Side effects and contraindications of antidiabetic medications

Treatment of diabetes mellitus cannot be designed by considering only the ability to lower blood glucose, because every glucose-lowering drug has a specific profile of adverse effects, contraindications, precautions, interactions, and clinical scenarios in which the balance between benefit and risk changes substantially. In modern clinical practice, therefore, there is no class that is absolutely “best”; rather, there is a more appropriate choice for the individual patient, determined by age, kidney function, hypoglycemic risk, body weight, the presence of heart failure, atherosclerotic cardiovascular disease, liver disease, frailty, sarcopenia, recurrent infections, gastrointestinal disorders, pregnancy, or possible autoimmune diabetes. Appropriate prescribing requires knowing not only what useful effects a drug can provide, but also in which setting it may cause harm.

This issue has become even more important as the therapeutic armamentarium has expanded. Newer classes have reduced hypoglycemic risk in many cases and improved cardiovascular or renal profiles, but they have introduced toxicities different from those of traditional drugs. At the same time, older agents retain important indications in selected subgroups, provided their limitations are understood and monitored precisely. Discussing the adverse effects and contraindications of antidiabetic medications therefore does not mean adding a list of warnings after treatment has been chosen, but defining from the outset which options are genuinely feasible and which should instead be excluded or used with particular caution.

The first principle is that the safety of an antidiabetic drug is not an abstract property, but the result of the interaction between the molecule and the clinical phenotype. The same therapy may be excellent in one patient and problematic in another. A sulfonylurea, for example, may be effective and simple in a person with preserved kidney function and regular meals, but become dangerous in a frail older adult with irregular food intake, kidney failure, or fall risk, because its main toxicity is hypoglycemia. Conversely, a sodium-glucose cotransporter 2 inhibitor may have substantial cardiorenal value in a selected patient, but require caution in a person predisposed to dehydration, genital infections, or ketosis.

The second principle is to distinguish precisely among contraindication, precaution, and adverse effect. A contraindication identifies a situation in which the drug must not be used, or should not be used except in highly justified exceptional circumstances. A precaution, by contrast, describes a scenario in which the drug may be used, but only with dose adjustment, monitoring, or careful patient selection. An adverse effect is a possible or frequent undesirable reaction that may be mild, moderate, or severe and does not necessarily amount to a prohibition on prescribing. Confusing these three levels leads to opposite errors: unjustified exclusion on the one hand and underestimation of genuine risk signals on the other.

The third principle is that toxicity is not limited to obvious acute events. Some undesirable effects are immediately perceptible, such as nausea, diarrhea, polyuria, or hypoglycemia, whereas others are more subtle and cumulative, such as vitamin B12 deficiency with metformin, weight gain with insulin or secretagogues, edema and heart failure risk with pioglitazone, or excessive loss of body mass in frail patients treated with drugs that markedly reduce caloric intake. Follow-up should therefore actively look for iatrogenic harm rather than being limited to glycated hemoglobin monitoring.

The fourth principle concerns the clinical setting in which therapy is modified. Hospitalization, prolonged fasting, severe infections, surgery, dehydration, acute kidney injury, administration of iodinated contrast medium, corticosteroid use, or abrupt dietary changes can profoundly alter the safety profile of antidiabetic medications. A therapy that is well tolerated under stable conditions may become inappropriate during an intercurrent illness. This is the basis of “sick day management,” namely the need to know which drugs should be temporarily withheld when the biological context changes.

Metformin remains the therapeutic foundation for many patients with type 2 diabetes mellitus because of its efficacy, cost, lack of hypoglycemia, and favorable weight profile, but its safety requires precise clinical interpretation. The most common adverse effects are gastrointestinal, particularly nausea, fullness, abdominal cramps, bloating, and diarrhea. These symptoms depend largely on the intestinal luminal concentration of the drug, the rate of titration, and individual susceptibility. Gradual introduction and use of extended-release formulations therefore often reduce symptoms and improve adherence.

The most feared adverse event is lactic acidosis, which is rare but potentially serious. Risk increases when the drug accumulates or when the patient develops conditions that promote tissue hypoxia, reduced renal clearance of the drug, or increased lactate production. The most relevant situations include severe kidney failure, acute deterioration of kidney function, sepsis, marked dehydration, cardiorespiratory failure with hypoperfusion, advanced liver disease, alcohol misuse, and severe catabolic states. Metformin does not itself “generate” lactic acidosis in a stable organism, but it may facilitate the problem when the patient enters a phase of metabolic or hemodynamic vulnerability.

Kidney function is now assessed using the estimated glomerular filtration rate rather than serum creatinine alone. Metformin can be used in many patients with a moderate reduction in filtration, but requires dose reduction and closer monitoring. It becomes inappropriate in more advanced disease or in situations involving acute deterioration of kidney function. The same applies to examinations with iodinated contrast medium in selected patients, in whom temporary discontinuation and subsequent reassessment of kidney function should be considered.

An often-overlooked issue is vitamin B12 deficiency, which may occur during prolonged treatment and contribute to neuropathy, macrocytic anemia, or worsening of neurologic symptoms already attributed to diabetes. Vitamin B12 monitoring should therefore be considered particularly in patients receiving chronic therapy, those with unexplained anemia, or those with neuropathic signs. In summary, metformin is a very safe drug when used in the correct setting, but its familiarity should not obscure the fact that the greatest risk emerges precisely when clinical conditions depart from stability.

Sulfonylureas and meglitinides share the mechanism of stimulating insulin secretion and, for this reason, their toxicity is dominated by the risk of hypoglycemia. This risk depends not only on dose, but on the mismatch between released insulin and the availability of glucose in the blood. It therefore increases when the patient skips meals, abruptly reduces caloric intake, consumes alcohol, undertakes unplanned physical activity, has kidney failure or liver disease, or has a limited ability to recognize the adrenergic symptoms of hypoglycemia. In older adults, a hypoglycemic event may present as confusion, a fall, syncope, or cognitive deterioration rather than the classic autonomic symptom complex.

Sulfonylurea toxicity is clinically relevant because hypoglycemia may be prolonged and severe, especially with longer-acting agents or reduced clearance. This makes these drugs less suitable for frail individuals, patients who live alone, those with irregular eating habits, or those with advanced kidney failure. Another common problem is weight gain, explained by the anabolic effect of insulin, disappearance of glycosuria, and sometimes compensatory carbohydrate intake to prevent or correct hypoglycemia.

Meglitinides generally have a shorter half-life and a greater focus on postprandial control, but they do not eliminate the problem of hypoglycemia and still require a reasonably regular relationship between drug administration and meals. Without this synchronization, their effect becomes unpredictable. Sulfonylureas should also be used cautiously in patients with glucose-6-phosphate dehydrogenase deficiency, because they may promote hemolytic anemia in predisposed individuals.

The clinical limitation of these classes is therefore not a single absolute contraindication, but the fact that their risk profile conflicts with many very common phenotypes of contemporary diabetes: older, multimorbid, obese patients with chronic kidney disease or high cardiovascular risk, in whom other classes can achieve glycemic control at a lower biological cost.

Pioglitazone improves insulin sensitivity and may have a strong rationale in some metabolic settings, but its use requires particular attention to safety. The most characteristic undesirable effect is fluid retention, which manifests as peripheral edema, weight gain, and, in predisposed individuals, worsening or unmasking of heart failure. This risk increases when the drug is combined with insulin or used in patients with already limited cardiac reserve. Pioglitazone is therefore contraindicated in patients with current or prior heart failure, NYHA class I to IV; signs of congestion or dyspnea also require immediate assessment.

Weight gain observed with pioglitazone is not a simple phenomenon because it reflects both fluid expansion and changes in the adipose compartment. Clinically, however, the result may be unfavorable in patients who are already obese or edematous. There is also the issue of fracture risk, reported particularly in women but more generally relevant in patients with osteopenia, advanced age, or fall risk. The drug should therefore be assessed cautiously when skeletal health is already vulnerable.

Another important consideration is the relationship with the bladder. The European summary of product characteristics contraindicates pioglitazone in patients with current or prior bladder cancer and unexplained macroscopic hematuria. Hepatic impairment and diabetic ketoacidosis are also contraindications. The biological significance of the oncologic signal has long been debated, but from a prescribing standpoint the urologic-oncologic history, liver function, and any hematuria must be checked before starting treatment.

Pioglitazone is therefore not a drug to be rejected wholesale, but a molecule to be used only when the expected pathophysiologic advantage clearly outweighs the risks of edema, heart failure, weight gain, and skeletal fragility. Its true practical contraindication is not merely the formal presence of a prohibited condition, but failure to select the appropriate patient.

Dipeptidyl peptidase 4 inhibitors are often considered “easy” drugs because they are oral, generally well tolerated, weight neutral, and unlikely to cause hypoglycemia unless combined with insulin or secretagogues. This assessment is largely correct, but risks becoming overly reassuring if their specific concerns are not considered. Common adverse effects are generally mild and include upper respiratory symptoms, headache, or modest gastrointestinal disturbances, but the main clinical issue concerns less frequent events and the need for dose adjustment in agents eliminated predominantly by the kidneys.

Regulatory labeling includes reports of acute pancreatitis with several drugs in the class. An absolute causal relationship remains difficult to establish in all clinical settings, but the precautionary principle is clear: in the presence of significant persistent abdominal pain, particularly if radiating to the back and associated with nausea or vomiting, the drug should be discontinued and the condition evaluated promptly. In patients with a history of pancreatitis, treatment is not always formally prohibited, but requires caution and individualized reasoning.

From a cardiorenal perspective, the class as a whole should be distinguished from individual molecules. The issue of heart failure emerged particularly with saxagliptin and alogliptin, to the extent that safety information recommends monitoring and caution in patients with prior cardiac or renal disease. This does not mean that the entire class should be treated as uniformly harmful to the heart, but it requires recognition that the molecules are not interchangeable. Kidney function also determines the dosage of several agents in the class, and failure to adjust the dose may increase exposure and adverse events.

Hypersensitivity reactions, severe arthralgia, and, more rarely, bullous pemphigoid have also been reported. These events are uncommon but important because they are often not recognized as drug related. The tolerability advantage of dipeptidyl peptidase 4 inhibitors remains real, but should not become the misleading notion that these drugs have no warning signs.

Glucagon-like peptide 1 receptor agonists and dual agonists of the glucose-dependent insulinotropic polypeptide and glucagon-like peptide 1 receptors have profoundly changed treatment of type 2 diabetes mellitus because of their glycemic efficacy, weight reduction, and, for some molecules, cardiovascular benefit. Their most frequent toxicity, however, is clearly gastrointestinal. Nausea, early satiety, vomiting, diarrhea, constipation, and dyspepsia are particularly common during the initial phases or when titration is too rapid. These symptoms arise from the combination of central effects on appetite and delayed gastric emptying.

In most patients, symptoms diminish over time and with gradual dose escalation, but in some cases they substantially impair food intake, adherence, or hydration status. This is particularly relevant in older or frail patients and in those with already inadequate protein intake, in whom an excessive anorectic effect may translate into loss of lean mass and functional deterioration. The adverse effect is therefore not merely “nausea,” but a possible imbalance between metabolic benefit and nutritional depletion.

Prescribing information also highlights the risk of acute pancreatitis and gallbladder disease, including cholelithiasis or cholecystitis. The mechanism cannot be fully reduced to a single biological pathway, but the combination of weight loss, biliary changes, and individual susceptibility makes this issue clinically concrete. Severe and persistent abdominal pain should therefore prompt immediate reassessment. For tirzepatide and semaglutide, as well as other molecules in the same therapeutic area, a history of pancreatitis is not always an absolute prohibition in every regulatory document, but it requires selective caution.

A crucial point is to distinguish between regulatory labels. U.S. FDA labels for semaglutide and tirzepatide contraindicate treatment in patients with a personal or family history of medullary thyroid carcinoma and in those with multiple endocrine neoplasia type 2, based on the signal of C-cell tumors observed in rodents. This contraindication does not appear in the European EMA summaries of product characteristics. In the Italian context, prescribing must therefore follow the European summary for the individual medicine, without automatically transferring U.S. restrictions.

Additional caution is needed in the presence of clinically significant gastroparesis or severe pre-existing digestive symptoms, because delayed gastric emptying may worsen an existing disorder. Furthermore, when these molecules are combined with insulin or sulfonylureas, hypoglycemic risk does not derive directly from the incretin agent but from the need to reduce concomitant therapy to avoid excessive lowering of blood glucose.

Sodium-glucose cotransporter 2 inhibitors have an extremely favorable profile in many patients because of cardiovascular, renal, and metabolic benefits, but their mechanism of action based on glycosuria explains much of their toxicity. The most common adverse events are genital mycotic infections, particularly vulvovaginal candidiasis and balanitis, promoted by the increased availability of glucose in the genitourinary tract. Urinary tract infections may occur, although the association is less consistent than for genital mycoses. Good local hygiene, advance patient education, and early recognition of symptoms greatly improve clinical manageability.

The second major issue is volume depletion. These drugs increase osmotic diuresis and natriuresis and may therefore promote hypotension, dizziness, functional deterioration of kidney function, or syncope, especially in older adults, patients receiving diuretics, individuals with reduced fluid intake, or those with intercurrent conditions such as vomiting, diarrhea, and fever. Risk is not constant, but depends on the hemodynamic context and the overall treatment regimen. Prescribing should therefore take into account not only glomerular filtration, but also vulnerability to dehydration.

The most insidious adverse event is diabetic ketoacidosis, which may occur even without markedly elevated blood glucose, namely in euglycemic or relatively low-glucose form. Reduced insulinemia, increased glucagon, fasting, drastic carbohydrate restriction, surgery, severe infections, alcohol misuse, prolonged vomiting, or excessive insulin reduction can shift metabolism toward ketogenesis. Sodium-glucose cotransporter 2 inhibitors should therefore be discontinued before planned surgical procedures and temporarily withheld during acute illness, reduced food intake, or suspected ketosis.

There are also rare but serious events, such as Fournier gangrene, which require a very high index of clinical suspicion, and a historical signal of increased amputation risk observed particularly with canagliflozin in specific trials, although the regulatory context was subsequently modified. This neither justifies ignoring the signal nor automatically generalizing it to the entire class, but calls for caution in patients with advanced peripheral arterial disease, foot ulcers, previous amputations, or severe diabetic foot infections.

From the standpoint of practical contraindications, these drugs are not indicated for glycemic control in type 1 diabetes mellitus in many regulatory jurisdictions because of the excess risk of ketoacidosis, and their usability also depends on kidney function and the specific clinical indication. It is therefore not enough to know that sodium-glucose cotransporter 2 inhibitors “protect the kidney,” because safety changes according to the stage of kidney disease, the purpose of treatment, and the acute or chronic setting in which they are used.

Alpha-glucosidase inhibitors, represented mainly by acarbose, are used more selectively than in the past but remain conceptually useful for controlling postprandial blood glucose. Their main limitation is almost entirely intestinal. Bloating, flatulence, abdominal distension, cramps, and diarrhea occur because incompletely absorbed carbohydrates reach the colon and are fermented by bacterial flora. This toxicity does not depend on a rare idiosyncrasy, but on the drug’s mechanism itself, and practical tolerability is therefore often the true determinant of adherence.

Unlike other classes, there are very concrete gastrointestinal contraindications. Acarbose should not be used in inflammatory bowel disease, colonic ulceration, intestinal obstruction or predisposition to obstruction, chronic intestinal diseases with major disorders of digestion and absorption, or situations in which increased gas production could worsen the clinical condition. Cirrhosis is also included among regulatory contraindications. It is therefore a drug that requires a genuine gastroenterologic history before prescribing.

An important practical detail concerns hypoglycemia during combination treatment. Acarbose alone rarely causes it, but if the patient is also receiving insulin or secretagogues, any hypoglycemia should be corrected with glucose, not sucrose, because absorption of disaccharides may be delayed. This is a simple but essential educational point.

Insulin remains the most powerful glucose-lowering drug and, in many settings, the most necessary. For this very reason, it is also the treatment that most clearly shows how pharmacologic efficacy can become dangerous when it exceeds the patient’s requirements. The key adverse event is hypoglycemia, which represents the main limitation of intensive insulin therapy. Risk depends on dose, timing, insulin type, kidney function, meals, physical activity, alcohol consumption, and the patient’s ability to adapt treatment correctly to everyday circumstances.

Severe hypoglycemia may cause seizures, trauma, arrhythmias, loss of consciousness, and, in the most serious cases, death. In patients with a long history of diabetes and impaired awareness of symptoms, the danger increases further. It would, however, be reductive to view hypoglycemia only as an unavoidable pharmacologic effect. In clinical practice, it is often the result of management errors, such as overestimating basal requirements, failing to reduce the dose when intake falls, inappropriate mealtime dosing, or persistence of overly rigid regimens in patients whose real lives are highly variable.

Another very common effect is weight gain, caused by correction of catabolism, reduction in urinary caloric loss, and the anabolic action of insulin. Weight gain is particularly relevant in type 2 diabetes mellitus with obesity, where it may worsen insulin resistance and fuel a cycle of dose escalation. Insulin may also cause hypokalemia by shifting potassium into cells, an especially important problem in patients at risk of arrhythmia, those receiving diuretics, or those with other causes of potassium depletion.

Local reactions, lipodystrophy from repeated injections at the same site, technique errors, and mix-ups between preparations are additional sources of preventable harm. Insulin safety therefore depends to a very large extent on therapeutic education. More than with other classes, the drug cannot be separated from patient behavior and the quality of the instruction received.

Correct management of contraindications requires systematic assessment of several major clinical domains. The first is kidney function, which directly affects metformin, dipeptidyl peptidase 4 inhibitors, sodium-glucose cotransporter 2 inhibitors, and the hypoglycemic risk of insulin or secretagogues. The second is cardiac function, essential for avoiding pioglitazone in patients with current or prior heart failure and for selecting other agents cautiously in those with cardiac vulnerability. The third is nutritional and functional vulnerability, which profoundly changes the clinical significance of nausea, weight loss, dehydration, or hypoglycemia.

Liver disease modifies the management of several classes, not only for metabolic reasons but also because of the body’s ability to compensate for adverse events. Advanced age is not itself a contraindication, but increases the biological cost of many adverse effects, particularly hypoglycemia, volume depletion, loss of muscle mass, and interactions related to polypharmacy. During pregnancy, the range of options narrows considerably and safety must be reconsidered according to specific indications, avoiding automatic transfer of management principles from the nonpregnant patient.

Intercurrent illness is ultimately the scenario in which medications most often become temporarily inappropriate. Vomiting, diarrhea, fasting, fever, sepsis, surgery, or reduced hydration require knowledge of which therapies to withhold to prevent lactic acidosis, ketoacidosis, functional kidney failure, or hypoglycemia. This is not a minor educational detail, but an integral component of preventing serious adverse events.

Clinical situations requiring particular caution

• Kidney failure, dehydration, sepsis, hypoxia, or iodinated contrast medium with metformin
• Frailty, irregular meals, kidney failure, and fall risk with sulfonylureas, meglitinides, and insulin
• Current or prior heart failure, hepatic impairment, diabetic ketoacidosis, current or prior bladder cancer, or unexplained macroscopic hematuria with pioglitazone
• History of pancreatitis or significant digestive symptoms with selected incretin therapies; medullary thyroid carcinoma or multiple endocrine neoplasia type 2 only when included in the applicable regulatory label, as in U.S. FDA labels
• Fasting, surgery, acute illness, ketosis, predisposition to dehydration, or genital infections with sodium-glucose cotransporter 2 inhibitors

Reducing adverse effects depends not only on the choice of molecule, but on the quality of therapeutic education. The patient should be able to recognize symptoms of hypoglycemia, dehydration, genital candidiasis, pancreatitis, ketosis, and clinically significant gastrointestinal intolerance. The patient should also know when a reaction is expected and manageable and when it instead requires discontinuation of the drug and urgent medical assessment. Without this clinical literacy, even correctly prescribed therapy becomes less safe.

Monitoring should be proportionate to risk. With metformin, estimated glomerular filtration rate should be reassessed periodically and vitamin B12 measurement considered during prolonged treatment. With insulin and secretagogues, the focus is on hypoglycemia, body weight, and the suitability of the regimen for actual meals. With incretin therapies and dual agonists, appetite, hydration, weight trajectory, abdominal symptoms, and gastrointestinal tolerability should be followed. With sodium-glucose cotransporter 2 inhibitors, the patient should be instructed about temporary discontinuation rules and symptoms suggesting ketosis or severe infection.

Prevention of serious adverse events therefore follows a simple but often neglected principle: the drug does not merely enter a prescription, but enters the patient’s real life. If the context changes, safety changes as well. Clinicians treating diabetes should therefore periodically reassess not only whether therapy works, but whether it remains appropriate for the person’s current clinical profile.