A number on a lab report is a poor thing to search the internet about, because the first answer back is usually a chart, and the chart rarely says which test it applies to. A fasting sample, a two-hour sample taken during a glucose tolerance test, and an HbA1c are three different measurements of three different things, read against three different thresholds.

Add the fact that Indian practice and the international bodies do not fully agree on where one of those lines sits, and a single reading can look reassuring in one column and concerning in the next. What follows is how to work out which column yours belongs in.

• Fasting glucose, the two-hour OGTT value, and HbA1c answer different clinical questions and are read against separate thresholds. They are not interchangeable.
• Indian professional bodies (RSSDI and ESI) follow the ADA definition of prediabetes, not the WHO definition. The two set the lower edge of impaired fasting glucose differently, at 100 mg/dL and 110 mg/dL respectively.
• A fasting glucose of 104 mg/dL sits inside the prediabetes band under ADA criteria and outside it under WHO criteria, without anything about the person changing.
• Age-banded blood sugar charts found online are treatment targets for people already on medication, not diagnostic thresholds for healthy undiagnosed adults.
• A single abnormal result is not a diagnosis. Indian criteria require confirmation on a separate day unless the value is unequivocally abnormal.
• HbA1c measures glucose exposure over 8 to 12 weeks, not insulin resistance. A normal HbA1c can sit alongside a metabolic picture that is not normal.

Three measurements do most of the work in adult metabolic screening, and they answer different questions.

A fasting plasma glucose is a single point in time after at least eight hours without food. A two-hour value is taken during a standardised 75 g oral glucose tolerance test, and it is the only one of the three that shows how the body handles a glucose load rather than how it sits at rest. HbA1c is a percentage rather than a concentration, and it reflects an average over the preceding weeks rather than any one morning.

One detail derails more comparisons than any other. Fingerprick capillary samples and venous plasma samples do not read the same, and Indian population surveys that use capillary glucose apply adapted thresholds because of it. A number from a home glucometer or a camp screening is therefore not directly comparable with a number from a laboratory plasma sample.

The bands below are the ones used for screening and diagnosis in adults who are not pregnant. They are stated for venous plasma and reflect the criteria in use as of 2026.

Normal: Blood sugar is considered normal when fasting plasma glucose is below 100 mg/dL, the two-hour value on an OGTT is below 140 mg/dL, and HbA1c is below 5.7%.

Prediabetes: The prediabetes range is a fasting plasma glucose of 100 to 125 mg/dL, a two-hour OGTT value of 140 to 199 mg/dL, or an HbA1c of 5.7% to 6.4%.

Diabetes: The diabetes threshold is a fasting plasma glucose of 126 mg/dL or above, a two-hour OGTT value of 200 mg/dL or above, or an HbA1c of 6.5% or above.

The fasting threshold is where the disagreement sits. Those figures are the American Diabetes Association's, and Indian professional guidance has largely followed them. The joint RSSDI-ESI recommendations explicitly endorse the ADA criteria for the Indian context.

The World Health Organization sets the lower edge of impaired fasting glucose higher, at 110 mg/dL, and does not accept HbA1c for identifying prediabetes at all. The reasoning behind the gap is that the lower threshold captures a considerably larger group, including many people whose glucose returns to normal without intervention, while the higher one identifies a smaller group carrying more progression risk. So a fasting glucose of 104 mg/dL genuinely sits inside the prediabetes band under one rule set and outside it under another, without anything about the person changing.

Search for normal blood sugar and age-banded tables appear immediately, often with wider ranges for children and different targets by decade. They are not screening thresholds, and applying them to a routine report is the most common mistake made with these numbers.

The diagnostic criteria published by the ADA, the WHO, and Indian professional bodies are stated as single adult thresholds. They are not banded by age. The age-specific numbers in circulation are mostly treatment targets: ranges set for a person already diagnosed with diabetes and already on glucose-lowering medication, where the target is individualised by age, other conditions, and the risk of the level falling too low. A target of that kind is set by the clinician managing that person, and it says nothing about whether an undiagnosed adult's result is normal.

Two practical consequences follow:

• A healthy adult in their sixties is read against the same diagnostic thresholds as a healthy adult in their thirties.
• A number pulled from a paediatric or a treatment-target chart will make an ordinary adult result look either alarming or falsely reassuring, depending on which chart was found first.

HbA1c measures the fraction of haemoglobin with glucose attached, reflecting average blood glucose over the preceding 8 to 12 weeks, weighted toward the most recent month. No fasting is needed, and it does not swing with what happened at breakfast.

India has done validation work behind the threshold it uses. RSSDI's expert panel recommends 6.5% as the diagnostic cut-off for diabetes in Indian patients, drawing on pooled data from four Indian centres. A Chandigarh community study found that 6.5% gives 88% specificity while 7.0% gives 92% sensitivity.

RSSDI cautions plainly that HbA1c cannot be used as the sole measurement for diagnosing diabetes in Indian settings. For prediabetes it has adopted the 5.7% to 6.4% band, while noting a South Indian study that proposed 5.9% instead, a figure it cites without adopting.

Values shift for reasons that have nothing to do with a person's baseline metabolism:

• Recent acute illness, physiological stress, and steroid exposure can all raise glucose transiently.
• How long the fast actually ran matters for the fasting sample.
• Sample handling matters too, since tubes need separating within roughly an hour and haemolysed samples are rejected for good reason.

Which is why Indian diagnostic criteria require an abnormal result to be repeated on a separate day before it is treated as diagnostic, unless the value is unequivocally abnormal or the person has classic symptoms. A repeat request is not a sign that something was missed. It is the criteria working as written.

The most useful questions to bring to a glucose result are which test produced it, which criteria it was read against, and whether it has been repeated. Those three settle most of the confusion that a chart creates, and none of them requires another test.

A result near a threshold is a reason for a conversation rather than a conclusion, since diagnosis in India rests with a registered medical practitioner after a consultation and not with a table or an app.

A glucose result near a threshold is a reason for a proper review, not a search engine query. A Voy clinician can place the number alongside your history and the rest of your metabolic panel and tell you what it actually means for you. Book a metabolic assessment.

This article is for general information and education only and is not medical advice, diagnosis, or treatment. GLP-1 and other medications referenced are prescription-only and are appropriate only for certain people under the supervision of a qualified clinician. Do not start, stop, or change any medication based on this article. Please consult a registered medical practitioner about your individual circumstances. Information reflects what was available at the time of review and may change.