Prediabetes: what do the criteria actually mean?
Prediabetes means that one or more validated glucose tests fall above the usual range but below the diabetes threshold. It signals higher future risk—not certainty, not a moral failure, and not ‘mild diabetes.’
The quick answer
Prediabetes is a heterogeneous risk state, not one uniform disease.
Fasting plasma glucose, A1C, and oral glucose tolerance capture different aspects of glucose regulation and can disagree. The closer a value is to the diabetes threshold—and the more abnormalities or risk factors present—the greater the concern. A useful response verifies context, assesses overall cardiometabolic risk, and offers an evidence-based prevention plan rather than predicting that diabetes is inevitable.
Three established windows
Fasting glucose, A1C, and 2-hour oral glucose testing identify overlapping—not identical—groups.
Risk is continuous
A result just inside the range does not carry the same risk as several values near the diabetes threshold.
Progression can be delayed
Structured lifestyle programs reduce diabetes incidence in high-risk adults; medication is selective, not automatic.
The central distinctionA threshold helps organize decisions. It does not convert a continuous risk into a guaranteed future diagnosis.
Diagnostic ranges
Each test defines a prediabetes range and has a different limitation.
Pattern recognition
Four results that should not receive the same message.
The number, test, trend, and person all matter. These examples organize the next question without diagnosing an individual.
Lower end of one risk range
Review test quality and A1C interference, then integrate age, family history, adiposity, medicines, and other risk factors.
Concordant higher-range pattern
Risk is generally greater when multiple tests are abnormal and close to diabetes thresholds; timely follow-up matters.
Tests can disagree
Post-challenge dysglycemia may be missed by A1C or fasting glucose. Discordance is a reason to investigate, not average the numbers.
Use the diabetes confirmation pathway
In an asymptomatic person, a diabetes-range result usually requires prompt confirmation according to the applicable guideline.
MERHI ONE Evidence Map
What is established—and what the label cannot prove.
The rating belongs to each precise indication or claim. A laboratory category does not automatically validate every test, diet, drug, or supplement marketed around it.
Fasting glucose, A1C, and 2-hour oral glucose testing identify guideline-defined prediabetes ranges.1,3,5
U.S. and Brazilian guidance use these tests, but they measure different physiology and may classify the same person differently.
An intensive, structured lifestyle program reduces progression to type 2 diabetes in high-risk adults.2,4,6
In the original DPP, a program targeting weight reduction, nutrition, physical activity, and behavior lowered diabetes incidence more than placebo during the randomized phase.
Prevention can delay diabetes over the long term, although the between-group effect narrows.2,7
DPPOS follow-up continued to show lower diabetes incidence in the original lifestyle and metformin groups, while crossover and later shared interventions complicate long-term comparisons.
Metformin can be considered for selected adults at particularly high risk.2,4,6
ADA and SBD do not recommend automatic medication for every prediabetes result. Age, BMI, prior gestational diabetes, glycemic level, safety, and preferences change the decision.
A CGM spike, fasting insulin value, or HOMA-IR alone diagnoses prediabetes.1,3
Prediabetes criteria are based on validated plasma glucose, A1C, and oral glucose testing pathways—not a universal insulin or consumer-sensor threshold.
Preventing the diabetes label with one intervention automatically prevents cardiovascular events.8
In 21-year DPPOS follow-up, the original lifestyle and metformin assignments did not significantly reduce major cardiovascular events, despite delaying diabetes.
Two native pathways
The core ranges agree; screening and oral-challenge pathways differ.
United States — ADA / USPSTF
- ADA uses fasting plasma glucose 100–125 mg/dL, A1C 5.7–6.4%, or 2-hour OGTT 140–199 mg/dL and recommends at least annual monitoring when prediabetes is present.
- ADA screening is broader from age 35; USPSTF specifically recommends screening asymptomatic adults 35–70 with overweight or obesity and preventive intervention when prediabetes is found.
- The USPSTF core screening tests are fasting glucose, A1C, and the oral glucose tolerance test.
Brazil — Sociedade Brasileira de Diabetes
- SBD uses the same fasting glucose, HbA1c, and 2-hour TTGO ranges and recommends annual reassessment for confirmed prediabetes.
- The current SBD pathway also includes 1-hour TTGO glucose of 155–208 mg/dL for detecting prediabetes and ≥209 mg/dL for diabetes.
- SBD recommends lifestyle intervention for all people with prediabetes and reserves medication discussion for selected higher-risk situations.
A useful response
Turn a result into a risk-reduction plan.
Verify the context
Check fasting status, acute illness, medicines, laboratory method, and whether A1C may be unreliable.
Estimate gradient of risk
Consider how close values are to diabetes thresholds, whether multiple tests agree, prior gestational diabetes, family history, adiposity, and trajectory.
Assess cardiovascular risk
Measure blood pressure and appropriate lipoproteins and address smoking and other established risk factors directly.
Look beyond glucose
Kidney function, albuminuria, MASLD risk, sleep apnea, and medicines may change both risk and priorities.
Use structured prevention
Build realistic nutrition, activity, resistance training, sleep, and weight goals with support and follow-up.
Individualize medication
Discuss metformin or other options only when guideline-supported benefits, risks, cost, and the person’s priorities are clear.
Common interpretation errors
Prediabetes is easy to exaggerate—and easy to dismiss.
Good communication protects against both overmedicalization and missed prevention.
‘You will become diabetic’
Progression is not inevitable and risk varies substantially.
‘It is only borderline’
Higher-range or repeated abnormalities can signal meaningful risk.
‘A1C 5.7% proves insulin resistance’
A1C is a glycemic exposure marker, not a direct insulin-sensitivity test.
‘Every glucose spike is prediabetes’
Consumer sensor patterns are not the diagnostic criteria.
‘One diet works for everyone’
Evidence supports structured patterns and sustainable energy balance, not one universal menu.
‘A supplement reverses prediabetes’
A biomarker change does not prove durable prevention or safety.
What poor interpretation can cause
- Anxiety and stigma
- Unnecessary restrictive diets
- Unvalidated testing and supplement cost
- Missed diabetes-range results
- Delayed cardiovascular risk treatment
- A plan that is impossible to sustain
A practical next step
Six questions to discuss after a prediabetes-range result.
- 01
Which validated test was abnormal, and could preparation or a known interference have changed it?
- 02
Is the result near the lower end or close to the diabetes threshold, and do other tests agree?
- 03
Should the test be repeated, confirmed, or complemented with an oral glucose tolerance test?
- 04
Which cardiovascular, liver, kidney, sleep, medicine, and family-history factors change overall risk?
- 05
What structured prevention program is realistic, accessible, culturally appropriate, and measurable?
- 06
When should follow-up occur, and is there a guideline-supported reason to discuss medication?
Frequently asked questions
Direct answers to the questions people actually ask.
Is prediabetes the same as diabetes?+
No. It is a glucose range below the diabetes threshold that signals increased future risk.
Will everyone with prediabetes develop diabetes?+
No. Risk varies, and some people remain stable or return below the range, especially with effective prevention.
Which test is best?+
There is no universal winner. Fasting glucose, A1C, and oral glucose testing answer different questions and can disagree.
Should an unexpected result be repeated?+
Often yes—especially when it is close to a threshold, conflicts with other data, or may have been affected by illness or an A1C interference. The appropriate confirmation pathway depends on the result and guideline.
Does everyone need metformin?+
No. It is considered for selected higher-risk adults after individualized review.
How often should prediabetes be checked?+
ADA and SBD guidance generally support at least annual monitoring, adjusted for the value, trajectory, risks, symptoms, and clinical context.
Scientific sources
A risk category needs context—not fear.
We prioritize current U.S. and Brazilian guidelines, public-health recommendations, and randomized prevention trials. Evidence levels apply to specific claims, not to the label as a whole.
- 01Open source ↗
American Diabetes Association · 2026
Standards of Care in Diabetes — Diagnosis and Classification
- 02Open source ↗
American Diabetes Association · 2026
Prevention or Delay of Diabetes and Associated Comorbidities
- 03Open source ↗
Sociedade Brasileira de Diabetes · 2026 edition
Diagnosis of diabetes mellitus — Brazilian guideline
- 04Open source ↗
Sociedade Brasileira de Diabetes · 2026 edition
Pharmacologic treatment of prediabetes — Brazilian guideline
- 05Open source ↗
U.S. Preventive Services Task Force · 2021
Screening for Prediabetes and Type 2 Diabetes
- 06Open source ↗
Diabetes Prevention Program Research Group · 2002
Reduction in the Incidence of Type 2 Diabetes With Lifestyle Intervention or Metformin
- 07Open source ↗
Diabetes Prevention Program Outcomes Study · 2015
Long-term effects of lifestyle intervention or metformin over 15 years
- 08Open source ↗
Diabetes Prevention Program Outcomes Study · 2022
Long-term interventions and cardiovascular events over 21 years
Editorial record