One piece a day since April · structured 90-day protocol since July 27

The open chew log.

I'm Max. I make this gum, and this is the record of me actually using it: one piece a day since April 2026. No photos, no filters, just dated notes, including the weeks where nothing happened. That's the deal with a self-test. It's slow, and it's honest.

The first three months, in summary

I didn't keep dated notes from the start. That discipline begins today, below. What I can say about the first months, in order of how much I trust the memory:

  • The immediate thing, from the first days: the clean feeling after coffee. That part isn't remineralization, that's chewing and saliva doing their job. It's also why the habit stuck.
  • Somewhere along the way, cold water and hot coffee stopped making my teeth flinch. I can't point at the exact week it changed. That's exactly why this log now exists.
  • My teeth look brighter than they did in spring. I had no shade guide and no photo protocol back then, so take that one with a grain of salt.
  • The honest bits: I skipped days. There were whole weeks where nothing changed and the only reason to keep chewing was that the box was open. If you're expecting a movie moment, this isn't it.

The 90-day protocol

From July 27 this stops being a diary I remember to write and becomes a protocol I can't wiggle out of. The rules:

  1. One entry every Sunday, dated, newest first, written the same day. No catch-up entries pretending they were on time.
  2. The same three checks each week: cold-water flinch (yes or no, one sip, straight from the fridge), look (same mirror, same daylight), anything odd (staining, taste, jaw, whatever shows up).
  3. Entries get added, never edited. Typos stay. Bad weeks stay. If I skip a Sunday, the next entry says so.
  4. Day 90 gets a verdict either way. Around October 25 I write what actually happened across the whole run, in the same font as everything else.
Day 1 of 90
What would make this a failure, stated up front: if by day 90 the cold-water flinch is back, or nothing distinguishes these teeth from where they were in spring, this page says exactly that and stays online. A self-test you can't fail isn't a test, it's an ad. One more honesty line: I make this gum, chewing is only part of the picture (I also brush like everyone should), and one person is not a study. The ingredient research lives on the transparency page; this is just me, on the record.

The log

Newest first.

Minvelle 90-day self-test, newest first: one piece of Minvelle nano-hydroxyapatite gum a day (5.7 mg nano-hydroxyapatite per 1.9 g piece), logged by Max, founder of Minvelle. The structured 90 started 27 July 2026; the verdict entry is due around 25 October 2026. One person, not a study.

2026-08-23 · week four of the structured 90

One piece of Minvelle nano-hydroxyapatite gum a day, 5.7 mg per 1.9 g piece. Self-test by Max, founder of Minvelle. One person, not a study.

On time this Sunday, after the 2026-08-17 entry ran a day late. Adherence: six of seven days again, one day without a piece, the same dent as the week before.

Cold water: no flinch. Not quieter, none. The first zero since the 2026-07-27 baseline, at the end of a line that read mild twinge, less, quieter, quieter again. A first zero is a point, not a destination: it has to keep being zero to mean anything.

The look: slightly whiter again, the second entry in a row saying that. Two points in the same direction are still two points; three make a trend, and this check is not there yet.

Anything odd: for the first time this check has content: the teeth themselves feel different, smoother, as if some of the film on them had thinned out. That is a feeling reported by the person who sells the gum, so it is logged as exactly that and nothing more.

2026-08-17 · week three of the structured 90

One piece of Minvelle nano-hydroxyapatite gum a day, 5.7 mg per 1.9 g piece. Self-test by Max, founder of Minvelle. One person, not a study.

A day late: this entry was due Sunday 2026-08-16 and is logged Monday. The protocol says a skipped Sunday gets named, so: named. The checks cover the week to 2026-08-16. Adherence took its first dent too: six of seven days, one day without a piece, the first missed day since the start on 2026-07-27.

Cold water: quieter again, quieter than 2026-08-09, which was already quieter than 2026-08-02. This log's own rule says three points make a trend, and this is the third point in the same direction. So, said carefully: the flinch line is the first thing here that qualifies as a trend. Still one person, still not a study.

The look: slightly whiter than 2026-08-09, when it was unchanged. Logged at exactly that size. The full line so far reads barely brighter, unchanged, slightly whiter: not a straight line, so no trend claim on this one.

Anything odd: nothing. Third quiet week on this check, logged anyway.

2026-08-09 · week two of the structured 90

Fourteen days, fourteen pieces, still no skipped days.

Cold water: quieter than last Sunday, which was already quieter than day one. That is two weeks moving the same direction, which is the most interesting line in the log so far, and it still is not a trend. Three points make a trend. This is two.

The look: same as last week. No change gets logged at the same size as change, otherwise this page is an ad with dates on it.

Anything odd: nothing. Second boring week in a row. The protocol does not care.

2026-08-02 · week one of the structured 90

Seven days, seven pieces, no skipped days. First regular Sunday entry, same three checks.

Cold water: less flinch than last Sunday. Straight from the fridge there is still something there, but quieter than the day-one twinge. One week is too fast to call that a trend, so this stays a data point, not a result.

The look: brighter, and I mean barely. Same mirror, same daylight, and I can see it, but partly because I am looking for it. It is there, so it gets logged, at exactly that size.

Anything odd: nothing. A boring week, logged anyway, because that is the protocol.

2026-07-27 · day one of the structured 90

Vacation week, so the log finally gets the upgrade it needed. From today: one entry every Sunday, the same three checks every time. The rules are in the protocol card above. Here is the day-one baseline, so the next 89 days have something to be measured against.

Cold water: a really, really mild twinge. Not the March feeling, but not nothing either. Writing that down precisely because it would be tempting to round it to zero.

The look: brighter than my memory of spring. Memory is a bad camera, which is exactly why this line exists: to be checked against later, not trusted now.

Anything odd: two things, logged even though neither is an enamel result. My jawline looks more defined than a few months ago, that is what daily chewing does to the muscles and has nothing to do with the mineral. And breath is fresher, right after chewing and in general, which is saliva flow and xylitol doing their normal, well-documented job.

If the next 90 days show nothing beyond that, it lands here in the same font as everything else.

2026-07-12 · roughly three months in

First dated entry. Starting the discipline I should have had in April: one note a week, same three checks every time (sensitivity, look, anything odd). Current state: cold water is boring again. In March it wasn't.

Entries get added, never edited. If something goes wrong, it goes in the log too.

The product side of the receipts (dose, certificates, batch lookup) lives on the transparency page. The data side lives in the gum database, the 36 brand comparisons and the acid-day checker. This page is the human side.