Currently, 8.3% of the US population (25.8 million people) has diabetes, including 18.8 million people who are diagnosed, plus approximately 7 million who remain undiagnosed.1 Type 2 diabetes mellitus (T2DM), characterized by insulin resistance, a reduction in insulin secretion by pancreatic β cells, and impaired incretin response, accounts for 90% to 95% of all diagnoses of diabetes.1 Inadequate T2DM management is a leading cause of cardiovascular disease (CVD), blindness, kidney disease, and amputations, and has also been associated with increased risks for developing cancer, depression, cognitive decline, liver disease, and arthritis, and other serious conditions.2,3 In addition, it is estimated that up to 90% of patients with T2DM have sleep apnea, many of whom are undiagnosed in primary care settings.4,5

Risk factors for T2DM include nonmodifiable and modifiable factors. The former group includes age, race, ethnicity, sex, and family history. Modifiable risk factors include smoking, hypertension, physical inactivity, and obesity.1 In fact, 53% of individuals with diabetes are obese,6 including approximately 88% of patients with T2DM.7 The risks associated with diabetes increase substantially with increasing body mass index.6

The increasing prevalence of T2DM and obesity in the United States has led to skyrocketing economic costs associated with T2DM treatment and associated complications, reaching ≥ $174 billion for diabetes care in 2007 alone.1 Thus, effective management strategies are clearly important. Strict glycemic control, measured most commonly by glycated hemoglobin (HbA1c) levels, with minimal adverse events (AEs) is the ultimate goal of any treatment regimen. The American Association of Clinical Endocrinologists (AACE) and the American College of Endocrinology (ACE) suggest an HbA1c level ≤ 6.5%,8 while the American College of Physicians (ACP),9 the American Diabetes Association (ADA), and the European Association for the Study of Diabetes (EASD) recommend that HbA1c levels be maintained near normal if possible or < 7%, with less stringent goals considered on a patient-by-patient basis. For example, the ADA and EASD suggest that a higher HbA1c level of 7.5% to 8% may be acceptable for patients with significant comorbidities, and that a lower HbA1c level of 6% to 6.5% may be recommended for patients without significant CVD.3 Importantly, studies suggest that weight reduction may help patients reach these HbA1c goals.10,11

The widening array of pharmaceutical products offers more choices for prescribers wishing to optimally manage patients with T2DM. However, the number of treatment guidelines, each with limited specificity in medication selection and overall strategy, also makes it difficult for clinicians to readily identify the most effective and best-tolerated treatment for each patient. For example, some treatment protocols result in weight gain and/or hypoglycemia, and, thus, should ideally be avoided. This article discusses the strengths and weaknesses of the available treatment options for patients with T2DM. We recommend an individualized MGI (metformin, glucagon-like peptide-1 receptor agonist, and insulin) treatment approach consisting of specific diet, exercise, and drug intervention strategies that simultaneously address the core pathophysiologic defects in patients with T2DM and obesity in an effort to normalize physiology in patients with T2DM.

Pathophysiology of Patients With T2DM

The pathogenic formula for T2DM represents a complex interplay of various abnormalities. The triumvirate theory of T2DM, which was proposed by DeFronzo12 in 1988, consisted of insulin resistance in the liver and muscle, causing glucose release, failure of glucose uptake, and pancreatic β-cell malfunction (Figure 1, I–III), resulting in decreased insulin secretion. More than 20 years later, DeFronzo13 revised this theory to describe T2DM pathogenesis as an “ominous octet,” which added the fat cell (accelerated lipolysis), gastrointestinal tract (incretin deficiency/resistance), α cell (hyperglucagonemia), kidney (increased glucose reabsorption), and brain (loss of insulin-induced appetite suppression) (Figure 1, IV–VIII). The revised theory suggests that the ideal T2DM therapy should aim to restore normal physiology through reversal of all 8 pathogenic abnormalities, not only HbA1c level.

View: Figure 1
The multifactorial pathogenesis of T2DM.

Large X: T2DM pathologies directly addressed by GLP-1 receptor agonists.

Small x: Anecdotal evidence suggests that weight loss induced by GLP-1 receptor agonists may correspond with positive changes in food and exercise choices.

Abbreviations: GLP-1, glucagon-like peptide-1; T2DM, type 2 diabetes mellitus.

Lifestyle Intervention Approaches

Lifestyle intervention strategies have long been recognized as cornerstones of T2DM management and prevention. In 2002, the Diabetes Prevention Program demonstrated that lifestyle modification alone decreased the incidence of T2DM by 58%, compared with a reduction of 31% among individuals taking first-line drug therapy.14 In motivated patients, specific diet and exercise regimens may be superior to drug interventions, often allowing for a delay in the initiation of drug therapy, dose reduction, or even elimination of T2DM medications.


The US Surgeon General suggests that exercising ≥ 30 minutes per day will reduce the risk for developing T2DM.15 Meta-analyses have proposed that structured exercise programs improve long-term glycemic control as measured by HbA1c levels.16,17 The glucoregulatory effects of exercise are thought to result from increased insulin sensitivity, which persists for up to 48 hours after exercise completion.18 Thus, not only does exercise help maintain a patient’s weight and improve glycemic control, but the benefits may also include overall CVD risk reduction,19 improvement in depressive symptoms,20 and reduced mortality rates.21 One study suggested that even moderate levels of regular exercise can reduce the risk of death by up to 38% in patients with diabetes,21 although this requires further investigation.22

In general, 2 modes of exercise are considered. Endurance exercise, defined as aerobic exercise involving several muscle groups with an overall goal of improving cardiorespiratory performance and reducing body weight, includes running, cycling, or swimming. In contrast, resistance exercise, or strength training, forces contraction of muscle groups against resistance, usually accomplished with free weights or weight machines.23 Although endurance exercise has been considered a critical element in diabetes intervention programs because of its success in reducing fat mass and improving weight maintenance,24 it may not be possible for some patients with T2DM, such as those with muscle weakness, severe CVD, peripheral neuropathy, reduced exercise tolerance, or arthritis, or elderly patients.25,26 These limitations resulted in the inclusion of moderate-to-vigorous resistance training for 2 to 3 days per week in the ADA/American College of Sports Medicine exercise guidelines.27 Addition of resistance training is further supported by 2 studies that found endurance plus resistance exercise in patients with T2DM to improve glucose control and insulin sensitivity to a similar degree.26,28 Moreover, a recent study suggested that men who lifted weights as little as 59 minutes per week reduced their risk for diabetes by 12% compared with men who reported no weight training,29 possibly as a result of increased glucose uptake and insulin sensitivity in the muscles.30

Patients with diabetes who can perform both aerobic and resistance training will undoubtedly obtain the best results. For example, it is estimated that individuals who lift weights for 150 minutes per week and perform 150 minutes of aerobic exercise per week (translating to 50 minutes for 6 days per week) decreased their risk for developing T2DM by as much as 59% compared with 34% for patients who performed weight training alone.29 In addition to glycemic control and weight loss, Tvoura et al31 found that combined strength and aerobic exercise programs also have antiatherogenic and anti-inflammatory effects, which likely reduces CVD risk and improves overall health status.

While it is clear that both resistance and endurance exercise should be integrated into intervention programs, detailed adjustment to provide optimal benefits will require additional research. For example, a recent study by Yardley et al32 suggested that the order in which exercises are performed may be important in patients with type 1 diabetes mellitus for glycemic stability and severity of post-exercise hypoglycemia. Whether this also applies to patients with T2DM has yet to be determined.


In the United States, diets rich in animal products have contributed to the increased incidence of obesity and T2DM. A recent review of 3 cohort studies, including data from Harvard’s Nurses’ Health Studies I/II and Health Professionals’ Follow-Up Study, found that individuals who ate red or processed meat had a higher risk for developing T2DM.33,34 Specifically, 3.5 oz of red meat or 1.8 oz of processed meat (ie, a hot dog or 2 slices of bacon) daily led to a 19% and 51% increase in T2DM risk, respectively. Furthermore, a small study suggested that eating processed foods results in increased energy absorption,35 which may contribute to associated weight gain and increased risk for developing diabetes. Thus, a diet with an increased plant-based component should be considered a primary target for diabetes management.

Plant-based diets consisting of whole grains, fruits, vegetables, legumes, and nuts, which have long been used for weight loss,36 increase insulin sensitivity and reduce CVD risk factors.11 A 22-week randomized trial in which 99 patients with well-controlled T2DM were assigned to a low-fat, low–glycemic index, vegan diet or a portion-controlled omnivorous diet following the 2003 ADA guidelines, illustrated that HbA1c levels decreased more in the vegan group than in the ADA diet group (1.23% vs 0.38%; P = 0.01). Moreover, the average weight loss was greater for patients on the vegan diet than those on the ADA diet (6.5 kg vs 3.1 kg; P < 0.001). Importantly, antidiabetic medication requirements declined further in patients on the vegan diet (43% vs 26%, respectively).37 These changes were generally maintained for > 18 months,11 illustrating the importance of food choices for T2DM management.

Despite some clear health benefits, a vegan diet is not realistic for most patients. Instead, as supported by the Harvard studies,33 a less restrictive, primarily plant-based diet that incorporates healthy protein (eg, fish and chicken) is prudent. The diet should also include an abundance of whole grains, vegetables, legumes, fruits, and healthy oils. Given the relationships between red and processed meats and T2DM and CVD, it is best to avoid these foods.

Together, these studies demonstrate the importance of lifestyle intervention strategies via a mainly plant-based diet plus a combination of aerobic and resistance exercise as part of T2DM management and prevention. Strong support from counselors who encouraged patients to adhere to these lifestyle changes was also central to many of the studies discussed. For most patients, such a support system may not be available, limiting the effectiveness of this type of intervention in the real world, particularly in the long-term. Even for patients who are able to adhere to strict diet and exercise regimens, these programs are sometimes insufficient to properly manage T2DM; it is often necessary to add drug therapy to more effectively control T2DM in patients who present with high blood glucose levels or do not meet their glycemic and weight targets with lifestyle intervention alone.

Drug Therapies

The treatment options available to manage patients with T2DM have evolved significantly over the past decade. While each novel option brings about new opportunities for better patient management, they also raise questions regarding how they fit into the current treatment paradigm and what advantages they have compared with existing therapies. Currently, approximately 10 classes of drugs are available for the pharmacologic treatment of T2DM (Figure 2). The most frequently prescribed are discussed in detail in the next section, and there are 2 new drug classes awaiting US Food and Drug Administration approval (Table 1). Based on clinical experience, the recommendations in this article represent my opinions on the use of currently available T2DM therapies, advocating medication pathways that promote lifestyle change, weight loss, and minimal hypoglycemic risk.

View: Figure 2
Development of drug classes for T2DM.86

aBiguanides became available in the United Kingdom in 1957, but were not available in the United States until 1995.

Source: American Heart Association.86

Abbreviations: AGI, alpha-glucosidase inhibitor; DPP-4, dipeptidyl peptidase-4; GLP-1, glucagon-like peptide-1; SU, sulfonylurea; T2DM, type 2 diabetes mellitus; TZD, thiazolidinedione.

View: Table 1
Medications Available to Treat T2DM3,13,40-42,50,53,54,81,99,100,103-105,113-127
Class (Medication) MOA/Effects Advantages Disadvantages Dosage Average Wholesale Price, $ (US)a
Biguanides (metformin) Activates AMPK Weight neutral GI AEs (diarrhea, cramping) 500–2550 mg daily 25–175
↓ Hepatic glucose production No to low risk of hypoglycemia Lactic acidosis (rare) deficiency
↑ Insulin sensitivity Vitamin B12 deficiency
↓ Insulin resistance ↓ CV events Contraindications: reduced kidney, liver, cardiac function, history of lactic acidosis, hypoxemia
↑ GLP-1 levels? Low cost
Oral administration
Long-term safety known
GLP-1 receptor agonists (exenatide LAR, exenatide bid, liraglutide) Activates GLP-1 receptors ↓ Weight GI Aes (nausea, vomiting) Exenatide bid: 10–20 μg daily; 160–470
↑ GLP-1 response No to low risk of hypoglycemia Cases of acute pancreatitis and medullary thyroid cell carcinoma observed
↑ Insulin secretion exenatide LAR: 2 mg weekly;
↓ Glucagon secretion Improved β-cell function/survival
↓ Gastric emptying Injectable liraglutide: 0.6, 1.2, 1.8 mg/day
↑ Satiety Potential CV benefit Long-term safety unknown
DPP-4 inhibitors (sitagliptin, saxagliptin, linagliptin) Inhibits DPP-4 activity Weight neutral Cases of pancreatitis observed Saxagliptin: 2.5–5 mg daily; 220–230
↑ Active GLP-1 concentration No to low risk of hypoglycemia Allergic and hypersensitivity reactions
↑ Insulin secretion Oral administration Long-term safety unknown sitagliptin: 100 mg daily;
↓ Glucagon secretion β-cell survival linagliptin: 5 mg daily
Long-acting insulin analogs (glargine, detemir) Directly activates insulin receptor Sustained glycemic improvements ↑ Weight Varies 95–190 (not including injection device)
↓ Hepatic glucose production Hypoglycemia
↓ Lipolysis Training required
Associated stigma
Fluid retention
Dopamine agonists (bromocriptine) Activates dopaminergic receptors Weight neutral Dizziness/syncope 0.8–4.8 mg daily N/A
Alters hypothalamic regulation of metabolism No hypoglycemia Nausea
↓ CV events Fatigue
↓ Hepatic glucose production Rhinitis
SGLT2 inhibitors (dapagliflozin, canagliflozin, empagliflozin) Inhibits SGLT2, responsible for glucose reabsorption ↓ Weight Vulvovaginitis, balanitis, and lower UTIs N/A N/A
No hypoglycemia
↑ Urinary glucose excretion
↓ Plasma glucose
Not Recommended
Sulfonylureas (glyburide, glipizide, glimepiride) Closes KATP channels on Low cost ↑ Weight 1–40 mg daily 15–160
β-cell plasma membranes Long-term safety established Hypoglycemia
↑ Insulin secretion HbA1c reductions not durable
Conflicting data on CVD risk
Thiazolidinediones (pioglitazone, rosiglitazone) Activates PPAR-γ No hypoglycemia ↑ Weight Pioglitazone: 15–45 mg daily 180–300
↑ Insulin sensitivity Durable Edema, heart failure
↑ HDL-C levels ↑ Risk of myocardial infarction Rosiglitazone: ≤ 8 mg/day
↓ TG levels Bone fractures
↓ All-cause mortality (pioglitazone) ↑ LDL-C levels
Alpha-glucosidase inhibitors (acarbose, miglitol) Inhibit carbohydrate degradation ↓ TG levels Hypoglycemia likely in combination with other drugs 25–100 mg 3 times daily N/A
↑ Secretion of GLP-1 ↓ Weight
No hypoglycemia GI AEs
Low cost
Amylin (pramlintide) Amylinomimetic; stimulates amylin receptors ↓ Weight GI AEs: nausea, vomiting 60–120 μg daily 232–440
↓ Gastric emptying Injectable
↑ Satiety Long-term safety unknown
↓ Glucagon secretion Only used with insulin
Administration frequency
Meglitinides (repaglinide, nateglinide) Closes KATP channels on β-cell plasma membranes Rapid, short acting ↑ Weight Repaglinide: 1.5–12 mg daily 195–505
Administration frequency Nateglinide: 120–360 mg daily
↑ Insulin secretion
Ultra–long-acting insulin analogs (insulin degludec) Multihexamer solubilization technology Less hypoglycemia than with long-acting insulin analogs ↑ Weight similar to long-acting analogs N/A N/A
Directly activates insulin receptor Injectable
Training required
↓ Hepatic glucose production Associated stigma
↓ Lipolysis

aPrices calculated as of November 27, 2012.

Abbreviations: AE, adverse event; AMPK, adenosine monophosphate–activated protein kinase; bid, twice daily; CV, cardiovascular; CVD, cardiovascular disease; DPP-4, dipeptidyl peptidase-4; GI, gastrointestinal; GLP-1, glucagon-like peptide-1; HbA1c, glycated hemoglobin; HDL-C, high-density lipoprotein cholesterol; KATP, adenosine triphosphate–sensitive potassium; LAR, long-acting release; LDL-C, low-density lipoprotein cholesterol; MOA, mechanism of action; N/A, not applicable; PPAR, peroxisome proliferator–activated receptor; SGLT2, sodium-glucose co-transporter 2; T2DM, type 2 diabetes mellitus; TG, triglyceride; UTI, urinary tract infection.

Recommended Medications
Biguanides (Metformin)

Currently, in the absence of contraindications, metformin is the first-line oral therapy for T2DM as recommended by the AACE/ACE, ADA/EASD, and ACP.3,9,38,39 Metformin, a low-cost therapy available in immediate-release and once-daily extended-release formulations,40 addresses items I, II, and V of the pathophysiologic abnormalities of T2DM (Figure 1).41 Glycated hemoglobin level reductions are typically 1% to 2%,38 with a neutral impact on weight. The extended-release formulation, which offers the added convenience of less frequent dose administration, should be initiated at a low level, with gradual dose increases to reduce potential gastrointestinal AEs and ensure that the minimum effective dose is prescribed.40 Initial gastrointestinal AEs (eg, nausea and diarrhea) are common, appear to be dose-related, and often resolve after dose reduction; > 50% of patients with T2DM can tolerate the maximum dose, but up to 5% are unable to tolerate any dose of metformin.42 Similar AEs have been reported with both immediate-release and extended-release formulations.40 Renal, hepatic, and heart failure are the main contraindications to metformin therapy, due in part to the risk of lactic acidosis, a very rare AE.41 Metformin monotherapy is sometimes insufficient to enable patients to reach their HbA1c level goals and is often used in combination ≥ 2 additional T2DM drugs, including incretin therapies, sulfonylureas (SUs), and insulin.

Incretin Therapies

If T2DM is not controlled with metformin alone, or if patients are unable to take metformin because of contraindications, incretin therapy is often the next line of treatment. The incretin system plays an important role in glucose homeostasis, largely through the actions of glucagon-like peptide-1 (GLP-1), an intestinal hormone with broad physiologic effects that result in reduced hepatic glucose production.13 Incretin therapies consist of GLP-1 receptor agonists and dipeptidyl peptidase-4 (DPP-4) inhibitors.

GLP-1 Receptor Agonist Therapy

The preferred next step after metformin as recommended by the AACE guidelines,8 GLP-1 receptor agonist therapy provides pharmacologic levels of GLP-1 (Figure 1, V). This stimulates insulin secretion by pancreatic β cells and enhances glucose sensitivity (Figure 1, III), prevents inflammation-mediated apoptosis of pancreatic β cells (Figure 1, III), reduces glucagon secretion by pancreatic α cells (Figure 1, VI), decreases hepatic glucose production (Figure 1, I), replaces a deficient GLP-1 response, delays gastric emptying, induces satiety (Figure 1, VIII), and results in overall weight loss.13,43-45 These results suggest that GLP-1 receptor agonists address 5 of the pathophysiologic defects. Moreover, GLP-1 receptor agonists reduce triglyceride levels and blood pressure.46-48 Additionally, evidence suggests that GLP-1 receptor agonists may facilitate better food choices, aiding in further weight loss,49 and, potentially, increases in exercise quantity (Figure 1). The most common AEs are gastrointestinal in nature, particularly early transient nausea and diarrhea.50 An increased risk of pancreatitis was also reported in a small number of patients and in postmarketing surveillance51; however, the existence of a direct link between GLP-1 receptor agonists and pancreatitis remains controversial.52 Moreover, an association between GLP-1 receptor agonists and thyroid C-cell tumors has been shown in rodents, although the relevance to humans has not been determined.53,54

Two GLP-1 receptor agonists are available in the United States: liraglutide (once daily; starting at 0.6 mg, increasing to 1.2 or 1.8 mg) and exenatide (twice daily [bid] 5–10 μg and long-acting release [LAR] once weekly; 2 mg). Liraglutide is a modified human GLP-1 that is 97% homologous to the endogenous form. In contrast, exenatide is a modified version of exendin-4 (originating from the Gila monster) that is 53% homologous to human GLP-1.55 Their roles in the treatment of T2DM have been characterized in large-scale clinical trials as either monotherapy,46,48,56 in combination with SUs57,58 or metformin,59-62 or as a 3-drug combination,63-67 resulting in HbA1c level reductions of 0.7% to 1.5% for liraglutide, 0.6% to 0.9% for exenatide bid, and 1.2% to 1.5% for exenatide LAR.

Clinical trials show differences between the GLP-1 receptor agonists with regard to efficacy, safety, and weight effects. Specifically, the Liraglutide Effect and Action in Diabetes (LEAD) 6 trial68 was a 26-week randomized study involving 464 patients that compared addition of liraglutide (1.8 mg) or exenatide bid (10 μg) with metformin and/or glimepiride therapy. Data from this study suggest that liraglutide is more effective than exenatide in lowering HbA1c levels (1.12% vs 0.79%), with approximately 54% of patients treated with liraglutide achieving levels ≤ 7% compared with 43% of patients treated with exenatide bid. Although both induced gastrointestinal AEs, the duration was significantly shorter with liraglutide (eg, nausea observed in 3% vs 9% of patients at week 26), enabling more patients to continue treatment. Liraglutide and exenatide bid elicited similar weight loss (3.2 kg vs 2.9 kg), with low rates of hypoglycemia.68

The Effects of Exenatide Long-Acting Release on Glucose Control and Safety in Subjects With Type 2 Diabetes Mellitus (DURATION-1) trial, a 30-week study comparing exenatide LAR with exenatide bid in 295 patients with T2DM, illustrated that patients administered exenatide LAR had significantly greater reductions in HbA1c levels than those administered exenatide bid (−1.9% vs −1.5%; P = 0.0023), with a greater proportion of patients achieving target HbA1c levels of 7% (77% vs 61%; P = 0.0039). As in the LEAD-6 study, both formulations produced similar weight reductions (4.0 kg vs 3.8 kg).69 Similarly, the DURATION-5 study showed analogous trends with regard to HbA1c levels (−1.6% vs −0.9%; P < 0.0001) and body weight (−2.3 kg vs −1.4 kg) for exenatide LAR and exenatide bid, respectively.70 In both the DURATION-1 and 5 studies, gastrointestinal AEs were less likely in the exenatide LAR group; 26% and 14% of patients experienced nausea with exenatide LAR compared with 35% in both studies with exenatide bid.69,70

Finally, the DURATION-6 trial, a 26-week study with 912 patients comparing liraglutide (1.8 mg) with exenatide LAR (2 mg) added to preexisting oral antidiabetic agents, showed a greater HbA1c level change (−1.48% vs −1.28%) and weight loss (3.6 vs 2.7 kg) with liraglutide compared with exenatide LAR. Similarly, more patients achieved HbA1c levels < 7% (60% vs 52%, P = 0.008).71 However, the number of patients reporting nausea was higher with liraglutide than with exenatide LAR (20% vs 9%). Together, these trials suggest that agents providing 24-hour effects, such as liraglutide and exenatide LAR, may be more effective at reducing HbA1c levels and facilitating weight loss, as well as controlling blood glucose levels.

DPP-4 Inhibitors

Inhibitors of DPP-4, an enzyme responsible for the metabolism of GLP-1, prolong the effect of endogenous GLP-1, addressing items I, III, V, and VI of the pathophysiologic defects of patients with T2DM (Figure 1).13 Treatment with saxagliptin, sitagliptin, or linagliptin,72 3 DPP-4 inhibitors, results in HbA1c level reductions of 0.5% to 1.2%, 0.7% to 0.9%, or 0.28% to 0.37%, respectively.73-75 While sitagliptin is administered twice daily with meals,76 saxagliptin and linagliptin offer the added convenience of once-daily dosing without regard to meals.77,78 A review of 45 clinical trials found that DPP-4 inhibitors are generally well tolerated; the most frequent AEs include headache, nasopharyngitis, upper respiratory tract infection, urinary tract infection, hypertension, and dizziness.79 Moreover, serious allergic and hypersensitivity reactions have been reported, including anaphylaxis and angioedema.76-78 As with GLP-1 receptor agonists, pancreatitis is observed with DPP-4 inhibitors in postmarketing surveillance.79 Dipeptidyl peptidase-4 inhibitors have no effect on gastric emptying,80 are weight neutral or associated with limited weight gain (< 1 kg),79 and do not cause hypoglycemia.13

GLP-1 Receptor Agonists Compared With DPP-4 Inhibitors: Head-to-Head Trials

A major disadvantage of DPP-4 inhibitors is that their efficacy relies on the presence of endogenous GLP-1, which has a short half-life and limited expression in patients with T2DM.6,81 Head-to-head comparisons illustrate the superiority of GLP-1 receptor agonists in treatment efficacy and overall weight loss.

For example, a 26-week prospective trial in 665 patients treated with metformin plus either liraglutide 1.2 or 1.8 mg compared with sitagliptin 100 mg showed significantly greater reductions in HbA1c level (−1.24% and −1.4% vs −0.9%) and body weight (−2.86 kg and −3.38 kg vs −0.96 kg).82 However, treatment with liraglutide resulted in an increased incidence of transient nausea and vomiting compared with sitagliptin, although the duration of nausea was shorter with liraglutide (8 and 13 days vs 26 days). One-year follow-up showed maintenance of glycemic effects, with estimated treatment differences between liraglutide and sitagliptin in HbA1c levels of −0.40% (1.2 mg) and −0.63% (1.8 mg; P < 0.0001), with a 1.6- and 2.5-kg greater weight loss.60 Moreover, an extension of this study showed that patients who switched to liraglutide from sitagliptin also had improved glycemic control (−0.2% [1.2 mg], −0.5% [1.8 mg]; P ≤ 0.006), weight reduction (−1.6 kg [1.2 mg], −2.5 kg [1.8 mg]; P < 0.001), and improved overall treatment satisfaction (P = 0.0189) after 52 weeks.83

Similarly, a 26-week, randomized, double-blind trial assessed the safety and efficacy of exenatide LAR compared with sitagliptin in 342 patients treated with metformin, and showed that exenatide LAR treatment reduced HbA1c levels significantly more than sitagliptin (−1.5% vs −0.9%), again with a greater weight loss (−2.3 kg vs −0.8 kg).62 An extension of this study showed that patients who switched to once-weekly exenatide LAR from sitagliptin also had improved or sustained glycemic control (least squares mean HbA1c level, −0.3%; P = 0.001), with modest weight loss (least squares mean, −1.1 kg; P = 0.0006).84

Overall, GLP-1 receptor agonists demonstrate greater comparative reduction in HbA1c levels, weight loss benefit, patient satisfaction, and coverage of the 8 pathophysiologic abnormalities than DPP-4 inhibitors (Figure 1); however, which product is better in a clinical setting is still debated because of differences in the mode of delivery (ie, injectable vs oral) and incidence of gastrointestinal AEs.81,85 Patient tolerability, the extent of the patient–physician relationship, and the overall level of physician expertise are important factors to consider when deciding between these 2 incretin therapies. Increasing the comfort level of physicians when initiating injectable medications may alleviate or reduce a patient’s aversion to GLP-1 receptor agonists.

Insulin Therapy

Basal insulins, which are currently the most effective therapy for lowering HbA1c levels (1.5%–3.5%), are recommended by the ADA/EASD as add-on therapy after other metformin combination therapies are no longer effective.38,86 Basal insulin can be provided using either neutral protamine Hagedorn (NPH) human insulin or long-acting insulin analogs. Long-acting analogs, such as insulin detemir and insulin glargine, have relatively flat time–action profiles from 17.5 to 25.6 hours, compared with 12 to 14 hours for NPH insulin,87 thus allowing once-daily administration. Moreover, long-acting insulins give the patient sufficient metabolic control, with reduced hypoglycemic risk and reduced weight gain,88 which are common with other insulin therapies.38 For example, a meta-analysis of 14 randomized controlled trials showed that long-acting insulin analogs induced less hypoglycemia than NPH insulin, with insulin detemir inducing the least weight gain; however, long-acting formulations did not provide better overall glycemic control than NPH analogs, although hypoglycemic event rates were reduced.89 Similarly, a 26-week study of 505 patients with T2DM taking insulin detemir or NPH insulin showed that insulin detemir was associated with significantly less weight gain (1.0 kg vs 1.8 kg; P = 0.017).90 A direct comparison between once-daily insulin detemir and insulin glargine in a 52-week trial in patients with T2DM showed significantly less weight gain with insulin detemir (2.8 kg vs 3.8 kg; P < 0.05), with similar changes in HbA1c levels.91 The lower rates of weight gain with insulin detemir compared with insulin glargine may be due to an anorectic effect caused by preferential uptake of insulin detemir in the brain rather than peripheral tissues.92 Additionally, insulin detemir is stable for 14 days longer than insulin glargine once opened, allowing for less insulin waste, with potentially lower patient cost.92 Overall, currently available long-acting insulin analogs have similar safety, efficacy, and pharmacokinetic profiles, providing powerful adjuncts to therapy regardless of which product is chosen. However, if the goal is less weight gain, insulin detemir may have a slight advantage.

Despite the benefits of the new longer-acting formulations, as with other injectables, patient adherence remains a major issue. However, the development of pen devices and smaller, fine-gauge needles has eased some of these concerns, with some studies showing advantages of pens over syringes that include greater accuracy, ease of use, patient satisfaction, quality of life, and adherence.93

Other Medications

According to T2DM treatment guidelines, SUs may be used as add-on therapy to metformin or as an alternative for patients with contraindications to metformin.3 Sulfonylureas function by binding to the SU receptor-1 on pancreatic β cells, resulting in depolarization and calcium influx that induce insulin secretion.50 Although SUs lower HbA1c levels by up to 1.5%, they are associated with hypoglycemia, weight gain (~2 kg),38 adverse cardiovascular outcomes, and increased mortality rates.94 For example, a retrospective cohort study of patients with T2DM suggested that the SUs glipizide, glyburide, and glimepiride are associated with increased risks of overall mortality (64%, 59%, and 68%, respectively) compared with metformin (P < 0.001).94 Moreover, long-term studies over > 1.5 years suggest that SUs do not produce durable HbA1c level reductions or protect β-cell function.13 Thus, despite their low cost and inclusion in the ADA/EASD and AACE guidelines as an appropriate therapy,3,8 the AEs and lack of benefit on pathophysiology give SUs limited value in long-term management of T2DM.


Thiazolidinediones are peroxisome proliferator–activated receptor-γ agonists that address items I, III, and IV of the pathophysiologic defects of patients with T2DM13,41 presented in Figure 1, and provide HbA1c level reductions of 0.7% to 1.6%.95 Although still recommended as a treatment option in the AACE and ADA/EASD guidelines, rosiglitazone and pioglitazone induce substantial weight gain (~1.9–4.5 kg) and edema96 and are associated with increased risks for fractures,97 CVD,41,95 and bladder cancer.98 Thus, while thiazolidinediones address certain pathophysiologic mechanisms of T2DM, their AEs make them unacceptable for routine use.

Dopamine Receptor Agonists

Bromocriptine, a dopamine receptor agonist, suppresses hepatic glucose production (Figure 1, I), reduces HbA1c levels (0.55%–0.9%), is weight neutral, does not induce hypoglycemia, and reduces cardiovascular risk.99 For example, in a 52-week trial in 3095 patients, bromocriptine (monotherapy or combination) reduced the risk of CVD by 40% (via an unknown mechanism), and showed rates of hypoglycemia and weight gain similar to those of placebo.100 While the cardiovascular data are intriguing, the high cost and limited glucose response rate of bromocriptine restrict its value.

Sodium-Glucose Co-Transporter 2 Inhibitors

Sodium-glucose co-transporter 2 (SGLT2) reabsorbs glucose from the glomerular filtrate in the renal proximal tubule. Thus, SLGT2 inhibitors (eg, canagliflozin, dapagliflozin, and empagliflozin) block glucose reabsorption and promote glycosuria (Figure 1, VII), resulting in weight loss.41,101 Although these agents have limited effects on the pathophysiologic defects in patients with T2DM, the mechanism of action is independent of insulin, making it potentially easier to combine SGLT2 inhibitors with other drugs. In a 12-week trial, dapagliflozin reduced HbA1c levels by −0.55% to −0.90% compared with placebo, with weight loss of 2.5% to 3.4%.102 In a 24-week trial of 182 patients with T2DM being treated with metformin, addition of dapagliflozin resulted in weight loss of –2.96 kg compared with −0.88 kg with placebo (P < 0.0001), in addition to a −0.28% greater HbA1c level reduction. Vulvovaginitis, balanitis, and lower urinary tract infections were classified as AEs and were observed more frequently in patients taking dapagliflozin than in those taking placebo.103 If further studies verify their efficacy and tolerability, their unique mechanism of action and weight loss potential will make SGLT2 inhibitors a promising new T2DM treatment option.

Emerging Treatments

Established therapies provide a variety of choices when choosing T2DM treatment regimens; however, potentially superior treatments are currently under development.

Ultra–Long-Acting Insulin Analogs

Currently in review by the US Food and Drug Administration, insulin degludec is an ultra–long-acting insulin that after subcutaneous injection forms soluble multihexamers that slowly and continuously release insulin into the bloodstream. This property gives insulin degludec a stable plasma profile at steady state and a half-life of > 25 hours, twice that of insulin glargine, and a duration of action > 40 hours.104 The 52-week, treat-to-target Switching From Glargine to Insulin Degludec in Subjects With Type 2 Diabetes Mellitus (BEGIN) study, which assessed the efficacy and safety of insulin degludec compared with insulin glargine in patients with T2DM, showed similar HbA1c level reductions (−1.10% with insulin degludec vs −1.18% with insulin glargine), but significantly lower rates of nocturnal hypoglycemia with insulin degludec, suggesting that it may offer a better-tolerated option for patients who require a basal insulin.105

Weight and Physiologic Treatment Recommendations

Although current consensus committee guidelines established by the ADA/EASD, ACP, and AACE/ACE cover nearly all treatments options available for T2DM in depth, they do not provide simple guiding principles for physicians to manage patients on a day-to-day basis. These principles include normalization of physiology, facilitation of weight loss, and limitation of hypoglycemic risk. Thus, recommendations based on clinical experience, in addition to key studies, are included herein. These recommendations advocate a patient-centric management approach that utilizes specific diet, exercise, and drug intervention strategies to simultaneously address the core pathophysiologic defects in patients with T2DM and obesity (Figure 3).

View: Figure 3
The MGI treatment approach: weight and pathophysiologic treatment recommendations.

*DPP-4 inhibitors can replace GLP-1 receptor agonists in injection-averse patients.

Long-acting SUs can replace long-acting insulin in injection-averse patients.

The borders of each step are imprecise, depending on comorbidities; the pathophysiologic defects (see Figure 1) addressed at each stage are indicated. The extent of “regression” along the continuum will also vary from patient to patient.

Abbreviations: DPP-4, dipeptidyl peptidase-4; GLP-1 RA, glucagon-like peptide-1 receptor agonist; HbA1c, glycated hemoglobin; MGI, metformin, glucagon-like peptide-1 receptor agonist, and insulin; SU, sulfonylurea.

Lifestyle Intervention Recommendations

Typically, successful lifestyle intervention therapy involves 3 core features: 1) a healthy, mainly plant-based diet that may optionally include healthy animal proteins and is high in vegetables, fruits, and whole grains; 2) regular physical activity combining both aerobic (≥ 30 minutes, 5 days/week) and resistance (≥ 20 minutes, 4–5 days/week) exercises, similar to the Diabetes Prevention Program and ADA/American College of Sports Medicine goals; and 3) frequent intervention from dietitians and lifestyle coaches.

It is necessary to adjust diet and exercise routines according to an individual’s needs. Thus, aerobic exercises should be tailored to an individual’s preference and overall comfort level. For example, aerobic exercise could include brisk walking, aerobic dance, skating, bicycle riding, or swimming. Resistance training should be of a moderate-to-vigorous intensity, targeting chest/triceps, back/biceps, legs, and shoulders on different days, with the abdomen being targeted 2 to 3 times per week (3 sets of 10–15 repetitions to fatigue). Limiting training to focused areas each day may help to increase the intensity at each session, improve strength and muscle mass, and facilitate patient adherence. Weights should be 60% of maximum capacity or enough weight to exhaust muscle within 20 repetitions. For patients who cannot perform aerobic exercise, a resistance exercise program should be completed. Because the average patient pays approximately $650 per month for diabetes care,106 physicians may help motivate patients to adhere to lifestyle intervention programs by explaining that these changes may reduce or even eliminate the need to take costly T2DM medications.

Drug Interventions

In patients for whom lifestyle intervention is not sufficient (in my opinion, in those with an HbA1c level > 5.7%), drug intervention is usually necessary.107 Some important decision points to consider when choosing a drug include overall efficacy (reduce HbA1c level, addressing pathophysiologic defects), safety (hypoglycemia, AEs, contraindications, CV risk factors), patient acceptance (mode of administration, tolerability, dosing regimen), mechanism of action (overlap with other therapies), and, importantly, weight effects.

In addition to these important points, cost is a significant issue when selecting treatments, especially considering the high price of incretin therapies. Thus, physicians should try to help patients understand that the benefits of incretin therapies may lead to reduced long-term costs for diabetes care. For example, studies suggest that the annual total cost of diabetes-related medical care with exenatide was significantly lower than that of insulin glargine ($7833 vs $8536; P < 0.001),108 including a substantially lower cost related to hypoglycemia ($7812 vs $19 637).109 Similarly, evidence indicates that liraglutide may reduce the total cost of diabetes-related care compared with glimepiride110 and sitagliptin.111 Moreover, costs may be covered by a patient’s insurance or manufacturer rebates and be of little direct cost to the patient. Despite long-term cost benefits, some patients may wish to select medications that are more immediately cost-effective, such as SUs, despite serious safety concerns. Thus, in addition to prescribing pharmacotherapy, physicians should discuss the implications of each therapeutic option with their patients.

The MGI Protocol

Figure 3 shows my treatment recommendations, arranged on a continuum of medications selected on the basis of HbA1c levels, comorbidities, and targets set by the patient and diabetes team. While HbA1c level triggers are used, the end targets should be individualized based on patient goals and health, and therapeutic initiatives should be considered as a continuum of options used in achieving individual targets. While the MGI protocol works for the majority of patients, it does not necessarily provide a solution for every patient. For example, in my experience, a small percentage (< 5%) of patients may require a short-acting insulin before meals; such exceptions are not comprehensively included in this discussion.

HbA1c Level 5.7% to 7%

Metformin (or a GLP-1 receptor agonist for patients with contraindications to metformin) should be added to the aggressive lifestyle-modification program described previously. Additionally, for patients with an HbA1c level of 6.5% to 7% who have weight-dependent comorbidities, such as sleep apnea, a GLP-1 receptor agonist is a consideration.

HbA1c Level 7% to 8%

To reverse the pathologies of T2DM as early as possible, patients with HbA1c levels ≥ 7% should be treated with a GLP-1 receptor agonist in combination with lifestyle intervention and metformin, in accordance with the AACE guidelines.8 For patients who are sensitive to gastrointestinal discomfort, consider dosage reduction, switching to a longer-acting product, or changing to a DPP-4 inhibitor.

HbA1c Level > 8%

If HbA1c levels are > 8% (or above the target set by the diabetes care team) in patients already treated with lifestyle intervention, metformin, and GLP-1 receptor agonists, a basal, long-acting insulin (or potentially an ultra–long-acting insulin in the future) should be added to optimize glycemic control. Patients should be started at a bedtime dose of 0.2 units/kg and titrated by 1 unit/day until morning blood glucose levels are < 120 mg/dL. Use of metformin, a GLP-1 receptor agonist, and long-acting insulin at diagnosis in patients with an HbA1c level > 10% is an excellent choice to control glucose and facilitate weight loss with minimal chance of hypoglycemia. In the future, it may also be acceptable to add an SGLT2 inhibitor.

Alternative Therapies for Injection-Averse Patients

The use of injectable medication is a concern for some patients. Thus, before initiating GLP-1 receptor agonist or insulin therapy, it is important for physicians to work to alleviate these concerns. This requires clinician confidence in the modality, coupled with mutual trust between the clinician and patient. Provider confidence depends on experience, awareness of the minimal discomfort associated with newer pen delivery systems, and knowledge of the significant potential benefit of GLP-1 receptor agonists. Patients should also be aware that, unlike insulin therapy, with appropriate sustained and consistent adherence to diet and exercise programs, withdrawal of treatment with GLP-1 receptor agonists may be feasible. Moreover, it is important to offer patients an alternative if they do not wish to initiate/continue injection therapy. Nevertheless, some patients may refuse to administer medications via injection. In these cases, in conjunction with lifestyle intervention and metformin, patients with an HbA1c level > 7% should be offered a DPP-4 inhibitor in place of the GLP-1 receptor agonist, and patients with an HbA1c level > 8% may benefit from a long-lasting SU instead of insulin therapy.

Case Study 1

J. T. is a 232-lb (105.5-kg), 5′3″ (160-cm), 43-year-old woman who has had T2DM for 3 years, with concomitant sleep apnea, hypothyroidism, hypertension, and hypercholesterolemia. She states that she has gained an average of 3 to 5 lb each year for the past 5 years, and “has never been able to lose weight.” Her current medications include metformin and glimepiride, as well as 2 medications to lower her blood pressure and cholesterol levels. Fasting blood glucose level readings range from 200 to 230 mg/dL and her HbA1c level is 8.7%. J. T. is counseled on diet and asked to begin walking for 20 minutes per day, which will increase as she becomes able to take on more exercise. Her metformin is continued, glimepiride is stopped, and liraglutide is started and titrated to 1.2 mg after 1 week. She seems eager to use a medication that will help with weight loss and lifestyle changes. After 6 weeks, J. T. has lost 6 lb, her fasting blood glucose level readings range from 150 to 180 mg/dL, and her HbA1c level is 8%. She gave up all sugary drinks, and is ready to begin resistance training. She is optimistic about continuing to lose weight now that her appetite is better controlled.

Case Study 2

M. R. is a 5′10″ (178-cm), 260-lb (118-kg), 51-year-old man newly diagnosed with T2DM. Complaining of excess thirst and tingling in his feet, M. R. was seen 1 day earlier in an urgent care facility and was started on metformin 500 mg daily. His blood glucose level is 342 mg/dL and HbA1c level is 11.8%. M. R. is immediately given 0.6 mg of liraglutide and 0.2 units/kg insulin detemir (24 U) and instructed to take an additional 24 units that night and increase detemir by 1 unit each night until his morning blood glucose level is < 120 mg/dL. He is told that it will likely require approximately 50 units for success. He is advised to increase his dose of metformin to 1000 mg bid over 2 weeks as tolerated. Counseling on diet and exercise has begun. He states that he will switch to whole grains and reduce excess butter consumption. Prior to leaving the office, a recheck of his blood glucose indicates that it has decreased to 270 mg/dL. Three days later, M. R. feels markedly better other than some mild nausea. His blood glucose level is 200 mg/dL, his excess thirst has resolved, and he has lost 2 lb. At this time, he is advised to increase his liraglutide dose to 1.2 mg in 3 days, if his nausea has resolved. Now on 25 units of insulin detemir, M. R. is reminded to continue to increase the detemir dose by 1 unit every day until his morning blood glucose levels are < 120 mg/dL and to continue to increase the metformin dose. Six weeks later, he returns to the clinic 6 lb (2.7 kg) lighter, feeling markedly better, with morning blood glucose levels of 120 to 150 mg/dL and an HbA1c level of 8.7%. He states that he has joined a gym.


Collectively, these recommendations represent a treatment paradigm shift in which immediately following lifestyle intervention, concomitant therapy is initiated to target the 8 pathophysiologic defects in patients with T2DM and the issue of weight by following a metformin, GLP-1 receptor agonist, and long-acting basal insulin protocol (Figure 3). Weight loss not only decreases T2DM symptoms and associated comorbidities such as sleep apnea,4,112 but may also give patients a better outlook on life. Patients who are losing weight as a result of medications are more likely to continue taking the medications and losing weight. Thus, in addition to diet and exercise, combination treatments work in concert to promote weight loss and target many of the 8 pathophysiologic defects in patients with T2DM to restore normal physiology.


Medical writing support was provided by Nicole Gudleski, PhD, at Complete Publication Solutions, LLC; this support was funded by Novo Nordisk A/S.

Open-access expire date: 
Tue, 06/24/2014