Sourdough instead of commercial yeast

Sourdough is worse than commercial yeast in every respect that a business cares about. It is slower, it is less predictable, it cannot be dosed, and it means the schedule is decided by a jar rather than by me.

I use it anyway, and I would rather explain that with the actual evidence than with the usual claims, because the usual claims are a mixture of things that are true, things that are true in a narrower sense than stated, and things that are simply wrong.

What is in the two of them

Commercial yeast is one organism, Saccharomyces cerevisiae, grown and dried to a specification. Add a measured amount, get a predictable amount of gas over a predictable time.

A sourdough starter is a partnership. Wild yeasts do most of the leavening. Lactic acid bacteria do the acidification, and the classic one is Fructilactobacillus sanfranciscensis. That bacterium is heterofermentative, meaning it does not only make lactic acid: it also produces acetic acid, ethanol and carbon dioxide. The acetic side is the reason sourdough has a sharpness that a yeasted dough simply does not have available to it.

They also break down proteins, produce polysaccharides that change the crumb, and generate a long list of aromatic compounds. Flavour in a yeasted bread comes mostly from the crust. In a sourdough it comes from the crumb as well.

What long fermentation genuinely does

It degrades phytate. Wheat holds phosphorus as phytate, which binds minerals. Acidifying the dough brings the pH down towards the optimum for wheat's own phytase enzyme, and the phytate is broken down by the flour's own enzyme, prompted by the acid, rather than by the bacteria directly. This is well established.

It does not follow that sourdough improves anyone's mineral status. A systematic review of human trials published this year found that single-meal studies do show better iron absorption, and that longer interventions do not reliably improve iron status at all. In one twelve-week trial the low-phytate group's ferritin went down. The reviewers were blunt about the quality of the evidence. So: better absorption at a meal, demonstrated. Better nourished, not demonstrated. Those are different sentences and bakeries routinely print the second while citing the first.

It reduces FODMAPs, conditionally. Fructans are the main FODMAP in wheat, and long fermentation reduces them substantially. Reported reductions run up to about ninety per cent. Monash University, who define the FODMAP diet, are specific about the condition: the dough has to prove for a prolonged period, typically more than twelve hours. Fast-proofed supermarket "sourdough" does not qualify, and neither does bread acidified with vinegar. Worth knowing too: a long yeasted ferment does much the same thing. This is a property of time, not of starters.

It keeps longer. Sourdough stales more slowly, and it moulds measurably later. That second part is not only about acidity: researchers have identified specific antifungal acids produced during fermentation, and shown that adding them to a chemically acidified dough extends shelf life beyond what the acidity alone achieves.

It weakens the dough. The same bacterial proteolysis that softens the crumb degrades gluten, and a ferment that runs too long produces a slack, unmanageable dough. This is a trade-off rather than a free gift, and it is the failure mode people hit when they decide that if long is good, longer must be better.

What it does not do

It is not safe for coeliacs. Not "less bad". Not safe.

Even aggressive fermentation removes something like seventy to eighty per cent of the gluten, which leaves thousands of parts per million against a gluten-free threshold of twenty. That is not close. Feeling better after eating it is not evidence of anything, because coeliac damage is not something you feel in the moment.

FODMAP research and the coeliac question get conflated constantly, including by people selling bread. Sourdough may genuinely help some people with IBS. It does nothing whatsoever for coeliac disease. Everything I bake contains gluten and I am not the person to ask about it.

Why I still do it

Because the bread tastes of something, and because the schedule it imposes is one I can actually work with. Long, cold and overnight suits one person with a domestic oven better than a system built for a production line.

Scheduling is genuinely harder, though. With yeast you set a dose and get a clock. With a starter you have three interacting variables, being temperature, how recently it was fed, and how much you use, plus a culture that behaves differently in February from how it behaved in July. The professional answer is to control the finished dough temperature to a target, usually somewhere in the mid-twenties, and in a home kitchen that mostly means paying attention.

There is no dose. There is a dough, and you read it.