Fasting glucose

What it actually measures

You have not eaten since last night. Somewhere between eight and twelve hours have passed, and your body has been running on stored fuel the entire time.

And yet your blood sugar has stayed in a remarkably narrow band the whole night. Not because nothing happened, but because a great deal happened. Your liver released stored glucose at exactly the rate you burned it. Insulin rose and fell to keep the release matched to the demand. All of it while you slept.

That is what a fasting glucose is actually measuring. Not what you ate. The quality of an overnight regulation job you were not awake for.

Which makes it a genuinely elegant test. One number, taken at the calmest possible moment, describing how well a system holds itself steady when nothing is being thrown at it.

And it makes the interpretation subtler than most people are told, because a system can hold steady while working extremely hard, and the number looks exactly the same either way.

Normal and optimal are different questions

The thresholds here are unusual among the markers on this site, because they are not statistical. They are diagnostic, chosen deliberately by guideline committees, largely on the basis of the risk of developing diabetes and its complications.

Under 100 mg/dL normal. 100 to 125 impaired fasting glucose. 126 and above, on two occasions, diabetes.

That looks precise, and here is the honest wrinkle: several international guidelines set the impaired fasting glucose floor at 110 rather than 100. So the exact same result of 105 gets called prediabetes in one country and normal in another. The biology did not change at the border.

The functional target most root-cause practitioners work toward is roughly 75 to 90, and the argument for it is straightforward. Risk relates to blood sugar continuously, not as a switch that flips at a threshold. Someone sitting at 98 has been on a different trajectory from someone sitting at 82, and both reports say normal.

There is a caution I want to attach to that, because chasing a low fasting glucose is not a virtue in itself. Prolonged fasting can raise the number as the liver releases stored glucose, stress can raise it, and a single value bounces. This is a marker to read as a trend and alongside its companions, not one to optimise obsessively in isolation.

The years this number cannot see

Alright. Here is the part that should reframe the whole page, so listen close.

Insulin resistance does not begin when your fasting glucose crosses 100. By the time it crosses 100, the process has usually been running for a long time.

Here is the sequence. Your cells become gradually less responsive to insulin. Your pancreas notices that glucose is drifting up and responds by producing more insulin. More insulin does the job. Glucose comes back down. Everything reads normal.

That compensation can go on for years. Your fasting glucose stays under 100 the entire time, and every annual physical returns the same reassuring sentence: your blood sugar is fine.

It is fine. It is being held fine, at an increasing cost that this test does not measure.

Fasting glucose only starts to rise when the compensation begins to fail. Which means, and I want to be precise about this rather than dramatic, that a rising fasting glucose is often a late signal rather than an early one.

So the question is not whether your fasting glucose is normal. It is what it is costing to keep it there, and there is a test for that. Fasting insulin measures the effort rather than the result.

That is one extra line on a lab order, on the same draw, answering the question the standard panel structurally cannot.

What the progression data actually shows

If your result does land in the prediabetes band, the useful question is what that predicts, and the answer is more nuanced than the label suggests.

A 2026 systematic review and meta-analysis in BMJ Nutrition, Prevention and Health screened 11,980 papers and included 40 studies of adults with prediabetes published between 2006 and 2024, specifically to work out which test or combination of tests best predicts progression to diabetes.

Two findings are worth carrying away.

First, the highest risk came from impaired fasting glucose in the 6.1 to 6.9 mmol/L band, with a hazard ratio of 9.0. That is not a small signal.

Second, and more practically, combinations mattered. People meeting multiple criteria at once, impaired fasting glucose plus impaired glucose tolerance plus an HbA1c of 6.0 to 6.4 percent, had the highest annual incidence at 15.2 percent. Generally, combination tests were associated with higher progression rates than any single test.

The takeaway is not that prediabetes means diabetes is coming. It is that how many doors are open matters more than any one of them. Which is a good reason to look at HbA1c and fasting insulin alongside this number rather than reading it alone.

And it is worth saying plainly: this trajectory responds to what you do. That is not motivational language, it is the reason the progression rates differ so much between groups in the first place.

What moves it

Down: resistance training, which builds the muscle that stores most of your glucose and is probably the highest return single change here. Walking after meals, which blunts the spike. Reducing refined carbohydrate and added sugar. Losing excess visceral fat. Adequate sleep, since even a few short nights measurably reduce insulin sensitivity. Managing chronic stress. And where prescribed, medication, which is a doctor's decision.

Up: insulin resistance progressing past the point of compensation. Poor sleep and chronic stress, both acting through cortisol. Corticosteroids, which raise it substantially, along with some other medications. Acute illness. Excess visceral fat and inactivity. And an unusually long fast before the draw, which raises rather than lowers it.

The honest hedge

I am not a doctor and I am not diagnosing anyone. Diabetes is a diagnosis, it requires a physician, and a single number on a page is not it.

Two things I want to hold together here. A fasting glucose in the diabetes range is a genuine medical finding that needs a doctor rather than a diet plan, and I would not want anyone reading this page to treat it as a project to handle alone.

At the same time, a normal fasting glucose is weaker reassurance than it sounds, for the structural reason above rather than out of caution. The test cannot see the compensation, and the compensation is where the years are.

I would also rather you did not chase this number in isolation. It bounces, it responds to a bad night's sleep, and a single value tells you much less than three values across eighteen months.

I am not saying your metabolism is failing. I am saying it is a variable, an early-detectable one, and the earliest part of it is measured by a test almost nobody orders.

So. What to actually do.

Tonight, at no cost. Line up your fasting glucose from every blood test you can find, with dates. A value of 94 means one thing on its own and something quite different if it read 82 five years ago.

Calculate your triglyceride to HDL ratio while you are in there. Your ratio is free, calculable from a lipid panel you already have, and one of the better free flags for insulin resistance.

Add the test that sees earlier. Fasting insulin measures the effort your body is spending to hold this number steady. Same draw, one extra line, and it is the number that moves first.

Add HbA1c for the three month view. HbA1c catches the after-meal spikes a single fasting value misses entirely.

Start with the muscle. Resistance training two or three times a week plus a ten minute walk after your largest meal. That combination moves this marker more reliably than any supplement, and both are free.

Take the set to someone who reads it together. A good functional medicine doctor, or your physician if the numbers are in diagnostic territory.

None of this has to happen this week. But tonight you can line up your old glucose values and see which way they have been walking.

You are not failing at blood sugar. You are a regulation system that has been holding a line all night, every night, for years, and being graded only on whether the line held rather than on what it cost.

Your body is not broken. It is blocked. And sometimes the block is years of quiet compensation, invisible behind a number that kept coming back normal because the compensation was working.

Go find out what it is costing.