Nutrition Research 102 (2022) 59–70

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/NTR

Probiotic kefir consumption improves serum

apolipoprotein A1 levels in metabolic syndrome
patients: a randomized controlled clinical trial

Ezgi Bellikci-Koyu a,b, Banu Pınar Sarer-Yurekli c, Cem Karagozlu d, Fadime Aydin-Kose e,
Ahmet Gokhan Ozgen c, Zehra Buyuktuncer a,∗
a Department of Nutrition and Dietetics, Faculty of Health Sciences, Hacettepe University, Ankara, Turkey
b Department of Nutrition and Dietetics, Faculty of Health Sciences, Izmir Katip Celebi University, Izmir, Turkey
c Department of Endocrinology, Faculty of Medicine, Ege University, Izmir, Turkey
d Department of Dairy Technology, Faculty of Agriculture, Ege University, Izmir, Turkey
e Department of Biochemistry, Faculty of Pharmacy, Izmir Katip Celebi University, Izmir, Turkey

a r t i c l e i n f o a b s t r a c t

Article history: Metabolic syndrome has become a major health hazard of the modern world. Studies inves-
Received 7 August 2021 tigating the effects of traditional fermented foods on metabolic syndrome are limited. We
Revised 25 February 2022 hypothesized that regular kefir consumption could improve the anthropometrical measure-
Accepted 28 February 2022 ments, glycemic control, lipid profile, blood pressure, and inflammatory status in patients
with metabolic syndrome. Sixty-two participants were randomly assigned to receive either
180 mL/d probiotic kefir or unfermented milk for 12 weeks. Dietary intake, anthropometri-
Keywords: cal measurements, biochemical status, and blood pressure were assessed at baseline and
Diabetes the end of weeks 4, 8, and 12. Serum apolipoprotein A1 concentration increased by 3.4%
Glycemic control in the kefir group, whereas it decreased by 2.4% in the milk group in 12 weeks (P = .03). A
Hyperlipidemia subgroup analysis for participants with low-density lipoprotein cholesterol (LDL-C) levels
Hypertension >130 mg/dL showed that serum LDL-C and apolipoprotein B concentrations (7.6% and 5.4%,
Kefir respectively) significantly decreased with kefir consumption compared with the baseline
Probiotic values at the 12th week (P < .05), but not compared with milk consumption (P > .05). Both
milk and kefir consumption was associated with lower systolic and diastolic blood pressure
compared with the baseline (P < .05). The 12-weeks of kefir administration also decreased
serum tumor necrosis factor-α, interleukin 6, interleukin 10, interferon-gamma, and homo-
cysteine concentrations significantly (P < .05). In conclusion, regular dairy consumption as
part of a well-balanced diet can provide favorable effects in the management of metabolic
syndrome, and probiotic kefir may deserve a special interest among dairy products. This
trial was registered at clinicaltrials.gov (NCT03966846).
© 2022 Elsevier Inc. All rights reserved.

Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; BMI, body
mass index; HbA1c, glycated hemoglobin; HDL-C, high-density lipoprotein cholesterol; HOMA-IR, homeostasis model assessment for
insulin resistance; IFN-γ , interferon gamma; IL, interleukin; LDL-C, low-density lipoprotein cholesterol; PAL, physical activity level; TNF-
α, tumor necrosis factor alpha.
∗
Corresponding author at: Hacettepe University, Faculty of Health Sciences, Department of Nutrition and Dietetics, 06230, Sihhiye
Ankara, Turkey. Tel: +903123051096 E-mail address: [email protected] (Z. Buyuktuncer).

https://doi.org/10.1016/j.nutres.2022.02.006
0271-5317/© 2022 Elsevier Inc. All rights reserved.
60 Nutrition Research 102 (2022) 59–70

the development of potential fermented food products with

1. INTRODUCTION different microbial compositions by the food industry.

Metabolic syndrome, a cluster of metabolic abnormali-

ties including abdominal obesity, insulin resistance, dys- 2. METHODS AND MATERIALS
lipidemia, and high blood pressure, has been regarded as
a global public health problem because of its high preva- 2.1. Subjects
lence and contribution to the increased risk of cardio-
vascular disease and type 2 diabetes mellitus [1,2]. Ab- This study was conducted on patients with metabolic syn-
dominal obesity and insulin resistance have been ac- drome who were recruited from the outpatient clinics in the
cepted as the main components of metabolic syndrome, Department of Internal Medicine, Endocrinology Division at
which trigger chronic low-grade inflammation. Although Ege University, Izmir, Turkey. Patients who were 18 to 65 years
both environmental and genetic factors contribute to the of age and met the criteria for the diagnosis of metabolic
development of metabolic syndrome, its pathophysiolog- syndrome published by the International Diabetes Federation
ical mechanisms still have not been fully understood were screened [23]. Metabolic syndrome was characterized by
[3,4]. the presence of at least 2 of the following components in ad-
Over the past 2 decades, the potential role of the gut mi- dition to abdominal obesity (waist circumference ≥94 cm for
crobiome has drawn attention because of its functions on en- men and ≥80 cm for women): (1) fasting plasma glucose con-
ergy homeostasis and substrate metabolism [5,6]. It has been centration: ≥100 mg/dL or previously diagnosed type 2 dia-
suggested that alterations in the microbial composition of the betes; (2) raised concentration of triglycerides: ≥150 mg/dL; (3)
gut microbiome and its functions might be involved in the reduced concentration of high-density lipoprotein cholesterol
progress of metabolic syndrome [7]. Thereby, the modulation (HDL-C): <40 mg/dL in men and <50 mg/dL in women; and
of the gut microbiome has become an emerging issue. Pro- (4) raised blood pressure: systolic blood pressure ≥130 mm Hg
biotics are the most promising agents in the modification of or diastolic blood pressure ≥85 mm Hg, or treatment of previ-
the gut microbiome [8]. The effects of probiotics on different ously diagnosed hypertension [23]. The exclusion criteria were
components of metabolic syndrome have been examined sep- having a dairy allergy, lactose intolerance, chronic gastroin-
arately. However, research focusing on patients with metabolic testinal system disease, type 1 diabetes, cancer, severe liver or
syndrome is limited [9–11]. kidney disease, abnormal thyroid hormone levels or immun-
Kefir is a traditional fermented milk product with a com- odeficiency, as well as being pregnant or lactating, taking med-
plex and symbiotic culture containing Lactobacillus kefiri, ication to regulate blood glucose (except regular use of met-
species of the genera Leuconostoc, Lactococcus, and Acetobacter formin) or lipid levels, having taken antibiotics until 1 month
as bacteria, and lactose fermenting (Kluyveromyces marxianus) before the study, consuming regular probiotic foods or sup-
and non-lactose fermenting (Saccharomyces unisporus, Saccha- plements, taking a supplement that may affect the metabolic
romyces cerevisiae, and Saccharomyces exiguus) species as yeast outcomes (such as prebiotics or omega-3 fatty acids), or di-
[12,13]. In previous studies, kefir and/or bacteria isolated from eting for weight loss or another disease. Participants taking
kefir grains showed several biological activities that can im- metformin or antihypertensive medication were allowed to
prove health and prevent diseases, including antihyperlipi- be included in the study only if their medical treatment had
demic [14], antidiabetic [15], antihypertensive [16,17], and an- not been modified in the past 3 months before the baseline
tiobesity activities [18,19]. Rosa et al. also reported some ben- of the study and throughout the 12-week intervention period.
eficial effects of kefir on metabolic syndrome components in The study protocol was approved by the Clinical Trials Ethical
rats [20]. However, these activities are mainly demonstrated in Board of Ege University Faculty of Medicine (15-2.1/14). The
animal models, and clinical investigations are still limited [21]. study was conducted according to the guidelines of the Dec-
Therefore, further clarifications with randomized controlled laration of Helsinki and written consent was obtained from
trials are necessary to prove their efficiency in humans. all participants. The trial was registered at clinicaltrials.gov
Our previous report focused on the alterations in gut mi- (NCT03966846).
crobial composition by the 12-week kefir consumption [22],
and this study aimed to examine the potential effects of kefir 2.2. Study Design
consumption on the anthropometrical and biochemical sta-
tus of patients with metabolic syndrome. We hypothesized This single-center, parallel-group, randomized, and controlled
that 12-week kefir consumption could improve the anthropo- trial was conducted between March 2015 and July 2017. Follow-
metrical measurements, glycemic control, lipid profile, blood ing the screening period, the participants were assigned into 2
pressure, and inflammatory status in patients with metabolic groups (1:1) using stratified randomization based on sex: kefir
syndrome. To test our hypothesis, we conducted a random- group or unfermented milk (control) group. The randomiza-
ized controlled clinical trial that assessed the effects of ke- tion list was provided using a computerized random-number
fir consumption on the biomarkers of each metabolic syn- generator by a statistician from Hacettepe University Depart-
drome component and compared it with milk consumption. ment of Biostatistics who was not involved in the study. Pa-
This study provides a better understanding of the potential ef- tients were enrolled in the study by the research physicians
fects of probiotics given in a matrix of fermented food on each who were blinded to interventions. The participants in the in-
component of metabolic syndrome. This is crucial both for the tervention group consumed a bottle of kefir (180 mL) daily,
implementation of evidence-based clinical interventions and whereas the control group consumed the same amount of
 Nutrition Research 102 (2022) 59–70 61

milk during the 12-week intervention period. All drinks were 2.5. Biochemical Analyses and Blood Pressure
served in identical bottles and delivered to the participants
twice per week in cold chain equipment. The drinks were con- Blood samples were collected following the 10-hour overnight
sumed by the participants at any time during the day with- fasting. Fasting serum glucose, insulin, total cholesterol, HDL-
out any restriction. The participants were requested to follow C, LDL-C, triglycerides, alanine transaminase (ALT), and aspar-
their habitual diet, physical activity, and other lifestyle behav- tate transaminase (AST) levels were analyzed at baseline and
iors during the intervention period. The sustainability of the at weeks 4, 8, and 12. Serum glycated hemoglobin (HbA1c),
habitual lifestyle during the study was assessed by face-to- apolipoprotein A1 (ApoA1), apolipoprotein B (ApoB), lipopro-
face interviews using the study diary and dietary recalls dur- tein (a), homocysteine, and high-sensitivity serum C-reactive
ing each visit. Participants visited the research center 5 times protein levels were analyzed at weeks 0 and 12. All analy-
in total. The first visit was for the screening of individuals ses were performed at Ege University, Department of Clini-
for the eligibility criteria. The second was carried out as the cal Biochemistry Laboratory, through the standard procedures.
baseline of the intervention (week 0), visits 3 and 4 were con- Serum glucose concentration was analyzed by the glucose
ducted, respectively, at weeks 4 and 8, and the fifth visit at hexokinase method using the HK Gen 3 kit (cat. no. 05168791
the end of the intervention (week 12). Compliance was de- 190), serum insulin concentration by the electrochemilumi-
fined as the consumption of >80% of the test drinks during the nescence immunoassay method using the Elecsys insulin as-
study period. The participants were asked to record their con- say (cat. no. 12017547 122), serum total cholesterol concentra-
sumption pattern of drinks on the study diary; this diary was tion by the enzymatic colorimetric method using the choles-
checked during each visit to the research center. Furthermore, terol Gen 2 kit (cat. no 05168538 190), serum HDL-C concen-
the number of consumed drinks was inquired and recorded at tration by homogeneous enzymatic colorimetric assay using
each delivery visit. The participants were also asked to record the HDL-cholesterol plus 3rd generation kit (cat. no. 05168805
any adverse effects of test drinks on the study diary, and the 190), serum triglyceride concentration by colorimetric assay
records were checked by the researchers at each visit. using the triglycerides kit (cat. no. 05171407 190), serum ALT
and AST concentrations using the ALT and AST kits (cat. no.
20764957 322 and 20764949 322) according to the International
Federation of Clinical Chemistry method without pyridoxal
2.3. Sample Size
phosphate activation, HbA1c concentration by the turbidi-
metric inhibition immunoassay method using the Tina-quant
Our primary outcome was HDL-C. The secondary outcomes
hemoglobin A1c kits (cat. no. 04528123 190), serum ApoA1 and
were lipid parameters (triglyceride, total cholesterol, low-
ApoB concentrations by the immunoturbidimetric method us-
density lipoprotein cholesterol [LDL-C]), glycemic parameters,
ing the Tina-quant ApoA1 version 2 kit (cat. no. 03032566 122)
inflammatory parameters, blood pressure, and anthropomet-
and the Tina-quant ApoB version 2 kit (cat. no. 03032574 122),
ric measurements. We used the biochemical analyses of a pi-
serum lipoprotein (a) concentration by the immunoturbidi-
lot group including 10 patients with metabolic syndrome to
metric method using the Tina-quant lipoprotein (a) Gen 2 kit
evaluate the effects of kefir consumption. The HDL-C mea-
(cat. no. 05852625 190), and high-sensitivity serum C-reactive
surements of the 2 groups with and without kefir consump-
protein level by the particle-enhanced immunoturbidimetric
tion over 1 month were 47.6 ± 8.4 mg/dL and 40.1 ± 10.8 mg/dL
method using the cardiac C-reactive protein (latex) high sen-
(effect size, 0.77). Using these values, we calculated the mini-
sitivity kit (cat. no. 04628918 190) of the Cobas system (Roche
mum required sample size for each study arm as 28 to achieve
Diagnostics GmbH, Mannheim, Germany). Samples were pro-
a minimum of 80% of study power and to identify a clinically
cessed and analyzed using Roche Cobas 8000 c502, c702, and
significant difference between groups at a 5% of type I error
e602 modular analyzer series. Serum homocysteine concen-
level and a 2-sided hypothesis testing design based on the
tration was analyzed by competitive chemiluminescent en-
HDL-C concentrations. Finally, we included 62 patients (31 in
zyme immunoassay using the Immulite 2000 homocysteine
each group) in the study. Sample size estimation calculations
kit (cat. no. L2KH02) on an Immulite 2000 analyzer (Siemens
were conducted using the G∗ Power 3.1 software.
Medical Solutions Diagnostics, Los Angeles, CA). Serum LDL-
C concentration was calculated based on the Friedewald for-
mula in patients with serum triglycerides levels <400 mg/dL
2.4. Assessment of Dietary Intake and Lifestyle Behaviors [25], whereas it was analyzed by homogeneous enzymatic col-
orimetric assay using the LDL-cholesterol Gen 2 kit (Cobas;
Dietary intake was assessed using the 24-hour dietary recall Roche Diagnostics GmbH) in patients with serum triglycerides
method conducted by a research dietician at each visit. A pho- levels >400 mg/dL. Homeostasis model assessment for insulin
tographic atlas was used to determine the portion sizes of resistance (HOMA-IR) was calculated based on the following
food items. Energy, macronutrient, and micronutrient intakes formula: fasting blood glucose (mg/dL) × fasting plasma in-
were calculated using the BeBIS Version 8 dietary analysis sulin/405 (μU/mL).
software (Turkish version of the BeBIS Nutrition Information Serum cytokine concentrations (tumor necrosis factor-
System). Information on lifestyle behaviors such as smoking, α [TNF-α], interleukin 6 [IL-6], interleukin 10 [IL-10], and
alcohol consumption, and physical activity levels (PALs) was interferon-gamma [IFN-γ ]) were analyzed at weeks 0 and
also recorded at the beginning of the study through face-to- 12. Blood samples were collected in a standard, sterile
face interviews. According to PALs, participants were classi- polystyrene vacuum tube. The samples were centrifugated
fied as sedentary, active, or very active [24]. at 2000g for 10 minutes, then serum samples were aliquoted
62 Nutrition Research 102 (2022) 59–70

into sterile plastic tubes and stored at -80°C until use. Serum microbial analysis of kefir was conducted at 4 times through-
cytokines were analyzed in fasting blood serum samples by out the study period, including the samples from the first and
a solid-phase type enzyme-linked immunosorbent assay us- fourth days of the production. It was shown that kefir provided
ing commercial kits (DIAsource ImmunoAssays, Louvain-la- a minimum of 106 CFU/g of Lactobacillus spp., Leuconostoc spp.,
Neuve, Belgium; cat. no. KAP1751, KAP1261, KAP1321, and and Lactococcus spp. during the storage period. Similarly, the
KAP1231, respectively) according to the manufacturer’s in- yeast amount was also constant (105 CFU/g) throughout the
structions. The assays were performed in 96-well microtiter study.
plates at room temperature. Then, the absorbance of plates
was read spectrophotometrically at 450 and 650 nm using a
2.8. Statistical Analyses
microplate reader (Thermo Varioskan Flash Multimode Mi-
croplate Reader; Thermo Fisher Scientific, Vantaa, Finland).
Statistical analyses were performed using SPSS version 22.0
The cytokine concentrations of the samples were calculated
(SPSS Inc., Chicago, IL). The categorical variables were pre-
with interpolation using the calibration curve prepared for
sented as numbers and proportions and analyzed by the χ 2
each parameter. IL-6, IL-10, and TNF-α concentrations were
or Fisher exact test. The continuous variables were presented
given as picograms per milliliter, whereas IFN-γ concentra-
as mean ± standard deviation unless otherwise specified. The
tion was given as international units per milliliter. All anal-
normality of the data was assessed with the Shapiro-Wilk test.
yses were run in duplicate.
Depending on the normality of the data, t test or the Mann-
The systolic and diastolic blood pressures of the partici-
Whitney U test was used for intergroup comparisons of the
pants were measured at the brachial artery of the right upper
continuous variables, whereas paired samples t test and the
arm using an automated blood pressure monitor (Omron M2
Wilcoxon signed-rank test were used for intragroup compar-
Intellisense; Omron Corp., Kyoto, Japan). All measurements
isons. For the comparison of the intervention effect between
were recorded in the sitting position after a 15-minute rest.
the groups, the differences between before and after the in-
tervention values were analyzed with the t test or the Mann-
2.6. Anthropometric Measurements Whitney U test. For all analyses, P < .05 was considered signif-
icant.
All anthropometrical measurements were taken by the re-
search dietician at the same time of the day during each visit,
and the participants were fasting when the measurements 3. RESULTS
were taken. Body mass index (BMI) was calculated as body
weight in kilograms divided by the square of height in me- 3.1. Participants
ters (kg/m2 ). Bioelectric impedance analysis was used to deter-
mine the fat mass and fat-free mass using a body composition Of the 94 participants invited to our study, 12 did not meet the
analyzer (TANITA BC 418; Tanita Corp., Tokyo, Japan). Waist cir- International Diabetes Federation metabolic syndrome crite-
cumference was measured at the midpoint between the lower ria. Three participants were excluded because of abnormal
rib and the iliac crest following normal breathing without any thyroid hormone levels, and 1 for receiving initial therapy with
compression or pressure. oral antidiabetic medication. Randomization was performed
for the remaining 78 participants, but 12 of them were ex-
cluded during the intervention because of various reasons.
2.7. Preparation of Kefir and Milk
The study was carried out with 62 participants who completed
the intervention (Fig. 1). The compliance rate was 97.8%, and
Kefir was produced from a bulk culture, which was prepared
the mean number of skipped servings was 1.8 ± 2.2 over the
from the DC1500I culture obtained from Danisco (Olsztyn,
12-week study period.
Poland). To prepare the bulk culture, 500 mL of reconstituted
milk with 12% dry matter prepared from skimmed milk pow-
der was sterilized in an autoclave at 115 °C for 10 minutes and 3.2. Baseline Characteristics and Dietary Intake
cooled to 25 °C. Then, 1 g of the culture was added to 500 mL
of milk. Afterward, the milk was thoroughly mixed and incu- Table 1 shows the baseline characteristics of the participants.
bated at 25 °C. The incubation was terminated when the pH of No significant difference was observed between the groups in
the bulk culture decreased to 4.6. For kefir preparation, the raw terms of participants’ age, smoking status, alcohol consump-
milk was heat-treated at 80 to 85 °C for 10 minutes, brought to tion, and PALs. The number of participants taking metformin
25 °C, and inoculated with the addition of 3.25% bulk culture. or hypertensive treatment was also similar between groups
The inoculated raw material was left to incubate in a stainless- (P> .05). The mean daily intake of metformin was 1473 ± 573
steel vessel. When the acidity of the kefir reached a pH of 4.7, mg in the kefir group and 1500 ± 519 mg in the milk group (P=
which lasted for approximately 18 hours from the culture ad- .90).
dition, the incubation was terminated and transferred to the The dietary intakes of participants throughout the study
cold storage (+4 °C) in 180-mL bottles. The microbial compo- are given in Table 2 (see also Supplementary Table S1). There
sition of the kefir included Lactococcus lactis subsp. lactis, Lac- were no significant differences in terms of energy, macronu-
tococcus lactis subsp. cremoris, Lactococcus lactis subsp. diacety- trient, and micronutrient intakes between the groups both at
lactis, Leuconostoc mesenteroides subsp. cremoris, Lactobacillus ke- the baseline and end of the study (P> .05 for each). During
fyr, Kluyveromyces marxianus, and Saccharomyces unisporus. The the study period, the calcium intake increased in both groups,
 Nutrition Research 102 (2022) 59–70 63

Table 1 – Baseline characteristics of the kefir and milk groups

Kefir group (n = 31) Milk group (n = 31)
n % n % P

Age, y (mean ± SD) 49.1 ± 8.5 50.5 ± 7.5 .51

Gender
Male 9 29.0 9 29.0 1.00
Female 22 71.0 22 71.0
Smoking status
Never 16 51.6 19 61.3 .76
Former 4 12.9 3 9.7
Current 11 35.5 9 29.0
Alcohol consumption
Nondrinker 18 58.1 23 74.2 .18
Drinker 13 41.9 8 25.8
Physical activity level
Sedentary 7 22.6 10 32.3 .39
Active 24 77.4 21 67.7
Presence of metabolic syndrome componentsa
IFG or type 2 diabetes
IFG 16 51.6 16 51.6
Type 2 diabetes 8 25.8 7 22.6 .94
Absent 7 22.6 8 25.8
Metformin treatment
Yes 11 35.5 14 45.2 .44
No 20 64.5 17 54.8
High blood pressure or hypertension
Present 21 67.7 22 71.0 .59
Absent 10 32.3 9 29.0
Hypertensive treatment
Yes 8 25.8 10 32.3 .58
No 23 74.2 21 67.7
High blood triglycerides
Present 21 67.7 21 67.7 1.00
Absent 10 32.3 10 32.3
Low HDL
Present 21 67.7 17 54.8 .30
Absent 10 32.3 14 45.2

Abbreviations: HDL-C, high-density cholesterol; IFG, impaired fasting glucose.

a
Categorizations (present/absent) of metabolic syndrome components were evaluated according to International Diabetes Federation 2005
criteria [23]. The kefir group consumed 180 mL/d probiotic kefir and the milk group consumed 180 mL/d unfermented milk for 12 weeks.
Abdominal obesity was present in all participants. Age was analyzed with t test. Smoking status was analyzed with Fisher exact test. For other
variables, the χ 2 test was used.

Table 2 – Dietary energy and nutrient intakes of the kefir and milk groups at the baseline and at the end of the study
Kefir group (n = 31) Milk group (n = 31)
x̄ ± SD P1 x̄ ± SD P2 Pbaseline Pend

Week 0 Week 12 Week 0 Week 12

Energy (kcal) 2008 ± 573 2087 ± 553 .45 1945 ± 756 2042 ± 560 .39 .47 .75
CHO (%) 42.6 ± 8.3 43.0 ± 7.5 .81 44.8 ± 8.6 44.4 ± 8.7 .55 .13 .50
Protein (%) 16.7 ± 2.7 15.7 ± 3.4 .08 15.9 ± 4.0 16.3 ± 3.7 .63 .34 .68
Fat (%) 40.8 ± 7.5 40.8 ± 6.5 .99 39.1 ± 7.4 39.4 ± 7.7 .86 .50 .45
Fiber (g) 26.5 ± 12.5 27.6 ± 11.8 .71 25.6 ± 8.9 26.4 ± 11.8 .74 .77 .80
Ca (mg) 887.3 ± 362.2 968.9 ± 297.6 .54 755.9 ± 324.0 1017.6 ± 271.4 <.01a .14 .50

Abbreviations: Ca, calcium; CHO, carbohydrates; SD, standard deviation.

Sixty-two patients with metabolic syndrome were randomly classified into 2 groups. The kefir group consumed 180 mL/d probiotic kefir and
the milk group consumed 180 mL/d unfermented milk for 12 weeks. Data are expressed as mean ± SD.
a
P < .05, P1 and P2 , within-group comparisons (week 0 and week 12), were performed using the paired sample t test or the Wilcoxon test.
Pbaseline was the week 0 comparisons of the groups and analyzed with either t test or the Mann-Whitney U test. Pend was the week 12 comparisons
of the groups and analyzed either with t test or the Mann-Whitney U test.
64 Nutrition Research 102 (2022) 59–70

Fig. 1 – Flow diagram of the study. This randomized, controlled clinical trial focused on the potential effects of regular kefir
and milk (as control) consumption on the components of metabolic syndrome. Patients who were 18 to 65 years old and
met the criteria for the diagnosis of metabolic syndrome published by the International Diabetes Federation were screened
[23]. The exclusion criteria were having a dairy allergy, lactose intolerance, chronic gastrointestinal system disease, type 1
diabetes, cancer, severe liver or kidney disease, abnormal thyroid hormone levels or immunodeficiency, as well as being
pregnant or lactating, taking medication to regulate blood glucose or lipid levels, having taken antibiotics until 1 month
before the study, consuming regular probiotic food or supplement, taking a supplement that may affect the metabolic
outcomes (such as prebiotics or omega-3 fatty acids), or dieting for weight loss or another disease. Of the 94 individuals
assessed for eligibility, 12 did not meet the International Diabetes Federation metabolic syndrome criteria. Three
participants were excluded because of their abnormal thyroid hormone levels, and 1 for receiving initial therapy with oral
antidiabetic medication. Randomization was performed for the remaining 78 participants who were classified into 2 groups
to receive either 180 mL/d probiotic kefir or milk for 12 weeks. Twelve participants were excluded after the allocation or
during the follow-up period because of taking antibiotics, nutritional supplements, or for personal reasons. The follow-up
was completed with a total of 62 participants, with 31 participants in each group.

but the increment was statistically significant only in the milk no significant change in fat mass and waist circumference ei-
group (P< .01). ther with kefir or milk consumption.

3.3. Anthropometrical Measurements

3.4. Biochemical Markers
The baseline anthropometrical measurements were similar
between the groups (P> .05 for each). However, body weight, The changes in the metabolic parameters are shown in Table 4
BMI, and total body water decreased significantly in the milk (see also Supplementary Table S3). Apart from the fasting
group compared with the kefir group following the 12-week serum glucose concentration, the biochemical parameters
intervention (Table 3 and Supplementary Table S2). There was of the participants showed a similar profile in each group at
 Nutrition Research 102 (2022) 59–70 65

Table 3 – Anthropometrical measurements of the kefir and milk groups at the baseline and at the end of the study
Kefir group (n = 31) Milk group (n = 31)
x̄ ± SD P1 x̄ ± SD P2 Pbaseline Pchange
Week 0 Week 12 Week 0 Week 12

Weight (kg) 84.7 ± 14.0 85.2 ± 14.5 .10 85.9 ± 14.3 85.2 ± 13.8 .03a .73 .01a
BMI (kg/m2 ) 32.3 ± 5.8 32.5 ± 5.9 .14 32.9 ± 5.4 32.7 ± 5.4 .03a .67 .01a
WC (cm) 106.2 ± 12.1 106.5 ± 12.4 .57 107.1 ± 9.5 106.7 ± 3.4 .42 .75 .32
FM (kg) 31.5 ± 11.3 31.8 ± 11.7 .50 31.7 ± 9.6 31.4 ± 9.5 .48 .95 .33
FFM (kg) 53.1 ± 9.9 53.5 ± 10.1 .11 54.2 ± 10.3 53.8 ± 10.2 .23 .55 .05
TBW (kg) 38.9 ± 7.2 39.2 ± 7.4 .13 39.9 ± 7.5 39.6 ± 7.5 .12 .44 .03a

Abbreviations: BMI, body mass index; FFM, fat-free mass; FM, fat mass; SD, standard deviation; TBW, total body water; WC, waist circumference.
Sixty-two patients with metabolic syndrome were randomly classified into 2 groups. The kefir group consumed 180 mL/d probiotic kefir and
the milk group consumed 180 mL/d unfermented milk for 12 weeks. Data are expressed as mean ± SD.
a
P < .05, P1 and P2 , intragroup comparisons (week 0 and week 12), were performed using the paired samples t test or the Wilcoxon test as
appropriate. Pbaseline was the week 0 comparisons of the groups and analyzed with either t test or the Mann-Whitney U test. Pchange compares the
changes from week 0 to week 12 between the groups and was analyzed with t test or the Mann-Whitney U test.

Table 4 – Biochemical profiles of the kefir and milk groups at the baseline and at the end of the study
Kefir group (n = 31) Milk group (n = 31)
x̄ ± SD P1 x̄ ± SD P2 Pbaseline Pchange
Week 0 Week 12 Week 0 Week 12

Glucose (mg/dL) 106.1 ± 10.9 107.5 ± 17.3 .85 101.3 ± 6.6 102.7 ± 9.9 .37 .04a .59
Insulin (mU/L) 16.6 ± 8.7 16.4 ± 10.1 .36 18.7 ± 9.4 19.8 ± 10.3 .14 .23 .12
HbA1c (%) 5.8 ± 0.5 5.8 ± 0.6 .12 5.6 ± 0.3 5.6 ± 0.4 .95 .09 .16
HOMA-IR 4.4 ± 2.5 4.4 ± 3.0 .53 4.7 ± 2.5 5.1 ± 2.8 .20 .52 .17
TC (mg/dL) 236.0 ± 40.6 229.7 ± 41.4 .20 231.4 ± 38.5 229.5 ± 35.7 .63 .64 .48
Triglycerides (mg/dL) 188.6 ± 68.1 209.8 ± 107.2 .13 182.0 ± 64.7 185.8 ± 87.3 .80 .70 .20
HDL-C (mg/dL) 43.9 ± 8.8 44.3 ± 10.4 .68 46.1 ± 9.3 45.3 ± 9.1 .45 .34 .40
LDL-C (mg/dL) 154.5 ± 34.4 145.7 ± 35.8 .08 148.9 ± 35.5 148.6 ± 32.9 .94 .52 .14
ApoA1 (mg/dL) 144.1 ± 15.9 148.9 ± 18.9 .06 149.1 ± 19.9 145.6 ± 20.7 .23 .27 .03a
ApoB (mg/dL) 141.6 ± 30.3 135.3 ± 28.2 .07 133.8 ± 27.5 131.7 ± 25.8 .70 .30 .49
Lipoprotein (a) (nmol/L) 53.6 ± 84.3 68.6 ± 99.6 .18 68.7 ± 151.8 40.2 ± 61.4 .92 .86 .54
Homocysteine (μmol/L) 11.58 ± 4.75 11.08 ± 4.57 .048a 11.59 ± 3.36 11.47 ± 2.72 .57 .53 .40
hs-CRP (mg/dL) 0.38 ± 0.42 0.32 ± 0.26 .98 0.36 ± 0.27 0.35 ± 0.26 .86 .34 .87
TNF-α (pg/mL) 17.58 ± 18.87 10.57 ± 12.67 .047a 13.95 ± 14.90 10.61 ± 11.50 .39 .44 .43
IL-6 (pg/mL) 19.69 ± 14.34 15.03 ± 13.50 .01a 21.55 ± 15.49 16.75 ± 11.39 .19 .53 .31
IL-10 (pg/mL) 48.89 ± 62.05 21.26 ± 35.01 <.01a 30.95 ± 33.99 29.75 ± 57.03 .33 .33 .17
IFN-γ (IU/mL) 1.24 ± 1.70 0.66 ± 1.09 <0.01a 1.19 ± 1.49 0.89 ± 1.68 .07 .57 .34
ALT (U/L) 23.3 ± 8.1 25.2 ± 7.9 0.35 28.8 ± 16.5 25.7 ± 12.0 .01a .37 .05
AST (U/L) 18.9 ± 4.3 19.1 ± 4.8 0.87 20.5 ± 6.7 19.1 ± 4.8 .06 .44 .24
SBP (mmHg) 128.9 ± 12.8 120.0 ± 15.9 <.01a 131.9 ± 16.7 123.0 ± 15.1 <.01a .43 .93
DBP (mmHg) 83.9 ± 7.4 79.5 ± 10.3 .04a 85.3 ± 9.9 81.5 ± 9.8 .02a .55 .79

Abbreviations: ALT, alanine transaminase; ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; AST, aspartate transaminase; DBP, diastolic blood
pressure; Hb, hemoglobin; HDL-C, high-density cholesterol; HOMA-IR, homeostasis model assessment for insulin resistance; hs-CRP, high-
sensitivity serum C- reactive protein; IFN-γ , interferon-gamma; IL, interleukin; LDL-C, low-density cholesterol; SBP, systolic blood pressure; SD,
standard deviation; TC, total cholesterol; TNF-α, tumor necrosis factor-α.
Sixty-two patients with metabolic syndrome were randomly classified into 2 groups. The kefir group consumed 180 mL/d probiotic kefir and
the milk group consumed 180 mL/d unfermented milk for 12 weeks. Data are expressed as mean ± SD.
a
P < .05, P1 and P2 , intragroup comparisons (week 0 and week 12), were performed using the paired samples t test or the Wilcoxon test as
appropriate. Pbaseline was the week 0 comparisons of the groups and analyzed with either t test or the Mann-Whitney U test. Pchange compares the
changes from week 0 to week 12 between the groups and was analyzed with t test or the Mann-Whitney U test.

the baseline. No significant change was observed in serum and diastolic blood pressures decreased significantly in both
glucose, insulin, HbA1c, HOMA-IR, total cholesterol, HDL-C, kefir and milk groups following the 12-week intervention
LDL-C, ApoB, triglycerides, or lipoprotein (a) levels by the period (Table 4). Among all biochemical parameters, the only
interventions. Following the 12-week kefir consumption, difference between the kefir and milk group was observed in
serum homocysteine and cytokine concentrations, including serum ApoA1, which was significantly increased in the kefir
TNF-α, IL-6, IL-10, and IFN-γ significantly decreased com- group compared with the milk group (P= .03).
pared with the baseline (P< .05), but not compared with A subgroup analysis was performed for the participants
milk consumption (P> .05 for each) (Table 4). The systolic with dyslipidemia (LDL-C >130 mg/dL). In the kefir group,
66 Nutrition Research 102 (2022) 59–70

Table 5 – Lipid profiles of the participants with dyslipidemia in kefir and milk groups at the baseline and at the end of the
study
Kefir group (n = 24) Milk group (n = 23)
x̄ ± SD P1 x̄ ± SD P2 Pbaseline Pchange
Week 0 Week 12 Week 0 Week 12

TC (mg/dL) 248.9 ± 35.7 239.8 ± 39.7 .10 245.2 ± 33.7 236.7 ± 35.6 .04a .72 .94
Triglycerides (mg/dL) 189.0 ± 59.6 201.0 ± 84.2 .22 173.7 ± 54.5 169.1 ± 79.7 .32 .36 .20
HDL-C (mg/dL) 45.3 ± 9.1 46.2 ± 10.7 .45 47.5 ± 8.6 46.4 ± 7.5 .34 .40 .23
LDL-C (mg/dL) 165.9 ± 30.2 153.4 ± 34.7 .03a 163.0 ± 27.3 158.4 ± 29.2 .19 .82 .21
ApoA1 (mg/dL) 145.5 ± 16.7 151.5 ± 20.1 .05 151.5 ± 17.7 146.4 ± 17.0 .16 .25 .02a
ApoB (mg/dL) 150.1 ± 28.2 142.0 ± 27.3 .03a 142.6 ± 23.6 137.8 ± 23.5 .11 .47 .58

Abbreviations: ApoA1, apolipoprotein A1; ApoB, apolipoprotein B; HDL-C, high-density cholesterol; LDL-C, low-density cholesterol; SD, standard
deviation; TC, total cholesterol.
A subgroup analysis was performed in the participants with dyslipidemia (LDL-C >130 mg/dL). Kefir group consumed 180 mL/d probiotic kefir
and milk group consumed 180 mL/d unfermented milk for 12 weeks. Data are expressed as mean ± SD.
a
P < .05, P1 and P2 , intragroup comparisons (week 0 and week 12), were performed using the paired samples t test or the Wilcoxon test as
appropriate. Pbaseline was the week 0 comparisons of the groups and analyzed either with t test or the Mann-Whitney U test. Pchange compares the
changes from week 0 to week 12 between the groups and was analyzed with t test or the Mann-Whitney U test.

serum LDL-C (7.6%) and ApoB levels (5.4%), and in the milk marxianus) reduced fat accumulation and weight gain in high-
group, total cholesterol (3.5%), significantly decreased com- fat-diet-induced obese mice [18,19]. Rosa et al. showed that
pared with the baseline (Table 5). a 10-week kefir supplementation reduced the abdominal cir-
cumference significantly without any differences in the BMI
3.5. Adverse Events of rats induced with metabolic syndrome [20]. On the other
hand, Chen et al. emphasized the importance of overall diet
Six participants in the unfermented milk group and 7 partici- composition as they demonstrated the beneficial effects of ke-
pants in the kefir group reported some minor gastrointestinal fir on body weight, glucose intolerance, and hepatic steato-
side effects such as bloating, nausea, and diarrhea, especially sis in obese mice fed with a high-fat diet, but not with a
in the first week of the intervention. The reported side effects Western diet [28]. According to the results of clinical stud-
were not different between the kefir and unfermented milk ies, kefir consumption has limited effects on body weight and
groups (P> .05). No serious adverse effect was reported. body composition [21,29-31]. A randomized, controlled clini-
cal study showed that the consumption of 2 servings/d of kefir
or lowfat milk in addition to the 2 servings/d of lowfat dairy
4. DISCUSSION products for 8 weeks provided similar reductions in weight,
BMI, and weight circumferences in overweight or obese pre-
The most promising outcome of this study was the signifi- menopausal women [29]. On the other hand, previous stud-
cant increase in serum ApoA1 concentrations provided by ke- ies conducted by da Silva Ghizi et al. [21], Ostadrahimi et al.
fir consumption compared with milk consumption; therefore, [30], and Pražnikar et al. [31] demonstrated that kefir con-
our hypothesis that kefir consumption could improve the lipid sumption did not cause any changes in the body weight or
profile in patients with metabolic syndrome was accepted. related anthropometrical measurements, which was in agree-
On the other hand, regular kefir consumption did not pro- ment with the results of this study. In our study, not ke-
vide superior effects compared with milk consumption on an- fir but regular milk consumption resulted in a decrease in
thropometrical measurements, glycemic control, inflamma- weight and fat-free mass. The significant increase in calcium
tory parameters, or blood pressure even if there were amelio- intake in the milk group may partially be an explanation for
rations in baseline traits. Therefore, our hypothesis that kefir these findings because calcium has a particular role in weight
consumption could improve the anthropometrical measure- loss [32]. It is well known that the activities of probiotics,
ments, glycemic control, blood pressure, and inflammatory including antiobesity activity, are specific to the species or
status in patients with metabolic syndrome was rejected. strains [33]. Therefore, one of the explanations for the dis-
It has been suggested that probiotic administration as a crepancy among the results of the studies conducted with ke-
part of metabolic syndrome therapy may improve the health fir could be the wide variety of species or strains that kefir
status of patients [26]. A meta-analysis conducted by Borger- comprises.
aas et al. reported that probiotic supplementation leads to a The efficiency of probiotics on glycemic control is also con-
moderate decrease in body weight, BMI, and fat percentage troversial. Some studies demonstrated the beneficial effects
(weighted mean difference: -0.60 kg, -0.27 kg/m2 , and -0.60%, of probiotic interventions on glycemia [34–37], whereas others
respectively) in overweight and obese subjects [27]. However, did not support these findings [9,38]. A recent meta-analysis
the findings of the studies that investigated the effect of ke- reported that probiotic yogurt consumption did not modulate
fir on body weight and body composition are conflicting. In fasting blood glucose, insulin, HbA1c, or HOMA-IR in people
animal models, kefir or kefir powder (Lactobacillus species, with type 2 diabetes or obesity [39]. The strain of the probi-
Leuconostoc mesenteroides, Lactobacillus kefiri, and Kluyveromyces otics, the initial compositions of the microbiota, and the base-
 Nutrition Research 102 (2022) 59–70 67

line glycemic status of the participants were listed as the fac- that both milk and kefir consumption had significantly posi-
tors that could modify the efficacy of probiotics on glycemic tive effects on systolic and diastolic blood pressure. Bioactive
control [40,41]. In our previous report, although our major fo- peptides with angiotensin I-converting enzyme inhibitory ac-
cus was to determine the change in the composition of gut tivity and mineral content (calcium, potassium, phosphorus,
microbiota provided by kefir consumption, we detected a bor- etc.) of dairy products are potential mediators in the regula-
derline significance in fasting insulin and HOMA-IR values tion of blood pressure [52].
(P= .05) following kefir consumption when compared with Previous studies demonstrated that probiotic supplemen-
the baseline, but not compared with milk consumption [22]. tation, including kefir, might modulate the inflammatory re-
Ostadrahimi et al. reported that 8 weeks of 600 mL/d kefir sponse by reducing the pro-inflammatory cytokine concentra-
consumption, containing Lactobacillus casei, Lactobacillus aci- tions [15,53]. In addition to probiotics and kefir, it was shown
dophilus, and Bifidobacteria, significantly reduced the fasting that nonfermented milk consumption might also decrease
blood glucose and HbA1c concentrations in diabetic patients the pro-inflammatory cytokines [49]. In line with the previ-
[30]. Another study has shown a decrease in fasting glycemia ous findings, this study demonstrated that kefir consumption
without a change in HbA1c concentration after kefir admin- could reduce the TNF-α, IFN-ɤ, and IL-6 levels. Interestingly,
istration of 1.6 mL/kg for men or 1.9 mL/kg for women in serum IL-10 concentrations, an anti-inflammatory cytokine,
metabolic syndrome patients [21]. In that study, kefir was pro- were also decreased by kefir consumption. The complex regu-
duced from kefir grains and contained more than 30 species lation of the cytokines and the stimulation of IL-10 production
of bacteria and more than 12 species of yeast and fungi. A with TNF-α may be responsible for the decrease in the IL-10
meta-analysis indicated that kefir could reduce the fasting concentration [54].
blood glucose and insulin levels without a significant effect This study has significant contributions to the literature
on HbA1c [42]. In our study, neither kefir nor milk consump- because it is one of the first randomized controlled clinical
tion for a 12-week period revealed any significant effects on studies that investigated a traditional fermented drink with
glycemic control. Although the test drinks were also called probiotic characteristics in metabolic syndrome. However, it
“kefir” in the previous studies, the microbial compositions of also had several limitations. First, the use of milk as the con-
the kefirs used in those studies were different from the ke- trol drink failed to provide a placebo effect in some aspects,
fir used in our study. It should also be noted that the daily- even though it has been used as the control drink of fermented
consumed amount of kefir varied widely between the studies, dairy products in a previous study [55]. Second, this study
potentially leading to the discrepancy in the findings of differ- could not be performed in a double-blind design because of
ent studies. the limitations of finding out a better placebo of kefir. Third,
The cholesterol-lowering effects of probiotics have been a larger study population might have provided a better dis-
shown in previous studies [9,43]. Several potential mecha- tribution of participants with different degrees of metabolic
nisms regarding probiotics’ efficiency have been proposed, in- abnormalities.
cluding modification of salt hydrolase activity, inhibition of In conclusion, a 12-week administration of 180 mL/d ke-
3-hydroxy-3-methylglutaryl coenzyme A reductase activity, fir provided beneficial effects on lipid parameters, blood pres-
binding of cholesterol to the cellular surface, incorporation of sure, and inflammatory response in patients with metabolic
cholesterol into the bacterial cell membrane, and conversion syndrome, compared with the baseline. Not surprisingly,
of cholesterol to coprostanol [44–47]. In our study, kefir con- milk consumption also showed some beneficial effects on
sumption led to a significant increase in serum ApoA1, the metabolic syndrome components, probably because of the
major protein of HDL particles, compared with milk consump- many bioactive components of milk. Therefore, kefir con-
tion. However, the improvement in the lipid profile provided by sumption showed superior effects only in serum ApoA1 lev-
kefir consumption was moderate in this study, compared with els compared with milk consumption. The efficiency of kefir
previous ones [9,20,21,43]. Fuentes et al. showed that subjects was associated with its microbial composition and daily con-
with higher cholesterol levels benefited more from probiotic sumption amount, as well as the baseline health status of the
supplementation [43]. Therefore, we performed a subgroup participants. Therefore, further studies are required to inves-
analysis of the participants with serum LDL-C concentrations tigate the most efficient microbial composition of kefir in pa-
higher than 130 mg/dL. The subgroup analysis confirmed the tients with different health statuses.
greater efficiency of kefir consumption in individuals with
worsening lipid profiles. Interestingly, some improvements in
the lipid profile were also observed in the control group. This Author Declarations
could be explained by the potential beneficial effects of regular
milk consumption on lipid profile because of its bioactive pep- The authors declare that they have no conflicts of interest.
tides and calcium content [48]. The decreases in total choles-
terol and LDL-C levels by the consumption of nonfermented
milk and conventional yogurt were also reported previously CRediT authorship contribution statement
[31,49,50].
The beneficial effects of dairy consumption on blood pres- Ezgi Bellikci-Koyu: Conceptualization, Methodology, Inves-
sure are well known. A meta-analysis showed that each ad- tigation, Formal analysis, Funding acquisition, Writing – orig-
ditional 200 g/d milk or total dairy consumption was associ- inal draft. Banu Pınar Sarer-Yurekli: Methodology, Investiga-
ated with a 4% or 3% of lower risk for hypertension, respec- tion, Writing – original draft. Cem Karagozlu: Methodology,
tively [51]. In parallel with these findings, our study showed Investigation, Writing – original draft. Fadime Aydin-Kose:
68 Nutrition Research 102 (2022) 59–70

Methodology, Investigation, Formal analysis, Writing – origi- Evaluation of the effect of Lactobacillus reuteri V3401 on
nal draft. Ahmet Gokhan Ozgen: Methodology, Investigation, biomarkers of inflammation, cardiovascular risk and liver
Writing – original draft. Zehra Buyuktuncer: Conceptualiza- steatosis in obese adults with metabolic syndrome: a
randomized clinical trial (PROSIR). BMC Complement Altern
tion, Methodology, Funding acquisition, Writing – review &
Med 2018;18:e306. doi:10.1186/s12906- 018- 2371- x.
editing, Supervision.
[11] Rezazadeh L, Gargari BP, Jafarabadi MA, Alipour B. Effects of
probiotic yogurt on glycemic indexes and endothelial
dysfunction markers in patients with metabolic syndrome.
Acknowledgment Nutrition 2019;62:162–8. doi:10.1016/j.nut.2018.12.011.
[12] Bellikci Koyu E, Buyuktuner Z. A functional food: kefir. J Nutr
This work was supported by the Teaching Staff Training Pro- Diet 2018;46:166–75. doi:10.33076/2018.BDD.301.
[13] Codex Alimentarius. Codex standard for fermented milks.
gram of the Turkish Council of Higher Education.
Accessed October 10, 2021 . https://www.fao.org/
fao- who- codexalimentarius/codex-texts/list-standards/en/.
[14] Fathi Y, Ghodrati N, Zibaeenezhad MJ, Faghih S. Kefir drink
Supplementary materials causes a significant yet similar improvement in serum lipid
profile, compared with low-fat milk, in a dairy-rich diet in
Supplementary material associated with this article can be overweight or obese premenopausal women: a randomized
found, in the online version, at doi:10.1016/j.nutres.2022.02. controlled trial. J Clin Lipidol 2017;11:136–46.
doi:10.1016/j.jacl.2016.10.016.
006.
[15] Hadisaputro S, Djokomoeljanto RR, Judiono Soesatyo MH.
The effects of oral plain kefir supplementation on
references proinflammatory cytokine properties of the hyperglycemia
Wistar rats induced by streptozotocin. Acta Med Indones
2012;44:100–4.
[1] Alberti KGMM, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, [16] Friques AG, Arpini CM, Kalil IC, Gava AL, Leal MA, Porto ML,
Donato KA, et al. Harmonizing the metabolic syndrome: a et al. Chronic administration of the probiotic kefir improves
joint interim statement of the International Diabetes the endothelial function in spontaneously hypertensive rats.
Federation Task Force on Epidemiology and Prevention; J Transl Med 2015;13:e390. doi:10.1186/s12967- 015- 0759- 7.
National Heart, Lung, and Blood Institute; American Heart [17] Kanbak G, Uzuner K, Kusat Ol K, Oglakci A, Kartkaya K,
Association; World Heart Federation; International Senturk H. Effect of kefir and low-dose aspirin on arterial
Atherosclerosis Society; and International Association for blood pressure measurements and renal apoptosis in
the Study of Obesity. Circulation 2009;120(16):1640–5. unhypertensive rats with 4 weeks salt diet. Clin Exp
doi:10.1161/CIRCULATIONAHA.109.192644. Hypertens 2014;36:1–8. doi:10.3109/10641963.2013.783046.
[2] Saklayen MG. The global epidemic of the metabolic [18] Choi JW, Kang HW, Lim WC, Kim MK, Lee IY, Cho HY. Kefir
syndrome. Curr Hypertens Rep 2018;20:e12. prevented excess fat accumulation in diet-induced obese
doi:10.1007/s11906- 018- 0812- z. mice. Biosci Biotechnol Biochem 2017;81:958–65.
[3] Grundy SM. Metabolic syndrome update. Trends Cardiovasc doi:10.1080/09168451.2016.1258984.
Med 2016;26:364–73. doi:10.1016/j.tcm.2015.10.004. [19] Kim DH, Kim H, Jeong D, Kang IB, Chon JW, Kim HS, et al.
[4] Elder SJ, Lichtenstein AH, Pittas AG, Roberts SB, Fuss PJ, Kefir alleviates obesity and hepatic steatosis in high-fat
Greenberg AS, et al. Genetic and environmental influences diet-fed mice by modulation of gut microbiota and
on factors associated with cardiovascular disease and the mycobiota: targeted and untargeted community analysis
metabolic syndrome. J Lipid Res 2009;50:1917–26. with correlation of biomarkers. J Nutr Biochem
doi:10.1194/jlr.P900033-JLR200. 2017;44:35–43. doi:10.1016/j.jnutbio.2017.02.014.
[5] Backhed F, Ding H, Wang T, Hooper LV, Koh GY, Nagy A, et al. [20] Rosa DD, Grześkowiak ŁM, Ferreira CL, Fonseca AC, Reis SA,
The gut microbiota as an environmental factor that Dias MM, et al. Kefir reduces insulin resistance and
regulates fat storage. Proc Natl Acad Sci USA. inflammatory cytokine expression in an animal model of
2004;101:15718–23. doi:10.1073/pnas.0407076101. metabolic syndrome. Food Funct 2016;7:3390–401.
[6] Ley RE, Turnbaugh PJ, Klein S, Gordon JI. Human gut microbes doi:10.1039/c6fo00339g.
associated with obesity. Nature 2006;444:1022–3. [21] da Silva Ghizi AC, de Almeida Silva M, da Andrade
doi:10.1038/4441022a. Moraes FS, da Silva CL, Endringer DC, Scherer R, et al. Kefir
[7] Vijay-Kumar M, Aitken JD, Carvalho FA, Cullender TC, improves blood parameters and reduces cardiovascular risks
Mwangi S, Srinivasan S, et al. Metabolic syndrome and in patients with metabolic syndrome. PharmaNutrition
altered gut microbiota in mice lacking Toll-like receptor 5. 2021;16:e100266. doi:10.1016/j.phanu.2021.100266.
Science 2010;328:228–31. doi:10.1126/science.1179721. [22] Bellikci-Koyu E, Sarer-Yurekli BP, Akyon Y, Aydin-Kose F,
[8] Hill C, Guarner F, Reid G, Gibson GR, Merenstein DJ, Pot B, Karagozlu C, Ozgen AG, et al. Effects of regular kefir
et al. Expert consensus document. The International consumption on gut microbiota in patients with metabolic
Scientific Association for Probiotics and Prebiotics consensus syndrome: a parallel-group, randomized, controlled study.
statement on the scope and appropriate use of the term Nutrients 2019;11:e2089. doi:10.3390/nu11092089.
probiotic. Nat Rev Gastroenterol Hepatol 2014;11:506–14. [23] Zimmet P, Alberti KGMM, Serrano Rios M. A new
doi:10.1038/nrgastro.2014.66. International Diabetes Federation worldwide definition of
[9] Bernini LJ, Simao AN, Alfieri DF, Lozovoy MA, Mari NL, de the metabolic syndrome: the rationale and the results. Rev
Souza CH, et al. Beneficial effects of Bifidobacterium lactis on Esp Cardiol 2005;58:1371–6 1016/S1885-5857(06)60742-1.
lipid profile and cytokines in patients with metabolic doi:10.1016/S1885- 5857(06)60742- 1.
syndrome: a randomized trial. Effects of probiotics on [24] Food and Agriculture Organization. Human energy
metabolic syndrome. Nutrition 2016;32:716–19. requirements. Report of a joint FAO/WHO/UNU expert
doi:10.1016/j.nut.2015.11.001. consultation. Accessed July 1, 2021. https://agris.fao.org/
[10] Tenorio-Jimenez C, Martinez-Ramirez MJ, Tercero-Lozano M, agris-search/search.do?recordID=XF2015034734.
Arraiza-Irigoyen C, Del Castillo-Codes I, Olza J, et al. [25] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the
 Nutrition Research 102 (2022) 59–70 69

concentration of low-density lipoprotein cholesterol in Anothaisintawee T. The effect of probiotic yogurt on

plasma, without use of the preparative ultracentrifuge. Clin glycemic control in type 2 diabetes or obesity: a
Chem 1972;8:499–502. meta-analysis of nine randomized controlled trials.
[26] Tenorio-Jiménez C, Martínez-Ramírez MJ, Gil Á, Nutrients 2019:e11. doi:10.3390/nu11030671.
Gómez-Llorente C. Effects of probiotics on metabolic [40] Mobini R, Tremaroli V, Stahlman M, Karlsson F, Levin M,
syndrome: a systematic review of randomized clinical trials. Ljungberg M, et al. Metabolic effects of Lactobacillus reuteri
Nutrients 2020;12:e124. doi:10.3390/nu12010124. DSM 17938 in people with type 2 diabetes: a randomized
[27] Borgeraas H, Johnson LK, Skattebu J, Hertel JK, Hjelmesaeth J. controlled trial. Diabetes Obes Metab 2017;19:579–89.
Effects of probiotics on body weight, body mass index, fat doi:10.1111/dom.12861.
mass and fat percentage in subjects with overweight or [41] Nikbakht E, Khalesi S, Singh I, Williams LT, West NP,
obesity: a systematic review and meta-analysis of Colson N. Effect of probiotics and synbiotics on blood
randomized controlled trials. Obes Rev 2018;19:219–32. glucose: a systematic review and meta-analysis of controlled
doi:10.1111/obr.12626. trials. Eur J Nutr 2018;57:95–106.
[28] Chen YT, Hsu AH, Chiou SY, Lin YC, Lin JS. AB-Kefir reduced doi:10.1007/s00394- 016- 1300- 3.
body weight and ameliorated inflammation in adipose tissue [42] Salari A, Ghodrat S, Gheflati A, Jarahi L, Hashemi M,
of obese mice fed a high-fat diet, but not a high-sucrose diet. Afshari A. Effect of kefir beverage consumption on glycemic
Nutrients 2021;13:e2182. doi:10.3390/nu13072182. control: a systematic review and meta-analysis of
[29] Fathi Y, Faghih S, Zibaeenezhad MJ, Tabatabaei SH. Kefir randomized controlled clinical trials. Complement Ther Clin
drink leads to a similar weight loss, compared with milk, in a Pract 2021;44:e101443. doi:10.1016/j.ctcp.2021.101443.
dairy-rich non-energy-restricted diet in overweight or obese [43] Fuentes MC, Lajo T, Carrion JM, Cune J. Cholesterol-lowering
premenopausal women: a randomized controlled trial. Eur J efficacy of Lactobacillus plantarum CECT 7527, 7528 and
Nutr 2016;55:295–304. doi:10.1007/s00394- 015- 0846- 9. 7529 in hypercholesterolaemic adults. Br J Nutr
[30] Ostadrahimi A, Taghizadeh A, Mobasseri M, Farrin N, 2013;109:1866–72. doi:10.1017/S000711451200373X.
Payahoo L, Beyramalipoor Gheshlaghi Z, et al. Effect of [44] Tanaka H, Doesburg K, Iwasaki T, Mierau I. Screening of lactic
probiotic fermented milk (kefir) on glycemic control and acid bacteria for bile salt hydrolase activity. J Dairy Sci
lipid profile in type 2 diabetic patients: a randomized 1999;82:2530–5. doi:10.3168/jds.S0022- 0302(99)75506- 2.
double-blind placebo-controlled clinical trial. Iran J Public [45] Chen K, Li S, Chen F, Li J, Luo X. Regulation of the
Health 2015;44:228–37. Lactobacillus strains on HMGCoA reductase gene
[31] Pražnikar ZJ, Kenig S, Vardjan T, Bizjak MČ, Petelin A. Effects transcription in human HepG2 cells via nuclear
of kefir or milk supplementation on zonulin in overweight factor-kappaB. J Microbiol Biotechnol 2016;26:402–7.
subjects. J Dairy Sci 2020;103:3961–70. doi:10.4014/jmb.1507.07086.
doi:10.3168/jds.2019-17696. [46] Lye H-S, Rahmat-Ali GR, Liong M-T. Mechanisms of
[32] Ilich JZ, Kelly OJ, Liu P-Y, Shin H, Kim Y, Chi Y, et al. Role of cholesterol removal by lactobacilli under conditions that
calcium and low-fat dairy foods in weight-loss outcomes mimic the human gastrointestinal tract. Int Dairy J
revisited: results from the randomized trial of effects on 2010;20:169–75. doi:10.1016/j.idairyj.2009.10.003.
bone and body composition in overweight/obese [47] Lye HS, Rusul G, Liong MT. Removal of cholesterol by
postmenopausal women. Nutrients 2019;11:e1157. lactobacilli via incorporation and conversion to coprostanol.
doi:10.3390/nu11051157. J Dairy Sci 2010;93:1383–92. doi:10.3168/jds.2009-2574.
[33] Crovesy L, Ostrowski M, Ferreira D, Rosado EL, [48] Marcone S, Belton O, Fitzgerald DJ. Milk-derived bioactive
Soares-Mota M. Effect of Lactobacillus on body weight and peptides and their health promoting effects: a potential role
body fat in overweight subjects: a systematic review of in atherosclerosis. Br J Clin Pharmacol 2017;83:152–62.
randomized controlled clinical trials. Int J Obes doi:10.1111/bcp.13002.
2017;41:1607–14 Lond. doi:10.1038/ijo.2017.161. [49] Barreto FM, Colado Simao AN, Morimoto HK, Batisti
[34] Kassaian N, Feizi A, Aminorroaya A, Jafari P, Ebrahimi MT, Lozovoy MA, Dichi I, Helena da Silva Miglioranza L. Beneficial
Amini M. The effects of probiotics and synbiotic effects of Lactobacillus plantarum on glycemia and
supplementation on glucose and insulin metabolism in homocysteine levels in postmenopausal women with
adults with prediabetes: a double-blind randomized clinical metabolic syndrome. Nutrition 2014;30:939–42.
trial. Acta Diabetol 2018;55:1019–28. doi:10.1016/j.nut.2013.12.004.
doi:10.1007/s00592- 018- 1175- 2. [50] Fabian E, Elmadfa I. Influence of daily consumption of
[35] Kobyliak N, Falalyeyeva T, Mykhalchyshyn G, Kyriienko D, probiotic and conventional yoghurt on the plasma lipid
Komissarenko I. Effect of alive probiotic on insulin resistance profile in young healthy women. Ann Nutr Metab
in type 2 diabetes patients: randomized clinical trial. 2006;50:387–93 2006. doi:10.1159/000094304.
Diabetes Metab Syndr 2018;12:617–24. [51] Soedamah-Muthu SS, Verberne LD, Ding EL, Engberink MF,
doi:10.1016/j.dsx.2018.04.015. Geleijnse JM. Dairy consumption and incidence of
[36] Tonucci LB, Olbrich Dos Santos KM, Licursi de Oliveira L, hypertension: a dose-response meta-analysis of prospective
Rocha Ribeiro SM, Duarte Martino HS. Clinical application of cohort studies. Hypertension 2012;60:1131–7.
probiotics in type 2 diabetes mellitus: a randomized, doi:10.1161/HYPERTENSIONAHA.112.195206.
double-blind, placebo-controlled study. Clin Nutr [52] McGrane MM, Essery E, Obbagy J, Lyon J, MacNeil P, Spahn J,
2017;36:85–92. doi:10.1016/j.clnu.2015.11.011. et al. Dairy consumption, blood pressure, and risk of
[37] Ejtahed HS, Mohtadi-Nia J, Homayouni-Rad A, Niafar M, hypertension: an evidence-based review of recent literature.
Asghari-Jafarabadi M, Mofid V. Probiotic yogurt improves Curr Cardiovasc Risk Rep 2011;5:287–98.
antioxidant status in type 2 diabetic patients. Nutrition doi:10.1007/s12170- 011- 0181- 5.
2012;28:539–43. doi:10.1016/j.nut.2011.08.013. [53] Tenorio-Jiménez C, Martínez-Ramírez MJ, Del
[38] Ivey KL, Hodgson JM, Kerr DA, Lewis JR, Thompson PL, Castillo-Codes I, Arraiza-Irigoyen C, Tercero-Lozano M,
Prince RL. The effects of probiotic bacteria on glycaemic Camacho J, et al. Lactobacillus reuteri V3401 reduces
control in overweight men and women: a randomised inflammatory biomarkers and modifies the gastrointestinal
controlled trial. Eur J Clin Nutr 2014;68:447–52. microbiome in adults with metabolic syndrome: the PROSIR
doi:10.1038/ejcn.2013.294. study. Nutrients 2019;11:e1761. doi:10.3390/nu11081761.
[39] Barengolts E, Smith ED, Reutrakul S, Tonucci L, [54] Stenvinkel P, Ketteler M, Johnson RJ, Lindholm B,
70 Nutrition Research 102 (2022) 59–70

Pecoits-Filho R, Riella M, et al. IL-10, IL-6, and TNF-α: central or cholesterol fractional synthesis rates relative to milk in
factors in the altered cytokine network of uremia—the good, hyperlipidemic men: a randomized controlled trial
the bad, and the ugly. Kidney Int 2005;67:1216–33. [ISRCTN10820810]. BMC Complement Altern Med 2012;2:e1.
doi:10.1111/j.1523-1755.2005.00200.x. doi:10.1186/1472- 6882- 2- 1.
[55] St-Onge MP, Farnworth ER, Savard T, Chabot D, Mafu A,
Jones PJ. Kefir consumption does not alter plasma lipid levels