Glucocorticoid-induced diabetes

Glucocorticoid-induced diabetes is a form of iatrogenic hyperglycemia that develops during glucocorticoid treatment and may present as worsening of pre-existing diabetes or as new-onset diabetes in individuals without a previous diagnosis. It is not merely a nonspecific biochemical adverse effect, but a defined clinical condition supported by recognizable pathophysiological mechanisms involving insulin resistance, increased hepatic glucose production, impaired beta-cell function, and changes in the glucose profile at different times of day. Recent documents and dedicated reviews emphasize that this form is probably underdiagnosed, especially when assessment is limited to fasting glucose alone, because hyperglycemia often tends to be more evident during the postprandial and afternoon hours.

Its clinical relevance is very high because glucocorticoids are used in a vast range of settings, including autoimmune diseases, respiratory disorders, oncology, hematology, transplantation, pain management, palliative care, and neurological conditions. Consequently, diabetologists, internists, and organ-system specialists must be able to distinguish transient hyperglycemia from genuinely persistent diabetes, recognize the characteristic glucose profiles of different steroid molecules, and treat the problem without compromising, whenever possible, the underlying therapy that required glucocorticoid use. Current guidelines and standards stress precisely this point: early recognition affects safety, prognosis, and the quality of overall clinical management.

The epidemiology of glucocorticoid-induced diabetes cannot be summarized by a single percentage applicable to everyone, because risk varies according to dose, treatment duration, glucocorticoid potency, route of administration, timing schedule, and the patient's metabolic susceptibility. The most recent reviews emphasize that steroid-induced hyperglycemia is one of the most common forms of iatrogenic dysmetabolism in clinical practice and that its true frequency is probably underestimated, particularly in outpatient settings and among patients who have not undergone prior glucose screening. Moreover, risk is not confined to very long treatment courses: even relatively brief exposures can precipitate clinically significant hyperglycemia in susceptible individuals.

The problem occurs both in patients with known diabetes, in whom steroid treatment tends to destabilize previously stable glycemic control, and in individuals without a previous diagnosis, in whom the glucocorticoid may act as a trigger or reveal a metabolic vulnerability that was already present but still subclinical. Older age, visceral obesity, a family history of type 2 diabetes mellitus, prediabetes, metabolic syndrome, hepatic steatosis, reduced physical activity, and chronic inflammation increase the likelihood that steroid exposure will result in overt diabetes. In hospitalized patients, the coexistence of acute stress, infections, artificial nutrition, or reduced mobility further amplifies this risk.

The epidemiological relevance is particularly evident in departments where glucocorticoids are an integral part of standard treatment. This applies to pulmonology, rheumatology, gastroenterology, nephrology, internal medicine, oncology, and hematology. The post-transplant setting is also high risk, although steroid diabetogenicity in that context often adds to the effects of other immunosuppressants. It should also be remembered that the problem is not limited to patients receiving chronic oral therapy. Risk also exists with high-dose intravenous pulses, long-acting steroids, and certain repeated intermittent regimens, in which the cyclical nature of exposure produces a more variable glucose profile that is therefore less intuitive to interpret.

From a public health perspective, glucocorticoid-induced diabetes carries a burden that extends beyond a laboratory abnormality, because it can promote infections, delay healing, increase the risk of hospitalization, prolong length of stay, and worsen the outcomes of already severe underlying diseases. In vulnerable patients, steroid-induced hyperglycemia may also interfere with treatment titration, forcing clinicians to balance anti-inflammatory or immunosuppressive efficacy against metabolic safety. For this reason, recent sources no longer treat the problem as an endocrinological curiosity, but as a cross-cutting issue in contemporary medicine.

The pathophysiology of glucocorticoid-induced diabetes is multifactorial and arises from the combined effects of these drugs on the liver, skeletal muscle, adipose tissue, endocrine pancreas, and counterregulatory control. The most important mechanism is increased insulin resistance. Glucocorticoids enhance hepatic glucose production by stimulating gluconeogenesis and reduce the suppression of hepatic output that normally occurs in the presence of insulin. At the same time, they limit peripheral glucose uptake, particularly in muscle, which is the main compartment responsible for postprandial glucose disposal. The result is a rise in glycemic load, especially after meals.

In adipose tissue, glucocorticoids promote lipolysis and fat redistribution, increasing the flux of free fatty acids and other metabolic substrates that further impair insulin signaling. This environment favors systemic insulin resistance that is not confined to glucose metabolism but involves overall energy integration. The accumulation of visceral adiposity during more prolonged exposure helps sustain the disorder beyond the initial treatment phase, especially in patients who are already predisposed. Steroid diabetogenicity is therefore not exclusively acute, but may become chronic when the drug acts on an already reduced metabolic reserve.

In addition to insulin resistance, there is a component of inadequate beta-cell compensation. Under physiological conditions, beta cells should increase insulin secretion to offset the greater glycemic load induced by steroids. Many patients, however, do not have sufficient secretory reserve. In some cases, glucocorticoids may also interfere directly with beta-cell function, reducing the efficiency of glucose-stimulated insulin secretion. This explains why glucocorticoid-induced diabetes develops more readily in individuals with prediabetes or a pre-existing latent beta-cell defect.

A very important pathophysiological feature is the time course of hyperglycemia. With intermediate-acting steroids administered in the morning, such as prednisone, the peak hyperglycemic effect tends to occur in the following hours and is most evident in the afternoon and after meals. With long-acting agents, such as dexamethasone, or with multiple doses distributed throughout the day, the disturbance may extend more uniformly across the 24-hour period and substantially affect fasting glucose as well. This feature is central to both diagnosis and treatment selection, because management must track the actual pattern rather than a generic concept of “cortisone-induced diabetes.”

In final pathophysiological terms, glucocorticoid-induced diabetes therefore represents the clinical expression of an imbalance between increased insulin requirements and reduced compensatory capacity. The drug shifts the system toward greater hepatic glucose production, lower peripheral utilization, increased availability of counterregulatory substrates, and relatively insufficient beta-cell secretion. When the sum of these effects exceeds the individual's compensatory threshold, clinically significant hyperglycemia appears; it may be transient or persistent depending on the duration of exposure and the patient's metabolic substrate.

The clinical manifestations of glucocorticoid-induced diabetes are highly variable and depend on the severity of hyperglycemia, speed of onset, duration of steroid therapy, and the patient's comorbidities. The history may range from a simple abnormality detected during monitoring to the classic symptoms of hyperglycemia, namely polyuria, polydipsia, fatigue, blurred vision, nocturia, weight loss, and increased susceptibility to infection. This form may nevertheless be clinically elusive, because hyperglycemia is often more marked during the postprandial and evening hours, whereas morning blood glucose may be less abnormal. Many patients are therefore missed when assessment is limited to isolated fasting measurements.

In patients with pre-existing diabetes, the most typical manifestation is worsening glycemic control, with higher postprandial values, greater day-to-day variability, and an increased insulin requirement or need to intensify treatment. Patients may report that their values “spike” especially after lunch and in the late afternoon, a pattern that is highly consistent with morning use of intermediate-acting steroids. In people without previous diabetes, symptoms may remain mild, particularly when treatment is brief, but in more pronounced cases the clinical onset may become evident within a few days.

There are no pathognomonic physical findings of glucocorticoid-induced diabetes. Dehydration, dry mucous membranes, weight loss, or worsening superficial infections may be present, but the most important finding is often the overall clinical context. The patient may have a cushingoid phenotype, central adiposity, elevated blood pressure, skin fragility, or signs of the underlying disease that required steroid treatment. Clinical interpretation therefore requires the correct logical sequence: first identify hyperglycemia, then relate it to the pharmacological profile, and finally assess its severity and systemic impact.

In the most severe cases, signs of acute metabolic decompensation may develop. Although glucocorticoid-induced diabetes is often dominated by hyperglycemia without ketosis, the literature reports the possibility of hyperosmolar hyperglycemic state and, less commonly, diabetic ketoacidosis, particularly in individuals with unrecognized diabetes, severe relative insulin deficiency, or high-dose exposure. In the presence of nausea, vomiting, abdominal pain, altered mental status, tachypnea, or marked dehydration, the condition must therefore be managed as a potential metabolic emergency rather than as a simple “effect of cortisone.”

Clinical suspicion should arise whenever hyperglycemia develops in a patient receiving glucocorticoids, especially when there is a close temporal relationship with treatment initiation or a recent dose increase. The likelihood is greater when the patient has predisposing factors such as obesity, a family history of diabetes, prediabetes, older age, cardiovascular disease, hepatic steatosis, or previous stress hyperglycemia. Even in patients with established diabetes, the finding should not be dismissed as nonspecific decompensation, because the steroid-induced pattern has temporal characteristics that directly influence monitoring and treatment.

A detailed medication history is essential. It should establish the specific drug used, steroid equivalence, route of administration, number of daily doses, timing of administration, expected treatment duration, and any recent changes. Prednisone, methylprednisolone, dexamethasone, and hydrocortisone do not produce the same glycemic profile. The generic information that a patient “takes cortisone” is insufficient to interpret the disorder. Clinicians must understand when the drug reaches its greatest metabolic effect, because that is when hyperglycemia tends to become most evident.

Features that strongly support suspicion of glucocorticoid-induced diabetes

• Onset or worsening of hyperglycemia after glucocorticoid initiation or dose escalation
• Higher blood glucose values after meals and in the afternoon than during the morning fast
• Presence of pre-existing metabolic risk factors, even without known diabetes
• Reduction in glucose values after glucocorticoid tapering or withdrawal, when clinically feasible

Suspicion should be even greater in hospitalized patients, because the steroid effect in hospital often adds to inflammatory stress, immobility, artificial nutrition, and fluctuations in treatment. A low threshold for attention is also needed in outpatient care, however, particularly during repeated courses or chronic therapy. The most recent practical reviews emphasize that the absence of fasting hyperglycemia does not exclude the condition and that targeted assessment of postprandial periods is often decisive for early detection.

Diagnosis of glucocorticoid-induced diabetes begins with the general criteria used to diagnose diabetes mellitus. According to the American Diabetes Association's Standards of Care in Diabetes 2026, 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 symptoms of hyperglycemia or a hyperglycemic crisis. In glucocorticoid-induced diabetes, however, the choice of test and especially the time at which it is performed are crucial, because the disorder is often not optimally expressed by fasting glucose alone.

The most rational diagnostic pathway begins by documenting the glycemic pattern. In patients receiving morning steroid dosing, postprandial glucose and the afternoon profile are often more informative than fasting glucose. JBDS guidelines and practical reviews therefore recommend targeted capillary monitoring during the hours when the hyperglycemic effect is greatest. HbA1c may help identify a pre-existing or chronic abnormality, but during recent exposure it may underestimate the problem because it reflects the average pattern over the preceding weeks rather than the immediate impact of the new treatment.

In the absence of separate, specific official diagnostic criteria for glucocorticoid-induced diabetes, etiological diagnosis is based on a combination of clinical and laboratory findings. According to the sources most widely used in practice, diagnosis requires documentation of hyperglycemia that meets the general criteria for diabetes or is otherwise clinically significant, in the presence of glucocorticoid exposure with a plausible temporal relationship and a glycemic profile consistent with the effect of the administered drug. The core of diagnosis is therefore not an exclusive biomarker, but the integration of chronology, pathophysiology, value trends, and clinical context.

The differential diagnosis should include unrecognized type 2 diabetes mellitus, stress hyperglycemia, enteral or parenteral nutrition, hyperglycemia-producing endocrinopathies, severe infections, and other concurrent iatrogenic causes. In patients with marked hyperglycemia, hyperosmolar state and ketoacidosis must also be excluded through clinical assessment, electrolytes, effective osmolality when indicated, blood or urine ketones, and acid-base balance. This step is fundamental because it transforms the evaluation from simple etiological classification into immediate clinical risk stratification.

From a practical standpoint, glucocorticoid-induced diabetes may be classified according to three dimensions: the time course of exposure, the resulting glycemic pattern, and the patient's metabolic substrate. A first useful distinction is between transient hyperglycemia during a short course and persistent diabetes associated with chronic therapy or the unmasking of a pre-existing diabetic predisposition. This distinction is not merely theoretical, because it affects follow-up intensity, likelihood of remission, and long-term treatment choices.

A second classification is based on the type of glucocorticoid and its clinical pharmacokinetics. With intermediate-acting steroids administered once in the morning, the typical picture is predominantly postprandial and afternoon hyperglycemia. With long-acting steroids or repeated doses, the profile tends to become more widespread across the 24-hour period, with a greater likelihood of involving the nighttime hours and fasting glucose on the following day. This way of classifying the problem is extremely useful because it guides selection of the most appropriate insulin regimen or monitoring strategy.

A third distinction concerns the relationship with pre-existing metabolism. In some patients, glucocorticoids produce a new metabolic disturbance that is relatively reversible after withdrawal. In others, they accelerate the emergence of type 2 diabetes mellitus that was already developing biologically. In still others, they aggravate known diabetes by increasing insulin requirements and glycemic variability. This classification does not change the fact that steroid therapy is the precipitating event, but it helps explain why some patients normalize their metabolism after tapering whereas others continue to meet the diagnostic criteria for diabetes over time.

Finally, there is a classification by clinical severity, distinguishing a simple rise in glucose detected through monitoring from forms requiring intensive pharmacological treatment or hospitalization for metabolic decompensation. This distinction is particularly important in medical and oncology patients, because the impact of the disorder depends not only on the absolute glucose value but also on the overall context, infection risk, fluid and electrolyte status, and whether high-dose steroid therapy must be continued.

Treatment of glucocorticoid-induced diabetes must begin with a simple principle: clinicians must treat not only the glucose value, but the glycemic pattern induced by the steroid. Merely intensifying therapy empirically without considering the type of glucocorticoid, timing of administration, and daily trend in values creates the dual risk of persistent hyperglycemia during critical hours and hypoglycemia as the pharmacological effect wanes. JBDS guidelines, the most recent practical reviews, and hospital-based sources emphasize that insulin often remains the most effective option, particularly for hospitalized patients, those with high values, clinically unstable individuals, and anyone requiring rapid adjustments.

In hospitalized patients or severe cases, insulin therapy is the cornerstone of management. The choice of regimen depends on the steroid profile. With morning prednisone, for example, it may be rational to provide coverage that matches the daytime and postprandial rise in glucose; with long-acting steroids or multiple doses, more extended coverage across the entire 24-hour period may be required. In critically ill patients, intravenous insulin remains the preferred method for tight control, as indicated in chapters devoted to inpatient diabetes care.

In clinically stable outpatients with nonsevere hyperglycemia and a predictable steroid regimen, non-insulin strategies may be considered in selected cases, but the decision must account for kidney function, hypoglycemia risk, the required speed of action, and expected duration of exposure. Practical reviews report the use of metformin and other medications in appropriate settings, while also highlighting the limitations of many classes when hyperglycemia develops very rapidly or flexible titration is needed in parallel with steroid tapering. For this reason, insulin often remains the most controllable solution in real-world practice.

A crucial issue is adjustment of therapy when the glucocorticoid dose is reduced. If clinicians correct hyperglycemia during the high-dose phase but fail to reduce glucose-lowering treatment promptly during tapering, the risk of clinically significant hypoglycemia increases. Proper management therefore requires close reassessment and precise instructions for the patient. Practical guidelines regard this aspect as being just as central as the initial intensification phase, because glucocorticoid-induced diabetes is a dynamic condition rather than a static metabolic state.

Treatment must ultimately form part of a broader strategy that includes patient education, self-monitoring, nutritional adjustment to the clinical context, and coordination with the specialist prescribing the steroid. When possible, dose reduction or use of the least diabetogenic regimen compatible with control of the underlying disease may be discussed, but this decision is never purely a diabetological one. The correct rationale is to preserve the therapeutic benefit of the glucocorticoid while reducing its metabolic cost through proactive rather than reactive management.

Monitoring is one of the most important steps, because this form of diabetes can easily be underestimated when values are measured at the wrong time of day. Practical sources therefore recommend targeted assessment during the periods when the steroid effect is greatest. In patients without known diabetes who start systemic glucocorticoids, capillary monitoring should be more intensive during the early phase of treatment and pay particular attention to postprandial values. In patients with known diabetes, especially those already receiving insulin, surveillance must be intensified immediately because requirements can change rapidly.

Follow-up should be tailored to the expected duration of steroid therapy. During short courses, the main objective is to detect and treat hyperglycemia during exposure, followed by reassessment of regression after withdrawal. With chronic therapy, by contrast, the patient should be followed as someone at persistent risk of overt diabetes and metabolic complications, with periodic assessment of HbA1c, body weight, blood pressure, and other cardiovascular risk factors. The most recent reviews emphasize that a proportion of patients do not return to normoglycemia after treatment ends, especially when steroids have unmasked evolving type 2 diabetes.

In hospitalized patients, follow-up must be integrated into the overall care setting. The same patient may require nearly daily adjustments according to steroid dose, the course of inflammation, resumption of oral intake, development of infections, or changes in concomitant therapy. For patients discharged while still receiving steroids, a clear transition of care is essential, with written instructions on monitoring times, alert thresholds, treatment changes if the glucocorticoid dose is reduced, and the timing of outpatient reassessment. This prevents a problem that was well managed in hospital from being overlooked after the patient returns home.

Education is a substantial component of follow-up. Patients must understand that their values may worsen temporarily but to a clinically important degree during the hours after a dose, that steroid withdrawal or dose reduction changes treatment requirements, and that symptoms such as intense thirst, polyuria, nausea, or marked fatigue require immediate attention. In glucocorticoid-induced diabetes, monitoring is not a technical accessory but the tool that makes it possible to track a disorder that is, by definition, variable over time.

Prognosis depends on the duration of steroid exposure, severity of hyperglycemia, presence of pre-existing diabetes or prediabetes, and quality of monitoring and treatment. In patients receiving short-term therapy who have good metabolic reserve, the disorder may be reversible after glucocorticoid withdrawal. In other individuals, especially those who are already predisposed, the iatrogenic event marks the transition from subclinical vulnerability to persistent overt diabetes. Prognosis therefore cannot be inferred solely from the fact that the steroid will subsequently be discontinued.

Acute complications include worsening infections, dehydration, severe hyperglycemic decompensation, hyperosmolar hyperglycemic state, and, less commonly, diabetic ketoacidosis. In hospital, glucocorticoid-induced hyperglycemia is also associated with poorer outcomes and greater management complexity. Prognosis is further affected by the glucocorticoid's impact on overall cardiovascular and metabolic risk factors, because hyperglycemia often occurs alongside weight gain, elevated blood pressure, and worsening body composition.

When exposure is prolonged, the patient enters the same continuum of risk as in other forms of diabetes, with possible development of microvascular and macrovascular complications if glycemic control remains inadequate. Although the literature on chronic damage specifically attributable to glucocorticoid-induced diabetes alone is less consistent, the biological principle does not change: persistent hyperglycemia promotes organ toxicity regardless of its origin. Proper management should therefore not be deferred on the assumption that the problem is secondary or, by definition, temporary.

In summary, prognosis is better the more promptly clinicians recognize the steroid-induced glycemic pattern, tailor monitoring to the pharmacological profile, and adjust treatment dynamically throughout the entire period of exposure. The true prognostic error is not only undertreatment, but treatment that is inconsistent with the temporal pathophysiology of the disorder.