Human Milk Oligosaccharides (HMOs)
Human milk oligosaccharides (HMOs) are carbohydrates that make up the third largest solid component of human breast milk, after lactose and lipids. HMOs in supplements and infant formula, such as 2'-fucosyllactose (2'-FL), lacto-N-neotetraose (LNnT) and sialyllactoses, are human-identical copies made by chemical or enzyme-based synthesis or by fermentation with genetically modified bacteria, not taken from breast milk. People cannot digest them, but some gut bacteria can, and in placebo-controlled trials in adults and children bifidobacteria rose at some doses and not others. In adults with IBS, a 2'-FL and LNnT blend did not change symptoms compared with placebo. Infant formula trials showed normal growth, and less parent-reported bronchitis was a secondary finding. Two sialyllactose trials missed their main outcomes. Most trials were funded by makers, and high doses can cause gas and loose stools.
Benefits
Bifidobacteria in the gut
In placebo-controlled trials, HMOs raised gut Bifidobacterium more than placebo at some doses: in healthy adults over 2 weeks with 10 g but not 5 or 20 g of 2'-FL, with 5, 10 and 20 g of LNnT and with 10 or 20 g of a 2'-FL and LNnT blend; in adults with IBS with 10 g of a blend at week 4 but not week 8; and in children with 2'-FL. This is a bacteria marker, not a health outcome, and most trials were maker funded.
Bowel comfort and bowel habits
In adults with IBS, 5 or 10 g a day of a 2'-FL and LNnT blend for 4 weeks did not change overall or individual gut symptoms compared with placebo. A larger 12-week trial of 5 g a day reported fewer abnormal stools and lower symptom scores than at baseline, but it had no placebo group and was sponsored by the maker.
Growth and illness reports in formula-fed infants
In healthy infants, formula with 2'-FL and LNnT gave weight gain no lower than standard formula, the main outcome. Parents reported less bronchitis and less antibiotic use in the HMO group as secondary outcomes, which the authors said need confirmation. An independent review of 12 trials in infants and young children found illness results varied and the trials were too small for serious illness.
Cognition studied only in breastfed infants
In an observational study of 50 mother-infant pairs, more 2'-FL in breast milk at 1 month was linked to higher cognitive scores at 24 months, with no link for 2'-FL at 6 months. A review found six such studies, all observational. No cited trial tested whether an HMO supplement affects thinking or memory at any age.
Knee comfort with 3'-sialyllactose
In 60 adults with knee osteoarthritis, 200 or 600 mg a day of 3'-SL for 12 weeks lowered pain and function scores from baseline, but neither score differed from placebo at any time point, the trial's main comparison. The trial was single-blind, the maker made the tablets, and three authors worked for it.
Training adaptations with 6'-sialyllactose
In men doing 12 weeks of resistance training, 900 mg a day of 6'-SL did not change strength, body composition or blood lactate compared with placebo, the main outcomes. Some aerobic measures differed between groups in both directions, and the authors call the results exploratory. Only 19 men completed it, and the maker funded it.
Mechanism of action
Food for bifidobacteria
People lack the enzymes to digest HMOs, so most reach the large intestine, where some bifidobacteria can use them. In the 2-week adult trial, a bacterium closely related to Bifidobacterium adolescentis was the main responder and stool short-chain fatty acids did not change; B. adolescentis also drove the rise in children.
Decoy receptor idea
Some HMOs resemble sugars on the surface of gut cells, and in laboratory and tissue studies they can act as decoys that stop certain microbes from attaching. Whether this happens in people taking an HMO supplement has not been shown by the trials cited here.
Immune development
Reviews describe HMOs as possible shapers of the developing immune system, mainly from laboratory work. In infant formula trials, inflammation and gut function markers shifted toward those of breastfed infants. These are markers, and no cited trial in older children or adults measured an immune outcome.
Clinical trials
Randomized, double-blind, placebo-controlled, parallel dose-finding trial over 2 weeks: ten groups of ten took 2'-FL alone, LNnT alone or a 2:1 blend of the two at 5, 10 or 20 g a day, or 2 g glucose placebo (NCT01927900). The 2'-FL was synthetic. Funded by Glycom A/S, which employed three authors and paid three others as consultants (Elison et al. 2016, Br J Nutr)
100 healthy adults aged 18 to 60 in Denmark; all completed.
Bifidobacterium rose more than on placebo with 10 g of 2'-FL alone but not with 5 or 20 g of 2'-FL; it also rose with LNnT and with 10 or 20 g of the blend. All 56 adverse events were mild. At 20 g of 2'-FL alone, nausea, rumbling, bloating, gas, diarrhea, loose stools and urgency scored higher than on placebo.
Randomized, double-blind, placebo-controlled parallel trial of a two-HMO blend (4:1 2'-FL and LNnT) at 5 or 10 g a day versus glucose for 4 weeks plus 4 weeks of follow-up; primary outcome fecal bifidobacteria (NCT02875847). Glycom A/S is listed as the grant source and employed four authors (Iribarren et al. 2020, Neurogastroenterol Motil)
60 adults with IBS diagnosed by Rome IV criteria in Sweden; 58 completed.
Bifidobacteria rose at week 4 with 10 g compared with the other groups, but not at week 8. Overall and individual gastrointestinal symptoms did not differ between groups at week 4 or 8, and symptoms did not worsen. 2'-FL was not tested on its own.
Open-label, single-arm multicenter trial at 17 US sites with no placebo group; aim was to normalize bowel habits (NCT03550742). Sponsored by Glycom, Inc., which helped design the trial and write the paper; three authors were Glycom employees (Palsson et al. 2020, Clin Transl Gastroenterol)
317 adults with IBS (70.7% women, ages 18 to 93) took a 4:1 blend of 2'-FL and LNnT; 245 completed per protocol.
The share of bowel movements with abnormal stool consistency fell from 90.7% to 57.2% and symptom severity scores fell from baseline over 12 weeks. With no placebo group, these changes cannot be separated from placebo effects, and the authors called for a placebo-controlled trial. Common side effects were mild gas, abdominal pain and distension.
Randomized, double-blind, placebo-controlled trial of 2'-FL alone, a 2'-FL and LNnT blend, or glucose placebo once a day; primary outcome fecal microbiota (NCT02786160, lead sponsor Glycom A/S). Three authors were Glycom employees; the daily dose is not given in the abstract (Fonvig et al. 2021, J Pediatr Gastroenterol Nutr)
75 children aged 6 to 12 with overweight, including obesity.
Compared with placebo, bifidobacteria rose on 2'-FL at weeks 4 and 8, while the blend showed only a trend versus placebo (p = 0.071). Symptom rating and stool form scores showed no tolerance problems, and blood tests raised no safety concerns.
Randomized, multicenter trial of formula with 1.0 g/L 2'-FL and 0.5 g/L LNnT, a two-HMO blend, versus the same formula without HMOs to 6 months; primary endpoint weight gain through 4 months (NCT01715246). Sponsored by Nestec Ltd., with several Nestle-employed authors (Puccio et al. 2017, J Pediatr Gastroenterol Nutr)
175 healthy infants aged 0 to 14 days (88 test, 87 control).
Weight gain met noninferiority. In secondary outcomes, parents reported less bronchitis and lower respiratory infection and less antibiotic and antipyretic use in the HMO group, which the authors said needs confirmation. Stools were softer at 2 months.
Observational cohort study, not a clinical trial; no supplement was given, and 19 HMOs naturally present in breast milk were measured. NIH-supported; the authors declared no competing interests (Berger et al. 2020, PLoS One)
50 Hispanic mother-infant pairs, with breast milk sampled at 1 and 6 months and infant cognition assessed at 24 months
2'-FL in breast milk at 1 month predicted higher cognitive scores at 24 months, but 2'-FL measured at 6 months was not associated with cognition. Because this is an observational study in breastfed infants, it cannot show that an HMO supplement improves cognition at any age.
Randomized, single-blind, three-arm pilot trial versus placebo tablets; primary outcome VAS pain. Supported by Korean government grants (MOTIE and KIAT); the 3'-SL tablets were made by GeneChem Inc., which employed three authors, one of whom held its stock (Park et al. 2024, Nutrients)
60 Korean adults over 40 with knee osteoarthritis (Kellgren and Lawrence grades 1 to 4), 20 per group.
Pain and KWOMAC scores fell from baseline at 12 weeks in the 3'-SL groups and not on placebo, but neither score differed between groups at any time point. No serious adverse events were reported.
Randomized, double-blind, placebo-controlled proof-of-concept trial; primary outcomes body composition, strength and blood lactate; registered in January 2026, a year after ethics approval (ISRCTN50097942). Funded by GeneChem through a service contract with Texas A&M University; the authors declared no conflicts (Estes et al. 2026, Nutrients)
28 healthy recreationally active men aged 18 to 40 were randomized; 19 completed.
Strength, body composition and blood lactate did not differ from placebo. Ventilatory threshold and oxygen uptake at that threshold were lower on 6'-SL at 12 weeks, while some peak aerobic measures rose more at 6 weeks; the authors call the results exploratory. Side effects did not differ between groups.
Side effects and drug interactions
Common Potential side effects
Important Drug interactions
Frequently asked questions about Human Milk Oligosaccharides (HMOs)
What is Human Milk Oligosaccharides used for?
HMOs are added to infant formula and sold as prebiotics for adults. In placebo-controlled trials they raised gut bifidobacteria at some doses, but a trial in adults with IBS found no symptom difference from placebo, and infant formula trials mainly showed normal growth. NutraSmarts does not list HMOs under a health category.
What is the recommended dosage of Human Milk Oligosaccharides?
There is no established dose. Adult trials used 5 to 20 g a day of 2'-FL, LNnT or a blend for 2 to 12 weeks, and sialyllactose trials used 200 mg to 6 g a day for 12 weeks. Digestive side effects rose at 20 g a day of 2'-FL alone. Follow the product label.
Is Human Milk Oligosaccharides safe, and does it have side effects?
In trials lasting up to 12 weeks, the most common effects were gas, bloating and loose stools, mostly at high doses, and reported side effects were mostly mild. EFSA has assessed specific HMOs, including 2'-FL, LNnT and 6'-SL, for use in foods. Long-term adult data are lacking, so ask a healthcare provider first, especially for children.
Is an HMO blend the same as 2'-FL alone?
No. Many products and trials combine 2'-FL with LNnT or other milk oligosaccharides, and results differ by product and dose: in healthy adults, bifidobacteria rose with 10 or 20 g of the blend but only with 10 g of 2'-FL alone. See the 2'-fucosyllactose page for trials of 2'-FL on its own.
How do HMO supplements differ from the HMOs in breast milk?
Mature breast milk has 10 to 15 g of HMOs per liter, with more than 200 different structures, and the mix varies between women. Supplements and formulas add only a few human-identical HMOs. Links between breast milk HMOs and infant development come from observational studies and do not show what a supplement does.
What is Human Milk Oligosaccharides?
Human milk oligosaccharides (HMOs) are carbohydrates that make up the third largest solid component of human breast milk, after lactose and lipids. HMOs in supplements and infant formula, such as 2'-fucosyllactose (2'-FL), lacto-N-neotetraose (LNnT) and sialyllactoses, are human-identical copies made by chemical or enz…
Does Human Milk Oligosaccharides interact with any medications?
Possible interactions include: No drug interaction studies of HMOs have been published. Antibiotics: the effect of HMO blends on gut bacteria during or after a course of antibiotics has not been tested. Take antibiotics as prescribed. If you take prescription medication, check with a pharmacist or doctor before using it.
How strong is the scientific evidence for Human Milk Oligosaccharides?
NutraSmarts rates the evidence for Human Milk Oligosaccharides as Limited (2 out of 5). It is backed by 8 study summaries and 13 cited references summarized on this page. A higher rating reflects more, larger, and better-designed human studies.
References(13 citations)
Evidence ratings on NutraSmarts are based on the totality of human clinical research, with emphasis on randomized controlled trials, meta-analyses, and systematic reviews. The references below directly support claims made throughout this page.
- Oral supplementation of healthy adults with 2'-O-fucosyllactose and lacto-N-neotetraose is well tolerated and shifts the intestinal microbiota. Br J Nutr. 2016;116(8):1356-1368. doi: 10.1017/S0007114516003354.PubMedUsed to support: Funded by Glycom A/S, which supplied synthetic 2'-FL (99.9% pure) and LNnT; three authors were Glycom employees and three were paid Glycom consultants (NCT01927900). Double-blind, placebo-controlled, dose-finding trial in 100 healthy adults, ten groups of ten, taking 2'-FL alone, LNnT alone or a 2:1 2'-FL and LNnT blend at 5, 10 or 20 g a day, or 2 g glucose placebo, for 2 weeks. Bifidobacterium rose more than on placebo with 10 g of 2'-FL alone (not with 5 or 20 g of 2'-FL), with all three LNnT doses and with 10 or 20 g of the blend. Fifty-six adverse events in 44 people were all mild; at 20 g of 2'-FL alone, nausea, rumbling, bloating, gas, diarrhea, loose stools and urgency scored higher than on placebo, though mean scores stayed at mild discomfort or below.
- Human milk oligosaccharide supplementation in irritable bowel syndrome patients: A parallel, randomized, double-blind, placebo-controlled study. Neurogastroenterol Motil. 2020;32(10):e13920. doi: 10.1111/nmo.13920.PubMedUsed to support: Glycom A/S is listed as the grant source and four authors were Glycom employees (NCT02875847). Double-blind trial in 60 adults with IBS (Rome IV) taking 5 g or 10 g a day of a 4:1 blend of 2'-FL and LNnT, a two-HMO blend, or glucose placebo for 4 weeks. Bifidobacteria, the registered primary outcome, rose at week 4 with 10 g compared with the other groups, but not at week 8. Overall and individual gastrointestinal symptoms did not differ between groups at week 4 or 8, and symptoms did not worsen.
- Human Milk Oligosaccharides Support Normal Bowel Function and Improve Symptoms of Irritable Bowel Syndrome: A Multicenter, Open-Label Trial. Clin Transl Gastroenterol. 2020;11(12):e00276. doi: 10.14309/ctg.0000000000000276.PubMedUsed to support: Sponsored by Glycom, Inc., which helped design the trial and write the paper; three authors were Glycom employees (NCT03550742). Open-label, single-arm trial with no placebo in 317 adults with IBS taking 5 g a day of a 4:1 blend of 2'-FL and LNnT for 12 weeks; 245 completed per protocol. Abnormal stool consistency and symptom scores fell from baseline. The authors called for a placebo-controlled trial. Common side effects were mild gas, abdominal pain and distension.
- Human Milk Oligosaccharides Modulate Fecal Microbiota and Are Safe for Use in Children With Overweight: A Randomized Controlled Trial. J Pediatr Gastroenterol Nutr. 2021;73(3):408-414. doi: 10.1097/MPG.0000000000003205.PubMedUsed to support: Glycom A/S is the lead sponsor of the registered trial (NCT02786160), and three authors were Glycom employees. Randomized, double-blind, placebo-controlled trial in 75 children aged 6 to 12 with overweight, including obesity, taking 2'-FL, a 2'-FL and LNnT blend, or glucose placebo once a day for 8 weeks; the primary outcome was fecal microbiota. Bifidobacteria rose more than on placebo with 2'-FL at weeks 4 and 8, while the blend showed only a trend (p = 0.071). Symptom rating scale, stool form and blood safety results raised no concerns.
- Effects of Infant Formula With Human Milk Oligosaccharides on Growth and Morbidity: A Randomized Multicenter Trial. J Pediatr Gastroenterol Nutr. 2017;64(4):624-631. doi: 10.1097/MPG.0000000000001520.PubMedUsed to support: Sponsored by Nestec Ltd., which supplied the formula; several authors were from Nestle units (NCT01715246). Healthy infants were randomized to formula with 1.0 g/L 2'-FL plus 0.5 g/L LNnT, a two-HMO blend (n = 88), or the same formula without HMOs (n = 87) to 6 months. Weight gain through 4 months, the primary endpoint, met noninferiority. Parent-reported bronchitis, lower respiratory infections and antibiotic and antipyretic use were lower in secondary analyses, which the authors said need confirmation.
- Human Milk Oligosaccharides: 2'-Fucosyllactose (2'-FL) and Lacto-N-Neotetraose (LNnT) in Infant Formula. Nutrients. 2018;10(9):1161. doi: 10.3390/nu10091161.PubMedUsed to support: Review of HMO biology and the early infant formula trials, mainly of 2'-FL. It describes HMOs as non-digestible carbohydrates and the third largest solid component of breast milk after lactose and lipids, with more than 200 structures and 10 to 15 g per liter in mature milk, and says the anti-adhesion (decoy) evidence comes from laboratory and tissue studies. It concludes that adding HMOs to formula appears safe and supports normal growth and that more randomized trials are needed. No external funding; two authors were from Nestlé units and the authors declared no conflict of interest.
- Limited Evidence of Benefits from Clinical Trials of Human-Identical Milk Oligosaccharides for Infants. Adv Nutr. 2026;17(3):100593. doi: 10.1016/j.advnut.2026.100593.PubMedUsed to support: Independent review of 12 randomized trials of human-identical milk oligosaccharides in infants and young children, most often 2'-FL alone or with other HMOs. All trials in healthy infants showed growth and tolerance no worse than control formula; results on illness were inconsistent and the trials were underpowered for serious illness. The authors reported no conflicts of interest.
- Human milk oligosaccharide 2'-fucosyllactose links feedings at 1 month to cognitive development at 24 months in infants of normal and overweight mothers. PLoS One. 2020;15(2):e0228323. doi: 10.1371/journal.pone.0228323.PubMedUsed to support: NIH-supported observational cohort of 50 Hispanic mother-infant pairs; the authors declared no competing interests. 2'-FL in breast milk at 1 month predicted higher Bayley-III cognitive scores at 24 months, while 2'-FL at 6 months showed no association. No supplement was given and no adults were studied.
- Human Milk Oligosaccharides and Infant Neurodevelopment: A Narrative Review. Nutrients. 2023;15(3):719. doi: 10.3390/nu15030719.PubMedUsed to support: Narrative review that found six studies of HMO exposure and infant neurodevelopment, all observational. The five in full-term infants reported positive associations with cognitive, language and motor scores, and the one preterm study found no significant results in the full cohort. It includes no supplement trials and is not evidence that HMO supplements affect cognition in children or adults.
- Effects of 3'-Sialyllactose on Symptom Improvement in Patients with Knee Osteoarthritis: A Randomized Pilot Study. Nutrients. 2024;16(19):3410. doi: 10.3390/nu16193410.PubMedUsed to support: Randomized, single-blind pilot trial in 60 adults over 40 with knee osteoarthritis taking 200 mg or 600 mg of 3'-SL or placebo once a day for 12 weeks. VAS pain (the primary outcome) and KWOMAC scores fell from baseline in the 3'-SL groups, but neither differed between groups at any time point. No serious adverse events occurred. Funded by Korean government grants (MOTIE and KIAT); the tablets were made by GeneChem Inc., which employed three authors, one of whom held GeneChem stock.
- Efficacy of Human Milk Oligosaccharide 6'-Sialyllactose Supplementation on Exercise Performance and Training Adaptations. Nutrients. 2026;18(11):1743. doi: 10.3390/nu18111743.PubMedUsed to support: Funded by GeneChem through a service contract with Texas A&M University; the authors declared no conflicts (ISRCTN50097942, registered January 2026). Randomized, double-blind, placebo-controlled proof-of-concept trial in recreationally active men taking 900 mg a day of 6'-SL or placebo during 12 weeks of resistance training; 19 of 28 randomized completed. Body composition, strength and blood lactate, the primary outcomes, did not differ between groups. Ventilatory threshold and oxygen uptake at that threshold were lower with 6'-SL at 12 weeks, while some peak aerobic measures rose more at 6 weeks; the authors call the results exploratory. Side effects did not differ between groups.
- Randomized, triple-blind, placebo-controlled study to evaluate the safety of 6'-Sialyllactose in healthy adults. Regul Toxicol Pharmacol. 2022;129:105110. doi: 10.1016/j.yrtph.2021.105110.PubMedUsed to support: Randomized, triple-blind, placebo-controlled safety trial in healthy adults taking 3 g of 6'-SL or maltodextrin twice a day for 12 weeks. No serious adverse reactions occurred, liver and kidney blood tests showed no clinically significant changes, and digestive complaints such as diarrhea, abdominal discomfort and bloating did not differ in frequency from placebo. The abstract notes EFSA has assessed 6'-SL sodium salt as a novel food; it does not give the number of participants or the funding source.
- Safety of the extension of use of 2'-fucosyllactose (2'-FL) as a novel food pursuant to Regulation (EU) 2015/2283. EFSA J. 2023;21(11):e08334. doi: 10.2903/j.efsa.2023.8334.PubMedUsed to support: EFSA opinion on raising the 2'-FL limits in infant and follow-on formula. It records that 2'-FL is on the EU Union list of novel foods, made by chemical synthesis or by fermentation with genetically modified strains of E. coli K-12 DH1, E. coli BL21(DE3) or Corynebacterium glutamicum, and concludes the extension of use is safe. EFSA notes it does not assess claimed benefits.