Kefir: what the trials measured

Kefir has one finding that holds up across pooled trials, and it is about blood sugar. A 2026 meta-analysis of 24 intervention studies in adults found fasting glucose 8.46 mg/dL lower on kefir, with no significant change in body measures, blood lipids or inflammation markers. An earlier pooling of 6 randomised trials in 323 people put the drop at 10.28 mg/dL, and the fall in HbA1c, the three-month average, did not reach significance. So the daily number moves, and the marker that matters over months has not yet been shown to follow it. Most of the rest of the science is single trials of a few dozen people, and several of them compared kefir with plain milk rather than with nothing. Where that was done, in 75 overweight women drinking four dairy servings a day for 8 weeks, cholesterol and weight fell against the control group and kefir did no better than milk. One trial on this page did not use a drink at all. The bone density study in 40 people with osteoporosis gave 1,600 mg a day of kefir-fermented milk alongside calcium, under two grams, so it says nothing about a glass. Who should be careful with kefir is the question our cards do not answer. The only safety trial we hold was run in healthy young adults.

The numbers, per 100 g

How much is actually in it

• One cup of kefir is 244 g in FNDDS, and by the single branded USDA record for low-fat plain kefir it carries about 0.7 µg of vitamin B12, 0.33 mg of riboflavin and 417 µg RAE of vitamin A, the last from fortification. — FNDDS Kefir (code 11115400, fdc_id 2705394) read against USDA SR Legacy Kefir, lowfat, plain, LIFEWAY (FDC 170904) (n = 2 household measures; 2 branded composition records) Evidence E source
  1 cup 244 g, 1 fl oz 30 g; per 100 g vitamin A 171 µg RAE (171 to 178 across the 2 records), vitamin B12 0.29 µg, riboflavin 0.135 mg; per cup about 417 µg, 0.71 µg and 0.33 mg

What your body absorbs

• The EU has authorised the claim that live cultures in yoghurt or fermented milk improve lactose digestion of the product in people who have difficulty digesting lactose. The register entry is written for yoghurt cultures and does not name kefir. — EU market; Article 13.1 claim for live yoghurt cultures (n = 1 claim) Evidence E source
  Claim status Authorised; applies to the product's own lactose, not to lactose from other foods
• To carry the EU lactose digestion claim, a yoghurt or fermented milk must contain at least 100 million live Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus per gram, so a kefir made with other cultures cannot use it. — EU market, Regulation (EU) No 432/2012 consolidated 2016, list of permitted health claims (n = 1 condition of use) Evidence E source
  Threshold 10^8 CFU per gram of the two named yoghurt starter organisms

What it does to your health

• A 2026 meta-analysis of 24 intervention studies found that kefir lowered fasting blood glucose by 8.46 mg/dL and HOMA-IR by 1.71 in adults, with no significant effect on body measures, blood lipids or inflammatory markers. — adults in 24 interventional studies identified from 702 articles, searched to 25 January 2025, subgroups by kefir dose (n = 24 studies) Evidence A source
  Fasting glucose MD -8.46 mg/dL (p = 0.006); HOMA-IR MD -1.71 (p = 0.004); anthropometrics, lipids and inflammation not significant
• A meta-analysis of 6 randomised trials in 323 people found that kefir lowered fasting blood sugar by 10.28 mg/dL (95% CI 4.02 to 16.53) and insulin by 2.87 units, while the drop in HbA1c of 0.64 was not significant. — 323 participants in 6 RCTs published to July 2020, varied duration and dose of kefir (n = 6 RCTs, 323) Evidence A source
  FBS WMD -10.28 (95% CI -16.53 to -4.02; P = 0.001); insulin WMD -2.87 (95% CI -3.96 to -1.78); HbA1c -0.64 (95% CI -1.36 to 0.08; P = 0.08)

Safety, limits and interactions

• In 65 healthy young adults, 250 mL of lactose-free kefir daily for 6 weeks reduced gastrointestinal symptoms and serum cholesterol, creatinine and uric acid (P < 0.05) but did not improve immunity or inflammation markers. — 65 healthy young adult volunteers, randomised controlled trial, 250 mL lactose-free kefir a day for 6 weeks, registered NCT06612164 (n = 65) Evidence B source
  Gastrointestinal symptoms, serum cholesterol, creatinine and uric acid decreased (P < 0.05); immunity, inflammation, glucose, CRP, sleep and mental health scores not significant

What people believe that the data does not support

• In a meta-analysis of 14 trials of fermented dairy, of which only 2 used kefir, C-reactive protein fell by a standardised mean difference of 0.21, a small effect, and TNF-alpha, IL-6 and IL-10 did not change. — 14 trials, 15 arms: yogurt 9, fermented milk 4, kefir 2; healthy and metabolically ill adults (n = 14 trials) Evidence A source
  CRP SMD -0.21 (95% CI -0.40 to -0.02; P = 0.033); IFN-gamma SMD +0.12; TNF-alpha, IL-6, IL-10, IL-2 not significant
• In 75 overweight women given 2 extra daily servings of kefir or low-fat milk for 8 weeks, total cholesterol fell 10.4 mg/dL and LDL 9.7 mg/dL against control, but kefir did no better than plain milk. — 75 overweight or obese Iranian women aged 25 to 45, 8-week parallel trial: control 2 dairy servings a day, milk or kefir arms 4 servings a day (n = 75) Evidence B source
  Kefir vs control: TC -10.4 mg/dL, LDL-C -9.7, non-HDL-C -11.5, TC/HDL -0.4, LDL/HDL -0.3 (all P < .05); kefir vs milk: no significant difference
• In the same 75 overweight women, 8 weeks of 4 daily dairy servings from kefir or milk without calorie restriction cut weight, BMI and waist more than 2 servings (all p < 0.01), and kefir and milk did not differ. — 75 overweight or obese premenopausal women in Iran, intention-to-treat, 58 completers, 8 weeks (n = 75) Evidence B source
  Weight, BMI and waist circumference reduced vs control (p < 0.01) in both kefir and milk arms; kefir vs milk not significant
• In a double-blind trial of 125 children aged 1 to 5 years on antibiotics, a daily kefir drink did not prevent diarrhoea: 18% versus 21.9% on heat-killed placebo, relative risk 0.82 (95% CI 0.54 to 1.43). — 125 children aged 1 to 5 years prescribed antibiotics in primary care, Washington DC area, 14-day follow-up (n = 125) Evidence B source
  Diarrhoea 18% kefir vs 21.9% placebo; RR 0.82 (95% CI 0.54 to 1.43); no difference in secondary outcomes

Who this works differently for

• In a 6-week double-blind trial in 53 UK children aged 8 to 13 with ADHD, daily kefir did not change ADHD symptom severity. Actigraphy showed 70 versus 90 minutes awake at night (p = 0.04), while the children reported more sleep problems. — 53 children with ADHD aged 8 to 13 years in the UK who completed a 6-week randomised double-blind placebo-controlled trial (n = 53) Evidence B source
  Parent and teacher SWAN scores: no significant effect; minutes awake in the down period 70.10 vs 89.72 (p = 0.04); self-reported sleep problems higher with kefir (p = 0.02); no change in microbiota diversity
• In 67 overweight men aged 65 and over, 240 mL a day of kefir fortified with 2 probiotic strains for 8 weeks lowered depression scores compared with regular kefir (p = 0.001), with no difference in inflammation or oxidative stress except total antioxidant capacity. — 67 Iranian overweight or obese men over 65, double-blind: 35 on regular kefir as placebo, 32 on probiotic-fortified kefir, 8 weeks (n = 67) Evidence B source
  GDS-15 depression score differed between groups after 8 weeks (p = 0.001); TAC change p = 0.009; other oxidative and inflammatory markers not significant
• In an open-label trial of 45 people with inflammatory bowel disease, 400 mL of kefir a day for 4 weeks raised faecal Lactobacillus and, in Crohn's disease, lowered ESR and CRP and reduced bloating scores (p = 0.012). — 45 IBD patients in Turkey, 25 kefir and 20 control, single centre, open label, 4 weeks (n = 45) Evidence B source
  Crohn's disease: ESR and CRP decreased, haemoglobin increased, bloating scores reduced in the last 2 weeks (p = 0.012), feeling-good scores up (p = 0.032); faecal Lactobacillus load rose (p = 0.001 UC, p = 0.005 CD)
• In 62 people with metabolic syndrome, 180 mL a day of probiotic kefir for 12 weeks raised apolipoprotein A1 by 3.4% while it fell 2.4% on unfermented milk (P = .03), and lowered TNF-alpha, IL-6 and homocysteine. — 62 adults with metabolic syndrome randomised to 180 mL/d probiotic kefir or unfermented milk for 12 weeks, NCT03966846 (n = 62) Evidence B source
  ApoA1 +3.4% kefir vs -2.4% milk (P = .03); in the LDL > 130 mg/dL subgroup LDL-C -7.6% and ApoB -5.4% vs baseline, not vs milk; blood pressure fell in both arms
• In a 6-month double-blind pilot in 40 people with osteoporosis, kefir-fermented milk 1,600 mg plus 1,500 mg calcium carbonate was associated with greater 6-month gains in hip bone density than calcium carbonate alone. — 40 osteoporosis patients, parallel double-blind trial, 6 months, kefir-fermented milk with CaCO3 vs CaCO3 alone, NCT02361372 (n = 40) Evidence B source
  Greater 6-month increase in hip BMD; beta-CTX decreased at 3 months in those with T-score > -1; osteocalcin turned positive at 6 months; PTH rose with kefir

What is still unknown

• A 2026 systematic review found only 8 human intervention studies on kefir and gut microbiota: changes were modest and heterogeneous, Bifidobacterium rose in 3 of 4 studies, and direct functional evidence in humans is limited. — 8 human interventional studies selected from 2,743 records in Web of Science, PubMed and Scopus; risk of bias by Cochrane tools (n = 8 studies) Evidence B source
  Minor changes at phylum and class level; Bifidobacterium increased in 3 of 4 studies; kefir-specific strains detected in faeces

The bottom line

If you like kefir, drink it as a dairy food. Nothing on this page turns it into a treatment. One cup is 244 g, and by the single branded USDA record for low-fat plain kefir it gives about 0.7 µg of vitamin B12, 0.33 mg of riboflavin and 417 µg RAE of vitamin A, and the vitamin A comes from fortification. On lactose, the EU allows a fermented milk to claim that its live cultures help people who struggle with lactose digest the lactose in the product itself. The claim belongs to two named yoghurt organisms at 100 million live cells per gram, so a kefir made with other cultures cannot carry it, and it never covered the lactose in the rest of your meal. The gut story is thinner than the marketing. A 2026 systematic review found only 8 human intervention studies on kefir and gut bacteria, with modest and inconsistent shifts, and Bifidobacterium rising in 3 of the 4 studies that looked. In 125 children aged 1 to 5 on antibiotics, a daily kefir drink did not prevent diarrhoea: 18 percent got it against 21.9 percent on heat-killed placebo, a gap the trial could not separate from chance. Inflammation reads much the same. Across 14 trials of fermented dairy, of which only 2 used kefir, C-reactive protein fell by a small amount and the other markers stayed put. Several trials in particular groups show something, and each has a catch. In 67 overweight men over 65, 240 mL a day for 8 weeks lowered depression scores, but the comparison was kefir fortified with two probiotic strains against ordinary kefir, so ordinary kefir was the placebo. In 62 people with metabolic syndrome, 180 mL a day for 12 weeks raised apolipoprotein A1 by 3.4 percent while it fell 2.4 percent on unfermented milk. In 45 people with inflammatory bowel disease, 400 mL a day for 4 weeks cut bloating scores in Crohn's disease, in an open-label trial where everyone knew what they were drinking. In 53 children with ADHD, 6 weeks of kefir did nothing to symptoms. The activity monitors recorded 70 rather than 90 minutes awake at night, and the children themselves reported more sleep problems. Two readings of the same nights point in opposite directions, and we would not act on either. Who should be careful is the gap on this page. Our only safety card is a 6-week trial of 250 mL of lactose-free kefir a day in 65 healthy young adults, where gut symptoms and cholesterol went down and immunity markers did not move. We hold nothing on kefir in pregnancy, in people whose immune system is weakened, alongside any medicine, or with milk allergy. If one of those is you, the question goes to the person treating you, and this page cannot answer it.

Sources

1. kef-c4-16 · reference dataset
   FNDDS 2705394 Kefir — 1 cup 244 g · 1 fl oz 30 g || USDA SR Legacy 170904 Kefir, lowfat, plain, LIFEWAY — vitamin A 171 µg RAE · vitamin B12 0.29 µg · riboflavin 0.135 mg per 100 g
   USDA FNDDS 2021-2023 and USDA SR Legacy (FDC 170904) · checked 2026-09-23
2. kef-c4-14 · regulatory register entry
   POL-HC-6400 Live yoghurt cultures Live cultures in yoghurt or fermented milk improve lactose digestion of the product in individuals who have difficulty digesting lactose Authorised
   EU Register on nutrition and health claims, claim POL-HC-6400, snapshot 2026-09-21 · checked 2026-09-23
3. kef-c4-15 · regulation
   In order to bear the claim, yoghurt or fermented milk should contain at least 10⁸ Colony Forming Units live starter microorganisms (Lactobacillus delbrueckii subsp. bulgaricus and Streptococcus thermophilus) per gram.
   Regulation (EU) No 432/2012, consolidated text, EUR-Lex · checked 2026-09-23
4. kef-c4-01 · meta-analysis
   Kefir consumption induced a significant reduction of fasting blood glucose (MD= -8.46 mg/dL, p = 0.006), and HOMA-IR (MD= -1.71, p = 0.004). However, no significant effects were observed on anthropometric measures, lipid profile, or inflammatory markers.
   Nutr Metab Cardiovasc Dis, 2026 · checked 2026-09-23
5. kef-c4-02 · meta-analysis of RCTs
   In total, six RCTs on a total sample size of 323 subjects were selected for the meta-analysis. ... Findings indicated a more significant reduction in the FBS (WMD = -10.28; 95% CI: -16.53-4.02; P = 0.001) and insulin (WMD = -2.87; 95% CI: -3.96-1.78; P < 0.00001) in the kefir treatment group.
   Complement Ther Clin Pract, 2021 · checked 2026-09-23
6. kef-c4-09 · RCT
   participants in the kefir group consumed 250 mL of lactose-free kefir daily for 6 wk. ... The results revealed a significant decrease in gastrointestinal symptoms, along with reduced serum cholesterol, creatinine, and uric acid concentrations, after kefir intervention (P < 0.05). However, no significant improvement in immunity or inflammation markers was observed.
   J Nutr, 2026 · checked 2026-09-23
7. kef-c4-03 · meta-analysis
   Finally, fourteen trials (15 arms) were included in this meta-analysis: yogurt (n = 9), fermented milk (n = 4), and kefir (n = 2). ... Firstly, the analysis indicated that FDPs' supplementation decreased the levels of C-reactive protein (CRP) (SMD = -0.21; 95% CI: -0.40, -0.02; P = 0.033)
   Nutr Metab Cardiovasc Dis, 2023 · checked 2026-09-23
8. kef-c4-04 · RCT
   mean between-group differences were -10.4 mg/dL, -9.7 mg/dL, -11.5 mg/dL, -0.4, and -0.3 for TC, LDLC, non-HDLC, TC/HDLC, and LDLC/HDLC, respectively; all P < .05). Similar results were observed in the milk group. However, no such significant differences were found between the kefir and milk groups.
   J Clin Lipidol, 2017 · checked 2026-09-23
9. kef-c4-05 · RCT
   subjects in the kefir and milk groups had significantly greater reductions in weight, BMI, and WC compared to those in the control group (all p < 0.01). However, no such significant differences were found between the kefir and milk groups.
   Eur J Nutr, 2016 · checked 2026-09-23
10. kef-c4-06 · RCT
    There were no differences in the rates of diarrhea per group, with 18% in the active group and 21.9% in the placebo group (relative risk, 0.82; 95% confidence interval, 0.54-1.43).
    Arch Pediatr Adolesc Med, 2009 · checked 2026-09-23
11. kef-c4-07 · RCT
    Kefir had no significant overall effect on parent or teacher-rated ADHD symptom severity. ... Actigraphy revealed the kefir group spent fewer minutes awake during the down period at week six (M = 70.10, SE = 0.09) than the placebo group (M = 89.72, SE = 0.07), p = 0.04.
    BMC Psychiatry, 2025 · checked 2026-09-23
12. kef-c4-08 · RCT
    One group (n = 35) received one bottle (240 cc) of regular kefir as a placebo, while the intervention group (n = 32) received one bottle of probiotic-fortified kefir for eight weeks. ... A significant difference in depression scores and the mean change between the two groups was observed after eight weeks (p = 0.001 and p = 0.042, respectively).
    J Health Popul Nutr, 2025 · checked 2026-09-23
13. kef-c4-10 · RCT, open label
    A 400 mL/day kefir was administered to the patients for 4 weeks day and night. ... For patients with CD, there was a significant decrease in erythrocyte sedimentation rate and C-reactive protein, whereas hemoglobin increased, and for the last 2 weeks, bloating scores were significantly reduced (p=0.012), and feeling good scores increased (p=0.032).
    Turk J Gastroenterol, 2019 · checked 2026-09-23
14. kef-c4-11 · RCT
    Sixty-two participants were randomly assigned to receive either 180 mL/d probiotic kefir or unfermented milk for 12 weeks. ... Serum apolipoprotein A1 concentration increased by 3.4% in the kefir group, whereas it decreased by 2.4% in the milk group in 12 weeks (P = .03).
    Nutr Res, 2022 · checked 2026-09-23
15. kef-c4-12 · RCT, pilot
    In this pilot study, we concluded that kefir-fermented milk therapy was associated with short-term changes in turnover and greater 6-month increases in hip BMD among osteoporotic patients.
    PLoS One, 2015 · checked 2026-09-23
16. kef-c4-13 · systematic review
    The search resulted in 2743 studies, of which eight studies met our eligibility criteria. Overall, kefir resulted in minor changes in phyla and class levels. On the other hand, the Bifidobacterium genus increased in 3 out of 4 studies.
    Microbiologyopen, 2026 · checked 2026-09-23