The short answer
A normal A1c is below 5.7% at every age. The cutoffs for prediabetes (5.7–6.4%) and diabetes (≥ 6.5%) don't change when you turn 50, 60, or 70. No major guideline — ADA, CDC, WHO — publishes age-specific diagnostic thresholds.
What does change with age is the treatment target for people who already have diabetes. A younger adult with a long life expectancy and no complications may aim for an A1c under 6.5–7%. A frail 80-year-old on five medications, living alone, with a history of falls — aiming for that same number risks hypoglycemia that's more dangerous than the high glucose would have been. Their target may reasonably be under 8% or even 8.5%.
This guide walks through both sets of numbers — the diagnostic ones (which don't change) and the treatment targets (which do) — so you know which applies to your question.
A1c diagnostic chart (same at every age)
These are the numbers used to diagnose prediabetes and diabetes. They apply to non-pregnant adults of any age.
| Category | A1c | What it means |
|---|---|---|
| Normal | Below 5.7% | No action needed based on A1c alone. |
| Prediabetes | 5.7% – 6.4% | Elevated risk of progressing to type 2 diabetes. Lifestyle change has strong evidence behind it. |
| Diabetes | 6.5% or higher | Needs confirmation on a separate day or with another abnormal test (fasting glucose ≥ 126, OGTT 2-hr ≥ 200, random ≥ 200 with symptoms). |
You will see articles online claiming that "normal A1c for a 60-year-old is 6.0%" or similar. They are wrong. They often confuse the slow, small population-level rise in A1c with age (discussed below) with a diagnostic target. A 60-year-old with an A1c of 6.0% has prediabetes by the same criteria as a 30-year-old with the same reading.
A1c treatment targets by age and health profile
If you already have diabetes, the target your clinician sets is individualized. The ADA's 2025 Standards of Care asks clinicians to weigh life expectancy, hypoglycemia risk, duration of diabetes, and comorbidities. Here are the profiles the guidelines describe, as rough anchors:
| Profile | Typical A1c target |
|---|---|
| Younger adult, recent-onset diabetes, no significant comorbidities | Under 7% (under 6.5% if achievable safely) |
| Most adults with type 2 diabetes | Under 7% |
| Older adult, healthy, long life expectancy, few medications | Under 7.0 – 7.5% |
| Older adult, multiple chronic conditions, intermediate life expectancy | Under 8% |
| Older adult, very complex health, limited life expectancy | Under 8.5% (or simply "avoid hypoglycemia and symptomatic hyperglycemia") |
| Pregnancy with pre-existing or gestational diabetes | Under 6.0 – 6.5% |
| Children and adolescents with type 1 diabetes | Under 7% (individualized) |
These are starting points for a conversation with a clinician, not rules. A 72-year-old who cycles 100 miles a week and takes one medication is not the same patient as a 72-year-old recovering from a stroke, even though they share a decade.
Why targets get looser with age
The logic behind looser targets in older adults is not "they matter less." It's that tight control in frail, older people produces more harm than benefit.
The ACCORD trial, published in the New England Journal of Medicine in 2008, randomized over 10,000 adults with type 2 diabetes (average age 62) to either intensive glucose control (A1c target under 6.0%) or standard control (7.0–7.9%). The intensive-control arm was stopped early because mortality was higher — not lower — in that group. The ADVANCE and VADT trials found similar neutral or negative results for aggressive lowering in older, long-duration diabetics.
The mechanism is hypoglycemia. Low blood sugar in an older adult causes falls, fractures, arrhythmias, and cognitive impairment, and it can be the direct cause of death. The calculation changes: if someone's life expectancy is 20+ years, an A1c of 7.8% matters — microvascular complications accumulate. If someone's life expectancy is 5 years and they live alone, an A1c of 7.8% probably does not meaningfully change their trajectory, but a single severe hypoglycemic episode can.
This is why clinicians ask about falls, cognitive status, and who is home with you — not because they're being nosy, but because those answers shift what "good control" means for this specific patient.
The small population-level rise with age
Large population studies — NHANES in the U.S., Framingham, Whitehall II — do show a small average rise in A1c as people get older. Roughly 0.1 percentage points per decade past age 40, in non-diabetic adults.
It's real, but it's small, and it does not justify relaxing the diagnostic cutoff. Most of the drift is explained by age-related changes in body composition (more visceral fat, less muscle) and the increasing prevalence of undiagnosed insulin resistance. When you control for BMI and activity level, the age effect largely disappears.
What this means practically: a 70-year-old with an A1c of 5.8% is in the prediabetes range. They have slightly higher odds of having landed there for "normal aging" reasons than a 30-year-old with the same reading, but the clinical implications are the same. Lifestyle intervention still works. The number still predicts risk.
Special cases
Children and adolescents
The same diagnostic cutoffs apply: below 5.7% is normal, 5.7–6.4% is prediabetes, ≥ 6.5% is diabetes. In a child, type 1 diabetes is more common than type 2, and the diagnosis is usually made from symptoms (rapid-onset thirst, frequent urination, weight loss) and a clearly elevated glucose, not from a screening A1c. For children already diagnosed with type 1 diabetes, the ADA target is generally under 7%, but is strongly individualized because hypoglycemia in kids carries its own set of concerns.
Pregnancy
A1c is less useful in pregnancy because red blood cell turnover speeds up and A1c values drift down by week 12–16. Gestational diabetes is diagnosed with the OGTT, not A1c. For women with pre-existing type 1 or type 2 diabetes who become pregnant, the ADA target is an A1c under 6.0–6.5% to minimize birth complications, with close attention to avoiding hypoglycemia.
Hemoglobin variants
Sickle cell trait, HbC trait, and related variants can make A1c unreliable depending on the assay method. Fructosamine or continuous glucose monitoring may be used instead. This affects an estimated 8% of Black Americans and a smaller fraction of other populations. If your A1c seems inconsistent with your finger-stick readings, it's worth asking about hemoglobin electrophoresis.
Anemia and recent blood loss
Iron deficiency anemia, B12/folate deficiency, and chronic kidney disease can all push A1c upward without a real change in glucose. Hemolytic anemia, recent transfusion, and pregnancy push it down. If your A1c and your fasting glucose disagree dramatically, one of these may be the reason.