Most people quit fibre before the side effects even start

Every article about fibre supplements ends the same way. Give it time. One to two weeks for the obvious effects, four weeks or more for anything metabolic. Be patient, the adjustment period passes.

It is good advice. It is also being given to people who are no longer in the room.

What the survival curve looked like

I ran a fifteen-day fibre programme with thirty participants. Below is how many were still logging each day.

Day 1    30 people   100%
Day 2    21 people    70%   ← nine gone after the first night
Day 4    16 people    53%   ← half gone inside four days
Day 9    16 people    53%
Day 11   13 people    43%   ← second wave
Day 15   10 people    33%

Ten people were still logging on day fifteen. Five had logged on thirteen days or more, which is the bar I set for completing, and that is the figure I use for completion throughout.

Now put that next to the discomfort timings from the same cohort. Averaged across those who reported discomfort, it arrived around day 3.7 for one constitutional group, day 3.9 for another, and day 7.0 for a third.

Nine people were already gone before the earliest of those averages.

They did not quit because it was uncomfortable. They quit before it was uncomfortable.

One thing the curve cannot show, and it is worth naming before going further. Everyone on it had logged at least once, which means everyone on it had started. A separate group of people signed up and never began at all. Something landed in the same week and they never took a first dose. They leave no first record, so they are not in the thirty and not in any figure below. Whatever that layer costs, it is invisible here.

The product was not the problem

This is the part that changed how I think about the programme.

When people logged a check-in, they had almost always taken the fibre. Morning compliance ran at 94 per cent of logged checks and evening at 97 per cent. Nobody was logging in to report that they had skipped it.

But the evening logs were thinner than the morning ones, around thirty per cent fewer records. And among those who dropped out, 44 per cent had missed their evening check immediately before disappearing.

So the evening was where it came apart. Not because evening is a bad time to take fibre; there is nothing in this data about that. Because evening is where the day’s attention runs out.

I should be careful about what that does and does not show. A missed evening log is a marker, not a proven cause — and it may not even be a marker of what I assumed.

Logging meant scanning a code on the product to open an app. I treated that as a non-event when I designed it. It was not. Enough participants found scanning a code and following a link into an app genuinely difficult that some of the thinness in those evening records is probably measuring how hard the app was to open rather than whether anybody took anything. It is entirely possible that people who were already drifting simply stopped logging first and stopped taking it second, in which case the log is a symptom rather than a lever. I ran a simulation on the cohort suggesting that onboarding changes plus an evening reminder could lift completion from 16 per cent to somewhere in the forties. That is a model, not a result. It has not been tested.

Where the constitutional difference showed up

The four Sasang constitutions did not fail in the same way.

Missed evening check-ins were not evenly distributed. One group, the So-yangin (spleen-dominant), missed half of theirs, well above the others. That same group was the slowest to report discomfort, averaging day 7.0 while two other groups hit theirs under day 4.

Which produces two different failure modes rather than one.

If discomfort arrives on day three, the risk is that you conclude the product does not suit you and stop. If discomfort does not arrive until day seven, the risk is different: nothing happens for a week, attention drifts, and you stop without ever forming an opinion. The first is a decision. The second is an absence of one.

In classical terms the So-yangin pattern runs hot and quick, with follow-through as the weak point rather than sensitivity. What the check-in data looks like is consistent with that. I would not put it more strongly than consistent — thirty people, seven in that group, self-reported logs.

How far the numbers carry

Thirty participants, no control group, no blinding, and self-reported logs throughout. Only five finished, so anything drawn from the later days rests on very few people. One constitutional group had a single participant and is therefore absent from every figure above. The simulation is a simulation.

What survives all that is narrow but, I think, real: the dropout happened earlier than the discomfort, and the two did not line up the way I assumed they would.

What I got wrong running it

Three things I had not accounted for, and each one sits underneath a number above.

Signing up and starting are not the same event. Several people registered and then could not begin alongside the group. Something else took that week. They never took a first dose, so they never appear in any of the numbers, and the loss they represent is one I only know about because I was the one taking the registrations. For the people who did fall away partway through, I never established why at the time.

The second is what sits between taking something and stopping it. In clinic this is the part I see constantly. There is no shortage of foods that are good for people, and saying so in a consultation room is not the difficult bit. Without something ongoing, adherence usually does not end cleanly. It degrades. People carry on taking it, roughly, when they remember. That state can run for weeks and it does not appear in a completion figure at all. A cohort split into finishers and quitters hides it entirely.

The third is the instrument. Logging meant scanning a code on the product and following a link into an app. I had assumed that was trivial. A fair number of people found it genuinely difficult. Which means that when I look at thin evening records and read fading attention into them, I may be reading the app.

None of that makes the timing finding go away. It does mean the cleanest thing I can say about this cohort is that the failure happened well before anyone’s gut had an opinion, and at least some of it happened before the product was ever opened.

Before you start the next bottle

If you have started a fibre supplement and stopped before, the useful question is not whether it worked. You probably never got far enough to find out.

Write down the day you stopped and what was happening that day. Not how your gut felt — whether you were travelling, whether the bottle was somewhere you could not see it, whether you simply forgot twice and decided that settled it. In this cohort, the point of failure was closer to that than to anything digestive.

If you are starting one now, decide in advance what day you will judge it on, and make it later than day four. Most of the people in this programme never reached the day their own body would have given them an answer.

And watch for the middle state, which is the one nobody warns you about. Stopping is obvious. Carrying on carelessly (most days, roughly, when you remember) is not, and it can go on for a long time while you believe you are still giving it a fair trial.