Fasting Glucose

Fasting glucose is the same blood sugar as any other glucose reading, measured under one strict rule: nothing but water for at least eight hours. That rule is what lets one morning's number be compared to the next.

Part of the Diabetes Panel — see all 9 values together, including Beta-Hydroxybutyrate, Estimated Average Glucose, Glycated Albumin.

Fasting glucose is not a different substance from the glucose on any other report. It is the same blood sugar, measured under one strict condition: nothing to eat or drink except water for at least 8 hours beforehand. The NIDDK and MedlinePlus both define the fast that plainly. The word "fasting" in the test's name isn't a label for the sample; it's the specification for how the sample was collected, and it does almost all the work of turning a reading into something you can act on.

Picture a wind tunnel. Engineers don't measure a car's drag on the open road, where a gust, a passing truck, or a damp patch of tarmac would change the answer every run. They put the car in a sealed tunnel with the air moving at a set speed, so two measurements taken a week apart describe the car and not the weather. The fast is that tunnel. Strip away the meals, the sugary coffee, the juice on the way out the door, and what's left is your blood sugar under controlled conditions, the same conditions you'll fast under next year. That comparability is the entire reason the rule exists.

It also means the reading is only ever as trustworthy as the conditions were. A number drawn after a leaky fast describes the gusts as much as the car, and that's where most fasting-glucose confusion lives.

What actually breaks the fast

The honest version is the one nobody volunteers at the counter. The classic broken fast isn't a forgotten breakfast; it's the small stuff that doesn't feel like eating:

  • coffee with milk, cream, or any sweetener (those carry calories, so the fast is over)
  • a "just a little juice" morning, or a sports drink on the commute
  • gum or mints, even sugar-free ones, which still set digestion going
  • a fast that was only six hours because the appointment moved earlier

Plain water is the one exception, and labs encourage it because a well-hydrated vein is an easier draw. Black coffee is the genuine gray area: it has almost no calories, but the CDC notes that some people's blood sugar is extra-sensitive to caffeine, so even unsweetened coffee can move the number. Most orders ask for water only precisely to delete that uncertainty rather than litigate it sample by sample. The practical rule is simpler than the worry: if you had anything but water, say so when the blood is drawn, and let the result be read for what it is rather than for what it was supposed to be.

Getting a clean fasted draw

  1. 1

    Count the hours honestly

    The NIDDK and MedlinePlus both define fasting as at least 8 hours with nothing but water. An overnight fast with a morning appointment is the easy way to clear that bar without thinking about it.

  2. 2

    Drink water, skip everything else

    Water is allowed and helps the draw go smoothly. Coffee, tea, juice, milk, and gum are not water, and any of them turns a fasting sample into a non-fasting one.

  3. 3

    Keep your routine otherwise normal

    The fast is the variable being controlled, so the days before should look like your ordinary life. A crash diet or a feast the night before makes the reading describe that, not your baseline.

  4. 4

    Speak up if the fast slipped

    If you ate, drank, or the gap was short, tell whoever draws the blood. A result read as fasted when it wasn't is the single most common way a number gets misjudged. A clean retest is usually a quick fix worth asking about.

Why the rule exists at all

Strip the conditions away and a glucose reading still tells you something true about that moment, but it stops being comparable to anything. After a meal, blood sugar climbs for a couple of hours in everyone, healthy or not, which is exactly what the body is built to do. A reading caught mid-climb can't be measured against the thresholds the NIDDK and the American Diabetes Association publish, because those numbers were drawn from fasted samples. The fast is what lines your result up with their ruler, and with your own result from last year.

This is also why a fasting glucose and an HbA1c are read together rather than against each other. The fasted number is one controlled measurement; HbA1c estimates the average of every uncontrolled hour over roughly three months. When a clean fasting result looks fine but the HbA1c runs higher, it often describes someone whose sugar behaves at 8 a.m. and drifts the rest of the day. The HbA1c versus fasting glucose comparison takes the two head to head.

The dawn phenomenon: a high reading under perfect rules

Here is the twist that catches careful people. The single most rigorously fasted draw of the day, taken first thing in the morning, can be the highest fasting reading you'll get. The CDC describes a surge in hormones in the early morning that happens whether or not a person has diabetes, the dawn phenomenon, and those hormones lift blood sugar as the body prepares to wake. Cortisol is one of them, rising in a predictable predawn climb. So an 8 a.m. fasted glucose can sit a few points above what the same fasted body would show at, say, 11 a.m.

The point isn't that morning draws are wrong. They're the standard, and the thresholds account for it. The point is that a fasted result near a borderline, a 100 or a 105, can reflect the hour as much as anything, which is one more reason a single fasting glucose is a prompt to confirm rather than a verdict. The general glucose page covers what the bands themselves mean once the fast is settled; this page is about getting the fast right so those bands apply.

What a properly fasted number can say

A clean fasting glucose earns the right to be read against the published cutoffs, and that's its whole value. After a genuine 8-hour fast, the NIDDK puts the normal band up to 99 mg/dL, calls 100 to 125 mg/dL the prediabetes range, and treats 126 mg/dL and above, confirmed on a repeat, as the diabetes threshold. To convert any of these, divide the mg/dL figure by 18, so 126 mg/dL is 7.0 mmol/L.

A random glucose can do none of that on its own. The exception the ADA allows is at the extreme: a glucose of 200 mg/dL or higher in someone with clear symptoms is diagnostic without a fast, because no amount of recent eating explains a number that high. Short of that, the fast is what gives the reading its meaning, which is why a borderline fasted value is usually rechecked under the same controlled conditions, often paired with HbA1c, before anyone reads a trend into it.

That single morning is also a poor judge on its own. A fasting glucose of 101 means one thing if last year's fasted draw was 88 and quite another if it was 99 and climbing. Because these numbers drift slowly, the diabetes panel guide walks through reading them in company, the guide to fasting for lab tests covers what the rule means across the whole draw, and as with most slow markers, the direction usually says more than any single value.

Try BloodSight

See your Fasting Glucose on one timeline.

BloodSight calibrates the reference range to your sex, age, and lab — and shows every value across every visit.

Fasting Glucose 5 visits
85 mg/dL −29
Mar Apr May Jun Jul

In your personal range

Normal ranges

Group Range Unit
Adult Male 70–99 mg/dL
Adult Female 70–99 mg/dL

Reference ranges may vary by laboratory and individual factors.

Fasting Glucose — Common Questions

What counts as fasting for a glucose test?
The NIDDK and MedlinePlus describe fasting the same way: nothing to eat or drink except sips of water for at least 8 hours before the draw. That rules out coffee, tea, juice, milk in your coffee, and anything with calories. Plain water is allowed and encouraged, since being well hydrated makes the draw easier. If your order doesn't say how long to fast, the lab can tell you, because the printed result is read as if that rule was followed.
Does coffee break a fasting glucose test?
Coffee with milk, sugar, or any sweetener clearly breaks the fast, because those have calories. Black coffee is the gray area: it has almost no calories, but the CDC notes that some people's blood sugar is extra-sensitive to caffeine, so even unsweetened coffee can nudge the number. Most labs ask for water only to remove the question entirely. If you had coffee, tell whoever draws the blood, so the result is read for what it is.
Why is my fasting glucose higher in the morning?
Often because of the dawn phenomenon. The CDC describes a surge in hormones in the early morning that happens whether or not someone has diabetes, and those hormones raise blood sugar. So the most carefully fasted draw of the day, taken first thing, can read higher than a fasted sample would later on. It's a normal feature of how the body wakes up, not a sign the fast failed.
What is the difference between fasting glucose and random glucose?
Same molecule, different rules. A fasting glucose is drawn after at least 8 hours of water only, so the diagnostic thresholds the NIDDK and ADA publish apply to it directly. A random glucose is drawn at any time, including after eating, and reads higher without meaning anything is wrong. The fast is what lets one morning's number be compared with another and with the published cutoffs.
I accidentally ate before my glucose test. Does it still count?
It becomes a non-fasting sample, which is read against different expectations than a fasted one. A short fast or a forgotten snack doesn't ruin your day, but it does change what the number means, so it shouldn't be measured against the fasting thresholds. The fix is simple: tell the lab what you had and when, and ask whether a clean fasted retest is worth booking.
How do I convert fasting glucose to mmol/L?
Divide the mg/dL figure by 18. A fasting 99 mg/dL is about 5.5 mmol/L, 100 is about 5.6, and the 126 mg/dL diabetes threshold is 7.0 mmol/L. US labs report mg/dL and much of the rest of the world uses mmol/L, so the same fasted sample reads as either number depending on where it was run.

Disclaimer

This content is for informational and educational purposes only. It is not intended as medical advice, diagnosis, or treatment recommendation. Reference ranges may vary by laboratory. Always discuss your results with a qualified healthcare professional.

Related Tests

Glucose

Glucose is a single photograph of your blood sugar, captured the instant the needle goes in. Whether you had eaten, the hour of day, even the stress of the draw can change what the picture shows.

Hemoglobin A1c HbA1c

One number that remembers everything: roughly three months of blood sugar, weighted toward recent weeks, immune to last-minute virtue. That long memory is also where the test can go wrong.

Insulin

Glucose tells you where your blood sugar landed. Fasting insulin tells you how hard your body worked to put it there, which is often the part the standard panel never shows.

Cortisol

Cortisol runs on a daily timetable, peaking around the moment you wake and thinning out by midnight. Read without the clock beside it, the number on your report can look alarming and mean almost nothing.

Estimated Average Glucose eAG

Estimated average glucose is your HbA1c spoken in the language of a glucose meter. No glucose was measured to produce it, which is why it almost never matches the average on your meter.

HOMA-IR HOMA-IR

HOMA-IR is not a test the lab runs. It is a number your report calculates from two others, an efficiency score for how much insulin your body burns to hold a normal blood sugar.

Albumin-to-Creatinine Ratio uACR

The uACR is a urine test, not a blood test, and it answers a question eGFR can't: whether your kidney filters are leaking. A trace of protein in urine can show up years before filtration ever slips.

Triglycerides

Triglycerides are the most movable number on the lipid panel. The fast everyone associates with a blood test exists, more than anything, to hold this one number still.

C-Peptide

Every insulin molecule the pancreas releases is paired with one C-peptide. Counting the stub tells you how much insulin your body makes, even when the insulin itself is gone in minutes or arrived from a syringe.