Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026

Key points

• Diabetes is diagnosed by A1C ≥6.5% (≥48 mmol/mol), fasting plasma glucose (FPG) ≥126 mg/dL (≥7.0 mmol/L), 2-h plasma glucose ≥200 mg/dL (≥11.1 mmol/L) during a 75-g oral glucose tolerance test (OGTT), or random plasma glucose ≥200 mg/dL (≥11.1 mmol/L) with classic symptoms of hyperglycemia or a hyperglycemic crisis.
• Without unequivocal hyperglycemia, the diagnosis needs two abnormal results: two different tests (they may come from the same sample) or the same test repeated.
• Prediabetes is defined as A1C 5.7–6.4% (39–47 mmol/mol), FPG 100–125 mg/dL (5.6–6.9 mmol/L) or 2-h plasma glucose 140–199 mg/dL (7.8–11.0 mmol/L).
• Plasma glucose, not A1C, should be used when the relationship between A1C and glycemia is altered: some hemoglobin variants, pregnancy, G6PD deficiency, HIV and conditions that change red blood cell turnover.
• Adults of any age with overweight or obesity and at least one risk factor should be considered for testing; everyone else should start at 35 years, repeating at least every 3 years if results are normal.
• Evidence is currently insufficient to use continuous glucose monitoring (CGM) for screening or diagnosis.
• Classification guides personalized management; standardized islet autoantibody tests are recommended when an adult has features that overlap with type 1 diabetes.

Overview

The American Diabetes Association (ADA) Standards of Care 2026 explain how to diagnose diabetes and prediabetes, how to classify the type of diabetes and whom to screen. Diagnosis is based on A1C or plasma glucose (fasting, 2 hours after a 75-g glucose load, or random glucose with symptoms), and outside a clear hyperglycemic presentation it must be confirmed.

The section also covers presymptomatic type 1 diabetes, screening of high-risk groups and people on certain medications, and testing before and during pregnancy.

Diagnostic tests

FPG, 2-h plasma glucose during a 75-g OGTT and A1C are all appropriate for screening and diagnosis. They reflect different aspects of glucose metabolism, so they do not always identify the same people; the 2-h value diagnoses more people with prediabetes and diabetes than FPG or A1C. Fasting means no caloric intake for at least 8 hours.

FPG or A1C is usually preferred for routine screening because they are easier to perform. The OGTT is more sensitive and is preferred for some conditions, such as cystic fibrosis–related diabetes and post-transplantation diabetes. Before an OGTT, people should eat a mixed diet with at least 150 g of carbohydrate per day for 3 days; restricting carbohydrate beforehand can falsely raise post-load glucose. Glucose samples must be processed promptly, because glycolysis falsely lowers the result.

Confirming the diagnosis

• Symptoms or crisis: classic symptoms of hyperglycemia (polyuria, polydipsia, unexplained weight loss) or a hyperglycemic crisis plus random plasma glucose ≥200 mg/dL (≥11.1 mmol/L) are enough for the diagnosis; A1C helps show how long hyperglycemia has been present.
• No symptoms: confirm with two abnormal results, either two different tests at the same time (for example A1C and FPG) or a repeat test performed in a timely manner.
• Discordant results: repeat the test that is above its cut-off. A person with two A1C results ≥6.5% (≥48 mmol/mol) but FPG <126 mg/dL (<7.0 mmol/L) has diabetes.
• Results near a threshold: explain the possible symptoms of hyperglycemia and repeat testing in 3–6 months.

When A1C can mislead

A1C reflects glucose exposure over the red blood cell life span (about 120 days), weighted toward recent weeks. It may not be suitable for diagnosis in anemia, during erythropoietin treatment, on hemodialysis or during HIV treatment. Some assays are affected by hemoglobin variants; in sickle cell trait an assay without this interference should be used.

The laboratory method should be certified by the National Glycohemoglobin Standardization Program (NGSP) and traceable to the DCCT reference assay. Point-of-care A1C should be used for diagnosis only with FDA-approved devices in appropriately certified laboratories. When glucose and A1C consistently and substantially disagree, clinicians should look for an analytical problem or interference with either test; in people with diabetes, fructosamine or glycated albumin can be considered for monitoring.

Classifying the type of diabetes

• Type 1 diabetes: autoimmune β-cell destruction, usually leading to absolute insulin deficiency; includes latent autoimmune diabetes in adults (LADA).
• Type 2 diabetes: non-autoimmune, progressive loss of insulin secretion, often on a background of insulin resistance.
• Specific types: monogenic diabetes, diseases of the exocrine pancreas, and drug- or chemical-induced diabetes.
• Gestational diabetes mellitus (GDM): diabetes diagnosed in the second or third trimester that was not clearly present before pregnancy.

Both main types occur at any age, and some people cannot be clearly classified at diagnosis. Features most useful for identifying type 1 diabetes are diagnosis before 35 years, BMI <25 kg/m², unintentional weight loss, ketoacidosis and plasma glucose >360 mg/dL (>20 mmol/L) at presentation. Some people with type 2 diabetes present with ketoacidosis (ketosis-prone diabetes); their initial need for insulin is usually transient.

The ADA flowchart for newly diagnosed adults, based on data from White European populations, adds practical steps:

• Measure GAD antibodies first; if negative, test IA-2 and/or ZnT8 where available. Before 35 years, a negative result does not exclude type 1 diabetes, because 5–10% of people with type 1 diabetes have no antibodies.
• C-peptide is indicated only in people on insulin. A random sample within 5 hours of eating, taken with a glucose measurement, is enough; a result ≥600 pmol/L (≥1.8 ng/mL) is valid whatever the circumstances, and very low results (for example <80 pmol/L) do not need repeating. Do not test within 2 weeks of a hyperglycemic emergency.
• Monogenic diabetes is suggested by A1C <7.5% (<58 mmol/mol) at diagnosis, one parent with diabetes, features of a specific monogenic cause, or a MODY prediction-model probability >5%.

Type 1 diabetes: stages and screening

• Stage 1: multiple islet autoantibodies with normal glucose and normal A1C (presymptomatic).
• Stage 2: islet autoantibodies with dysglycemia: impaired fasting glucose, impaired glucose tolerance, A1C 5.7–6.4% (39–47 mmol/mol) or an A1C rise of ≥10% (presymptomatic).
• Stage 3: diabetes by standard criteria (symptomatic); autoantibodies may become absent.

Screening for presymptomatic type 1 diabetes uses autoantibodies against insulin, GAD, IA-2 or ZnT8 and should be offered to people with a family history of type 1 diabetes or a known high genetic risk. Anyone with one or more positive autoantibodies should be evaluated for stage 3 diabetes with A1C, urinalysis and/or plasma glucose. People with multiple confirmed autoantibodies should be referred to a specialized center for metabolic staging, education and consideration of prevention trials or approved treatments (for example teplizumab). A single confirmed IA-2 autoantibody is monitored like multiple autoantibodies; another single autoantibody is retested every 6 months to 3 years, depending on age.

Who to screen for prediabetes and type 2 diabetes

Testing should be considered in adults with overweight or obesity (BMI ≥25 kg/m², or ≥23 kg/m² in people of Asian ancestry) who have at least one risk factor:

• a first-degree relative with diabetes
• high-risk race, ethnicity or ancestry
• a history of cardiovascular disease
• hypertension (≥130/80 mmHg or on treatment)
• HDL cholesterol <35 mg/dL (<0.9 mmol/L) and/or triglycerides >250 mg/dL (>2.8 mmol/L)
• polycystic ovary syndrome
• physical inactivity
• other conditions associated with insulin resistance (for example severe obesity, acanthosis nigricans, metabolic dysfunction-associated steatotic liver disease)

For everyone else, testing begins at 35 years. People with prediabetes should be tested yearly and those with previous GDM at least every 1–3 years. After normal results, testing is repeated at a minimum of 3-year intervals, or sooner with symptoms or a change in risk such as weight gain. In children and adolescents, risk-based screening should be considered after the onset of puberty or after 10 years of age, whichever is earlier, if BMI is ≥85th percentile and there is at least one risk factor.

Specific situations

• Medications: consider screening people taking statins, thiazide diuretics or some HIV medications. With second-generation antipsychotics, test at baseline, 12–16 weeks after starting and then yearly. With recurrent or long-term glucocorticoids, monitor postprandial or random glucose.
• HIV: check FPG before starting or switching antiretroviral therapy and 3–6 months afterwards, then yearly if normal.
• Cancer therapy: with immune checkpoint inhibitors, test glucose before treatment and at each visit. With PI3Kα inhibitors, test glucose and A1C before treatment, random glucose weekly for 2 weeks and then every 4 weeks, and consider A1C every 3 months. With mTOR inhibitors, test glucose before treatment and at each visit, and consider A1C every 3 months.
• Pancreatitis: screen 3–6 months after acute pancreatitis and yearly thereafter; screen yearly in chronic pancreatitis.
• Cystic fibrosis: annual screening from 10 years of age, preferably with an OGTT. If an OGTT is not feasible, A1C 5.5–6.4% (37–47 mmol/mol) should lead to an OGTT within 3 months, and A1C ≥6.5% (≥48 mmol/mol) is consistent with cystic fibrosis–related diabetes.
• Organ transplantation: screen for hyperglycemia; diagnose post-transplantation diabetes once immunosuppression is stable and there is no acute infection, preferably with an OGTT.
• Genetic testing: recommended for everyone diagnosed with diabetes in the first 6 months of life (neonatal diabetes) and for children and young adults with atypical diabetes and diabetes in successive generations of the family (MODY).

Pregnancy

• Before pregnancy: screen people with risk factors and consider testing everyone of childbearing potential.
• Before 15 weeks: if not screened before conception, test people with risk factors (and consider testing everyone) for undiagnosed diabetes. Also screen for abnormal glucose metabolism, defined as A1C 5.9–6.4% (41–47 mmol/mol) or FPG 110–125 mg/dL (6.1–6.9 mmol/L), which marks a higher risk of adverse outcomes and of later GDM.
• 24–28 weeks: screen for GDM in everyone not already found to have diabetes or high-risk abnormal glucose metabolism earlier in the pregnancy.
• After GDM: a 75-g OGTT at 4–12 weeks postpartum, then lifelong screening every 1–3 years.

Questions and answers

Can A1C and fasting glucose from one blood sample confirm diabetes? Yes. If both are above the diagnostic threshold in the same sample, the diagnosis is confirmed without a second visit.

What if A1C and glucose results disagree? Repeat the test that is above its cut-off and consider factors that can distort A1C or glucose. The diagnosis is based on the confirmed test: two A1C results ≥6.5% (≥48 mmol/mol) mean diabetes even if FPG is below 126 mg/dL (7.0 mmol/L).

Can continuous glucose monitoring be used for diagnosis? No. The ADA considers the current evidence insufficient to support CGM for screening or diagnosing prediabetes or diabetes.

Source

American Diabetes Association Professional Practice Committee for Diabetes. 2. Diagnosis and Classification of Diabetes: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl 1):S27–S49. Published online 8 December 2025. DOI: 10.2337/dc26-S002. © 2025 American Diabetes Association; the source may be used for educational, noncommercial purposes if properly cited and unaltered. This overview is an independent summary by Medpresso and does not replace the original publication.