How Is Yoghurt Made? Two Bacteria That Need Each Other

A product cannot legally be called yoghurt in most jurisdictions unless it is fermented with two specific bacteria, Streptococcus thermophilus and Lactobacillus delbrueckii subspecies bulgaricus, and their relationship is the reason the process works as fast as it does. The streptococcus grows first, consuming oxygen and producing formic acid and carbon dioxide, which the lactobacillus needs. The lactobacillus in turn breaks milk proteins down into peptides and amino acids, which the streptococcus cannot obtain for itself and which it needs to keep growing. Each therefore accelerates the other, a relationship called protocooperation, and a culture containing both acidifies milk considerably faster than either alone. Both are thermophilic, working best between forty and forty-five degrees, which is why the mixture must be held warm. The lactobacillus also generates acetaldehyde, which is the compound most responsible for the characteristic yoghurt aroma.

The setting is a protein effect rather than a thickening one. Milk contains casein proteins arranged in structures called micelles, which are kept apart and suspended because they carry a negative charge and repel each other, stabilised by a hairy outer layer of one particular casein. As the bacteria convert lactose into lactic acid, the pH falls from about 6.7 toward 4.6, which is the isoelectric point of casein, the pH at which its net charge is zero. With the repulsion gone, the micelles aggregate into a continuous three-dimensional network that traps the water and the fat, which is the gel. This is the same chemistry as acid cheesemaking, and the difference between yoghurt and a soft acid-set cheese is largely whether the whey is drained. It also explains the common failures: disturbing the container while the network is forming breaks it and produces a grainy result, and overshooting the fermentation produces excessive acid and a curd that contracts and expels whey.

The domestic and industrial processes are the same sequence at different scales:

• •Standardising the milk, adjusting fat and adding milk powder in commercial production, since more protein gives a firmer set
• •Heating to around 85 degrees for half an hour or 90 for a few minutes, which is hotter than pasteurisation and is done for a different reason: it denatures the whey proteins so that they participate in the gel, giving a firmer and less watery result, and it kills competing organisms
• •Cooling to about 43 degrees, the working temperature of the cultures
• •Inoculating with a starter, either a commercial culture or a spoonful of live yoghurt, and holding the temperature
• •Incubating for four to eight hours until the pH reaches about 4.6, which is when the set occurs, tested by tilting rather than stirring
• •Chilling immediately, which slows the bacteria and stops the acidity rising further, and after which the product keeps developing slowly in the refrigerator

The differences between products come from a small number of choices. Set yoghurt is incubated in its final pot so the gel forms undisturbed; stirred yoghurt is fermented in a tank and broken up when packed, which is why it pours. Greek-style yoghurt is strained through cloth to remove whey, which concentrates the protein and produces a thicker product and a substantial quantity of acid whey that the industry has struggled to dispose of; many supermarket products labelled Greek-style are thickened with milk powder or starch instead, which is cheaper and produces a different thing. Skyr from Iceland and quark from central Europe are related strained products with their own cultures. Kefir uses an entirely different starter, a grain containing dozens of bacterial and yeast species, which produces a thinner, slightly effervescent and mildly alcoholic drink. Labneh is yoghurt strained until it is a spreadable cheese. The ubiquitous fruit yoghurt adds sugar in quantities that frequently exceed those in a soft drink, which is worth reading the label for.

Two are reasonably supported and one is oversold. Lactose reduction is real: the bacteria consume some of the lactose during fermentation and, more importantly, supply the enzyme that digests it, which continues working in the gut, and controlled studies find that lactose-intolerant people tolerate yoghurt considerably better than milk. Nutritional density is straightforward, since yoghurt supplies protein, calcium, phosphorus and B vitamins in a form many people find easier to eat than milk. The probiotic claim is more complicated: the two standard yoghurt cultures largely do not survive to colonise the gut, which is why products marketed as probiotic add other species, typically bifidobacteria or specific lactobacillus strains selected for survival, and the evidence for benefit is strain-specific and condition-specific rather than general, with the strongest results for antibiotic-associated diarrhoea and much weaker support for the broad claims made in advertising. Regulators in Europe have rejected most health claims submitted for probiotic products on the grounds that the evidence did not support them.

Two thermophilic bacteria ferment milk faster together than either does alone, each supplying compounds the other needs, and the lactic acid they produce lowers the pH to about 4.6, the point at which casein proteins lose their mutual repulsion and aggregate into a gel that traps water and fat. Heating the milk above pasteurisation temperature first denatures whey proteins so the set is firmer. Greek-style yoghurt is strained, kefir uses a completely different mixed culture, and the lactose-digestion benefit is well supported while general probiotic claims are not.