Her bloodwork was normal for nine years.
It was also true for nine years. That's the part nobody explains.
Every spring, the same routine. Fast overnight, sit for the draw, wait a few days, get the call that says everything looks good. Glucose fine. A1C fine. Cholesterol fine. See you next year.
And then somewhere around year ten the numbers finally moved, and the word that came back was prediabetes — as though it had started that month.
It had been underway the whole time. The panel just wasn't measuring the part that was changing.
This isn't a story about a bad doctor or a bad lab. Every one of those results was accurate. Every one of them was inside its printed reference range. Nothing was misread and nothing was ignored.
The test that would have caught it years earlier simply wasn't on the page.
A1C is the last thing to break. Not the first.
Insulin is the key that lets glucose out of your bloodstream and into your cells. Insulin resistance means the lock has gotten stiffer — the same key doesn't turn as easily as it used to.
Your body's response to a stiff lock is not to give up. It's to press harder. The pancreas makes more insulin, and more insulin still gets the job done. Glucose goes where it's supposed to go. Blood sugar stays exactly where it should be.
This is called the compensated phase, and it can run for years.
During all of it, a fasting glucose is normal — because glucose is being cleared. An A1C is normal — because A1C is a three-month average of that same cleared glucose. Both tests are working correctly. They are measuring the result. What has changed is the effort.
Glucose only rises when the pancreas finally can't keep up.
That's the moment the panel notices. And by then the thing it noticed has often been developing for a long time — quietly, at full strength, while every annual report said normal.
Which is why an abnormal A1C is a late signal rather than an early warning. It's not a smoke detector. It's the fire department arriving.
| Test | Result | Reference range | Flag |
|---|---|---|---|
| Glucose, fasting | 94 | 70–99 mg/dL | Normal |
| Hemoglobin A1C | 5.4 | 4.0–5.6 % | Normal |
| Creatinine | 0.82 | 0.57–1.00 mg/dL | Normal |
| ALT | 26 | 7–35 U/L | Normal |
| Cholesterol, total | 186 | <200 mg/dL | Normal |
| Triglycerides | 142 | <150 mg/dL | Normal |
| HDL cholesterol | 51 | >50 mg/dL | Normal |
| LDL cholesterol | 106 | <130 mg/dL | Normal |
| TSH | 2.10 | 0.450–4.500 uIU/mL | Normal |
| Vitamin D, 25-OH | 38 | 30–100 ng/mL | Normal |
| Insulin, fasting | — | — | Not ordered |
Nothing on this panel is wrong. Every result falls inside its range, so nothing gets flagged, and nothing gets flagged means nobody looks further. The range is what ends the conversation.
There are two, and you want both.
One is a measurement. The other is that measurement put in context.
How much insulin your pancreas is producing after an overnight fast, when it should be at its quietest. It's a simple blood draw and it can be added to the same fasting sample you're already giving.
Most labs print a reference range somewhere in the neighbourhood of 2–25 uIU/mL, which is wide enough that a result can sit comfortably inside it while still being higher than it was ten years ago. Ranges vary between labs, so read yours off your own report rather than off anyone's chart.
Not a separate draw. It's a calculation that pairs your fasting insulin with your fasting glucose, which is what makes it more useful than either number alone:
(insulin × glucose) ÷ 405
This is the value most of the research actually uses. Thresholds differ between studies and populations, and it isn't a diagnostic cut-off the way an A1C is — it's a way of watching a direction over time. Which is exactly what makes it worth having a first one on file.
“Next time I'm in for fasting labs, could you add a fasting insulin? I'd like to have a HOMA-IR to track.”
Your doctor didn't miss this.
I want to be careful here, because there's a version of this page that turns into an accusation, and that version would be wrong.
Fasting insulin is not standard of care for screening. The major diabetes guidelines don't recommend it for that purpose. So a physician following current guidance to the letter orders glucose and A1C and stops — and that is not negligence, that's the protocol working as written.
The gap isn't a person. It's the protocol.
And there are real reasons for the caution. Insulin assays aren't standardised the way glucose assays are, so results aren't perfectly comparable between labs. Reference ranges are broad. There's genuine disagreement about where an “optimal” value sits, and anyone quoting you a single confident number is quoting an opinion.
So here's where it honestly lands: fasting insulin and HOMA-IR are well-supported for tracking a trend in yourself over time, and not established as a screening test with clean diagnostic thresholds. Both of those are true at once.
That's still worth having. A first value now is what makes a second value in two years mean something. Without a baseline, you're doing what the annual panel already does — waiting for a result to cross a line.
Ask for it as a baseline, not as a verdict. That framing is also the one most likely to get a yes.
A number with no next step is just anxiety.
So before you get it, know what it would actually change.
The reason this test is worth asking for is that insulin sensitivity responds to ordinary things — and it responds early, which is the whole advantage of finding out early.
Muscle is where glucose goes.
Skeletal muscle is the largest destination for glucose in your body, and contracting muscle pulls glucose in through a pathway that doesn't need insulin to open the door. More muscle means more room. A used muscle means an open door.
This is also why a ten-minute walk after eating does more than the same ten minutes at another hour. You're giving the meal somewhere to go while it's arriving.
Sleep moves this number, and moves it fast.
Short sleep measurably reduces insulin sensitivity, and it doesn't take months — studies have produced meaningful changes in healthy people within a handful of nights. It's one of the few inputs here that shifts on a timescale you can actually feel.
If you're going to change one thing before you retest, this is the one with the shortest distance between the change and the result.
The order of a meal changes its size.
The same food, eaten in a different order, doesn't land the same way. Protein, fat and fibre ahead of the starch blunts how sharply glucose — and therefore insulin — rises afterward. Nothing is removed from the plate. The plate is just sequenced.
It's a small thing that costs nothing, which is exactly the kind of thing worth doing before you attempt a large thing that costs a lot.
Some of this isn't about food at all.
Chronically elevated cortisol raises blood glucose by design — that's the hormone doing its job — and sustained stress therefore sits upstream of this whole picture. Falling oestrogen through perimenopause also shifts insulin sensitivity, which is part of why a body that handled carbohydrate fine at forty can feel like a different body at fifty.
And some medications move it directly. If a doctor has ever told you a prescription might affect your blood sugar, that's the same conversation as this one.
Which of these matters most depends on what you're already doing.
Four levers, and they don't weigh the same for everybody. If you're already lifting three times a week, sleep is your lever. If you're sleeping seven and a half hours and still crashing at three, the meal order is doing more work than you think.
That's what My Timing Map is for. It reads your actual stack, lays your day out in five blocks, and flags what's blocking what — including the things landing in a window that's working against you.
It also tells you what you already got right without anybody telling you.
Try every tool free for 7 days.
Drop your email and unlock all 12 tools — unlimited use, no credit card, no commitment. The peptide calculator that does the math. The supplement timing that actually matters. The interaction checker that catches what your pharmacy missed.
No spam · No hard sells · Cancel anytime