Drug-induced diabetes

Drug-induced diabetes is a form of persistent hyperglycemia or overt diabetes that develops in temporal association with exposure to a pharmacological treatment capable of significantly disrupting glucose homeostasis. It is not a single homogeneous disease, but a broad clinical and etiological category in which drugs that differ greatly in therapeutic indication, molecular target, and metabolic profile may converge on a common outcome: elevated blood glucose caused by reduced insulin secretion, increased insulin resistance, increased hepatic glucose production, altered incretin signaling, direct or immune-mediated injury to pancreatic beta cells, or variable combinations of these mechanisms. In some patients, the drug unmasks diabetes in a person who was already predisposed but still subclinical, whereas in others it is the predominant causal factor.

Clinically, this topic requires a very precise framework because simply observing elevated blood glucose is not enough to define the problem. It is necessary to determine whether the patient has transient hyperglycemia, a reversible but clinically relevant form, persistent diabetes that continues after drug withdrawal, or a condition in which the medication has accelerated the emergence of type 2 diabetes or pre-existing latent autoimmune diabetes. Its practical relevance is very high: glucocorticoids, calcineurin inhibitors, mammalian target of rapamycin inhibitors, second-generation antipsychotics, phosphatidylinositol 3-kinase alpha inhibitors, immune checkpoint inhibitors, certain antiretrovirals, L-asparaginase, and other pharmacological classes can worsen glycemic control, increase the risk of acute complications, and alter the prognosis of the underlying disease for which the treatment was required.

The epidemiology of drug-induced diabetes is difficult to define precisely because it depends on the drug used, dose, duration of exposure, clinical setting, and especially the patient's metabolic vulnerability. There is therefore no single incidence applicable to all forms. The problem is probably underestimated in clinical practice because many cases of hyperglycemia are attributed generically to illness-related stress, artificial nutrition, reduced physical activity, systemic inflammation, or progression of the underlying disease, while the role of the drug remains secondary. In reality, the metabolic impact of medications is common and clinically tangible on medical and oncology wards, in immunosuppressed patients, after transplantation, and in inflammatory diseases treated with steroids.

Glucocorticoids are the most common paradigm of iatrogenic hyperglycemia, both because they are widely prescribed and because they simultaneously affect several nodes of glucose metabolism. The problem develops both in patients with known diabetes, whose glucose values tend to worsen rapidly, and in individuals without a previous diagnosis, in whom hyperglycemia may first appear during treatment. The rise in postprandial and afternoon values is particularly characteristic, which is why fasting glucose alone may underestimate the disorder. Alongside steroids, increasing importance is now attributed to molecularly targeted cancer drugs, immunotherapies, and immunosuppressants used in transplantation.

In post-transplant diabetes, now more precisely termed post-transplantation diabetes mellitus, risk depends on the interaction among diabetogenic immunosuppression, inflammatory status, viral infections, obesity, age, genetic profile, pre-existing beta-cell function, and the recipient's immunometabolic state. In this context, tacrolimus has a particularly important diabetogenic effect, often greater than that of cyclosporine, especially in people whose beta-cell reserve is already reduced. In oncology, immune checkpoint inhibitors represent a model opposite to steroid-induced diabetes: they do not predominantly cause insulin resistance, but can produce a rapid and sometimes fulminant loss of insulin secretion through an autoimmune mechanism, frequently presenting with diabetic ketoacidosis.

Second-generation antipsychotics also have major epidemiological relevance. Increased diabetes risk has been documented particularly with clozapine and olanzapine, although other drugs in the same class are affected to varying degrees. In this case, the clinical burden derives not only from hyperglycemia, but from the fact that it often occurs within a broader metabolic syndrome involving weight gain, visceral adiposity, dyslipidemia, and worsening cardiovascular risk. Taken together, these conditions explain why drug-induced diabetes should not be regarded as a classificatory curiosity, but as a category of major practical importance in endocrinology, internal medicine, oncology, psychiatry, and transplant medicine.

The pathophysiology of drug-induced diabetes can be understood only by distinguishing the predominant mechanisms through which treatment disrupts the balance among insulin secretion, insulin sensitivity, and endogenous glucose production. One major category includes drugs that increase insulin resistance. Glucocorticoids are the classic example: they increase hepatic gluconeogenesis, reduce peripheral glucose utilization in skeletal muscle, promote lipolysis and accumulation of gluconeogenic substrates, alter adipose-tissue distribution, and enhance the action of counterregulatory hormones. The final result is an increased secretory demand on the beta cell, which predisposed individuals cannot adequately meet.

A second category includes drugs that directly suppress beta-cell function. Calcineurin inhibitors, especially tacrolimus, reduce insulin secretion by interfering with intracellular pathways that are crucial for maintaining the beta-cell phenotype, insulin transcription, and granule exocytosis. In these patients, the principal defect is not only insulin resistance within the post-transplant setting, but also genuine secretory failure. Similar effects, although through different mechanisms, may be observed with certain immunosuppressants, L-asparaginase, some pancreatic toxins, and drugs that disrupt cellular mitochondrial homeostasis or endoplasmic reticulum function.

A third category comprises drugs that induce weight gain and neuroendocrine appetite dysfunction. Second-generation antipsychotics act through multiple receptor pathways, including histaminergic, serotonergic, and dopaminergic systems, promoting hyperphagia, reduced energy expenditure, visceral fat accumulation, adipose-tissue inflammation, and impaired insulin signaling. Their effect is not explained solely by weight gain, however. The literature also indicates direct mechanisms involving glucose transport, insulin signaling, and beta-cell secretion, so the drug may promote hyperglycemia even before weight gain becomes apparent.

The pathogenesis of diabetes caused by immune checkpoint inhibitors is very different. Here, the dominant mechanism is immune-mediated: blockade of programmed cell death protein 1, programmed death ligand 1, or cytotoxic T-lymphocyte-associated protein 4 may disrupt immune tolerance and peripheral control of autoreactive T lymphocytes, permitting rapid pancreatic beta-cell injury. Clinical presentation is often abrupt, with a collapse in C-peptide, glycated hemoglobin that may be only modestly elevated relative to the severity of current hyperglycemia, and frequent onset with diabetic ketoacidosis. The relatively modest glycated hemoglobin reflects the fact that beta-cell destruction may occur very rapidly.

There are also more specific mechanisms related to oncological targets. Phosphatidylinositol 3-kinase alpha inhibitors, such as alpelisib, cause hyperglycemia by interfering with a central insulin-signaling pathway. In essence, the drug blocks one of the main downstream intracellular circuits through which insulin promotes glucose uptake and suppresses hepatic production. Hyperglycemia is therefore not a nonspecific adverse effect but a direct consequence of the same pathway being targeted in the malignant cell. This explains why the glycemic disturbance may be marked, occur early, and closely depend on the intensity of pharmacological inhibition.

From a general pathophysiological perspective, drug-induced diabetes therefore arises from the interaction between pharmacological insult and individual metabolic reserve. A patient with visceral obesity, hepatic steatosis, a family history of diabetes, prediabetes, reduced beta-cell mass, older age, or chronic inflammation has less compensatory capacity. In such individuals, the drug does not act in isolation, but within a system already close to its decompensation threshold. This is why some treatments produce only modest glucose abnormalities in metabolically healthy people but can precipitate overt diabetes in predisposed patients.

Clinical manifestations depend on the speed of onset, dominant pathophysiological mechanism, and presence or absence of pre-existing diabetes. Patients may report the classic symptoms of hyperglycemia, namely polyuria, polydipsia, unintentional weight loss, fatigue, blurred vision, nocturia, worsening skin or genital infections, and delayed wound healing. In steroid-related forms, however, symptoms may remain subtle for some time, especially when hyperglycemia is predominantly postprandial or more pronounced in the afternoon. In these cases, patients may report only post-meal fatigue, greater evening thirst, or nonspecific deterioration in general well-being.

When the predominant mechanism is severe insulin resistance, as often occurs with high-dose glucocorticoids or certain targeted cancer treatments, the clinical picture may be dominated by marked hyperglycemia without necessarily causing ketosis. By contrast, with immune checkpoint inhibitors or rare forms of acute beta-cell toxicity, onset may be dramatic, with nausea, vomiting, abdominal pain, dehydration, tachypnea, confusion, or overt diabetic ketoacidosis. In this subgroup, progression from initial symptoms to metabolic decompensation may be rapid, and the window for detecting the problem before an acute event may be brief.

In patients receiving antipsychotics, the history often reveals a more gradual trajectory. Clinicians may identify increased appetite, weight gain, growing physical inactivity, worsening lipid profile, increased waist circumference, and progressive development of hyperglycemia. Here, diabetes often occurs within a broader metabolic syndrome. After transplantation, by contrast, the clinical picture is frequently masked by the complexity of care. Hyperglycemia may emerge during routine monitoring, intercurrent infections, adjustment of immunosuppressive therapy, or reduced insulin sensitivity caused by surgical stress and corticosteroids.

The physical examination may initially reveal few specific signs. Findings may include dehydration, weight loss, mucocutaneous candidiasis, worsening hypertension, central adiposity, or cushingoid facies and phenotype in patients exposed to steroids, as well as indirect signs of the causal setting such as inflammation, immunosuppression, oncological toxicity, or nutritional abnormalities. In patients with severe decompensation, signs of hemodynamic compromise, Kussmaul breathing, obtundation, abdominal pain, and altered consciousness must be sought immediately. Clinical observation should therefore proceed in a temporally coherent sequence: first identify the symptom, then relate it to current treatment, and finally estimate its severity and acute risk.

Suspicion of drug-induced diabetes arises from the combination of three elements: documented hyperglycemia, exposure to a plausibly diabetogenic drug, and a compatible temporal relationship between treatment initiation, dose escalation, or continuation and the development of the glucose abnormality. This does not mean that every case of hyperglycemia during treatment should automatically be attributed to the drug. Clinicians must always assess whether values were already abnormal beforehand, whether the patient had major metabolic risk factors, and whether concurrent infections, enteral or parenteral nutrition, surgery, severe stress, or intercurrent endocrinopathies are present. The correct judgment is therefore probabilistic and clinical rather than purely mechanical.

During initial assessment, several high-probability settings are useful to identify. The first is initiation of systemic glucocorticoids, especially at moderate to high doses, in individuals with obesity, older age, prediabetes, or known diabetes. The second is the period after solid-organ transplantation with tacrolimus, corticosteroids, and other immunosuppressants. The third is initiation of antipsychotics with a high metabolic burden, particularly when the patient already has weight gain or a positive family history. The fourth is cancer immunotherapy with immune checkpoint inhibitors, which requires attention not so much to gradual glycemic deterioration as to the risk of sudden onset and ketoacidosis. The fifth is use of phosphatidylinositol 3-kinase alpha inhibitors, which may cause very early hyperglycemia.

When drug-induced disease is suspected, the medication history must be extremely detailed. It is not enough to ask whether the patient takes “cortisone” or “cancer therapy.” The specific drug, dose, formulation, administration time, duration of exposure, any intermittent courses, recent treatment changes, and associated medications must be established. Chronology is often decisive: hyperglycemia appearing a few days after introduction of prednisone or dexamethasone, rising values in the weeks after transplantation while receiving tacrolimus, and abrupt onset after only a few immunotherapy infusions are very different scenarios and guide clinical reasoning in different ways.

Situations that strongly support a pharmacological etiology

• Onset of hyperglycemia after initiation or dose escalation of a drug with a recognized diabetogenic effect
• A glycemic profile consistent with the drug, such as predominantly postprandial hyperglycemia during steroid treatment
• No documented previous history of diabetes or marked worsening of previously stable diabetes
• Partial or complete resolution after treatment reduction, substitution, or withdrawal, when clinically feasible

Once suspicion has been raised, the next step is to distinguish the stable patient from one at immediate risk. If ketonemia, gastrointestinal symptoms, dehydration, tachypnea, altered mental status, or very high glucose values are present, the problem is no longer only etiological but a metabolic emergency. In the absence of acute decompensation, assessment continues by defining the glycemic pattern, probable beta-cell reserve, predisposing factors, and whether the causal treatment needs to be modified.

Diagnosis is based on the same general biochemical criteria used for diabetes mellitus, but assessment must be more detailed when an iatrogenic cause is suspected because the decisive issue is not only whether the patient has diabetes, but why it developed and whether the condition is reversible, persistent, or attributable to another subtype. According to ADA standards, diabetes may be diagnosed using fasting plasma glucose, 2-hour plasma glucose after an oral glucose tolerance test (OGTT), glycated hemoglobin (HbA1c), or random plasma glucose in the presence of classic hyperglycemic symptoms. In drug-induced diabetes, however, these tools do not all have the same sensitivity at different stages of the condition.

In glucocorticoid-induced diabetes, for example, fasting glucose may remain relatively preserved while postprandial and afternoon values rise markedly. Morning testing alone may therefore miss the condition. Serial capillary profiles, postprandial measurements, and, in selected cases, continuous glucose monitoring are particularly useful in these situations. HbA1c is most useful when exposure is chronic or a pre-existing abnormality is suspected, but it may provide little information in very recent-onset forms. Conversely, in immune checkpoint inhibitor-induced diabetes, HbA1c may be surprisingly modest despite marked current hyperglycemia, precisely because secretory loss may have occurred rapidly.

A rational diagnostic pathway first confirms the glucose abnormality and defines its pattern, then seeks evidence indicating the mechanism. When severe insulin deficiency is suspected, blood or urine ketones, blood gas analysis when indicated, electrolytes, kidney function, and acid-base balance should be assessed. If an autoimmune or fulminant form is suspected, as in checkpoint inhibitor-induced diabetes, C-peptide and autoantibodies associated with autoimmune diabetes are useful, while recognizing that serology is positive in only some cases. After transplantation, the glucose findings must be integrated with the treatment setting, timing in relation to transplantation, use of tacrolimus or steroids, presence of infections, and any relevant drug levels.

In the absence of specific, universal official diagnostic criteria for drug-induced diabetes, the correct approach is to formulate an integrated clinical etiological diagnosis. In practice, drug-induced diabetes is diagnosed when a patient meets the biochemical criteria for diabetes or clinically significant hyperglycemia, is exposed to a recognized diabetogenic medication, has a coherent temporal relationship, and no more convincing alternative explanation emerges. A definitive diagnosis therefore does not derive from a single pathognomonic biomarker, but from convergence among chronology, plausible pathophysiology, the glycemic pattern, and the clinical course. When feasible, regression of hyperglycemia after drug reduction or withdrawal retrospectively strengthens the causal relationship, but failure to regress does not exclude the diagnosis because the drug may already have unmasked persistent metabolic vulnerability.

The differential diagnosis primarily includes previously unrecognized type 2 diabetes, autoimmune diabetes of adulthood, stress hyperglycemia, pancreatic disorders, hyperglycemia-producing endocrinopathies, artificial nutrition, sepsis, and recurrence or progression of the underlying disease. Assessment should therefore not stop at the current medication, but should include the previous metabolic history, body weight, family history, prior HbA1c, lipid profile, pancreatic function when indicated, and the overall clinical picture. Only this pathway leads to a diagnosis that is genuinely useful for management.

The most useful practical classification is not purely nosological, but pathophysiological and therapeutic. One category includes drugs that predominantly cause insulin resistance, such as glucocorticoids and certain antipsychotics. A second includes drugs with a predominantly beta-cell-toxic or beta-cell-damaging effect, such as tacrolimus and, to some extent, other immunosuppressants. A third comprises treatments that cause genuinely rapid autoimmune diabetes, such as immune checkpoint inhibitors. A fourth includes targeted drugs that block intracellular pathways central to insulin action, such as alpelisib. A fifth comprises drugs or combinations with mixed effects, combining weight gain, insulin resistance, incretin abnormalities, and secretory impairment.

Glucocorticoids are the most frequent and clinically cross-cutting category. Hyperglycemia is often dose-dependent, but may also occur at moderate doses in predisposed individuals. The profile varies with the drug's half-life and administration time: with prednisone or equivalent agents given in the morning, glucose tends to rise mainly during the following hours, whereas with dexamethasone or long-acting steroids the disturbance may persist more broadly across the 24-hour period. Intermittent therapy, high-dose pulses, and association with systemic stress or artificial nutrition may amplify the condition.

Among immunosuppressants, tacrolimus is the drug most classically associated with post-transplant diabetes. Sirolimus and other mammalian target of rapamycin inhibitors may also contribute, often worsening an already fragile situation. Here, the iatrogenic component adds to transplant-specific factors, including inflammatory responses, sarcopenia, viral infections, pre-existing loss of beta-cell mass, and concomitant steroid use. Overall, the result is a form of diabetes in which insulin resistance and relative or absolute insulin deficiency coexist to variable degrees.

Second-generation antipsychotics are a particularly important category because risk depends not only on dose but also on the individual drug. Clozapine and olanzapine have the least favorable metabolic profile, whereas aripiprazole, ziprasidone, and lurasidone generally have a more limited impact, although none is absolutely metabolically neutral. Here too, classification must remain clinical: actual risk depends on the interaction among the drug, individual vulnerability, duration of exposure, and cumulative weight gain.

Among cancer treatments, immune checkpoint inhibitors and phosphatidylinositol 3-kinase alpha inhibitors deserve particular attention. The former may induce an often irreversible autoimmune insulin-deficient state. The latter cause mechanistically expected hyperglycemia because they interrupt insulin signal transduction. In hematology and oncology, L-asparaginase should also be remembered because it can impair pancreatic function and glucose metabolism, especially in intensive protocols and when combined with steroids.

Finally, some drugs have a more modest or less consistent diabetogenic effect, including certain thiazides, some beta-blockers, niacin, certain antiretrovirals, protease inhibitors, and other therapies that promote weight gain or worsen insulin sensitivity. In many of these cases, the drug rarely acts as the sole cause, but may push the system beyond the clinical threshold for diabetes. Precisely because of this heterogeneity, the category of “drug-induced diabetes” should be understood as a family of iatrogenic metabolic syndromes rather than a monolithic entity.

Treatment must pursue two simultaneous goals: correct hyperglycemia and maintain indispensable causal therapy whenever possible. This is the decisive point. In many cases, the diabetogenic drug cannot simply be stopped because it is required to control cancer, prevent transplant rejection, treat a severe autoimmune disease, or maintain psychiatric stability. The therapeutic strategy must therefore be individualized according to the mechanism of hyperglycemia, severity of the condition, and the real scope for modifying the causal drug. The first question is not only “which glucose-lowering medication should be used,” but “can the responsible treatment be reduced, replaced, adjusted, or must it be continued unchanged?”

In glucocorticoid-induced diabetes, correction of the glycemic pattern requires particular attention to timing. Because hyperglycemia is often postprandial and more pronounced in the afternoon, regimens focused exclusively on fasting glucose may be inadequate. Oral or other non-insulin medications may be considered in selected mild and stable cases, but in clinical practice insulin frequently remains the most effective option, especially when glucose is high, steroid doses are substantial, the patient is hospitalized, or the metabolic course is highly variable. Titration must track the profile of the glucocorticoid used and the intensity of the glucose response, with continuous reassessment during steroid tapering to prevent hypoglycemia.

Treatment of immune checkpoint inhibitor-induced diabetes is conceptually different. When the condition is insulin-deficient, insulin is the cornerstone of therapy, often permanently. Glucocorticoids do not correct autoimmune beta-cell loss and are not a specific treatment for diabetes in this setting. When presentation includes diabetic ketoacidosis, management follows standard protocols for the acute complication, with fluids, insulin, electrolyte correction, and intensive monitoring. The patient subsequently requires structured education, determination of insulin requirements, and close coordination with the oncology team regarding continuation of immunotherapy.

In the post-transplant setting, management must always be shared with the transplant center. Optimization of glycemic control cannot disregard the immunological safety of the graft. In some situations, modification of the immunosuppressive regimen may reduce the diabetogenic burden, but this decision must balance metabolic risk against rejection risk. Insulin is often the safest initial choice during the perioperative period and early post-transplant phase, when clinical variability is high. Other glucose-lowering medications may subsequently be considered in selected patients according to kidney function, infection risk, interactions, and graft stability.

In antipsychotic-associated diabetes, the most effective strategy is often twofold: correction of the metabolic disorder and reassessment of the psychiatric medication when the clinical situation permits. Switching to agents with a lower metabolic burden may be very useful, but must be decided only in conjunction with the treating psychiatrist. Metformin may have a role particularly when insulin resistance and weight gain predominate, whereas insulin is indicated in more severe or decompensated cases. In all cases, nutrition, physical activity compatible with the clinical context, and regular body-weight monitoring are structural rather than optional components of treatment.

In patients treated with alpelisib, hyperglycemia should be anticipated rather than chased. Recent literature emphasizes the importance of pretreatment metabolic assessment, intensive monitoring during the first weeks, and prompt intervention, often using metformin as the initial foundation when not contraindicated, with further intensification according to severity. Here too, the goal is not only to normalize glucose but to prevent unnecessary dose reductions or discontinuation of cancer therapy because of unmanaged metabolic toxicity. The general principle applicable to all forms is that treatment of drug-induced diabetes must be integrated, dynamic, and closely tailored to the responsible medication.

Follow-up should be tailored to the individual patient's risk profile and the responsible drug. There is no single universal schedule, but one principle is constant: monitoring must be more intensive when treatment is initiated, during dose increases, during periods of clinical stress, and while the diabetogenic treatment is being reduced, when the opposite problem—hypoglycemia caused by excessive correction—may develop. This applies particularly to steroids, cancer immunotherapy, and targeted treatments with a high metabolic burden.

In patients exposed to glucocorticoids, monitoring that captures the postprandial peak rather than fasting glucose alone is often useful. In hospitalized patients this means serial profiles; in outpatients it may mean capillary checks at selected times of day or use of continuous glucose monitoring when appropriate. Follow-up must be even closer in patients with pre-existing diabetes who start steroids because metabolic deterioration may be rapid and require daily treatment adjustments. After transplantation, follow-up must be integrated with monitoring of drug levels, kidney function, infection status, and the immunosuppressive regimen.

In antipsychotic-associated diabetes, follow-up should not be limited to glucose. Weight, body mass index, waist circumference, blood pressure, and the lipid profile should also be monitored because the metabolic injury is multidimensional. In immune checkpoint inhibitor-induced diabetes, by contrast, follow-up should focus primarily on early detection of hyperglycemic symptoms and evidence of insulin deficiency, with a very low threshold for urgent glucose and ketone testing in the presence of polyuria, polydipsia, nausea, or marked fatigue. Once insulin therapy is established, the patient should be followed as having insulin-deficient diabetes, with carbohydrate-counting education when indicated, prevention of hypoglycemia, and sick-day management.

Medium-term follow-up must also establish whether the glucose abnormality is resolving, stabilizing, or becoming chronic. Regression is possible after short courses of steroids, but is not guaranteed, particularly when treatment has revealed a pre-existing diabetic predisposition. After drug withdrawal, a few lower glucose readings are therefore insufficient: glucose metabolism must be reassessed formally to determine whether there is complete remission, persistent prediabetes, or overt diabetes requiring structured long-term follow-up.

Prognosis depends on four main variables: severity of the glucose disorder, reversibility of the pharmacological effect, feasibility of modifying the causal treatment, and presence of pre-existing metabolic vulnerability. In forms caused by glucocorticoids or other drugs acting predominantly through insulin resistance, metabolic prognosis may be good if exposure is limited, recognition is early, and the patient did not have a major diabetogenic substrate. A substantial proportion of individuals nevertheless fail to return to stable normoglycemia because treatment has unmasked type 2 diabetes that was already biologically developing. Reversibility should therefore never be assumed automatically.

In insulin-deficient forms, especially those caused by immune checkpoint inhibitors, prognosis is different. The secretory defect is often persistent, and the patient may remain insulin-dependent over the long term. Metabolic prognosis then depends on the quality of therapeutic education, prevention of ketoacidosis, adherence to insulin therapy, and integration with cancer care. After transplantation, poorly managed diabetes is associated with increased cardiovascular and infectious risk and sometimes poorer graft outcomes, making glycemic control an integral component of transplant prognosis.

Acute complications include diabetic ketoacidosis, hyperosmolar hyperglycemic state, severe dehydration, electrolyte disturbances, and worsening active infections. Ketoacidosis is particularly concerning in checkpoint inhibitor-induced forms, but may also occur in other settings of marked insulin deficiency or unrecognized decompensation. Chronic complications depend on the duration of hyperglycemia: if the disorder persists, the patient enters the same continuum of microvascular and macrovascular risk as other forms of diabetes, with renal, retinal, neurological, and cardiovascular involvement.

There is also a less obvious but clinically decisive complication: interference with treatment of the underlying disease. Unrecognized or poorly controlled iatrogenic diabetes may force dose reductions, immunotherapy interruptions, delays in cancer treatment, greater psychiatric instability when changing the antipsychotic is difficult, or deterioration of the post-transplant course. Drug-induced diabetes is therefore not merely a metabolic toxicity to be treated in parallel, but a clinical issue that can directly affect the ability to treat the primary disease effectively.

In summary, prognosis is better the earlier clinicians recognize the etiological relationship, define the dominant pathophysiological mechanism, and establish treatment that is genuinely consistent with the responsible drug. The key concept is not simply “treating blood glucose,” but managing a pharmacometabolic interaction that, if neglected, may worsen both diabetes and the condition for which the drug was prescribed.