90 days, three blood draws, one number: what HbA1c actually showed

The number on a January morning

Sam was 38 the morning his patient portal sent him a number. The number was 6.0. The portal called it borderline. His doctor's note, attached at the bottom, was three sentences. Let's monitor for three months. Lifestyle adjustments first. Recheck in March, and we'll see where we are. Sam read the note in his car in a parking lot.

His dad had been diagnosed with Type 2 in his early fifties. Sam knew the family math. He drove home and did not say anything to his wife until the next morning.

He decided to do something most patients in his position don't do. He decided to test his HbA1c himself, monthly, out of pocket, through a diabetes panel that ran around eighty dollars a draw. He wanted three numbers, not one.

Why monthly when HbA1c is a 90-day average

The question was reasonable, and Sam asked it of himself before he asked it of anyone else.

HbA1c reflects the average glucose exposure of his red blood cells, which live around ninety days. A single monthly draw, then, would be heavily weighted toward the most recent month but still pulled by the previous two. Any one number would be noisy. He understood that.

He wanted the noise. Three noisy numbers, plotted in order, would tell him something a single number in March would not. He could see direction. He could see whether month one had moved anything at all. He could see whether the curve was the curve he was hoping for.

He called the cost tuition. He said it to himself once and never said it out loud.

Month 1: what changed in the routine

He didn't make a plan. He made small adjustments, most of them not technically about HbA1c.

Walks after dinner became a default — twenty-five minutes, sometimes thirty. The dog needed walking, and Sam started looking forward to the walks more than he had expected.

The candy jar that had lived on his desk since 2019 moved to a coworker's desk on a Tuesday. The coworker was unsentimental about candy. Sam stopped opening the drawer where he had moved his backup snacks.

Breakfast, briefly, moved later — 9:30 instead of 7:00, on the theory he had read about somewhere. By 10:30 most mornings he was hungry and irritable, so by week three he had reverted to 7:30, just without the cereal that had been the reason he ate breakfast at 7:00 in the first place.

He didn't count carbs. He didn't do keto. He didn't change dinner much. He still drank a beer on Friday and a different beer on Saturday.

He bought a digital food scale and used it twice.

Draw 2: 5.9

The mid-February draw came back at 5.9. Sam had quietly hoped for 5.7. The portal note was unchanged in tone — borderline — and he sat with the number for a longer minute than he wanted to admit.

Then he remembered why he was testing monthly. HbA1c averages roughly ninety days. The hardest changes he had made were in week two and week three of January. By the February blood draw, those weeks were maybe a quarter of the average. The number was not yet allowed to know what he had been doing.

A different number on the same panel had moved more meaningfully. His fasting glucose, which he had been spot-checking with a cheap drugstore meter, had drifted from 102 in the first week to 94 by the time of the second draw. His insulin was slightly lower as well, and his HOMA-IR followed it down. The leading indicators were moving while the lagging one was still catching up.

He did not call his wife with the 5.9. He did call her with the 94.

Months 2 and 3: what stuck and what didn't

The walks stuck. They had stopped feeling like effort by mid-February. The dog was happier and Sam slept slightly better, neither of which he could attribute to the walks specifically, but he had stopped trying to attribute things.

Breakfast time stayed at 7:30. The cereal did not come back. He ate eggs, sometimes oatmeal, sometimes nothing if the morning was strange.

The candy jar was, by then, fully forgotten. He occasionally remembered it the way you remember an old roommate.

The spot fasting glucose readings continued descending: 94, 91, 88. Around mid-March he stopped checking the meter daily. It had started feeling like a chore, and he did not want to make this a chore — that had been a quiet promise to himself in January. Once a week, he decided, was the maximum he would tolerate before the testing became the thing he was managing instead of the thing it was supposed to measure.

Draw 3: 5.6

April. Same lab, same Tuesday morning slot, fasted twelve hours.

HbA1c: 5.6. Below the 5.7 threshold the labs use as the prediabetes line. His insulin was lower again. His estimated average glucose had moved from 126 to 114.

He sent the screenshot to his doctor through the portal. The reply came that afternoon. Good. Same plan. Recheck in six months.

It was not cured. Sam knew what cured looked like, and this wasn't it. He was back across the line, which was a different thing. The line moved easily, in either direction. He had spent ninety days getting his number from one side of it to the other, and the line was, in some sense, three weeks of a different routine away from being on the other side again.

His wife was happy with the 5.6. She also, he noticed, did not treat it as the end of anything, which was the response he had been hoping for and the response he would not have known how to ask for.

What Sam learned about the number itself

By the third draw, Sam thought he understood the metric better than he had in January.

HbA1c is a slow mirror. By the time it moves, the change is already weeks behind. Fasting glucose is the leading indicator — by months two and three, his fasting was well down, and the A1c was still catching up to what the fasting had already told him.

Monthly HbA1c testing isn't necessary for most people. Twice a year is enough, and Sam knew that. What the monthly testing did for him was specific: it gave him structure for what would otherwise have been a year of guessing whether anything he was doing was working.

Most of what changed for him, he came to think, wasn't food in any deep sense. It was structure. When he ate. Where the snacks lived. What happened after dinner. The food scale, used twice, was the thing that had pretended to be the answer. The walks, undertaken because the dog needed them, were closer.

It is October. Sam still walks after dinner. The dog is older now and walks more slowly than in January, and Sam walks at her pace, which is most days about right for him too. His next HbA1c is in three months. He doesn't think about it daily anymore.

The number gave him structure for a year that would otherwise have been guessing. The structure outlasted the year.