The short answer: medicine is not automatic
No. An HbA1c result in the prediabetes range is not a ready-made prescription, and it does not mean that everyone should start metformin that day. It means average glucose exposure is above the usual range but remains below the diabetes threshold used by the ADA for a nonpregnant adult. A sound decision starts by confirming what the result means, then considering risk factors, health conditions and the person’s ability to follow and review a plan [1, 3, 4, 5].
For many people, the first plan is a realistic programme for food, activity, sleep and weight management when excess weight is present, with a clear review date. For a higher-risk group, a clinician may discuss metformin as an addition rather than a replacement for that foundation. Large trials found that intensive lifestyle intervention and metformin delayed diabetes in selected high-risk adults, with lifestyle intervention having the larger overall effect in the original DPP; trial results are not a promise of one person’s outcome [2, 9, 10].
The useful question is therefore not only, “Will I take medicine?” It is: is the result reliable, what is my level of risk, which step can I sustain, and when will we review its effect? If medicine is discussed, it then becomes part of a shared decision with a clear reason rather than a reaction to fear about one number [2, 8].
Understand the result first: HbA1c is not the whole picture
HbA1c summarises average glucose exposure over roughly the previous three months and does not require fasting. Under ADA criteria, the prediabetes range for a nonpregnant adult is 5.7% to 6.4%. Alternative measures are fasting plasma glucose from 100 to 125 mg/dL or two-hour glucose during an oral glucose tolerance test from 140 to 199 mg/dL. These tests measure different aspects of glucose regulation, so they do not always agree completely [1, 5].
If a result is close to the diabetes threshold or does not fit symptoms or glucose measurements, a clinician may repeat it promptly or select fasting glucose or a tolerance test. In the absence of unequivocal hyperglycaemia with classic symptoms or a glucose crisis, a diabetes diagnosis requires confirmation with another abnormal result. This avoids treating one measurement as a final diagnosis while still taking a result that needs follow-up seriously [1].
HbA1c can be misleading when red-cell lifespan changes, after recent bleeding or transfusion, with some anaemias or haemoglobin variants, during pregnancy, and in some kidney or liver conditions. Tell the clinician about these circumstances, medicines and supplements; plasma-glucose criteria may sometimes be more appropriate. Do not use a home meter to diagnose yourself. Home monitoring has specific treatment roles, but diagnosis requires a laboratory test interpreted in clinical context [1, 5, 6].
What makes risk higher than the HbA1c number alone?
The number matters, but its position and direction within the range matter too. An HbA1c nearer the upper boundary, higher fasting glucose, or abnormality on more than one test may justify closer follow-up. Family history, low activity, excess weight (particularly abdominal fat) high blood pressure, unhealthy blood lipids, polycystic ovary syndrome, previous gestational diabetes and some medicines such as long-term glucocorticoids also enter the assessment [1, 4, 7].
This does not mean that a person in a larger body will inevitably develop diabetes or that a lean person is protected. It means the clinician assembles the picture rather than judging by appearance or one test. If the terms and tests are confusing, signs of insulin resistance explains why tiredness, hunger or weight gain cannot confirm the diagnosis. When lipids are raised, blood pressure and lipids explains why cardiovascular prevention is not decided from one isolated item [1, 4].
Prediabetes is also an opportunity to review blood pressure, lipids, smoking, sleep and medicines because health risk is broader than progression to diabetes alone. This is not a reason for endless testing; it is a reason to choose measurements that can genuinely change the plan. Record previous results and their dates, because a trend over time says more than a number without context [1, 7, 8].
No medicine does not mean no plan
Lifestyle change here is not the vague instruction to “just stop sugar.” The DPP used an intensive programme with ongoing support to change eating, increase activity and achieve moderate weight loss among participants who had excess weight. Its results support structured intervention as a way to reduce or delay diabetes in high-risk people, and a prevention difference persisted during longer follow-up even as effects between groups changed over time [9, 10].
In practice, start with one or two measurable aims: suitable walking or activity on most days, a meal built around vegetables, protein and fibre instead of a highly processed option, fewer sugar-sweetened drinks, more regular sleep, or a change that supports management of excess weight without a crash diet. There is no need to remove every carbohydrate. Type, amount, meal pattern and context all matter, and the plan should fit work, culture, medicines and other health conditions [2, 4, 8].
If you need help turning this into a routine, clinical nutrition can start with household food and available time instead of an idealised menu that will not last. Agree on something to review: a specific habit, weight when appropriate, blood pressure, or a laboratory result at the planned time. If the plan stalls, review the barrier rather than assigning blame. Redesign is part of treatment and does not mean that you failed [2, 8].
When might metformin enter the discussion?
The 2026 ADA Standards say metformin for prevention should be considered in adults at high risk, especially those resembling higher-risk DPP participants: age 25 to 59 years, BMI at least 35, fasting glucose higher within the range, HbA1c at least 6.0%, or a history of gestational diabetes. These factors open a discussion; they are not a checklist in which one item makes medicine compulsory [2].
The clinician also asks whether the result is reliable, whether kidney function and overall health permit treatment, which adverse effects or other medicines matter, what benefit is expected for this person, and how progress will be judged. With long-term use, the ADA advises considering periodic vitamin B12 assessment, particularly when anaemia or peripheral-neuropathy symptoms are present. Starting, dosing and monitoring belong with the treating clinician; do not use another person’s prescription [2].
Evidence shows that metformin delayed diabetes in the high-risk DPP population, but the ADA notes that long-term data do not currently support other glucose-lowering medicines solely to prevent type 2 diabetes. Weight-management medicines may have a separate indication for selected people with overweight or obesity, but that is a distinct treatment decision with its own assessment, benefits and harms. A prediabetes-range HbA1c alone should not be turned into a reason for one particular medicine [2, 9, 10].
Follow-up: make a clear agreement instead of testing at random
The ADA recommends testing people with prediabetes yearly for progression to diabetes, and NICE similarly recommends at least annual blood testing for people confirmed to be at high risk. Timing may be sooner when the value is close to the diagnostic threshold, symptoms appear, a relevant medicine is started, or the first result needs confirmation. The clinician decides whether repeating the same test will best show a trend or whether another test is needed to answer a specific question [1, 4, 8].
There is no automatic need to check glucose at home every day or buy a continuous glucose monitor solely because of a prediabetes result. The ADA states that evidence is insufficient to use CGM for screening or diagnosis of prediabetes or diabetes. Frequent measurements without a defined question may increase anxiety and produce values that are difficult to interpret because meals, timing and device technique all affect them [1].
Leave the review with four written points: what the result means, one or two actions you will take, whether medicine is being considered and why, and the date of the next test or visit. If the diagnosis changes to confirmed diabetes, follow-up becomes broader. A whole-health diabetes review covers areas such as blood pressure, kidneys, eyes and feet; its full scope should not automatically be imposed on prediabetes [1, 2]. If you have type 2 diabetes and your numbers have improved, read why diabetes remission is not a cure.
When should you not wait for routine follow-up?
Prediabetes usually causes no clear symptoms. If marked thirst, frequent urination, new blurred vision, unintentional weight loss, recurrent infections or increasing fatigue appears, arrange prompt assessment and glucose testing instead of assuming this is simply the old result. These symptoms can accompany high glucose, but an article cannot diagnose their cause [1, 4].
Seek urgent care for persistent vomiting, inability to keep fluids down, clear dehydration, rapid or difficult breathing, severe abdominal pain, unusual drowsiness, confusion, fainting or rapid deterioration. These are not symptoms for a diet experiment or a laboratory test months later. If you are unsure how urgent the situation is, contact an urgent medical service rather than trying to decide alone [1].
Pregnancy follows a different pathway, and gestational-diabetes thresholds and management are not the same as a prediabetes decision in a nonpregnant adult. Tell the clinician about pregnancy, plans for pregnancy, or previous gestational diabetes because these change assessment, follow-up and the medicine discussion. Likewise, do not stop a glucocorticoid, antipsychotic or other prescribed treatment merely because it may affect glucose; ask the prescriber how to review it safely [1, 2, 4].
The next step: prepare for a review that ends with a decision
Bring the HbA1c result and date, any fasting-glucose or older results, and a list of medicines and supplements with doses. Note changes in weight, activity or sleep, and any family history of diabetes, previous gestational diabetes, high blood pressure, abnormal lipids or PCOS. This helps the clinician distinguish a result needing confirmation, risk needing a stronger programme, and a situation in which a medicine discussion may be reasonable [1, 2, 7].
Ask directly: is this truly prediabetes; could anything make my HbA1c inaccurate; which modifiable risk factor matters most; is metformin expected to help in my circumstances and how would safety be monitored; when should the test be repeated; and what should bring me back sooner? Write down the answers so the plan does not remain a set of vague instructions [1, 2, 5, 6].
If you need an individual assessment, you can book an appointment at the clinic to review the results and health history and agree on a measurable next step. The clinic is mentioned here only to organise the final next step, not as a promise of an outcome or a substitute for emergency care. Until review, do not start medicine independently, and do not treat a later improved value as permission to abandon follow-up; the trend and plan belong in review with the treating clinician [2, 8].
