A 46-year-old bank officer comes for a routine health check. He has no symptoms. His report shows a fasting plasma glucose of 131 mg/dL and an HbA1c of 6.1 per cent, both from the same sample. The intern looks at the two numbers and asks, “Sir, is he diabetic or not? One says yes and one says no.” This is an illustrative case, not a real patient.
The Diabetes Diagnosis tool answers that question at the bedside. It reads each value against the current thresholds, applies the confirmation rule and the HbA1c exclusions, and draws each value beside the treatment target. What it no longer carries is most of the why: where the thresholds came from, why two respected bodies still disagree, and the Indian studies that show the tests do not find the same people. Those have moved here, so that the tool stays quick. By the end of this article you will know why the numbers sit where they do, and what to do when two of them disagree.
Use this at the bedside
Open the tool: Diabetes Diagnosis
Get the CDSS app: install the MEDiscuss CDSS on your phone
1. Where the numbers came from
“Why 126? It looks so arbitrary.” It is not. A diagnostic threshold is set where the risk of diabetes-specific damage, retinopathy above all, begins to rise. The fasting threshold was once 140. It came down to 126 in 1997 because that is where the prevalence of retinopathy rises, and where it matches a 2-hour value of 200. Here is how each threshold arrived, as the tool carried it on 4 November 2026.
| Year | Document | What moved |
|---|---|---|
| 1979 | National Diabetes Data Group | Fasting 140 mg/dL, 2-hour 200 mg/dL, impaired glucose tolerance defined |
| 1980 and 1985 | WHO Expert Committee and Study Group | Adopted the same frame internationally |
| 1997 | ADA Expert Committee | Fasting lowered to 126 mg/dL, chosen because that is where the prevalence of retinopathy rises and where it matches the 2-hour 200; impaired fasting glucose introduced at 110 to 125 |
| 1999 | WHO | Adopted 126 and 110 to 125; a single value never diagnoses an asymptomatic person |
| 2003 | ADA follow-up report | Impaired fasting glucose lowered to 100. WHO did not follow, in 2006 or since |
| 2009 and 2010 | International Expert Committee; ADA | HbA1c 6.5 per cent for diagnosis; the ADA added 5.7 to 6.4 |
| 2011 | WHO; DETECT-2 | WHO accepted 6.5 with stringent quality assurance and no intermediate range. DETECT-2 pooled 44,623 people and found the retinopathy threshold at a fasting glucose of 6.5 mmol/L and an HbA1c of 6.5 per cent |
| 2024 | International Diabetes Federation | Proposed the 1-hour value, 155 and 209 mg/dL. Not adopted by ADA 2026 or by any Indian document |
Read the 2011 row twice. DETECT-2 pooled 44,623 people and found the retinopathy threshold at a fasting glucose of 6.5 mmol/L and an HbA1c of 6.5 per cent. That is the evidence that put the HbA1c on the same footing as the glucose tests.
2. Why the ADA and WHO still disagree at 100 and 110
The upper end of normal fasting glucose has two values, and neither is a mistake. The ADA lowered the floor from 110 to 100 mg/dL in 2003 to make the fasting category as sensitive as impaired glucose tolerance. WHO kept 110 in 2006 because the lower floor more than doubles the number of people labelled, without evidence that the added group benefits from intervention. RSSDI 2022 follows WHO; ICMR 2018 prints both.
So a fasting glucose of 105 mg/dL is prediabetes in one clinic and normal in the next. The common mistake is to argue about which is right. Write down which range you used, and be consistent with your patient.
3. The tables in full
The tool now shows its threshold and target tables as smaller tables of three columns or fewer, so that they fit a phone. Here they are whole, with the source of each row beside it.
Diagnostic thresholds, non-pregnant adults.
| Test | Normal | Intermediate | Diabetes | Source |
|---|---|---|---|---|
| Fasting plasma glucose | <100 mg/dL (5.6 mmol/L) on the ADA range; <110 mg/dL (6.1) on the WHO and RSSDI range | Impaired fasting glucose: 100 to 125 mg/dL (ADA); 110 to 125 mg/dL (WHO 1999 and 2006, RSSDI 2022). ICMR 2018 prints both and adjudicates neither | ≥126 mg/dL (7.0 mmol/L) | ADA 2026, WHO 2006, ICMR 2018, RSSDI 2022 |
| 2-hour plasma glucose, 75 g OGTT | <140 mg/dL (7.8 mmol/L) | Impaired glucose tolerance: 140 to 199 mg/dL (7.8 to 11.0 mmol/L) | ≥200 mg/dL (11.1 mmol/L) | All four agree |
| HbA1c | <5.7 per cent (39 mmol/mol) on the ADA range. WHO defines no intermediate range | 5.7 to 6.4 per cent (39 to 47 mmol/mol), ADA and RSSDI | ≥6.5 per cent (48 mmol/mol), NGSP-certified assay | International Expert Committee 2009, ADA 2010 onward, WHO 2011, ICMR 2018, RSSDI 2022 |
| Random plasma glucose | Not a screening threshold | Not defined. A value of 140 to 199 mg/dL calls for a fasting test or an HbA1c | ≥200 mg/dL (11.1 mmol/L) with classic symptoms or a hyperglycaemic crisis | ADA 2026, ICMR 2018 |
| 1-hour plasma glucose, 75 g OGTT | <155 mg/dL (8.6 mmol/L) | ≥155 mg/dL: intermediate hyperglycaemia | ≥209 mg/dL (11.6 mmol/L), to be confirmed by repeat | IDF position statement 2024 only. Not adopted by ADA 2026, WHO, ICMR or RSSDI |
Thresholds in pregnancy.
| Test | Gestational diabetes | Notes | Source |
|---|---|---|---|
| DIPSI: 75 g glucose, irrespective of the last meal, 2-hour plasma glucose | ≥140 mg/dL (7.8 mmol/L) | 120 to 139 mg/dL is gestational glucose intolerance and is followed up. The national test at the first contact and at 24 to 28 weeks. Against IADPSG its sensitivity was 22.6 per cent in 1,031 women | MoHFW 2018, DIPSI |
| IADPSG one-step: fasting 75 g OGTT | Fasting ≥92 mg/dL (5.1), 1-hour ≥180 (10.0), 2-hour ≥153 (8.5). One value diagnoses | Derived from HAPO at 1.75 times the odds of an adverse outcome. Raises the diagnosed prevalence roughly one- to threefold | IADPSG 2010, WHO 2013, ADA 2026 |
| Two-step: 50 g screen then 100 g OGTT | Screen positive at 130, 135 or 140 mg/dL at 1 hour; then two of fasting 95, 1-hour 180, 2-hour 155, 3-hour 140 (Carpenter-Coustan) | The United States alternative. Not used in India; printed for reference only | ADA 2026 Table 2.8 |
| Overt diabetes in pregnancy | The non-pregnant criteria: fasting ≥126, 2-hour ≥200, random ≥200 with symptoms, and HbA1c ≥6.5 per cent before 15 weeks (ADA 2.31a). From 15 weeks the HbA1c is set aside and plasma glucose decides (ADA 2.4) | Classified as diabetes complicating pregnancy, not GDM, whenever it is found. Reclassify every GDM at 4 to 12 weeks post partum (ADA), 6 weeks under the national programme, using the non-pregnant criteria | ADA 2026, MoHFW 2018 |
Treatment targets, which are not normal values.
| Population | HbA1c | Fasting or pre-meal | After meals | Source |
|---|---|---|---|---|
| Many non-pregnant adults with diabetes | <7.0 per cent (53 mmol/mol); <6.5 may be appropriate where it can be reached without hypoglycaemia | 80 to 130 mg/dL (4.4 to 7.2 mmol/L) | <180 mg/dL (10.0 mmol/L) at the peak, 1 to 2 hours after the start of the meal | ADA 2026, section 6 |
| Continuous glucose monitoring | Time in range 70 to 180 mg/dL above 70 per cent; below 70 mg/dL under 4 per cent; below 54 under 1 per cent | A time in range of 70 per cent corresponds to an HbA1c near 7 per cent | ADA 2026, section 6 | |
| ICMR 2018 | Ideal <7 per cent; satisfactory 7 to under 8 | Ideal 80 to 110 mg/dL; satisfactory 111 to 125 | 2-hour ideal 120 to 140 mg/dL; satisfactory 141 to 180 | ICMR 2018, section 4.1 |
| RSSDI 2022 | <7.0 per cent | 160 mg/dL, as long as hypoglycaemia is avoided | RSSDI 2022 | |
| Older adult, generally healthy | <7.0 to 7.5 per cent | 80 to 130 mg/dL | Bedtime 80 to 180 mg/dL | ADA 2026, section 13 |
| Older adult, complex or intermediate health | <8.0 per cent (64 mmol/mol) | 90 to 150 mg/dL | Bedtime 100 to 180 mg/dL | ADA 2026, section 13 |
| Older adult, very complex or poor health | Do not rely on HbA1c; avoid hypoglycaemia and symptomatic hyperglycaemia | 100 to 180 mg/dL | Bedtime 110 to 200 mg/dL | ADA 2026, section 13 |
| Pregnancy, any diabetes | <6 per cent (42 mmol/mol) if achievable without hypoglycaemia; relaxed to <7 if necessary | <95 mg/dL (5.3 mmol/L) | 1-hour <140 mg/dL or 2-hour <120 mg/dL | ADA 2026, section 15 |
“So if his HbA1c on treatment is 6.8, is he normal now?” No. That is the difference the tool is built to show. A target is set where the harm of pushing glucose lower, hypoglycaemia above all, begins to outweigh the benefit. It sits inside the range that the diagnostic threshold calls abnormal. A person at target still has diabetes.
4. The three tests do not find the same people
In the CURES cohort, the prevalence of diabetes in 2,188 adults was 6.1 per cent by fasting glucose, 10.1 per cent by the 2-hour value and 12.8 per cent by HbA1c. Only 121 of the 281 diagnosed met all three criteria. The ADA records that the 2-hour value diagnoses more people than either of the others, and that the trials preventing progression were done in people with impaired glucose tolerance, not in isolated impaired fasting glucose or HbA1c-defined prediabetes.
“Then why not use an Indian HbA1c cut-off?” One was looked for. The same Chennai laboratory found HbA1c cut points of 6.1 per cent against the 2-hour criterion and 6.4 against the fasting one, and 5.6 for the intermediate group at under 70 per cent accuracy. These are printed as context. The tool applies the international 6.5 and 5.7, as ICMR and RSSDI do.
The bedside lesson is short. A fasting glucose alone misses roughly a third of Indian diabetes. When the fasting value is in the impaired range, the 2-hour test is the one that changes management.
5. Why the HbA1c misleads more often in India
HbA1c is a weighted average over the life of the red cell, so anything that shortens or lengthens that life moves the number without any change in glucose. Think of it as a diary kept by each red cell. If the cells live longer, as in iron deficiency, the diary covers more weeks and reads high. If they die early, after haemolysis, bleeding or a transfusion, the diary is short and reads low.
Iron deficiency is the one that matters most here. NFHS-5 found anaemia in 57 per cent of women aged 15 to 49. In a woman with iron deficiency, diagnose on plasma glucose, and do not let a “high” HbA1c make the diagnosis alone.
6. How common is it, and where does prediabetes lead?
- ICMR-INDIAB 2023, 113,043 adults across every state: diabetes 11.4 per cent, prediabetes 15.3 per cent, both by WHO criteria. In many states with a lower human development index there is more prediabetes than diabetes.
- CURES 10-year follow-up, Chennai: 58.9 per cent of people with prediabetes converted to diabetes, and 19.4 per cent of those with normal glucose tolerance converted directly to diabetes within a decade. Among the highest conversion rates recorded anywhere.
- IDPP-1, 531 Indian adults with impaired glucose tolerance: 55 per cent of the control group had diabetes within 3 years. Lifestyle advice reduced that by 28.5 per cent and metformin by 26.4 per cent; combining them added nothing. Number needed to treat 6.4.
This is why “borderline sugar” is not a reassurance in India. It is the most useful moment to act.
7. Back to the bank officer
Here is how the confirmation rule settles him, step by step. A 46-year-old man with no symptoms has a fasting glucose of 131 mg/dL and an HbA1c of 6.1 per cent from the same sample. The fasting value is in the diabetes range; the HbA1c is not. The tests disagree, so the fasting glucose is repeated on another day. If the repeat is 126 or more, he has diabetes, whatever the HbA1c says. If it is 119, he has impaired fasting glucose and is retested in 3 to 6 months. Neither test outranks the other; the repeated one decides.
“Couldn’t we just believe the HbA1c? It is the newer test.” It is tempting, because it feels more stable. But the rule is that the test above the threshold is the one repeated, and the repeated result decides. His repeat fasting glucose came back at 129 mg/dL. He has diabetes.
8. Pregnancy, and DIPSI
India tests pregnant women with DIPSI because it needs no fasting and one visit. Against IADPSG it is insensitive: 22.6 per cent in 1,031 women in one Chennai series. So a negative DIPSI in a woman with risk factors does not settle the question, and a fasting 75 g test with IADPSG thresholds is better where it is feasible. The other Indian studies of DIPSI, and what they mean, are in Diabetes Classification: The Evidence Behind the Pathway, section 6.
Try this
A 32-year-old woman, not pregnant, has a haemoglobin of 9.1 g/dL with a low ferritin. Her HbA1c is 6.6 per cent and her fasting glucose is 98 mg/dL. Does she have diabetes? Illustrative case.
Answer. Not on these results. Iron deficiency makes the HbA1c read falsely high, so it cannot be used for her diagnosis. Diagnose on plasma glucose. Her fasting value is normal on both the ADA and the WHO range. If you are still worried, a 75 g glucose tolerance test with a 2-hour value is the test that finds the most people. And treat the iron deficiency.
Take-home points
- A normal value and a treatment target answer different questions; a person at target still has diabetes.
- Without unequivocal hyperglycaemia, two abnormal results are needed; when two tests disagree, repeat the one above the threshold.
- A fasting glucose of 100 to 109 mg/dL is prediabetes on the ADA range and normal on the WHO range; record which you used.
- In iron deficiency, haemoglobin variants, advanced kidney disease and pregnancy from 15 weeks, diagnose on plasma glucose, not HbA1c.
- A negative DIPSI in a woman with risk factors does not settle the question.
Read with this: Diabetes Classification: The Evidence Behind the Pathway, for what comes after the diagnosis: which type of diabetes it is; Pregnancy Dating and the Antenatal Schedule: The Evidence Behind the Tool, for when to test in pregnancy; and The Smart Plate: How to Eat Well With Diabetes, the Indian Way, for your patient.
Use this at the bedside
Open the tool: Diabetes Diagnosis
Get the CDSS app: install the MEDiscuss CDSS on your phone
Declaration: the author built and maintains the tool. It is free to use. The figures and tables in this article are those the tool carried in its Criteria and Evidence tabs on 4 November 2026, with the sources it named. The DIPSI sensitivity of 22.6 per cent is from a Chennai series (Mohan and colleagues, Acta Diabetologica 2014), which the CDSS cites in full in its Pregnancy and Antenatal Scheduler rather than in this tool’s own list.
References
- American Diabetes Association Professional Practice Committee. 2. Diagnosis and classification of diabetes: Standards of Care in Diabetes-2026. Diabetes Care 2026;49(Suppl 1):S27-S49.
- American Diabetes Association Professional Practice Committee. 6. Glycemic goals, hypoglycemia, and hyperglycemic crises: Standards of Care in Diabetes-2026. Diabetes Care 2026;49(Suppl 1):S132-S149.
- American Diabetes Association Professional Practice Committee. 13. Older adults: Standards of Care in Diabetes-2026. Diabetes Care 2026;49(Suppl 1):S277-S296.
- American Diabetes Association Professional Practice Committee. 15. Management of diabetes in pregnancy: Standards of Care in Diabetes-2026. Diabetes Care 2026;49(Suppl 1):S321-S338.
- World Health Organization, International Diabetes Federation. Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia: report of a WHO/IDF consultation. Geneva: WHO; 2006.
- World Health Organization. Use of glycated haemoglobin (HbA1c) in the diagnosis of diabetes mellitus: abbreviated report of a WHO consultation. Geneva: WHO; 2011. WHO/NMH/CHP/CPM/11.1.
- Indian Council of Medical Research. ICMR guidelines for management of type 2 diabetes 2018. New Delhi: ICMR; 2018.
- Research Society for the Study of Diabetes in India. RSSDI clinical practice recommendations for the management of type 2 diabetes mellitus 2022. Int J Diabetes Dev Ctries 2022;42(Suppl 1):1-143.
- Maternal Health Division, Ministry of Health and Family Welfare, Government of India. Diagnosis and management of gestational diabetes mellitus: technical and operational guidelines for the implementation of maternal health programme. New Delhi: MoHFW; 2018.
- International Association of Diabetes and Pregnancy Study Groups Consensus Panel, Metzger BE, Gabbe SG, et al. International Association of Diabetes and Pregnancy Study Groups recommendations on the diagnosis and classification of hyperglycemia in pregnancy. Diabetes Care 2010;33(3):676-82.
- Bergman M, Manco M, Satman I, et al. International Diabetes Federation position statement on the 1-hour post-load plasma glucose for the diagnosis of intermediate hyperglycaemia and type 2 diabetes. Diabetes Res Clin Pract 2024;209:111589.
- Riddle MC, Cefalu WT, Evans PH, et al. Consensus report: definition and interpretation of remission in type 2 diabetes. Diabetes Care 2021;44(10):2438-44.
- National Diabetes Data Group. Classification and diagnosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 1979;28(12):1039-57.
- Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997;20(7):1183-97.
- Genuth S, Alberti KG, Bennett P, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care 2003;26(11):3160-7.
- International Expert Committee. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care 2009;32(7):1327-34.
- Colagiuri S, Lee CM, Wong TY, Balkau B, Shaw JE, Borch-Johnsen K; DETECT-2 Collaboration Writing Group. Glycemic thresholds for diabetes-specific retinopathy: implications for diagnostic criteria for diabetes. Diabetes Care 2011;34(1):145-50.
- Anjana RM, Unnikrishnan R, Deepa M, et al. Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study (ICMR-INDIAB-17). Lancet Diabetes Endocrinol 2023;11(7):474-89.
- Anjana RM, Shanthi Rani CS, Deepa M, et al. Incidence of diabetes and prediabetes and predictors of progression among Asian Indians: 10-year follow-up of the Chennai Urban Rural Epidemiology Study (CURES). Diabetes Care 2015;38(8):1441-8.
- Mohan V, Vijayachandrika V, Gokulakrishnan K, et al. A1C cut points to define various glucose intolerance groups in Asian Indians. Diabetes Care 2010;33(3):515-9, read with Nazir A, Papita R, Anbalagan VP, Anjana RM, Deepa M, Mohan V. Prevalence of diabetes in Asian Indians based on glycated hemoglobin and fasting and 2-h post-load (75-g) plasma glucose (CURES-120). Diabetes Technol Ther 2012;14(8):665-8.
- Ramachandran A, Snehalatha C, Mary S, Mukesh B, Bhaskar AD, Vijay V; Indian Diabetes Prevention Programme. The Indian Diabetes Prevention Programme shows that lifestyle modification and metformin prevent type 2 diabetes in Asian Indian subjects with impaired glucose tolerance (IDPP-1). Diabetologia 2006;49(2):289-97.
Leave a Reply