Benefits
Stomach comfort and upper abdominal pain ratings
In 174 healthy adults with mild fullness after meals or early satiety, 85 g of yogurt with at least 1 billion CFU a day for 12 weeks lowered upper stomach pain ratings more than the same yogurt without the strain at 6 and 12 weeks. Pain was one of ten symptoms scored at two time points with no correction for multiple testing; fullness, early satiety and the overall score did not differ. The maker funded it and employs all authors.
After-meal fullness and bloating ratings
In 106 adults with recurring upper stomach symptoms and no stomach infection, the yogurt daily for 12 weeks did not significantly improve the overall symptom rating, the main outcome. More people lost their four main symptoms (35% vs 17%), and fullness after meals and bloating eased at some visits; pain did not. The maker paid for the trial.
Gastric emptying time on a breath test
In 28 healthy adults with mildly slow stomach emptying, 12 weeks of the yogurt did not change emptying time, the main outcome, or the overall stomach rating compared with yogurt without the strain, and the authors said the data did not support a benefit. Salivary amylase, a stress-linked marker, fell more. The maker paid for the trial.
Breath test markers of a common stomach bacterium
In 124 analyzed adults carrying H. pylori, one cup a day for 12 weeks did not lower urea breath test or stool antigen values, the main outcome, versus yogurt without the strain. An earlier small study with no randomized comparison reported lower breath test values. These are infection markers, and the yogurt has not been shown to clear the infection.
Sleep length and morning sleepiness during exam stress
In 78 medical students facing two exams, 85 g of yogurt with the strain daily for 8 weeks improved sleepiness on rising at 2 and 4 weeks and lengthened wrist-tracker sleep at 2, 4 and 7 weeks versus yogurt without it. Anxiety scores did not differ, the summary gives no result for the overall sleep quality index, and most authors work for the maker.
Mechanism of action
Reaching the stomach lining
Biopsies from two volunteers who drank the yogurt found the strain's DNA in the stomach's mucus layer, and lab work reported in a maker-coauthored review found it survives acid near stomach pH. This suggests it can stay in the stomach for a time, but it does not show a health effect.
Competing with a stomach bacterium in lab and mouse work
In cell tests the strain slowed the growth of H. pylori, including antibiotic-resistant strains, and lowered the inflammatory signal IL-8; in infected mice it reduced colonization. A controlled human trial at one cup a day did not reduce breath test markers, so these findings have not carried over to people.
Stomach fluid bacteria in single-group studies
In people with recurring indigestion, stomach fluid held a different bacterial mix than in healthy controls, with more Prevotella, and after 12 weeks of the yogurt the mix moved toward that of controls. There was no placebo group, so the change cannot be credited to the strain.
Ghrelin signals in stomach cells
In a human stomach cell line, live or heat-treated cells of the strain and their culture fluid raised gene expression and release of ghrelin, a stomach hormone tied to appetite and stomach movement. This is exploratory cell work by maker employees and has not been measured in people.
Clinical trials
Randomized, double-blind, placebo-controlled parallel trial (UMIN000042422) of 85 g of yogurt a day with at least 1 billion CFU of OLL2716 versus the same yogurt without the strain; registered primary outcome the severity of stomach and chest symptoms at 12 weeks, reported as changes in ten symptom scores and three combined scores at 6 and 12 weeks with no correction for multiple testing; funded by Meiji Co., Ltd., all seven authors employees of Meiji Co., Ltd. or Meiji Holdings (Yamada et al. 2024, Nutrients)
174 healthy Japanese adults aged 20 to 64 with mild to moderate fullness after meals or early satiety, excluding those meeting functional dyspepsia criteria; all analyzed.
Epigastric pain improved more than on placebo at 6 weeks (p = 0.020) and 12 weeks (p = 0.028), and the epigastric pain-like score at 6 weeks (p = 0.028); none of these would pass a Bonferroni cutoff for ten symptoms (0.005). The combined dyspepsia-like score (p = 0.207 at 6 weeks, 0.377 at 12 weeks), fullness and early satiety (the enrollment symptoms used to size the trial) and the other symptoms did not differ. Baseline pain was higher on OLL2716 (1.6 vs 1.3). Some stress questionnaire items at 6 weeks and the GSRS overall responder count and hunger rating at 12 weeks favored OLL2716, while GSRS score changes were similar. No adverse events.
Randomized, double-blind, placebo-controlled parallel trial (UMIN000025356) of 85 g of yogurt a day with at least 1 billion CFU of OLL2716 versus the same yogurt without the strain; primary endpoint the global impression of gastric symptoms; Meiji Co., Ltd. covered the costs, three authors (Ohtsu, Kawashima, Uchida) Meiji employees and others receiving Meiji grants or honoraria (Ohtsu et al. 2017, Digestion)
116 Japanese adults with functional dyspepsia by Rome III criteria and without H. pylori infection; 106 analyzed per protocol.
The primary global impression did not reach significance (p = 0.073). Elimination of the four main symptoms was 35.3% vs 17.3% (p = 0.048), and fullness after meals (8 and 12 weeks) and upper abdominal bloating (8 and 12 weeks) improved, while epigastric pain did not differ at any visit. Efficacy was analyzed per protocol only, with no correction for multiple testing. Adverse events were 2 vs 5, none judged related.
Randomized, double-blind, placebo-controlled parallel trial (UMIN000035057) of 85 g of yogurt a day with at least 1 billion CFU of OLL2716 versus the same yogurt without the strain; primary endpoint change in gastric emptying time (Tmax area under the curve) on a 13C breath test; costs paid by Meiji Co., Ltd., Ohtsu a Meiji employee and Takagi a Meiji grant recipient (Ohtsu et al. 2021, Nutrients)
28 healthy Japanese adults (4 men, 24 women) with stomach upset and mild to moderate delayed gastric emptying; 27 analyzed per protocol.
The primary endpoint and the overall impression of gastric condition did not differ between groups; the odds ratio for improved emptying was 4.1 (95% CI 0.8 to 20.2), and the authors state the data were not enough to support a benefit. Salivary amylase fell more on OLL2716. Adverse events were 19 vs 13, all non-serious and none judged related.
Multicenter, randomized, double-blind, placebo-controlled trial (UMIN000016746) of 90 g of yogurt a day with at least 1 billion CFU of OLL2716 versus the same yogurt without the strain; primary endpoint a fall in H. pylori load by urea breath test and stool antigen; fully funded by Meiji Co., Ltd., one author (Ohtsu) a Meiji employee (Takagi et al. 2016, Gastroenterol Res Pract)
131 Japanese adults aged 30 or older with H. pylori infection found at health checks; 124 analyzed after 7 were excluded for prohibited medicines.
Breath test and stool antigen values, the primary endpoint, did not differ between or within groups, and the pepsinogen I/II ratio was unchanged; these are infection and stomach lining markers. Fullness after meals fell within the OLL2716 group only, and fewer people on OLL2716 had a bloating score above 10 (p < 0.05). No adverse events.
Short report of yogurt with OLL2716 (LG21) for 8 weeks; the abstract describes no randomized comparison, a later paper by the group calls it nonrandomized and twice daily, and a Meiji-coauthored review gives 1 billion CFU a day; the PubMed record lists no funding (Sakamoto et al. 2001, J Antimicrob Chemother)
31 adults infected with H. pylori.
Breath test values and serum pepsinogens improved after the yogurt, with no numbers or between-group results in the abstract. These are infection and stomach lining markers, not health outcomes, and the later review states no subject cleared the infection 3 months after stopping.
Randomized controlled trial of 1-week triple drug therapy alone versus the same therapy plus 112 g of yogurt with at least 1 billion CFU of OLL2716 twice daily for 4 weeks (3 weeks before and 1 week during the drugs); no placebo yogurt in the comparison group; outcome eradication 8 weeks later; yogurt supplied by Meiji Dairies, no funding statement found (Deguchi et al. 2012, J Gastroenterol Hepatol)
229 Japanese patients with H. pylori infection, most with peptic ulcer disease.
Eradication was 82.6% vs 69.3% by intention to treat (P = 0.018) and 85.6% vs 74.5% per protocol (P = 0.041). For clarithromycin-resistant strains it was 38.5% vs 28.0%, not significant (the abstract gives P = 0.458 and the text P = 0.428). Side effects did not differ. This is drug co-treatment, so the result cannot be credited to the yogurt alone.
Randomized, double-blind, placebo-controlled trial of 85 g of yogurt a day with OLL2716 versus placebo yogurt (CFU not given in the abstract); primary endpoint subjective sleep on the OSA-MA inventory and the Pittsburgh Sleep Quality Index; five of seven authors Meiji Co., Ltd. employees; Europe PMC lists UMIN000052184 (Yoshizawa et al. 2026, J Dairy Sci)
78 healthy Japanese fourth-year medical students sitting two consecutive exams.
Sleepiness on rising improved more than on placebo at 2 and 4 weeks and dreaming at 8 weeks; wrist-tracker sleep length was longer at 2, 4 and 7 weeks; salivary chromogranin A (4 weeks) and cortisol (5 weeks) were lower. Anxiety scores were similar, and the abstract reports no between-group result for the sleep quality index.
Randomized, double-blind, placebo-controlled trial of 112 mL of yogurt containing OLL2716 or placebo twice daily, with capsule endoscopy before and after; the abstract gives no funding statement, and co-authors Ohtsu and Asami are listed at Meiji in the group's 2019 review (Suzuki et al. 2017, Digestion)
64 patients who had taken aspirin for more than a month.
After 6 weeks the OLL2716 group had significantly fewer small bowel mucosal breaks and reddened lesions (p < 0.01), unlike the placebo group. Reflux and digestive symptom questionnaire scores improved within the OLL2716 group but not the placebo group; the abstract reports these as within-group changes.