Two years ago I started reducing the sugar in the sweets I make. Not dramatically, and not all at once — by maybe five percent at a time, over about a year.

The end point is roughly a third less than where I started. Everything tastes right to me. And a sweet made to my original recipe is now genuinely difficult to eat, which is the part I did not expect.

The adaptation is real and it goes both ways

Perception of sweetness adjusts to habitual exposure. This is well described and I had assumed it was a small effect.

It is not small. After about six months of gradual reduction, sweets that I had previously considered normal read as cloying. Not slightly — actively unpleasant, in the way that over-salted food is unpleasant.

And it works in the other direction. A period of eating a lot of very sweet things resets the baseline upward within a couple of weeks, and things that had tasted fine start tasting thin.

Which means there is no objective correct sweetness. There is only what your recent eating has calibrated you to.

Why gradual worked and abrupt did not

I had tried this before by cutting a third at once, and it failed within a fortnight.

An abrupt reduction produces something that tastes wrong immediately, and wrong in a way that reads as a deficiency rather than as a different balance. You notice the absence.

Five percent is below the threshold at which most people notice anything. Each step feels normal, and by the time you have taken ten of them you are somewhere you could not have reached directly.

This is presumably how food manufacturers move in the other direction, and it works equally well running backwards.

What sugar was doing besides sweetness

The complication, and the reason it is not simply a matter of using less.

In baked goods sugar affects texture, browning and moisture retention. Reducing it by a third changes the crumb, the colour and the keeping quality, not just the taste.

In milk-based sweets it affects viscosity and how the mixture sets.

In anything preserved, it is doing preservation work, and reducing it shortens the shelf life materially. Traditional sweets that keep for weeks keep for weeks because of the sugar.

So the reduction has to be compensated for in some cases. More fat, or a shorter shelf life, or accepting a different texture.

Where I could reduce most

Anything where sugar was doing only sweetness turned out to have the most room.

Milk-based sweets took the reduction easily, and the result is that they taste more of milk, which is an improvement.

Syrups took it very well. Most syrup recipes are considerably sweeter than they need to be, and reducing lets the aromatics in the syrup come forward.

Fruit preparations took it best of all, because the fruit brings its own sweetness and the added sugar was largely masking it.

Where I could not

Anything relying on sugar structurally. Certain confections depend on crystallisation behaviour that changes entirely at different concentrations, and they simply do not work with less.

Anything intended to keep, unless you are content to eat it within a few days.

And, honestly, one or two traditional sweets that are meant to be intensely sweet and are not improved by moderation. Some things are what they are and a restrained version is just a worse version.

The social problem

The part nobody warns you about.

Food made at your new calibration tastes under-sweet to everybody else, and people say so.

I have had things politely described as not quite finished, which is a specific and accurate description of what a reduced-sugar sweet tastes like to somebody calibrated higher.

What I do now is make things at something like a fifteen percent reduction when other people are eating them, and at my own level when they are not. That is a compromise rather than a solution and it avoids the conversation.

Whether it was worth it

For flavour, unambiguously. The things I make taste more of their actual ingredients, which is the entire argument.

For anything else, I would not claim anything. I did not do this for health reasons and I am not in a position to say what a third less sugar in occasional sweets does or does not do to anybody.

What I can say is that the adaptation is real, it is faster than expected, and it is reversible in both directions, which means anybody can run the experiment themselves at no cost.

Where the sugar actually comes from

One thing the year clarified. Almost none of my sugar intake was from the sweets I make.

It was from the ordinary things — bought bread, sauces, drinks, biscuits at work, anything packaged and savoury.

Which means that a year of carefully reducing sugar in home baking, while changing nothing else, would have been mostly symbolic.

I mention it because the flavour argument for reducing sugar in what you make is strong and entirely separate from any argument about intake, and conflating them leads people to feel virtuous about the wrong thing.