Thiazolidinediones (TZDs), also called glitazones, are a class of oral antidiabetic agents that act as agonists at PPAR-gamma (peroxisome proliferator-activated receptor gamma), a ligand-activated nuclear transcription factor expressed predominantly in adipose tissue but also in skeletal muscle, liver, macrophages, and vascular endothelium. By activating PPAR-gamma, TZDs reprogram cellular metabolic gene expression toward enhanced insulin sensitivity rather than stimulating insulin secretion, making them the only oral antidiabetic class that directly targets insulin resistance at the nuclear transcriptional level. This mechanism produces durable glycemic improvement, favorable effects on lipid partitioning, and anti-inflammatory properties, but also produces the class-defining adverse effects of fluid retention, weight gain, and bone fragility, and generated one of the most consequential drug safety controversies in recent endocrinology history.
PPAR-gamma belongs to the nuclear receptor superfamily and functions as a ligand-dependent transcription factor that, upon activation, heterodimerizes with RXR (retinoid X receptor) and binds to PPRE (PPAR response elements, specific DNA sequences) in the promoters of target genes. In adipocytes, PPAR-gamma activation drives adipogenesis, fatty acid uptake and storage, and the upregulation of FABP4 (fatty acid-binding protein 4, also called aP2), adiponectin, and GLUT4 (glucose transporter type 4). The net metabolic effect in adipose tissue is a redistribution of lipid from ectopic depots in liver and skeletal muscle (where free fatty acids impair insulin signaling) toward subcutaneous adipose tissue (where lipid is safely stored), reducing lipotoxicity and improving insulin receptor signaling in peripheral organs. In skeletal muscle, PPAR-gamma activation increases GLUT4 expression and translocation, enhancing insulin-stimulated glucose uptake. In the liver, TZDs reduce hepatic glucose production by suppressing gluconeogenic gene expression and improving hepatic insulin sensitivity, though this effect is secondary to peripheral fat redistribution rather than a primary hepatic action.1
Two TZDs are currently available: pioglitazone and rosiglitazone. A third agent, troglitazone, was withdrawn worldwide in 2000 due to idiosyncratic hepatotoxicity. Pioglitazone activates both PPAR-gamma and, to a lesser degree, PPAR-alpha (peroxisome proliferator-activated receptor alpha, the hepatic lipid-oxidation receptor), producing a more favorable lipid profile than rosiglitazone: pioglitazone raises HDL-C (high-density lipoprotein cholesterol) modestly and lowers triglycerides, while rosiglitazone raises both LDL-C (low-density lipoprotein cholesterol) and HDL-C without the triglyceride-lowering effect. Both agents are well absorbed orally, highly protein-bound (greater than 99 percent), and extensively metabolized by CYP2C8 (cytochrome P450 2C8) and CYP3A4 (cytochrome P450 3A4); their active metabolites contribute to pharmacological effect. Onset of maximal glycemic effect is delayed 6 to 12 weeks, reflecting the requirement for PPAR-gamma-driven changes in gene expression and protein synthesis, a distinctly different time course from secretagogues or metformin, with durability of glycemic effect shown in long-term comparative trials.215
The cardiovascular safety trajectories of pioglitazone and rosiglitazone diverged dramatically in the 2000s. Rosiglitazone's cardiovascular risk was raised by a 2007 meta-analysis showing a statistically significant increase in myocardial infarction risk, which was subsequently confirmed in the RECORD (Rosiglitazone Evaluated for Cardiovascular Outcomes in Oral Agent Combination Therapy for Type 2 Diabetes) trial. The FDA imposed severe prescribing restrictions on rosiglitazone in 2010, subsequently partially lifted in 2013 after additional data review, but rosiglitazone remains rarely used clinically. Pioglitazone, by contrast, showed cardiovascular benefit in the PROactive (Prospective Pioglitazone Clinical Trial in Macrovascular Events) trial — a reduction in the composite secondary endpoint of all-cause mortality, MI (myocardial infarction), and stroke — though the primary composite endpoint missed statistical significance. Pioglitazone additionally reduces the risk of progression from impaired glucose tolerance to T2DM (type 2 diabetes mellitus) and shows a cardiovascular-protective signal in secondary prevention. The key divergence is mechanistic: pioglitazone's dual PPAR-alpha/gamma agonism improves the atherogenic lipid profile; rosiglitazone's more selective PPAR-gamma action does not.3
Fluid retention is a class effect of TZDs arising from PPAR-gamma-mediated upregulation of the epithelial sodium channel ENaC (epithelial sodium channel) in the collecting duct of the kidney, increasing renal sodium reabsorption and expanding extracellular fluid volume. This mechanism is independent of blood pressure and aldosterone. Clinically, TZDs cause weight gain of 2 to 5 kg (fluid and fat), peripheral edema in 5 to 15 percent of patients, and a 2 to 3-fold increased risk of heart failure hospitalization. TZDs are contraindicated in NYHA (New York Heart Association) class III or IV heart failure and should be used with extreme caution in NYHA class I or II. They worsen macular edema in patients with pre-existing diabetic retinopathy and should be avoided in that setting. The bone fracture risk associated with TZDs is significant: both pioglitazone and rosiglitazone suppress osteoblast differentiation (via PPAR-gamma activation in bone marrow mesenchymal stem cells) and stimulate osteoclastogenesis, increasing fracture risk particularly at distal extremity sites (wrist, foot, ankle) in postmenopausal women. Fracture risk reduction strategies include adequate calcium and vitamin D supplementation, avoidance of TZDs in patients already on bisphosphonates or with low bone density, and preference for alternative agents in higher-risk patients.11
Observational studies and a French nationwide cohort study reported an elevated bladder cancer risk with pioglitazone use exceeding 24 months, leading France to withdraw pioglitazone in 2011. The FDA added a bladder cancer warning to pioglitazone prescribing information. Subsequent meta-analyses showed a modest signal (relative risk approximately 1.2 to 1.4) that was not confirmed in all datasets, leading to substantial ongoing debate. Current consensus: pioglitazone should be avoided in patients with active bladder cancer or uninvestigated hematuria, and used with caution in patients with a history of bladder cancer. The benefit-risk calculation is generally favorable in patients with established ASCVD (atherosclerotic cardiovascular disease), non-alcoholic steatohepatitis, or polycystic ovary syndrome where evidence of pioglitazone benefit is well-established.
DPP-4 (dipeptidyl peptidase-4) inhibitors, also called gliptins, enhance the action of endogenous incretin hormones by blocking the enzyme responsible for their rapid degradation. The incretin system amplifies glucose-stimulated insulin secretion in a glucose-dependent manner — meaning the incretin-potentiated secretion is extinguished as blood glucose normalizes — which gives DPP-4 inhibitors a categorically safer hypoglycemia profile than sulfonylureas. This glucose-dependence, combined with weight neutrality, good tolerability, and renal dosing adjustability, has established DPP-4 inhibitors as a widely used second-line oral agent across all major T2DM (type 2 diabetes mellitus) guideline frameworks.
GLP-1 (glucagon-like peptide-1) and GIP (glucose-dependent insulinotropic polypeptide, also called gastric inhibitory polypeptide) are the two primary incretin hormones. GLP-1 is secreted by enteroendocrine L cells in the distal ileum and colon in response to nutrient ingestion; GIP is secreted by K cells in the proximal duodenum and jejunum. Both act on their cognate G protein-coupled receptors on pancreatic beta cells to stimulate glucose-dependent insulin secretion, meaning they potentiate insulin release only when blood glucose is elevated. GLP-1 additionally suppresses glucagon secretion from alpha cells (in a glucose-dependent manner), delays gastric emptying, and reduces appetite via central GLP-1 receptor activation in the hypothalamus and brainstem. GIP acts primarily on beta cells and adipocytes; its glucagon-suppressing and gastric-slowing effects are weaker than GLP-1. Both GLP-1 and GIP are rapidly inactivated by DPP-4, a serine protease expressed on endothelial cells, lymphocytes, macrophages, and epithelial cells that cleaves the N-terminal dipeptide from GLP-1 and GIP within 1 to 2 minutes of secretion, reducing their plasma half-life to 2 to 3 minutes. Physiological postprandial GLP-1 and GIP levels are therefore largely captured within the portal vein before reaching the systemic circulation in biologically active form.4
DPP-4 inhibitors competitively inhibit the DPP-4 enzyme, extending the half-life of endogenous GLP-1 and GIP from 2 to 3 minutes to 6 to 8 minutes and raising their postprandial plasma concentrations approximately 2-fold. This modest pharmacodynamic augmentation (relative to the 5 to 8-fold GLP-1 elevation achievable with injectable GLP-1 receptor agonists) explains why DPP-4 inhibitors produce more modest HbA1c reductions (0.5 to 1.0 percent) compared to GLP-1 receptor agonists (1.0 to 2.0 percent or greater) and do not produce the significant weight loss, gastric emptying slowing, or appetite reduction characteristic of pharmacological GLP-1 receptor agonism. The glucose-dependent mechanism means DPP-4 inhibitor-mediated insulin secretion ceases as glucose normalizes, explaining the low monotherapy hypoglycemia rate. DPP-4 also cleaves multiple endogenous substrates including stromal cell-derived factor-1 (SDF-1/CXCL12), BNP (B-type natriuretic peptide), neuropeptide Y, and substance P; the pharmacological consequences of blocking cleavage of these substrates are incompletely understood but may contribute to some off-target effects of DPP-4 inhibition, including possible immune modulation.5
Five DPP-4 inhibitors are approved in the United States: sitagliptin, saxagliptin, alogliptin, linagliptin, and trelagliptin (weekly dosing, Japan-approved). All are once-daily oral agents except trelagliptin. Sitagliptin was the first approved (2006) and has the largest safety database. All agents in the class require renal dose adjustment except linagliptin, which is eliminated primarily by biliary/fecal excretion unchanged and requires no renal dose adjustment — making it the preferred gliptin in CKD (chronic kidney disease) of any severity. Sitagliptin, saxagliptin, and alogliptin are primarily renally excreted and require progressive dose reduction as eGFR (estimated glomerular filtration rate) falls. Sitagliptin can be used at 25 mg once daily (versus the standard 100 mg) in patients with eGFR 15 to 30 mL/min/1.73m2, including those on dialysis. Saxagliptin requires dose reduction (2.5 mg) when eGFR falls below 45 mL/min/1.73m2 and is avoided in dialysis. Drug interactions are limited: saxagliptin is a CYP3A4 (cytochrome P450 3A4) substrate and requires dose reduction when co-administered with potent CYP3A4 inhibitors; the others are not significantly metabolized by cytochrome P450 enzymes and have minimal interaction profiles.6
Cardiovascular outcome trials (CVOTs) required by the FDA for all new T2DM agents since 2008 have demonstrated cardiovascular non-inferiority for all gliptins versus placebo when added to standard care. The Saxagliptin Assessment of Vascular Outcomes Recorded in Patients with Diabetes Mellitus trial (SAVOR-TIMI 53) and EXAMINE (Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care) trial showed non-inferiority for the primary cardiovascular endpoint but raised the notable safety concern of heart failure hospitalization: saxagliptin increased heart failure hospitalization by 27 percent compared with placebo in SAVOR-TIMI 53 (hazard ratio 1.27; p=0.007), a signal also seen with alogliptin in EXAMINE. Sitagliptin in the TECOS (Trial Evaluating Cardiovascular Outcomes with Sitagliptin) trial showed no increase in heart failure hospitalization. Linagliptin in the CAROLINA (CARdiovascular Outcome trial of LINAgliptin versus Glimepiride) and CARMELINA (Cardiovascular and Renal Microvascular Outcome Study with Linagliptin) trials showed cardiovascular non-inferiority without excess heart failure signal. The heart failure signal with saxagliptin and alogliptin likely relates to inhibition of DPP-4 cleavage of SDF-1 and BNP, altering myocardial biology, though the exact mechanism remains under investigation. The practical clinical implication is that saxagliptin and alogliptin should be used with caution in patients with established heart failure, while sitagliptin and linagliptin have a more reassuring heart failure safety profile.7
Alpha-glucosidase inhibitors (AGIs) work at the intestinal brush border to slow the digestion and absorption of complex carbohydrates, attenuating postprandial glucose excursions without stimulating insulin secretion or altering insulin sensitivity. Their mechanism is the simplest of any antidiabetic drug class — competitive inhibition of an enzyme — and produces a predictable and dose-dependent pharmacodynamic effect. They do not cause hypoglycemia as monotherapy, do not affect weight, and have essentially no systemic pharmacological activity, making them conceptually among the safest antidiabetic agents. Their clinical utility is limited in practice almost entirely by dose-dependent gastrointestinal adverse effects that render them poorly tolerated in Western populations, though they remain more widely used in East Asian countries where lower-carbohydrate meal compositions produce less gas and bloating.
Dietary complex carbohydrates (starch, sucrose, maltose, lactose) are not absorbed directly across the intestinal epithelium; they must first be hydrolyzed to monosaccharides by alpha-glucosidases located on the brush border of small intestinal enterocytes. The principal brush-border alpha-glucosidases are maltase, sucrase, isomaltase, and glucoamylase; pancreatic amylase provides proximal starch hydrolysis. Acarbose and miglitol competitively inhibit these alpha-glucosidase enzymes by binding reversibly at the active site with higher affinity than their natural oligosaccharide substrates. The result is delayed and reduced monosaccharide generation in the proximal intestine, shifting carbohydrate absorption to the more distal small intestine and colon. This attenuation of the rate of glucose appearance in the portal circulation reduces postprandial glucose excursions by 40 to 60 mg/dL on average, without altering fasting glucose significantly. The preferential effect on postprandial glucose makes AGIs particularly valuable as add-on therapy in patients with adequate fasting glucose control but persistent elevated postprandial values. The undigested oligosaccharides reaching the colon are fermented by colonic bacteria, producing short-chain fatty acids and gas — the mechanistic basis of the flatulence, bloating, diarrhea, and abdominal discomfort that characterize AGI (alpha-glucosidase inhibitor) use.8
Acarbose is a pseudotetrasaccharide derived from Actinoplanes utahensis fermentation. It has essentially no systemic absorption (less than 2 percent of the dose reaches the systemic circulation as parent drug), acting entirely at the intestinal brush border as a luminal agent. Degradation products generated by bacterial cleavage in the colon are absorbed and excreted renally, which is why acarbose is contraindicated in patients with significant renal impairment (serum creatinine above 2.0 mg/dL or eGFR below 30 mL/min/1.73m2), even though the parent drug is not renally cleared; the metabolites can accumulate. Miglitol, unlike acarbose, is substantially absorbed systemically (bioavailability approximately 50 to 70 percent at low doses) and is excreted unchanged by the kidneys, making it similarly contraindicated in significant renal impairment. However, miglitol has no systemic pharmacological effects at circulating concentrations and inhibits alpha-glucosidases exclusively at the intestinal brush border during absorption; the systemic exposure is pharmacologically inert. Both agents should be administered with the first bite of each main meal, as the competitive inhibition must be present when carbohydrate substrates arrive at the brush border.8
Monotherapy HbA1c reduction with AGIs is modest, approximately 0.5 to 1.0 percent, lower than sulfonylureas, metformin, or DPP-4 (dipeptidyl peptidase-4) inhibitors. Their most compelling clinical efficacy evidence comes from the STOP-NIDDM (Study to Prevent Non-Insulin-Dependent Diabetes Mellitus) trial, in which acarbose reduced the relative risk of progression from impaired glucose tolerance to T2DM (type 2 diabetes mellitus) by 25 percent and, as a secondary finding, reduced the risk of cardiovascular events.9 The cardiovascular signal from STOP-NIDDM was not the primary endpoint and was not replicated in a dedicated cardiovascular outcomes trial, limiting its clinical weight. In patients already taking insulin or a sulfonylurea, AGIs can cause hypoglycemia by reducing glucose absorption below the threshold maintained by the secretagogue; such hypoglycemia cannot be treated with sucrose or starch (which the AGI prevents from being absorbed) and must be treated with glucose tablets or gel. This is a critical prescribing education point for patients combining alpha-glucosidase inhibitors with secretagogues or insulin.8
Alpha-glucosidase inhibitors are prescribed substantially more in East Asian countries (Japan, China, Korea) than in Western countries. Several pharmacological and dietary factors explain this: East Asian dietary patterns contain higher proportions of rice, noodles, and other complex carbohydrates with lower fat and protein content, producing a meal composition where postprandial glucose excursions are more pronounced and more amenable to alpha-glucosidase inhibition. Additionally, lower daily dietary fat intake reduces colonic fermentation gas production, resulting in substantially lower rates of flatulence and GI adverse effects compared with typical Western dietary patterns. The result is that AGIs are better tolerated and more clinically relevant in East Asian T2DM management, where they occupy a more prominent guideline position than in North American or European frameworks.
Understanding the adverse effect profiles of these three drug classes requires distinguishing class effects from agent-specific effects, and separating mechanistically plausible risks from those supported by robust outcome trial evidence. The safety landscape has been shaped by post-marketing surveillance, dedicated cardiovascular outcome trials, and regulatory actions that are as instructive for what was not found as for what was.
Heart failure risk is the most clinically significant safety concern in this drug group, and the risk varies markedly by agent. TZDs carry a well-established class-effect risk of heart failure hospitalization driven by renal sodium retention and volume expansion, with a relative risk approximately 2 to 3-fold above background and absolute risk approximately 1 to 3 additional HF (heart failure) hospitalizations per 100 patient-years. This risk is contraindicated in NYHA (New York Heart Association) class III/IV (severe to end-stage heart failure) and warrants extreme caution in class I/II. Among DPP-4 (dipeptidyl peptidase-4) inhibitors, saxagliptin shows a confirmed 27 percent increase in HF hospitalization in the Saxagliptin Assessment of Vascular Outcomes (SAVOR-TIMI 53) cardiovascular (CV) outcomes trial (p=0.007), and alogliptin a directionally similar but statistically non-significant signal in the EXAMINE (alogliptin CV outcomes) trial. Sitagliptin in the TECOS (Trial Evaluating Cardiovascular Outcomes with Sitagliptin) study and linagliptin (CARMELINA) show no excess HF signal.14 The mechanism underlying the saxagliptin signal may involve DPP-4 substrate effects on natriuretic peptides and cardiac remodeling pathways, but has not been definitively established. For patients with established HF or high HF risk, sitagliptin or linagliptin are preferred within the gliptin class; saxagliptin and alogliptin should be avoided or used with monitoring.7
The pancreatitis signal associated with incretin-based therapies (DPP-4 inhibitors and GLP-1 receptor agonists) generated intense regulatory and scientific scrutiny from 2013 to 2016. Initial case reports and autopsy studies suggested pancreatic ductal metaplasia with exocrine GLP-1 receptor activation, and pharmacovigilance databases showed disproportionate pancreatitis reporting with incretins. However, the dedicated cardiovascular outcome trials (SAVOR-TIMI 53, EXAMINE, TECOS, ELIXA) collectively enrolled over 45,000 patients and showed no statistically significant increase in confirmed acute pancreatitis with DPP-4 inhibitors or liraglutide compared with placebo.10 The FDA concluded that the pancreatitis risk does not represent a class-level pharmacological hazard, though it remains listed as a precaution given the mechanistic plausibility and the pre-existing association of T2DM (type 2 diabetes mellitus) itself with pancreatitis risk. Clinically, DPP-4 inhibitors and GLP-1 agents should be discontinued and not restarted if acute pancreatitis is confirmed, but they need not be avoided in patients with a history of pancreatitis without ongoing pancreatic disease, according to current ADA (American Diabetes Association) guidance.11
TZD (thiazolidinedione) bone fracture risk is a class effect with well-established epidemiological and mechanistic support. The PROactive trial showed a significant increase in fracture rates with pioglitazone versus placebo. Large cohort studies confirm a 40 to 60 percent increased fracture risk at distal extremity sites (wrist, foot, ankle) in postmenopausal women on TZDs, with a more modest and less consistent signal in men. The mechanism involves PPAR-gamma (peroxisome proliferator-activated receptor gamma) activation in bone marrow mesenchymal stem cells, which favors adipocyte over osteoblast differentiation, reducing bone formation rate while PPAR-gamma also suppresses osteoprotegerin expression, increasing osteoclast activity and bone resorption. The result is a reduction in bone mineral density, particularly at trabecular sites. Management: TZDs should be avoided in patients with osteoporosis or prior fragility fracture; in patients who require TZD therapy, adequate calcium (1,000 to 1,200 mg/day) and vitamin D (1,500 to 2,000 IU/day) supplementation and monitoring of bone density are appropriate. Pioglitazone is preferred over rosiglitazone for overall benefit-risk, but both carry the same bone fracture liability.2
Drug interactions relevant to this drug group are summarized by class. TZDs are CYP2C8 substrates; gemfibrozil (a potent CYP2C8 inhibitor) increases pioglitazone and rosiglitazone AUC (area under the curve) substantially, increasing adverse effect risk; rifampin (CYP2C8 inducer) reduces TZD exposure. Saxagliptin and alogliptin are CYP3A4 (cytochrome P450 3A4) substrates; potent CYP3A4 inhibitors (ketoconazole, clarithromycin, ritonavir) require dose reduction of saxagliptin from 5 mg to 2.5 mg. Sitagliptin and linagliptin have minimal clinically significant drug interactions. AGIs can reduce the absorption of digoxin and metformin if co-administered simultaneously; patients on digoxin should have serum digoxin levels monitored when alpha-glucosidase inhibitor (AGI) therapy is initiated. The combination of AGIs with insulin or sulfonylureas requires patient education on treating hypoglycemia with glucose only (tablets, gel) rather than sucrose-containing foods because sucrose absorption will be inhibited by the AGI, leading to inadequate glucose recovery.9
Troglitazone was the first TZD approved (1997) and was withdrawn from all markets by 2000 after approximately 100 deaths from acute fulminant hepatic failure. Unlike pioglitazone and rosiglitazone, troglitazone contains a vitamin E (alpha-tocopherol) moiety that undergoes oxidative metabolism to a reactive quinone intermediate capable of forming protein adducts and causing mitochondrial dysfunction in susceptible individuals. The hepatotoxicity was idiosyncratic (not dose-related at therapeutic doses) and affected approximately 1 in 50,000 patients, making it undetectable in pre-approval clinical trials. Pioglitazone and rosiglitazone lack the tocopherol moiety and do not generate the toxic metabolite; post-marketing surveillance and large outcome trials have not identified a hepatotoxicity signal for these agents. The troglitazone case remains a benchmark example of structural metabolism-based toxicity that escapes pre-marketing detection.
The current ADA/EASD (American Diabetes Association/European Association for the Study of Diabetes) guideline framework positions antidiabetic agents primarily according to comorbidity-driven benefit rather than HbA1c efficacy rank. Within this framework, TZDs (thiazolidinediones), DPP-4 (dipeptidyl peptidase-4) inhibitors, and AGIs (alpha-glucosidase inhibitors) occupy differentiated but distinctly secondary roles defined by their adverse effect profiles, specific patient subgroup benefits, and position after agents with proven cardiovascular or cardiorenal outcome benefits. Understanding the specific niches where each class excels is more useful clinically than attempting to rank them by glycemic potency alone.
Pioglitazone occupies a unique niche among oral antidiabetics because of well-established benefits beyond glycemia in three specific conditions: non-alcoholic steatohepatitis (NASH), where it reduces hepatic steatosis and inflammation in randomized trials and is recommended in the ADA/AASLD (American Association for the Study of Liver Diseases) guidelines for T2DM (type 2 diabetes mellitus) patients with biopsy-proven NASH; PCOS (polycystic ovary syndrome) with T2DM or impaired glucose tolerance, where PPAR-gamma (peroxisome proliferator-activated receptor gamma)-mediated reduction in hyperinsulinemia and androgen excess provides metabolic and hormonal benefit; and secondary prevention in patients with established cerebrovascular disease, where the IRIS (Insulin Resistance Intervention after Stroke) trial demonstrated a 24 percent reduction in the primary composite endpoint of stroke or MI (myocardial infarction) in insulin-resistant patients with recent TIA (transient ischemic attack) or stroke randomized to pioglitazone versus placebo.13 In patients with T2DM and established ASCVD (atherosclerotic cardiovascular disease), pioglitazone may be considered as add-on to metformin when SGLT-2 (sodium-glucose cotransporter-2) inhibitors or GLP-1 (glucagon-like peptide-1) receptor agonists are not accessible or tolerated, though it is no longer the preferred agent for established CV (cardiovascular) disease given the weight of evidence favoring the newer classes. Rosiglitazone should generally be avoided in clinical practice given its unfavorable CV profile and the availability of pioglitazone with a better-characterized safety record.3
DPP-4 inhibitors occupy a well-defined niche as second- or third-line agents in patients where hypoglycemia avoidance and weight neutrality are primary concerns. They are particularly well-suited to: elderly patients with T2DM in whom hypoglycemia carries high risk of falls and cardiovascular events; patients with CKD (chronic kidney disease) in whom metformin is contraindicated or dose-limited, since gliptins are dose-adjustable across all CKD stages (linagliptin without any adjustment); patients with cardiac disease where saxagliptin and alogliptin are avoided but sitagliptin or linagliptin represent safe alternatives; and patients who cannot tolerate the gastrointestinal effects of GLP-1 receptor agonists or SGLT-2 inhibitors. The CAROLINA (CARdiovascular Outcome trial of LINAgliptin versus Glimepiride) trial demonstrated that in patients without established cardiovascular disease, the cardiovascular safety of glimepiride was non-inferior to linagliptin, supporting the continued use of affordable sulfonylureas as an alternative to gliptins in cost-sensitive settings. DPP-4 inhibitors may be combined with metformin (fixed-dose combinations available for sitagliptin-metformin and others), with sulfonylureas (with appropriate hypoglycemia precautions since adding a non-glucose-dependent secretagogue removes the glucose-dependence safety feature of the gliptin), with TZDs, and with insulin. They are not combined with GLP-1 receptor agonists, as dual incretin targeting shows no additive benefit and the combination is not approved.11
Alpha-glucosidase inhibitors have a limited but well-defined role in specific scenarios. The strongest evidence base supports their use in: patients with predominantly postprandial hyperglycemia on a background of acceptable fasting glucose control; patients with impaired glucose tolerance requiring pharmacological intervention for diabetes prevention when lifestyle modification alone is insufficient; and patients in whom weight neutrality and absence of systemic adverse effects are paramount and who are willing to accept the GI (gastrointestinal) tolerability challenge. In practice, AGIs are rarely initiated in North American clinical care due to tolerability; they are more commonly encountered as existing therapy in patients who have relocated from East Asian countries where they are more routinely prescribed. When initiating an alpha-glucosidase inhibitor (AGI), a critical practical step is starting at the lowest dose (acarbose 25 mg with the first bite of one meal per day) and increasing very gradually over weeks to months, as rapid dose escalation is the principal cause of intolerable GI adverse effects. Reducing dietary complex carbohydrate intake modestly while on an alpha-glucosidase inhibitor can substantially improve tolerability without compromising glycemic benefit, since the drug's mechanism is concentration-dependent on substrate availability.11
The combination logic for agents in this module follows the same comorbidity-driven framework established in Diab-02. When adding a second agent to metformin: in the absence of established ASCVD, HF (heart failure), or CKD, the choice among DPP-4 inhibitor, TZD (thiazolidinedione), AGI, sulfonylurea, SGLT-2 inhibitor, or GLP-1 receptor agonist is guided by the patient's specific circumstances. DPP-4 inhibitors are preferred when hypoglycemia risk is the dominant concern (elderly, irregular meals, driving). TZDs (thiazolidinediones) are preferred when NASH is present, when the patient has had a cerebrovascular event and is insulin-resistant, or when weight-neutral insulin-sensitizing therapy without hypoglycemia is desired and HF risk is low. AGIs are preferred when postprandial glucose is the primary target and the patient is committed to tolerability management. For all three classes, combination with metformin is the most common and rational sequence; triple combinations of metformin plus DPP-4 inhibitor plus sulfonylurea or TZD are supported by evidence but require careful monitoring of hypoglycemia, fluid status, and bone health as applicable. The fundamental hierarchy remains: agents with proven CV or cardiorenal outcome benefit first in patients with the relevant comorbidities; glycemic-efficacy and tolerability considerations guide selection among the remaining options.11
Pioglitazone is the only oral antidiabetic agent with consistent randomized trial evidence of histological improvement in NASH. The landmark PIVENS (Pioglitazone Versus Vitamin E Versus Placebo for the Treatment of Nondiabetic NASH) trial in non-diabetic patients and multiple trials in T2DM with NASH showed that pioglitazone significantly reduces hepatic steatosis, lobular inflammation, and hepatocyte ballooning, with some evidence of fibrosis regression.12 Current AASLD guidelines recommend pioglitazone (with informed consent regarding adverse effects) for T2DM patients with biopsy-proven NASH. This represents a clinically meaningful application beyond glycemia where pioglitazone's PPAR-gamma-mediated reduction of hepatic lipotoxicity and insulin sensitization directly targets the core disease mechanism.
Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, Kliewer SA. An antidiabetic thiazolidinedione is a high affinity ligand for peroxisome proliferator-activated receptor gamma (PPAR gamma). J Biol Chem. 1995;270(22):12953–12956.
doi:10.1074/jbc.270.22.12953Nesto RW, Bell D, Bonow RO, et al. Thiazolidinedione use, fluid retention, and congestive heart failure. Circulation. 2003;108(23):2941–2948.
doi:10.1161/01.CIR.0000103683.99399.7EDormandy JA, Charbonnel B, Eckland DJ, et al. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive Study (PROspective pioglitAzone Clinical Trial In macroVascular Events): a randomised controlled trial. Lancet. 2005;366(9493):1279–1289.
doi:10.1016/S0140-6736(05)67528-9Drucker DJ, Nauck MA. The incretin system: glucagon-like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhibitors in type 2 diabetes. Lancet. 2006;368(9548):1696–1705.
doi:10.1016/S0140-6736(06)69705-5Deacon CF. Physiology and pharmacology of DPP-4 in glucose homeostasis and the treatment of type 2 diabetes. Front Endocrinol (Lausanne). 2019;10:80.
doi:10.3389/fendo.2019.00080Scheen AJ. Pharmacokinetics of dipeptidylpeptidase-4 inhibitors. Diabetes Obes Metab. 2010;12(8):648–658.
doi:10.1111/j.1463-1326.2010.01212.xScirica BM, Bhatt DL, Braunwald E, et al. Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013;369(14):1317–1326.
doi:10.1056/NEJMoa1307684Bischoff H. Pharmacology of alpha-glucosidase inhibition. Eur J Clin Invest. 1994;24(Suppl 3):3–10.
doi:10.1111/j.1365-2362.1994.tb02249.xChiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M; STOP-NIDDM Trial Research Group. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet. 2002;359(9323):2072–2077.
doi:10.1016/S0140-6736(02)08905-5Monami M, Dicembrini I, Martelli D, Mannucci E. Safety of dipeptidyl peptidase-4 inhibitors: a meta-analysis of randomized clinical trials. Curr Med Res Opin. 2011;27(Suppl 3):57–64.
doi:10.1185/03007995.2011.602964American Diabetes Association Professional Practice Committee. Standards of Care in Diabetes—2024. Diabetes Care. 2024;47(Suppl 1):S1–S321.
doi:10.2337/dc24-SINTSanyal AJ, Chalasani N, Kowdley KV, et al. Pioglitazone, vitamin E, or placebo for nonalcoholic steatohepatitis. N Engl J Med. 2010;362(18):1675–1685.
doi:10.1056/NEJMoa0907929Kernan WN, Viscoli CM, Furie KL, et al. Pioglitazone after ischemic stroke or transient ischemic attack. N Engl J Med. 2016;374(14):1321–1331.
doi:10.1056/NEJMoa1506930Green JB, Bethel MA, Armstrong PW, et al. Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2015;373(3):232–242.
doi:10.1056/NEJMoa1501352Kahn SE, Haffner SM, Heise MA, et al. Glycemic durability of rosiglitazone, metformin, or glyburide monotherapy. N Engl J Med. 2006;355(23):2427–2443.
doi:10.1056/NEJMoa066224