Probiotics and Strain Specificity: Why 'Take a Probiotic' Is Not Useful Advice

Walk into any health shop and you will find shelves of probiotics making broad claims. Better digestion. Stronger immunity. Improved mood. The labels rarely mention that the evidence behind those claims is strain-specific, not category-wide. A strain trialled in antibiotic-associated diarrhoea tells you nothing about a strain trialled in irritable bowel syndrome. One that moved a score in one trial may have shown nothing on a different measure in another. No probiotic strain has authorised wording on the GB nutrition and health claims register, so "probiotic" on a label is a category name rather than a claim anyone has substantiated.

This matters because most people buy 'a probiotic' as though they are all interchangeable. They are not. The difference between Lactobacillus rhamnosus GG and Lactobacillus rhamnosus HN001 is not a branding choice. These are functionally different organisms with different clinical profiles. Getting this wrong does not usually cause harm, but it does waste money on something unlikely to help with your specific goal.

Why the strain matters more than the species

Probiotic taxonomy works at three levels: genus (Lactobacillus), species (rhamnosus), and strain (GG). Most marketing stops at species. Most evidence operates at the strain level. This disconnect is the core problem.

Two strains within the same species can have different adhesion properties, different metabolite production, different survival rates through stomach acid, and different effects on the immune system. A 2018 systematic review in Gastroenterology found that pooling probiotic trials by genus or species produced inconsistent results, while strain-specific analyses showed clearer effect sizes. The evidence is strain-level or it is noise.

Strains with strong evidence

Lactobacillus rhamnosus GG

One of the most studied strains in existence. Strong evidence for preventing and shortening acute gastroenteritis in children. Moderate evidence for reducing antibiotic-associated diarrhea in adults. The GG strain adheres to intestinal mucosa better than most Lactobacillus strains, which partly explains its GI tract specificity. Typical studied dose: 10 to 20 billion CFU daily.

Saccharomyces boulardii

A yeast-based probiotic, not a bacterium. This gives it a practical advantage: antibiotics do not kill it. It is the best-studied probiotic for preventing antibiotic-associated diarrhea, with a 2017 Cochrane review finding a relative risk reduction of roughly 40% at doses of 250 to 500 mg (approximately 5 to 10 billion CFU) twice daily. Also studied for Clostridioides difficile recurrence, where the evidence is moderate. Because it is a yeast, it does not colonize permanently and clears within days of stopping.

Bifidobacterium infantis 35624

Studied specifically for irritable bowel syndrome. A 2006 trial published in Gastroenterology showed significant improvement in abdominal pain, bloating, and bowel movement difficulty compared to placebo. The dose used was 1 billion CFU daily, notably lower than many commercial products suggest. This is a case where more is not necessarily better.

Lactobacillus reuteri DSM 17938

Evidence for infantile colic. A 2018 meta-analysis of seven RCTs found that this specific strain reduced crying time in breastfed infants with colic by an average of about 50 minutes per day. Evidence in formula-fed infants is weaker. This strain also has moderate evidence for H. pylori eradication when used alongside standard triple therapy.

Lactobacillus helveticus R0052 and Bifidobacterium longum R0175

This two-strain combination (sometimes marketed as Cerebiome) has been studied for psychological outcomes. A 2011 RCT found significant reductions in global stress, anxiety, and depression scores over 30 days compared to placebo. The doses used were roughly 3 billion CFU total. Gut-brain axis research is still young, but this specific combination has more human data than most mood-related probiotic claims.

CFU counts: more is not always better

Colony-forming units measure the number of viable organisms per dose. Marketing pushes higher numbers because they look impressive on packaging. The evidence does not support this arms race.

Bifidobacterium infantis 35624 showed effects at 1 billion CFU. Saccharomyces boulardii works at 5 to 10 billion CFU. Lactobacillus rhamnosus GG is typically studied at 10 to 20 billion CFU. The effective dose depends on the strain, not on a universal 'more is better' principle. A 50 billion CFU multi-strain blend with no clinical data behind its specific combination is not inherently superior to a 1 billion CFU single-strain product with three positive RCTs.

What does matter: the CFU count should reflect viable organisms at the time of use, not at the time of manufacture. Some products list CFU at manufacture and lose significant viability before you open the bottle. Look for products that guarantee CFU through the expiration date.

Storage and viability

Some probiotic strains require refrigeration. Others are shelf-stable. This is not a quality indicator either way; it depends on the strain and the manufacturing process.

• Refrigerated strains include most traditional Lactobacillus and Bifidobacterium products. Heat and humidity degrade them. If your probiotic shipped unrefrigerated for three days in July, viability may already be compromised.
• Shelf-stable strains include Saccharomyces boulardii (naturally heat-tolerant) and many newer formulations using microencapsulation or lyophilization technology. These survive at room temperature because of the manufacturing process, not because they are fundamentally hardier organisms.
• Packaging matters. Blister packs protect individual doses from moisture. Bottles expose the entire supply to humidity every time you open the cap. For moisture-sensitive strains, blister packs maintain viability longer.

Multi-strain blends: the marketing problem

Many commercial probiotics contain 10 to 15 strains. The implied logic is that more strains equals broader coverage. The problem: almost none of these specific combinations have been tested together.

Strains can compete with each other. They can produce bacteriocins that inhibit nearby organisms. A 2019 study in Cell found that some multi-strain formulations colonized the gut less effectively than single-strain products, because inter-strain competition reduced overall survival. The blend on the label does not tell you what survives to reach your colon.

This does not mean multi-strain products never work. Some specific combinations have been studied (like the two-strain Cerebiome combination above, or VSL#3's eight-strain blend for ulcerative colitis). But a product combining 12 strains based on a proprietary logic with no published trials behind that specific blend is a marketing exercise, not evidence-based practice.

Matching strains to goals

The practical question: how do you pick a probiotic for a specific outcome? Start with the strain, not the brand.

• Antibiotic-associated diarrhea: Saccharomyces boulardii (250-500 mg twice daily) or Lactobacillus rhamnosus GG (10-20 billion CFU daily). Start on day one of antibiotics, continue for one week after finishing.
• Studied in IBS: Bifidobacterium infantis 35624 was the strain used in the irritable bowel syndrome trials, at 1 billion CFU daily over four weeks. IBS is diagnosed and managed by a clinician, so that is a trial fact rather than something to take for it.
• General gut maintenance: Lactobacillus rhamnosus GG (10 billion CFU daily) has the broadest GI evidence base.
• Stress and mood support: L. helveticus R0052 + B. longum R0175 (3 billion CFU daily). Evidence is promising but still early.
• Infant colic (breastfed): L. reuteri DSM 17938. Consult your paediatrician first.

How Stack Almanac handles this

Stack Almanac keeps the strain designation in your record where the label gives one, so a probiotic in your routine is that specific product rather than a category. The strain-by-strain evidence this page describes is on the ingredient page with its sources, the antibiotic timing question included. The app does not rule on whether your product matches your goal, and antibiotics are a prescription. Check with your pharmacist.

The record holds strain-level detail, not just 'probiotic.' Scan a product or enter it by hand and you can keep the specific strains and the CFU count. Over time that gives you a dated picture of what you have actually been taking, which is the thing you need before you can read it against the trials on this page.