Weight, body mass index (BMI), and distribution of body fat are three of the core issues in type 2 diabetes.
Weight, body mass index (BMI), and distribution of body fat are three of the core issues in type 2 diabetes. Weight gain,1 a high BMI,2 and central obesity3 are all risk factors for the development of this condition and its comorbidities, such as vascular complications and organ damage.4,5 The majority of individuals with type 2 diabetes are overweight, many at the time of diagnosis.6 These patients can often experience additional metabolic abnormalities, leading to further decreases in glycemic control and worsening of diabetes symptoms.7 Diabetes, obesity, dyslipidemia, and hypertension frequently occur in the same individual in a cluster of conditions collectively known as the metabolic syndrome.8
Weight loss in obese individuals has been shown to delay the onset of diabetes.9 In those who have already developed type 2 diabetes, metabolic control can be improved if weight is reduced.10,11 Properly managed lifestyle interventions such as exercise and diet plans are highly effective, and are considered as part of first-line therapy.12 This requires careful planning: according to the American Diabetes Association (ADA), standard weightreduction diets (controlled at 500–1,000 fewer calories per day than normal) do not work when used alone.13 Intensive programs of lifestyle modification that include a low-calorie, low-fat diet, structured physical activity, and one-on- one educational sessions—such as that used by the Diabetes Prevention Program (DPP)—can be even more effective in aiding weight loss than pharmacological intervention.9
Despite these relationships between diabetes and BMI, weight, and fat distribution, many of the traditional therapies to treat type 2 diabetes actually result in weight gain.7 With various insulins, for example, weight gain can be in the order of 1–4kg per patient, often proportional to the correction of glycemia.12 However, it should be noted that the newer basal analog insulins such as insuline glargine and detemir are associated with lower weight gain compared with traditional insuins. Nevertheless, as glycemic control is improved with these medications, many clinicians consider that the weight gain is an acceptable side effect. However, an increase in weight can set up an unfortunate barrier to the effective use of insulin in type 2 diabetes. Most diabetes patients are overweight already, and further weight gain may be problematic. This results in some patients deliberately not complying with their medication regimen in order to avoid such weight gain.6,14
As more types of drug become available for the treatment of type 2 diabetes, the clinician’s options will increase. Through a combination of lifestyle modification, bariatric surgery, antiobesity medication, and antidiabetic medication, a patient will be able to both gain glycemic control and lose weight. This review will examine each of these options.
Bariatric surgery includes a range of interventions in the gastrointestinal tract that aim to reduce food consumption and/or nutrient absorption. Techniques include fitting a gastric band to restrict the size of the stomach and various forms of bypass surgery; the gold standard is gastric bypass surgery, which has been performed for more than 50 years in the US. Another common procedure is sleeve gastrectomy with duodenal switch, an irreversible procedure that involves resection of the stomach and gall bladder and bypass of much of the small intestine. A meta-analysis of the effect of surgical weight-loss procedures reported that in a general population of patients with a BMI of at least 40, surgery is more effective than non-surgical treatments for weight loss, resulting in an average loss of around 20–30kg over 10 years. For patients with a lower BMI, the results were inconclusive. Mortality rates from the procedure were less than 1%.15 Type 2 diabetes is a common comorbidity in the morbidly obese, and following bariatric surgery it was completely resolved in 77% of patients (baseline BMI of 47) and at least improved in a further 9%.16
The hypothesis that gastric bands are an effective treatment for obese patients (BMI >30 <40) with type 2 diabetes was specifically tested in a randomized controlled trial. Of the 55 patients who completed the two years, 26 achieved remission of type 2 diabetes: 22 in the surgery group and four through lifestyle intervention (control group). Remission was related to weight loss, which was 20% for the surgery group compared with 1.4% in the control group.17 However, despite the increasing interest, bariatric surgery is currently an elective procedure, and is not mentioned in either the ADA’s Standards of Medical Care in Diabetes document18 or the Medical Guidelines for Clinical Practice for the Management of Diabetes Mellitus from the American Association of Clinical Endocrinologists (AACE)19
Cannabinoid Receptor Type 1 Blockers
Cannabinoid receptor type 1 (CB-1) is one of the most common G-proteincoupled receptors in the brain, and is also expressed elsewhere in the body, including the pituitary and adrenal glands, fat, muscle, and liver cells, and in the digestive tract, lungs, and kidneys. Within the gastrointestinal tract, CB- 1 plays a role in hunger, and the effect of CB-1 antagonists is to reduce appetite. The first and only CB-1 blocker is rimonabant, which was licensed in Europe in 2006 as an adjunct to diet and exercise for the treatment of obese patients with associated risk factors such as type 2 diabetes. The oneyear Rimonabant In Obesity (RIO-diabetes) trial in overweight or obese patients showed that patients taking rimonabant 20mg lost 3.9kg more weight than those on placebo over the course of one year. Furthermore, more patients on rimonabant lost at least 10% of their bodyweight than those taking placebo (16 versus 2%). Some important secondary end-points, including waist circumference, blood pressure, triglycerides, high-density lipoprotein (HDL) cholesterol, and glycated hemoglobin (HbA1c), were also significantly improved. As a result, the proportion of individuals with the metabolic syndrome was substantially reduced in those taking rimonabant.20
However, there was a high drop-out rate for all of the trials (40% at one year), although such high rates can be common in obesity studies. Nevertheless, because of adverse events, approximately twice as many people stopped taking rimonabant than placebo (14 versus 7%). Such events included psychiatric disorders, including depression and anxiety, as well as the common adverse events of insomnia, nausea, vomiting, diarrhea, and fatigue.21 Partly for these reasons, rimonabant has not been approved by the US Food and Drug Administration (FDA).
Various anorectic medications have also been studied in type 2 diabetes patients for weight loss. Fenfluramine and phentermine were effective in producing weight loss and improved glycemic control,22 but fenfluramine has been removed from markets worldwide owing to a link with valvular heart disease, and phentermine has a large number of contraindications. Sibutramine, a serotonin–norepinephrine re-uptake inhibitor, helps reduce appetite and can result in 5.5kg weight loss over one year in patients with type 2 diabetes.23 However, once again there can be cardiovascular complications, for example tachycardia. Orlistat is an inhibitor of lipase activity that causes malabsorption of a portion of dietary fat. In the four-year XENical in the prevention of diabetes in obese subjects (XENDOS) study, mean weight loss was significantly greater with orlistat than with placebo (6 versus 3kg). The cumulative incidence of diabetes was 9% with placebo and 6% with orlistat, corresponding to a risk reduction of 37.3% (p=0.0032). Weight loss produced by orlistat also resulted in modest improvements in glycemia. Furthermore, the drop-out rate of trial subjects was lower for the orlistat than the placebo group (52 versus 34%).24 Thus, currently available weight-loss medications produce clinically significant weight loss and modest reductions in glycemia, but most are expensive and must be continued indefinitely to maintain weight loss. For a patient already taking antidiabetic drugs, this can substantially increase the pill burden.
Alpha-glucosidase inhibitors are oral antidiabetic drugs that prevent the digestion of carbohydrates. Thus, their primary use is in controlling postprandial blood-glucose levels. They act as competitive inhibitors of the carbohydrate digestive enzymes, particularly those in the small intestines, and can be used as monotherapy alongside appropriate diet and exercise regimens, or added to other medications. The three available inhibitors are acarbose, miglitol, and voglibose. Acarbose is an oligosaccharide, while miglitol is closer to a monosaccharide. In addition, acarbose inhibits pancreatic alpha-amylase. A meta-analysis of randomized controlled trials studying alpha-glucosidase inhibitors found that all of the drugs reduce HbA1c only modestly (the greatest effect being 0.8% with acarbose). There were also adverse events associated with these agents, notably ‘hyperflatulence.’25 Effects on bodyweight also appear negligible.26
Metformin is the only widely available biguanide—a class of drug that acts primarily by reducing hepatic glucose output and increasing uptake of glucose in the peripheral tissues. It typically reduces HbA1c27 Its mechanism of action is not fully elucidated, although it is classified as an insulin sensitizer.28 Metformin does not stimulate insulin secretion.
Overall, metformin is not associated with as much weight gain as some of the other type 2 diabetes treatments, and in fact can often result in weight loss. Although metformin therapy is not as effective overall as lifestyle intervention, the DPP trial showed that it nevertheless reduced the risk for diabetes by 31% and led to an average weight loss of 2.1kg.9 Administering metformin alongside insulin can reduce or negate the weight gain associated with the latter.27 There are also non-glycemic benefits, as metformin can help reduce low-density lipoprotein (LDL), triglycerides, and antifibrinolytic factor plasminogen activator inhibitor.
However, despite these positive aspects of metformin, there are downsides. The most serious complication is lactic acidosis, which is rare but can be fatal. Two of the three other biguanides (phenformin and buformin) were withdrawn in the 1980s owing to this side effect.29 Therefore, metformin is limited to patients with sufficient renal function to avoid drug accumulation. It is also contraindicated in patients with cardiac or respiratory insufficiency, or other conditions associated with hypoxia.
Thiazolidinediones (TZDs) increase insulin sensitivity in adipose tissue, muscle, and the liver by binding to peroxisome proliferator-activated receptor gamma (PPARγ), which is involved in the transcription of genes that regulate glucose and fat metabolism. As with metformin, the mechanism of action of TZDs is not completely understood, although it is thought to be distinct.30 There is a delay in onset of action with TZDs, and it can often take up to three months for their full effect to be felt.
With TZDs, notably rosiglitazone and pioglitazone, there appears to be a trade-off between increase in weight (roughly 1–4kg) and reduction in HbA1c (around 0.5–1.5%). Some of the weight gain is explained by water retention, and a small proportion of patients develop leg edema.31 There is evidence that any adipogenesis caused by TZDs does not occur around the internal organs.12,29 Given that it is visceral fat deposits that are most closely associated with insulin resistance, the weight gain caused by TZDs is unfortunate but probably not of great concern metabolically. The effect of TZDs on atherogenic lipid profiles is, however, more complex.32
For secondary events, the PROspective pioglitAzone Clinical Trial In macroVascular Events (Proactive) in high-risk patients with type 2 diabetes found that pioglitazone was associated with a statistically significant 16% reduction in all-cause mortality, non-fatal myocardial infarction (MI), or stroke. However, in contrast Proactive also highlighted an increased frequency of heart failure and edema and a 4kg weight increase compared with placebo.33 In Europe, the use of TZDs in combination with insulin is contraindicated owing to a perceived increased risk of heart failure. US guidelines also urge a cautious approach to TZD use in patients with evidence of heart failure.
Sulfonylureas have been available for more than 50 years. They work only in patients who have sufficient β-cell function to support the increased insulin output. Treatment with this class of antidiabetic agent reduces HbA1c by approximately 1–2%.34 There is also evidence for a decrease in the risk for microvascular complications and some other cardiovascular complications, such as MI.35
Sulfonylureas do increase the risk for hypoglycemia. This is a particular concern in older people,18 with glyburide giving the highest rate.36 However, perhaps the biggest drawback of sulfonylureas—certainly with the firstgeneration agents (chlorpropamide, tolbutamide, acetohexamide, and tolazamide)—is the associated weight gain, averaging around 1–4kg within the first six months. Second-generation agents, including glyburide and glipizide, are more potent, but are also associated with hypoglycemia and weight gain. Longer-acting sulfonylureas, such as glimepiride and extendedrelease glipizide, have more favorable profiles with less hypoglycemia and less weight gain. Subjects in the UK Prospective Diabetes Study (UKPDS) experienced weight gain of 1.7kg with glyburide and 2.6kg with chlorpropamide.35 Overall, sulfonylureas may be used with metformin as first-line therapy as they have a long history of use and are cost-effective.37
Meglitinides, also known as rapid- or short-acting secretagogs, are distinct from the sulfonylureas but have a comparable mechanism of action. These agents also exhibit a change in weight that is roughly equivalent to that achieved with sulfonylureas.38 The two meglitinides available are repaglinide and nateglinide. Nateglinide has a much lower frequency of hypoglycaemia compared with repaglinide; on the other hand, repaglinide has more potent effects on lowering HbA1c. A head-to-head trial of the two agents as monotherapy resulted in mean weight gain of 1.8kg in the repaglinide group compared with 0.7kg in the nateglinide group.39 There is currently a dearth of evidence regarding the effect of meglitinides on long-term micro- and macrovascular risk.
Amylin is a peptide secreted by pancreatic β cells at the same time as insulin. It inhibits the appearance of nutrients (especially glucose) in the plasma by reducing food intake, slowing gastric emptying, and inhibiting secretion of digestive juices. It also acts to inhibit secretion of glucagon. Synthetic amylin, or pramlintide, was approved in 2005 for use in patients with type 2 diabetes as an adjunct treatment to mealtime insulin therapy, with or without a concurrent sulfonylurea and/or metformin. A review of five studies, ranging from four to 52 weeks’ duration, examined the effect of administering pre-meal pramlintide in patients with type 2 diabetes. In these trials, pramlintide 90–450μg per day reduced HbA1c by 0.3–0.6%, with the main adverse events being nausea and hypoglycemia. There is a high risk for hypoglycemia in the first four weeks of therapy with pramlintide; thus, the manufacturer recommends reducing the dose of pre-meal insulin by 50% when starting this therapy.40 In trials of pramlintide, reductions in HbA1c have been accompanied by significant reductions in bodyweight of around 1.4kg from baseline, independent of nausea,41 possibly caused by promoting the feeling of satiety and thus reducing caloric intake.42
Like insulin, glucagon is an important regulator of carbohydrate metabolism that works by stimulating glucose production in the liver. In type 2 diabetes, glucagon levels are often raised in the plasma, causing an increase in glucose output. Newer agents have moved the focus from controlling insulin and glucagon levels to look at other factors in glucose metabolism. Levels of insulin and glucagon are both influenced by the incretin hormones glucagon-like peptide-1 (GLP-1), which is secreted by the enteroendocrine L cells as a response to food intake, and glucose-dependent insulinotropic polypeptide (GIP), which is secreted by intestinal K cells. In addition to affecting levels of insulin and glucagon, GLP-1 and GIP affect the rate of gastric emptying and act as a mediator of satiety. Thus, the overall picture of glucose homeostasis and bodyweight control is a very complex model involving multiple organs and hormones.43 GLP-1 protects pancreatic β cells and inhibits glucagon secretion, gastric emptying, and food intake, which all promote weight loss. However, GLP-1 is rapidly degraded by the enzyme dipeptidyl peptidase-4 (DPP-4), which renders it unsuitable for use as a therapeutic. There are therefore two strategies that allow this mechanism to be used therapeutically: GLP-1 mimetics that resist the action of DDP-4, and inhibitors of DPP-4 that allow endogenous levels of GLP-1 to rise.44
Glucagon-like Peptide-1 Receptor Agonists
The first-in-class GLP-1 mimetic is exanatide, a synthetic form of exendin-4, which is found in the saliva of the Gila monster. Although exendin-4 has roughly 50% sequence homology with the human peptide, it is a full agonist for its receptor. Furthermore, it is stable against DPP-4 and is eliminated via the kidneys by glomerular filtration, with a plasma half-life of 30 minutes. It also suppresses glucagon, has an inhibitory effect on gastric emptying, and has been shown to reduce food intake in humans.45
Exenatide 5 and 10μg administered twice a day has been shown to reduce HbA1c by approximately 1% and to simulate weight loss of 0.9kg when given alongside sulfonylurea and metformin,46,47 and as much as 2.5kg when added to metformin alone.48 This weight loss appears to be durable: a longer trial in 150 overweight patients found that HbA1c reduction was around 1% at 30 weeks and 1.3% at 82 weeks, while weight loss was 3.0 and 5.3kg at the same time periods, respectively.49
One study has explicitly looked at the safety of substituting exenatide for insulin in patients with type 2 diabetes. A total of 49 obese patients with BMI >30 either substituted exenatide for insulin or remained on their normal insulin regimen; 45 completed the 16-week study. In the exenatide group (n=29), mean HbA1c increase was 0.3%, while in the insulin group (n=16) HcA1c decreased by 0.1%. These changes were not significantly different. More than 60% of patients in the exenatide group maintained glycemic control compared with just over 80% in the insulin group. C-peptide and baseline bodyweight were the best predictors of successful glycemic control. In terms of weight loss, switching from insulin to exenatide resulted in a reduction in weight for the majority of patients (27 of 29; 93%). Slightly more than one-third (six of 16; 38%) of the patients in the insulin group lost weight. This difference was statistically significant (p<0.001), and was similar for all patients, regardless of retention of glycemic control.50 However, this study emphasizes the need for insulin replacement therapy in patients with long-standing type 2 diabetes who also have significant insulinopenia.
Severe adverse events that may limit patient compliance with exenatide therapy are rare. Hypoglycemia occurs in up to one-third of patients taking both exenatide and a sulfonylurea. There have also been reports of gastrointestinal adverse events, including nausea and vomiting.51 In the exenatide substitution study, 26 of 33 exenatide patients (79%) reported a treatment-emergent adverse event compared with nine of 16 (56%) in the reference group. Most of the adverse events were mild to moderate and were predominantly gastrointestine-related, with nausea the most common. Five patients discontinued because of an adverse event and one patient required hospitalization. The incidence of hypoglycemia was similar between the two groups, and most of the patients in the exenatide group who experienced hypoglycemia were also taking a sulfonylurea.50
Exenatide is currently administered in twice-daily injections, making it less appealing to patients. A long-acting release (LAR) formulation of exenatide, administered once a week, is under review with the FDA. Compared with (long-acting) placebo, exenatide 2mg once weekly reduced mean HbA1c by 1.7% from baseline, with 86% of patients achieving HbA1c ≤7%. Weight loss in this group of patients was 3.8kg compared with 0kg with placebo.52 In a 30-week head-to-head, randomized, open-label comparison of LAR exenatide with twice-daily exenatide, the LAR formulation showed significantly greater improvements in HbA1c (1.9% compared with 1.5%). More patients achieved HbA1c ≤7%, and both groups lost approximately 4kg in weight. Furthermore, side effects such as nausea were far less common with the LAR formulation.53
Behind exenatide, in phase II development is liraglutide, an acylated GLP-1 mimetic. It has a longer half-life and can be given once a day rather than via the twice-daily injections that are currently required for exenatide. In a study of overweight patients, liraglutide at the highest dose of 1.9mg led to a 1.5% reduction of HbA1c and weight loss of 3.0kg after 14 weeks. Triglyceride levels also decreased by 22% compared with placebo.54
A meta-analysis of incretin mimetics has confirmed that GLP-1 analogs significantly lower HbA1c compared with placebo, and have an equivalent effect to insulin glargine. In trials that reported weight changes, GLP-1 analogs led to a statistically significant weight loss compared with other agents. This effect was even greater in comparison with insulin. Furthermore, for at least 30 weeks the weight loss is progressive and dose-dependent. To date, trials have shown more weight loss with exenatide than with liraglutide. Nausea may contribute in part to the weight loss, although it cannot explain all of it: participants who did not report any nausea also had weight loss. Regarding lipid profiles, the meta-analysis covered only three trials that reported on this factor, and overall there were no major effects.55
Dipeptidyl Peptidase-4 Inhibitors
Sitagliptin is the only DPP-4 inhibitor currently approved in the US. Vildagliptin is approved for use in Europe but is still under review in the US. Both agents have been shown to be effective at reducing HbA1c when used as monotherapy, and can also be added to metformin, sulfonylureas, or TZDs. Typical reduction is in the range of 0.5–1.1%.56 Patients with higher starting HbA1c typically achieve greater reductions.57 Nevertheless, overall glucose is not lowered to as great an extent as with many existing therapies, although DPP-4 inhibitors tend to be well tolerated with low incidences of hypoglycemia and gastrointestinal adverse events.
In a meal tolerance test, sitagliptin has been shown to significantly decrease two-hour post-prandial glucose. However, there is little evidence of meaningful bodyweight changes from baseline: DPP-4 inhibitors as monotherapy tend to be weight-neutral.58 One study even found a greater decrease in bodyweight with placebo than with sitagliptin.57 A comparison of vildagliptin with rosiglitazone has shown that while there was a negligible change in bodyweight in vildagliptin-treated patients, there was a corresponding increase in weight in those treated with rosiglitazone. Vildagliptin did significantly decrease triglycerides, total cholesterol, and LDL.59 There is evidence that, as with HbA1c, there is more weight loss in patients with very high baseline BMI.56 Conversely, sitagliptin therapy as an add-on therapy to glimepiride has been shown to lead to modest weight gain in addition to improved glycemic control.60
A trial of combination treatment with sitagliptin 100mg and metformin 2,000mg versus either agent alone or placebo showed that this combination provides substantial and additive glycemic improvement and is generally well tolerated, with a low incidence of hypoglycemia.61
A meta-analysis of DPP-4 studies showed that overall these agents led to greater reductions in HbA1c as both monotherapy and add-on therapy compared with placebo. There have been no direct comparisons to date, but effect size seems comparable between the two main inhibitors. Compared with other hypoglycemic agents, DPP-4 inhibitors appear to be marginally less effective. In terms of weight, there was a small increase in weight with DPP-4 inhibitors compared with placebo, but sitagliptin had a favorable weight profile compared with glipizide, while vildagliptin had a favorable weight profile compared with TZDs but not with metformin. Lipid profiles were not noticeably affected compared with placebo, but showed improvements against rosiglitazone, as mentioned above.55
Summary and Conclusions
There are many physiological changes within the diabetic patient that all need to be addressed when considering treatment. One of the major contributing factors to the rise in incidence of diabetes and its persistence within the population is obesity. Thus, while the existing armamentarium is effective when it comes to controlling blood glucose levels, it is now time to consider the effect of weight and lipid levels as well. While many antiobesity drugs are effective at lowering weight in patients with type 2 diabetes, they may be limited by safety, tolerability, and cost concerns. For obese patients, bariatric surgery is associated with good weight loss, improved glycemic values, and, in many cases, a reduction in the incidence of diabetes. However, at the moment they are still elective procedures and have hardly been studied other than in the morbidly obese. Of the traditional agents for type 2 diabetes, metformin therapy can lead to weight loss. Incretin therapy offers an alternative option to currently available hypoglycemic agents. On the whole, the GLP-1 mimetics and DPP-4 inhibitors are equally effective in lowering HbA1c, but there is promising evidence that they also have other, more desirable, effects. Notably, the weight loss associated with exenatide, which is progressive and consistent, is a valuable tool for use in conjunction with lifestyle modifications to help to ameliorate one of the underlying causes of diabetes and the metabolic condition. However, these new classes of agent will need continued evaluation both in long-term trials and in clinical practice to assess their effectiveness and safety profiles.■