework of an Eduzendium course and needs to be assessed for quality. If this is done, this {{EZnotice}} can be removed.

Introduction

Diabesity was first coined by Ethan Sims in 1973, to describe the close relationship between T2D (type 2 diabetes) and Obesity (1,2). Their findings suggested that by overfeeding young men, with no previous family history of diabetes, the initial signs of diabetes were induced. This excess consuption led to increases in insulin production, plasma glucose, triglycerides and eventually impaired glucose tolerance; all signs predisposing one to T2D and obesity (1).

Following its coinage over 30 years ago, a very large research project called Diabesity has now been set up and funded by the European Union, involving over 27 institutions from 24 member countries. It aims to identify new genes implicated in obesity and identify several new drug targets for the treatment and prevention of diabesity (4). The University of Edinburgh plays an key role in this project, with Professor Seckl and Professor Leng important collaborators. (Maybe leave this for end? as "European Project - Diabesity")

1. EAH Sims, E Danforth, ES Horton, GA Bray, JA Glennon and LB Salans, Endocrine and metabolic effects of experimental obesity in man, Recent Prog Horm Res 29 (1973), pp. 457–496.

2 "Obesity". Haslam DW, James WP (October 2005). Lancet 366 (9492): pp: 1197–209.

3 Ellenberg & Rifkin's diabetes mellitus Daniel Porte,Robert S. Sherwin,Alain Baron,Max Ellenberg,Harold Rifkin McGraw-Hill Professional; 6 edition (September 23, 2002)

4 http://www.eurodiabesity.org/ </ref>

The Genetics of Diabesity

Thrifty gene hypothesis

Since the 1960s, there has been a gathering of scientific evidence to strongly implicate genetics in the role of obesity. Whilst many hypotheses have been suggested, the thrifty gene hypothesis has captured the attention of the scientific community, and remains a popular explanation for the growing rise of diabetes and obesity. In 1962, Neel proposed the thrifty gene hypothesis to explain the high prevalence of T2D in recently westernized nations. Neel suggested that the “thrifty” genes that predispose to T2D and obesity are evolutionary advantageous (1). He proposed that in the primitive age, where famine and food shortages would be common, individuals with “thrifty genes” were more likely to survive. This is because these individuals could store a high percentage of their energy as fat, and use this energy store to survive times of famine. However, in the modern world there is a constant abundance of food; the thrifty genes prepare us for a famine that never comes, resulting in a huge rise in T2D and obesity rates (2). It is also interesting to observe that those from ethnic populations who had experienced famine and poor nutrition until relatively recently, are shown to have the highest rates in the world (3). For example, research on the Pima Indians has contributed in providing clear evidence that a sudden abundance in food can lead to a substantial rise in obesity rates as well as T2D rates. The research found that 50% of adult Pima Indians have Diabetes, and 95% of these diabetic subjects were overweight. (4)

1. The original Paper that proposes the Thrifty genotype. Neel JV (1962). "Diabetes mellitus: a "thrifty" genotype rendered detrimental by "progress"?". Am. J. Hum. Genet. 14: 353–62. 2. Joffe, Barry; Paul Zimmet. 1998. Thrifty genotype in Type 2 diabetes. And obesity. J "The thrifty genotype in type 2 diabetes". Endocrine 9 (2): 139–141.

3. Relates genetic and environmental factors. Thrifty gene hypothesis summarised

Noel Cameron, Nicholos G Norgan, George Ellison, Childhood Obesity Contemporary Issues 2006, CRC press, Taylor and Francis Group Obesity relates very closely to Diabetes.

Website which describes the study on Pima Indians - rates of Obesity and Diabetes 4. http://diabetes.niddk.nih.gov/DM/pubs/pima/obesity/obesity.htm

Visceral fat accumulation and type 2 diabetes

Changes in lifestyle and behaviour recently have caused a massive increase in the incidence of T2DM . In recent times it has been discovered that excess fat within the abdomen, known as visceral adiposity, creates a superior independent health risk of metabolic complications than accumulation of adipose tissue in other regions. [2] Visceral adiposity is related with an increase in future insulin resistance, whereas abdominal subcutaneous fat is not. [3] The lipoprotein profile related to obesity and insulin resistance is largely due to intra-abdominal fat. [4] There are better measures of obesity and in particular visceral obesity that can predict diabetes. These include Waist Circumference (WC), the Waist to hip ratio (WHR) and insulin resistance (IR).

It is known that T2DM is a heterogeneous disease and that visceral adiposity is at the centre of the problem. Fat cells show elevated hydrolysis of stored triglycerides and increased free fatty acids(FFAs) into the blood. IR reduces the antilipolytic effect of insulin which in turn leads to reduced glucose uptake and increased release of FFAs and glycerol. [5]

Excess FFAs are taken by the portal vein to the liver. The liver is overwhelmed by the FFAs and starts up typical IR metabolic processes. The liver responds by increasing gluconeogenesis, increasing triglyceride (TG), apolipoprotein B and very low density lipoprotein (VLDL) production. This in turn increases the production of LDLs and the reduction of HDLs. This lipid profile is known as atherogenic dyslipidaemia as it eventually leads to atherosclerosis.

Intramyocellular lipids (IMCL) are more closely associated with IR than BMI, WHR, or total body fat. High FFA and VLDL levels are a key cause of fat accumulation in muscle cells and IMCL increases have been seen in patients with IR.

Surgical removal of visceral fat had a positive effect on the hepatic and peripheral insulin sensitivity, and on leptin and TNFα levels. [6]

Long term exposure of beta cells to increased FA levels causes damaging effects, such as increased insulin secretion at low glucose concentrations, decreased proinsulin production, exhaustion of insulin reserves and reduced response to concentrations of glucose stimulus. Solomon and Mayer [2] were first to associate glucocorticoids as a required factor in genetic obesity, observing that obesity was avoided after bilateral adrenalectomy (ADX) and completely restored by cortisol replacement.

[2]J. Solomon and J. Mayer, The effect of adrenalectomy on the development of the obese- hyperglycemic syndrome in ob-ob mice, Endocrinology 93 (2) (1973), pp. 510–512.

11β-Hydroxysteroid dehydrogenase type 1 (11HSD-1) catalyses the conversion of inactive cortisone to active cortisol, a potent glucocorticoid. It is extensive throughout the body and a highly regulated enzyme which increases the ligand accessibility for glucocorticoid receptors.

Excessive glucocorticoid exposure causes central obesity, hypertension, and dyslipidaemia and insulin resistance, as seen in Cushing’s syndrome. Transgenic mice over-expressing 11HSD1 in their white adipose tissue have these features as well.

Further, 11HSD1 knockout mice are protected from these metabolic abnormalities. In human idiopathic obesity, circulating cortisol levels are not elevated but 11HSD1 mRNA and activity is increased in subcutaneous adipose. The impact of increased adipose 11HSD1 on pathways leading to metabolic complications remains unclear in humans. Pharmacological inhibition of 11HSD1 has been achieved in liver with carbenoxolone, which enhances hepatic insulin sensitivity.

Visceral obesity may be secondary to enhanced local activation of cortisol via increased levels and activity of 11β-HSD-1 in adipose tissue that result in abnormally high levels of cortisol in adipose tissue. Obesity is distinct from Cushing's syndrome in that the source of the elevated glucocorticoids is adipose tissue as opposed to the adrenal cortex.[7]

[1] Zimmet P, Alberti K, Shaw J (2001) Global and societal implications of the diabetes epidemic. Nature 414:782–787 [2] Langendonk JG, Kok P, Frölich M et al (2006) Decrease in visceral fat following diet-induced weight loss in upper body compared to lower body obese premenopausal women. Eur J Intern Med 17:465–469 [3] Palaniappan L, Camethon MR, Wang Y et al (2004) Predictors of the incident metabolic syndrome in adults. Diabetes Care 27:788–793 [4] Nieves DJ, Cnop M, Retzlaff B et al (2003) The atherogenic lipoprotein profile associated with obesity and insulin resistance is largely attributable to intra-abdominal fat. Diabetes 52:172–179. [5] Turner NC, Clapham JC (1998) Insulin resistance, impaired glucose tolerance and non-insulin dependent diabetes, pathologic mechanism and treatment: current status and therapeutic possibility. Drug Res 51:36–94 [6] Gabriely I, Ma XH, Yang XM et al (2002) Removal of visceral fat prevents insulin resistance and glucose intolerance of aging. Diabetes 51:2951–2958 [7] London E and Castonguay T. Diet and the role of 11β-hydroxysteroid dehydrogenase-1 on obesity. Nutritional Biochemistry Volume 20, Issue 7, July 2009, Pages 485-493

Causes of type 2 diabetes in obese patients

Endoplasmic reticulum stress causing hyper-activation of Jun kinases (JNKs), which leads to phosphorylation of insulin receptor substrates (IRSs), inhibiting insulin signaling

Endoplasmic reticulum stress can be classed as a molecular-level link connecting obesity, insulin resistance, and type 2 diabetes. In related research, mice lacking X-box-binding protein-1 (XBP-1), a transcription factor used to modulate the body’s response to ER stress, as well as mice that had induced ER stress via pharmacological means, showed the development of insulin resistance.

ER stress or down-regulation of XBP-1 causes the suppression of insulin receptor signaling in the body’s cells via activation of Jun kinases (JNKs). In mouse studies this insulin receptor suppression led to increased insulin resistance and the development of type 2 diabetes.

It is thought that increased activity of JNKs causes the phosphorylation of insulin receptor substrates (IRSs) within important tissues such as liver, muscle and fat. As well as insulin resistance, studies have shown that increased JNK activity can result in insulin production inhibition in the pancreatic β cells. This hypothesis is strengthened by studies of mice which lack JNKs such as JNK1. In such animals, obesity induced obesity prevalence is reduced, and in general such animals also benefit from reduced adiposity.

In summary there is a key process which controls the detection of obesity-induced ER stress, causing an inhibition of insulin action that ultimately leads to insulin resistance and type 2 diabetes. It is thought that ER stress is a precursor to cell inflammation as a result of obesity. This then leads to complete breakdown of glucose homeostasis.

Dysfunction of the pancreatic β-cells, which do not produce or secrete enough insulin to compensate for insulin resistance

Many studies have shown the importance of insulin secretory capability in the formation of type 2 diabetes. If an inadequate volume of insulin is secreted by the pancreatic β cells, then adequate glucose uptake cannot occur. Couple this with increased cell insulin resistance correlated to levels of obesity, and you have the root cause of why incidence of type 2 diabetes increases with increasing levels of obesity.

In mice fed on a high fat diet (HFD), studies have shown that the subsequent diagnosis of type 2 diabetes was at least party due to reduced insulin secretion in response to greater insulin resistance.

Analysis of insulin secretion from isolated pancreatic islets of HFD mice found disfunction in the islets for the production and/or secretion of insulin. Average islet insulin contents of HFD mice were found to be significantly lower than a control group. In addition, the islets from the HFD mice showed significantly lower insulin secretion than the control group. An increase in glucagon-positive cells within the islets was also discovered in the HFD group. These physiological changes are all present in human cases of type 2 diabetes.

It is also possible that there is increased pancreatic β cell apoptosis, induced by increasing levels of obesity, reducing the level of insulin secretion. This reduced insulin secretion cannot then cope with the increasing cell insulin resistance caused by obesity.

Further research has shown that disfunction and death of the pancreatic β cells may be as a result of cell inflammation due to hyperglycemia, dyslipidemia and increased levels of adipokines.

The hormone resistin, which is thought to cause resistance to insulin

The discovery of resistin came about through the development of a new class of anti-diabetic drugs called thiazolidinediones. (TZDs) These drugs act by increasing a target tissue’s sensitivity to insulin. They function by acting as a ligand for a nuclear receptor called peroxisome proliferator activated receptor-ϒ (PPARϒ) which is found in abundance in adipocytes. Tests showed a high correlation between TZD/PPARϒ binding and glucose lowering in vivo. However, the target genes of TZD-bound PPARϒ are unknown. To try and discover whether insulin resistance might be controlled by a TZD - controlled, adipocyte-originating substance, a genetic screen was carried out for genes induced by adipocyte formation but down regulated when treated with TZDs. This screen produced evidence of a TZD- regulated protein, called resistin.

Resistin gene expression increases when adipocytes differentiate, and decreases when treated with TZD drugs such as rosiglitazone, pioglitazone and troglitazon. Mouse studies show that resistin gene expression occurs almost exclusively in white adipose tissue, with highest expression in female gonadal fat. An amino acid sequence expressed in humans with a large similarity to resistin was also found.

In mice serum analysis, levels of resistin which decrease with fasting and are restored with re-feeding were discovered. In mice fed on a 45% fat content diet for 8 weeks, the levels of resistin in serum are greatly elevated, initially increasing within four weeks of the diet being adopted, the same point as when the mice become obese and insulin resistant. Higher than normal resistin levels can also be detected in ob/ob and db/db mice, individuals genetically predisposed to obesity and diabetes.

Intraperitoneal administration of resistin to test mice results in impaired insulin sensitivity, while insulin levels remain normal. Both in vitro and in vivo studies show that neutralization of resistin causes enhanced insulin action and glucose uptake. In obese, diabetic mice, resistin neutralization causes reduced levels of hyperglycemia by increasing insulin sensitivity.

At present the molecular target of resistin is not known, but it is hypothesized that it modulates at least one step in the insulin signaling pathway. Due to it’s recent discovery, much work has still to be conducted to fully understand it’s function. At present there is great debate within the scientific community as to whether levels of resistin have a significant effect on insulin activity in humans.

In humans, resistin is thought to be secreted by macrophages not adipocytes. Despite this, there is still a strong correlation in humans between heightened levels of resistin, obesity, and type 2 diabetes.

The immunology of obesity

Type 2 diabetes has long been thought of as primarily a metabolic disease. A series of recent studies have challenged this dogma and implicated an unlikely candidate system in the promotion of disease onset - the immune system. Mild inflammation of fat tissue in obese patients reportedly acts through immune-cell processes to impair insulin signalling in adipocytes (CITE). This work therefore provides a novel way of understanding the link between obesity and type 2 diabetes.

The development of insulin insensitivity involves the malfunction of several organs; however, the focus of this article will be on that of adipose tissue. Adipocytes, the cells that comprise the adipose tissue, have a duplicitous role in that they are both a storage depot and endocrine organ. Obesity impairs the performance of adipose tissue and can induce a state of chronic, low-grade inflammation (Feuerer et al., 2009). However, unlike other forms of inflammation which are subject to control mechanisms, fat inflammation appears to escape immune regulation. The factors that initiate this inflammatory cascade are poorly understood. Several recent studies from Japanese, American and American groups have examined how T cell phenotypic changes in adipose tissue may underlie inflammation, and how adipose tissue inflammation may ultimately give rise to type 2 diabetes .

Depending on its state, adipose tissue will activate various phenotypes of T-cells (‘non-obese’ CD4 or ‘obese’ CD8), which in turn regulate (or fail to regulate) the infiltration of macrophages. It is this permeation of macrophages and their production of proinflammatory cytokines that results in chronic inflammation. The induction of chronic inflammation impairs insulin signalling in the adipocytes, which in turn leads to lipolysis and the release of non-esterified fatty acids (NEFAs) into the circulation. These fatty acids induce insulin resistance in the liver and skeletal muscles and impair B-cell function. Several groups have recently targeted the different T cell populations and both reversed and prevented the onset of obesity-induced type 2 diabetes.

Feuerer et al. (2009) isolated a specific phenotype of T cell, CD4, which is enriched in the adipose tissue of lean mice but dramatically reduced in that of obese, insulin-resistant mice. Through loss-of-function experiments it was shown that CD4 cells are functionally active and their absence results in inflammatory cytokine production and reduced glucose uptake. Importantly, CD4 T cells were only shown to behave in this manner in visceral fat stores, which, unlike subcutaneous stores, are associated with the development of type 2 diabetes. A complementary study, performed by Winer et al. (2009), demonstrated that CD4 T cell transfer reverses weight gain and insulin resistance in null mutants. Nevertheless, because the mice lost weight after the CD4 T cell transfer, it makes conclusive interpretation of the data difficult. This data led the authors to conclude that obesity-associated metabolic abnormalities are under the pathophysiological control of CD4 T cells, which inversely control the infiltration of problematic macrophages .

A separate study by Nishimura et al. (2009) revealed that a different type of T cell, CD8, are increased in obese mice and precede chronic inflammation observed in adipose tissue. Similarly, the adoptive transfer of CD8 T cells resulted in adipose inflammation. Together this evidence led the authors to propose that obese adipose tissue activates CD8+ T cells, which drive the recruitment of macrophages and their differentiation into an inflammatory rather than anti-inflammatory phenotype.

Another aspect of the immune system that has been implicated in type 2 diabetes onset is mast cell function. Mast cells respond to allergic and parasitic challenge by releasing inflammatory mediators, thus playing an integral protective role. An abundance of mast cells beyond that of what is immunologically necessary can lead to mast cell instability and inflammation. Shi et al. observed that the white adipose tissue of obese mice possesses a significantly greater number of mast cells when compared to that of lean equivalents. This led the authors to ask whether the manipulation of mast cell number, achieved through genetic reduction and pharmacological equalization, can reduce the onset of obesity and type 2 diabetes. In the first set of experiments, genetically mast cell-deficient mice and control mice were fed on a Western-diet for three months. Loss of mass cell function appeared to be having the effect of lowering serum leptin, increasing glucose tolerance and increasing insulin sensitivity in comparison to the control group. In the second strand of experimentation, mice were treated with mast cell-stabilizing medication to ask whether diet-induced obesity and diabetes could be inhibited. After two months on a Western-diet, mice were either switched to a healthy diet, supplied with mast-cell stabilizing medication, or a combination of both. While the dietary adjustment caused minor improvements, the medication stimulated significant restitution and the combination of both allowed a near full recovery in comparison to control group who continued on a Western-diet. Shi has since signed a contract with a pharmaceutical company to develop this mast cell-stabilizing drug for testing in humans.

Both of these drugs are already used to treat other medical conditions and are therefore both safe and available, however the question that remains to be answered is do Zaditor and cromolyn offer similar protection against diabetes in humans?

At present the application from model to human appears positive. A study into T cell concentration in human abdominal fat tissue by Winer et al. (2009) has revealed an abundance of protective CD4 T cells in normal weight individuals when compared to that of obese, diabetic patients, as well a reflection of inverse number of macrophages.

Normalization of obesity-associated insulin resistance through immunotherapy http://www.nature.com/nm/journal/v15/n8/full/nm.2001.html

Genetic deficiency and pharmacological stabilization of mast cells reduce diet-induced obesity and diabetes in mice http://www.nature.com/nm/journal/v15/n8/full/nm.1994.html

Lean, but not obese, fat is enriched for a unique population of regulatory T cells that affect metabolic parameters http://www.nature.com/nm/journal/v15/n8/full/nm.2002.html

CD8+ effector T cells contribute to macrophage recruitment and adipose tissue inflammation in obesity http://www.nature.com/nm/journal/v15/n8/full/nm.1964.html

The protein kinase IKKepsilon regulates energy balance in obese mice http://www.cell.com/retrieve/pii/S0092867409007934

T-ing up inflammation in fat http://www.nature.com/nm/journal/v15/n8/full/nm0809-846.html

Obesity in pregnant women: implications for both mother and fetus

As previously discussed obesity precipitates Non-Insulin Dependent Diabetes Mellitus (NIDDM). Pregnancy itself is an insulin-resistant condition; a state that allows for adequate fetal nutrition[1,2]. Obesity in pregnancy, given its link with NIDDM, has been considered a risk factor for gestational diabetes (GDM); a condition of glucose intolerance that affects around 7% of pregnancies[5,8]. The magnitude of the risk associated with obesity is not yet clear; Yogev and Visser report it as being as high as 20%, and in the US GDM has dramatically increased in line with the alarming increase in obese woman, suggesting a correlation and possible causal link.

It has been shown that for pregnant women being even moderately overweight can lead to gestational diabetes[1,4,5]. The increased production of hormones in pregnancy, such as HPL, progesterone, cortisol and prolactin, interferes with the action of insulin in its management of glucose. If the mother’s pancreas is unable to produce enough insulin (3 times as much as in a non-pregnancy state) in response to such changes in pregnancy the subsequent rise in sugar levels result in GDM[7]. Factors influencing the ability of the pancreas include genetics and obesity.

The correlation between obesity and GDM has implications for both the mother and fetus. Mothers that have suffered from GDM have an increased risk of developing NIDDM later on[8,9,10], as well as the Metabolic Syndrome(MS)* which itself can lead to NIDDM. For the fetus the risk of NIDDM in adolescence is higher, as is the prevalence of MS in early childhood. The fetus in obese and GDM mothers may be exposed to hyperglycaemia which can lead to hyperinsulinaemia; this in turn could be linked to this increased risk of developing MS and NIDDM, as well as cardiovascular disease.

Maternal obesity and its association with an increase risk in GDM is well established; the relationship and subsequent risk from GDM to NIDDM in mother and/or offspring is not as well defined although under investigation. Its implications for subsequent generations could be huge.

• metabolic syndrome refers to obesity-associated conditions such as dyslipidaemia, glucose intolerance and hypertension.

Treatment of obesity and type 2 diabetes

Obesity, alongside genetic predisposition is one of the most significant risk factors for the development of diabetes mellitus. Lifestyle changes provide the basis of treatment in all obese patients. When lifestyle changes fail to reduce the weight in obese patients, anti-obesity drugs are used. [1] There are few well tolerated drugs which have been proven to have long term efficacy in maintaining weight loss. Current available medications include sibutramine and orlistat. Sibutramine reduces body weight and appetite and increases satiety. Numerous prospective randomised controlled trials have shown it to be effective, with one trial finding that patients on sibutramine lost 4.3kg or 4.6% more weight than those taking the placebo. The most common adverse effects are dry mouth, constipation and insomnia. [2]) Orlistat acts by inhibiting pancreatic and gastrointestinal lipases, preventing absorption of about 30% of dietary fat. Randomized controlled trials have shown that patients taking this have lost 2.7kg or 2.9% more weight than controls. As orlistat reduces LDL and cholesterol levels independently of reductions in body weight, it also retards the progression to a diabetic state and aids glycemic control in patients with diabetes. Side effects include fecal urgency and abdominal cramping. [2]

Patients with impaired glucose tolerance, impaired fasting glucose and obesity are all at a high risk of developing type 2 diabetes, therefore combination therapy for glycaemic control and weight management is often required. [1] Several strategies are used, including the promotion of weight loss through lifestyle modificantions and anti-obesity drugs, improving glycemic control through the reduction of insulin resistance and the treatment of common associated risk factors such as hypertension and dyslipidaemia to improve cardiovascular prognosis. [1]

There is significant evidence that the development of type 2 diabetes can be prevented or delayed through the instigation of lifestyle modification and drugs such as metformin and orlistat [3]) Physical excerise and weight loss are amongst the most effective methods for preventing the onset of diabetes [3] and a large randomised study concluded that lifestyle intervention was more effective that metformin [4]. However lifestyle modification is often found to be difficult to sustain by obese patients.

In treating type 2 diabetes, main aims are to return metabolic disturbances to normal, achieve good glycemic control and assist with weight management. Dietary management in diabetic patients is particularly important, in order to reduce the cardiovascular risks associated with central obesity. Type 2 diabetic patients need to restrict carbohydrate and total calorific intake and eat foods of low glycemic index, to reduce the post prandial rise in blood glucose. [5] When dietary management is not successful, pharmacological intervention is added, including anti-diabetic drugs to prevent hyperglycaemia, ACE inhibitors to treat hypertension and statins or fibrates to treat hyperlipidaemia. [1]

Metformin is recommended as first line treatment in type 2 diabetic patients. When this fails, other agents are added to provide combination therapy. The majority of type 2 diabetic patients require combination therapy, because monotherapy with metformin usually only maintains good metabolic control in the short term. Treatments which can be added include sulphonylureas, acarbose, glucagon-like peptide-1 (GLP-1) analogues, thiazolidinediones, glinides, or insulin. [6]

In conclusion, a multi-strategy approach is used in the management of an obese diabetic patient. Therapy focuses on weight reduction, which is imperative as it simultaneously improves glycemic control and vascular risk factors, but often includes pharmacological treatment to reduce hyperglycemia and correct vascular risk factors. [1]

Key references

[1] A.J. Scheen. Treatment of diabetes in patients with severe obesity. Biomed Pharmacother. Volume 54, Issue 2, March 2000, Pages 74-79

[2] Chaputy JP and Tremblay A. Current and novel approaches to the drug therapy of obesity. Eur J Clin Pharmacol. 2006; 62

[3] Jermendy G. Can type 2 diabetes mellitus be considered preventable? Diabetess res clin pr. 2005; 68, S73-S81

[4] Diabetes Prevention Program Research Group, Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin, N. Engl. J. Med. 346 (2002) 393–403.

[5] Lean MEJ, Powrie JK, Anderson AS, et al. Obesity, weight loss and prognosis in type 2 diabetes. Diabetic Med. 1990; 7:228-233.

[6] Monami M, Lamanna C, Marchionni N et al. Comparison of different drugs as add-on treatments to metformin in type 2 diabetes: a meta-analysis. Diabetes Res Clin Pr. 2008; 79: 196–203.